Nanos2 promotes differentiation of male germ cells basing on thenegativeregulationofFoxd3andthetreatmentof5‐Azadc and TSA

Wenhui Zhang1* | Yulin Bi1* | Yingjie Wang1 | Man Wang1 | Dong Li2 | Shaoze Cheng1 | Jing Jin1 | Tingting Li1 | Bichun Li1 | Yani Zhang1

Abstract
The transcription factor positioning in promoter regions relate to gene regulation, and the level of DNA methylation and histone acetylation also impact the promoter activity. In this study, we tested and verified the core promoter region and key transcription factor of Nanos2 which is a male‐critical gene in the differentiation of embryonic stem cells to male germ cells, meanwhile, epigenetic effects by mean of 5‐Aza‐2′‐deoxycytidine (5‐Azadc) and Trichostin A (TSA) on the activity of Nanos2 promoter were detected. The results reveal that key transcription factor Foxd3 is a negative regulator of Nanos2, which suggests that loss‐of‐function of Foxd3 causes strong expression of Nanos2 responsive to large amounts of primordial germ cells and spermatogonial stem cells, whereas its overexpression causes the opposite effect. Furthermore, both 5‐Azadc and TSA can provoke responses of Nanos2, but the combination effect of the two is better.

K E Y W O R D S
chicken, epigenetic modification, male germ cell, Nanos2, transcription factor

Introduction

Embryonic stem cells (ESCs) which are isolated from early embryos or primitive gonads have the potential to differentiate into various types of cells, especially, the differentiation into male germ cells represents an exciting area of the current research. It is coordinated by various factors, such as inducible factors, key genes, and signaling pathways. Therefore, induction efficiency of male germ cells in vitro could be effectively improved if only the regulation mechanisms of these regulatory factors in male germ cell differentiation were Nanos2 is a highly specific sex‐determining endogenous male reproductive gene, which plays key roles in the self‐renewal and maintenance of primordial germ cells (PGCs) and spermatogonial stem cells (SSCs) (Murray, Yang, & Van doren, 2010; Pui & Saga, 2018). Although Nanos2 expressed differently in various tissues of dairy goat, yak, and cow, it showed higher expression in testis. It has been reported that Nanos2 is involved in the formation of spermatids and meiosis that can inhibit the expression of the meiosis initiation gene Stra8 (Yao et al., 2014; Zeng et al., 2013). Our previous study found that overexpression of Nanos2 in vitro can promote the production of PGCs and SSCs, while knockout of Nanos2 can result in the inhibition of gonadal development and thus inhibit the production of male germ cells (Zhang et al., 2017). In addition, MiR‐34c can target the 3′ untranslated region (UTR) of Nanos2 to regulate the differentiation of SSCs (Yu et al., 2013), this possibility is suggested that, in principle, Nanos2 is epigenetically modified in the regulation of male germ cell production. Thus, it need to be further explored that how these epigenetic factors can regulate the expression of Nanos2 and affect the process of male germ cells.
Epigenetic modifications refer to the regulation of gene expression by epigenetic changes (DNA methylation, acetylation, histone modifications, and noncoding RNAs such as microRNAs) that do not depend on changes in gene sequence and can be expressed epigenetically.
DNA methylation is an important form of DNA epigenetic modification discovered earlier. Specifically, under the catalysis of DNA methyltransferase, the methyl groups in organism are selectively added to the cytosine of two nucleotide CGs of the DNA sequence to form 5‐methyl cytosine. Methylation can lead to tight junctions between nucleosomes and thus suppress gene expression. However, the gene can be re‐expressed by demethylating the gene by adding a small molecule compound. 5‐Aza‐2′‐deoxycytidine (5‐Azadc) is a clinically popular and well‐acting methylation inhibitor. A large number of studies support that 5‐Azadc can be used to repress the expression of a gene and might be responsible for gene silencing (Li, Wong, Chan, Chng, & Chim, 2018; Lmb et al., 2018).
Histone acetylation is a reversible reaction involving histone acetyltransferase and histone deacetylase (HDAC), both of which interact with each other. Acetylation of histone protein facilitates the dissociation of DNA and histone octamer, and the relaxation of nucleosome structure, so that various transcription factors and co‐transcription factors can specifically bind to DNA binding sites and activate gene transcription. HDAC allows them to bind tightly to negatively charged DNA, lead to dense curl of chromatin, and gene transcription is suppressed. Trichostin A (TSA) binds to HDAC, inhibits histone deacetylation, and promotes gene expression (Cao et al., 2017; Fila‐Danilow, Borkowska, PaulSamojedny, Kowalczyk, & Kowalski, 2017).
Here, the effect of the sequence immediately upstream of the chicken Nanos2 on its transcription activity was studied. Nucleotide deletions were introduced into the 5′‐flanking sequence and the obtained plasmids were used to transfect DF‐1 cells where the relative level of Nanos2 was measured. This allowed us to identify a region crucial for the transcription of Nanos2. On the basis of the early induction system that retinoic acid (RA) promotes the differentiation of ESCs into male germ cells, exploring the effect of 5‐Azadc and TSA on Nanos2 promoter and to verify the transcriptional regulation mechanism of Nanos2 by Foxd3, to make better use of this gene in vitro to promote the differentiation of ESCs into the male germ cells.

2 | MATERIALS AND METHODS

2.1 | Animals

All procedures involving the care and use of animals conformed to US National Institute of Health guidelines (NIH Pub. No. 85‐23, revised 1996) and were approved by the Laboratory Animal Management and Experimental Animal Ethics Committee of Yangzhou University.
Fertilized eggs of Rugao yellow chickens (Gallus gallus domesticus ) were collected shortly after fertilization from the Poultry Institute of the Chinese Academy of Agricultural Sciences Experimental Poultry Farm.

2.2 | Bioinformatics prediction

To predict Nanos2 promoter region: http://genome.ucsc.edu/, http://www.ncbi.nlm.nih.gov/gene/, http://www.cbrc.jp/rese‐arch/ db/TFSEARCH, http//wwwibimas.cit.nih.gov/molbio/proscan
To predict the transcription factor binding sites of Nanos2: The JASPAR database（http://jaspar.genereg.net/）
To predict the CpG island of Nanos2 promoter: http://www. urogene.org/cgi‐bin/methprimer/methprimer.cgi

2.3 | Plasmid construction

5′‐flanking sequence of Nanos2 were generated by polymerase chain reaction (PCR) according to the predicted promoter region, and the specific primers were designed: F: 5′‐CGCATTAATCCAAGCTTGCCCAGTAGAAAA‐3′ (underlined is the Ase I restriction site), R: 5′‐CGGGGTACCTCAGTCTCTCCTGCTCTGCAT‐3′ (underlined is Kpn I restriction site). The amplified fragment was cloned into the pEGFP‐N1 and the constructed plasmid named pNanos2‐EGFP was verified by restriction enzyme digestion and sent for sequencing (Beijing TsingKe Biotech Co., Ltd., Nanjing, China).
The primers for amplifying the 5′‐flanking sequence of Nanos2 promoter with different length were designed as shown in Table 1. According to the software design results, when designing these missing different sequences, the regulatory elements would be avoided. The amplified fragments with different lengths were cloned into pGL3‐basic vector. Follow‐up steps as per Fila‐Danilow et al. (2017).
According to the homologous recombination technology, Foxd3 and Nobox binding site deletion vectors and Foxd3 binding site overexpression vector were constructed. The fragment amplification primers were designed as shown in Tables . The amplified fragment was ligated with the pGL3‐basic vector double‐digested in the above process using a homologous recombinase TreliefTM SoSoo Cloning Kit (Beijing TsingKe Biotech Co., Ltd., Nanjing, China). Follow‐up steps as per Fila‐Danilow et al. (2017).

2.4 | Cell separation and culture

After fresh fertilized eggs were washed and disinfected, single blastoderms were obtained in a sterile environment, rinsed 2–3 times in phosphate buffered saline (PBS), centrifuged at 1,200 rpm/min for 6 min and then resuspended with PBS, filtered, inoculated into petri dishes and cultured with differentiation inhibitor medium (high glucose Dulbecco modified Eagle medium + 0.1 mmol/L β‐mercaptoethanol + 5 ng/μL stem cell factor (SCF) + 10 ng/mL bFGF + 1,000 IU/mL LIF + 15% FBS (fetal calf serum) + 2% Serum) (Sigma, Shanghai, China). After 24 hr, the cells were purified and harvested and then inoculated into a 24‐well plate with 105 cells per well.

2.5 | In vitro induced differentiation

ESCs were transfected with plasmid at the ratio of transfection reagent (V/μL):plasmid (m/ng) = 3:1 (FuGENE® HD Transfection Reagent, Promega, Shanghai, China) in a 24‐well plate. Groups were as follows: Control group (transfected with pGL3/493 vector), Foxd3 deletion group (transfected with pGL3/493 vector lacking Foxd3 binding site), and Foxd3 overexpression group (transfected with pGL3/493 vector overexpressing Foxd3 binding site). After 48 hr of transfection, cells were cultured with 10−5 mol/L RA‐induced differentiation medium. Cell morphology was observed every 2 days, and the cells were detected by quantitative real time PCR (qRT‐PCR), indirect immunofluorescence, and flow cytometry analysis.
When cell confluency was about 70%, ESCs were transfected with pGL3‐1187 at the ratio of transfection reagent (V/μL): plasmid (m/ng) = 3:1. After 48 hr of transfection, the groups were changed to medium containing RA and 5‐Azadc or TSA. Groups were as follows: Control group (10−5 mol/L RA), 5‐Azadc group (10−5 mol/L RA + 5 μmol/L 5‐Azadc), TSA group (10−5 mol/L RA + 1.5μmol/L TSA), and 5‐Azadc + TSA group (10−5 mol/L RA + 5μmol/ L 5‐Azadc + 1.5μmol/L TSA). Follow‐up experiment as per Fila‐Danilow et al. (2017).

2.6 | Luciferase reporter assay

The dual‐luciferase reporter gene vectors constructs were generated by cloning the entire 5′UTR or the mutant 5′UTR of Nanos2 into pGL3‐Basic vector. The Nanos2 5′UTR fragment cloning was performed using PCR. The Firefly luciferase vector was used for internal reference. pGL3‐Basic empty vector was used as negative control. pRL‐SV40 were cotransfected into DF‐1 cells in a 24‐well plate using FuGENE® HD Transfection Reagent (Promega). After 48 hr, all the target validation assays were performed with the dual‐luciferase reporter system according to the Promega dual luciferase reporter assay kit instructions. The activities were measured by a microplate reader.

2.7 | CHIP assay

The binding of Foxd3 to Nanos2 was detected by a CHIP assay according to protocol of a EZ‐ChIP™ Chromatin Immunoprecipitation Kit (MILLIPORE, Temecula，USA). Cells were collected to crosslinking and lysis, used sonication to shear DNA by Ultrasonic Cell Crusher (RTPIO, Nanjing, China), carried out immunoprecipitation of crosslinked protein/DNA, and anti‐Foxd3 was purchased from abcam (Cambridge, UK), after elution of protein/DNA complexes, conducting reverse crosslinks of protein/DNA complexes to free DNA, using spin columns to purify DNA and then performed PCR.

2.8 | Bisulfite sequencing PCR

Genome was treated with bisulfite according to protocol of a EZ DNA Methylation‐GoldTM Kit (ZYMO RESEARCH, CA), then PCR primers were designed using a MethPrimer software (Table 4) and PCR amplification was performed by a Zymo TaqTM premix (ZYMO RESEARCH). The amplified fragment was cloned into the pMDTM 19‐T, the products were transformed into DH5α competent cells, coated plates, and selected 20 monoclonal colonies to sequencing.

2.9 | Quantitative real time PCR assay

Total RNAs were extracted from cell pellet by the Trizol method and then transcribed into complementary DNA (FastQuant RT Kit with gDNase, TIANGEN, Beijing, China), according to the manufacturer’s protocol. Real‐time PCR assays were performed using a SYBR fluorescence reagent and a 7500 System Fluorescence Quantitation Instrument (Applied Biosystems, Carlsbad, CA) according to the instructions of a SuperReal PreMix Plus (SYBR Green, TIANGEN, Beijing, China). Finally, the 2‐ΔΔCt relative quantification method was used to analyze the experimental data in Microsoft Excel software.

2.10 | Cell morphology and cell immunochemical detection

The cells cultured in a 24‐well plate were fixed with 4% formaldehyde 10 min at room temperature (RT) and were washed with PBS for two times, 5 min each. The cells were permeabilized by 0.1% Triton X‐100 for 10 min at RT and were blocked for minimum 30 min with 1% BSA at RT. Then the cells were incubated with primary antibodies specific against CVH (1:400), C‐KIT (1:400), ITGA6 (1:500), ITGβ1 (1:500) (all from Abcam, Cambridge, UK), respectively for overnight at 4°C. The appropriate FITC‐ or TRITCconjugated secondary antibodies were used following the manufacturer’s manual (1:1,000, Abcam). Concurrently, the negative controls were stained with conjugated secondary antibodies alone. All cells were incubated for 2 hr at 37°C in the dark, washed 3 times with PBS‐T. The nuclei of cells were stained by 4′,6-diamidino-2phenylindole (DAPI) for 15 min. Images were captured with fluorescence microscope.

2.11 | Flow cytometry assay

Cells were stained with antibodies against CVH, C‐KIT (all from Abcam), to assess the changes in PGC‐related gene expression. To assess the level of SSC‐related gene expression, cells were stained with ITGA6 and ITGβ1 (all from Abcam). Cells were analyzed by a BD FACS Aria flow cytometer (BD Biosciences).

2.12 | Statistical analysis

All data were expressed as mean ± SE (X ± SE), using GraphPad Prism 5 statistical software, using t test between two sample means, p < 0.05 for significant difference; p < 0.01 for significant difference.

3 | RESULTS

3.1 | Verification of Nanos2 promoter region (positions −2 to −1189)

The amplified fragments (positions −2 to −1189) were isolated using electrophoresis in 1% agarose gel (Supporting Information Figure S1A). After identifying the recombinant plasmid pNanos2EGFP by Ase I, Kpn I double enzyme digestion, and bacterial liquid PCR, the nucleic acid sequencing results were compared with the Nanos2 5′‐flanking sequences in GeneBank to indicate that the recombinant vector was constructed successfully (Supporting Information Figure S1B). pNanos2‐EGFP, pEGFP‐N1, and pLinker‐EGFP were respectively transfected into DF‐1 cells. After transfection for 24 hr, pNanos2‐EGFP group expressed GFP at a rate of 40.5%, but its fluorescence intensity was weaker than that of the positive control pEGFP‐N1 group, whereas the negative control pLinkerEGFP group which deleted promoter region has no expression of GFP (Supporting Information Figure S1C). Therefore, the cloned fragment of the Nanos2 5′‐flanking sequences can initiate GFP expression with promoter activity.

3.2 | Detection of Nanos2 core promoter region (positions −157 to −495 bp)

The different length fragments (1,187 bp, 959 bp, 788 bp, 493 bp, and 155 bp) of Nanos2 promoter were amplified by PCR (Supporting Information Figure S2A). After identifying a series of incomplete promoter vectors by double enzyme digestion and bacterial liquid PCR, the positive samples sequencing results were consistent with expectations, indicating that the vectors were successfully constructed (Supporting Information Figure S2B), and respectively named pGL3‐1187, pGL3‐959, pGL3‐788, pGL3‐493, and pGL3‐155.
These vectors were respectively cotransfected with pRL‐SV40 and the pGL3‐basic group, which were used as a negative control. Dual‐luciferase reporter assays showed that the activity of −2 to −157 bp was essentially lost and no difference with the negative control. The activity of −157 to −495 bp was significantly higher than that of −2 to −157 bp, meanwhile −790 to −961 bp significantly increased once again (Figure 1a), indicating that there were important positive regulatory elements at −157 to −495 bp and −790 to −961 bp, especially at −157 to −495 bp which was estimated the core promoter region. For further verifying whether the −157 ~ −495 bp was the core region, constructing deletion vector named pGL3‐849 (−157 to −495 bp deletion) and vector pGL3‐694 (−495 to −1,189 bp deletion) (Supporting Information Figure S2C,D), then respectively cotransfecting DF‐1 cells with pRL‐SV40 compared with pGL3‐1187 and pGL3‐155 (Figure 1b). Results showed that promoter activity decreased significantly and almost lost after deleting −157 to −495 bp (Figure 1c), indicating that −157 to −495 bp is indeed the core promoter region of Nanos2.

3.3 | Foxd3 is a key transcription factor of Nanos2 core promoter region

According to the candidate transcription factor score, matching degree, species and some other standards, we screened out Foxd3 and Nobox these two highest scoring transcription factors in −157 to −495 bp of Nanos2 that most likely be the key transcription factor of Nanos2 (Figure 2a). In our preliminary screening for the transcription factors, all factors we screened are in different sites and have no cross, so they do not affect the other transcription sites. After that, we used homologous recombination technology to construct Foxd3 and Nobox binding site deletion vectors. Sequencing results indicated that the vectors were successfully constructed (Supporting Information Figure S3A–C), severally named pGL3/493‐Foxd3 and pGL3/ 493‐Nobox, and then cotransfected DF‐1 cells with pRL‐SV40. The pGL3/493 group was set as a positive control, whereas the pGL3basic group was a negative control. Dual luciferase reporter assays results showed that promoter activity was significantly upregulated after deleting the Foxd3 binding site, but there was no significant change after deleting the Nobox binding site (Figure 2b), indicating that Foxd3 is a key transcription factor of Nanos2 core promoter region, and plays a negative role, then the Foxd3 binding site deletion and overexpression vectors were constructed separately (Supporting Information Figure S3D,E) to further verify the combination of Foxd3 and Nanos2 promoters, the luciferase reporter assay results showed that the overexpression of Foxd3 binding sites significantly reduced  the activation activity (p < 0.01), which further indicated that Foxd3 is a negative regulatory factor (Figure 2c). CHIP results showed that Foxd3 can directly bind to the Nanos2 promoter, and the loss of the Foxd3 binding sites significantly decreased the binding to the Nanos2 promoter (p < 0.01), while the overexpression was opposite (Figure 2d).

3.4 | Foxd3 suppresses the expression of Nanos2 thus restrains the differentiation of ESCs into SSCs

The Foxd3 binding site deletion and overexpression vectors were validated in vitro using the RA induction system established in our laboratory.
Through cell morphology observation on Days 2–10, we can see that, in the control group, small EBs (Embryoid Bodies) began to appear on Day 4, and then the number of EBs gradually increased; EBs were broken on Day 8; on Day 10, most EBs completely ruptured and released cells. In the Foxd3 DEL group, larger EBs appeared on Day 4, the number of EBs was also more than that of the control group, and then the EBs increased; the EBs were broken on Day 6; on Day 8, the majority of EBs completely ruptured and released cells; SSC‐like cells appeared on Day 10. In the Foxd3 OE group, there were no obvious EBs on Days 2–6; on Days 8–10, apparent small EBs appeared but was less than the control group (Figure 3a).
On Day 0, 4, and 10 after induction, qRT‐PCR was performed on each group and the relative expression of related genes was  detected. On Day 4, CVH and Nanos2 expression in the Foxd3 DEL group was significantly higher than those in the control group (p < 0.05), and c‐kit expression in the Foxd3 OE group was significantly lower than those in the control group (p < 0.05), Nanos2 expression was significantly downregulated compared to the control group (p < 0.01). On Day 10, integrin α6 and integrin β1 expression in the Foxd3 DEL group was not significantly different from the control group (p > 0.05), Nanos2 expression was significantly higher than the control group (p < 0.01). Integrin α6 expression in the Foxd3 OE group was significantly downregulated compared to the control group (p < 0.05), and Nanos2 expression was significantly downregulated compared to the control group (p < 0.01) (Figure 3b).
Further detection by immunofluorescence staining found that the expression of PGC marker CVH and C‐KIT in the Foxd3 DEL group was higher than that in the control group on Day 4, whereas the Foxd3 OE group was lower than the control group. On Day 10, the expression of the SSC marker ITGA6 and ITGβ1 in the Foxd3 DEL group was higher than that in the control group, while the Foxd3 OE group was lower than the control group (Figure 4a,b).
To further detect changes in germ cells, cells were respectively labeled with CVH, C‐KIT, and ITGA6, ITGβ1 for flow cytometric analysis. The results showed that the positive cell rates of the control group, the Foxd3 DEL group, and the Foxd3 OE group were 14.91% ± 1.23%, 21.40% ± 2.67%, and 13.83% ± 0.74%, respectively. The Foxd3 DEL group was significantly higher than the control group and Foxd3 OE group (p < 0.05), and there was no significant difference between the Foxd3 OE group and the control group (p > 0.05). On Day 10, the positive cell rates of the control group, the Foxd3 DEL group, and the Foxd3 OE group were 6.04%. ± 0.49%, 8.19% ± 1.27%, 3.04 ± 0.11%, respectively. The positive cell rate of the Foxd3 DEL group was significantly higher than the control group and the Foxd3 OE group (p < 0.05), whereas the Foxd3 OE group was significantly lower than the control group (p < 0.05) (Figure 4c,d).
The results indicated that the loss of Foxd3 binding site could upregulate the expression of Nanos2 thus promote the differentiation of chicken ESCs into SSCs, whereas its overexpression caused the opposite effect.

3.5 | 5‐Azadc and TSA co‐induction can upregulate the expression of Nanos2 thus promote the differentiation of ESCs into SSCs

To further analyze the epigenetic function to Nanos2, we predicted the CpG island of Nanos2 promoter online. The prediction showed that there was no CpG island in this region, but existed scattered CpG sites (Supporting Information Figure S4A). After bisulfite treatment, the genome was amplified by 5 pairs of different methylation amplification primers, the third amplification product was the brightest (Supporting Information Figure S4B), the area was −143 to −409 bp, just contained Nanos2 core promoter region, there were five CG sites, respectively in −165 bp, −245 bp, −285 bp, −345 bp, and −381 bp. The sequencing results showed that C in these sites were not converted to T in some cases (Figure 5a), it explained that Nanos2 promoter was modified by methylation.
To further explore the effect of epigenetic modification on Nanos2, we screened the best inducing concentration of 5‐Azadc and TSA to induce better. The results showed that 5‐Azadc had the best induction effect at 5 μmol·L−1, which was significantly higher than other induced concentrations (p < 0.01). Simultaneously, the induction effect of TSA at 1.5 μmol·L−1 was the best (Figure 5b,c).
Dual luciferase assay was used to examine the effect of 5‐Azadc and TSA on the activity of Nanos2 promoter. The results showed that both 5‐Azadc and TSA could significantly enhance the Nanos2 promoter activity compared with the control group (p < 0.01), and the combinative induction of 5‐Azadc and TSA significantly enhanced the activity than either of them alone (p < 0.01) (Figure 5d).
Through cell morphology observation on Days 2–10, we can see that, in the control group, small EBs began to appear on Day 4, then gradually increased; on Day 8, the EBs were broken; on Day 10, they gradually disappeared and completely ruptured to release cells. In the 5‐Azadc group, larger EBs appeared on Day 4, and the quantity was more than the control group; on Day 6, the EBs began ruptured and released cells; The majority of EBs completely ruptured on Days 8–10, left EBs fragments. In the TSA group, there was no obvious EB on Day 2 and the number of EBs increased gradually on Day 4–6, and more than the control group; most EBs were broken on Day 6; most of the EBs ruptured completely and released cells on Days 8–10. In the 5‐Azadc + TSA group, small EBs began to appear on Day 2, and EBs were broken on Day 4. EBs increased significantly on Days 4–6, and more than other groups. On Days 8–10, most EBs were completely ruptured, left EBs fragments (Figure 6a).
On Day 0, 4, and 10 after induction, qRT‐PCR was performed on each group of cells and the relative expression of related genes was detected. The results showed that CVH, c‐kit, and Nanos2 expression in the 5‐Azadc + TSA group were significantly higher than other groups on Day 4 (p < 0.05); CVH and c‐kit expression in 5‐Azadc group and TSA group were not significantly different from the control group (p > 0.05), Nanos2 expression was significantly higher than the control group (p > 0.05). On Day 10, the relative expression of integrin β1 and Nanos2 in the 5‐Azadc + TSA group were significantly higher than other groups (p < 0.01), and integrin α6 expression was significantly higher than the control group (p < 0.05); integrin α6, integrin β1, and Nanos2 relative expression in the 5‐Azadc group and TSA group were significantly higher than the control group (p < 0.05); while the relative expression of integrin β1 in the 5‐Azadc group was significantly higher than the TSA group, and no significant changes between these two group in the relative expression of integrin α6 and Nanos2 (Figure 6b).
To further detect changes in germ cells, cells were respectively labeled with C‐KIT and ITGA6 for flow cytometric analysis. On Day 4, the results showed that the positive cell rates in the control group, the 5‐Azadc group, the TSA group, and the 5‐Azadc + TSA group were 44.6% ± 0.21%, 66.5% ± 0.17%, 72.0% ± 0.45%, and 78.5 ± 0.38%, respectively. The positive cells rates in the 5‐Azadc group, TSA group, and 5‐Azadc + TSA group were significantly higher than the control group (p < 0.05), and the positive cell rate in the 5‐Azadc + TSA group was significantly higher than the 5‐Azadc group and TSA group (p < 0.05). On Day 10, the positive cells rates in the control group, 5‐Azadc group, TSA group, and 5‐Azadc + TSA group, respectively were 46.6% ± 0.36%, 54.4% ± 0.94%, 54.8% ± 0.12%, and 59.2% ± 2.59%, the positive cells in 5‐Azadc group, TSA group, and 5‐Azadc + TSA group were significantly higher than the control group (p < 0.05), while there was no significant difference between 5‐Azadc group and TSA group (p > 0.05). The 5‐Azadc + TSA group was significantly higher than the 5‐Azadc group and the TSA group (p < 0.05) (Figure 7c,d).
It showed that the expression of Nanos2 was modified by DNA methylation and histone acetylation. 5‐Azadc and TSA could upregulate the expression of Nanos2 thus promote the differentiation of ESCs to SSCs, and the joint induction of both can enhance the induction effect.

4 | DISCUSSION

Although abundant studies have shown that ESCs can be induced to differentiate into male germ cells in vitro (Cai et al., 2013; Shirazi, Zarnani, Soleimani, Nayernia, & Ragerdi Kashani, 2016), the induction efficiency is uneven due to differences in the induction and culture systems.The occurrence of male germ cells lacks thorough and meticulous research on the regulatory mechanisms, and that make it impossible to continuously obtain high purity and quality male germ cells in vitro, further making it difficult to be applied to clinical medical care or actual production. Based on the results of previous transcriptome sequencing of ESCs, PGCs, and SSCs, we identified several key genes and signaling pathways that regulate the differentiation of chicken ESCs into male germ cells, Nanos2 is one of them. Previous studies have shown that Nanos2 can promote chicken ESCs differentiate into male germ cells, but what molecular mechanism Nanos2 uses to regulate the transcription and expression needs further exploration. This study aims to elucidate the expression regulation mechanism of Nanos2 in chicken male germ cell formation to establish a stable and efficient differentiation, and culture system in vitro and obtain high quality male germ cells, providing theoretical support for the further use of male germ cells to produce genetically modified chickens.
Our previous work found that the loss of Nanos2 could cause the fail of gonad mature and inhibit the production of PGCs and SSCs. However, some studies have shown that the action mechanism of Nanos2 in different species is not the same. In mice, Nanos2 only expresses in males, has important regulation on the differentiation and maintenance of SSCs. In buffalos, Nanos2 is abundant in embryonic and prepubertal testis, but is less expressed in testis of adult buffalo and is limited to prolonged sperm growth, indicating that Nanos2 can be used as potential spermatogonia molecule markers (Li et al., 2017). In dairy goats, large yellow croaker, and other species, this gene is not a male reproductive specific gene (Han, Chen, & Mingyi, 2018), but it is still essential for the generation of male germ cells. In pigs, knocking out the Nanos2 single allele of male pigs can improve their fertility, whereas knocking out one pair of Nanos2 alleles of female pigs will increase fertility, which may be related to the expression regulation mechanism of Nanos2 (Park, Kaucher, & Powell, 2017). Overexpression of miR‐34c inhibits the expression of Nanos2 and thus inhibiting the differentiation of SSCs (Yu et al., 2013), suggesting that Nanos2 may be regulated by epigenetic factors. Currently, whether or not the Nanos2 is affected by epigenetic regulation (DNA methylation and histone acetylation) has not been reported. This study found that Foxd3 binds to the Nanos2 core promoter region (−157 to −495), and negatively regulates the expression of Nanos2; 5 μmol·L−1 5‐Azadc and 1.5 μmol·L−1 TSA combinatorial induction can significantly increase Nanos2 expression thus promotes differentiation of ESCs to SSCs in vitro.
It is known that Nanos2 acts as an inhibitor of genes that involved in meiosis, such as Stra8, and it inhibits meiosis Trichostatin A and initiates a male differentiation program (Saba, Kato, & Saga, 2014). To further improve the regulatory network of Nanos2, we speculate that Nanos2 is regulated by transcription factors and epigenetic modification during the suppression of meiosis: When the Foxd3 binding site is deleted in Nanos2 promoter, the expression restriction of Nanos2 is relieved. Nanos2 inhibits the expression of Stra8 after it is expressed in large quantities, thereby inhibiting meiosis and promoting differentiation of male germ cells; when the Foxd3 binding site increases, excessive Foxd3 binds to the Nanos2 promoter and inhibits the expression of Nanos2, and then the expression of Stra8 increases and promotes the occurrence of meiosis; DNA methylation can inhibit Nanos2 promoter activity, whereas 5‐Azadc can reduce the methylation degree of Nanos2 so as to improve its promoter activity and promote its expression, TSA also can upregulate the Nanos2 promoter activity, thereby promoting its expression (Figure 8). This study preliminarily clarified the expression and regulation mechanism of Nanos2, providing basis for the use of this gene to increase the production efficiency of male germ cells, and also providing a reference for using male germ cells to product genetically modified chickens in animal husbandry.

REFERENCES

Cai, H., Xia, X., Wang, L., Liu, Y., He, Z., Guo, Q., & Xu, C. (2013). In vitro and in vivo differentiation of induced pluripotent stem cells into male germ cells. Biochemical and Biophysical Research Communications, 433, 286–291.
Cao, Z., Hong, R., Ding, B., Zuo, X., Li, H., Ding, J., … Zhang, Y. (2017). TSA and BIX‐01294 induced normal DNA and histone methylation and increased protein expression in porcine somatic cell nuclear transfer embryos. PLoS One, 12(1), e0169092. 23
Fila‐Danilow, A., Borkowska, P., Paul‐Samojedny, M., Kowalczyk, M., & Kowalski, J. (2017). The influence of TSA and VPA on the in vitro differentiation of bone marrow mesenchymal stem cells into neuronal lineage cells: Gene expression studies. Postepy Higieny i Medycyny Doswiadczalnej, 71(0), 236–242.
Han, Kunhuang, Chen, Shihai, & Mingyi, Cai, et al. (2018). Nanos3 not nanos1 and nanos2 is a germ cell marker gene in large yellow croaker during embryogenesis. Comparative Biochemistry and Physiology. Part B, Biochemistry and Molecular Biology, S1096‐S4959(18), 30002–2.
Li, K., Guo, Q., Yang, J., Chen, H., Hu, K., Zhao, J., … Li, L. (2017). FOXD3 is a tumor suppressor of colon cancer by inhibiting EGFR‐Ras‐Raf‐MEKERK signal pathway. Oncotarget, 8(3), 5048–5056.
Li, Z., Wong, K. Y., Chan, G. C. F., Chng, W. J., & Chim, C. S. (2018). Epigenetic silencing of EVL/miR‐342 in multiple myeloma. Translational Research, 192, 46–53.
Lmb, L. M. B., Chequin, A., Gcm, G. C. M., Biembengut, I. V., Toledo, M. B., Baura, V. A., … Klassen, G. (2018). MMP9 gene expression regulation by intragenic epigenetic modifications in breast cancer. Gene, 642, 461–466.
Murray, S. M., Yang, S. Y., & Van doren, M. (2010). Germ cell sex determination: A collaboration between soma and germline. Current Opinion in Cell Biology, 22(6), 722–729.
Park, Ki‐Eun, Kaucher, Amy V., & Powell, Anne, et al. (2017). Generation of germline ablated male pigs by CRISPR/Cas9 editing of the NANOS2 gene. Scientific Reports, 25(2), 209–210.
Pui, Han Pin, & Saga., Yumiko (2018). NANOS2 acts as an intrinsic regulator of gonocytes‐tospermatogonia transition in the murine testes. Mechanisms of Development, S0925‐4773(17), 30772–30774.
Saba, R., Kato, Y., & Saga, Y., et al. (2014). NANOS2 promotes male germ cell development independent of meiosis suppression. Developmental Biology, 385(1), 32–40.
Shirazi, R., Zarnani, A. H., Soleimani, M., Nayernia, K., & Ragerdi Kashani, I. (2016). Differentiation of bone marrow‐derived stage‐specific embryonic antigen 1 positive pluripotent stem cells into male germ cells. Microscopy Research Technique, 80(4), 430–440.
Zeng, Xianbin, Ji, Qiumei, & Ma, Zhijie, et al. (2013). Self‐renewal and differentiation regulation of yak and dzo spermatogonial stem cells. Proceedings of the 9th National Member Congress and Academic Symposium of the Chinese Society of Genetics, S823.851, 268–268.
Yao, X., Tang, F., Yu, M., Zhu, H., Chu, Z., Li, M., … Peng, S. (2014). Expression profile of Nanos2 gene in dairy goat and its inhibitory effect on Stra8 during meiosis. Cell Proliferation, 47(5), 396–405.
Yu, M., Mu, H., Niu, Z., Chu, Z., Zhu, H., & Hua, J. (2013). miR‐34c enhances mouse spermatogonial stem cells differentiation by targeting Nanos2. Journal of Cellular Biochemistry, 115(2), 232–242.
Zhang, Wenhui, Bi, Yulin, & Wang, Yingjie, et al. (2017). Nanos2 promotes differentiation of chicken (Gallus gallus) embryonic stem cells to male germ cells. Journal of Cellular Biochemistry, 10(1002), 26528.