Chromatin remodeling and H3K4me3 depletion regulate germline specification from pluripotency

Abstract

Background

Germ cells are the only cells capable of transmitting heritable genetic material to future generations. Epigenetic mechanisms that regulate germ cell formation are essential for optimizing offspring production, which is particularly important in farm animals like chicken. Primordial germ cells (PGCs), the precursors of gametes, could be derived from the pluripotent blastoderm cells (BC) or embryonic stem cell (ESCs) in chicken but the germline induction efficiency remain low and require further improvements.

Methods

We systematically profiled key histone modifications and chromatin states during the germ/soma specification from chicken pluripotent blastoderm cells to either PGCs or fibroblasts to uncover the chromatin regulators that direct the germline specification. The histone methyltransferase was perturbed during germ cell differentiation to assess the effect of histone modification on germline induction.

Results

The specific alterations of chromatin states could instruct the expression of germline genes and repress the pluripotency or somatic gene program in distinct cell types. The dynamic chromatin activation at both promoters and enhancers contribute to germline induction from pluripotency. Interestingly, the diminished active histone modification H3K4me3 regulate the transitions of bivalent states into repressive to facilitate the specification of the germ cell lineage. We demonstrated that selectively erase the H3K4me3 modifications could block the expression of BMP signaling antagonists, thereby enhancing the creation of PGC-like cells (PGCLCs) in chicken.

Conclusions

The comprehensive analysis of gene expression and chromatin regulation patterns during germ/soma segregation reveals that chromatin reprogramming mechanisms play crucial roles in controlling germline specification. This research also provides new epigenetic strategies to enhance the production of germ cells.

Graphical Abstract

Supplementary Information

The online version contains supplementary material available at 10.1186/s13287-025-04584-z.

Background

Primordial germ cells (PGCs) are the precursors of sperm and eggs, serve the unique role of transmitting genetic information to the next generation. Constituting only a small proportion of cells in the early developing embryo, PGCs are distinctly separate from somatic cells. Throughout their specification, development and migration, PGC continuously engage in dynamic interactions with neighboring somatic cells and surrounding microenvironment. These interactions necessitate the tight control over the germline programming to prevent inappropriate deviations from germline lineage to differentiate into alternative cell fates [1]. A comprehensive understanding of the underlying epigenetic regulation mechanisms and the corresponding gene expression programs during germ cell specification and development is prerequisite for the derivation of healthy and competent gametes. Particularly, the developmental trajectory and epigenetic features of avian PGCs remain less well-characterized compare to mammalian counterparts.

Avian species, particularly chickens, have been extensively utilized in developmental biology and germ cell research. Chicken PGCs serve as a robust cellular model for studying germline development, genetic manipulations, and germplasm conservation. While significant progress has been made in uncovering the regulatory mechanisms of chicken PGC development, particularly in transcriptome analysis as well as DNA methylation patterns [2–5], the regulations of chromatin structure and its corresponding transcriptional output during this process is not fully understood. Few studies reported the distinct histone modifications between chicken pluripotent blastoderm cells and specified PGCs [6, 7], but the detailed chromatin landscape and comprehensive chromatin reprogramming during germline specification are not investigated.

In birds, PGCs are believed to originate via the preformation model, but they can also be induced from pluripotent blastoderm cells [8]. Similar to mammals, chicken pluripotent blastoderm cells or embryonic stem cells (ESCs) are capable of both self-renewal and differentiate into either somatic or the germ cells [8–10]. Different labs have successfully generated primordial germ cell-like cells (PGCLCs) from chicken ESCs in culture, but with various efficiencies [11–13]. Nonetheless, these PGCLCs have demonstrated the ability to adopt the germ cell fate when transplanted into recipient embryos and produce viable offsprings [12–14]. However, it remains unclear whether these induced PGCLCs can faithfully reconstitute the epigenome modifications similar to the native PGCs. Moreover, the in vitro mass production of PGCLCs offers the opportunity to obtain large quantities of viable gametes for targeted genetic manipulations, thereby bypassing the need to collect the limited number of endogenous PGCs from each embryo. Despite these favorable features, induced PGCLCs typically exhibit compromised competency in gonad colonization and offspring production compare to native genuine PGCs, indicate incomplete germline programming in vitro. Conducting the comparative analysis of chromatin regulations among the pluripotent blastoderm cells, germline restricted PGCs and somatic fibroblast cells could aid in designing epigenetic strategies to enhance the germline specification in chickens.

In this study, we investigated the chromatin reprogramming, in particular the histone modifications, during the differentiation from pluripotent cells to either somatic or germ cells. Through systematic comparison of chromatin states and the associated gene expression patterns in chicken blastoderm cells, PGCs and chicken embryo fibroblasts (CEFs), we identified the cis-regulatory elements and the putative epigenetic factors that could specify the germ cell lineage. Notably, the histone modification H3K4me3 was reduced in native PGCs, and its decline leads to diminished expression of pluripotency and somatic genes, as well as the BMP antagonists. Therefore, we purposely inhibited the H3K4me3 deposition during the differentiation from blastoderm-derived ESCs to PGCLCs, which greatly enhance the germline induction efficiency.

Method and materials

Animal and tissue

All eggs were purchased from a local farm. The use and care of animals complied with the guidelines of the Animal Advisory Committee at Shandong Agricultural University. The ethics application was approved by the Ethics Committee of Shandong Agricultural University.

Primary cell isolation and culture of chicken blastoderm, PGC and CEF cells

Isolation and culture of stage X pluripotent blastoderm cells

Fresh fertilized eggs (Hy-line brown chickens) were disinfected by 75% alcohol and were broken by tweezers. Then the EG&K stage X blastoderm was taken out by scissors and spoons and was washed with PBS for three times to remove the yolk and vitelline membrane. The tissue was dissociated into single cells using mechanical pipetting. The blastoderm cells were filtered with a 40 μm mesh nylon cloth and centrifuged for 5 min with 1000×g. The blastoderm cells were incubated for 10–20 min with SSEA-1 (stage-specific embryonic antigen 1) antibody (Santa Cruz, sc-21702), then the cells were washed with 1 mL of magnetic-activated cell-sorter (MACS) buffer (0.5% BSA and 2 mM EDTA), and the supernatant was completely removed by centrifugation. The cell pellet was mixed with 50 µL MACS buffer supplemented with 20 µL anti-mouse IgM microbeads (Beyotime) at room temperature for 20–30 min, and the microbeads could bind to the target cells during incubation, followed by magnetic separation of the beads-binding cells using a magnetic stand. The cells were washed by MACS buffer and collected. For RNA-seq and ChIP-seq, the stage X SSEA1+ blastoderm cells were directly subjected to dissociation or fixation. For immunostaining and different treatments, the cultured pluripotent blastoderm cells (BC) were then designated as embryonic stem cells (ESCs) and maintained in knockout Dulbecco’s modified Eagle’s medium (KO-DMEM, Gibco) supplemented with 10% fetal bovine serum (Gibco), 1% GlutaMAX-I supplement (Gibco), 0.02% β-mercaptoethanol (Solarbio), 0.4% non-essential amino acids (NEAA, Gibco), 2% chicken serum (Solarbio), 1% antibiotic antimycotic (Gibco), 10 ng/mL Leukemia inhibitory factor (LIF, PeproTech), 10 ng/mL stem cell factor (SCF, PeproTech) and 20 ng/mL bFGF (Solarbio), and the CEF as feeder cells.

Isolation and culture of PGCs

For the collection of PGCs, gonads at HH stage 27–28 were dissociated using 0.25% trypsin–EDTA, and the gonadal cells were incubated for 10–20 min with SSEA1(Santa Cruz, sc-21702), then the cells were purified as the same MACS protocol of BCs. For RNA-seq and ChIP-seq, the cells were directly subjected to dissociation or fixation. For immunostaining and different treatments, the isolated PGCs were cultured and amplified in the medium of knockout Dulbecco’s modified Eagle’s medium (KO-DMEM, Gibco) 39 °C in 5% CO₂ with saturated humidity. The culture medium was modified from previous protocols [15, 16]. In brief, the KO-DMEM supplemented with 0.2% chicken serum (Solarbio), 0.5% GlutaMAX-I supplement (Gibco), 0.5% antibiotic antimycotic (Gibco), 0.5% non-essential amino acids (NEAA, Gibco), 10 ng/mL LIF (PeproTech), 10ng/mL basic fibroblast growth factor (Sigma), 0.05% β-mercaptoethanol (Solarbio), and 100 µg/mL heparin sodium (Solarbio).

Isolation and culture of CEFs

CEF was isolated from E5.5 chicken embryos (HH St. 27). Fertilized eggs were sterilized with 75% alcohol, after 5.5 days of incubation at 37.8℃ and 60% relative humidity. Chicken embryos were taken out and washed with PBS for three times. After removing the head, limbs, tail, internal organs and bones, the remaining tissue was crushed and digested with 0.25% trypsin for 15 min. An equal volume of RPMI 1640 medium (Gibco) was added to terminate digestion. The cell suspension was filtered with a 40 μm mesh nylon cloth and centrifuged for 5 min with 1000×g. The cell pallet was resuspended by the RPMI 1640 medium containing 10% fetal bovine serum (Gibco) and cultured at 39 °C in 5% CO₂ with saturated humidity [17]. The cells were cultured overnight for the subsequent RNA-seq, ChIP-seq and immunostaining assays.

ESCs induction into PGCLCs

The ESCs induction into PGCLCs was modified from previous protocols [14]. Briefly, the cultured ESC spherical colonies were centrifuged and mechanically dissociated into single cells by pipetting, and the cells were seeded in a 12-well plate with ESC medium. The embryoid body-like structures (EBLSs) formed within 2 days of culture. Subsequently, these EBLSs were centrifuged and dissociated into single cells by digestion with 0.25% trypsin, and the cells were plated and co-cultured with the CEF feeders in ESC medium supplemented with 40 ng/mL BMP4 (ProSpec, Rehovot, Israel, CYT-361), 40 ng/mL BMP8b (ProSpec, Rehovot, Israel, CYT-830) and 50 ng/mL EGF (ProSpec, Rehovot, Israel, CYT-217), and the treatment group (TRT) was additionally added with 50 µΜ OICR-9429 (APE×BIO, Houston, B6168). Half of the medium is renewed every day.

Induction of meiosis

Induction of meiosis was achieved by aggregation of PGCs or PGCLCs with the addition of meiosis inducers. For PGCLCs preinduction, ESCs were plated on 12-well plates in medium supplemented with BMP4, BMP8b, EGF and OICR-9429, and the medium was changed after 4 days preinduction. Aggregates were cultured for 48 h to induce meiosis in the PGCs medium or ESCs induction medium, containing 1 μm retinoic acid (Solarbio). CEFs and ESCs as the negative controls for inducing meiosis, and they were also treated with retinoic acid-containing medium for 48 h.

Immunofluorescence staining

The cells were fixed for 10 min with 0.4% PFA and were washed with PBS for three times, and then incubated with 0.5% TritonX-100 for 10 min and washed with PBS for three times again. Then Blocking Buffer was used to block the cells for 1–2 h. After washing with PBS, the cells were incubated with primary antibody at 4℃ for overnight. The next day, the cells were washed with PBS for three times and incubated with the Alexa Fluor-labelled secondary antibody at 1:500 dilution at room temperature for 2 h. Finally, 50 ng/mL DAPI was used for nuclear staining for 10 min at room temperature and were observed by the fluorescence microscope [18]. The antibodies are listed in Supplementary Table S2.

mRNA extraction and real-time quantitative PCR

Total RNA was extracted by QIAsymphony RNA Kit (Qiagen, 931636). RNA concentration was measured on NanoDrop2000 (Thermo Scientific, USA). 1 µg of RNA was reverse transcribed using PrimeScript™ RT reagent Kit (TAKARA, #RR047A). Real-time quantitative PCR (RT-qPCR) was performed using SYBR Green Mix (Abclonal, #RK21203) following manufacturer’s instructions.Expression was normalized to GAPDH using 2-ΔΔCt method. The error bars indicate the SEM. The RT-qPCR primers are described in Supplementary Table S3.

RNA extraction and RNA-seq analysis

RNA was extracted using TRIzol (Simgen, 5301100); two replicates were used for each sample. RNA-seq data were generated using Illumina sequencing. Clean reads were obtained by removing reads containing adapter and ploy-N and low-quality reads from raw data using FastQC and Trimmomatic (version 0.39). Hisat2 (version 2.1.0) was used to align reads against the galGal6 reference genomic assembly of gallus gallus. The gene read counts were calculated using featureCounts (version 2.0.1) and normalised to transcripts per kilobase million (TPM), which were annotated using Gallus_gallus.GRCg6a.105.chr.gtf. Differential gene expression analysis between two groups was performed using the R package DESeq2 (|log2FC| ≥ 1, P-value < 0.01). The Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) functional enrichment analyses were implemented using the Metascape database (http://metascape.org/gp/index.html) and David (https://david.ncifcrf.gov/). The genes enriched in the items related to PGC (GO:0007281, GO:0008406 etc.) [19, 20] development were collected and used for further analysis. Then relationship was performed within gene read counts to analyze the similarity between different cell populations. R packages were used for the graphical representation of principal component analysis (PCA) plots, correlationship and heatmaps, respectively. Genes with TPM ≥ 1 in at least one sample were defined as detected genes.

Chromatin Immunoprecipitation followed by next-generation sequencing (ChIP-Seq) library construction and analysis

ChIP experiments were performed in the isolated BC, PGC and CEF in duplicates. First, the cells were cross-linked using 1% formaldehyde at room temperature, followed by quenching using 200-mM glycine. Sonication was set at 34% power and 20-s ON and 300-s OFF cycles for 6 min. Chromatin was incubated at 4℃ overnight with different histone antibodies. The libraries were purified and subjected to sequencing using HiSeq2000 (Illumina) according to the manufacturer’s instructions. Clean data were checked using FastQC and trimmed to remove Illumina Nextera adapter sequences using trim_galore with the ‘--q 25, --nextera and --fastqc’ options. Low quality reads were removed using Trimmomatic with default parameters. Bowtie2 (version 2.3.5.1) [21] was used to align the reads to the galGal6 using the ‘--no-mixed, --no-discordant, --no-unal, --time and --omit-sec-seq’ options. Samtools was used to sort and isolate uniquely mapped reads. Picard MarkDuplicates was subsequently used to remove duplicates using the ‘VALIDATION_STRINGENCY = LENIENT’ option. The final BAM files were generated using Samtools with the ‘view -bS -q 30 -F 1804 -f 0 × 2’ option. This resulted in the final aligned, de-duplicated BAM file that was used for all downstream analyses. The two duplicates within each group were highly reproducible and then combined to call peaks. The peaks were identified using MACS2, with a P-value threshold of 0.01 (version 2.1.1). All alignment results were converted to coverage bigwig files and normalized to the corresponding input using deepTools (version 3.0.2) [22]. R packages were used for the graphical representation of correlation analysis of these samples. The R package ChIPseeker was used to identify the nearest genes around the peak and annotate genomic regions of the peaks [23]. deepTools was used for PCA and Correlation analysis and for assessing the enrichment profiles of all the ChIP-seq data. The bigwig formats can be visualised using the Integrative Genomics Viewer (IGV) software.

Chromatin state analysis and alluvial plotting

Chromatin states were identified and characterized using ChromHMM (v1.11) [24]. The alignment files of H3K4me3, H3K27ac and H3K27me3 modifications across 3 cells were binned into 200-bp bins using the BinarizeBam command, with the input alignment file as the control. Chromatin states were analyzed at a fold threshold of 10. Next, we trained the model with 7 emission states using 200-bp resolution and default parameters using the LearnModel command. Finally, at each stage, the whole genome was classified into seven states. Among these states, TssFlnk, Enh and Biv states were used for further alluvial plotting.

Alluvial diagrams of cell differentiation lineages were plotted using the alluvial function in R to show the transitions. The total regions of each chromatin state were counted once the 100-bp intervals had been marked by the specific states in that lineage, and the percentages of the specified intervals in each stage were plotted to show the global trend of that specific chromatin state. The alluvial diagrams showed the percentage changes of chromatin states during each transition; the lines from the present stage to next stage cannot be traced, as they represent different genomic locations. State associated genes were subsequently annotated using ChIPseeker.

Motif analysis

The ‘findMotifsGenome’ function of HOMER (version v3.1) [25] was used to identify motifs significantly enriched in different chromatin state regions in three types of cells, and the ‘mergePeaks’ function was used to overlap the peaks to identify the special and shared peaks among the samples.

Statistics and reproducibility

The animal experiments were not randomized. No statistical method was used to predetermine sample size. No samples were excluded from the statistical analysis. Statistical analysis was performed using the independent samples Student’s s-test (two-sided). Data were presented as mean ± standard error of mean (SEM), *P < 0.05, **P < 0.01, ***P < 0.001. Sample sizes for each experiment were indicated in the relevant results section or figure legends. Statistical analysis was performed in R (v3.4.1) and GraphPad Prism v8 software.

Results

Global transcriptome analysis revealed the expression of germline-specific genes

In order to discover the molecular mechanisms regulating chicken germ cell specification from the pluripotent stem cells, we first analysed the gene expression patterns of stage X pluripotent blastoderm cells (BC) and gonadal primordial germ cells (PGC) by RNA-seq, and then compare to the somatic chicken embryonic fibroblast cells (CEF) to identify the germline-specific gene expression features. First, the chicken BC, PGC and CEF cells were isolated and their identities and purities were examined by detecting the expression of specific markers (Fig. 1A). The pluripotent BC cells were purified by MACS with SSEA-1 antibody and they showed unanimously expression of the pluripotent stem cell marker SSEA-1. Gonadal PGCs exhibited strong positivity for both SSEA-1 and the germ cell-specific marker VASA (also named DDX4, dead box polypeptide 4). Both BCs and PGCs are negative for the fibroblast marker vimentin. In contrast, CEFs were negative for SSEA-1 and DDX4, while vimentin was homogeneously expressed in these fibroblast cells (Figure S1A, S1B, S1C). Hence, the immunofluorescence detection confirmed the cell identities of pluripotent, germline and somatic cells, respectively.

Fig. 1.

Transcriptome analysis of chicken pluripotent blastoderm cells (BC, St.X), gonadal primordial germ cell (PGC) and somatic chicken embryonic fibroblast (CEF). (A) Isolated EG&K stage X blastoderm cells (BC, St.X) and PGCs were cultured in vitro, the immunofluorescence staining of SSEA-1, DDX4 and vimentin were performed. Scale bars = 100 μm. (B) PCA analysis of the transcriptome of three types of cells (BCs, PGCs and CEFs). The biological replicates of RNA-seq showed highly reproducible results. (C) Correlation analysis of the three types of cells. The BCs and PGCs showed similar transcriptional profiles, consistent with PCA analysis. (D) Heatmap demonstrate different transcriptional patterns in each cell type after cluster analysis. Only genes expressed at least in one stage or group (TPM ≥ 1) were included. (E) GO analysis of cell-specific enriched DEGs of the three cells were illustrated. (F) GO analysis of genes that are enriched in both BCs and PGCs

Next, we analyzed the transcriptome of the three types of cells by RNA-seq to identify the cell-type specific gene expressions to discover the features of germline and somatic lineage segregation in chicken. Principal component analysis (PCA) of the overall transcriptome showed that germline PGCs were more similar to the pluripotent BCs other than somatic CEFs (Fig. 1B), which was further confirmed by the correlation analysis (Fig. 1C). In order to identify the specialized regulatory pathways controlling the cell identity of each cell type, we screened the differentially expressed genes (DEGs) (|logFC| >2-fold; P-value < 0.01) between different samples and further clustered the cell type-enriched DEGs using the heatmap to illustrate the unique gene expression patterns in different types of cells (Fig. 1D). The Gene Ontology (GO) analysis demonstrate signaling pathways that are enriched in each cell type. For instance, “regulation of cell adhesion”, “establishment or maintenance of cell polarity” related genes were expressed higher in pluripotent BCs, while “male meiotic unclear division” were specifically higher in germline PGCs. In contrast, “cell junction organization” and “extracellular matrix organization” related pathway genes were enriched in somatic CEFs (Fig. 1E). Moreover, BCs and PGCs shared many pluripotency and germline genes which are enriched in terms of stem cell development, such as “multicellular organismal homeostasis” and “embryonic morphogenesis” (Fig. 1F). We further investigated the expression pattern of PGC associated genes (GO:0007281 and GO:0008406) [17, 18] among different types of cells and found that these genes were expressed similarly between BCs and PGCs (r = 0.97), but very different between PGCs and CEFs (r = 0.76) (Figure S1D).

Therefore, the transcriptional profiles of chicken BCs, PGCs and CEFs indicate that specific cell type-enriched genes exhibit lineage characteristics of the respective pluripotency, germline and somatic features. We confirmed that stage X BCs exhibited similar transcriptional characteristics to PGCs, indicating a close biological relationship between these two types of cells.

Overview of chromatin histone modification profiles of pluripotent BC, germline PGC and somatic CEF

Histone modifications play critical roles in regulating gene expression and cell lineage changes during the early embryonic development. In order to find the unique epigenetic regulatory mechanisms that direct the chicken germline specification, we analyzed the global chromatin features by interrogating the canonical histone modification markers in the freshly isolated BC, PGC and CEF cells. ChIP-Seq was performed to profile the active and repressive chromatin histone modifications, including H3K4me3 (associated with promoters), H3K27ac (active enhancers), and H3K27me3 (repression by polycomb complex 2, PRC2). Principal component analysis (PCA) of the ChIP-Seq data revealed that the H3K27me3 and H3K27ac profiles clustered closely across all three cell types, whereas the H3K4me3 profiles were more distinctly separated (Fig. 2A). As expected, the repressive H3K27me3 modification mark showed a negative correlation with the active marks H3K4me3 and H3K27ac. In addition, the observation of a more distinct distribution of H3K4me3 among different cell types compare to other two histone modification markers was further supported by correlation analysis (Fig. 2B). It indicates that the H3K4me3 chromatin distributions show greatest variations among different cells and may play an important role in the germ/soma lineage segregations.

Fig. 2.

The general features of H3K4me3, H3K27ac, and H3K27me3 histone modifications in chicken pluripotent BCs, germline PGCs and somatic CEFs. (A) PCA analysis of the histone modifications in different types of cells including pluripotent blastoderm cells (BC, St.X), germline-restricted PGC, and somatic CEF. Please note that H3K4me3 showed bigger variations among different cells compare to H3K27ac, and H3K27me3. (B) Correlation analysis of the histone modifications in three types of cells. Please note that H3K4me3 exhibit negative correlation with other two histone markers. (C) The genomic distribution of H3K4me3, H3K27ac, and H3K27me3 peaks. (D) The signal profiles of H3K4me3, H3K27ac, and H3K27me3 modifications around the Transcription Start Site (TSS) of all genes in three types of cells (ratio to input). (E) Heatmap of the joint analysis of histone modifications and transcriptome data showed that chromatin activation correlates with gene expression. Please note that genes with broad depositions of H3K4me3 and H3K27ac exhibited the highest expression (top). Genes enriched with H3K4me3 and H3K27ac only at the TSS region showed moderate expression (middle), whereas genes marked by H3K27me3 showed lowest expression (bottom). (F) Representative IGV browser shows the mRNA expression levels were negatively correlated with the H3K27me3 depositions, as shown in blue and red framed regions

We further analyzed the distribution and intensity of these histone modifications across the genome. The H3K27ac and H3K27me3 were more or less evenly distributed to gene promoters, intergenic regions, introns and others. But the active chromatin mark H3K4me3 was predominantly enriched at gene promoters (30%-55%) (Fig. 2C, Figure S2). Accordingly, the signal intensity of H3K4me3 was peaked at the transcription start site (TSS) of all genes and the BCs exhibited much higher H3K4me3 deposition around TSS region compare to that in PGCs and CEFs. In contrast, the H3K27ac was less enriched at the TSS and the H3K27me3 was evenly low around the genes in all three types of cells (Fig. 2D). Therefore, the H3K4me3 deposition at the TSS could be a major driver to instruct the distinct gene expression program in different cells.

To understand the regulation of gene expression by chromatin modifications, we conducted an integrated analysis of transcriptome and histone depositions across all genes. Our findings revealed a close association between gene expression levels and specific chromatin modifications. For instance, genes with broad depositions of H3K4me3 and H3K27ac exhibited the highest expression levels. Genes enriched with H3K4me3 and H3K27ac only at the transcription start site (TSS) region showed moderate expression, whereas genes marked by H3K27me3 showed lowest expression (Fig. 2E). It indicates that these selective histone modifications could control gene expression. We also securitized the ChIP-seq and RNA-seq signals at specific genomic regions via Integrative Genomics Viewer (IGV) to verify the correlation of chromatin modification with gene expression. We confirmed that, in all three types of cells, the genes marked by high levels of H3K27me3 displayed minimal to no expression, irrespective of the levels of H3K4me3 and H3K27ac modifications (Fig. 2F). Hence, the chromatin histone modification features were well-correlated with gene expression patterns.

Changes in chromatin states during germline specification from pluripotency

To define unique aspects of the germline chromatin landscape, we wished to contrast PGC chromatin states with those in pluripotent stem cells and differentiated somatic cells. We identified active or inactive genome regions that correlate with particular combinations of histone modifications and associate these chromatin states with genomic features such as promoters, enhancers, exons, and introns. Next, we annotated chromatin states across the genome using ChromHMM [24]. Here, to discover chromatin features associated with germ/soma lineage segregation in an unbiased manner, we trained a 7-state model consisting of 3 active states and 4 repressed states (Fig. 3A). The active chromatin states, linked to expressed genes, include the active proximal promoter states around TSS region: TssA (H3K4me3 and H3K27ac) and TssFlnk (H3K4me3), as well as the enhancer state: Enh (H3K27ac). The inactive chromatin states include constitutive bivalent regulatory states: Biv (H3K4me3 and H3K27me3), repressed Polycomb states: ReprPC (strong H3K27me3), weak repressed Polycomb: ReprPCWk (weak H3K27me3), and the quiescent state: Quies (no mark).

Fig. 3.

Chromatin states change during the differentiation from pluripotent BCs to germline PGCs. (A) Seven chromatin states were defined using chromHMM. The activate states (TssA, red; TssFlnk, orange; Enh, pink), repressed states (Biv, green; ReprPC, blue; ReprPCWk, gray) and quiescent state (Quies, light brown) were shown. The genome coverage of each chromatin states was illustrated and the largest proportion among the three types of cells of each chromatin state was highlighted with blue. (B) The expression level of the genes associated with different chromatin states in three cell types were illustrated using boxplot diagram. TssA state induced higher gene expression than all the other states. (C) The correlation of each chromatin states in BCs and PGCs were illustrated and each state of PGCs were compared to each state in BCs to define the similarity between two cells. (D) Venn diagram showing the TssA-associated genes in three types of cells, BCs and PGCs shared more TssA-associated genes. (E) GO analysis of overlapped TssA-associated genes between BCs and PGCs. Biological processes were selected based on P value (P value < 1 × 10^− 10). (F) Alluvial plots showing the global dynamics of TssFlnk regions during BCs (left panel) differentiating to PGCs (right panel). Each line represents a 200-bp bin defined on the ChromHMM categories. (G) The distribution of the cell specific and shared TssFlnk regions were illustrated using ChIPseeker. (H) TF motif enrichment identified from BC and PGC specific TssFlnk regions

We first analyzed the distribution patterns of active and repressive chromatin states in different cell types (Fig. 3A). The active chromatin states including active/flanking TSS and enhancer exhibited similar distributions, covering 9.8–12.5% of the entire genome, with enhancers enriched in approximately 5.8–7.3% in each cell type. The repressive Bivalent and repressed Polycomb regions comprised 12.7%-15.4% of the genome among different cells. Interestingly, the pluripotent BCs have the highest proportions of bivalent regions, which is a typical signature in stem cells poised for differentiation [26]. As to the expression levels of genes associated with different chromatin states, we found that the genes in the active chromatin states at the TSS region, particularly the TssA region, showed highest transcripts abundance among all cell types (Fig. 3B). Hence, the active chromatin states indeed enhance the gene expression.

Next, we compared the same chromatin states among different cells to identify the most conserved and dynamic chromatin states during the germline specification. We found that the TssA state showed highest similarity between BCs and PGCs, with approximately 60% of BC_TssA regions remained stable in PGCs. Interestingly, the Tss-flank state exhibited great variations between BCs and PGCs, with only 8% of TSS-flank region remain unchanged during germline differentiation (Fig. 3C). Since the TssA modified regions are nearest to the genes and directly affect gene transcription, we examined the TssA-associated genes and the related biological pathways. A total of 4,858 TssA-associated genes were shared between BCs and PGCs, indicating that these genes were continuously activated by the active chromatin states in their promoters during germline specification (Fig. 3D). The GO analysis showed that these TssA-associated genes were mostly involved in “cell cycle”, “chromatin organization”, “fatty acid beta-oxidation” related processes that were important for germ cell function (Fig. 3E). Accordingly, among all the germline associated genes, more than one third of them were constantly activated by the TssA states (Figure S3). Hence, the conserved TssA chromatin states contribute to the germline maintenance during PGC specification.

Among all the different chromatin states, the TssFlnk showed the biggest variations between BCs and PGCs, with only approximately 8% of TssFlnk regions were shared between these two cell types (Fig. 2C). We found that 59.3% of newly formed TssFlnk regions in PGCs were derived from quiescent state (Quies) in the BCs, whereas 37.6% of TssFlnk in the BCs were de-activated into Quies in the PGCs (Fig. 3F). The BC_TssFlnk specific regions were predominantly located in the promoter regions (75.5%), whereas the majority of PGCs_TssFlnk specific regions (73.2%) were enriched in intronic and intergenic regions (Fig. 3G). Therefore, these dynamic chromatin states switched at the TssFlnk regions prompted us to identify the potential transcription factors (TFs) could modulate the changes of TssFlnk states, since TF-promoter interactions are major drivers for chromatin activation and gene expression. Critical pluripotency TFs such as SOX2, SOX4, and NANOG were found to have high-frequency binding sites in BCs. In PGCs, we also observed the enrichment of germline TFs, such as PRDM1 and NR5A2 (Fig. 3H). These important TFs enriched in the TssFlnk regions of PGCs were associated with BMP and WNT signaling pathways which play important roles in PGC formation.

In summary, both conserved and changed chromatin states were observed during the germline differentiation from BCs to PGCs. The conserved TssA states contribute to the stable expression of germline genes in both cells, while a group of pluripotency transcription factors could regulate the chromatin states switches during germline specification.

Reprogramming of enhancers during PGCs formation

Next, we analyzed the Enhancer (Enh) state in PGCs compared to BCs, since the newly generated Enh states would serve as distal regulatory elements to induce gene expression to regulate PGC formation (Fig. 4A). Approximately half of the enhancer regions were shared between the two types of cells (Fig. 4B). We further examined the chromatin activation and gene expression levels across the enhancer regions. As expected, higher H3K27ac signals were observed in BC-specific and PGC-specific enhancer regions compared to the same regions in the other cell type. Interestingly, we also found higher H3K27ac depositions in the shared enhancers in PGCs compared to BCs (Fig. 4C). Moreover, the PGC-specific enhancers induced more robust gene expression than the BC-specific enhancers (Fig. 4D). These findings underscore the crucial role of enhancers in PGC formation, consistent with previous studies in mammals [27, 28].

Fig. 4.

Changes of enhancers during differentiation from pluripotent BCs to germline PGCs. (A) Alluvial plots showing the global alterations of Enh regions during BCs (left panel) differentiating to PGCs (right panel). Each line represents a 200-bp bin defined on the ChromHMM categories. (B) Venn diagram of the Enh regions of BCs and PGCs. (C) Average signal profiles of H3K27ac modification in shared and specific enhancer regions of BCs and PGCs (ratio to input). (D) GO analysis of the genes associated with nascent enhancers in PGCs. Biological processes were selected based on P value (P value < 1 × 10^− 10). (E) Gene expression level of the genes associated with BC- or PGC-specific enhancers were calculated. The total transcriptome was used as control. (F) IGV browser examples of the enhancer associated gene during BCs differentiation to PGCs. Noted the significant increase of H3K27ac (green) around the enhancer regions of PGC associated genes

Furthermore, we detected the newly formed Enh regions in PGCs, which accounted for 55.3% of the total enhancer regions in PGCs. These enhancers were targeting genes related to germline functions such as “regulation of transcription by RNA polymerase II” and “signal transduction” “positive regulation of canonical Wnt signaling pathway” and “epidermal growth factor receptor signaling pathway” (Fig. 4E). Representative examples from IGV demonstrated that elevated enhancer activity, indicated by H3K27ac deposition, correlated with increased expression of target genes such as SMAD5 [29], STRA8 [30], and RFX2 [31], which are critical for germ cell development (Fig. 4F, S4A). Notably, we also observed heightened H3K27ac levels at regulatory regions of BMP4 in PGCs, suggesting its significant role in PGC formation (Fig. S4B). Thus, the presence of nascent enhancers in PGCs appears to be instrumental in facilitating PGC specification and development.

Loss of H3K4me3 at the bivalent state chromatin regions underlying cell-fate specification of germline lineage

Bivalent domains, which are defined by the paradoxical coexistence of the permissive H3K4me3 and repressive H3K27me3, are thought to play an important role in pluripotency by keeping developmental genes in a silenced state poised for activation later [26]. This bivalent configuration is thought to maintain lineage commitment programs in a poised state that prepare for future activation or repression. Thus, we analyzed the bivalent chromatin regions and the associated genes in both BCs and PGCs (Fig. 5A). We found that about 29% of bivalent chromatin regions remain stable during the BC to PGC specification, and these regions contain genes associated with “cell differentiation”, “somitogenesis” and “Notch signaling pathway” that are involved in somatic tissue development (Fig. 5B). It indicates that these somatic pathways are stably repressed in both pluripotency and germline cells.

Fig. 5.

The changes of bivalent chromatin state during germline specification. (A) Alluvial plots showing the alterations of bivalent regions during the differentiation from pluripotent blastoderm cells (BC) to germline PGCs. Each line represents a 200-bp bin defined on the ChromHMM categories, and the total regions are those classified as this state in at least one analysed cell. (B) GO analysis of the genes that are located in the stable bivalent regions between BCs and PGCs. (C) GO analysis of the genes enriched in the bivalent chromatin regions specific for BCs (lost in PGC). Biological processes were selected based on P value (P value < 1 × 10^− 10). (D) IGV browser examples of the multipotency genes, such as GATA5 and SOX2 in BCs and PGCs. Please not the significant decline of H3K4me3 around the regulatory region of the genes. (E) Profile plot showing the bivalent elimination. H3K4me3 were significantly declined from BCs (orange) to PGCs (red), while H3K27me3 remained relatively unchanged during BCs (dark green) differentiating to PGCs (light green)

We also observed over half (54.8%) of the bivalent regions in BCs transit into repressed states in PGCs, including ReprPC: 56.0%, ReprPCWk: 9.3%. Accordingly, 778 genes lose their bivalent state during germline specification from BCs into PGCs. These genes were mainly involved in pluripotency competency such as “embryonic system morphogenesis” and multi-tissue development, which are not required in the germ cells (Fig. 5C). For instance, the regulatory regions of pluripotency genes SOX2 and GATA5 were marked by both H3K4me3 and H3K27me3 in the ESCs, but in PGCs the H3K4me3 deposition was greatly reduced and the H3K27me3 become dominant to repress the gene expression (Fig. 5D). Overall, during the differentiation of BCs into PGCs, the elimination of bivalent domains was mainly driven by the decline of H3K4me3 depositions, whereas the H3K27me3 remain relatively stable (Fig. 5E).

In summary, the alterations of bivalent states are primarily influenced by fluctuated H3K4me3 modifications within the context of relatively stable H3K27me3. The reduction of H3K4me3 from pluripotent BCs to germline-restricted PGCs predominantly drives the elimination of bivalent regions. This change regulates gene expression potential, particularly by repressing stemness-related multipotency genes. Consequently, we hypothesize that the decline in H3K4me3 facilitates the specification of the germ cell lineage.

The induction efficiency of ESCs to PGCLCs was enhanced by depleting H3K4me3 at BMP signaling antagonist genes

Chicken PGCs are frequently utilized in gene editing and species conservation applications; however, their usage is hindered by the scarcity of cell populations and the difficulty of handling. Alternatively, the pluripotent blastoderm cell derived ESCs can be induced to differentiate into PGC-like cells (PGCLC), albeit with low efficiency. Considering the generalized decrease in H3K4me3 levels around gene promoters in PGCs compared to pluripotent BCs and ESCs, we hypothesized that the induction from ESC to PGC-like cells could be improved by purposely manipulating the H3K4me3 levels. Hence, we tried to inhibit the H3K4me3 deposition during the course of ESC-to-PGCLC differentiation to examine the germ cell induction efficiency. The small molecule OICR-9429 is a potent antagonist of WDR5, which is a core component of histone methyltransferase complex and is essential for the methylation of histone H3K4 [32, 33]. We asked whether H3K4me3 inhibitor OICR-9429 treatment could accelerate the process of germ cell specification through modulating the H3K4me3 levels in targeted genes.

We utilized the chicken PGCLC induction system to conduct a comparative analysis with and without H3K4me3 inhibitor OICR-9429 during germ cell derivation from ESCs. Unlike in mammals, where mesodermal specification precedes germ cell induction, chicken ESCs can be directly induced toward a germ cell fate without the need for mesodermal priming [2, 34]. During the early stage of germ cell induction process in chicken, we observed the embryoid body-like structures from day 2 in both OICR-9429 treatment and control groups. Notably, OICR-9429 treatment significantly increased the number of embryoid body-like structures from day 4 (Fig. 6A). Furthermore, RT-qPCR results showed that OICR-9429 treatment significantly upregulated the PGC marker genes (DDX4, DAZL, CXCR4, and BLIMP1) (Fig. 6B). The immunostaining of the PGC-specific protein DDX4 also demonstrated that OICR-9429 increased the efficiency of PGCLC generation from 20 to 29% (Fig. 6C). These findings indicate that the addition of H3K4me3 inhibitor OICR-9429 enhances the efficiency of PGCLCs induction from ESCs. Furthermore, we also examined the cell quality of the PGCLC upon the chemical inhibitor treatment. These cells can respond to the meiotic stimulations as shown by the significant upregulation of meiosis marker genes (SCP3, STRA8, SPO11) upon retinoic acid treatment in both PGCs and PGCLCs. Meanwhile, we observed an apparent downregulation of cellular apoptosis in PGC and induced PGCLCs compared to the ESC (Figure S5), demonstrating that WDR5-inhibited PGCLCs retained germ cell properties and cell viability (Figure S5).

Fig. 6.

PGC-like cells (PGCLC) induction was enhanced by depleting H3K4me3. (A) Chicken blastoderm-derived ESCs were induced to differentiate by BMP signaling and H3K4me3 inhibitor for 2d, 4d and 6d and the extent of differentiation was quantified by the number of embryoid body-like structures. n = 3; Scale bars = 100 μm. (B) The expression of PGC marker genes (DDX4, DAZL, CXCR4, BLIMP1) were assayed in ESCs, PGCs and the PGCLCs with/without H3K4me3 inhibition. (C) Representative images showing the immunofluorescence staining of DDX4 in the ESCs and PGCLCs with/without H3K4me3 inhibition. The germline induction efficiency was quantified by counting the DDX4 + cells. Scale bars = 100 μm

Since the BMP signaling was the major driver of germ cell specification in chicken [35, 36], we examined the H3K4me3 deposition at the core genes of the BMP pathway to determine if the OICR-9429 treatment could enhance the BMP signaling via modifying the chromatin activation status of these loci. Utilizing IGV browser, we first scrutinized the histone modifications at the BMP-associated genes in the native BCs and PGCs, respectively. Notably, we observed that H3K4me3 was enriched at the TSS region of certain BMP antagonist genes such as FSTL1, BMPER and NBL1 in pluripotent BCs, accompanied by elevated H3K27ac signals at the same loci. In contrast, germline-restricted PGCs displayed lower levels of H3K4me3 marks around these genes. Accordingly, the expression levels of these BMP antagonist genes were significantly reduced in PGCs compare to pluripotent BCs (Fig. 7A). These data led us to propose that the decreased H3K4me3 deposition in the germ cells may alleviate the antagonist effect on BMP signaling, thereby promoting germ cell formation.

Fig. 7.

H3K4me3 depletion reduced the BMP antagonist genes and somatic genes. (A) IGV browser examples of the antagonist genes of the BMP signaling pathway (FSTL1, BMPER and NBL1) illustrate the deposition of H3K4me3 (top; red), H3K27me3 (middle; blue) and H3K27ac (bottom; green) in ESCs and PGCs. Noted the dramatic reduction of H3K4me3 around the regulatory regions of the genes in PGCs compare to ESCs. (B) ChIP-qPCR analysis confirmed that the reduction of H3K4me3 deposition at the TSS regions of BMP antagonist genes upon OICR-9429 treatment. (C) RT-qPCR analysis show the gene expression of FSTL1, BMPER and NBL1 were significantly decreased after H3K4me3 depletion by OICR-9429. (D) IGV show the H3K4me3, H3K27me3 and H3K27ac deposition at the somatic differentiation genes (HOXB9, HOXD13, COL1A1) in ESCs and PGCs. (E) ChIP-qPCR analysis confirmed that the reduction of H3K4me3 deposition at the TSS regions of somatic genes HOXB9 and HOXD13 upon OICR-9429 treatment. (F) RT-qPCR analysis show the gene expression of HOXB9 and HOXD13 were significantly decreased after H3K4me3 depletion by OICR-9429. Data are expressed as the mean ± SEM. n ≥ 3. n.s. not significant, *P < 0.05, **P < 0.01, ***P < 0.001

However, during the in vitro induction of PGCLC from ESC, ChIP-qPCR showed that the H3K4me3 signal was otherwise increased at these BMP antagonist loci, potentially impairing BMP signaling (Fig. 7B). Furthermore, following OICR-9429 treatment, we observed a significant reduction in H3K4me3 levels at the regulatory regions of FSTL1, BMPER, and NBL1 compared to the non-treatment control group (Fig. 7B) and subsequent downregulation of these BMP antagonists (Fig. 7C). This confirms the crucial role of H3K4me3 depletion in promoting BMP signaling and inducing germline formation. Hence, elevated H3K4me3 at FSTL1, NBL1, and BMPER in pluripotent BCs or ESCs promotes the transcription of these BMP antagonists, thereby suppressing BMP signaling and hindering PGCLC formation. Conversely, inhibition of H3K4me3 leads to transcriptional silencing of BMP antagonists, restoring BMP pathway activity and enabling efficient PGCLC induction.

On the other hand, the somatic cell gene expression program should be suppressed during the progression of the germ cell lineage. Accordingly, we observed reduced H3K4me3 depositions in PGCs compared to BCs for certain known regulators of somatic tissue morphogenesis, such as the HOX family genes (Fig. 7D). However, the PGCLCs exhibited elevated H3K4me3 signals at the promoter regions of HOXB9 and HOXD13 during germline induction, potentially interfering with germline specification. Subsequently, the addition of OICR-9429 reduced H3K4me3 levels around these Hox genes to levels comparable to those in native PGCs (Fig. 7E), leading to significant downregulation of HOXB9 and HOXD13 gene expression (Fig. 7F). Thus, H3K4me3 depletion also inhibited the somatic gene program and contributed to germline specification.

In summary, depleting the H3K4me3 with histone methyltransferase inhibitor OICR-9429 could promote PGCLCs induction during ESCs differentiation in chicken. The downregulation of BMP signaling antagonist genes, coupled with the silencing of genes related to organism morphogenesis, enhanced the BMP-driven germline specification, and in turn promoted ESCs differentiation towards germ cells (Fig. 8).

Fig. 8.

The schematic diagram of the mechanism of H3K4me3-associated BMP signaling regulation during PGC formation in chicken

Discussion

Investigating the specification of chicken PGCs holds promise for gene editing, transgenesis and the conservation of endangered avian species. The induction of PGC-like cells from embryonic pluripotent stem cells has been successfully achieved in mammals [37], birds [35], and fishes [38] with various efficiencies. Subsequently, the induced PGCs (iPGCs) can be further differentiated into functional gametes after transplantation into the host animals. In chicken, the in vitro germline differentiation efficiency was only 10–20%, underscoring the need for a deeper understanding of the mechanisms governing PGC formation. Lacking of comprehensive exploration into the epigenetic regulation mechanisms limits our ability to effectively induce PGC formation in vitro. In this study, we performed a thorough analysis of different histone modifications and chromatin states in chicken stem cells, germ cells and somatic cells. Our goal was to elucidate the chromatin regulation mechanisms underlying PGC specification from the pluripotent stem cells and how is it separated from the somatic lineage differentiation. Our findings highlighted the important role of H3K4me3 chromatin modification in chicken PGCs derivation from pluripotent BCs or ESCs and confirmed that reducing H3K4me3 histone modification, in conjunction with BMP signaling factors, can improve the efficiency of ESC-PGC differentiation.

The comparative transcriptome analysis in chicken pluripotent BCs, germline PGCs and somatic CEFs provide novel insights into signaling pathways regulating stem cell function and germ/soma segregation, highlighting the enrichment of “cell polarity” and “cell adhesion” genes in pluripotent BCs, which suggests their important roles in avian pluripotency. Additionally, the early activation of meiotic genes in PGCs from 5.5d gonads suggests that meiotic priming in avian germ cells may occur earlier than previously thought. Then, we scrutinized the RNA-seq data from different cells and found several chromatin organization-related genes were highly enriched in PGCs, indicating chromatin reprogramming during the germ/soma lineage segregation from pluripotency. Notably, we found PGC expressed abundant KDM1B and KDM3A, which demethylate ‘Lys-4’ of histone H3, indicate the critical roles of H3K4me3 histone modification for PGC specification. Subsequently, we mainly focused on H3K4me3 marked chromatin regions and design strategies to manipulate H3K4me3 depositions during germ cell differentiation to improve its efficiency.

The methylation of the histone H3 in the residue K4 (H3K4) is a critical epigenetic marker for transcriptional-dependent functions with strongly activate gene expression [39]. While previous studies in chicken [6] and mammals [40] identified a role of H3K4me2 in PGC development, our current data suggest that the H3K4me3 also plays a critical function in chicken PGC formation. In mammals, the establishment of repressive chromatin states are critical for the proper suppression of multipotency and somatic genes during germline specification [41, 42]. Notably, distinct differences of histone modification landscape has been reported between chicken and mammals [7, 43–46]. For instance, we observed a significant reduction in H3K4me3 during chicken PGC specification, which is associated with the regulation of bivalent domains (co-localized H3K4me3 and H3K27me3). In contrast, human PGCs show no global changes in H3K4me3 or H3K27ac levels during specification [44]. We also identified a global decrease in H3K27me3 in chicken PGCs, consistent with prior immunostaining results [7, 43]. During the differentiation of chicken pluripotent cells into germ cells, H3K27me3 levels decline specifically at heterochromatic regions [7]. In contrast, murine PGCs show increased H3K27me3 deposition at euchromatic loci during specification (e.g., E8.5), indicating evolutionary divergence in Polycomb-mediated chromatin repression [40, 45, 47]. These substantial differences underscore the importance of characterizing the histone modification landscape during chicken PGC formation.

A recent study showed that high levels of histone acetylation modifications promote the formation of PGCLCs from cultured ESCs in chicken [48], highlighted the important roles of histone acetylation in germline induction. By performing ChIP-qPCR, we also found that H3K27ac deposition was elevated in some germline genes in both PGCs and induced PGCLCs (Figure S6), confirming the important roles of chromatin acetylation during germline development. Regarding other histone modifications, we found suppression of certain H3K4me3 marked chromatin regions, also contribute to the germline induction through multiple mechanisms. First, the higher H3K4me3 deposition at the transcription start site (TSS) in pluripotent BCs compared to PGCs and CEFs suggests that this modification may drive distinct gene expression programs across these cell types. Second, we observed that the reduction of H3K4me3 leads to the loss of the bivalent state in many chromatin regions. During early embryo development, late developmental regulator genes are typically not fully repressed but exist in a bivalent state [7, 49]. The decrease in H3K4me3 in the germline primarily disrupts this bivalent state, resulting in the permanent repression of somatic development-related genes, which facilitates germ cell lineage specification and maintenance. Furthermore, the germ cell induction experiments confirmed our hypothesis that the depletion of H3K4me3 deposition in certain chromatin regions may be a key step in inducing chicken PGCLCs formation from pluripotent BCs as well as cultured ESCs. In the endogenous native chicken PGCs, the active H3K4me3 levels decreased at specific chromatin regions of BMP inhibitors to release the antagonistic effect on BMP signaling to ensure efficient germ cell differentiation. However, during the in vitro artificial induction of PGCLC, the H3K4me3 deposition at these BMP inhibitor loci were conversely increased and resulted in unfavorable suppression of BMP signaling. Therefore, we utilized OICR-9429 to blocked the MLL/SET histone methyltransferase to diminish the H3K4me3 deposition in order to reduce the antagonistic effect on BMP signaling, which increased the germ line differentiation efficiency by 50%. At the same time, a group of somatic differentiation genes were also repressed by the diminished H3K4me3 deposition, further enhanced the germline maintenance.

In this study, our results validated the critical role of H3K4me3 in PGC specification; and we established an effective induction system for PGCLCs by integrating H3K4me3 downregulation in combination with the BMP4 factor. The use of epigenetic inhibitors in conjunction with key transcription factors or secreted factors may offer valuable insights into the induction of germline formation from pluripotent stem cells. Further investigations into the specific mechanisms and regulatory networks that govern the histone modifications will be essential for enhancing our understanding of germ cell fate determination and developmental processes.

Conclusion

In summary, we elucidate the critical role of chromatin histone modifications in pluripotent blastoderm cells, germline primordial germ cells, somatic fibroblast cells and the dynamic chromatin remodeling during chicken germline specification, especially for H3K4me3. We identified cis-regulatory elements and inferred transcription factors capable of specifying germ cell lineages. By strategically removing H3K4me3, we enhance the efficiency of producing PGC-like cells (PGCLCs) through decreasing in BMP signaling antagonists. Our comprehensive epigenomic analysis provides valuable insights into stem cell differentiation, paving the way for advancements in germ cell production and species conservation efforts.

Supplementary Information

Below is the link to the electronic supplementary material.

Authors’ contributions

SW and LM conducted most of the experiment and analysis; LM, XH, ZP, YH, YL, RR conducted bioinformatic analysis; HW and GZ designed the experiments and revised the manuscript. All authors read and approved the manuscript.

Funding

This work was supported by Key R&D Program of Shandong Province (2024LZGC018), Natural Science Foundation of Shandong Province (ZR2023MC123), NSFC (32102517), Scientific Research Innovation Team of Young Scholar of Shandong, Taishan Scholar program.

Data availability

RNA-seq and ChIP-seq data of this study are available in the NCBI SRA database PRJNA1089707 and PRJNA1089782.

Declarations

Ethics approval and consent to participate

The study under the experimental project “chicken germ cell development and function” was approved by the Animal Care and Use Committee of Shandong Agricultural University under permit number SDAUA-2022-91. Data of Approval: 2022-09-13.

Consent for publication

All authors have approved the manuscript for publication.

Competing interests

The authors declare no competing interests.