Single‐cell transcriptome analysis of male chicken germ cells reveals changes in signaling pathway‐related gene expression profiles during mitotic arrest

Mitotic arrest is necessary for the embryonic development of germ cells, and thus, it is important to understand the signaling pathways that regulate mitotic arrest. Here, we investigated the signaling pathway dynamics of male embryonic chicken germ cells during mitotic arrest by single‐cell transcriptome analysis using germ‐cell tracing models. We identified signaling pathways that change at the transcriptional level during chicken male germ‐cell development after sex determination. We found that several components of the BMP, Notch, and JAK–STAT signaling pathways were downregulated at the mitotic‐arrest stage and were reactivated 1 week after hatching when all germ cells are quiescent after entering mitotic arrest. In addition, the transcriptional levels of components of the MAPK, Hedgehog, and thyroid‐hormone signaling pathways were steadily upregulated after mitotic arrest. This suggests the cooperation of multiple signaling pathways during entry into mitotic arrest and subsequent quiescence of chicken male germ cells.

Mitotic arrest is necessary for the embryonic development of germ cells, and thus, it is important to understand the signaling pathways that regulate mitotic arrest. Here, we investigated the signaling pathway dynamics of male embryonic chicken germ cells during mitotic arrest by single-cell transcriptome analysis using germ-cell tracing models. We identified signaling pathways that change at the transcriptional level during chicken male germ-cell development after sex determination. We found that several components of the BMP, Notch, and JAK-STAT signaling pathways were downregulated at the mitotic-arrest stage and were reactivated 1 week after hatching when all germ cells are quiescent after entering mitotic arrest. In addition, the transcriptional levels of components of the MAPK, Hedgehog, and thyroid-hormone signaling pathways were steadily upregulated after mitotic arrest. This suggests the cooperation of multiple signaling pathways during entry into mitotic arrest and subsequent quiescence of chicken male germ cells.
Mitotic arrest is a unique essential event that occurs during the development of male germ cells from the embryonic stage into a spermatocyte pool [1][2][3]. In chickens, during the differentiation of primordial gem cells (PGCs) into sex-specific germ cells, female germ cells enter meiosis and maintain meiotic arrest in G2/M phase until hatching, whereas male germ cells asynchronously enter mitotic arrest on embryonic day 14 (E14) and then enter into the resting phase (mitotic quiescence state). At 10 weeks after hatching, they re-enter the cell cycle and begin to differentiate [3][4][5]. The mitotic arrest in germ cells is regulated by multiple signaling pathways in diverse species [6][7][8][9][10]. Previously signaling pathways involved in differentiating germ cells were analyzed at different stages of chicken embryo development by genome and transcriptome-wide analyses after their separation from somatic cells. This approach revealed that signaling pathways such as TGF-b, JAK-STAT, and Hedgehog are involved in chicken germ-cell mitotic arrest [11]. However, we have yet to understand the pathways that regulate male germ-cell-specific differentiation, including mitotic arrest, during in vivo development.
Single-cell RNA sequencing (scRNA-seq) has enabled the study of the signaling pathways responsible for germ-cell mitotic arrest in several species. In humans, fetal male germ cells enter mitotic arrest 9 weeks after fertilization, and TGF-b-pathwayrelated genes including those related to the BMP and NODAL pathways are highly expressed in mitotically arrested fetal germ cells. Moreover, modulation of the expression of some ligands AMH and NODAL may be important for germ-cell mitotic arrest [10]. In mice, Notch signaling is more active in PGCs transitioning to mitotic arrest, and blocking the expression of HES1 and NOTCH1 belonging to the Notch signaling prevents PGCs from properly entering mitotic arrest [9]. Hence, Notch signaling may initiate mitotic arrest in mouse male germ cells and may be essential for maintaining germ-cell identity. Also, in mice, Hippo signaling is transiently more active 2 days after birth when male germ cells exit mitotic arrest [12].
Recently, we developed germ-cell tracing chicken models expressing germ-cell-specific fluorescent markers and were able to perform scRNA-seq of germ cells at each stage of embryonic development using single-cell transcriptomic approach [13]. Using this model, we identified the mitotic-arrested prospermatogonia cluster at three time points (E12, E16, and hatch) and found an epigenetic reprogramming schedule of mitotic-arrested chicken prospermatogonia [5]. As a follow-up study, here we investigate signaling pathways that regulate male embryonic germ cells during mitotic arrest for the first time using a single-cell transcriptomic approach. We found dynamic changes in multiple signaling pathways during mitotic arrest of male embryonic chicken germ cells, which were also observed when the developmental stage was extended beyond hatching. It is proposed that multiple pathways whose activities change before and after mitotic arrest cooperatively regulate mitotic arrest and subsequent quiescent state of male germ cells in chickens.

Experimental animals and animal care
All experimental procedures and care of chickens were approved by the Institute of Laboratory Animal Resources, Seoul National University. All methods were performed in accordance with ARRIVE (Animal Research: Reporting of In Vivo Experiments) guidelines and approved by the Institutional Animal Care and Use Committee (IACUC, SNU-190401-1-3) of Seoul National University.

Single-cell RNA-seq data processing and analysis
Raw fastq files were processed using the CELLRANGER PIPE-LINE, version 3.1.0 (10x Genomics, Pleasanton, CA, USA). The fasta and GTF files for chicken genome (GRCg6a) were modified to include the DAZL::GFP insert sequence. The cDNA sequences were mapped to the modified-chicken genome by using STAR (version 2.5.1b) aligner (GitHub, Inc., San Francisco, CA, USA) with the GRCg6a.99 GTF file. After aggregating gene-by-cell count matrixes of E12, E16, and hatch to remove cell-specific biases, cells were clustered by using the quickCluster function of the SCRAN (version 1.16.0) R package (Bioconductor, Boston, MA, USA) [14]. Differentially expressed genes (DEGs) among five clusters (clusters 1, 2, 3, 4, and 5) were identified by using the Find-Markers function of the SEURAT R package (FDR < 0.05). Gene Ontology (GO) terms enrichment analysis in DEGs (FDR < 0.05) of cluster 1 was performed using PANTHER [15], and 'GO biological process complete' was the annotation dataset. Among these, terms with fewer than three included genes or terms with associated P-values ≥ 1 were excluded. Count matrixes of E2.5 to 1 week after hatch were further aggregated and re-normalized with SCRAN R package. Detailed information is shown in our previous studies [5].

Immunohistochemistry
Immunohistochemistry was performed to identify the expression of GFP and DAZL in germ cells of DAZL:: GFP chick testis. Testes of chicks at hatching day were paraffin-embedded and sectioned (thickness, 9-10 lm). After deparaffinization, sections were washed three times with PBS and blocked with a blocking buffer (5% goat serum and 1% bovine serum albumin in PBS) for 1 h at room temperature. Sections were then incubated at 4°C overnight with primary antibodies, rabbit anti-GFP (Invitrogen, Thermo Fisher Scientific Inc., Carlsbad, CA, USA), or rabbit anti-DAZL [16]. After washing three times with PBS, sections were incubated with fluorescence-conjugated secondary antibodies (Alexa Fluor 488 or 568) for 1 h at room temperature. After washing three times, sections were mounted with ProLong Gold antifade reagent with DAPI (Vector Laboratories, Burlingame, CA, USA) and visualized on a fluorescence microscope.

RT-PCR and quantitative RT-PCR (qRT-PCR)
Total RNA samples from GFP + or GFP À cells at E12, E16, and hatch were extracted using the ReliaPrep RNA Miniprep Systems (Promega, Madison, WI, USA). Total RNA samples were then reverse transcribed into cDNAs using the SuperScript III Reverse Transcription Kit (Invitrogen). The cDNAs were amplified by PCR using specific primer sets (30 cycles at 95°C for 30 s, 60°C for 30 s, and 72°C for 30 s). Quantitative RT-PCR (qRT-PCR) was performed to measure the expression of genes involved in the identified pathways in germ cells and somatic gonadal cells, respectively. qRT-PCR was performed in triplicate using a StepOnePlus RT-PCR system (Applied Biosystems, Foster City, CA, USA). These reactions followed thermocycling conditions: 5 min at 95°C; 40 cycles of 30 s at 95°C, 30 s at 60°C, and 30 s at 72°C; and at the melting temperatures. Relative gene expression was calculated using the following formula: DCt = Ct of the target gene -Ct of GAPDH. Primer sequences are listed in Table S1.

Statistical analysis
Statistical analysis was performed using GRAPHPAD PRISM 9 (GraphPad Software, San Diego, CA, USA). Significant differences between groups were determined by one-way ANOVA with Tukey's multiple comparisons. Statistical significance was ranked as *P < 0.05 and **P < 0.01.

Identification of mitotic-arrested prospermatogonia and dynamic signaling pathways after initiation of mitotic arrest
We first validated the expression of GFP and the normal expression of DAZL in the germ cells of DAZL:: GFP chicks (Fig. 1A,B). To identify mitotic germ cells and mitotic-arrested prospermatogonia, we used scRNA-seq data of germ cells from male gonads at E12, E16, and hatch chicks [13]. We visualized results in the space calculated by using uniform manifold approximation and projection (UMAP) (Fig. 1C). We identified 5 clusters of cells according to gene expression profiles (Fig. 1D). All clusters robustly expressed a germ-cell marker DAZL (Fig. 1E) [5]. Cells in clusters 3 and 5 robustly expressed a proliferation-marker gene MKI67 during mitotic entry at both E16 and hatch, indicating these cells are mitotic germ cells. Cells in cluster 1 corresponding to Fig. 1E expressed lower levels of MKI67 and WEE1 than did cells in other clusters; these genes are minimally expressed in mitotically arrested PGCs [9]. However, cluster 1 cells robustly expressed NFIC, a co-marker of mitotic-arrested germ cells in both humans and mice [9]. Therefore, we considered cluster 1 to represent nonproliferating mitoticarrested prospermatogonia (Fig. 1E). To confirm transcriptional changes in signaling pathway-related genes, we identified differentially expressed genes (DEGs) whose levels in cluster 1 were higher or lower than those in other clusters (Table S2). Among cluster 1 DEGs, we found that downregulated genes had higher absolute log 2 FC values than upregulated genes (Tables S3 and S4). Expression of HES5, ID1, HES4, JAK1, and DLL4 was drastically downregulated, and the expression of PDGFD, ATP2A2, RXRA, GRK3, and TGFB2 was upregulated in cluster 1 (Fig. 1F,G, respectively). Next, we performed enriched gene ontology (GO) terms analysis by using the PANTHER database. Among cluster 1-enriched GO terms, we extracted only signaling pathway-related terms to identify changes in signal transduction involved in the transition of mitotic germ cells into mitotic-arrested prospermatogonia ( Fig. 1H and Table S5). We confirmed that many signaling pathways such as cell-cycle checkpoint, smoothened (also known as Hedgehog), thyroidhormone mediated, TGF-b, FGF, JAK-STAT, BMP, and MAPK were altered in mitotic-arrested prospermatogonia. We therefore hypothesized that multiple signaling pathways might be dynamically involved in the mitotic arrest in male chicken germ cells.
Expression changes in signaling pathway-related genes after mitotic arrest in chicken male germ cells We screened the expression of genes encoding receptors, ligands, effectors, activators, targets, and inhibitors of canonical signaling pathways. We found that genes related to BMP, Notch, and JAK-STAT signaling were downregulated after mitotic arrest ( Fig. 2A). Genes related to the BMP-signaling pathway including receptor BMPR1, ligand BMP2 and targets ID1 and ID4 were downregulated from E12 to hatch. On the other hand, inhibitors NOG, NBL1, and SMAD6 were upregulated after initiation of mitotic arrest (Fig. 2B).  In the Notch-signaling pathway, receptor NOTCH2, ligand DLL4, effectors ADAM17, RFNG, and MAML2, and targets HES4 and HES5 were downregulated from E12 to hatch. On the other hand, the genes coding for the inhibitors JAG2, DVL1, and DVL3, and co-repressors NCOR2 and CTBP1 were progressively upregulated (Fig. 2C). In the JAK-STAT signaling pathway, receptors JAK1 and IL10RB, ligand PXDN, effectors SPRY2, and target SOCS6 were downregulated after E12 (Fig. 2D).
We also found that gene sets for several signaling pathways were preferentially expressed after mitotic arrest (Fig. 2E). In the classic MAP-kinase pathway, receptor FGFR3, ligands FGF18, PDGFD, and EFNA5, effectors RRAS and RAF1, and target ETV1 were upregulated from E12 to hatch (Fig. 2F). In the p38 MAP-kinase pathway, receptor and ligand genes AMHR2 and TGFB2, respectively, were upregulated from E12 to hatch (Fig. 2F). In the Hedgehogsignaling pathway, activator genes (EVC2 and GRK3) and dual-function effector-and-target genes GLI2 and GLI3 were upregulated after mitotic arrest (Fig. 2G). Hedgehog-pathway activation requires both ligand expression and PTCH1-receptor repression [17]. However, three ligand genes (DHH, IHH, and SHH) were virtually unexpressed in GFP + germ cells (Fig. S1A). Hence, we measured the expression of one of these ligands DHH secreted from somatic gonadal cells. We selected the DHH, because it is expressed in mouse Sertoli cells [18] and regulates germ-cell differentiation [17]. By using RT-PCR, we confirmed that GFP À somatic gonadal cells at E12, E16, and hatch expressed DHH (Fig. S1B). This suggested that the DHH ligand secreted from somatic gonadal cells during mitotic arrest activates Hedgehog signaling in germ cells. We also found that thyroid-hormone signaling pathwayrelated genes coding for receptors THRA and RXRA, effectors PLCD1 and RAF1, and targets CREBBP and ATP2A2 progressively upregulated from E12 to hatch (Fig. 2H). All these genes demonstrated the highest expression in cluster 1, which corresponds to mitotically arrested cells. This suggests that their related pathways may be involved in the quiescent state of male germ cells.

Downregulated signaling pathways in mitoticarrested prospermatogonia appeared to be suppressed after transient activation
In order to investigate more detailed expression patterns of identified signaling pathway-related genes, we performed additional analysis on five time points: E8, E12 (before the onset of mitotic arrest), E16 (entering asynchronously into mitotic arrest), hatch (mostly mitotic arrest in G0/G1), and 1 week after hatch (maintaining mitotic arrest in G0/G1). This analysis confirmed the expression profiles identified by singlecell transcriptomic analysis for selected pathwayrelated genes. Among these genes, several genes were expressed at higher levels at E12 than at E8 and gradually downregulated from E12 to hatch. This expression pattern was found for the genes involved in the activation of the BMP-signaling pathway (such as BMPR1B, BMP4, ID1, ID2, ID3, and ID4; Fig. 3A), Notch-signaling pathway (DLL4, MAML2, HES4, and HES5; Fig. 3B), and JAK-STAT signaling pathway (IL10RB, PXDN, and SPRY2; Fig. 3C). We speculate that activation of these pathways may be required for mitotic-arrest initiation. The genes related to these pathways were next downregulated and upregulated again 1 week after hatch (Table S6). Among genes related to the BMP-signaling pathway, the expression was upregulated in the case of BMPR1B, BMP2, BMP4, ID1, ID2, and ID4, whereas that of gene encoding BMP antagonist Noggin (NOG) downregulated 1 week after hatch [19] (Fig. 3A). Among genes related to the Notch-signaling pathway, the expression of NOTCH2 and HES4 downregulated at E16 and upregulated again at 1 week after hatch, while expression of the co-repressors NCOR2 and CTBP1 downregulated at the same time (Fig. 3B). The expression pattern of genes related to the JAK-STAT pathway was similar: IL10RB, SPRY2, and SOCS6 expression downregulated at E16 and next upregulated again 1 week after hatch (Fig. 3C). Thus, the expression of genes related to BMP, NOTCH, and JAK-STAT pathways peaked at E12 (just before mitotic arrest), downregulated at E16, then upregulated again 1 week after hatch. We therefore suggest that transient changes in these signaling pathways may induce male germ cells to enter mitotic arrest.

Upregulated signaling pathways increase steadily during mitotic arrest
We examined the expression patterns of genes involved in MAPK, Hedgehog, and thyroid-hormone signaling. Expression levels of these genes were similar between E8 and E12 except for PDGFD. Expression of many genes upregulated after mitotic arrest and persisted after hatch. The expression of genes involved in MAPK pathway including FGFR3, FGF18, EFNA5, ETV1, RRAS, and TGFB2, in Hedgehog-signaling pathway including EVC, EVC2, GRK3, GLI2, and BCL2, and in thyroid-hormone signaling pathway including RXRA, CREBBP, and ATP2A2 upregulated steadily from E12  to 1 week after hatch ( Fig. 4A-C, respectively). PTCH1 and HHIP are both targets and antagonists of the Hedgehog-signaling pathway. PTCH1 levels downregulated at E12-E16, the period of mitotic-arrest transition, and upregulated significantly at 1 week after hatch, whereas HHIP gradually upregulated until 1 week after hatch (Fig. 4B). Considering these collective results, we propose that these upregulated signaling pathways may be involved in both mitotic arrest and subsequent quiescent state of male germ cells.

Interactions among germ cells and somatic gonadal cells regulate BMP-, Notch-, and MAPKsignaling pathways
Germ cells develop in interaction with surrounding somatic cells [20]. We thus investigated the role of somatic gonadal cells in the signaling pathway regulation of mitotic arrest. We selected BMP-, Notch-, and MAPK-signaling pathways as representative pathways, and we isolated GFP + cells (germ cells) and GFP À cells (somatic gonadal cells) from testes of germ-cell tracing chicken models [13] by FACS at E12, E16, and hatch. After using RT-PCR to confirm DAZL-expression levels in GFP + germ cells (Fig. 5A), we used qRT-PCR to measure the expression of genes involved in these pathways. We found that the expression of BMPR1B, a receptor of BMP signaling, downregulated in germ cells after E12. Expression of BMP4, a ligand of BMP signaling, downregulated in somatic gonadal cells from E12 to hatch. Expression of NOG, an antagonist of BMP signaling, upregulated in somatic gonadal cells from E12 to hatch (Fig. 5B). Expression of Notchpathway target genes HES4 and HES5 significantly downregulated in germ cells after E12, and expression of the inhibitor gene JAG2 upregulated in somatic gonadal cells from E12 to hatch (Fig. 5C). Expression of FGFR3, a receptor gene of MAPK pathway, upregulated in germ cells after hatch. Also, the expression of PDGFD, one of the ligand genes in MAPK pathway, upregulated in somatic gonadal cells after hatch (Fig. 5D). BMP2 and FGF18 ligand genes were also expressed in somatic cells, although the expression of these genes did not significantly differ among the three time points (Fig. 5B,D). These results indicate that changes in the expression of signal molecules related to BMP, Notch, and MAPK pathways in the somatic gonadal cells may be involved in the regulation of mitotic arrest in male germ cells.

Discussion
Male germ-cell differentiation is controlled by various signal transduction and transcription factors and strongly dependent on surrounding somatic gonadal cells after their sex differentiation [3,4]. Although various signaling pathways are involved in the development of chicken germ cells [11,21], studies on the mechanisms regulating mitotically arrested male germ cells are lacking [22]. In this study, we identified signaling pathways involved in mitotic arrest and subsequent quiescent state of male germ cells by single-cell transcriptome analysis of germ cells isolated from a germcell tracing chicken model. Stem cell differentiation requires the downregulation of genes essential for self-renewal. For example, the downregulation of Hes-family genes regulates proper neural stem-cell differentiation [23]. BMP-, NODAL-, and Notch-signaling pathways occur in mitotically arrested human fetal germ cells, and these signaling pathways are regulated by interaction with somatic cells [10]. In the mouse PGCs, the Notch-signaling pathway is activated in mitotically arrested cells, and NOTCH1 expression increases from E12.5 (just before the onset of mitotic arrest) and downregulated at E16.5. In these PGCs, the expression of the target HES1 gradually increases from E13.5 to E16.5. Blocking Notch signaling increases the number of MKI67-positive cells, suggesting that Notch signaling is essential for regulating mitotic arrest in male PGCs [9]. We found lower HES4 and HES5 expression levels in mitotic-arrested prospermatogonia than in mitotic cells. Thus, changes in Notch-signaling pathway in mitotic-arrested cells in chicken male embryonic germ cells differed from those in human and mouse. Many genes related to BMP-, Notch-, and JAK-STAT signaling pathways were suppressed after initiation of mitotic arrest. We propose that these three signaling pathways, which were transiently upregulated before mitotic arrest and subsequently downregulated, may be involved in mitotic arrest in male chicken germ cells.
We found MAPK, Hedgehog, and thyroid-hormone pathways were continuously upregulated in male germ cells during mitotic arrest. In mice, the transition from gonocyte to spermatogonia begins at E18.5. Moreover, FGF signaling prominently regulates pathways upstream of GDNF, retinoic acid, MEK/ERK, and PI3K/Akt signaling [1]. In addition, the main function of Fgf9 expressed in both somatic and germ cells in mice is to repress female-promoting genes [24], and p38-MAPK signaling also represses retinoic acid signaling [25]. Thus, pathways upregulated after mitotic arrest (including the MAPK pathway) may facilitate the early development of prospermatogonia into spermatogonia and inhibit the female-specific pathway. In mice and rats, ligand (such as Dhh) secreted from Sertoli cells activates Hedgehog signaling in Leydig cells, which regulates testis development and germ-cell survival [26,27]. Thus, Hedgehog-signaling activation in mitotic-arrested chicken prospermatogonia may be important for the survival and differentiation of germ cells and may be a central signal that regulates mitotic-arrest initiation because upregulated expression of PTCH1 and HHIP activates negative feedback. Thyroid hormone is involved in testis development and spermatogenesis in several vertebrates [28] and in the development of prospermatogonia in mice [12]. Thyroid hormones, especially T 4 activates ERK1/2 to initiate activities of TRb1, ERa, and STAT, and these hormones regulate transcriptional activities, cytokines, and growth factors [29]. We speculate that the thyroidhormone-related genes activated in mitotic-arrested prospermatogonia of chickens may be important for activating MAPK/ERK signals and promoting differentiation, but further research is required.

Conclusion
In this study, we identified changes in the gene expression profile of components of BMP-, Notch-, JAK-STAT, MAPK, Hedgehog, and thyroid-hormone signaling pathways before, during, and after mitoticarrest transition in male chicken germ cells. We propose that these pathways are involved in mitotic arrest of male germ cells during embryonic development and subsequent quiescent state after hatching. Signaling pathways that are upregulated before the initiation of mitotic arrest, and downregulated thereafter, may initiate mitotic arrest and male-specific development. Overall, our findings give insight into the regulation of signaling pathways involved in different steps of mitotic arrest of male germ cells in chickens.

Supporting information
Additional supporting information may be found online in the Supporting Information section at the end of the article. Fig. S1. Confirmation of ligand genes for Hedgehog signaling in DAZL::GFP-positive cells and DAZL:: GFP-negative cells during mitotic arrest. Table S1. List of primers used for research.  .  Table S6. List of genes whose expression was changed at 1 week after hatch compared with hatch in downregulated signaling pathways. One week after hatch/ hatch represents the mean value of 1 week after hatch divided by the mean value of hatch for each normalized gene expression. Table S2. List of DEGs in cluster 1 (Cluster 1 vs others).