Pregranulosa cell–derived FGF23 protects oocytes from premature apoptosis during primordial follicle formation by inhibiting p38 MAPK in mice

A large number of oocytes in the perinatal ovary in rodents get lost for unknown reasons. The granulosa cell–oocyte mutual communication is pivotal for directing formation of the primordial follicle; however, little is known if paracrine factors participate in modulating programmed oocyte death perinatally. We report here that pregranulosa cell–derived fibroblast growth factor 23 (FGF23) functioned in preventing oocyte apoptosis in the perinatal mouse ovary. Our results showed that FGF23 was exclusively expressed in pregranulosa cells, while fibroblast growth factor receptors (FGFRs) were specifically expressed in the oocytes in perinatal ovaries. FGFR1 was one of the representative receptors in mediating FGF23 signaling during the formation of the primordial follicle. In cultured ovaries, the number of live oocytes declines significantly, accompanied by the activation of the p38 mitogen-activated protein kinase signaling pathway, under the condition of FGFR1 disruption by specific inhibitors of FGFR1 or silencing of Fgf23. As a result, oocyte apoptosis increased and eventually led to a decrease in the number of germ cells in perinatal ovaries following the treatments. In the perinatal mouse ovary, pregranulosa cell–derived FGF23 binds to FGFR1 and activates at least the p38 mitogen-activated protein kinase signaling pathway, thereby regulating the level of apoptosis during primordial follicle formation. This study reemphasizes the importance of granulosa cell–oocyte mutual communication in modulating primordial follicle formation and supporting oocyte survival under physiological conditions. A large number of oocytes in the perinatal ovary in rodents get lost for unknown reasons. The granulosa cell–oocyte mutual communication is pivotal for directing formation of the primordial follicle; however, little is known if paracrine factors participate in modulating programmed oocyte death perinatally. We report here that pregranulosa cell–derived fibroblast growth factor 23 (FGF23) functioned in preventing oocyte apoptosis in the perinatal mouse ovary. Our results showed that FGF23 was exclusively expressed in pregranulosa cells, while fibroblast growth factor receptors (FGFRs) were specifically expressed in the oocytes in perinatal ovaries. FGFR1 was one of the representative receptors in mediating FGF23 signaling during the formation of the primordial follicle. In cultured ovaries, the number of live oocytes declines significantly, accompanied by the activation of the p38 mitogen-activated protein kinase signaling pathway, under the condition of FGFR1 disruption by specific inhibitors of FGFR1 or silencing of Fgf23. As a result, oocyte apoptosis increased and eventually led to a decrease in the number of germ cells in perinatal ovaries following the treatments. In the perinatal mouse ovary, pregranulosa cell–derived FGF23 binds to FGFR1 and activates at least the p38 mitogen-activated protein kinase signaling pathway, thereby regulating the level of apoptosis during primordial follicle formation. This study reemphasizes the importance of granulosa cell–oocyte mutual communication in modulating primordial follicle formation and supporting oocyte survival under physiological conditions. In female mammals, primordial follicle is the basic structural and functional unit of reproduction, which provides an essential structural basis and microenvironment to maintain a dormant state of oocyte in ovary. The primordial follicle pool is considered to be nonrenewable reproductive resources of female mammals (1Zhang H. Li L. Liu，X. Shen Y. Busayavalasa K. Hovatta O. et al.Life-long in vivo cell-lineage tracing shows that no oogenesis originates from putative germline stem cells in adult mice.Proc. Natl. Acad. Sci. U. S.A. 2014; 111: 17983-17988Google Scholar, 2Kezele P. Nilsson E. Skinner M.K. Cell-cell interactions in primordial follicle assembly and development.Front Biosci. 2002; 7: d1990-d1996Google Scholar). Once the pool is established, the total number of follicles in the ovary reaches to peak (3Oktem O. Urman B. Understanding follicle growth in vivo.Hum. Reprod. 2010; 25: 2944-2954Google Scholar, 4Findlay J.K. Hutt K.J. Hickey M. Andersonet R.A. How is the number of primordial follicles in the ovarian reserve established?.Biol. Reprod. 2015; 93: 111Google Scholar). More than half of oocytes are lost during the formation of primordial follicles (5Wang C. Zhou B. Xia G. Mechanisms controlling germline cyst breakdown and primordial follicle formation.Cell. Mol. Life Sci. 2017; 74: 2547-2566Google Scholar, 6Alton M. Taketo T. Switch from BAX-dependent to BAX-independent germ cell loss during the development of fetal mouse ovaries.J. Cell Sci. 2007; 120: 417-424Google Scholar). The fate of oocytes in the fetal ovary seems to be determined by different protective strategies through the timely control of apoptosis or autophagy. According to present studies, the oocytes have developed mature protective mechanisms before meiosis prophase I has been successfully achieved, including the progressively controlled activity of glycogen synthase kinase-3 beta (GSK-3β), the levels of lysine-specific demethylase 1, and the balanced relationship between X-linked inhibitor of apoptosis and CASP9, to prevent premature oocyte loss (7Wen J. Yan H. He M. Zhang T. Mu X. Wang H. et al.GSK-3beta protects fetal oocytes from premature death via modulating TAp63 expression in mice.BMC Biol. 2019; 17: 23Google Scholar, 8He M. Zhang T. Zhu Z. Qin S. Wang H. Zhao L. et al.LSD1 contributes to programmed oocyte death by regulating the transcription of autophagy adaptor SQSTM1/p62.Aging Cell. 2020; 19e13102Google Scholar). Despite that, the endogenous and exogenous factors that cause the two typical programmed cell death (PCD) in oocytes remain elusive (6Alton M. Taketo T. Switch from BAX-dependent to BAX-independent germ cell loss during the development of fetal mouse ovaries.J. Cell Sci. 2007; 120: 417-424Google Scholar, 9Gawriluk T.R. Hale A.N. Flaws J.A. Dillon C.P. Green D.R. Rucker E.R. Autophagy is a cell survival program for female germ cells in the murine ovary.Reproduction. 2011; 141: 759-765Google Scholar). Hormones and growth factors are vital for blocking default cell death pathway (10Noda T. Suzuki K. Ohsumi Y. Yeast autophagosomes: de novo formation of a membrane structure.Trends Cell Biol. 2002; 12: 231-235Google Scholar). For instances, S100A8 is an oocyte-expressed chemotaxin which functions by directing ovarian somatic cell (OSC) migration during primordial follicle formation (11Teng Z. Wang C. Wang Y. Huang K. Xiang X. Niu W. et al.S100A8, an oocyte-specific chemokine, directs the migration of ovarian somatic cells during mouse primordial follicle assembly.J. Cell. Physiol. 2015; 230: 2998-3008Google Scholar, 12Cai H. Liu B. Wang H. Sun G. Feng L. Chen Z. et al.SP1 governs primordial folliculogenesis by regulating pregranulosa cell development in mice.J. Mol. Cell Biol. 2020; 12: 230-244Google Scholar). Oocyte-derived factor Jagged1 governs the primordial follicle formation by controlling the development of ovarian pregranulosa cells via ADAM10–Notch signaling (13Feng L. Wang Y. Cai H. Sun G. Niu W. Xin Q. et al.ADAM10-Notch signaling governs the recruitment of ovarian pregranulosa cells and controls folliculogenesis in mice.J. Cell Sci. 2016; 129: 2202-2212Google Scholar). Besides, Zp1, Zp2, Zp3, Figa, Cx43, and other genes related to oocyte/follicle function are parts of the determining factors for normal primordial follicle formation (14Lei L. Zhang H. Jin S. Wang F. Fu M. Wang H. et al.Stage-specific germ-somatic cell interaction directs the primordial folliculogenesis in mouse fetal ovaries.J. Cell. Physiol. 2006; 208: 640-647Google Scholar). Taken together, the recruitment of OSCs essential for establishment of the ovarian reserve depends on multiple paracrine factors derived from oocytes. Although one of our studies have uncovered the role of progesterone in modulating primordial follicle formation (15Guo M. Zhang C. Wang Y. Feng L. Wang Z. Niu W. et al.Progesterone receptor membrane component 1 mediates progesterone-induced suppression of oocyte meiotic prophase I and primordial folliculogenesis.Sci. Rep. 2016; 6: 36869Google Scholar), little is known if extracellular molecules, especially paracrine molecules, get involved in oocyte death during formation of the primordial follicles. The generic mitogen-activated protein kinases (MAPKs) have been proved to be pivotal for regulating cellular growth, proliferation, differentiation, migration, and apoptosis (16Sui X. Kong N. Ye L. Han W. Zhou J. Zhang Q. et al.p38 and JNK MAPK pathways control the balance of apoptosis and autophagy in response to chemotherapeutic agents.Cancer Lett. 2014; 344: 174-179Google Scholar, 17Sun Y. Liu W.Z. Liu T. Feng X. Yang N. Zhou H.F. Signaling pathway of MAPK/ERK in cell proliferation, differentiation, migration, senescence and apoptosis.J. Recept. Signal Transduct. Res. 2015; 35: 600-604Google Scholar). This signaling pathway consists of the extracellular signal–related kinases (ERK1/2), Jun amino-terminal kinases (JNK1/2/3), p38 MAPK, and ERK5. Abundant evidences indicated that p38 MAPK involves in apoptosis as well (18Xia Z. Dickens M. Raingeaud J. Davis R.J. Greenberg M.E. Opposing effects of ERK and JNK-p38 MAP kinases on apoptosis.Science. 1995; 270: 1326-1331Google Scholar, 19Juo P. Kuo C.J. Reynolds S.E. Konz R.F. Raingeaud J. Davis R.J. et al.Fas activation of the p38 mitogen-activated protein kinase signalling pathway requires ICE/CED-3 family proteases.Mol. Cell. Biol. 1997; 17: 24-35Google Scholar, 20Fernandes-Alnemri T. Armstrong R.C. Krebs J. Srinivasula S.M. Wang L. Bullrich F. et al.In vitro activation of CPP32 and Mch3 by Mch4, a novel human apoptotic cysteine protease containing two FADD-like domains.Proc. Natl. Acad. Sci. U. S. A. 1996; 93: 7464-7469Google Scholar, 21Cahill M.A. Peter M.E. Kischkel F.C. Chinnaiyan A.M. Dixit V.M. Krammer P.H. et al.CD95 (APO-1/Fas) induces activation of SAP kinases downstream of ICE-like proteases.Oncogene. 1996; 13: 2087-2096Google Scholar). Studies have approved that inhibition of caspases blocks p38 MAPK activation through Fas crosslinking (22Huang S. Jiang Y. Li Z. Nishida E. Mathias P. Lin S. et al.Apoptosis signaling pathway in T cells is composed of ICE/Ced-3 family proteases and MAP kinase kinase 6b.Immunity. 1997; 6: 739-749Google Scholar). Alternatively, overexpression of dominant active mitogen-activated protein kinase kinase 6 also induce caspase activity and cell death, implying that p38 may function both upstream and downstream of caspases in apoptosis (23Cardone M.H. Salvesen G.S. Widmann C. Johnson G. Frisch S.M. The regulation of anoikis: MEKK-1 activation requires cleavage by caspases.Cell. 1997; 90: 315-323Google Scholar). Therefore, the role of p38 MAPK in apoptosis could be both cell type and stimulus dependent (24Zarubin T. Han J. Activation and signaling of the p38 MAP kinase pathway.Cell Res. 2005; 15: 11-18Google Scholar). Whether p38 MAPK plays important roles in the oocytes during primordial follicle formation remains elusive. Fibroblast growth factor (FGF) plays significant roles not only in the oocyte survival but in granulosa cell differentiation in mammals (25Beenken A. Mohammadi M. The FGF family: biology, pathophysiology and therapy.Nat. Rev. Drug Discov. 2009; 8: 235-253Google Scholar, 26Presta M. Dell’era P. Mitola S. Moroni E. Ronca R. Rusnati M. Fibroblast growth factor/fibroblast growth factor receptor system in angiogenesis.Cytokine Growth Factor Rev. 2005; 16: 159-178Google Scholar, 27Ronca R. Giacomini A. Rusnati M. Presta M. The potential of fibroblast growth factor/fibroblast growth factor receptor signaling as a therapeutic target in tumor angiogenesis.Expert Opin. Ther. Targets. 2015; 19: 1361-1377Google Scholar, 28Turner N. Grose R. Fibroblast growth factor signalling: from development to cancer.Nat. Rev. Cancer. 2010; 10: 116-129Google Scholar). Of all the 22 members of FGFs, FGF23 was firstly found in patients with autosomal dominant hypophosphatemic rickets (29White K.E. Evans W.E. O’riordan J.L.H. Speer M.C. Econs M.J. Lorenz-Depiereux B. et al.Autosomal dominant hypophosphataemic rickets is associated with mutations in FGF23.Nat. Genet. 2000; 26: 345-348Google Scholar, 30Martin A. David V. Quarles L.D. Regulation and function of the FGF23/klotho endocrine pathways.Physiol. Rev. 2012; 92: 131-155Google Scholar). As a pleiotropic multiple functional hormone, FGF23 is involved in the phosphate reabsorption, the regulation of vitamin D production, and the conserving of calcium and sodium level in the kidney (31Silver J. Naveh-Many T. FGF23 and the parathyroid.Adv. Exp. Med. Biol. 2012; 728: 92-99Google Scholar, 32Chen G. Liu Y. Goetz R. Fu L. Jayaraman S. Hu M.C. et al.alpha-Klotho is a non-enzymatic molecular scaffold for FGF23 hormone signalling.Nature. 2018; 553: 461-466Google Scholar, 33Quarles L.D. Fibroblast growth factor 23 and alpha-Klotho co-dependent and independent functions.Curr. Opin. Nephrol. Hypertens. 2019; 28: 16-25Google Scholar, 34Haussler M.R. Livingston S. Sabir Z.L. Haussler C.A. Jurutka P.W. Vitamin D receptor mediates a myriad of biological Actions dependent on its 1,25-Dihydroxyvitamin D Ligand: distinct regulatory Themes revealed by Induction of klotho and fibroblast growth factor-23.JBMR Plus. 2021; 5e10432Google Scholar). There are four receptors of FGFs available (35Itoh N. Ornitz D.M. Functional evolutionary history of the mouse Fgf gene family.Dev. Dyn. 2008; 237: 18-27Google Scholar, 36Itoh N. The Fgf families in humans, mice, and Zebrafish: their evolutional processes and roles in development, metabolism, and disease.Biol. Pharm. Bull. 2007; 30: 1819-1825Google Scholar, 37Itoh N. ORNITZ D.M. Evolution of the Fgf and Fgfr gene families.Trends Genet. 2004; 20: 563-569Google Scholar). At physiological conditions, α-Klotho improves the binding affinity of FGF23 to fibroblast growth factor receptors (FGFRs) (38Razzaque M.S. The FGF23-klotho axis: endocrine regulation of phosphate homeostasis.Nat. Rev. Endocrinol. 2009; 5: 611-619Google Scholar, 39Ito S. Fujimori T. Hayashizaki Y. Nabeshima Y. Identification of a novel mouse membrane-bound family 1 glycosidase-like protein, which carries an atypical active site structure.Biochim. Biophys. Acta. 2002; 1576: 341-345Google Scholar, 40Kuro-o M. Matsumura Y. Aizawa H. Kawaguchi H. Suga T. Utsugi T. et al.Mutation of the mouse klotho gene leads to a syndrome resembling ageing.Nature. 1997; 390: 45-51Google Scholar). Interestingly, mice with depleted Fgf23 display significantly high serum phosphate with increased renal phosphate reabsorption, premature death, kidney and bone defects, and infertility (41Bowe A.E. Finnegan R. Jan D.B.S. Cho J. Levine M.A. Kumar R. et al.FGF-23 inhibits renal tubular phosphate transport and is a PHEX substrate.Biochem. Biophys. Res. Commun. 2001; 284: 977-981Google Scholar, 42Ben-Dov I.Z. Galitzer H. Lavi-Moshayoff V. Goetz R. Kuro-o M. Mohammadi M. The parathyroid is a target organ for FGF23 in rats.J. Clin. Invest. 2007; 117: 4003-4008Google Scholar). However, how does FGF23 correlates to fertility remains unknown so far. Here, we have shown that pregranulosa cell–derived FGF23, which binds to FGFR1, regulates the level of apoptosis by activating the p38 MAPK signaling pathway and contributes to determining the fate of oocyte during primordial follicle formation in mice ovary. In order to address the physiological significance of FGF23 during primordial follicle formation, we firstly detected the location and expression pattern of FGF23 in ovaries around the time of birth in mice. The results showed the expression level of FGF23 retains at a relative higher level from 18.5 days postcoitum (dpc) to 1 day postpartum (dpp, Fig. 1, A and B). The conclusion was reapproved by our quantitative reverse transcription-PCR (qRT-PCR) results (Fig. 1C). Immunofluorescence staining results showed that FGF23 was predominantly localized to the cytoplasm of (pre)-granulosa cells in mice ovaries (Fig. 1D). Further, to confirm the localization of FGF23 protein within the ovaries, we isolated mouse ovarian granulosa cells and oocytes from the ovaries that have developed to different stages. DEAD-box helicase 4 (DDX4) and Forkhead box L2 (FOXL2) are specifically expressed by oocytes and granulosa cells, respectively. The results showed that the expression level of FGF23 was much higher in the granulosa cells than in the oocytes (Figs. 1E and S1). To study the function of FGF23 in regulating primordial follicle formation, we established an in vitro culture model of perinatal mouse ovaries pretreated with or without foreign plasmid injection (Fig. S2). To knockdown the Fgf23 gene, shRNA-Fgf23 was cloned into pSicoR-GFP (43Zhang T. He M. Zhao L. Qin S. Zhu Z. Du X. et al.HDAC6 regulates primordial follicle activation through mTOR signaling pathway.Cell Death Dis. 2021; 12: 559Google Scholar). Based on this model, we constructed Fgf23-shRNA vectors (shRNA-Fgf23) and injected the shRNA-Fgf23 plasmids into 16.5 dpc ovaries and cultured for 3 days (equal to 1 dpp in vivo) (Fig. S3). We then applied the plasmid with the highest knockdown efficiency to perform the subsequent experiments. It showed that the number of available oocytes in the shRNA-Fgf23 group were significantly less than the control group after RNAi treatment for 6 days (equal to 4 dpp in vivo) (Fig. 2, A and B). Additionally, specific knocking down of Fgf23 raised the level of cellular apoptosis in the cultured tissues (Fig. 2C). Collectively, these results implied that silencing of Fgf23 accelerated the loss of primordial follicles in perinatal ovaries in vitro. To further verify if granulosa cell–derived FGF23 contributes to rescue oocyte death, we conducted a granulosa cell–specific knockdown of Fgf23 (Foxl2-shRNA-Fgf23). We injected the negative control vector (control) and the Foxl2-shRNA-Fgf23 vector into 16.5 dpc ovaries, respectively. Both the knockdown efficiency of Fgf23 and the primordial follicle formation of the cultured ovaries were examined, which showed that the FGF23 protein level was significantly decreased after 3 days of culture in the Foxl2-shRNA-Fgf23 group (Fig. 2D). The number of the oocytes was significantly less than the control in the 6-days culture group (Fig. 2G). Subsequently, the pure FGF23 protein was supplemented into the culture media either immediately or 3 days after injection of Foxl2-shRNA-Fgf23. The level of FGF23 protein in each group was detected after 6 days. The results showed that the level of the FGF23 protein increased in both culture strategies, namely supplementing FGF23 for either the first 3 days (0–3 days period) or for the last 3 days (3–6 days) (Fig. 2E). In addition, the loss of oocytes by RNAi was partially rescued by FGF23 after 3 days of culture, while the loss of oocytes could be effectively rescued by replenishing FGF23 immediately after RNAi (Fig. 2, F and G). These results further proved that FGF23 does play an important role in regulating primordial follicle formation in mice. Furthermore, to verify if additional FGF23 contributes to preserve more germ cells, we cultured 16.5 dpc ovaries with various concentrations of FGF23 protein for 6 days before histological analysis. The results showed that the number of total follicles within each ovary in different groups was comparable to the control (Fig. S4), which implies that excessive FGF23 may reach a saturation state with no significant effect on the number of primordia follicles. To investigate the function of FGFRs in primordial follicle formation, we detect the cellular localization and expressing dynamics of FGFRs in ovaries around the time of birth, respectively. The results showed that the expression level of FGFR1 and FGFR4 in the ovaries retain at a relative higher level from 18.5 dpc to 1 dpp, while the expression levels of FGFR2 and FGFR3 did not change significantly from 15.5 dpc to 9 dpp (Figs. 3A and S5A). The results were further approved by qRT-PCR analysis, which showed the expressions of the mRNA of Fgfr1 and Fgfr4 retain at relative higher level from 18.5 dpc to 3 dpp. On the contrary, the expression levels Fgfr2 and Fgfr3 decreased before birth but increased after birth (Fig. S5B). We next studied whether FGFRs were involved in the process of primordial follicle formation in ovaries. Based on the culture model, 17.5 dpc ovaries were cultured with FIIN-2, one of pan-FGFRs inhibitors, for 5 days (equal to 4 dpp in vivo). The statistical analysis of total ovaries showed that inhibition of FGFRs led to a decrease in the number of oocytes (Fig. 3, C and D). To detect the cause of the loss of oocytes, we cultured 17.5 dpc ovaries for 3 days. Both reverse transcription-PCR as well as Western blotting results showed that apoptosis levels were upregulated in the ovaries after in vitro inhibition of FGFRs, whereas, there was no significant change in autophagy levels (Fig. 3, E and F). In order to clarify if inhibition of FGFRs would affect oocyte meiosis progression in perinatal mice ovaries, 13.5 dpc ovaries were cultured with FIIN-2 for 4 days (equal to 17.5 dpc in vivo). The results showed that the ratio of colocation of Y box Protein 2 (MSY2) and DDX4 between the FIIN-2 group and the control was insignificant (Fig. S6, A and B), implying that inhibition of the FGFRs in fetal mice ovaries may not influence the progress of oocyte meiosis. In order to identify the most important FGFRs in the regulation of oocyte number during primordial follicle formation, we further specifically inhibited either FGFR1 or FGFR4 in vitro. Briefly, 16.5 dpc ovaries were cultured with or without PD173074, one of the specific inhibitors of FGFR1, for 3 days. The results showed that the level of the phosphorylated FGFR1 protein was decreased in PD173074 group as compared to the control (Fig. 4A). Moreover, after 6 days of culture by PD173074, much fewer oocytes were available in the experimental group than that in the control ovaries (equal to 4 dpp in vivo) (Fig. 4, B and C). Then, the dynamic change of oocytes number influenced by PD173074 was recorded after 16.5 dpc ovaries were cultured for different days. The results showed that remarkable influence of FGFR1 on oocytes number inhibition showed up on the fourth day of culture (Fig. 4D). However, there was no significant change in the number of oocytes if FGFR4 was specifically inhibited by H3B-6527 (Fig. S7). The results showed that FGFR1 is more important than FGFR4 in mediating FGF23 signaling in perinatal mouse ovaries. Then, we carried out a more detailed detection on the positioning of FGFR1. The results of granulosa cells and oocytes isolated from different developmental stages of mouse ovaries showed that the expression of FGFR1 in the oocytes was significantly higher than that in the granulosa cells (Figs. 3B and S5C), suggesting that FGF23 is secreted from granulosa cells and acts on oocytes as a paracrine factor to protect oocyte premature loss during primordial follicle formation. To further verify the function of FGFR1, we performed the in vitro reconstitution system. The ovaries were cultured for 4 days with PD173074 and then digested into single cells for reconstruction (Fig. S8A). The statistical analysis showed that inhibition of FGFR1 significantly influenced primordial follicle formation (Fig. S8, B–D). Then, we selected the burosumab (neutralizing antibody of FGF23), FNab10417 (neutralizing antibody of FGFR1), and 6F10 (neutralizing antibody of FGFR1) to further examine the role of FGF23-FGFR1 in primordial follicular formation. Briefly, neutralizing of FGFR1 leads to a decrease in the number of oocytes (Fig. S9, A and B). In order to further verify the function of FGF23 and FGFR1, we performed the in vitro reconstitution primordial follicle–like structure system. The results showed that the ovaries were cultured for 4 days with neutralizing antibodies and then digested into single cells for reconstruction (Fig. S9C). The results showed that neutralizing of FGFR1 or FGF23 blocked the primordial follicular–like structures’ formation efficiency, in which the majority of the cells were aggregated into reconstructed cell masses (Fig. S9, D–F). All these results preliminarily suggest that FGF23 and FGFR1 function during primordial follicle formation. To verify the effect of FGFR1 on oocyte meiosis progression in vitro, we used synaptonemal complex protein 3 (SYCP3) to mark the meiosis process in the oocytes and performed immunofluorescence costaining to further determine the effect of FGFR1 on oocyte meiosis progression (44Wang Y. Teng Z. Li G. Mu X. Wang Z. Feng L. et al.Cyclic AMP in oocytes controls meiotic prophase I and primordial folliculogenesis in the perinatal mouse ovary.Development. 2015; 142: 343-351Google Scholar, 45Dai Y. Bo Y. Wang P. Xu X. Singh M. Jia L. et al.Asynchronous embryonic germ cell development leads to a heterogeneity of postnatal ovarian follicle activation and may influence the timing of puberty onset in mice.BMC Biol. 2022; 20: 109Google Scholar). By combining the staining of SYCP3 and DDX4 with a high-resolution confocal imaging system, it showed that the meiotic phases of oocytes were well-identified after the ovaries were cultured with PD173074 from 13.5 dpc for 4 days (equal to 17.5 dpc in vivo). Consistent with the previous results, most of oocytes have entered into pachytene and diplotene stages (Fig. S6C), and a small portion of germ cells were already arrested at dictyate stage (Fig. S6C). Statistical analysis further confirmed that activation of FGFR1 is not essential for oocyte meiosis progression (Fig. S6D). To further ascertain whether apoptosis or autophagy was involved in oocytes death correlated to FGF23 signals, we firstly examined the changes of LC3B in our in vitro culture model. The level of autophagy in the cultured ovaries treated by FIIN-2 was unaffected, while the apoptosis level indicated by the phosphorylated caspase 3 was significantly elevated, as compared to the control (Fig. 3, E and F). The results were in line with the RNAi assay (Fig. 2A). Therefore, apoptosis, instead of autophagy, is more responsible for explaining why inhibition of FGF23-FGFRs induced significant oocyte loss in vitro. To explore the reasons for the massive loss of oocytes after inhibiting FGFR1, we performed the following assays. Firstly, the level of active caspase 3, which was located in the oocytes, was obviously higher in PD173074-treated group than in the control (Fig. 5A). Secondly, the number of TUNEL-positive signals was increased after inhibiting FGFR1 by PD173074 (Fig. 5, B and C). Thirdly, the level of the phosphorylated histone H2AX at Ser139 (referred to as γ-H2AX), which marks DNA double-strand breaks (DSBs), was used to test the completeness of the repair of DNA damage. The results showed that the percentage of γ-H2AX signaling increased significantly when 16.5 dpc ovaries were cultured with PD173074 for 6 days. Fourthly, the number of apoptotic cells, as were indicated by both the Annexin Ⅴ– and propidium iodide–positive cells, increased time dependently by the PD173074 treatment in vitro (Fig. S10A). Further, to test the dynamic changes in the number of positive signals after being treated with PD173074, 16.5 dpc ovaries were cultured for different days. The statistical analysis showed that the signal of positive γ-H2AX became significant since the third day till the sixth day, as compared with the control (Fig. 5, D and E). When we transfected Fgf23-shRNA vectors (shRNA-Fgf23) into either KGN cells or KK1 cells and cultured the cells for 2 days, the cell viability in both the assays was decreased as well (Fig. S10, B–D). Together with the findings in Figure 5, it is assumed that FGF23 in granulosa cells is most possibly involved in the regulation of granulosa cells apoptosis. Subsequently, to determine whether apoptosis was the main cause for oocyte loss after FGF23 was blocked, Z-VAD-FMK, a pan-caspase inhibitor, was added to the culture media of 16.5 dpc ovaries with PD173074 and cocultured for 6 days. The results showed that the number of oocytes was not significantly changed (Fig. S10, E and F). Therefore, inhibition of FGFR1 eventually leaded to apoptosis of germ cells, but apoptosis was not the cause of mass loss of germ cells. To explore the intracellular mechanism of the FGF23-FGFR1 signaling in regulating oocyte apoptosis, 16.5 dpc ovaries were cultured with or without PD173074 for 3 days in vitro. The preliminary analysis of RNA-seq data indicated significant changes in cellular adhesion and MAPK signaling pathways induced by PD173074 (Fig. 6, A and B). In line with these findings, most of the mRNA levels of the MAPK signaling pathway–related molecules were significantly upregulated after the 16.5 dpc ovaries were cultured for 3 days (Fig. 6C). Furthermore, phosphorylated p38 MAPK (p-p38 MAPK) levels were significantly upregulated, as compared to the control (Fig. 6D). However, Western blotting analyses revealed that after inhibition of FGFR1, other proteins, such as p62, AKT/p-AKT, and mTOR/p-mTOR, did not change significantly (Fig. S11). To verify whether p38 MAPK signaling pathway was the key to germ cells depletion, pamapimod, one of the p38 MAPK specific inhibitors, was added to the culture media of 16.5 dpc ovaries. The results indicated that the p-p38 MAPK level was significantly decreased by either 1 μM or 5 μM of pamapimod (Figs. S12 and 6D). When 16.5 dpc ovaries were cultured for 6 days, more oocytes were observed in PD173074 plus pamapimod treatment than those in the PD173074 treatment alone (Fig. 6, E and F). Therefore, specific inhibition of p-p38 MAPK reversed the mass loss of oocytes induced by PD173074. In most mammals, programmed loss of oocytes occurs briefly around the time of birth, but the reasons and the mechanisms remain uncertain (5Wang C. Zhou B. Xia G. Mechanisms controlling germline cyst breakdown and primordial follicle formation.Cell. Mol. Life Sci. 2017; 74: 2547-2566Google Scholar, 46Byun S. Kim Y.C. Zhang Y. Kong B. Guo G. Sadoshima J. et al.A postprandial FGF19-SHP-LSD1 regulatory axis mediates epigenetic repression of hepatic autophagy.EMBO J. 2017; 36: 1755-1769Google Scholar). Cellular apoptosis is the first recognized process responsible for oocyte PCD (6Alton M. Taketo T. Switch from BAX-dependent to BAX-independent germ cell loss during the development of fetal mouse ovaries.J. Cell Sci. 2007; 120: 417-424Google Scholar, 47Felici A. Giorgio M. Krauzewicz N. Della R.C. Santoro M. Rovere P. et al.Medullary thyroid carcinomas in transgenic mice expressing a Polyoma carboxyl-terminal truncated middle-T and wild type small-T antigens.Oncogene. 1999; 18: 2387-2395Google Scholar, 48Coucouvanis E.C. Jones P.P. Changes in protooncogene expression correlated with general and sex-specific differentiation in murine primordial germ cells.Mech. Dev. 1993; 42: 49-58Google Scholar). This is not only because low-level apoptosis occurs before birth but because increase or decrease in follicle numbers occurred in postnatal mice bearing systemic deletions of proapoptotic or antiapoptotic genes (49Matikainen T.M. Moriyama T. Morita Y. Perez G.I. Korsmeyer S.J. Sherr D.H. et al.Ligand activation of the aromatic hydrocarbon receptor transcription factor drives Bax-dependent apoptosis in developing fetal ovarian germ cells.Endocrinology. 2002; 143: 615-620Google Scholar). Most recently, we have proved that GSK-3β is essential for sustaining fetal oocyte survival by fine-tuning the cytoplasmic-nuclear translocation of β-catenin, which in turn modulates TAp63 expression timely during meiotic prophase I mice promptly (7Wen J. Yan H. He M. Zhang T. Mu X. Wang H. et al.GSK-3beta protects fetal oocytes from premature death via modulating TAp63 expression in mice.BMC Biol. 2019; 17: 23Google Scholar). However, it is unknown what kinds of extracellular molecules participate in oocyte apoptosis during primordial follicle formation. Here, we have proved that FGF23 as a pregranulosa cell–derived paracrine factor prevents premature oocyte apoptosis and protecting oocytes from massive loss at physiological condition. During primordial follicle formation, the granulosa cells gradually envelop the dominant oocyte. During this process, many paracrine or endocrine factors are involved to ensure proper assembly of the primordial follicle. For example, maternal progesterone levels in midpregnancy inhibit the expression of Jagged2 and Notch1 in mouse fetal mouse, which markedly inhibited nest breakdown and follicle formation (50Guo M. Zhang H. Bian F. Li G. Mu X. Wen J. et al.P4 down-regulates Jagged2 and Notch1 expression during primordial folliculogenesis.Front. Biosci. (Elite Ed.). 2012; 4: 2631-2644Google Scholar). In addition, the interaction of granulosa cell–derived KitL with oocyte- and membrane-derived Kit has an important role in the migration and proliferation of primordial germ cells and follicle development. All these studies showed that the close communication between granulosa cells and oocytes is important for primordial follicle formation (51Merkwitz C. Lochhead P. Tsikolia N. Koch D. Sygnecka K. Sakurai M. et al.Expression of KIT in the ovary, and the role of somatic precursor cells.Prog. Histochem. Cytochem. 2011; 46: 131-184Google Scholar, 52Choi Y. Rajkovic A. Genetics of early mammalian folliculogenesis.Cell. Mol. Life Sci. 2006; 63: 579-590Google Scholar, 53Buratini J. Price C.A. Follicular somatic cell factors and follicle development.Reprod. Fertil. Dev. 2011; 23: 32-39Google Scholar). The results from this study emphasized the importance in the role of granulosa cells–drove paracrine factor FGF23 in protecting oocytes from premature loss in primordial follicle formation in mice. In vitro, impaired function of FGF23 before primordial follicle formation leads to activated MAPK signal pathway in mice oocytes, which induces significant oocyte apoptosis in ovaries, implying that FGF23 is pivotal for supporting DSB repair completion. Although our results suggest that the mutual spatiotemporal expression patterns of FGF23 and FGFR1 might be closely related to the PCD of oocytes in the mouse ovary around the time of birth, the underlying mechanism that coordinates the substantially programmed meiotic DSB existence and the DNA damage checkpoint in fetal oocytes is still elusive. This study has proved that at least one of the FGFRs participates in mediating the function of FGF23 during primordial follicle formation in vitro. Previous studies have shown that intact FGF23 signaling activates FRS2/RAS/RAF/MEK/ERK1/2 (54Ornitz D.M. Itoh N. The fibroblast growth factor signaling pathway.Wiley Interdiscip. Rev. Dev. Biol. 2015; 4: 215-266Google Scholar, 55Sen A. Yokokura T. Kankel M.W. Dimlich D.N. Manent J. Sanyal S. et al.Modeling spinal muscular atrophy in Drosophila links Smn to FGF signaling.J. Cell Biol. 2011; 192: 481-495Google Scholar). The results of our study have shown that the mRNA levels of FRS2/RAS/ERK1/2 were changed systematically (Fig. 6C), but the level of the protein expression and the activity (phosphorylation) of ERK1/2 did not change significantly. In contrast, in mouse ovaries, FGF23 activates P38 through activating, at least FGFR1, and plays a pivotal role in supporting primordial follicle formation. In addition, we noticed no significant changes in p62 protein level after knockdown of Fgf23, no changes in either lysine-specific demethylase or p62 protein levels, and no changes in relative mRNA expression level of GSK-3β after FGFRs inhibition (Fig. S13). This also provides preliminary evidence that the FGF23–FGFR1 pathway is different from the pathway previously studied, which also provides a new mechanism for regulating oocyte number during primordial follicle formation. Whether there are other signaling pathways that are pivotal for conducting FGF23 remains unclear. Despite these, plenty of works are needed to provide substantial data that are helpful to give precise explanations upon the FGF23-involved oocyte fate determination in mice. For instances, although this study highlights that FGF23 mainly binds to FGFR1 to protect oocytes from mass loss, the specific roles of FGFR2, FGFR3, and FGFR4 have not been excluded (56Takashi Y. Sawatsubashi S. Endo I. Ohnishi Y. Abe M. Matsuhisa M. et al.Skeletal FGFR1 signaling is necessary for regulation of serum phosphate level by FGF23 and normal life span.Biochem. Biophys. Rep. 2021; 27: 101107Google Scholar). Due to the high homology of FGFR1-FGFR4, the knockdown effect of single receptor constructed in this study needs optimization, and more detailed experiments, including in vivo assays, are needed for further verification. Moreover, the function of the FGF23 in mice ovaries is obviously different from the conventionally reported so far, it is therefore necessary to further explore if the metabolism of the mineral particles, such as Na+, Ca2+ and phosphorus, are affected by the FGF23 in the ovaries (57Rodriguez M. FGF23: is it another biomarker for phosphate-calcium metabolism?.Adv. Ther. 2020; 37: 73-79Google Scholar, 58Imura A. Tsuji Y. Murata M. Maeda R. Kubota K. Iwano A. et al.alpha-Klotho as a regulator of calcium homeostasis(J).Science. 2007; 316: 1615-1618Google Scholar). Lastly, although the follicle developmental, stage-depended roles of FGF23 was not systematically studied here, the expression patterns of FGFRs suggested that FGF23 might be active in affecting the fate of the growing follicles as well (results not shown). Hence, the physiological roles of FGF23 in adult ovaries need further study through constructing conditionally modified animals in the future (59Białka-Kosiec A. Orszulak D. Gawlik A. Drosdzol Cop A. The relationship between the level of vitamin D, leptin and FGF23 in girls and young women with polycystic ovary syndrome.Front. Endocrinol. (Lausanne). 2022; 13: 1000261Google Scholar). We approved here that pregranulosa cell–secreted FGF23 prevents oocyte from premature apoptosis through activation of the p38 MAPK pathway downstream of FGFR1 in perinatal mice ovaries (Fig. 7). This study emphasizes not only the importance of the mutual communication between the germ cells and the OSCs during primordial follicle formation but is helpful to explain the primordial follicle size-determining mechanisms that are pivotal for understanding the physiological regulation of primordial follicle formation.