Germ cells are fundamental to human reproduction. Follicle growth occurs in two stages. The initial slow phase lasts several weeks in small rodents and several months in large mammals, characterized by the proliferation of granulosa cells and an increase in the diameter of both the follicle and the oocyte. In the second stage, the follicles respond to follicle-stimulating hormone (FSH) and luteinizing hormone (LH), leading to follicular cavity formation, steroid hormone synthesis, and the pre-ovulatory phase [7].

In modern infertility treatments, improving the quality and quantity of oocytes is crucial. In particular, there is a need to develop new technologies to produce human gametes, improve oocyte quality and quantity, and integrate these methods into assisted reproduction [8].

Inducing stem cells to form primordial germ cells

Stem cell therapy is a cornerstone of regenerative medicine. Based on the mechanisms underlying oocyte development into gametes, research focuses on developing methods to induce the maturation of secondary oocytes from stem cells that are capable of fertilization. In mice, stem cell development occurs in two stages: spontaneous and induced differentiation. Initially, researchers obtained primordial germ cells (PGCs) by allowing stem cells to differentiate spontaneously. In 2003, Hübner et al. [9] discovered that mouse embryonic stem cells (mESCs) cultured in heated inactivated serum could develop into oocytes, undergoing meiosis and forming blastocyst-like structures. Toyooka and colleagues [10] later used the testis capsule to generate sperm from mESCs in a medium containing lymphocyte inhibitory factor (LIF), although they did not assess the sperm’s reproductive function. In 2004, Clark et al. [11] studied the spontaneous differentiation of human embryonic stem cells (hESCs) into embryoid bodies (EBs) that expressed germ cell-specific markers. However, the efficiency of these spontaneous differentiation methods for generating PGCs from stem cells was quite low, ranging from only 0.5–3.6% per EB.

To improve induction efficiency, research has increasingly focused on inducing PGCs from induced pluripotent stem cells (iPSCs). In 2008, Ying et al. [12] advocated for culturing pluripotent stem cells (PSCs) in a combination of MAPK, GSK, and LIF (known as 2i) to maintain germline-competent naive PSCs in what they referred to as the “ground state.” Building on this, several induction methods have been proposed. In 2009, Ohinata et al. [13] used BMP4, LIF, SCF, BMP8b, and EGF to induce germ cell characteristics in mouse epiblast cells, achieving an induction success rate of up to 45%. In 2011, Hayashi et al. [14] reported the successful derivation of PGCs from mESCs for the first time. This technique was used in conjunction with co-culturing gonadal cells (testicular mesothelial cells/ovarian somatic cells) to reconstruct mouse ovaries and testes, ultimately producing offspring. However, due to chromosomal abnormalities after meiosis, the offspring birth rate was low [15]. In 2014, Kimura et al. [16] found that inhibiting the ERK signaling pathway using MEK inhibitors mesodermal differentiation system while promoting germ cell differentiation in the mesodermal differentiation system of OP9 feeder cells. Although functional sperm could not be produced, this finding offered new insights for inducing iPSCs. Mouse PGC induction serves as a model to validate hypotheses on stem cell induction while adhering to ethical restrictions on human research, laying the foundation for human-induced PSCs (hiPSCs).

Regarding human germ cell development, Chuang et al. [17] demonstrated that hESCs could spontaneously differentiate into germ cell-like cells in a differentiation medium. Similar to methods used in animal stem cells, the synergistic action of BMP4 and WNT3A promoted the differentiation of hESCs into germ-like cells, achieving a differentiation rate of 31–47%. A significant advancement came when Irie et al. [18] established defined culture conditions to standardize the derivation of human primordial germ cell-like cells (hPGCLCs) from hESCs and hiPSCs. They maintained near-ground-state pluripotency by pre-culturing PSCs in a medium containing four inhibitors: CHIR99021 (a GSK-3 inhibitor), PD0325901 (a MEK inhibitor), SB203580 (a p38 MAPK inhibitor), and SP600125 (a JNK inhibitor). To induce hPGCLCs, they cultured the PSCs in this medium for two days with bFGF/TGFβ, then transitioned to a suspension culture containing BMP2 (or BMP4), LIF, SCF, and EGF. The derived hPGCLCs exhibited gene expression and epigenetic patterns similar to human PGCs (hPGCs). Subsequent studies introduced various induction processes with different factors. However, due to ethical constraints regarding human subjects, the reproductive function of these induced cells could not be verified.

In 2015, Sasaki et al. [19] applied hiPSCs, initially using the GSK3 inhibitor (CHIR99021) and Activin A to derive induced mesoderm-like cells (iMeLCs). These cells were then converted into BVSC-positive hPGCLCs using Irie’s method. In 2021, researchers from National Taiwan University reported the induction of early germ cells from hESC and hiPSCs using RA, BMP4, and Activin A. When these interacted with human granulosa cells and were cultivated in vivo in mice, they developed into follicle-like cells [8] ( Fig. 1).

Human gamete development is a complex process involving numerous transcription factors that regulate differentiation. Various combinations of transcription factors are currently used to induce human stem cells into PGCs, with induction efficiencies ranging from 27–47%—below the ideal level. Future research should focus on the epigenetics and developmental biology of human gamete differentiation, identifying key transcription factors for stem cell differentiation into gametes and optimizing their combinations to improve induction efficiency. Due to the uniqueness of the human population, reproductive function validation has not yet been achieved. However, with the advancement of genomic research, future studies may be able to verify genetic stability through omics approaches, paving the way for potential clinical applications.

Fig. 1

Process of inducing iPSCs into PGCLCs (A. Mouse induction process; B. Human induction process)

Primordial germ cells differentiating into gametes

To generate mature follicles, ongoing research focuses on establishing in vitro conditions for inducing PGCs into more advanced germ cells. Initially, researchers co-cultured PGCs with ovarian somatic cells or testicular supporting cells to achieve oocyte maturation and meiosis. For example, in 2016, researchers reported the use of Transwell-COL culture to co-culture mESCs and female gonadal somatic cells to form mouse ovaries, which then induced functional gamete production [20]. In 2018, Yamashiro et al. [21] co-cultured in vitro induced hPSCs with mouse ovarian somatic cells for 120 days. This process resulted in the gradual development of oocyte-like cells entering prophase of meiosis, although their survival rate was less than 10%. In 2022, Yang et al. [22] used fetal ovarian somatic cells from miscarried fetuses (7–8 weeks old) and hPGCs/oogonia derived from hiPSCs to create allogeneic recombinant ovaries (IrOvaries). These ovaries were reconstructed under ovarian capsules in severe combined immune deficiency( SCID) mice, generating haploid human oocytes in vivo, with a 3.2% transformation rate from mitosis to meiosis. However, the low induction efficiency, long induction times, and ethical concerns regarding the use of allogeneic fetal ovaries, along with the complexity and high costs of the technique, severely limit their potential for clinical translation. Therefore, efficient and simplified induction methods are urgently needed.

Researchers aim to identify specific induction and transcription factors that drive early germ cells into meiosis. Some studies have highlighted factors promoting germ cell differentiation into meiosis. In 2011, Eguizabal et al. [23] used retinoic acid (RA) to pre-induce primordial germ cell-like cells (PGCLCs) from human iPSCs. Subsequently, they applied human LIF (hLIF), forskolin (FRSK), bFGF, and a CYP26 inhibitor (R115866) to induce meiosis, producing haploid-like cells with an efficiency of 0.4–2.3%. This demonstrated that transcription factors could induce meiosis, although the resulting cells resembled sperm-like structures. In 2012, researchers found that overexpressing VASA and/or DAZL in hESCs and iPSCs increased meiosis-related marker expression, promoting meiotic progression in vitro [24]. Chuang et al. [17] discovered that BMP4 and Wnt3A could further induce hESC differentiation into germ cells reaching the pachytene stage of meiosis, which were successfully implanted into kidney capsules. However, their reproductive functionality was not validated.

In 2017, Jung’s team [25] used DAZL and BOULE to induce hESCs, regulating the exit from pluripotency and entry into meiosis. They further applied recombinant human GDF9 and BMP15, which induced these meiotic germ cells to form follicle-like cells (FLCs). In 2020, Abdyyev et al. [26] used BMP4 and recombinant human Activin A (rhActA) to culture hiPSCs and hESCs into late-stage germ cells, followed by RA induction. Although RNA-seq analysis revealed changes in meiosis-related genes, haploid cells were not successfully generated. These results indicate that current in vitro conditions are insufficient to complete meiosis.

Identifying optimal conditions for this process is crucial for ovarian organoid development, as producing mature gametes remains a key goal. Disorders in oocyte development are among the primary causes of infertility. In vitro gametogenesis can enable genetic screening, treat infertility, and improve reproductive health. Future research should focus on elucidating germ cell differentiation mechanisms both in vivo and in vitro, identifying factors beyond RA that promote haploid formation, and ensuring genetic and epigenetic stability. These challenges remain central to future research in this field.

Regulatory network of hPGC development

Understanding the signaling pathways and transcription factor networks governing human germ cell specification is crucial for efficient female gamete production. Identifying and inducing key upstream factors can activate germ cell development genes, leading to the efficient production of the desired female gametes. Current studies suggest that GATA2 or GATA3, SOX17, and TFAP2C are critical transcription factors in the process of inducing hPGCs from hiPSCs. These factors suppress somatic differentiation while promoting epigenetic reprogramming and germ cell development (Fig. 2) [27, 28].

In the induction of female gametes from hPGCs, transcription factor TAF4b is vital for meiosis I and the establishment of a healthy primordial follicle reserve by regulating key genes, including Stra8, Figla, Nobox, and Dazl [29]. The BMP-ZGLP1 axis and RA-STRA8 axis guide female germ cells into meiosis, while the STRA8-RB axis facilitates S-phase progression during meiosis, significantly contributing to female germ cell differentiation (Fig. 2) [30]. However, the expression, regulation, and interactions of downstream genes involved in meiosis remain poorly understood. Further exploration of these processes is crucial not only for understanding fetal oocyte development but also for gaining insights into oocyte aging, both of which are key to achieving successful in vitro differentiation.

Fig. 2

Signaling and transcription factor network for human germ cell development (divided into two regulatory processes involving signals and transcription factors: induction of hiPSCs into hPGCs and completion of meiosis by hPGCs. Arrows indicate stimulation or upregulation, while lines indicate inhibition.) (Created with BioRender.com)

Future research should prioritize constructing a more comprehensive signaling network for hPGC development, focusing on the intricate interactions between pathways and their specific contributions to germ cell fate determination. This approach will help identify precise signaling factors, ultimately improving the efficiency and accuracy of in vitro gametogenesis. These advancements will enable infertility treatments and support the development of reproductive medicine and regenerative biology.

Animal models for gamete development

Most research on in vitro gametogenesis (IVG) relies on mouse models. Human germ cells and mouse PGCs exhibit similar expression profiles for many key genes, including OCT4, NANOG, BLIMP1, TFAP2C, LIN28, SSEA1, cKIT, NANOS3, DAZL, and VASA [31]. However, gametogenesis and developmental mechanisms differ between species. For example, the key regulator Sox17 in hPGCs is expressed only transiently in mouse PGCs (mPGCs), whereas SOX15 expression is closely associated with hPGCs. Another difference lies in the expression of Sox2, which is downregulated in hPGCs but restored in mPGCs [32]. The activity of Blimp1 also varies: It suppresses mesodermal programs in mPGCs but inhibits neuronal development programs in hPGCs.

Despite these differences, mouse models remain valuable for genetic studies and provide significant reference value for researching human gametogenesis. Other animal models, such as rats, pigs, cows, and goats, can also be used to induce PGCLCs from iPSCs [32, 33]. Among these, pigs have shown particularly promising results. Porcine iPSCs, first induced to EpiLCs with Activin A and bFGF and then differentiated into PGCLCs, are treated with BMP4, BMP8B, SCF, hLIF, and EGF, similar to the mouse model for PGCLC induction. Notably, porcine PGCLCs exhibit SOX17 regulatory trajectories similar to those of humans [34].

Research on gametogenesis induced from mammalian stem cells provides valuable insights for human stem cell studies. Animals like pigs, with developmental biology similarities with humans, help validate regulatory factors and ensure genetic stability. Meanwhile, small mammals, being cost-effective and easier to handle, play an important role in advancing reproductive research. Selecting suitable animal models based on species traits is critical for advancing stem cell meiosis induction. Ultimately, this research aims to develop methods for generating gametes from hiPSCs in vitro. Studying induction processes in different animals will help establish a foundation for identifying appropriate stem cell sources, organoid carriers, and optimized methodologies [12].

Stem cell-induced ovarian somatic cells

The formation of ovarian organoids requires the participation of both germ cells and ovarian somatic cells, making it crucial to generate ovarian somatic cell-like cells in vitro. Previous studies have used ovarian somatic cells derived from germ cell-depleted sources, which pose challenges such as genetic instability and immune rejection during clinical translation. Novel techniques for generating fetal ovarian somatic cell-like cells from ESCs/iPSCs could eliminate the need for embryonic somatic cells. Yoshino et al. [35, 36] used several signaling factors (WNT, BMP, SHH, and RA) to stimulate signaling pathways guiding the differentiation of mouse pluripotent cells into E12.5 ovarian somatic cell-like cells. When co-cultured with mPGCLCs, these FOSLCs form reconstituted ovarian-like compounds (rOvarioids) capable of producing functional oocytes.

In 2021, Wang et al. [37] found that treating cells with a combination of AM580 and vitamin C induced the expression of Foxl2 and Gata4, activating RA- and Wnt-related pathways to produce E12.5 gonadal somatic cell-like cells (E12.5 GSCLCs). Co-culturing these cells with PGCs activated meiosis in vitro, although with much lower efficiency than that achieved with E12.5 GSCs.

Producing ovarian somatic cells is a relatively challenging process because these cells constitute a complex group, including ovarian surface epithelium (OSE) cells, granulosa cells, and theca cells. Some studies have reported successful induction of certain types of OSE cells. In 2013, Liu et al. [38] reported that inducing microRNA-17-3P expression while inhibiting vimentin expression transformed hiPSCs into estrogen-sensitive ovarian epithelial-like cells (OSE-like cells). These cells were transplanted into a mouse model of POF, where they secreted estradiol.

Only a few studies have successfully generated granulosa cells in vitro. In a 12-day multi-step protocol, Lan et al. [39, 40] used recombinant signaling proteins to induce hESCs to differentiate into functional ovarian granulosa-like cells, although yields were low (12–36%). In 2016, Liu et al. [41] utilized various cell growth factors (TGF-β and human growth hormone) and hormones (E2, AMH, inhibin α, and inhibin β) at different time points to induce iPSCs to generate granulosa-like cells after 12 days. These cells were validated in a mouse model of POF, where they repaired ovarian damage and improved ovarian function. Similarly, Lipskind et al. [33] found that EBs derived from mESCs and human amniocyte-derived induced PSCs (hAdiPSCs) could serve as biologically relevant models for ovarian granulosa cell differentiation and steroid cell production, although conversion efficiency was relatively low. In 2023, Smela et al. [39] demonstrated that overexpressing transcription factors NR5A1 and RUNX1 or RUNX2 was sufficient to produce granulosa-like cells, with an induction efficiency of up to 70%, and tested their endocrine functions. There are no reports of generating theca cells from PSCs, with current reports limited to theca cells derived from theca stem cells [42].

While the induction of ovarian somatic cells from stem cells still requires optimization, recent progress in understanding granulosa cell development has led to advancements in restoring ovarian function in endocrine disorders. However, human-like pulsatile hormone secretion has not yet been replicated, and high induction rates and full-type differentiation of ovarian somatic cells have yet to be achieved. Future research efforts should focus on replacing exogenous gonadal cells with defined synthetic factors or iPSC-derived gonadal cells to pave the way for constructing ovarian organoids.

Current research on human ovarian somatic cell development focuses mainly on transcriptomics, highlighting stage-specific gene expression. Reported key transcription factors such as FOXL2, GATA4/6, NR5A1, WT1, and SOX9 are important for this process but require further investigation [43]. This is partly due to the limited progress in inducing human ovarian somatic cells, making it difficult to validate their roles in development.The interaction between follicular and somatic cells is crucial, influenced by hormones like FSH and LH, as well as gene interactions. Transcriptomic studies suggest that the NOTCH signaling pathway regulates oocyte-mediated granulosa cell proliferation and differentiation [44]. Further research is needed to identify key regulatory factors and understand hormonal regulation, laying the groundwork for in vitro induction of ovarian somatic cells.

In conclusion, recent studies have demonstrated the induction of ovarian germ cells, granulosa cells, and epithelial cells from hiPSCs. However, ethical concerns prevent full verification of their genetic and epigenetic stability, raising concerns about potential mutations and germ cell tumors. The simultaneous induction of multiple ovarian cell types remains a challenge, and meiosis in the absence of ovarian somatic cells has yet to be achieved. Further research is needed to develop functional artificial ovaries with endocrine and reproductive capabilities to treat female infertility and improve reproductive health.

Extracellular matrix construction

The construction of organoids requires stem cell induction and a suitable ECM for mechanical and nutritional support in 3D culture. The ECM is a vital network that supports stem cell growth, structural integrity, and the exchange of essential substances between organoids and culture medium. It regulates cell behavior and maintains tissue-specific functions, making ECM selection critical for successful organoid construction [6, 45]. ECM characteristics have been extensively leveraged in tissue engineering and regenerative medicine research, aiming to restore the function of damaged or dysfunctional tissues. The ovarian ECM environment is both complex and dynamic, regulated by periodic changes in endocrine factors from both systemic circulation and local signals. Its properties enhance interactions between follicles and somatic cells, making it a key focus in regenerative reproductive medicine [46].

Matrigel, derived from mouse Engelbreth–Holm–Swarm (EHS) sarcoma, is the most commonly used ECM for organoid construction. It supports cell growth with essential proteins and growth factors and has been widely utilized in stem cell differentiation and tumor organoid studies. However, Matrigel’s tumor origin, batch variability, instability, immunogenicity, limited tunability, and lack of scalability limit its use in advancing organoid culture technologie [47].

To address these limitations, two innovative approaches have been proposed. The first involves using biomaterials with well-controlled components and mechanical properties, while the other employs tissue decellularization techniques to create tissue-specific matrices [1].

Biomaterials, with their defined biochemical composition, mechanical properties, and low cost, are expected to become an alternative to Matrigel. Dadashzadeh et al. [48] reviewed the literature on various 3D culture systems designed to support follicle development. Natural polymers such as alginate, tyramine-based HA hydrogel, collagen, fibrin, alginate–collagen mixtures, and plasma clots have been employed to establish these systems. These approaches have enhanced follicle integrity, diameter, and oocyte maturation. Synthetic hydrogels, such as polyethylene glycol (PEG) derivatives, have also been used to mimic natural ovarian ECM. However, recent research remains limited to the follicle stage, and further studies are required to evaluate their potential for early stem cell attachment and gamete induction. Significant challenges, including reproducibility, hypoxia, insufficient nutrient supply, and immune rejection, must be addressed before clinical application [7].

To enhance follicle survival, researchers have proposed adding angiogenic factors like VEGF, bFGF, and EGF to scaffolds, which can promote vascularization by attracting endothelial cells. Limited studies have confirmed that VEGF and bFGF can induce angiogenesis and improve follicle survival [48,49,50,51]. Optimizing the mechanical and biological properties of synthetic materials, combined with appropriate regulatory factors, could create a microenvironment conducive to vascularization and cell growth, potentially enabling synthetic biomaterials to replace Matrigel for efficient culture and industrial-scale production.

Decellularized ECM (dECM) is another promising alternative, as it removes cellular components—primarily DNA—from tissues while preserving the tissue’s biochemical and structural properties. With low immunogenicity and a non-tumor origin, dECM is ideal for organoid construction, providing scaffolds with optimal porosity, stiffness, and elasticity for tissue repair. Methods for decellularizing ovarian tissues and their application in ovarian research are summarized in Table 1. The combination of dECM and hydrogels has also shown promise. Studies indicate that ovarian cells grown in recombinant dECM and alginate hydrogels exhibit high biocompatibility. Notably, depleting elastin microfibril interface-1 (EMILIN1) from synthetic polymers combined with decellularized mouse ovarian tissue has been shown to support follicle survival in vitro [52]. Nikniaz and colleagues [53] corroborated these findings, reporting an 85.9% viability rate for transferred follicles within the ECM-alginate scaffold after seven days.

Decellularized ovarian scaffold hydrogels provide an ideal platform for supporting follicle development and restoring ovarian function, offering both physical and chemical support for oocyte maturation. Due to ethical limitations, obtaining dECM from human ovaries is challenging. Therefore, the most feasible approach involves using dECM from other species for human applications. Current research on ovarian dECM hydrogels is restricted to in vitro studies focused on promoting follicle growth and oocyte maturation, with no in vivo studies available. Further exploration of suitable hydrogel materials and their characteristics is necessary to complete the development of ovarian organoids [54, 55].

Table 1 Studies related to ovarian decellularization

Products derived from dECM from other organs, such as cartilage, adipose tissue, and muscle, including decellularized tissue sheets, have already entered clinical practice. However, the application of dECM for ovarian tissue remains in its early stages. One significant challenge is the recellularization process, which requires complex co-culture systems to reseed all necessary cell types to reconstruct the ovarian microstructure. Furthermore, maintaining an artificial organ in culture demands substantial nutritional and oxygen support, and extensive research is needed to overcome these limitations [66].

3D printing has introduced innovative methods for creating ovarian-like organ ECMs, although ovarian 3D printing is still in its infancy. In 2017, Laronda et al. [67] attempted to generate prosthetic ovaries using 3D-printed microporous scaffolds made from gelatin. They reported partial restoration of ovarian functions, including hormone secretion and follicle production, in sterilized mice, culminating in the birth of healthy offspring. More recently, in 2022, T Wu et al. [4] designed and fabricated a 3D artificial ovary using a gelatin-methacryloyl (GelMA) bio-ink extrusion method. In the same year, Zheng et al. [68] employed 3D printing to create dECM scaffolds from pig ovaries, which were co-cultured with mouse ovarian fragments to construct artificial ovaries. In vivo and ex vivo experiments with mice confirmed the significant role of these constructs in repairing damaged ovaries. Given its excellent biomimetic capabilities, 3D printing is expected to play a critical role in developing supportive ovarian components in future research.

In summary, while Matrigel remains essential for research, bio-synthetic materials offer unmatched advantages in tissue engineering and regenerative medicine due to their controllability and scalability. For ovarian tissue engineering, biomaterials must balance biological factors (e.g., cell attachment and differentiation) with tissue-like mechanical properties to support organoid formation and improve transplantation outcomes. Optimizing material type, concentration, and stiffness is crucial for preserving and reconstructing the ovarian environment. Combining 3D printing with dECM technology holds great promise for constructing ovarian organoids.

Current status of ovarian organoids

The ovary is a critical organ for maintaining female reproductive and endocrine functions. Current fertility preservation strategies include cryopreservation of embryos, immature or mature oocytes, and ovarian tissue cryopreservation and xenotransplantation. Hormone replacement therapy (HRT) is often used to maintain endocrine function. However, these approaches have notable limitations, including economic burden and increase health risks. For example, infertility patients may be unable to produce viable embryos, while HRT increases the risk of ovarian cancer and is unsuitable for patients with estrogen-dependent cancers or a history of deep vein thrombosis [69]. Additionally, transplanted ovarian tissue lasts only 4–5 years and may provoke strong immune responses [57]. To address these challenges, ovarian organoids are emerging as a promising alternative to restore ovarian function without the burden of the original disease [64].

Rapid advancements in developmental biology and cellular regulation have shed light on the developmental mechanisms of germ cells in different species, enabling the directed differentiation of iPSCs to construct ovarian organoids. One of the latest trends in bioengineering, organoid development involves the 3D arrangement and culture of cells to replicate tissue and organ functions in vitro. Ovarian organoids have diverse applications, such as PODs for studying disease mechanisms, drug screening (e.g., for ovarian insufficiency, endometriosis, ovarian cancer), and immunotherapy. They can also replace ovarian endocrine function, induce the production of mature oocytes in vitro for infertility treatment, and preserve fertility, offering vast potential for clinical applications (Fig.