In vitro oogenesis breaks free of the ovary

Abstract

In the absence of ovarian somatic cells, Saitou and colleagues differentiated mouse pluripotent stem cells to large germinal-vesicle stage oocyte-like cells. Their new paradigm advances the field toward clinical translation and offers mechanistic insights into oogonial cyst breakdown, X chromosome dynamics, and factors required for oocyte growth and meiotic resumption.

The prospect of generating sperm or eggs from patient-specific pluripotent stem cells (PSCs)– in vitro gametogenesis— could revolutionize infertility care, enabling any person to produce biological offspring, even without functional gonads or gametes. In vitro oogenesis presents particular challenges, as oocytes communicate intimately with ovarian somatic cells from prenatal follicle formation through maturation and ovulation. Prior approaches relied upon mouse fetal ovarian somatic cells to instruct PSC-derived gametic precursors, called primordial germ cell-like cells (PGCLCs), to adopt an oogenic fate, subsequently encapsulating them within follicles to support growth and maturation¹. However, co-culture with mouse ovarian soma² or human PSC-derived ovarian somatic cells³ is insufficient for human PGCLC differentiation to oocyte-like cells.

Nosaka et al.⁴ bypass the need for ovarian somatic cells to produce germinal-vesicle (GV)-stage mouse oocyte-like cells from PSCs, achieving a breakthrough toward clinical translation of in vitro oogenesis. Starting with mouse embryonic stem cells, they induced PGCLCs⁵ and co-cultured them not with fetal ovary cells, but instead with a fibroblast feeder line known as m220 (Figure 1). To induce commitment to oogenesis, they added a pulse of retinoic acid (RA) and bone morphogenetic protein (BMP) 2. By nine days of culture, oocyte-like cells robustly initiated meiotic prophase I, and by 13 days, ~70% reached diplotene arrest, as expected for fetal oocytes. Throughout epigenetic reprogramming, meiotic entry, and prophase I progression, oocyte-like cells matched in vivo counterparts in tempo and transcription. On gas-permeable plates with m220 feeders, Oocyte-like cells grew dramatically – from ~13μm to ~65μm in diameter. Remarkably, these sizes match mouse oocytes in pre-antral and early antral stage follicles, achieved without granulosa cell encapsulation. Finally, upon in vitro maturation, 30% of oocyte-like cells resumed meiosis and completed GV breakdown and <10% of those extruded a polar body-like structure, but none progressed to metaphase II of meiosis as do mature oocytes.

Figure 1: Near-complete in vitro oogenesis without ovarian somatic cells.

Schematic illustrating the steps of Nosaka and colleagues’ ovarian-soma free in vitro oogenesis protocol alongside analogous steps in in vivo oogenesis. Figure generated with Biorender.

Despite incomplete maturation, these follicle-free oocyte-like cells provide valuable mechanistic insights into oocyte developmental dynamics and the contributions of ovarian somatic cells to meiotic competence. BMP2 synergy with RA to promote oogenic fate corroborates the importance of the BMP-responsive transcription factor ZGLP1 in this lineage commitment⁶. Surprisingly, RA promoted meiotic progression best as a transient early pulse and RA target genes were not upregulated until subsequent BMP2 addition, suggesting that RA primes meiotic gene targets for later BMP2 responsiveness. During culture on m220 feeders, PGCLCs expanded nearly 30-fold and generated interconnected cyst-like structures of ~30 cells; following meiotic entry, oocyte-like cells rapidly declined in number as their cyst-like structures broke down to ~10 cells, recapitulating dynamics observed in vivo⁷. This challenges the notion that pregranulosa cells are strictly required for germ cell elimination and cyst breakdown. Finally, this work provides insights on the role for X chromosome reactivation in oocyte development, stemming from the known propensity for mouse XX PGCLCs to lose one X chromosome during induction from PSCs. The observed depletion of XO oocyte-like cells from 45% at culture day seven to <10% at day 13 may reflect the importance of moderate X-chromosome inactivation followed by reactivation to confer competence for meiotic entry⁸.

Through meticulous optimization of culture conditions that promote the growth and maturation of oocyte-like cells, the authors shed new light on how somatic cells grant developmental competence to the oocyte. Dramatic improvement of oocyte-like cell growth in gas-permeable plates underscores the known importance of oxygen sensing in regulating oocyte growth and follicle activation. Of the media supplementations tested, SCF had the largest effect, highlighting the role for secreted SCF from ovarian somatic cells in coordinating oocyte growth in vivo. Strikingly, direct cell-cell contact was required between m220 feeder cells and oocyte-like cells, implicating an additional role of membrane-SCF and other membrane-bound factors as critical cues for growth. Finally, oocyte-like cells were unable to reach their maximum size without chemical inhibition of meiotic resumption; this suggests that ovarian follicles facilitate ongoing growth while maintaining oocyte meiotic arrest in vivo.

Differences between Nosaka and colleagues’ most mature oocyte-like cells and in vivo oocytes across stages may inform further optimization. Though the epigenetic reprogramming achieved during PGCLC expansion was adequate for subsequent BMP2 and RA-responsiveness, PGCLCs and oocyte-like cells in this study maintained elevated genomic methylation through meiotic onset; this is potentially attributable to premature upregulation of Dnmt3l, a methylation regulator typically not expressed until preantral/antral stages in oocytes. Furthermore, decreased expression of the critical DNA methyltransferase DNMT3A in growing oocyte-like cells may explain their reduced methylation programming. Despite transcriptional concordance with stage-matched in vivo oocytes, these epigenetic abnormalities in oocyte-like cells resemble those reported in oocyte-like cells directly induced from mouse PSCs by overexpression of key transcription factors⁹, highlighting the separability of the transcriptional network driving oocyte growth from epigenetic reprogramming and de novo methylation. Nosaka and colleagues’ oocyte-like cells upregulated genes involved in regulation of dendrite development including RADIXIN compared to pre-antral oocytes. These abnormalities likely reflect the absence of granulosa cells that interact directly with oocytes through transzonal projections perforating the zona pellucida, and suggest that the oocyte-like cells were seeking interaction partners through increased actin-driven membrane protrusion activity. Furthermore, oocyte-like cells harbored fewer endolysosomal vesicular assemblies (ELVAs), structures that sequester aggregated proteins for degradation after meiotic maturation¹⁰. This suggests that contact with ovarian somatic cells provides yet-unknown critical factors that prepare the oocyte for successful embryogenesis.

In defining culture conditions for the generation of GV-stage oocyte-like cells from mouse PSCs, Nosaka and colleagues lay the groundwork for further investigation into the molecular bases of PGC epigenetic reprogramming, epigenetic programming in oocytes, and growth and meiotic resumption competence of oocytes— all of which are critical to understand for the eventual generation of functional human oocytes in vitro. Since the use of human fetal ovarian somatic cells is ethically fraught, circumventing the need for these cells to generate oocytes in vitro is a major step toward future clinical application of in vitro oogenesis. However, successful meiotic resumption and chromosome segregation are essential for generating viable euploid embryos. Despite remarkable advances, the inability of this system to generate fully mature metaphase II mouse oocytes reaffirms that clinical application remains on the horizon. For now, scientists and clinicians eager to spur progress toward that lofty goal should seek to harness Nosaka and colleagues’ system for further exploration of fundamental oocyte and ovarian biology.