Infertility is a big issue in reproductive biology and medicine that plagues many couples, often via gametogenesis failure (1). Zhou et al. aims to overcome this obstacle by producing gametes in vitro. Previous studies explored generating haploid gametes in vitro from pluripotent stem cells (2, 3) or from a 3D culture system (4) but these studies have not evaluated the generated gametes’ function or reached the “gold standard” of meiosis: having correct DNA content in various meiotic stages, having appropriate protein localization during synapsis and recombination, and having the ability to produce functional, fertile progenies. Thus, Zhou el al. explores techniques to generate spermatid-like male gametes from murine embryonic stem cells(ESC)-derived primordial germ cell-like cells (PGCLCs).
The presented data is convincing, and there does not seem to be any faulty logic. However, the authors conducted this experiment to combat human infertility, though the entire study is limited to mice, and they mentioned in the paper that “[i]n humans, the risk for tumorigenesis prohibits in vivo transplantation approaches so that the capacity of human PGCLCs to form spermatozoa remains unexplored” (1). Given this information, the authors should suggest how their experiment may translate to human subjects so they can solve the problem they describe in the bigger picture.
The authors hypothesize: if appropriate morphogens could be added to the culture containing ESCs, then ground state ESCs can differentiate, through multiple steps, into PGCLCs that resembles in vivo, functional, PGCs. The experiment includes several major steps: generating mouse transgenic ESC lines that have fluorescent reporters to label key marker genes of each stage of germ cell development and meiosis, using morphogens to differentiate ESCs first into epiblast-like cells(EpiLCs) and then into PGCLCs, co-culturing these PGCLCs with early postnatal testicular cells leading to meiosis initiation and formation of spermatid like cells(SLCs), and functional tests to ensure the viability of the generated male spermatids. These steps are included on the timeline in Figure 1A.
For this experiment, the authors chose two lines of ESCs: Blimp1-mVenus and Stella-ECFP(BVSCs) and Stra8-EGFP and Prm1-DsRed(SGPDs). These two lines have different expressions controlled by promoters that are active at different stages of development: BVSCs to identify cells resembling PGCs, Stra8 to identify early-stage spermatogonia and spermatocytes, and Prm1 to identify post-meiotic spermatids. The lines are proven to be totipotent via tetraploid complementation.
The specification of PGCLCs from ESCs in vitro initiates with applying actin A and basic fibroblast growth factor(bFGF) to promoter EpiLC differentiation. This change results in a downregulation of pluripotency markers NANOG and SOX2 expression within 2 days of EpiLC culture yet OCT4 remains expressed, as seen in Figure S2A. To differentiate EpiLCs into PGCLCs, Hayashi et al.’s protocol is used: N2B27, containing precursor to retinoic acid, vitamin A, and insulin, serves as a basal medium for PGCLC induction. This results in Blimp1 and Stella induction in Figure 1B on day 4 and day 6 of PGCLC culture, suggesting the formation of PGCLCs. Changes in gene expression profile and histone modification in PGCLC resemble that of in vivo PGCs.
Then, these PGCLCs are co-cultured with early postnatal testicular cells from KITw/KITw-v mouse in order to mimic the in vitro environment male gametes during develop in. The mouse was a germ cell-deficient model to ensure the donor origin of these haploid cells. Combinations of morphogens: activin A, BMPs, and retinoic acid are in this culture; within three days, Stra8-EGEP-expressing cells become apparent in Figure 2A, particularly 3 days after co-culturing. Loss of BV and SC expression in Figure 3SB shows differentiation from a PGC-like stage. This helps to conclude that meiosis has been initiated.
After the initiation of meiosis, sex hormones including follicle-stimulating hormone(FSH) and testosterone(T) are added to regulate meiosis’s progression. Bovine pituitary extract(BPE) is also added to the extract. Three morphogens previously added into the co-culture are removed. This helps to move the cells into a postmeiotic state, and it is successful by the expression of Prm1-DsRed on day 10 in Figure 3A. SLCs are formed from the cells, and assays are conducted to ensure the cells have correct 1C DNA content to show their identity as haploid male SLCs.
Tests remain to determine whether the SLCs generated from ESCs in vitro comply to the “gold standard” of meiosis. These assessments include checking chromosomal synapsis and recombination/cross-over, events that happen naturally in meiosis. Results show that these cells can repair their double strand breaks through homologous recombination repair. The expression of SYCP1 in the in vitro spermatocytes’ nuclei indicates synaptonemal complex’s presence. Normal histone distribution can be found, and the estimated conversion rate from PGCLCs to SLCs was about 10% (Zhou et al., 2016).
Finally, these SLCs are tested for their viability. Indicated by global transcription profile cluster analyses, in vitro SLCs are similar to in vivo spermatids. Intracytoplasmic sperm injection(ICSI) was preformed into donor eggs, in which most developed into the two-cell stage embryos. The viable embryos are transferred into a recipient host, in which six pups were born as a result. These pups have a normal DNA methylatioin level of about 50%. Furthermore, these pups developed into adulthood and have pups themselves, indicating that the gametes developed are functional and this experiment is successful.
To further validate their argument, Zhou et al. should perform these experiments in a different mouse line to show that they can replicate these data and make the general conclusion that both SGPD and BVSC ESC lines can generate functional spermatids in vitro. Additionally, it was mentioned that SLCs was generated in 14 days, the maturation confirmed by the upregulation of spermatid markers such as Tp1, Prm1, acrosin, and haprin. However, spermatid generation takes up to a month in normal development. There is a lack of discussion and interpretation in Zhou et al.’s paper for why this would be, but this generation time difference may indicate some undiscovered differences in generated SLCs versus normal sperm cells. The pass the key function tests (giving rise to offsprings) but more study should be done to investigate possible SLCs/WT sperm cell differences.
For their next steps after validating spermatids generation in mice, Zhou et al. can try to use iPSCs to generate functional male gametes in vitro, try to generate functional female gametes in vitro, and eventually try to generate functional male and female gametes in vitro for humans, using ESC, iPSCs, or other cell types to combat human infertility resulting from defects in gametogenesis.
(1) Zhou, Q. et al., 2016, Complete Meiosis from Embryonic Stem Cell-Derived Germ Cells In Vitro, Cell Stem Cell, http://dx.doi.org/10.1016/j.stem.2016.01.017
(2) Eguizabal, C. et al., 2011, Complete Meiosis from Human Induced Pluripotent Stem Cells, Stem Cells 29, 1186-1195
(3) Geijsen, N. et al., 2004, Derivation of embryonic germ cells and male gametes from embryonic stem cells, Nature 427, 148-154
(4) Yokonishi, T., et al., 2013, In vitro reconstruction of mouse seminiferous tubules supporting germ cell differentiation, Biol. Reprod. 89, 15.