Germ cells of the mammalian female: A limited or renewable resource?†

Abstract

In many non-mammalian organisms, a population of germ-line stem cells supports continuing production of gametes during post-natal life, and germ-line stem cells are also present and functional in male mammals. Traditionally, however, they have been thought not to exist in female mammals, who instead generate all their germ cells during fetal life. Over the last several years, this dogma has been challenged by several reports, while being supported by others. We describe and compare these conflicting studies with the aim of understanding how they came to opposing conclusions. We first consider studies that, by examining marker-gene expression, the fate of genetically marked cells, and consequences of depleting the oocyte population, addressed whether ovaries of post-natal females contain oogonial stem cells that give rise to new oocytes. We next discuss whether ovaries contain cells that, even if inactive under physiological conditions, nonetheless possess oogonial stem cell properties that can be revealed through cell culture. We then examine studies of whether cells harvested after long-term culture of cells obtained from ovaries can, following transplantation into ovaries of recipient females, give rise to oocytes and offspring. Finally, we note studies where somatic cells have been re-programmed to acquire a female germ-cell fate. We conclude that the weight of evidence strongly supports the traditional interpretation that germ-line stem cells do not exist post-natally in female mammals. However, the ability to generate germ cells from somatic cells in vitro establishes a method to generate new gametes from cells of post-natal mammalian females.

Keywords: fertility; follicle; mammals; oocyte; stem cells.

Figure 1

Outline of female germ cell development. Following their arrival at the genital ridge, germ cells undergo mitotic proliferation accompanied by incomplete cytokinesis to generate oogonial cysts that are surrounded by pre-granulosa cells. The oogonia in the cysts subsequently enter meiosis and progress to late diplotene. Near the time of birth, the cysts break down and some oocytes become enclosed by granulosa cells in primordial follicles, whereas others do not and degenerate. Following birth, individual primordial follicles become activated to enter the growth phase, marked by an increase in oocyte size and mitotic proliferation of the granulosa cells. As the oocyte continues to grow, it elaborates the zona pellucida, and as the granulosa cells continue to proliferate and generate multiple layers, a basement membrane is laid down distally and layers of thecal cells are recruited. The development of the fluid-filled antrum during late follicular growth separates the granulosa cells into mural and cumulus sub-populations, and large antral follicles are selected for ovulation.

Figure 2

Searching for OSCs within the intact ovary. (A) Left: as described in [19], mice were generated that carry a gene encoding Cre recombinase (Cre) fused to a form of the estrogen receptor that has been modified so that it responds to tamoxifen (Tmx) but not to estrogen, under the control of a ubiquitously expressed promoter, and a gene encoding EYFP at the ROSA26 locus with a transcription arrest sequence flanked by loxp sequences. A low dose of Tmx will induce Cre expression in some cells, deletion of the stop sequence, and stable activation of Eyfp expression, thereby marking the cells. Right: following tamoxifen withdrawal, the total number of follicles and the number containing a marked oocyte were followed over time. A more rapid drop in marked follicles compared to total follicles would suggest that new follicles were generated, whereas no difference would suggest they were not. Graph is redrawn from [19]. (B) Left: as described in [20], mice were generated that carry a gene encoding a Tmx-regulated Cre, as above, inserted at the endogenous Foxl2 locus and the mTmG gene where a loxp-flanked stop sequence separates sequences encoding tdTomato and EGFP. Exposure to Tmx will induce recombination in a fraction of the Foxl2-expressing cells, leading to a stable switch from red to green fluorescence in them. If new granulosa cells are derived from pre-granulosa cells in which the Foxl2 promoter was not active at the time of exposure to Tmx, they will be red; thus, follicles containing oocytes produced after birth will contain red granulosa cells. The image from [20] shows that all follicles contain only green granulosa cells. IRES: internal ribosomal entry sequence. (C) As described in [21], mice were generated that carry genes encoding a Tet activator (rtTA) under the control of a 1.4-kb fragment of the Stra8 promoter, encoding Cre under the control of a Tet-response element (TRE) and encoding YFP preceded by a loxp-flanked stop sequence at the ROSA26 locus. Exposure to doxycycline will stably activate YFP expression in cells where the Stra8 promoter fragment is concurrently active. Right: an image from [22] shows a YFP-expressing oocyte in a pre-antral follicle. (D) Left: activity of the Stra8 promoter fragment in pre-meiotic germ cells will generate labeled oocytes. Right: activity of the fragment in oocytes in primordial follicles will also generate labeled oocytes. All images and figures reproduced with permission from the copyright holder.

Figure 3

Searching for OSCs after depleting oocytes within the ovary. (A) As described in [20], mice were generated that carry a gene encoding Cre under the control of the Gdf9 promoter and the diptheria toxin receptor (DTR) preceded by a loxp-flanked stop sequence at the ROSA26 locus. If OSCs exist, new oocytes and follicles will be generated following withdrawal of the toxin. Image on the right from [20] shows that new follicles are not observed. (B) As proposed in [22], mice were generated that carry a gene encoding thymidine kinase (Gfp-HSVtk) under the control of a 1.4-kb fragment of the Stra8 promoter. Under steady-state conditions, the generation of new follicles combined with the loss of existing follicles would create a steady turnover in the follicle population. Exposure to ganciclovir will kill cells in which the Stra8 promoter fragment is concurrently active. This leads to a loss of about 1500 follicles over a period of 3 weeks. Following withdrawal of ganciclovir, the follicle population is restored to its pre-treatment level. For this to occur, the rate of generating new follicles must substantially exceed the rate of loss. (C) Staining pattern obtained using anti-SYCP3 antibody in putative new oocytes (upper, [22]) and in embryonic germ cells (lower, [29]). Stained structures are green, as indicated by white arrows. (D) As described in [30], embryonic ovarian cells of mice carrying a GFP marker were injected into the ovaries of adult mice. Follicles consisting of GFP-positive oocytes and GFP-positive granulosa cells were observed, but no follicles containing a mix of GFP-negative (host) and GFP-positive cells were observed. Image is from [30]. All images and figures reproduced with permission from the copyright holder.

Figure 4

Gene expression in putative OSCs. Each row shows the genes reported in the study that is indicated in the left-most column. The second column indicates the method of purification, except for the top two rows, which show expression in ESCs and BM cells. Results obtained using RT-PCR are shown for freshly isolated ovarian cells (blue boxes), cultured cells (red boxes), and large cells observed after culture (red boxes in lower panel). +: expression detected; (+): weak expression detected; −: no expression detected.

Figure 5

Characteristics of putative OSCs. (A) Images of cultured putative OSCs. Images are at the same magnification and from references, clock-wise from upper left: [28, 32, 37, 41, 42, 52]. (B) Expression score of indicated marker genes in different cell types isolated from human ovary and subjected to single-cell RNA-sequencing. Reproduced from [7]. (C) Co-staining of perivascular cells in human ovarian fragments by anti-MCAM and anti-DDX4 antibodies. Asterisks show oocytes within follicles that are stained by anti-DDX4 antibody. Reproduced from [7]. (D) Images of large cells that arose during cultured of putative OSCs. Images are at the same magnification and from references, clock-wise from upper left: [27, 28, 38, 43, 52–54]. All images and figures reproduced with permission from the copyright holder.

Figure 6

Fate of putative OSCs after transplantation into ovaries. (A) Detection of genetically marked cells, which are brown in immunohistochemical images and green in immunofluorescent images. Images are at the same magnification and from references, clock-wise from upper left: [27, 38, 42, 48, 52, 59]. Arrow embedded in original image shows marked oocyte. (B) Immunoreactive oocyte in a pre-antral follicle [43]. (C) Image of ovary 1 day after transplantation of GFP-labeled putative OSCs [60]. (D) Immunoreactive oocytes following transplantation of GFP-labeled putative OSCs into ovaries [32, 60, 61]. Arrows embedded in original image shows marked oocytes. (E) Left: fluorescent egg and two-cell embryo following transplantation of GFP-labeled putative OSCs into ovaries [27]. Right: fluorescent egg [62] and two-cell embryo [63] from transgenic mice carrying a gene encoding GFP. All images and figures reproduced with permission from the copyright holder.