| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
review-article |
Department of Obstetrics and Gynecology and Division of Experimental Medicine, McGill University, Montreal, Quebec, Canada H3A 1A1
ABSTRACT
Germ cells constitute a highly specialized cell population that is indispensable for the continuation and evolution of the species. Recently, several research groups have shown that these unique cells can be produced in vitro from pluripotent stem cells. Furthermore, live births of offspring using induced germ cells have been reported in one study. These results suggest that it may be possible to investigate germ cell development ex vivo and to establish novel reproductive technologies. To this end, it is critical to assess if gamete induction processes in vitro faithfully recapitulate normal germ cell development in vivo. Here, this issue is discussed with a focus on the germ line specification and the sex-specific development of pre- and postnatal germ cells. The aim of this paper is to concisely summarize the past progress and to present some future issues for the investigation into in vitro gamete production from pluripotent stem cells.
assisted reproductive technology, developmental biology, gametogenesis
How truth would irk you, if you really sought it:For who can think of truth or trash to say,
But someone in the ancient world has thought it? (from "Faust" by Goethe [1])
To provide an overview of in vitro gamete induction, it will be informative first to briefly summarize the life cycle of germ cells (Fig. 1). The epiblast cells, which arise from the inner cell mass (ICM) of blastocysts, are the precursors of all cells that constitute an embryo. Until ~6 days post coitum (dpc) in mice, all epiblast cells are apparently equivalent in their developmental potential and are able to enter the germ line. However, regional specialization occurs in the epiblast shortly before gastrulation begins, and only the cells in the proximal epiblast, which directly faces extraembryonic ectoderm, acquire the ability to contribute to the germ line.
|
This specification is induced by extrinsic factors secreted by the extraembryonic cells, most prominently bone morphogenetic protein (BMP) 4 [3, 4]. BMP4 augments the expression of the Ifitm3 (fragilis) gene in the proximal epiblast, marking the first step to the germ line commitment: before this, Ifitm3 is weakly expressed throughout the epiblast [5]. Soon after gastrulation begins, around 7 dpc in mice, presumptive germ cells lose their responsiveness to BMP4, and a subset of Ifitm3+ cells that migrated to the extraembryonic mesoderm start expressing the Dpp3a (stella) gene. They represent the first population of fetal germ cells (primordial germ cells or PGCs), which can be identified based on their alkaline phosphatase activity. During this process, the expression of a gene associated with pluripotency, Pou5f1 (also known as Oct4), becomes restricted to the germ line: The Pou5f1 expression is seen in the ICM, epiblast, and then PGCs, while it becomes repressed in somatic cell lineages [6]. These are the key events known to occur during germ line specification.
PGCs that emerge in the extraembryonic space are then transferred back into the embryo per se, and migrate through the hindgut to the genital ridge, the future ovary and testis. The migration of PGCs coincides with their active proliferation. During migration, PGCs come into contact with a diverse array of somatic cells, which might modify PGCs (for example, see [7–9]). Upon completion of migration by 11.5 dpc, PGCs start expressing a marker gene for postmigratory germ cells, Ddx4 (mouse vasa homologue or Mvh) [10]. By 13.5 dpc, the ovary and testis become morphologically distinguishable, and PGCs initiate sex-specific development [11, 12] (Fig. 1). Female PGCs enter meiosis and become arrested in meiotic prophase, followed by periodic recruitment to folliculogenesis after birth. On the other hand, male PGCs enter mitotic arrest following migration to the genital ridge, and after birth, male germ cells are reactivated to initiate spermatogenesis. A uniqueness of the male germ line is that it retains self-renewing stem cells throughout life.
Two recent studies reported that the decision of meiotic entry or mitotic arrest of postmigratory PGCs is regulated by retinoic acid (RA) [13, 14]: male PGCs do not enter meiosis, because an enzyme (Cyp26b1) expressed in somatic cells in the male genital ridge degrades RA. In contrast, the repression of Cyp26b1 expression in female or its lack in null-mutant embryos allows PGCs to enter meiosis. Although the meiotic figure in female or Cyp26b1-null PGCs has not convincingly been shown, these studies strongly suggest that RA may promote meiotic entry of PGCs in the genital ridge. In this context, it is of note that PGCs that migrate to an ectopic site (e.g., adrenal gland) spontaneously enter meiosis, regardless of their genetic sex [15, 16]. It should also be noted that in vitro, RA can stimulate mitotic proliferation of PGCs up to 13.5 dpc [17]. Therefore, while migratory and postmigratory PGCs appear to respond differently to RA, the gonadal somatic environment also has an important role in regulating sex-specific germ cell development.
Thus, there are three critical events during the life cycle of germ cells: specification, migration/proliferation, and pre- and postnatal sex-specific development. Here, the main focus of discussion is placed on the specification and sex-specific development of germ cells.
PROGRESS OF IN VITRO SYSTEMS FOR GERM CELL INDUCTION FROM PLURIPOTENT STEM CELLS
Embryonic stem cells (ESCs) are derived from the inner cell mass of blastocysts and can be maintained and expanded indefinitely in culture, while retaining their ability to produce all cell types in the body (pluripotency) [18]. ESCs are tumorigenic, as they produce teratocarcinomas when transplanted into immunocompatible hosts. When introduced into preimplantation embryos, however, ESCs can integrate into normal embryonic development, with no tumorigenesis, and contribute to all cell lineages in the body, including the germ line [18]. Such a robust differentiation potential of ESCs provides a unique tool to generate various cell types, which could be beneficial in regenerative medicine.
A significant difficulty in studies of in vitro gamete generation from ESCs arises from the fact that many PGC markers are identical to ESC markers [18]; thus, it is a challenging task to distinguish early embryonic germ line cells from ESCs. In order to identify germ cells in vitro, therefore, researchers have relied on molecules expressed in germ cells from postmigratory to meiotic stages, morphology of cells produced, or the function of germ cells (e.g., generation of live offspring), as described below.
Hübner et al. reported the first study of in vitro gamete production [19] (Table 1). These authors cultured ESCs carrying a Pou5f1-reporter gene without feeder cells or added growth factors. The formation of ovarian follicle-like structures was observed, which included oocyte-like cells expressing Pou5f1 and Ddx4. These cells also expressed meiotic marker genes (e.g., Dmc1, Sycp3), and were released from follicles after 3–4 wk of culture. Although the germ cell potential of oocyte-like cells was not tested by fertilization, structures that closely resemble blastocysts emerged in ~6 wk, suggesting spontaneous parthenogeneic activation. Thus, this study suggested that ESC-derived cells have a potential to undergo gametogenesis in vitro.
|
Toyooka et al. next reported the derivation of male germ cells from mouse ESCs [22]. These authors took a different approach from Hübner's, and used ‘embryoid bodies' as starting material. When ESCs are aggregated and cultured in suspension without leukemia inhibitory factor, the cells form a structure resembling a preimplantation embryo, termed an embryoid body (EB), in which differentiation occurs randomly. Using ESCs carrying a Ddx4-reporter construct, the authors induced EBs and found that Ddx4+ cells spontaneously appeared, suggesting the presence of cells with characteristics of postmigratory PGCs in EBs. Ddx4+ cells were purified and aggregated with male genital ridge cells of wild-type embryos. Following implantation into adult mouse testes, the cell aggregates formed seminiferous tubules that supported complete spermatogenesis derived from purified Ddx4+ cells. Interestingly, the authors also found that the exposure of EBs to BMP4 led to the emergence of Ddx4+ cells within 24h. Although the ability of spermatozoa to activate eggs was not examined, this study suggested that the germ line specification and the emergence of postmigratory PGCs can occur spontaneously or be induced in EBs.
Using EBs, Geijsen et al. also suggested a spontaneous emergence of male PGCs from mouse ESCs in vitro [23]. Furthermore, the authors detected and isolated haploid cells from EBs, and showed that the injection of the cells into eggs resulted in the formation of blastocyst-like structures. Although the isolated cells did not resemble spermatozoa and analyses of further embryonic development was not completed, this study suggested that male PGCs arise from ESCs and spontaneously become postmeiotic cells that are capable of activating eggs.
Recently, Nayernia et al. reported not only the induction of male gametes from ESCs but also successful production of offspring [24]. These authors cultured mouse ESCs, which were transfected with a Stra8 (stimulated by retinoic acid gene 8)-reporter construct, and later, also with a Prm1 (protamine 1)-reporter construct, in monolayer culture. Following repeated RA treatments and selection of Stra8+ cells, they isolated Prm1+ cells and injected them into eggs. When the resulting 65 embryos were transferred to surrogate mothers, seven live pups carrying the Prm1-reporter gene were derived, which apparently had growth abnormalities and died within 5 mo after birth. Although the production of progeny needs to be confirmed by other laboratories, this study presented the potential of haploid male germ cells derived from mouse ESCs to activate the egg and lead to live births of progeny.
Using a similar RA-based approach, Nayernia et al. also reported male germ cell production from embryonal carcinoma cells (ECCs) and bone-marrow-derived mesenchymal stem cells (MSCs) in the mouse [25, 26]. In the ECC study [25], following RA-treatment of F9 ECCs, injection of cultured cells into eggs resulted in the formation of morula-like structures. The authors also transplanted cultured cells into infertile mouse testes to examine their ability to regenerate spermatogenesis (i.e., stem cell activity), and detected haploid cells in the transplants by DNA image cytometry analysis. The use of F9 ECCs may be a concern in this study, as this ECC line frequently contains chromosomal abnormalities, and its differentiation ability is often limited [27, 28]. The transplantation assay to detect activity of putative germ cells to regenerate spermatogenesis was also used in the MSC study [26]. The regeneration of spermatogenesis upon transplantation was not complete, but genes that are expressed in premeiotic germ cells (e.g., Dazl) were detected in the transplants. Although the results supporting in vitro generation of functional germ cells are not extensive, these studies represent the attempts to identify postnatal male germ cells, especially stem cells, produced in vitro.
Dyce et al. reported the potential of somatic stem cells to transform to oocytes [20]. The authors first produced skin-derived precursors (SKPs, pronounced as "skips") from pig fetuses, which can differentiate into multiple cell lineages [29], and cultured them with ovarian follicular fluid. Structures resembling the cumulus-oocyte-complex appeared by 50 days of culture, and upon gonadotropin stimulation, oocyte-like cells were released from the complex in 10 days. Although these cells were very fragile and thus appeared to have structural defects, some of them spontaneously formed blastocyst-like structures, suggesting parthenogeneic activation. Thus, this study suggested that the in vitro production of germ cells is not limited to mouse cells, and that the derivation of germ cells may be possible from somatic cells. The application of the method in mice should provide more genetic tools to understand the mechanism underlying transdifferentiation of somatic cells to germ cells.
The ability of human ESCs to enter the germ line was examined by Clark et al. [21]. These authors generated EBs from female and male human ESCs and found using PCR or immunohistochemistry that some cells in EBs expressed marker genes specific to different stages of germ line development (e.g., DDX4, DAZL). Furthermore, cells expressing a meiotic marker gene, SYCP3, were identified in human EBs. Thus, this study suggested the possibility that human ESCs of both sexes may spontaneously enter the germ line and undergo meiosis.
These studies collectively indicate that the derivation of mammalian germ cells, both female and male, is possible in vitro from pluripotent stem cells. These in vitro systems can be used to improve our knowledge about germ cell development and also to develop novel reproductive technology.
Germ cell development has been difficult to study, because important early events occur after implantation. This difficulty is most evident in humans, when ethical issues surrounding embryo manipulations are considered. An ESC-based in vitro system could provide a valuable platform to study some aspects of germ cell development, while circumventing or minimizing these problems. To this end, however, two interrelated issues need to be addressed. The first is that germ cell induction thus far reported appears to be a rare event. Thus, we need a system that allows directed induction and maintenance of germ cell development. The second is that it is yet to be determined if the in vitro process faithfully recapitulates normal germ cell development. In fact, as described below, the study of in vitro germ cell induction is still at a stage where we apply knowledge obtained in studies in vivo to situations in vitro, but not vice versa. Hence, we need to assess if the three critical steps in germ cell development (specification, migration/proliferation, and sex-specific pre- and postnatal development) indeed take place in the in vitro systems.
For the induction of germ cells in vitro, specification is the first, and perhaps, most critical step. As described earlier, specification of germ line begins with an increased expression of Ifitm3 (fragilis) in the proximal epiblast, followed by the expression of Dpp3a (stella) in a subset of Ifitm3+ cells [3, 4]. Interestingly, mouse and human ESCs express both of these genes at the undifferentiated state [19, 21–24]. The Ifitm3 expression in ESCs may not be a surprise, as it is seen in the entire epiblast, and ESCs can be derived from and are functionally similar to epiblast cells [30]. However, the Dpp3a expression could be unexpected, because it is apparently not expressed in the epiblast [5]. Although expression of these genes has been verified only by PCR and not confirmed at the protein level, the results suggest that some, if not all, of undifferentiated ESCs may have characteristics of presumptive germ cells. Identification of the cells expressing Ifitm3 and Dpp3a in ESC colonies and during in vitro germ cell induction, using in situ hybridization or immunocytochemistry, may be necessary as an approach toward characterizing "germ-cell-ness" of ESCs and comparing germ line specification processes in vitro and in vivo.
Recent in vivo studies suggest that germ line specification occurs by avoiding entry into the somatic lineage [3, 4]. Presumptive germ cells appear to initially have some characteristics of somatic cells—in particular, of mesodermal lineages [4]. Mesodermal markers, such as brachyury and Fgf8, are expressed in Dpp3a+ PGCs as well as in neighboring somatic cells. Prdm1 (Blimp1) expressed in Ifitm3+ cells repress gene expression that leads to the somatic lineage, such as the Hox genes [3, 31]. In this context, it should be noted that although the culture of ESCs is now commonly done, they are very difficult cells to maintain at the undifferentiated state, and readily differentiate unless sufficient extrinsic cues are provided. Similarly, EBs are already undergoing differentiation. Hence, the germ line specification in vitro should require a mechanism that prevents ESCs from spontaneous differentiation to somatic cell lineages.
Although it is unknown what this mechanism is, it might not arise solely within presumptive germ cells (e.g., through cell-autonomous transcriptional regulation), but could involve their communication with somatic cells in the vicinity, similar to how BMP4 secreted by the extraembryonic ectoderm increases Ifitm3 expression in the nearby epiblast. While localization of presumptive germ cells has been attempted in most of these studies, it may also be informative to identify the characteristics of neighboring non-germ cells and compare them with those observed in vivo.
As for PGC migration/proliferation process, it is not well documented whether or not the process takes place or is indispensable during in vitro gamete production. Toyooka et al. showed that Ddx4+ (Mvh+) cells increased in number and formed clusters in EBs [22]. The Geijsen study also reported a cluster of PGCs in an EB [23]. Likewise, the cluster formation was observed in monolayer culture systems [19, 24]. Considering that Ddx4+ (i.e., presumed postmigratory) cells emerge in EBs within 24 h after BMP4 treatment [22], induced PGCs might readily acquire postmigratory characteristics without the migration process in vitro. However, it may still be necessary to evaluate if the clusters form as a result of proliferation of induced germ cells and/or their migration/aggregation after specification.
The second critical step during in vitro induction of germ cells is their sex-specific development (Fig. 1). It is of note that PGCs enter meiosis unless encapsulated in a testicular environment, and ectopic PGCs can become oocytes, regardless of their genetic sex [11–16]. Therefore, it is legitimate to ask whether an appropriate testicular environment was generated in vitro to prevent meiotic entry of PGCs and to support spermatogenesis. If such an environment is not produced, male PGCs generated in vitro may enter meiosis and become haploid, while bypassing normal spermatogenic development.
However, Geijsen et al. [23] detected haploid cells in EBs using an antibody (FE-J1) that specifically reacts with postnatal male germ cells from pachytene spermatocyte to spermatozoa stages [32]. Nayernia et al. selected the cells expressing a Prm1-reporter gene [24]. Although the number of meiotic markers used for cell identification/selection was limited, these cells were able to activate the egg and induce early embryonic development or generate offspring [23, 24]. Thus, these observations suggest the possible entry of PGCs generated in vitro into the male-specific pathway, and the potential development of postnatal male germ cells. In turn, the results also suggest that cultured cells may generate an environment where such male-like germ cell behaviors can proceed [23]. Nonetheless, the Nayernia study indicated that presumptive postmeiotic Prm1+ cells emerged within 72 hr after RA treatment [24]. Considering that ~12 days are required for Type A spermatogonia to develop to round spermatids during normal spermatogenesis in mice [33], spermatogenic processes that may take place in vitro apparently do not follow normal patterns.
To further analyze the male-specific germ cell development in vitro, it may be informative to examine the development of pre- and postnatal testicular somatic cells and their interaction with induced germ cells. It is also noted that the studies of germ cell derivation in vitro thus far published have not convincingly shown normal chromosome pairing, which may be an infrequent event to occur in vitro. Perhaps, faulty meiosis could be one of the reasons for only rare success of in vitro germ cell derivation, and may be associated with a defective somatic cell environment generated in vitro. It is also important, as described below, to evaluate epigenetic imprinting status of induced germ cells and to verify the presence of male germ line stem cells.
Geijsen et al. detected erasure of DNA methylation at the H19/Igf2 locus in ESC-derived PGCs [23]. Nayernia et al. reported that some reprogramming of DNA methylation appeared to occur but involved errors [24]. The growth abnormalities and the poor viability of progeny produced using germ cells induced in vitro also suggest imprinting defects [24]. Thus, epigenetic programming associated with male-specific germ cell development apparently did not occur properly in vitro. However, the number of genes and the timing of reprogramming analyzed during culture are limited, and chromatin modification status has not been examined [34]. More extensive studies of epigenetic reprogramming status in induced germ cells of both sexes should provide important information about sex-specific germ cell development in vitro.
If male germ cells indeed undergo normal development before entering meiosis, then stem cells should be present among induced germ cells, which can be examined using spermatogonial transplantation [35]. The results of the transplantation assay available to date do not clearly support the generation of functional stem cells in vitro, as normal spermatogenesis was not established or maintained [24–26]. It is noted, however, that transplantation experiments could be difficult, because unless ESC differentiation is sufficiently induced, undifferentiated ESCs remaining in culture may cause teratomas after transplantation, as Nayernia et al. reported [24]. An alternative approach could be to selectively amplify stem cells based on recently established stem cell culture systems [36, 37]. Further studies to confirm the presence or absence of stem cells are required to verify the recapitulation of normal development of male germ cells in vitro.
As noted previously, in vitro germ cell induction systems have not sufficiently evolved to allow us to investigate normal development of germ cells ex vivo. Therefore, further in vivo studies are critical in order to establish an in vitro system as a reliable functional assay of germ cell development. In vivo studies should also provide target genes for assessing developmental processes/stages of germ cells generated in vitro. Once a series of in vitro processes is confirmed to recapitulate that in vivo, then more detailed studies in a defined process could efficiently be done using ESC-based systems.
In this respect, gene expression profiling data [38–41] should provide important information not only to understand normal germ cell development but also to generate various tools to apply in germ cell induction systems. For example, more marker molecules for specific stages of germ cell development should be identified. This is particularly important to identify PGCs at an early stage of germ cell induction from ESCs. With such data, we can also create reporter cell lines (e.g., Table 1, Dpp3a-reporter line [42]) that can be effective to isolate induced germ cells at a specific developmental stage. Candidate genes can also be identified to mutate in ESCs and evaluate their functions during germ cell development; this should be more efficient than creating transgenic animals and analyzing them. This approach could also be a method for genetic studies of human germ cell development in the future. The efficiency of in vitro germ cell induction may also be improved by controlling the expression of such genes. The establishment of in vitro systems in animal models should provide a solid foundation for the studies of human germ cell development by overcoming technical difficulties and minimizing ethical conflicts associated with manipulation of human embryos and germ cells.
Clinically, in vitro germ cell induction systems have the potential of developing into a novel assisted reproductive technology (ART). A clinical impact of such a technology would be more valuable for female germ cells than for their male counterparts, as the supply of oocytes is a major limiting factor in ART. Considering that only one study has been reported on oocyte derivation from ESCs [19], more intensive studies are essential to efficiently produce healthy oocytes in vitro.
The study by Dyce et al. also involves unique clinical implications, suggesting the possibility of deriving germ cells from somatic cells that are readily accessible [20]. Although the results raise many questions [43, 44], if this is possible in humans, the impact should be enormous. For instance, it could provide another source of oocytes for infertility patients or cancer survivors. Recipient oocytes for nuclear transfer may also be obtained with a minimal ethical conflict. Since human SKPs have been derived [29], further improvement of the method described by Dyce et al. will provide an important basis for the future use of skin as a potential source of oocytes.
I wish to conclude this essay with the words that Goethe let Wagner proclaim in "Faust" [1]:
The thing in Nature as high mystery prized,This has our science probed beyond a doubt;
What Nature by slow process organized,
That have we grasped, and crystallized it out.
There still are a number of questions left unanswered and obstacles to be overcome about gamete production in vitro. However, the excellent studies reported recently have paved the path that should lead us to grasping a long-lasting mystery of Nature. With a recent report on the identification of factors required for converting fibroblasts to ESCs [45], nothing seems to be impossible in science.
ACKNOWLEDGMENTS
I apologize to those whose studies could not be included, or were cited only partially or indirectly, because of the space constraints. I thank Drs. Hugh Clarke, Bernard Robaire, Teruko Taketo, and Jacquetta Trasler, as well as Frances Clerk and Kevin Ebata for their careful reading of the manuscript and their suggestions.
FOOTNOTES
1Supported by CIHR (MOP-49444) and FRSQ (Burse de career). 
Correspondence: 2Makoto Nagano, Royal Victoria Hospital, Room F3.07, 687 Pine Avenue West, Montreal, Quebec H3A 1A1, Canada. FAX: 514 843 1662; e-mail: makoto.nagano{at}muhc.mcgill.ca
Received: 22 October 2006.
First decision: 21 November 2006.
Accepted: 4 January 2007.
REFERENCES
This article has been cited by other articles:
|
|
 |
|
  J. L. Tilly, Y. Niikura, and B. R. Rueda The Current Status of Evidence for and Against Postnatal Oogenesis in Mammals: A Case of Ovarian Optimism Versus Pessimism? Biol Reprod, January 1, 2009; 80(1): 2 - 12. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. E. Gargett Review Article: Stem Cells in Human Reproduction Reproductive Sciences, July 1, 2007; 14(5): 405 - 424. [Abstract] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
