|Home | About | Journals | Submit | Contact Us | Français|
A fundamental step in embryonic development is cell differentiation whereby highly specialised cell types are developed from a single undifferentiated, fertilised egg. One of the earliest lineages to form in the mammalian conceptus is the trophoblast, which contributes exclusively to the extraembryonic structures that form the placenta. Trophoblast giant cells (TGCs) in the rodent placenta form the outermost layer of the extraembryonic compartment, establish direct contact with maternal cells, and produce a number of pregnancy-specific cytokine hormones. Giant cells differentiate from proliferative trophoblasts as they exit the cell cycle and enter a genome-amplifying endocycle. Normal differentiation of secondary TGCs is a critical step toward the formation of the placenta and normal embryonic development. Trophoblast development is also of particular interest to the developmental biologist and immunobiologist, as these cells constitute the immediate cellular boundary between the embryonic and maternal tissues. Abnormalities in the development of secondary TGCs results in severe malfunction of the placenta. Herein we review new information that has been accumulated recently regarding the molecular and cellular regulation of trophoblast and placenta development. In particular, we discuss the molecular aspects of murine TGC differentiation. We also focus on the role of growth and transcription factors in TGC development.
By the end of the pre-implantation period of development the murine blastocyst cavity is surrounded by the trophectoderm (TE), which develops into different types of trophoblast cells after implantation (Figs. 1 and and2).2). Polar TE overlies the inner cell mass (ICM) and gives rise to trophoblast stem cells (TSCs). Shortly after implantation, the mural TE, on the side away from the ICM, differentiates into primary TGCs. Polar TE differentiate later into secondary TGCs (Figs. 1 and and2).2). Primary TGCs form an anastomosing network of blood sinuses at the periphery of the embryo and facilitate the diffusion of oxygen and nutrients from the maternal circulation into the embryo (Bevilacqua and Abrahamsohn, 1988). Secondary TGCs arise from the polar TE-derived ectoplacental cone (EPC) from around day 7–7.5 post-coitum (p.c.) onwards (Figs. 2 and and33).
TGCs have characteristic large polyploid nuclei due to continuous rounds of DNA replication, without intervening mitosis (Gardner and Davies, 1993; MacAuley et al., 1998; Nakayama et al., 1998.) There are no direct cell lineage tracing experiments to date to support the differentiation of the EPC into secondary TGCs. However, spontaneous differentiation of cultured EPCs to secondary TGCs provides supportive evidence for this model (Romagnano and Babiarz, 1990; Sutherland et al., 1991, 1993; El-Hashash and Kimber, 2004, 2006).
TGCs mediate embryo implantation and conceptus invasion into the uterus, which are facilitated by proteinases, such as urokinase-type plasminogen activator (UPA) and gelatinase A,B/matrix metalloproteinase-9 (MMP-9) (Teesalu et al., 1996; Alexander et al., 1996), as well as by chorionic gonadotrophin hormone (Prast et al., 2008). In contrast, decidual metalloproteinase inhibitors, such as TIMP-1 and TIMP3 counteract TGC-derived proteinases and limit TGC invasion into the uterine stroma to several hundred microns in rodents (Alexander et al., 1996; Cross et al., 2003). This is a critical process, since unlimited TGC invasion could extend beyond the endometrium and thus give rise to tumour-like outgrowths, whereas the embryo will not receive enough nutrients if invasion is insufficient.
Most recently, Kuang et al. (2009) identified CXCL14, a member of the chemokine family, as an important paracrine/autocrine modulator regulating trophoblast outgrowth at the maternal-fetal interface during the process of pregnancy establishment. Uterine CXCL14 expression was specifically induced at the embryo implantation site and expanded with the subsequent decidualization process in a spatiotemporal manner. The implanting embryo also showed a highlighted expression of CXCL14 in the blastocyst TE and its derived EPCs during postimplantation development. In vitro functional study revealed that CXCL14 could significantly inhibit both primary and secondary TGC attachment and outgrowth, correlated with a stage-dependant downregulation of MMP-2 and/or MMP-9 activity. Moreover, it was found that biotinylated CXCL14 could specifically bind to trophoblast cells in vitro and in vivo, suggesting trophoblast, perhaps expressing the unidentified CXCL14 receptor, is a bioactive target of CXCL14 (Kuang et al., 2009).
Furthermore, recent studies found that brain-derived neurotrophic factor (BDNF) promotes implantation and subsequent placental development by stimulating trophoblast cell growth and survival. BDNF and neurotrophin-4/5, together with their receptor, tyrosine kinase B (TrkB), are expressed in TE cells of blastocysts during implantation. Treatment with BDNF promotes blastocyst outgrowth, but not adhesion, in vitro and increased levels of the cell invasion marker MMP-9 in cultured blastocysts through the phosphatidylinositol 3-kinase pathway (Kawamura et al., 2009).
In addition, primary and secondary TGCs produce several cytokines and prolactins, which upregulate maternal blood flow to the implantation site, and promote the production of ovarian progesterone and lactogens. The latter are necessary for physiological adaptations in the mother during pregnancy (Linzer and Fisher, 1999; Cross et al., 2002). Recently, the previously assumed to be homogenous population of TGCs has been shown to consist of several distinct subsets based on expression of placental lactogens, placental specific cathepsin and location, while each of these population now appears to have a separate origin (Simmons et al., 2007).
The chorioallantoic placenta is composed of an outer trophoblast-derived epithelium, and an underlying vascular network and stroma, which are mesodermal in origin (Downs, 1998; Cross et al., 2003). By day 9.5–10 p.c., the allantois with its associated vessels fuses with the chorionic plate to form a highly folded labyrinthine layer, which provides a large surface area for gas and nutrient exchange. The labyrinthine layer divides into two syncytiotrophoblast layers that separate the fetal blood vessels from the maternal blood spaces (Figs. 2 and and33).
Three distinctive subtypes of trophoblast cells form the placenta: the innermost labyrinthine layer, intermediate spongiotrophoblast and the outermost layer of secondary TGCs, each with specialised endocrine, vascular, immunological or transport functions during gestation (Figs. 2 and and3;3; Rinkenberger et al., 1997; Cross, 2000; Cross et al., 2003). Secondary TGCs appear on day 7.5 p.c. to first mediate implantation. The spongiotrophoblast layer forms around day 9.5 p.c. as the result of the expansion and flattening of the EPC. The EPC origin of spongiotrophoblasts is supported by their expression of several EPC-restricted genes, e.g. Flt1 and Tpbpa (Cross et al., 2002).
Glycogen trophoblasts, a suggested specialised subtype of spongiotrophoblast cells, appear within the spongiotrophoblast after day 12.5 p.c., and express typical spongiotrophoblast markers (Adamson et al., 2002). Another trophoblast subtype, syncytiotrophoblast cells, constitutes the nutrient transport surface within the labyrinthine layer (Cross, 2000). By day 10 p.c., the murine placenta consists of: (from inside to outside) the chorionic plate, the branched labyrinthine layer containing maternal blood, the junctional zone of spongiotrophoblast layer, and the TGC layer (Fig. 2).
Several in vitro and gene knockout models have been developed to study the development and differentiation of murine trophoblast cells. Mouse blastocyst can grow in vitro in a serum-containing medium on a defined culture substrate such as vitronectin, fibronectin, laminin or collagen. This allows study of the role particular substrates, e.g. fibronectin, laminin and collagen; play in primary trophoblast development (Armant et al., 1986; Sutherland et al., 1988). It is possible to investigate the interaction between trophoblast cells and molecules usually present in their adhesion and migration pathway, such as within the epithelial basement membrane or stromal extracellular matrix (ECM). Such studies led to the discovery of the importance of the proteoglycans, integrins and other ECM molecules in mouse implantation (Sutherland et al., 1993; Rout et al., 2004). They also facilitated investigation of the effect of growth factors on primary trophoblast outgrowth, including Tumour Necrosis Factor-α (TNF-α; Whiteside et al., 2003), Transforming Growth Factor-α and -β (TGF-α and TGF-β), Parathyroid Hormone-related protein (PTHrP; Nowak et al., 1999), Epidermal Growth Factor (EGF), and Fibroblast Growth Factor-2 (FGF-2; Haimovici and Anderson, 1993).
In contrast to primary TGCs, studies on secondary TGC differentiation remain limited, probably due to the difficulties of isolating EPC explants free of any attached maternal or embryonic tissues. In these studies, EPC explants were isolated from murine embryos, dissected from the extraembryonic ectoderm (EXE) and attached trophoblast cells and grown in a culture medium supplemented with serum on an ECM substrate (Sutherland et al., 1993; Parast et al., 2001; Goncalves et al., 2003). We (El-Hashash and Kimber, 2004) and others (Romagnano and Babiarz, 1990) have successfully grown EPC explants in a serum-free medium to characterize and study secondary TGC differentiation. The explants attach to the ECM substrate and form outgrowths that spread as a monolayer of flattened TGCs (Fig. 4A–D). Primary/secondary TGCs are identified by their characteristic large nuclei due to DNA endoreduplication and expression of several TGC-specific genetic markers (Fig. 5).
Furthermore, advancements in understanding trophoblast differentiation have been achieved through development of model systems, e.g. the Rcho-1 and TSC lines, and improvements in knockout and knockdown technology. Homologous recombination-based gene targeting has enhanced analysis of gene function by generation of mice with defined germ-line mutations, e.g. null mutation of individual genes. This has lead to identification of the function of several trophoblast-regulating genes; e.g. Hand1, Mash-2 and AP-2γ. Moreover, site-specific recombinases and their target sequences such as the Cre/loxP system provides a tool to study gene function by inactivating genes in a tissue specific and/or inducible manner (Kuhn and Torres, 2002), where tissue specific promoter sequences are established.
The successful establishment of placentation depends critically on the orderly differentiation of TGCs, which is thought to coincide with a type of the epithelial–mesenchymal transition (EMT; reviewed in Sutherland 2003). EMT is characterized by destabilised cell–cell interactions and an increase in cell protrusive activity, followed by trophoblast reepithelisation, little or no cell motility, and stabilised cell–cell and cell–matrix interactions (Parast et al., 2001). Other characteristics of TGC differentiation include rearrangement of the cytoskeletal architecture, exit from the cell cycle, DNA endoreduplication and production of trophoblast-specific proteins such as placental prolactins.
Like other Eukaryotic cells, trophoblasts are able to adopt a variety of shapes, and carry out coordinated movements depending on their highly dynamic cytoskeleton: actin filaments, microtubules and intermediate filaments (Fig. 4E and F). Undifferentiated trophoblast cells contain little organised actin and few peripheral focal complexes, and exhibit high membrane protrusive activity in vitro (Parast et al., 2001). In differentiated TGCs, cell junctions and the actin cytoskeleton become highly organised (Bevilacqua and Abrahamsohn, 1988, 1989; Parast et al., 2001). Both stability and organisation of cell–cell interactions increase, in correlation with changes in RhoGTPase activity (Parast et al., 2001). A few factors have been identified to regulate trophoblast cellular differentiation. Insulin growth factor-1 (IGF-1) increases the formation and organisation of actin stress fibres and lamellipodia (Kabir-Salmani et al., 2002). Leukaemia inhibitory factor (LIF) regulates TGC differentiation via Janus kinase 1-signal transducer and activator of transcription-suppressor of cytokine signalling 3 pathways (Takahashi et al., 2008). Moreover, PTHrP increases actin expression and stress fibre formation and organisation in trophoblast cells in vitro (El-Hashash and Kimber, 2006).
The two distinct cell populations within the EPC, giant and diploid non-giant cells, exhibit different cellular programmes. Mitotic cycles of non-giant spongiotrophoblasts provide a continuous source of cells. In contrast, the increase of cell size, with accompanying polyploidization due to DNA endoreduplication, is a common mode of growth for TGCs in euthe-rian mammals (Ilgren, 1983; Zybina and Zybina, 1996). DNA endoreduplication is linked to TGC differentiation and has been proposed to facilitate the acceleration of cell differentiation and to increase cell growth by shortening the cell cycle and simultaneously increasing the number of chromosomes (Gardner and Davies, 1993; Goncalves et al., 2003).
Normal cell cycle regulators, e.g. Cyclins, are expressed by murine and human trophoblasts and uncoupled during DNA endoreduplication leading to a shift from the mitotic to the endocycle (Lilly and Spradling, 1996; Korgun et al., 2006). Inhibition of Cyclin B1 protein translation, together with Cyclin D1 expression, has been suggested as a mechanism for the trophoblast endocycle (Palazon et al., 1998). Trophoblast cells fail to assemble Cyclin B/p34CDK1 complexes at the time of mitotic cycle arrest during the first endocycle, and suppress Cyclin B protein expression in the subsequent endocycles (MacAuley et al., 1998). Cyclin D3 expression decreases while Cyclin D1 increases during the differentiation into TGCs in vitro (MacAuley et al., 1998). Moreover, the initiation of S phase is associated with increasing Cyclin A/E synthesis during the endocycle, while termination of S phase involves a loss of Cyclin A/E in Rcho-1 trophoblast cells (Wilhide et al., 1995; MacAuley et al., 1998). Similarly, the progression into endocycle is due to the absence of transcription for mitotic Cyclins A/B (Lehner and O’Farrell, 1990), and periodic expression of the S phase-gene Cyclin E, which is also required continuously through S phase (Lilly and Spradling, 1996) in Drosophila.
Furthermore, more critical steps in programmed endoreduplication in trophoblast cells have been discovered recently; with the finding that Cyclin-dependent kinase (CDK) activity regulates the transition from mitotic cell cycles to endocycles when TSCs differentiate into TGCs (Ullah et al., 2008). Selective inhibition of CDK1 activity, either by a chemical inhibitor called RO3306, or by developmentally programmed induction of the CDK-specific inhibitor p57 causes TSCs to endoreduplicate and differentiate into TGCs. Multiple rounds of genome duplication are then sustained by CDK2 and oscillating levels of p57. Induction of the CDK-specific inhibitor p21 serves to prevent apoptosis in cells undergoing multiple rounds of endoreduplication. Thus, endoreduplication in TSCs is triggered by p57 inhibition of CDK1 with concomitant suppression of the DNA damage response by p21 (Ullah et al., 2008).
In addition to accommodating the nutritional and growth-regulatory needs of the developing foetus, trophoblast cells are unique endocrine organ, which produce prolactin (PRL)-related and growth hormones, including prolactin like proteins (PLP-A,-C,-E,-F,-N.), proliferin (PLF), and PLF-related protein. These proteins are expressed during mid- and late gestation and have been intensively reviewed (Soares et al., 2007; Simmons et al., 2008). Among PRL hormones, placental lactogen-1 and -2 (PL-I and PL-II) are of particular interest for TGC differentiation (Figs. 4G and H and Fig. 6G and H).
The PL-II is a non-glycosylated single-chain polypeptide, and shares significant amino acid sequence homology with PL-I (Duckworth et al., 1986; Robertson et al., 1982). While PL-I synthesis occurs during early-mid pregnancy, PL-II is specifically produced by secondary TGCs and appears in the murine maternal circulation on day 9 p.c., and increases until about day 14 p.c. It then either remains relatively constant in some mouse strains or continues increasing until the end of pregnancy in others (Faria et al., 1991; Soares et al., 1982; Soares and Talamantes, 1983). PL-II enters the fetal compartment and binds to the widely expressed PRL receptor in the developing foetus (Soares and Talamantes, 1983; Ogren and Talamantes, 1988), and co-localizes with IGF-2 in murine placenta (Ishida et al., 2007).
PL-I and PL-II display a wide variety of PRL-like activities on different maternal target tissues. They promote mammary gland development, lactogenesis and maintenance of the corpus luteum and its production of progesterone (Colosi et al., 1988; Galosy and Talamantes, 1995). Several proteins regulate PL-II production, including EGF (Yamaguchi et al., 1992), interleukins-1 and -6 (IL-1, IL-6; Yamaguchi et al., 1996), and decidual-derived calcyclin (Richard et al., 1998).
In the last two decades, the purification, cloning and characterisation of several transcription factors, together with improving transgenic mouse technologies, have increased the knowledge of the genetic basis of trophoblast differentiation. Many recently discovered cell type specific transcription factors are now known to regulate trophoblast cell development. This has been extensively reviewed (Cross et al., 2002, 2003; Simmons and Cross, 2005), and a brief summary of the most investigated factors is described in Table 1.
The general consensus is that the molecular differentiation of secondary TGCs is a multistep process. This has encouraged numerous studies on the expression and function of transcription factors that have been implicated in each step (Fig. 6; Cross, 2000; Simmons and Cross, 2005). Major transcription factors, including Mash2, Eomes, cdx2, and Err2, as well as Id1, Id2 and E-factors regulate the development of TSCs of day 6.5 p.c. EXE. Differentiation of the EXE into the EPC is stimulated by the suppression of the above transcription factors except Mash2, and upregulation of Hand1 and mSNA. Downregulation of Mash2 and mSNA, together with upregulation of Stra13, may trigger the differentiation of the EPC and spongiotrophoblasts into secondary TGCs (Fig 6, Table 1).
Increasing attention has also been paid to the role of growth factors and cytokines in trophoblast development. This review will not provide a comprehensive list of growth factors/cytokines, which are required for trophoblast development because the topic is well covered elsewhere (Morrish et al., 2007; Guzeloglu-Kayisli et al., 2009). A brief summary of the effect of some of the intensively studied growth factors on trophoblast cells is shown in Table 2. The role of growth factors/cytokines in controlling trophoblast development, together with trophoblast-regulatory genes, will be discussed for TSCs and each trophoblast subtypes below.
TSCs can be isolated from murine blastocysts or at an early postimplantation stage of development in the chorion, and can give rise to all differentiated trophoblast cell subtypes (Tanaka et al., 1998; Oda et al., 2006), a process regulated by the MAPK (e.g. MAPK14; Winger et al., 2007) signalling pathway. FGF-4, which is a paracrine factor produced by the ICM/epiblast and signals through MAPK, controls trophoblast proliferation. Its cognate receptor, FGFR2, is expressed in trophoblastic tissues, including the EXE and chorion (Rossant and Cross, 2001). Survival of mouse blastocyst-derived TSCs requires FGF-4 and embryonic fibroblast-conditioned medium (supplying TGF-β and Activin-A; Erlebacher et al., 2004). In the presence of these factors, mouse TSCs exhibit sustained undifferentiated proliferation, without significant expression of the phenotypic markers of placental trophoblasts, such as PL-1/PL-II, or placental prolactin-related proteins. Removal of FGF-4 results in the arrest of cell proliferation, rapid TGC formation and onset of hormone gene transcription (Tanaka et al., 1998). Consistent with these findings, FGF-4/FGR2 mutant mice show failure in trophoblast proliferation (Feldman et al., 1995; Goldin and Papaioannou, 2003). Fgf-4 expression is regulated by the transcription factor Sox-2. If Sox-2 is knocked down, fgf-4 and fgf-r2 expression is greatly diminished, and trophectoderm does not form (Keramari M and Kimber SJ, unpublished data).
Most Recently, Abell et al. (2009) have shown that TSC maintenance by FGF-4 requires MEKK4 activation of Jun N-terminal kinase. Mice harboring a point mutation that renders inactive the mitogen-activated protein kinase kinase kinase-4 (MEKK4) exhibit dysregulated placental development with increased trophoblast invasion. Isolated MEKK4 kinase-inactive TSCs cultured under undifferentiating, self-renewing conditions in the presence of FGF-4 display increased expression of Slug, Twist, and MMP-2, loss of E-cadherin, and hyper-invasion of ECM, each a hallmark of EMT (Abell et al., 2009).
Furthermore, it has been shown that SUMO-specific protease 2 (SENP2) is essential for modulating p53-Mdm2 in development of TSC niches and lineages (Chiu et al., 2008). SENP2 modifies proteins by removing SUMO from its substrates, and is highly expressed in trophoblast cells. SENP2 null mice have a deficiency in cell cycle progression, particularly in the G–S transition, which is required for mitotic and endoreduplication cell cycles in trophoblast proliferation and differentiation, respectively. Also, SENP2 ablation disturbs the p53-Mdm2 pathway, affecting the expansion of trophoblast progenitors and their maturation (Chiu et al., 2008).
The T-box protein eomesodermin (Eomes) and the caudal-type homeodomain protein Cdx2 transcription factors are expressed in the TE and maintain TSC fate, and both Eomes and Cdx2 homozygous mutant embryos die around the time of implantation (Russ et al., 2000; Chawengsaksophak et al., 1997; Tanaka et al., 1998) A block in early TE differentiation occurs in Eomes mutant blastocysts; while in the complete absence of Cdx2, TE cell identity cannot be maintained in the blastocyst (Strumpf et al., 2005). The latter study demonstrated that Cdx2 is required for correct cell fate specification and differentiation of the TE and is essential for segregation of the ICM and TE lineages at the blastocyst stage by ensuring repression of Oct4 and Nanog in the TE (Strumpf et al., 2005). A recent study also showed that Cdx2 plays a critical role in cell polarity and in cell allocation and specification of TE and ICM in the mouse embryo by upregulating atypical PKC (aPKC) and thus controlling asymmetric cell division (Jedrusik et al., 2008). TSC differentiation into TGCs or spongiotrophoblasts is associated with suppression of Eomes and cdx2 expression (Tanaka et al., 1998).
Oct4, is a member of the POU family of transcription factors, produced by the preimplantation embryo and high in ICM and inhibits trophoblast cell fate (Palmieri et al., 1994). All blastomeres adopt a trophoblast cell fate, regardless of their position in the conceptus of Oct4 null mutants, suggesting the importance of Oct4 for maintaining of ICM fate (Nichols et al., 1998). Ets2, a regulator of cdx2, is essential for TSC self-renewal (Wen et al., 2007), while Tead4 is an early transcription factor required for specification and development of the TE lineage, which includes expression of Cdx2 (Yagi et al., 2007; Nishioka et al., 2008). In addition, Err2 transcription factor is controlled by FGF-10 signalling, and is an important regulator of TSC fate (Luo et al., 1997; Tremblay et al., 2001).
AP-2γ, an essential regulator of TSC fate controlled by FGF signalling is of particular interest for TGC differentiation, since it activates the human PL promoter (Richardson et al., 2000). It is present in all embryonic cells during pre-implantation development and, together with Eomes, is expressed in all trophoblast lineages throughout placental development (Auman et al., 2002). In postimplantation stages, AP-2γ expression is restricted to the extraembryonic lineages: primary TGCs and diploid cells of the polar TE, EXE and EPC, and increases in all the derivatives of the trophoblast lineage during chorioallantoic fusion (Sapin et al., 2000). AP-2γ knockout conceptuses show a reduced cell proliferation of the EXE and the EPC, low number of TGCs, failure in the formation of the labyrinth layer, as well as repression of the FGF-4 responsive genes: Eomes and cdx2 (Auman et al., 2002). This also leads to a significant retardation of the growth of mutant embryos by day 7.5 p.c. and their re-absorption by day 9.5 p.c. (Werling and Schorle, 2002).
Nodal, a member of TGF-β family, is expressed by the epiblast as well as later on in TGCs (Ma et al., 2001), and maintains FGF-4 expression in the epiblast. It also induces proteases (Furin and Pace-4) in the EXE, which cleave Nodal to its active form. Nodal and FGF targets in the EXE include TSC genes indicating that they co-operate to maintain TSC fate (Guzman-Ayala et al., 2004). Differentiated TGCs increase in Nodal-deficient mouse embryo, indicating its importance for TGC differentiation (Ma et al., 2001). Furthermore, FGF/Nodal signalling actively inhibits TSC differentiation into giant cells (Hughes et al., 2004). In addition, recent studies showed that Hand1 release from the nucleolus controls differentiation of TSCs into TGCs, and this switch is controlled by the antagonistic activities of the I-mfa domain-containing protein HICp40 and Polo-like kinase-4 (Plk4; Martindill et al., 2007; Martindill and Riley, 2008).
Recently, there have been efforts to directly derive TSCs from human embryos (Rossant, 2001). However, the methods developed to successfully derive TSCs from mouse embryos have not been reported to yielded comparable cells from human blastocysts. This may be due to the diverse growth factor requirements and regulatory pathways for embryonic stem cell (ESC) derivation and differentiation between mouse and human. For example, studies on nonhuman primate (Thomson et al., 1995) and human (Thomson et al., 1998) showed that whereas LIF is critical for sustaining mouse ESCs, it is ineffectual in the human system. Subsequent studies showed that the pathways that sustain undifferentiated growth in human ESCs have rely primarily on FGF-2 and TGF-β/Activin signalling, directing suppression of BMP signalling, for undifferentiated growth of ESCs (Xu et al., 2005). In the continuous presence of FGF-2 and mouse embryonic fibroblast-conditioned medium (which typically sustain undifferentiated human ESC growth), addition of BMP-4 (or related ligands for BMP receptor IB, BMP-2, -7, or GDF-5) resultes in the differentiation of human ESCs to cells of a uniform epithelial appearance (Xu et al., 2002). Thus, It is reasonable to predict that TSCs could eventually be derived from human embryos given sufficient material with which to evaluate a broad range of culture conditions; however, the development of alternative approaches becomes necessary for progress beyond the mouse model (Douglas et al., 2009).
The attributed origin of spongiotrophoblast from the EPC is largely based on the similarity in gene expression patterns, e.g. Mash-2, Tpbp/4311 and Flt1. Mash2, a basic helix-loop-helix (bHLH) transcription factor, is predominantly expressed throughout murine pre- and postimplantation stages in the EPC, spongiotrophoblasts and the chorion, but is undetectable in primary and secondary TGCs and the yolk sac (Guillemot et al., 1994; Nakayama et al., 1997). Mash-2 transiently maintains the proliferation of trophoblast cells in an FGF-4-independent manner (Hughes et al., 2004), and sustains spongiotrophoblast proliferation and survival (Guillemot et al., 1994; Cross et al., 1995; Tanaka et al., 1997). Mash-2-deficient conceptuses show premature loss of the EPC, absence of spongiotrophoblasts by E10.5 p.c., enlargement of TGC layer and a reduction in the formation of labyrinth layer (Guillemot et al., 1994; Tanaka et al., 1997). Mash-2 mutant mice can be rescued through aggregation of diploid Mash-2−/−, and tetraploid wild-type embryos and are viable and fertile, suggesting no function for Mash-2 in the developing embryo itself (Guillemot et al., 1994). Neither zygotic nor maternal Mash-2 transcripts are required for early trophoblast development, as indicated by the normal development of mutant embryos derived from Mash-2−/− females (Rossant et al., 1998).
Two bHLH factors, Alf1/HEB and Itf2, are obligate DNA binding partners for Mash-2. They are only expressed in the inner region of the EPC and chorion (Scott et al., 2000), implying that Mash-2 function is limited to the proliferative inner EPC cells that probably maintain spongiotrophoblast layers (Cross et al., 2003). The genes Tpbp/4311 (Lescisin et al., 1988) and Flt1 (He et al., 1999) are also expressed by both the EPC and spongiotrophoblast cells, but are absent in the EXE, TGCs and labyrinth. The expression levels of Tpbp/4311 and Flt1 increase with embryonic development, and their functions are still unclear. In addition, recent studies identified the putative transcriptional regulator BPTF/FAC1, which is expressed in embryonic and extra-embryonic tissues of the early mouse conceptus, as an essential regulator for the differentiation of EPC-derived trophoblast cells (Goller et al., 2008). BPTF/FAC1-deficient mouse embryos form apparently normal blastocysts that implant and develop epiblast, visceral endoderm, and the EXE including TSCs. However, the EPC is drastically reduced or absent in mutants, which may cause the embryonic lethality (Goller et al., 2008).
Furthermore, Epidermal Growth Factor Receptor (EGFR) has been recently shown to promote growth of the placental spongiotrophoblast layer in mice. Also, EGFR expressed in the uterine stroma may play an underappreciated role in preparation of the uterus for embryo implantation (Dackor et al., 2009a). Egfr homozygous placentas have a reduced spongiotrophoblast layer in some strains, while spongiotrophoblasts and glycogen cells are almost completely absent in others (Dackor et al., 2009b).
In the absence of FGF, cultured primary trophoblast cells spontaneously differentiate into giant cells, suggesting that TGC differentiation is a “default pathway” of trophoblast differentiation. Overexpression of Hand1, a bHLH transcription factor, induces giant cell differentiation in Rcho-1 cells in vitro, which is inhibited upon ectopic expression of Mash-2 (Cross et al., 1995; Scott et al., 2000). Hand1 null embryos arrest in development around day 7.5–8 p.c., and show a diminished number of primary and secondary TGCs, a reduced EPC, and the absence of PL expression (Riley et al., 1998). Hand1−/− TSCs show reduced levels of TGC differentiation (Hemberger et al., 2004).
The expression pattern and activity of Hand1 and Mash-2 have been the subject of controversy. Hand1 expression overlaps the expression of Mash-2 in the EPC and spongiotrophoblasts, and layers of the placenta containing giant cells precursors in murine postimplantation stages (Scott et al., 2000). Contrary to Mash-2, Hand1 is highly expressed in primary and secondary TGCs, but is absent from the chorion (Scott et al., 2000). Hand1 and Mash-2 show mutually antagonistic activities on trophoblast cell differentiation in vitro, where Hand1 competes with Mash-2 for binding to E-factors such as ITF-2, ALF1 or E47, and thus competes away Mash-2 function (Scott et al., 2000). Hand1 is likely to be required for turning off Mash-2 expression, since Hand1 knockout mice show persistent Mash-2 expression (Riley et al., 1998). However, Scott et al. (2000) suggested that the function of Hand1 is merely to inhibit Mash-2 activity, as the defects of Hand1 mutants are not rescued in Hand1; Mash-2 double mutant mice. Ectopic expression of Hand1 induces TGC differentiation even in the presence of FGF-4 (Hughes et al., 2004).
Groucho-related gene 3 (Grg3) is a mouse homologue of the Drosophila Groucho corepressor protein, which, together with the other Notch signalling pathway member Hairy/Enhancer of Split (homologous to mammalian Hes), is involved in developmental processes, usually by downregulation of the Hairy/Enhancer of Split downstream gene Achaete-scute (mammalian Mash). Several Notch signalling pathway members were shown to be expressed in mouse placenta, including Notch2, Hes2, Hes3, Grg2, Grg3 and Mash2 (Nakayama et al., 1997). Grg3 is expressed in secondary TGCs and in the labyrinth, but not in the spongiotrophoblast layer. Homozygous knockout mice for Grg3 dying in utero between day 13.5 p.c. and day 15.5 p.c. show marked growth retardation, which is most probably related to the observed underdevelopment of the Grg3 deficient trophoblast. In Grg3 deficient placenta, the number of secondary TGCs is greatly reduced, suggesting that Grg3 is necessary for proper differentiation into secondary TGCs (Morrish et al., 2007). The hypothetical mechanism of Grg3 action could be interaction with Hes2 and downregulation of Mash2 expression, which would inhibit differentiation into spongiotrophoblast and shift it towards secondary TGCs (Morrish et al., 2007).
Retinoic acid and its inducible bHLH factor, Stra13, stimulate TGC differentiation. RA-treated TSCs show arrested cell proliferation and enhanced giant cell differentiation (Yan et al., 2001). Stra13 is expressed by TGCs in vivo (Boudjelal et al., 1997) and increases during TGC differentiation (Hughes et al., 2004). Overexpression of Stra13 overrides FGF signalling in order to promote TSC differentiation into TGCs (Hughes et al., 2004). Similarly, GATA2 and GATA3 have been implicated in the regulation of trophoblast-specific genes. Recently, Ray et al. (2009) found that GATA3 directly represses Gata2 in undifferentiated trophoblast cells, and a switch in chromatin occupancy between GATA3 and GATA2 (GATA3/GATA2 switch) induces transcription during trophoblast differentiation. Thus, GATA3/GATA2 switch provides an important mechanism for the transcriptional regulation of other trophoblast-specific genes.
In addition, BDNF and neurotrophin-4/5 proteins as well as their receptor TrkB, which are expressed after implantation in trophoblast cells and placentas, are also essential for trophoblast cell development (Kawamura et al., 2009). Treatment of cultured trophoblast cells with the TrkB ectodomain, or a Trk receptor inhibitor K252a, suppresses cell growth as reflected by decreased proliferation and increased apoptosis. Studies using the specific inhibitors indicated the importance of the phosphatidylinositol 3-kinase/Akt pathway in mediating the action of TrkB ligands. Also, in vivo studies in pregnant mice further demonstrated that treatment with K252a suppresses placental development accompanied by increases in trophoblast cell apoptosis and decreases in placental labyrinth zone at mid-gestation (Kawamura et al., 2009).
Furthermore a recent study by Schulz et al. (2009) suggested that leptin influences TGC differentiation. Leptin accelerates disappearance of non-giant cells while inhibiting terminal differentiation of committed giant cells, possibly by maintaining cells in an intermediate stage of differentiation. In trophoblast culture, leptin stimulates the phosphorylation of MEK (MAP2K1), increases the concentration of the suppressor of cytokine signalling 3 (SOCS3) protein, and upregulates metalloproteinase activity. In addition, leptin stimulates the activity of some genes with roles in cell motility, including Stmn, a gene linked to invasiveness in other cell types, as well as several genes associated with MAPK and RhoGTPase signalling (Schulz et al., 2009).
Evidence from our laboratory that PTHrP growth factor regulates TGC differentiation has shed new light on the mechanisms by which PTHrP can regulate cell differentiation. PTHrP, a 139- to 173-amino acid multifunctional protein, is widely expressed in different cell types, in which it acts as a paracrine, autocrine or intracrine factor, signalling through its cognate PTHR1 receptor (Wysolmerski and Stewart, 1998). PTHrP acts in a stage-specific manner on the EPC precursors to induce their differentiation into secondary TGCs through several parallel processes. Firstly, PTHrP shifts the trophoblast cell cycle into endocycle (El-Hashash et al., 2005). DNA endoreduplication of TGCs, a process closely linked to the onset of their differentiation, is accompanied with upregulation of Cyclin D1 and repression of Cyclin B expression (MacAuley et al., 1998; Palazon et al., 1998). PTHrP dramatically downregulates Cyclin B1, increases Cyclin D1 expression, and inhibits the expression of mSNA, a transcription factor that represses switching from mitosis to endoreduplication and TGC differentiation (Nakayama et al., 1998; El-Hashash et al., 2005). This possible stimulation of the mitotic-endocycle transition is supported by increasing DNA synthesis, number of TGCs, and upregulation of RhoA, a stimulator of Cyclin D1 expression in mid-G1 phase and of G1 phase progression (Welsh et al., 2001) in PTHrP–treated cultures (El-Hashash et al., 2005; El-Hashash and Kimber, 2006).
Secondly, PTHrP affects critical regulators of TGC differentiation. We identified several target genes for PTHrP that have not been reported in other cell types before, e.g. Eomes, Mash-2, mSNA, AP-2γ and Stra13. PTHrP downregulates TSC-maintaining Eomes and spongiotrophoblast-specific genes Mash-2 and mSNA, and upregulates TGC differentiation-stimulating genes AP-2γ, Stra13 and PL-II (El-Hashash et al., 2005). Thus, PTHrP may regulate TGC differentiation directly from trophoblast precursors in the EPC and by terminal differentiation of intermediate diploid trophoblast populations of the spongiotrophoblast. Finally, PTHrP induces the organisation of actin fibres, probably through controlling EphB/EphrinB-RhoA signalling, resulting in the characteristic flattened, fibroblastic morphology of differentiated TGCs. This is accompanied by downregulation of E-cadherin expression, and an increase in cell spreading in culture (El-Hashash and Kimber, 2006). Several other factors regulate TGC development, including LIF, TGF-α, EGF, FGF-2 HB-EGF, FGF-7, HGF and PDGF, and are summarized in Table 2.
It is widely thought that glycogen trophoblast cells (GTCs) are a specialised subtype of spongiotrophoblast, since they express the Tpbp/4311, a gene exclusively expressed by spongiotrophoblast cells (Adamson et al., 2002). GTCs first appear within the spongiotrophoblast layer, from which they migrate into the maternal decidual tissue (Adamson et al., 2002). They specifically express Connexin31 and increase in number from day 12.5 p.c. to day 16.5 p.c. in mouse (Coan et al., 2006). Little is known about the regulators of GTC development; however, GTC number is remarkably changed in Igf2 (Lopez et al., 1996), and p57kip2 (Takahashi et al., 2000) mutant mice, suggesting direct or indirect regulation by these factors. Moreover, a recent study showed that genetic deficiency in granulocyte–macrophage colony-stimulating factor (CSF2) alters placental gene expression and GTC and giant cell abundance in mice (Sferruzzi-Perri et al., 2009). The CSF2 ligand and the CSF2 receptor alpha subunit are predominantly synthesized in the placental junctional zone. Altered placental structure in Csf2 null mice at E15 is characterized by an expanded junctional zone and by increased Cx31(+) glycogen cells and cyclin-dependent kinase inhibitor 1C (CDKN1C(+), P57(Kip2+)) giant cells, accompanied by elevated junctional zone transcription of genes controlling spongiotrophoblast and giant cell differentiation and secretory function (Ascl2, Hand1, Prl3d1, and Prl2c2; Sferruzzi-Perri et al., 2009).
In most mammalian species, a key process of placenta development is the fusion of trophoblast cells into a highly specialised, multinucleated syncytiotrophoblast layer, through which most of the materno-fetal exchanges take place. Little is known about this process; however, a recent study has suggested an essential role for syncytin-A in trophoblast cell differentiation and syncytiotrophoblast morphogenesis during placenta development (Dupressoir et al., 2009). Mouse syncytin-A and syncytin-B are two fusogenic placenta-specific murine envelope genes of retroviral origin, and have in vitro cell–cell fusion activity (Dupressoir et al., 2005). Using gene knockout approaches, Dupressoir et al. (2009) found that homozygous syncytin-A null mouse embryos die in utero between 11.5 and 13.5 days of gestation, with specific disruption of the architecture of the syncytiotrophoblast-containing labyrinth, and the trophoblast cells failing to fuse into an interhemal syncytial layer. Lack of syncytin-A-mediated trophoblast cell fusion is associated with cell overexpansion at the expense of fetal blood vessel spaces and with apoptosis, adding to the observed materno-fetal interface structural defects to provoke decreased vascularization, inhibition of placental transport, and fetal growth retardation (Dupressoir et al., 2009). Similarly, syncytin glycoproteins are expressed in human trophoblastic cells and are involved in trophoblastic cell fusion (Malassine et al., 2005, 2007; Frendo et al., 2003).
Differentiated syncytiotrophoblasts (SynTs) comprise the bulk of the trophoblast component in the labyrinthine layer after fusion of allantoic mesoderm with the basal surface of the chorion (Cross, 2000). Gcm1 is a key transcription factor, which regulates both morphogenesis and differentiation of SynTs (Anson-Cartwright et al., 2000). Gcm1-deficient mouse embryos show failure of SynT formation and villous morphogenesis is not initiated (Anson-Cartwright et al., 2000). Gcm1 expression is mainly restricted to the murine placenta starting from day 7.5 p.c. (Basyuk et al., 1999; Anson-Cartwright et al., 2000) and is maintained by unknown signals from the allantois (Hunter et al., 1999). Gcm1 and Syncytin-A, a gene expressed in SynTs, stimulates TSC differentiation into SynTs, but not toward TGCs (Anson-Cartwright et al., 2000; Gong et al., 2007). Recently, the RNA-binding protein HuR has been found to control the differentiation of SynTs. HuR’s absence impairs the invagination of allantoic capillaries into the chorionic trophoblast layer and the differentiation of SynTs that controls the morphogenesis and vascularization of the placental labyrinth and fetal support. (Katsanou et al., 2009). Other genes, including Esx1 (Li and Behringer, 1998), Dlx3 (Morasso et al., 1999) and Pparγ (Barak et al., 1999) are also expressed exclusively in the labyrinthine layer of mature placenta and are critical for its development. The regulation and function of these genes are still unknown.
Like other trophoblast subtypes, SynT differentiation and branching of chorionic villi are regulated by intercellular signalling, including FGF/Fgfr, HGF/Met, EGF/Egfr and Wnt/Fzd (Cross et al., 2003; Cross, 2005). A hypomorphic mutation in Fgfr2 results in defects of chorioallantoic branching (Xu et al., 1998). Likewise, mutant mice for Hgf (Uehara et al., 1995) or its c-Met receptor (Bladt et al., 1995), Fra1, JunB or genes encoding signalling adaptor proteins (Grb2, Mek1, p38, Sos1; downstream targets of receptor tyrosine kinases; Rossant and Cross, 2001) show a similar phenotype. Recently, Nadeau et al. (2009) found that Map2k1 and Map2k2, which encode dual-specificity kinases responsible for ERK/MAP kinase activation that is essential for growth factor signalling, are also required for normal development of syncytiotrophoblasts during placentation. Map2k2 haploinsufficiency affects the normal development of placenta in the absence of one Map2k1 allele. Most Map2k1+/− Map2k2+/− double embryos die during gestation because of placenta defects restricted to extra-embryonic tissues. Moreover, the deletion of one or both Map2k2 alleles in the context of one null Map2k1 allele leads to the formation of multinucleated trophoblast giant (MTG) cells. MTG cells are derived from Gcm1-expressing SynTs, which are affected in their ability to form the uniform SynT layer II lining the maternal sinuses (Nadeau et al., 2009). In addition, Wnt signalling is required for chorioallantoic attachment and labyrinth development. A failure in chorioallantoic attachment is reported in Wnt7b−/− mouse embryos (Parr et al., 2001). A gene encoding a Wnt receptor, Fzd5, is expressed in labyrinth trophoblasts and is essential for placental angiogenesis and yolk sac formation (Ishikawa et al., 2001), while allantois-derived Wnt2 is required later in labyrinth development (Monkley et al., 1996).
Furthermore, studies by Dackor et al. (2007, 2009a,2009b) highlighted the importance of EGFR, which is a member of the ERBB family of receptor tyrosine kinases, in placental development. In mice that are Egfr(tm1Mag) nullizygous, placentas have fewer proliferative trophoblasts than wild-type and exhibit strain-specific defects in the spongiotrophoblast and labyrinth layers (Dackor et al., 2007, 2009a). Moreover, increasing the levels of EGFR signalling by using hypermorphic Egfr(Dsk5) allele results in larger placental size with a more prominent spongiotrophoblast layer and increased expression of glycogen cell-specific genes (Dackor et al., 2009b). This study also demonstrated that mice with increased levels of EGFR signalling exhibit an extensive level of genetic background-dependent phenotypic variability, and EGFR expressed in the uterine stroma may play an underappreciated role in preparation of the uterus for embryo implantation (Dackor et al., 2009b).
Recent studies showed that several members of the Notch signalling pathway, as well as the potential Notch target Mash2, are necessary for the proper development of a functional placenta. Notch2 regulates maternal blood sinuses formation; and Notch1, Notch1/4, Hey1/2, Dll4 and RBPJκ play important roles in vascular morphogenesis of the placenta. Additionally there are several reports describing an expression of other Notch pathway members: Dll1, Hes1, Tle1–3, in the placenta. However, the molecular processes by which members of the Notch pathway exert their roles are still not well understood (Gasperowicz and Otto, 2008).
Rapid progress in understanding placental development and its regulatory molecules has been achieved in the last decade. However, much more work to complete the analysis of both the molecular and cellular differentiation of TGCs remains to be done. Our current understanding of the process of trophoblast differentiation confirms that it has several aspects that are highly relevant to other morphogenetic and differentiation events. This is clearly evidenced by recent findings that trophoblast differentiation involves several molecular and cellular mechanisms that have been discovered during the differentiation of other embryonic tissues. This includes, among others, Hand1 expression, which is important for TGC and cardiac differentiation, and Eomes expression that is common between trophoblast and mesodermal cells. In addition, an analysis with a focus on placenta development of already existing Notch pathway mutants as well as the generation of mouse strains carrying targeted mutations in further members of the pathway is awaited for a better understanding of the Notch pathway function in placentation.
Several other general conclusions have been revealed from recent accumulated evidence of placenta and trophoblast development. Placental development, as does development of other embryonic organs, progresses through several steps, and several key regulators of these steps have now been identified. However, further work will be needed to discover up-stream and downstream genes involved in each step. Another exciting aspect of murine placenta regulatory genes is the discovery of human homologues of several of them. This will allow scientists to explore whether human placental pathologies are associated with changes in the expression of these genes. Further studies are required to understand the crosstalk between the epiblast and the developing trophoblast population during early trophoblast differentiation and the signalling exchange between the maternal system and the different cell types of the developing placenta. In addition, linking the molecular and cellular mechanisms of trophoblast differentiation will not only help in fully understanding trophoblast differentiation but also will facilitate discovery of critical changes that lead to developmental defects in the placenta.
This work was supported by the University of Manchester Research Scholarship (URS) and Overseas Research Scholarship (ORS) awards to AHKE.