Diverse subtypes and developmental origins of trophoblast giant cells in the mouse placenta

Developmental Biology 304 (2007) 567 – 578 www.elsevier.com/locate/ydbio

Diverse subtypes and developmental origins of trophoblast giant cells in the mouse placenta David G. Simmons a , Amanda L. Fortier b , James C. Cross a,⁎ a

Department of Biochemistry and Molecular Biology, University of Calgary, Faculty of Medicine, HSC Room 2279, 3330 Hospital Drive, N.W., Calgary, Alberta, Canada T2N 4N1 , b Department of Human Genetics, Montreal Children s Hospital Research Institute, Montreal, Quebec, Canada H3Z 2Z3 Received for publication 28 September 2006; revised 19 December 2006; accepted 4 January 2007 Available online 10 January 2007

Abstract Trophoblast giant cells (TGCs) are the first terminally differentiated subtype to form in the trophoblast cell lineage in rodents. In addition to mediating implantation, they are the main endocrine cells of the placenta, producing several hormones which regulate the maternal endocrine and immune systems and promote maternal blood flow to the implantation site. Generally considered a homogeneous population, TGCs have been identified by their expression of genes encoding placental lactogen 1 or proliferin. In the present study, we have identified a number of TGC subtypes, based on morphology and molecular criteria and demonstrated a previously underappreciated diversity of TGCs. In addition to TGCs that surround the implantation site and form the interface with the maternal deciduas, we demonstrate at least three other unique TGC subtypes: spiral artery-associated TGCs, maternal blood canal-associated TGCs and a TGC within the sinusoidal spaces of the labyrinth layer of the placenta. All four TGC subtypes could be identified based on the expression patterns of four genes: Pl1, Pl2, Plf (encoded by genes of the prolactin/prolactin-like protein/placental lactogen gene locus), and Ctsq (from a placental-specific cathepsin gene locus). Each of these subtypes was detected in differentiated trophoblast stem cell cultures and can be differentially regulated; treatment with retinoic acid induces Pl1/Plf + TGCs preferentially. Furthermore, cell lineage tracing studies indicated unique origins for different TGC subtypes, in contrast with previous suggestions that secondary TGCs all arise from Tpbpa+ ectoplacental cone precursors. © 2007 Elsevier Inc. All rights reserved. Keywords: Murine placenta; Trophoblast; Trophoblast giant cell; Retinoic acid

Introduction The placenta is composed of a variety of differentiated epithelial cell types (trophoblast) each having specialized functions during pregnancy. The trophoblast lineage is best understood in mice and is derived from the trophectoderm, which forms at the blastocyst stage of development. The trophectoderm forms an outer shell of cells surrounding the inner cell mass. After implantation, trophectoderm cells not in direct contact with the inner cell mass, the mural trophectoderm, stop dividing and differentiate to form trophoblast giant cells (TGCs) which line the implantation chamber, anastomosing to form a diffuse network of blood sinuses for the early transport and exchange of nutrients and endocrine signals (Bevilacqua ⁎ Corresponding author. Fax: +1 403 270 0737. E-mail address: [email protected] (J.C. Cross). 0012-1606/$ - see front matter © 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.ydbio.2007.01.009

and Abrahamsohn, 1988). In contrast, the trophectoderm immediately overlying the inner cell mass, the polar trophectoderm, continues to proliferate and gives rise to all the remaining trophoblast cell types of the placenta (Cross et al., 1994), including spongiotrophoblast, glycogen trophoblast cells, several labyrinth trophoblast cell types, and a later wave of TGCs (called ‘secondary’ to distinguish them from the initial ‘primary’ group) (Simmons and Cross, 2005). Secondary TGCs are thought to derive from the differentiation of ectoplacental cone precursors, since cultured trophoblast stem cells first pass through a progenitor phase characterized by the expression of Mash2 and Tpbpa (markers of ectoplacental cone) before expressing TGC-specific genes (Carney et al., 1993; Scott et al., 2000). TGCs are large, highly polyploid cells that form through the process of endoreduplication (Gardner and Davies, 1993; MacAuley et al., 1998). They are inherently invasive and

568

D.G. Simmons et al. / Developmental Biology 304 (2007) 567–578

phagocytic and first mediate the invasion of the implanted conceptus into the maternal decidua (Zybina et al., 2000). In addition, they produce paracrine and endocrine factors including angiogenic factors (Lee et al., 1988; Carney et al., 1993; Voss et al., 2000), vasodilators (Yotsumoto et al., 1998; Gagioti et al., 2000), and anticoagulants (Weiler-Guettler et al., 1996), presumably to promote blood flow to the implantation site, luteotrophic and lactogenic hormones to influence maternal responses to pregnancy (Linzer and Fisher, 1999; Cross et al., 2002), and interferon that affects the differentiation of decidual stromal cells (Bany and Cross, 2006). Terminally differentiated TGCs express several members of the prolactin/placental lactogen gene family (Wiemers et al., 2003), including placental lactogen 1 (Pl1/Csh1), placental lactogen 2 (Pl2/Csh2), and proliferin (Plf) (Lee et al., 1988; Faria et al., 1990, 1991; Yamaguchi et al., 1992, 1994; Carney et al., 1993; Hamlin et al., 1994), genes widely used as markers of TGCs. In addition, TGCs also express members of the cathepsin gene family that are clustered on mouse chromosome 13 (Hemberger et al., 2000; Ishida et al., 2004) which, like the prolactin/placental lactogen gene locus (Wiemers et al., 2003), represents a rodent-specific expansion of genes that appear to be exclusively expressed in the placenta (Deussing et al., 2002; Sol-Church et al., 2002). Differentiation of TGCs is often considered a “default pathway” as factors which promote trophoblast stem cell selfrenewal such as FGF, Nodal, Activin, and TGFβ actively suppress TGC formation and withdrawal of these factors results in rapid differentiation into TGCs (Tanaka et al., 1998; Erlebacher et al., 2004; Guzman-Ayala et al., 2004; Hughes et al., 2004). Despite this, numerous exogenous factors such as retinoic acid (RA) (Yan et al., 2001; Hemberger et al., 2004), diethylstilbestrol (DES) (Tremblay et al., 2001), parathyroid hormone-related protein (PTHrP) (El-Hashash et al., 2005), and nerve growth factor (NGF) (Kanai-Azuma et al., 1997) have all been shown to promote TGC formation. Whether these paracrine or endocrine factors are required in vivo is unclear. The Hand1 transcription factor is required for TGC differentiation as Hand1 deficient conceptuses die between embryonic day (E) 7.5 and 8.5 due to a block in TGC formation (Riley et al., 1998; Scott et al., 2000). Hand1 appears to be required for both primary and secondary TGC differentiation since there are fewer trophoblast cells lining the implantation site and the ones that are present are smaller than normal and have reduced expression of TGC-specific genes (Scott et al., 2000). In vitro, Hand1 mutant trophoblast cells are also less invasive (Hemberger et al., 2004). Recent observations indicate that TGCs may not be a homogeneous population. For example, detailed studies of the uterine vasculature during gestation identified Plf-positive/Pl1negative trophoblast cells which have replaced the maternal endothelial cells in the lumen of the spiral arteries (Adamson et al., 2002; Cross et al., 2002; Hemberger et al., 2003). Interestingly, the endovascular trophoblast cells express only Plf and not Pl1, even at a time when Pl1 is still strongly expressed in TGCs at the barrier between the decidua and ectoplacental cone, indicating that there may be more than one type of TGC. In further support of this idea, there have been several reports of

cells that express Pl2, as well as several cathepsin genes, within the labyrinth layer of the placenta during the second half of gestation (Campbell et al., 1989; Deb et al., 1991; Dai et al., 2000; Sahgal et al., 2000; Lee et al., 2003; Ishida et al., 2004; Simmons and Cross, 2005). These cells appear to be different than the TGCs lining the implantation site and invading spiral arteries because, while they express Pl2, they never express Pl1 or Plf. Together these observations suggest the existence of several TGCs subtypes within the murine placenta. In the course of further studying the mouse placenta in detail, we have indeed characterized new subtypes of TGCs. Based on characteristic cell morphologies, ploidy analysis, localization, and differential expression of genes (Pl1, Pl2, Plf, and Ctsq), we have distinguished four TGC subtypes both in vivo and in vitro. Using cell lineage tracing, we have found that the TGC subtypes have different developmental origins. Materials and methods Animals For wild-type conceptuses, CD1 mice (Charles River) were crossed and pregnant females dissected at E8.5 through E18.5 (noon of the day of vaginal plug is designated E0.5). For cell lineage tracing studies, Tpbpa-Cre-GFP female mice (Fortier et al., submitted for publication) were crossed with male Z/AP mice (Lobe et al., 1999) and pregnant females were dissected at E14.5. Briefly, Tpbpa-Cre-GFP mice were constructed by placing a Cre recombinaseIRES-EGFP cassette downstream of 5.4 kb of the Tpbpa promoter (Calzonetti et al., 1995). The construct was then used to generate transgenic mice via pronuclear injection. Three lines were generated which produced similar temporal and spatial expression of Cre-EGFP in accordance with endogenous Tpbpa. Line 5 was used for our studies because it displayed the earliest onset of expression (∼ E8.0) which coincides with the onset of endogenous expression (Calzonetti et al., 1995). Expression at this time is restricted to the tip of the ectoplacental cone but later is detectable throughout the spongiotrophoblast. Cre recombinase activity was also detectable as early as E8 (Fortier et al., submitted for publication). All animals were housed under normal light conditions (12 h light/12 h dark) with free access to food and water. All animal procedures were carried out in accordance with the University of Calgary Animal Care Committee.

Cell culture Trophoblast stem (TS) cells (tgRs26 provided by Tilo Kunath and Janet Rossant) were grown as previously described (Tanaka et al., 1998; Hughes et al., 2004) at 37 °C under 95% air and 5% CO2. TS cell medium contained 20% fetal bovine serum (CanSera), 1 mM sodium pyruvate (Invitrogen), 50 μg/ml penicillin/streptomycin, 5.5 × 10− 5 M β-mercaptoethanol (Invitrogen), 25 ng/ml basic fibroblast growth factor (Sigma), and 1 μg/ml heparin (Fisher Scientific) in RPMI 1640 (Invitrogen), with 70% of the medium being preconditioned by incubating on embryonic fibroblasts for 48 h. Differentiating medium consisted of the TS medium but without basic fibroblast growth factor, heparin, or embryonic fibroblast preconditioning. Hand1 heterozygous and homozygous mutant TS cells were cultured under the same conditions as tgRs26 TS cells and their generation has been previously described (Hemberger et al., 2004). In some experiments, retinoic acid (RA, Sigma) was dissolved in ethanol and added to TS medium to a final concentration of 5 μM (RA), ethanol only was used as a control.

RNA isolation and northern blot analysis Total RNA was collected from ectoplacental cone plus chorion tissues at E8.5 and E9.5 and placentas from E10.5 through E18.5 by homogenization in Trizol Reagent (Invitrogen), following the manufacturer's instructions. Total

D.G. Simmons et al. / Developmental Biology 304 (2007) 567–578 RNA from TS cell cultures was isolated using one QIAshredder and RNeasy column (Qiagen) per well of a six-well plate following the manufacturer's instructions. Ten micrograms of total RNA was separated on a 1.1% formaldehyde agarose gel, blotted onto GeneScreen nylon membrane (Perkin Elmer), and UV cross-linked. Random-primed DNA labeling of cDNA probes for Pl1, Pl2, Plf, and Ctsq was carried out with 25 μCi 32P-dCTP and isolated on Sephadex G-50 columns (Amersham Biosciences). Hybridizations were done at 60 °C overnight in hybridization buffer as previously described (Church and Gilbert, 1984). Signals were detected by exposure to BioMax MR film (Kodak) at − 80 °C.

Tissue preparation and in situ hybridization Implantation sites (E9.0) and placentas (E14.5) were dissected in cold phosphate buffered saline (PBS) and fixed overnight in 4% paraformaldehyde (PFA) in PBS at 4 °C. After 3× rinsing in PBS, tissues went through graded sucrose solutions (10% in PBS and 25% in PBS) before being embedded in OCT (Tissue Tek). Ten micron sections were cut on a cryostat (Leica), mounted on Super Frost Plus slides (VWR), and stored at −80 °C. For in situ hybridization, sections were re-hydrated in PBS, post-fixed in 4% PFA for 10 min, treated with proteinase K (15 μg/ml for 5 min at room temperature for E9.0 implantation sites and 30 μg/ml for 10 min at room temperature for E14.5 placentas), acetylated for 10 min (acetic anhydride, 0.25%; Sigma), and hybridized with digoxigeninlabeled probes overnight at 65 °C. Digoxigenin labeling was done according to the manufacturers instructions (Roche). Hybridization buffer contained 1× salts (200 mM sodium choride, 13 mM tris, 5 mM sodium phosphate monobasic, 5 mM sodium phosphate dibasic, 5 mM EDTA), 50% formamide, 10% dextran sulfate, 1 mg/ml yeast tRNA (Roche), 1× Denhardt's (1% w/v bovine serum albumin, 1% w/v Ficoll, 1% w/v polyvinylpyrrolidone), and DIG-labeled probe (final dilution of 1:2000 from reaction with 1 μg template DNA). Two 65 °C post-hybridization washes were carried out (1× SSC, 50% formamide, 0.1% tween-20) followed by two RT washes in 1× MABT (150 mM sodium chloride, 100 mM maleic acid, 0.1% tween-20, pH 7.5), and 30 min RNAse treatment (400 mM sodium chloride, 10 mM tris pH7.5, 5 mM EDTA, 20 μg/ml RNAse A). Sections were blocked in 1× MABT, 2% blocking reagent (Roche), 20% heat inactivated goat serum for 1 h, and incubated overnight in block with antiDIG antibody (Roche) at a 1:2500 dilution. After four 20 min washes in 1× MABT, slides were rinsed in 1× NTMT (100 mM NaCl, 50 mM MgCl, 100 mM tris pH 9.5, 0.1% tween-20) and incubated in NBT/BCIP in NTMT according to the manufacturer's instructions (Promega). Slides were counterstained with nuclear fast red, dehydrated and cleared in xylene, and mounted in cytoseal mounting medium (VWR). For double in situ experiments on TS cell cultures, the same protocol was carried out with minor modifications. Cells were fixed in 4% PFA for 15 min and were not proteinase K treated. Also, probes were labeled with fluorescein as well as digoxigenin according to the manufacturer's instructions and added to the hybridization mix. After the digoxigenin probe was detected with NBT/BCIP, the enzyme was inactivated with heating to 65 °C for 30 min in 1× MABT, followed by 30 min incubation in 0.1 M glycine pH 2.2, blocked for 1 h in blocking solution and incubated overnight at 4 °C with anti-fluorescein antibody (Roche, 1:2500). Fluorescein probes were detected by incubation with INT/ BCIP (Roche) until a brown precipitate formed. Reactions were stopped in PBS, counterstained with nuclear fast red, and mounted under 50% glycerol/PBS for microscopy.

β-Galactosidase and alkaline phosphatase staining β-Galactosidase and alkaline phosphatase staining was carried out as previously described (Lobe et al., 1999). Briefly, placentas were fixed in 2% PFA/0.2% glutaraldehyde for 4 h with rocking (bisected after the first hour) and processed through a series of sucrose gradients and embedded in OCT (Tissue Tek). Ten microgram cryosections were then stained for either β-galactosidase or alkaline phosphatase activity using NBT/BCIP (Roche) as a substrate for β-galactosidase or INT/BCIP (Roche) as a substrate for human alkaline phosphatase as previously described (Lobe et al., 1999). For double staining on the same section, β-galactosidase staining was done first because detection of human alkaline phosphatase activity driven from the transgene required a 30 min heat inactivation of endogenous alkaline phosphatase activity, and this step

569

inactivates β-galactosidase activity. Counterstaining was done with nuclear fast red and mounting was done using 50% glycerol/PBS.

Nuclear area and estimation of DNA content Fluorescent in situ hybridizations were done on 10 μm cryosections of E14.5 placentas with probes for Plf and Pl2 as described above with the following modifications. Detection of the digoxigenin labeled probes was done with horseradish peroxidase conjugated anti-DIG antibodies (Roche, 1:2500) followed by amplification with fluorescein or TMR conjugated tyramide (Perkin Elmer). Nuclei were counterstained with DAPI. Nuclear area and DNA content of Plf + or Pl2+ TGCs were estimated using OpenLab software (Improvision). As a control (representing 2C nuclei), Plf −/Pl2− endothelial cell nuclei or syncytiotrophoblast nuclei were also quantified. Statistical significance was evaluated using a one-way ANOVA.

Results Expression of genes from the prolactin/placental lactogen and cathepsin gene families can distinguish TGC subtypes In the course of using various molecular markers to characterize the different trophoblast cell subtypes within the mouse placenta, we focused on four that emerged as important ones for distinguishing different cell phenotypes (Pl1, Pl2, Plf, and Ctsq) (Fig. 1A). Plf expression was detected early in gestation, beginning at E6.5, in TGCs that line the wall of the implantation site separating the ectoplacental cone from the maternal decidua, herein referred to as “parietal TGCs” or P-TGCs (Fig. 1B, data not shown). Later in gestation, Plf expression was also detected in the trophoblast cells lining the spiral arteries (SpA-TGCs) and canal spaces (C-TGCs) (Figs. 1F, J, N), which result from the coalescence of the spiral arteries (Fig. 1A). Plf expression was not detectable within the labyrinth layer of the placenta (Fig. 1R) although scattered staining could be seen in the spongiotrophoblast layer (Fig. 1F). Similar to Plf, Pl1 expression could be detected early in gestation in P-TGCs (Fig. 1C) and was absent from the labyrinth (Fig. 1S). However, unlike Plf, Pl1 expression was not detected in SpA-TGCs (Fig. 1K) or C-TGCs (Fig. 1O) and expression was not maintained in P-TGCs past E10.5 (Fig. 2). Pl2 expression was not detected until ∼ E9.5 and was observed in P-TGCs and spongiotrophoblast (Figs. 1D and 2, data not shown). Pl2 expression was not detected in SpA-TGCs (Fig. 1L). In the mature placenta, Pl2 expression was detected in P-TGCs, spongiotrophoblast (Fig. 1H), C-TGCs (Fig. 1L), and cells within the labyrinth layer (Fig. 1T). Pl2 expression within the labyrinth was restricted to the mononuclear trophoblast cells which line the maternal sinusoids (Fig. 1T) (Simmons and Cross, 2005). Ctsq was expressed exclusively in these Pl2+ mononuclear trophoblast cells lining the maternal sinusoids (Figs. 1E, I, M, Q, U), cells we will refer to as sinusoidal TGCs (S-TGCs). Ctsq expression was not detectable before ∼ E11.5 (Figs. 1E and 2) but was evident until the end of gestation (Fig. 2). Based on these observations, four distinct TGC subtypes could be identified using gene expression patterns and spatial location; Pl1 and Ctsq exclusively marked P-TGCs and S-TGCs, respectively. SpA-TGCs were identified by expression of Plf alone and C-TGCs by expression of both Plf and Pl2, but not Pl1 or Ctsq.

570

D.G. Simmons et al. / Developmental Biology 304 (2007) 567–578

Fig. 1. Expression of genes from the prolactin and cathepsin gene families define subsets of trophoblast giant cells. (A) Diagram depicting the structures and cell types of the murine placenta. (B–U) In situ hybridization for Pl1 (B, F, J, N, R), Pl2 (C, G, K, O, S), Plf (D, H, L, P, T), and Ctsq (E, I, M, Q, U) on E9.0 implantation sites (B–E) and E14.5 placentas (F–U). TGCs lining the implantation site, separating the maternal decidua from the ectoplacental cone (and later the spongiotrophoblast layer), P-TGCs (B-E), express Pl1, Plf, and Pl2 (after E9.5, not shown), but not Ctsq. Spiral artery-associated TGCs, SpA-TGCs, express Plf exclusively. Maternal canal-associated TGCs, C-TGCs, express both Pl2 and Plf but do not express Pl1 or Ctsq. TGCs within the labyrinth layer, S-TGC, express both Pl2 and Ctsq but not Pl1 or Plf. C—canal lumen, SpA—spiral artery lumen, *—SpA-TGC, black arrow head—C-TGC, black arrows—S-TGC.

D.G. Simmons et al. / Developmental Biology 304 (2007) 567–578

571

in differentiating TS cell cultures as well, although triple labeling would be required to definitively identify them. Hand1 regulates the formation of all TGC subtypes

Fig. 2. Temporal expression of trophoblast giant cell subtype markers. Northern blot analysis was performed using probes for Pl1, Pl2, Plf, and Ctsq and RNA from isolated ectoplacental cone/chorion tissue (E8.5–9.5) and placentas (E10.5–18.5).

All TGC subtypes are mononuclear but polyploid TGCs were first defined as large cells with high ploidy as a result of endoreduplication (Gardner et al., 1973; Ilgren, 1983; Zybina and Zybina, 1996). Therefore, nuclear size and relative DNA content of the different TGC subtypes were measured and compared with diploid control cells (endothelial and syncytiotrophoblast cells of the labyrinth) in sections of E14.5 placenta. Fluorescent in situ hybridization for Plf and Pl2 was used to identify each TGC subtype based on a positive signal and spatial location, and DAPI staining was used to visualize their nuclear size and estimate DNA content (Fig. 3C). The nuclear area of all TGC subtypes was significantly larger than that of diploid controls (Fig. 3A, p < 0.001). In addition, all TGC subtypes had significantly higher estimated DNA content than diploid control cells (Fig. 3B, p < 0.001). Notably, P-TGCs had larger nuclei and a higher DNA content than the other three TGC subtypes (Figs. 3A, B, p < 0.001).

Previous studies have indicated a critical role for the bHLH transcription factor Hand1 during TGC differentiation (Riley et al., 1998; Scott et al., 2000; Hughes et al., 2004). However, the analysis to date has been limited to what we now define as PTGCs because Hand1 mutants die before the time that SpATGCs, C-TGCs, and S-TGCs develop. In situ hybridization demonstrated that Hand1 was expressed in all four TGC subtypes (Figs. 5C–F). The one exception is that Hand1 expression was not detected in SpA-TGCs that had invaded furthest up into the spiral arteries, cells that do still express Plf, implying that expression is lost by the distal end of their invasion at least at E14.5 (Figs. 5A, B). To explore the functional significance of Hand1, we assessed the expression of Plf, Pl1, Pl2, and Ctsq in Hand1 mutant TS cells. Northern blot analysis of differentiating Hand1−/− TS cells showed reduced expression

All TGC subtypes appear during differentiation of trophoblast stem cell cultures To investigate whether all four TGC subtypes appear during differentiation of trophoblast stem cell (TS) cultures, we first analyzed the expression of Plf, Pl1, Pl2, and Ctsq by northern blotting. Expression of all four genes was undetectable in stem cells and during early differentiation. Strong expression was detected by day 6 of differentiation (Fig. 4A). Based on in vivo expression patterns, Pl1 and Ctsq expression indicated the presence of P-TGCs and S-TGCs, respectively. In situ hybridization confirmed Pl1 as well as Ctsq expressing TGCs in day 6 differentiated cultures (Figs. 4B, D, E). In situ hybridizations with single probes for Pl1, Pl2, Plf, or Ctsq never showed staining in more than ∼ 40% of TGCs, also suggesting that TGCs are heterogeneous in differentiating TS cell cultures. In addition, Ctsq expressing TGCs were less abundant in day 6 differentiated cultures than TGCs expressing Pl1, Pl2, or Plf. To determine whether SpA-TGC or C-TGC was present in the differentiating TS cell cultures, colocalization was required. Dual color in situ hybridization identified TGCs which were Plf +/Pl− (Fig. 4D), Plf +/Pl2− (Fig. 4C), and Pl2+/Plf + (Fig. 4C). Together these observations indicated that SpA-TGCs and C-TGCs are most likely present

Fig. 3. Trophoblast giant cell subtypes are polyploid. (A) SpA-TGCs, C-TGCs, and STGCs have significantly larger nuclei at E14.5 than endothelial or syncytiotrophoblast cells (p < 0.001). P-TGCs have significantly larger nuclei than SpA-TGC, C-TGCs, STGCs, and endothelial or syncytiotrophoblast cells (p < 0.001). (B) SpA-TGCs, C-TGCs, and S-TGCs have significantly higher DNA content at E14.5 than endothelial or syncytiotrophoblast cells (p < 0.001). P-TGCs have significantly higher DNA content than SpA-TGC, C-TGCs, STGCs, and endothelial or syncytiotrophoblast cells (p < 0.001). DAPI intensity measured in arbitrary units (AU). (C) Representative florescent in situ hybridization of Plf (green) or Pl2 (red), indicating Plf (+) P-TGC, Plf (+) SpA-TGC, Pl2 (+) C-TGC and Pl2 (+) S-TGC. Scale bar = 0.1 mm.

572

D.G. Simmons et al. / Developmental Biology 304 (2007) 567–578

(in the absence of FGF4 and feeder cell conditioned medium), exposure to RA induced Pl1 and Plf expression but suppressed the expression of Pl2 and Ctsq (Fig. 6). Together, these data suggested that P-TGCs are preferentially induced by RA but that S-TGCs are suppressed. Not all TGC subtypes arise from Tpbpa+ precursors

Fig. 4. Trophoblast giant cell subtypes appear in differentiating TS cell cultures. (A) Northern blot analysis of Pl1, Pl2, Plf, and Ctsq mRNA expression in TS cells over 6 days of culture in the presence (proliferating) or absence (differentiating) of FGF, heparin, and fibroblast conditioned medium (CM). (B) In situ hybridization of TGC subtypes in differentiating TS cell cultures (day 6 of differentiation in the absence of FGF/CM). Double in situ hybridizations indicate Pl1+ TGCs (B, D), Pl2+ TGCs (B, C), Plf+ TGCs (C, D), Pl1/Pl2+ TGCs (B), Pl1/Plf+ TGCs (C), and Pl2/Plf+ TGCs (D). Single in situ hybridization indicates the expression of Ctsq+ TGCs (E). Scale bar = 0.1 mm.

It has been hypothesized that differentiation of secondary TGCs involves intermediate steps, with cells passing through a progenitor phase characterized by the expression of Mash2 and Tpbpa before terminal differentiation (Carney et al., 1993; Scott et al., 2000) and that these precursors reside within the ectoplacental cone (and later the spongiotrophoblast layer). To determine if all TGC subtypes originate from Tpbpa + precursors, we took advantage of a transgenic mouse driving Cre recombinase under the control of the Tpbpa promoter. By crossing female Tpbpa-Cre mice with male Z/AP dual reporter mice (Lobe et al., 1999), Cre induced the removal of the LacZ gene encoding β-galactosidase, and therefore the subsequent activation of human alkaline phosphatase (hAP), in all cells that had previously or continued to express Tpbpa. Staining for hAP activity therefore identified cells which originated from Tpbpa+ precursors and β-galactosidase activity indicates those cells which did not (Fig. 7A). Examination of E14.5 placentas indicated that none of the S-TGCs in the labyrinth layer expressed hAP (Figs. 7B, F, J, N and 8). In contrast, virtually all SpA-TGCs observed were positive for hAP (Figs. 7D, H, L, P and 8). Interestingly, the origins of P-TGCs (Figs. 7C, G, K, O) and C-TGCs (Figs. 7E, I, M, Q) were mixed, with approximately half of each subtype originating from Tpbpa+ precursors (Fig. 8). Based on these data, TGC subtypes had different developmental origins. In addition, TGCs of the same subtype could also have different developmental origins. Discussion

of all TGC marker genes compared to control (Hand1+/−) cells (Fig. 5G), indicating that loss of Hand1 impairs the formation of all TGCs regardless of subtype. In addition, the expression patterns for the trophoblast stem cell markers Esrrb and Eomes (Fig. 5G) were not altered in differentiating Hand1 mutant TS cells when compared with control. Retinoic acid has differential effects on TGC subtype differentiation Retinoic acid (RA) has been shown to promote differentiation of TS cells into Pl1-expressing TGCs (Yan et al., 2001; Hemberger et al., 2004). We examined the expression of other TGC marker genes and, strikingly, found that RA did not induce all genes uniformly. Instead, RA-induced Pl1 and Plf only (indicative of P-TGCs), indicating that S-TGCs and C-TGCs were not present in these cultures (Fig. 6). As expected, RA treatment induced TGC formation even under proliferating conditions (in the presence of FGF4 and feeder cell conditioned medium). Interestingly, under differentiating culture conditions

While a few studies have previously described heterogeneity in gene expression within TGCs (Weiler-Guettler et al., 1996; Hunter et al., 1999; Hemberger et al., 2000; Ma and Linzer, 2000), most have generally regarded TGCs as a homogeneous single cell population that lines the implantation site. These cells, what we now call parietal TGCs (P-TGCs), form the barrier between the maternal decidua and the ectoplacental cone and later the spongiotrophoblast layer. We have demonstrated the existence of additional TGC subtypes within the mature placenta. The four different TGC subtypes we have described are all postmitotic and polyploid, but can be distinguished by differential location, gene expression, cell lineage origins, and regulation of differentiation (Fig. 9). Endovascular trophoblast cells have previously been implied based on the expression of Plf (Adamson et al., 2002; Hemberger et al., 2003), but we have expanded this to now show that these cells are indeed polyploid. Pl2 expressing cells have been identified in the mouse and rat labyrinth layer, but they have not previously been confirmed to be TGCs (Campbell et al., 1989; Deb et al., 1991;

D.G. Simmons et al. / Developmental Biology 304 (2007) 567–578

573

Fig. 5. Hand1 mRNA is expressed in all trophoblast giant cell subtypes and regulates trophoblast giant cell differentiation. Hand1 expression can be detected in all four trophoblast giant cell subtypes (C–F). Hand1 expression is detected in SpA-TGCs proximal to the placenta (C) but appears absent from more distal SpA-TGCs (A) which still express Plf (B). Hand1 mutant TS cells show reduced expression of marker genes for all TGC subtypes in vitro, but do not have altered expression of the TS cell markers Err2 and Eomes (G). C—canal lumen, SpA—spiral artery lumen. Scale bar = 0.1 mm.

Dai et al., 2000; Sahgal et al., 2000; Lee et al., 2003; Ishida et al., 2004; Simmons and Cross, 2005), though a recent study showed that the mononuclear trophoblast cells in the labyrinth are indeed polyploid (Coan et al., 2005). We have brought these different observations together and added significant new details. With this knowledge, and the availability of new markers, mouse mutants with placental phenotypes can now be described in greater detail. More importantly, the discovery of several new TGC subtypes, some with overlapping patterns of hormone gene expression, raises significant questions about the functions of these cells.

The TGC subtypes that we have identified have distinct patterns of gene expression, although only two have exclusive markers. Pl1 is unique to P-TGCs and Ctsq is unique to S-TGCs. For the other subtypes, we have not yet identified cell subtype-specific genes. In addition, some genes are expressed in multiple subtypes. For example, Plf is expressed in SpA-TGCs, P-TGCs, and C-TGCs while Pl2 is expressed in P-TGCs, C-TGCs, and S-TGCs. Distinct patterns of gene expression not only serve to identify TGC subtypes, but also indicate different programs of differentiation. Previous studies have shown that Pl1 expression precedes Pl2 in P-TGCs (Carney et al., 1993).

Fig. 6. Formation of trophoblast giant cell subtypes is differentially regulated by retinoic acid. Northern blot analysis demonstrates that retinoic acid (RA) can induce trophoblast giant cell formation in TS cell cultures, even in the presence of FGF and embryonic fibroblast conditioned medium (CM). However, not all markers of trophoblast giant cell subtypes are uniformly up-regulated (compare with marker profile of control) indicating preferential induction of P-TGCs and possibly SpATGCs. * Indicates TS cell morphology and arrowhead indicates TGC morphology. Scale bar = 0.1 mm.

574

D.G. Simmons et al. / Developmental Biology 304 (2007) 567–578

Fig. 7. Not all trophoblast giant cells arise from Tpbpa+ precursors. Female mice carrying a transgene containing a Cre recombinase-IRES-EGFP cassette under the control of the Tpbpa promoter were crossed with male Z/AP reporter mice to trace the fate of ectoplacental cone/spongiotrophoblast precursors. Placentas were stained at E14.5 for either β-galactosidase (blue) and/or human alkaline phosphatase (brown). Almost all SpA-TGCs arise from Tpbpa (+) precursors (D, H, L, P) while none of the S-TGCs do (B, F, J, N). P-TGC (C, G, K, O) and C-TGCs (E, I, M, Q) appear to arise from either Tpbpa (+) precursors or Tpbpa (−) precursors.

However, Pl1 expression is not detectable in the developing labyrinth layer, and therefore Pl2 expression in S-TGCs appears without first going through a transient Pl1 expressing stage. In addition, SpA-TGCs turn off Pl1 expression before P-TGCs and, unlike P-TGCs, never express Pl2. These observations indicate a unique differentiation program for each TGC subtype. The results of our study have not only added new cell types, but have indicated a need to revise the past model of how the trophoblast cell lineage develops into the various differentiated cell types. Previously, the P-TGC population was thought to arise both from the mural trophectoderm and from the secondary differentiation of precursors in the

ectoplacental cone and later the spongiotrophoblast layer (Rossant and Ofer, 1977; Johnson and Rossant, 1981; Carney et al., 1993). Our results have indicated, however, that ∼ 50% of P-TGCs in the midgestation placenta are derived from Tpbpa− cells. There are only ∼ 50 mural trophectoderm cells that could form the primary wave of P-TGCs and yet there are several hundred P-TGCs lining the implantation site by E8.5 (Scott et al., 2000). As a result, there must be an additional source of secondary P-TGCs in addition to Tpbpa+ precursors of the ectoplacental cone, perhaps even from outside the ectoplacental cone/spongiotrophoblast. The caveat with the strict interpretation of the cell numbers is that when using Cremediated transgene activation, there may be some delay

D.G. Simmons et al. / Developmental Biology 304 (2007) 567–578

Fig. 8. Different trophoblast giant cell subtypes have different developmental origins. Quantification of trophoblast giant cell subtypes originating from Tpbpa (+) (hPLAP positive—brown) versus Tpbpa (−) (β-galactosidase positive—blue) precursors. Error bars = ±S.E.M.

between the onset of Cre expression and detection of the hAP activity (Nagy, 2000). We previously addressed this possibility by looking at whether there are cells in the spongiotrophoblast layer that express Tpbpa mRNA but which do not also express hPLAP, but we found a perfect congruence (A. Fortier et al., unpublished). We have also found that the Cre transgene begins expression and has associated recombinase activity as early as E8.5. Therefore, if there is a delay, it probably is not significant and though may account for some unlabeled cells, it is unlikely to account for the hundreds of P-TGCs that cannot be accounted for by mural trophectoderm precursors alone. The potential additional source of P-TGCs is either the direct differentiation of trophoblast stem cells into P-TGCs or differentiation from a previously uncharacterized precursor population in the ectoplacental cone that is devoid of Tpbpa expression. Previous studies have shown that retinoic acid treatment of TS cell cultures can induce TGC differentiation (based on morphology and Pl1 expression) without first expressing Tpbpa/4311 (Yan et al., 2001), clearly suggesting an alternative pathway of TGC differentiation. Our observations show that the type of TGC produced by RA treatment is specifically the P-TGC subtype. Tpbpa expression at E8.5, when expression is initiated, does not encompass the chorion or even the whole of the ectoplacental cone, particularly at the base of the ectoplacental cone nearest the chorion. Therefore, it is possible that P-TGC precursors reside in this area. The origin of the other TGC subtypes is specific to each subtype. In contrast to P-TGCs, virtually all of the SpA-TGCs we observed expressed hPLAP, indicating their origin from Tpbpa+ precursors. This makes some sense as the ectoplacental cone is oriented to grow towards the incoming spiral arteries. In contrast, none of the S-TGCs was positive for hPLAP indicating an alternative developmental origin. As Tpbpa expression does not encompass the whole of the ectoplacental cone, particularly at the base closest to the chorion, it seems likely that S-TGC cells arise from precursors either within the Tpbpa− area of the ectoplacental cone or from chorion trophoblast directly. The different origins of SpA-TGCs and S-TGCs are not surprising considering the spatial localization of the two subtypes in the mature placenta. C-TGCs have mixed developmental origins, with approximately half arising from Tpbpa+ precursors, similar to the findings with P-TGCs. Interestingly, hPLAP+ C-TGCs cells are intermixed with β-

575

galactosidase+ C-TGCs cells along the entire length of the canals, with no clear demarcations. The ability to culture trophoblast stem (TS) cells has been invaluable for the study of trophoblast differentiation and TGC formation in particular. In this system, FGF, heparin, and embryonic fibroblast conditioned medium is required to maintain proliferating conditions and promote TS cell selfrenewal (Tanaka et al., 1998). Given the obvious utility of this system for the study of TGC differentiation, we used markers for Plf, Pl1, Pl2, and Ctsq which can distinguish TGC subtypes in vivo to probe northern blots of differentiating TS cell cultures. All four markers were induced by day 6 of differentiation, indicating the presence of multiple TGC subtypes within the culture system. TGC formation relies on the interplay of numerous cell intrinsic factors to guide proper differentiation, most notably basic helix–loop–helix (bHLH) transcription factors and their interacting partners including Hand1, Mash2, Alf1, Itf2, Id1/2, and I-mfa (Cross et al., 1995; Jen et al., 1997; Kraut et al., 1998; Riley et al., 1998; Scott et al., 2000; Hughes et al., 2004). Hand1 in particular is essential for proper TGC differentiation as Hand1 deficient conceptuses die by E7.5 due to a block in TGC formation (Riley et al., 1998; Scott et al., 2000). Results of the current study indicated that Hand1 is expressed in all four TGC subtypes and also that it regulates the differentiation of all TGC subtypes. Expression of all the various TGC subtype markers was significantly impaired in differentiating Hand1 mutant TS cell cultures (Fig. 5G). The ability of the cells to begin to differentiate in general was not affected in these cells as indicated by the correct downregulation of the trophoblast stem cell markers Esrrb and Eomes. Differentiated Hand1 mutant TS cells have previously been shown to contain significantly fewer invasive TGCs, demonstrating an intrinsic deficiency in TGC formation (Hemberger et al., 2004). In vivo, Hand1 is required both for differentiation of mural trophectoderm into TGCs and to promote secondary TGC differentiation (Riley et al., 1998; Scott et al., 2000). It would be of obvious interest to see whether other cell intrinsic and paracrine factors known to regulate P-TGC differentiation are also required for differentiation of other TGC subtypes. Retinoic acid treatment did not induce Pl2 or Ctsq expression, and in fact suppressed the induction of these markers under differentiating conditions, indicating that retinoic acid promotes P-TGCs (and possibly SpA-TGCs also) but not C-TGCs or S-TGCs. These differential effects imply that while there are some common features, the different TGC subtypes are also regulated by their own mechanisms. The existence of numerous TGC subtypes with unique gene expression patterns and spatial localization begs the question, what is the functional significance of having multiple TGC subtypes? One could speculate that the specific milieu of hormones and growth factors produced by each TGC subtype would have particular functions intimately tied to their location within the placenta. For example, SpA-TGCs are known to secrete factors which regulate blood flow to the implantation site such as vasodilators and angiogenic factors. Proliferin (Plf), an angiogenic factor known to stimulate endothelial cell

576

D.G. Simmons et al. / Developmental Biology 304 (2007) 567–578

Fig. 9. Summary of the trophoblast cell lineage and origins of different TGC subtypes.

migration (Jackson et al., 1994) and proliferin-related protein (Plfr), an anti-angiogenic factor (Jackson et al., 1994), are both expressed in SpA-TGCs. The combination of hormones and cytokines secreted by P-TGCs may preferentially target decidual cells or infiltrating leukocytes such as natural killer cells. Prlpa is expressed predominantly by P-TGCs (Ma and Linzer, 2000) and has been shown to modulate uterine natural killer cells (Muller et al., 1999; Ain et al., 2003). S-TGCs, located within the labyrinth layer closest to the fetal circulation, are well situated to facilitate delivery of hormones and/or cytokines into the fetal circulation or to produce molecules which effectively clear, or modulate the bioactivity, of hormones before they enter the fetal circulation. This could potentially be an important function for Ctsq, a protease uniquely expressed in S-TGC, as it has been hypothesized to cleave members of the prolactin-like protein family such as Pl2 (Ishida et al., 2004). Cleavage of prolactin, itself an angiogenic factor, by cathepsin D for example, results in a shorter fragment with anti-angiogenic activity (Struman et al., 1999; Lkhider et al., 2004; Piwnica et al., 2004). The potential cleavage of other members of the prolactin family locus, of which there are at least 23 known to be expressed within the placenta (Wiemers et al., 2003), possibly by placentally expressed cathepsins, may prove to be a critical regulatory system within the placenta to augment hormone bioactivity before entering the fetal circulation or re-entering the maternal circulation. Some genes have multiple sites of expression, such as Pl2 and Plf. However, the biological action of these gene products likely depends on the local environment they are secreted into, with different TGC subtypes using the same molecule to different ends. For example, Plf can stimulate angiogenesis and endothelial cell migration (Jackson et al., 1994), a function required in the immediate vicinity of SpA-TGCs. But Plf secretion from P-TGCs may be critical for stimulating uterine cell proliferation (Nelson et al., 1995), something which occurs

widely within the decidua bordering P-TGCs. Likewise, Pl2 is expressed in multiple TGC subtypes and is known to act through the prolactin receptor to regulate maternal adaptations to pregnancy such as mammary gland development and corpus luteum maintenance (Soares, 2004) or to bind receptors in fetal liver, pancreas, kidney, and intestine influencing fetal development (Freemark et al., 1993). TGC subtype-specific expression may determine which compartment, fetal or maternal, is targeted. Pl2 is expressed early in gestation from P-TGCs, when maternal Pl2 serum levels are high and in the second half of gestation from S-TGCs when Pl2 first appears in the fetal circulation (Kishi et al., 1991). In conclusion, the present study demonstrates a previously underappreciated level of TGC sub-specialization and provides a panel of markers which allows for the identification of at least four TGC subtypes. In addition, cell lineage tracing studies show that TGC subtypes can have different developmental origins, and even TGCs of the same subtype can have more than one developmental origin. There are likely more subtypes to be idenified as a number of genes expressed in P-TGCs, for example, are not uniformly expressed within this population, including Mrj (Hunter et al., 1999), Thrombomodulin (WeilerGuettler et al., 1996), and members of the Cathepsin family (Hemberger et al., 2000). In addition, Ma and Linzer describe Prlpa expression in secondary TGCs but not in primary TGCs, adding a marker distinguishing TGCs from these two different trophectoderm lineages (Ma and Linzer, 2000). Studies investigating trophoblast differentiation or mutant mice with placental phenotypes have often relied on the expression of just a few TGC markers, particularly Pl1, Pl2, or Plf, as a read out of TGC formation and function. In light of our findings, we now know that expression of these specific genes does not reflect the whole of the TGC population. Clearly, future studies of trophoblast differentiation or gene targeted mice with placental phenotypes should account for the diversity of trophoblast subtypes.

D.G. Simmons et al. / Developmental Biology 304 (2007) 567–578

Acknowledgments The authors would like to thank David Natale for his technical assistance and valuable discussions and Martha Hughes and Colleen Geary for excellent technical assistance. D.S. was supported by fellowships from the Lalor Foundation and the Alberta Heritage Foundation for Medical Research (AHFMR). The work was supported by operating grants from the Canadian Institutes of Health Research (CIHR) (to J.C.C.). J.C.C. is a CIHR Investigator and an AHFMR Scientist.

References Adamson, S.L., Lu, Y., Whiteley, K.J., Holmyard, D., Hemberger, M., Pfarrer, C., Cross, J.C., 2002. Interactions between trophoblast cells and the maternal and fetal circulation in the mouse placenta. Dev. Biol. 250, 358–373. Ain, R., Tash, J.S., Soares, M.J., 2003. Prolactin-like protein-A is a functional modulator of natural killer cells at the maternal-fetal interface. Mol. Cell. Endocrinol. 204, 65–74. Bany, B.M., Cross, J.C., 2006. Post-implantation mouse conceptuses produce paracrine signals that regulate the uterine endometrium undergoing decidualization. Dev. Biol. 294, 445–456. Bevilacqua, E.M., Abrahamsohn, P.A., 1988. Ultrastructure of trophoblast giant cell transformation during the invasive stage of implantation of the mouse embryo. J. Morphol. 198, 341–351. Calzonetti, T., Stevenson, L., Rossant, J., 1995. A novel regulatory region is required for trophoblast-specific transcription in transgenic mice. Dev. Biol. 171, 615–626. Campbell, W.J., Deb, S., Kwok, S.C., Joslin, J.A., Soares, M.J., 1989. Differential expression of placental lactogen-II and prolactin-like protein-A in the rat chorioallantoic placenta. Endocrinology 125, 1565–1574. Carney, E.W., Prideaux, V., Lye, S.J., Rossant, J., 1993. Progressive expression of trophoblast-specific genes during formation of mouse trophoblast giant cells in vitro. Mol. Reprod. Dev. 34, 357–368. Church, G.M., Gilbert, W., 1984. Genomic sequencing. Proc. Natl. Acad. Sci. U. S. A. 81, 1991–1995. Coan, P.M., Ferguson-Smith, A.C., Burton, G.J., 2005. Ultrastructural changes in the interhaemal membrane and junctional zone of the murine chorioallantoic placenta across gestation. J. Anat. 207, 783–796. Cross, J.C., Werb, Z., Fisher, S.J., 1994. Implantation and the placenta: key pieces of the development puzzle. Science 266, 1508–1518. Cross, J.C., Flannery, M.L., Blanar, M.A., Steingrimsson, E., Jenkins, N.A., Copeland, N.G., Rutter, W.J., Werb, Z., 1995. Hxt encodes a basic helix– loop–helix transcription factor that regulates trophoblast cell development. Development 121, 2513–2523. Cross, J.C., Hemberger, M., Lu, Y., Nozaki, T., Whiteley, K., Masutani, M., Adamson, S.L., 2002. Trophoblast functions, angiogenesis and remodeling of the maternal vasculature in the placenta. Mol. Cell. Endocrinol. 187, 207–212. Dai, G., Wang, D., Liu, B., Kasik, J.W., Muller, H., White, R.A., Hummel, G.S., Soares, M.J., 2000. Three novel paralogs of the rodent prolactin gene family. J. Endocrinol. 166, 63–75. Deb, S., Faria, T.N., Roby, K.F., Larsen, D., Kwok, S.C., Talamantes, F., Soares, M.J., 1991. Identification and characterization of a new member of the prolactin family, placental lactogen-I variant. J. Biol. Chem. 266, 1605–1610. Deussing, J., Kouadio, M., Rehman, S., Werber, I., Schwinde, A., Peters, C., 2002. Identification and characterization of a dense cluster of placentaspecific cysteine peptidase genes and related genes on mouse chromosome 13. Genomics 79, 225–240. El-Hashash, A.H., Esbrit, P., Kimber, S.J., 2005. PTHrP promotes murine secondary trophoblast giant cell differentiation through induction of endocycle, upregulation of giant-cell-promoting transcription factors and suppression of other trophoblast cell types. Differentiation 73, 154–174.

577

Erlebacher, A., Price, K.A., Glimcher, L.H., 2004. Maintenance of mouse trophoblast stem cell proliferation by TGF-beta/activin. Dev. Biol. 275, 158–169. Faria, T.N., Deb, S., Kwok, S.C., Talamantes, F., Soares, M.J., 1990. Ontogeny of placental lactogen-I and placental lactogen-II expression in the developing rat placenta. Dev. Biol. 141, 279–291. Faria, T.N., Ogren, L., Talamantes, F., Linzer, D.I., Soares, M.J., 1991. Localization of placental lactogen-I in trophoblast giant cells of the mouse placenta. Biol. Reprod. 44, 327–331. Fortier, A.L., Yamamoto, H., Simmons, D.G., Cross, J.C., submitted for publication. Placental spongiotrophoblast-specific activity of Tpbpa-Cre transgenic mice. Freemark, M., Kirk, K., Pihoker, C., Robertson, M.C., Shiu, R.P., Driscoll, P., 1993. Pregnancy lactogens in the rat conceptus and fetus: circulating levels, distribution of binding, and expression of receptor messenger ribonucleic acid. Endocrinology 133, 1830–1842. Gagioti, S., Scavone, C., Bevilacqua, E., 2000. Participation of the mouse implanting trophoblast in nitric oxide production during pregnancy. Biol. Reprod. 62, 260–268. Gardner, R.L., Davies, T.J., 1993. Lack of coupling between onset of giant transformation and genome endoreduplication in the mural trophectoderm of the mouse blastocyst. J. Exp. Zool. 265, 54–60. Gardner, R.L., Papaioannou, V.E., Barton, S.C., 1973. Origin of the ectoplacental cone and secondary giant cells in mouse blastocysts reconstituted from isolated trophoblast and inner cell mass. J. Embryol. Exp. Morphol. 30, 561–572. Guzman-Ayala, M., Ben-Haim, N., Beck, S., Constam, D.B., 2004. Nodal protein processing and fibroblast growth factor 4 synergize to maintain a trophoblast stem cell microenvironment. Proc. Natl. Acad. Sci. U. S. A. 101, 15656–15660. Hamlin, G.P., Lu, X.J., Roby, K.F., Soares, M.J., 1994. Recapitulation of the pathway for trophoblast giant cell differentiation in vitro: stage-specific expression of members of the prolactin gene family. Endocrinology 134, 2390–2396. Hemberger, M., Himmelbauer, H., Ruschmann, J., Zeitz, C., Fundele, R., 2000. cDNA subtraction cloning reveals novel genes whose temporal and spatial expression indicates association with trophoblast invasion. Dev. Biol. 222, 158–169. Hemberger, M., Nozaki, T., Masutani, M., Cross, J.C., 2003. Differential expression of angiogenic and vasodilatory factors by invasive trophoblast giant cells depending on depth of invasion. Dev. Dyn. 227, 185–191. Hemberger, M., Hughes, M., Cross, J.C., 2004. Trophoblast stem cells differentiate in vitro into invasive trophoblast giant cells. Dev. Biol. 271, 362–371. Hughes, M., Dobric, N., Scott, I.C., Su, L., Starovic, M., St-Pierre, B., Egan, S.E., Kingdom, J.C., Cross, J.C., 2004. The Hand1, Stra13 and Gcm1 transcription factors override FGF signaling to promote terminal differentiation of trophoblast stem cells. Dev. Biol. 271, 26–37. Hunter, P.J., Swanson, B.J., Haendel, M.A., Lyons, G.E., Cross, J.C., 1999. Mrj encodes a DnaJ-related co-chaperone that is essential for murine placental development. Development 126, 1247–1258. Ilgren, E.B., 1983. Review article: control of trophoblastic growth. Placenta 4, 307–328. Ishida, M., Ono, K., Taguchi, S., Ohashi, S., Naito, J., Horiguchi, K., Harigaya, T., 2004. Cathepsin gene expression in mouse placenta during the latter half of pregnancy. J. Reprod. Dev. 50, 515–523. Jackson, D., Volpert, O.V., Bouck, N., Linzer, D.I., 1994. Stimulation and inhibition of angiogenesis by placental proliferin and proliferin-related protein. Science 266, 1581–1584. Jen, Y., Manova, K., Benezra, R., 1997. Each member of the Id gene family exhibits a unique expression pattern in mouse gastrulation and neurogenesis. Dev. Dyn. 208, 92–106. Johnson, M.H., Rossant, J., 1981. Molecular studies on cells of the trophectodermal lineage of the postimplantation mouse embryo. J. Embryol. Exp. Morphol. 61, 103–116. Kanai-Azuma, M., Kanai, Y., Matsuda, H., Kurohmaru, M., Tachi, C., Yazaki, K., Hayashi, Y., 1997. Nerve growth factor promotes giant-cell transformation of

578

D.G. Simmons et al. / Developmental Biology 304 (2007) 567–578

mouse trophoblast cells in vitro. Biochem. Biophys. Res. Commun. 231, 309–315. Kishi, K., Hirashiba, M., Hasegawa, Y., 1991. Gestational profiles of rat placental lactogen-II (rPL-II) and growth hormone (GH) in maternal and fetal serum, amniotic fluid, and placental tissue. Endocrinol. Jpn. 38, 589–595. Kraut, N., Snider, L., Chen, C.M., Tapscott, S.J., Groudine, M., 1998. Requirement of the mouse I-mfa gene for placental development and skeletal patterning. EMBO J. 17, 6276–6288. Lee, S.J., Talamantes, F., Wilder, E., Linzer, D.I., Nathans, D., 1988. Trophoblastic giant cells of the mouse placenta as the site of proliferin synthesis. Endocrinology 122, 1761–1768. Lee, C.K., Moon, D.H., Shin, C.S., Kim, H., Yoon, Y.D., Kang, H.S., Lee, B.J., Kang, S.G., 2003. Circadian expression of Mel1a and PL-II genes in placenta: effects of melatonin on the PL-II gene expression in the rat placenta. Mol. Cell. Endocrinol. 200, 57–66. Linzer, D.I., Fisher, S.J., 1999. The placenta and the prolactin family of hormones: regulation of the physiology of pregnancy. Mol. Endocrinol. 13, 837–840. Lkhider, M., Castino, R., Bouguyon, E., Isidoro, C., Ollivier-Bousquet, M., 2004. Cathepsin D released by lactating rat mammary epithelial cells is involved in prolactin cleavage under physiological conditions. J. Cell Sci. 117, 5155–5164. Lobe, C.G., Koop, K.E., Kreppner, W., Lomeli, H., Gertsenstein, M., Nagy, A., 1999. Z/AP, a double reporter for cre-mediated recombination. Dev. Biol. 208, 281–292. Ma, G.T., Linzer, D.I., 2000. GATA-2 restricts prolactin-like protein A expression to secondary trophoblast giant cells in the mouse. Biol. Reprod. 63, 570–574. MacAuley, A., Cross, J.C., Werb, Z., 1998. Reprogramming the cell cycle for endoreduplication in rodent trophoblast cells. Mol. Biol. Cell 9, 795–807. Muller, H., Liu, B., Croy, B.A., Head, J.R., Hunt, J.S., Dai, G., Soares, M.J., 1999. Uterine natural killer cells are targets for a trophoblast cell-specific cytokine, prolactin-like protein A. Endocrinology 140, 2711–2720. Nagy, A., 2000. Cre recombinase: the universal reagent for genome tailoring. Genesis 26, 99–109. Nelson, J.T., Rosenzweig, N., Nilsen-Hamilton, M., 1995. Characterization of the mitogen-regulated protein (proliferin) receptor. Endocrinology 136, 283–288. Piwnica, D., Touraine, P., Struman, I., Tabruyn, S., Bolbach, G., Clapp, C., Martial, J.A., Kelly, P.A., Goffin, V., 2004. Cathepsin D processes human prolactin into multiple 16K-like N-terminal fragments: study of their antiangiogenic properties and physiological relevance. Mol. Endocrinol. 18, 2522–2542. Riley, P., Anson-Cartwright, L., Cross, J.C., 1998. The Hand1 bHLH transcription factor is essential for placentation and cardiac morphogenesis. Nat. Genet. 18, 271–275. Rossant, J., Ofer, L., 1977. Properties of extra-embryonic ectoderm isolated from postimplantation mouse embryos. J. Embryol. Exp. Morphol. 39, 183–194.

Sahgal, N., Knipp, G.T., Liu, B., Chapman, B.M., Dai, G., Soares, M.J., 2000. Identification of two new nonclassical members of the rat prolactin family. J. Mol. Endocrinol. 24, 95–108. Scott, I.C., Anson-Cartwright, L., Riley, P., Reda, D., Cross, J.C., 2000. The HAND1 basic helix–loop–helix transcription factor regulates trophoblast differentiation via multiple mechanisms. Mol. Cell. Biol. 20, 530–541. Simmons, D.G., Cross, J.C., 2005. Determinants of trophoblast lineage and cell subtype specification in the mouse placenta. Dev. Biol. 284, 12–24. Soares, M.J., 2004. The prolactin and growth hormone families: pregnancyspecific hormones/cytokines at the maternal–fetal interface. Reprod. Biol. Endocrinol. 2, 51. Sol-Church, K., Picerno, G.N., Stabley, D.L., Frenck, J., Xing, S., Bertenshaw, G.P., Mason, R.W., 2002. Evolution of placentally expressed cathepsins. Biochem. Biophys. Res. Commun. 293, 23–29. Struman, I., Bentzien, F., Lee, H., Mainfroid, V., D'Angelo, G., Goffin, V., Weiner, R.I., Martial, J.A., 1999. Opposing actions of intact and N-terminal fragments of the human prolactin/growth hormone family members on angiogenesis: an efficient mechanism for the regulation of angiogenesis. Proc. Natl. Acad. Sci. U. S. A. 96, 1246–1251. Tanaka, S., Kunath, T., Hadjantonakis, A.K., Nagy, A., Rossant, J., 1998. Promotion of trophoblast stem cell proliferation by FGF4. Science 282, 2072–2075. Tremblay, G.B., Kunath, T., Bergeron, D., Lapointe, L., Champigny, C., Bader, J.A., Rossant, J., Giguere, V., 2001. Diethylstilbestrol regulates trophoblast stem cell differentiation as a ligand of orphan nuclear receptor ERR beta. Genes Dev. 15, 833–838. Voss, A.K., Thomas, T., Gruss, P., 2000. Mice lacking HSP90beta fail to develop a placental labyrinth. Development 127, 1–11. Weiler-Guettler, H., Aird, W.C., Rayburn, H., Husain, M., Rosenberg, R.D., 1996. Developmentally regulated gene expression of thrombomodulin in postimplantation mouse embryos. Development 122, 2271–2281. Wiemers, D.O., Shao, L.J., Ain, R., Dai, G., Soares, M.J., 2003. The mouse prolactin gene family locus. Endocrinology 144, 313–325. Yamaguchi, M., Ogren, L., Endo, H., Thordarson, G., Bigsby, R.M., Talamantes, F., 1992. Production of mouse placental lactogen-I and placental lactogen-II by the same giant cell. Endocrinology 131, 1595–1602. Yamaguchi, M., Ogren, L., Endo, H., Soares, M.J., Talamantes, F., 1994. Colocalization of placental lactogen-I, placental lactogen-II, and proliferin in the mouse placenta at midpregnancy. Biol. Reprod. 51, 1188–1192. Yan, J., Tanaka, S., Oda, M., Makino, T., Ohgane, J., Shiota, K., 2001. Retinoic acid promotes differentiation of trophoblast stem cells to a giant cell fate. Dev. Biol. 235, 422–432. Yotsumoto, S., Shimada, T., Cui, C.Y., Nakashima, H., Fujiwara, H., Ko, M.S., 1998. Expression of adrenomedullin, a hypotensive peptide, in the trophoblast giant cells at the embryo implantation site in mouse. Dev. Biol. 203, 264–275. Zybina, E.V., Zybina, T.G., 1996. Polytene chromosomes in mammalian cells. Int. Rev. Cytol. 165, 53–119. Zybina, E.V., Zybina, T.G., Stein, G.I., 2000. Trophoblast cell invasiveness and capability for the cell and genome reproduction in rat placenta. Early Pregnancy 4, 39–57.