Post-implantation mouse conceptuses produce paracrine signals that regulate the uterine endometrium undergoing decidualization

Post-implantation mouse conceptuses produce paracrine signals that regulate the uterine endometrium undergoing decidualization

Developmental Biology 294 (2006) 445 – 456 www.elsevier.com/locate/ydbio Post-implantation mouse conceptuses produce paracrine signals that regulate ...

641KB Sizes 0 Downloads 46 Views

Developmental Biology 294 (2006) 445 – 456 www.elsevier.com/locate/ydbio

Post-implantation mouse conceptuses produce paracrine signals that regulate the uterine endometrium undergoing decidualization Brent M. Bany a,⁎, James C. Cross a,b a

Department of Biochemistry and Molecular Biology, University of Calgary, Calgary, Canada b Department of Obstetrics and Gynaecology, University of Calgary, Calgary, Canada Received for publication 21 October 2005; revised 9 February 2006; accepted 6 March 2006 Available online 17 April 2006

Abstract The uterus undergoes a series of dramatic changes in response to an implanting conceptus that, in some mammalian species, includes differentiation of the endometrial stroma into decidual tissue. This process, called decidualization, can be induced artificially in rodents indicating that the conceptus may not be essential for a proper maternal response in early pregnancy. In order to test this hypothesis, we determined if and how the conceptus affects uterine gene expression. We identified 5 genes (Angpt1, Angpt2, Dtprp, G1p2 and Prlpa) whose steady-state levels in the uterus undergoing decidualization depends on the presence of a conceptus. In situ hybridization revealed region-specific effects which suggested that various components of the conceptus and more than one signal from the conceptus are likely responsible for altering decidual cell function. Using cell culture models we found that trophoblast giant cells secrete a type I interferon-like molecule which can induce G1p2 expression in endometrial stromal cells. Finally, decidual Prlpa expression was reduced in the uterus adjacent to Hand1- and Ets2-deficient embryos, suggesting that normal trophoblast giant cells in the placenta are required for the conceptus-dependent effects on Prlpa expression in the mesometrial decidua. Overall, these results provide support for the hypothesis that molecular signals from the mouse conceptus have local effects on uterine gene expression during decidualization. © 2006 Elsevier Inc. All rights reserved. Keywords: Embryo implantation; Uterus; Trophoblast; Conceptus

Introduction Implantation is the process by which the conceptus comes into intimate contact with endometrium so the mother can provide nutrients to the developing fetus. Although the precise nature of implantation can vary significantly between mammalian species, it begins with the apposition of the blastocyst to the luminal epithelium of the endometrium and culminates in the formation of the definitive placenta (Weitlauf, 1988). In rodents and humans, the decidual reaction occurs during early pregnancy in which the fibroblast-like endometrial stromal cells transiently proliferate and then differentiate into large, polyploid decidual cells (DeFeo, 1967; Finn, 1996, 1998). The implanting rodent

⁎ Corresponding author. Present address: Department of Physiology, Southern Illinois University School of Medicine, Carbondale, IL 62901, USA. E-mail address: [email protected] (B.M. Bany). 0012-1606/$ - see front matter © 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.ydbio.2006.03.006

conceptus stimulates decidualization in a specific spatial– temporal pattern (Abrahamsohn and Zorn, 1993). It begins in the anti-mesometrial region adjacent to the conceptus, forming the primary decidual zone. Decidualization then spreads throughout most of the remainder of the antimesometrial area, forming the secondary decidual zone. Finally, decidualization begins in the mesometrial area of the endometrium and this is the area where the definitive placenta forms. Although implantation results in profound changes in the endometrium, it is less clear if specific signals from the conceptus are required since many of the changes that occur in the endometrium during decidualization do not absolutely require a conceptus. This is illustrated by a large number of reports, beginning almost 100 years ago (Loeb, 1907, 1908), that decidualization of the endometrium can occur in response to an artificial stimulus such as scratching the endometrium or injection of small beads or droplets of oil into the uterine lumen (reviewed in (DeFeo, 1967)). The endometrial tissue that

446

B.M. Bany, J.C. Cross / Developmental Biology 294 (2006) 445–456

forms in response to an artificial deciduogenic stimulus is called a deciduoma to distinguish it from the term decidua. A few studies have indicated that the decidua and artificial deciduoma show differences at the morphological level or timing (Deanesly, 1971; Lundkvist and Nilsson, 1982; Welsh and Enders, 1985), but these differences are subtle. In terms of changes in gene expression, overwhelming evidence indicates that gene expression in the developing decidua and deciduoma are quite similar. For example, expression levels of several genes are comparable between decidual and deciduomal tissue, such as heart and neural crest derivatives expressed transcript 2 (Hand2) (Austin et al., 2003) and prolactin-like protein I (Prlpi) (formerly called prolactin-like protein J) (Dai et al., 2000; Toft and Linzer, 1999). To our knowledge, fibroblast growth factor 2 (Fgf2) (Carlone and Rider, 1993) and interferon alpha-inducible protein (G1p2) (formerly called interferon-stimulated gene 15) (Austin et al., 2003) are the only two genes whose expression levels have been shown to be different between decidual and deciduomal tissues in rodents. In other species with rather different modes of implantation compared to rodents, there are some well-documented effects of conceptus-derived factors on endometrial gene expression and function during early pregnancy (Austin et al., 1999; Cameo et al., 2004; Emond et al., 2000; Hansen et al., 1999; Johnson et al., 2001; Kayisli et al., 2003; Licht et al., 2001a, b; Nagaoka et al., 2003; Parent et al., 2002). Our recent discovery that expression of G1p2, an interferon-inducible gene, is conserved in the pregnant endometrium of cattle and mice suggested that the mouse conceptus may indeed actively produce paracrine signals that alter endometrium differentiation or function (Austin et al., 2003). Further, although many reviews have been published about blastocyst–endometrial interactions at the onset of implantation (e.g., (Armant et al., 2000; Kimber, 2000; Paria et al., 2002)), relatively little is known about such interactions later. Therefore, our objective here was to use gene expression arrays to broadly determine whether the conceptus in mice alters endometrial function long after the onset of implantation and, if so, to identify the molecular signals.

Materials and methods Materials Evans blue, sesame oil, normal rabbit serum, biotinylated Dolichos biflorus agglutinin (DBA) lectin, estradiol-17β and progesterone were purchased from Sigma (St. Louis, MO). Oligotex mRNA isolation kits were purchased from Qiagen (Valencia, CA). The UCDNA Core Facility at the University of Calgary synthesized the random hexamer and reverse transcription-polymerase chain reaction oligonucleotides. Trizol Reagent, superscript II reverse transcriptase, DMEM:F12 and heat-inactivated fetal calf serum were purchased from InVitrogen (Carlsbad, CA). Mouse interferon-α (IFNα) was from MP Biomedical (Irvine, CA). Hybond N+ membrane, 32P-dCTP, Nick™ columns (prepacked Sephadex™ G-50 DNA grade gel exclusion columns) were purchased from Amersham Biosciences Corporation (Piscataway, NJ). The source for the protein A/G plus agarose was Santa Cruz Biotechnology (Santa Cruz, CA). Rabbit anti-mouse IFNα antiserum (anti-IFNα), mouse insulin-like growth factor I (IGF1), transforming growth factor beta-1 (TGFβ1), insulin-like growth factor II (IGF2), transforming growth factor alpha (TGFα) and leukemia inhibitory factor (LIF) were from R&D Systems (Minneapolis, MN). The PGEMteasy kit was purchased from Promega (Madison, WI). Plasmids containing cDNAs for Angpt1, Angpt2 and Eif4g2 were purchased from InVitrogen and corresponded to IMAGE clones 1163088, 2123060 and 3672168, respectively. Mouse Dtprp, Prlpa and Prlpi plasmids (Lin et al., 1997; Toft and Linzer, 1999) were kindly provided by Dr. Daniel Linzer (Northwestern University, Evanston). The mouse 18S rRNA plasmid was a generous gift from Dr. Dylan Edwards (University of East Anglia, Norwich). The Vegfa cDNA used was a RT-PCR product (primers 5′-CAGGCTGCACCCACGACA-3′, 5′TCTATGTGCTGGCTTTGG-3′) cloned into PGEMteasy. The Hand2 and G1p2 cDNAs used have been described elsewhere (Austin et al., 2003; Cross et al., 1995). The G1p2 and Fgfr1 cDNAs were a generous gifts from Drs. Thomas Hansen (Colorado State University, Fort Collins, CO) and Janet Rossant (Samuel Lunenfeld Research Institute, Toronto, ON), respectively. The source, length and method of generation of all other cDNAs used in this study are available upon request.

Animals and sample collection All procedures involving animals were approved by the University of Calgary and Southern Illinois University Animal Care Committees. CD1 mice (6–10 weeks), obtained from Charles River Laboratories (Wilmington, MA), were kept under light-controlled conditions (lights-on from 07:00 to 19:00 h) with free access to food and water. Females were placed with fertile males, and the morning that a vaginal plug was observed was considered to be day 0.5 of pregnancy. As outlined in Fig. 1, mice were sacrificed by cervical dislocation and their uteri dissected at 09:00 h on days 3.5 to 9.5 of pregnancy which

Fig. 1. Schematic representation of the protocol used to obtain pregnant uteri (top time-course) and those undergoing artificially-induced decidualization (lower timecourse) in order to collect decidua- and deciduoma-bearing uterine tissue, respectively. (A) Pregnant animals were killed at 09:00 days −1 to 5 of implantation which corresponds to days of pregnancy 3.5–9.5 and implantation and non-implantation segments were dissected. (B) Ovariectomized animals were give a series of subcutaneous injections of estradiol-17β (E2) and/or progesterone (P4). In order to artificially-induce decidualization 15 μl of sesame oil (artificial stimulus) was injected into the lumen of one uterine horn. Uterine horns were collected at 1–4 days after artificially-inducing decidualization.

B.M. Bany, J.C. Cross / Developmental Biology 294 (2006) 445–456 corresponds to −1 to 5 days after the onset of implantation (Fig. 1). Interimplantation (non-implantation sites) and implantation (implantation sites) segments (includes mesometrial plus anti-mesometrial sides) containing endometrium plus myometrium were collected. For total RNA isolation, conceptus tissues (embryo and trophoblast) were carefully removed from these implantation segments prior to processing. To discern implantation from nonimplantation segments of the uterus at day 4.5 of pregnancy, 0.1 ml of Evan's blue dye solution (2% in saline) was injected into the tail vein 5 min before sacrifice. Due to the increased vascular permeability which occurs in the rodent endometrium at the onset of implantation (Psychoyos, 1960), the implantation sites were identified as areas of intense blue coloration along the uterus. The artificial model of decidualization was utilized as described before (Bany and Schultz, 2001a) (Fig. 1). Briefly, mice were ovariectomized, allowed to recover, and then injected with a regimen of estradiol-17beta and progesterone. The morning that the uteri were optimally sensitized to respond to a deciduogenic stimulus, sesame oil (15 μl) was injected into the lumen of one uterine horn (stimulated) while the other horn served as a control (nonstimulated). At 1, 2, 3 or 4 days after artificially-inducing decidualization, the mice were sacrificed and uterine horns were dissected. These days after induction were chosen for comparisons to the tissues collected from pregnant mice described above at days 1–4 after the onset of implantation. In order to assess the potential function of giant cells in conceptus-mediated alterations in endometrial gene expression, Hand1 and Ets2 mutant mice were used. Hand1- and Ets2-deficient embryos fail to develop past day 7.5–9 (Riley et al., 1998) and 6.5–7.5 (Riley et al., 1998; Yamamoto et al., 1998), respectively. Up to that time the size and morphology of the decidua appears normal (Bany, unpublished observations) and the failure of the embryo appears to be due to either a lack of trophoblast giant cells (Hand1 mutant) or trophoblast giant cell function (Ets2 mutant) (Riley et al., 1998; Yamamoto et al., 1998). Mice heterozygous for either mutations in Hand1 (Riley et al., 1998) or Ets2 (kind gift from Robert Oshima, Burnham Institute) (Yamamoto et al., 1998) were kept in a pathogen-free double-barrier facility under light-controlled conditions with free access to food and water. Heterozygous animals were intercrossed in order to obtain timed pregnant females carrying wildtype, heterozygous and homozygous mutant Hand1 or Ets2 embryos. On day 7.5 of pregnancy, approximately 3 days after the onset of implantation, segments of the uterus containing implanted embryos were dissected. The embryos and placental tissue were dissected and the embryos were used for genotyping using polymerase chain reaction methods as previously described (Riley et al., 1998; Yamamoto et al., 1998). Uterine tissues were either homogenized in Trizol reagent for RNA isolation or fixed in 4% paraformaldehyde in PBS pH 7.2 for 24 h for in situ hybridization or histochemistry. The fixed tissue was dehydrated with ethanol, cleared in xylene, embedded in paraffin and sectioned (5 μm) using standard methods. The tissue sections were mounted on acid-washed silanized glass slides. For all experiments, tissue was collected from at least three independent mice at each time-point for each group.

Complementary DNA membrane arrays, hybridization and analysis Membranes spotted with 83 different cDNAs (probes) representing transcription factors, signaling molecules and hormones that are known to be expressed in the early embryo and uterus, along with Gapd cDNA, were prepared using a 96-well dot blot manifold and vacuum pump. Native plasmids were also spotted on the membranes to act as negative controls. Purified cDNAs (1 μg, in 0.15 ml of 2 × SSC) were spotted onto pre-wetted Hybond N+ membrane (Amersham). After washing twice with 0.2 ml of 2 × SSC, the membranes were removed from the manifold and the probes were denatured using denaturing buffer (0.5 M NaOH, 1.5 M NaCl) followed by neutralization buffer (1.5 M NaCl, 0.5 M Tris, pH 7.5) then cross-linked to the membranes by UV exposure using a Stratalinker (Stratagene) and protocols suggested by the manufacturer. Previously we have shown that by day 3 after the onset of implantation there is a conceptus-dependent increase in G1p2 gene expression in the endometrium (Austin et al., 2003). Therefore, we used gene expression arrays to compare mRNA levels in samples collected 3 days after implantation or artificial stimulation. Total RNA was prepared from decidual (implantation segments with embryo and extraembryonic membranes

447

removed) and deciduomal (stimulated uterine horns) at 3 days after the onset of implantation and from uteri at 3 days after artificial-induction of decidualization, respectively. Poly A+ RNA was isolated from this total RNA using the Oligotex mRNA isolation kit as directed by the manufacturer. The poly A+ RNA (1 μg) was reverse transcribed using random hexamers and Superscript II reverse transcriptase as recommended by the manufacturer in the presence of 32P-dCTP and the resulting 32Plabeled “target” cDNA was purified using Nick™ columns. Hybridization of the 32P-labeled target cDNA to membranes was carried out in modified Church buffer (Church and Gilbert, 1984) overnight at 65°C. The membranes were washed twice in wash buffer 1 (1 × SSC, 0.1% SDS), then twice in buffer 2 (0.2 × SSC, 0.1% SDS) for 15 min at 65°C. The hybridized radioactivity was detected using a Storm Phosphorimager and quantitatively analyzed using ImagQuant software (Molecular Dynamics). Hybridization signals for every gene in the arrays were measured and were determined to be significantly (P < 0.05) above background or not using t tests. For the first group of genes, where signals above background for both the decidua and deciduoma were found, density volumes (product of the average intensity above background and area of the spots) for their hybridization signal were calculated then normalized to the density volumes obtained for Gapd on each membrane. T tests were then used to determine if there was a significant difference between mean normalized hybridization signals. All statistical analysis was carried out using SigmaStat software (SPSS, Chicago). In a second group of genes, signals above background were obtained using the target RNA from pregnant tissue but not that of uterine horns undergoing artificially-induced decidualization. In a final group of genes, signals above background were not obtained using either target RNA. For these last two groups of genes, volume densities and further statistical analyses could not be carried out.

Northern blot and in situ hybridization analyses Northern blots of total RNA were prepared and screened with 32Plabeled probes using methods previously described (Bany and Schultz, 2001a). Hybridization signals for mRNA and 18 S rRNA were quantitated by densitometry. To localize mRNA in tissue sections, in situ hybridization was carried out using antisense DIG-labeled riboprobes as described previously (Bany and Schultz, 2001a). Specificity of the hybridizations for all probes was confirmed using sense DIG-labeled riboprobes as negative controls. The sections were inspected and photographed using a light microscope equipped with a CoolSNAP digital camera (Leica Microsytems Canada Inc., Thornhill, ON).

Cell cultures Endometrial stromal and deciduoma cells were isolated from uteri sensitized for a deciduogenic stimulus and from deciduomas, respectively, as described elsewhere (Afonso et al., 2002; Bany and Schultz, 2001b). The cells were cultured in medium (DMEM:F12 containing 10% charcoalstripped heat-inactivated fetal calf serum) with or without various treatments (IFNα, 100 IU/ml; normal rabbit serum; rabbit anti-mouse IFNα antiserum at 600 neutralizing units/ml ; various cytokines: IGF1, TGFβ1, IGF2, TGFα and LIF at 150 ng/ml; 30% trophoblast cell conditioned-media) for 24 h. After treatments the endometrial cells were homogenized in Trizol for RNA isolation. Trophoblast stem cells were maintained in stem cell culture medium and were then allowed to differentiate into trophoblast giant cells in differentiation medium as previously described (Tanaka et al., 1998). Trophoblast stem cellconditioned medium was collected after incubation for 24 h in medium lacking FGF4 and feeder cell-conditioned medium. Trophoblast giant cellconditioned medium was collected from cells that had differentiated for 4 days. Immunodepletion studies involved incubating trophoblast-conditioned media without or with combinations of rabbit anti-mouse IFNα antiserum (800 neutralization units/ml), normal rabbit serum (40 μl/ml) and IFNα (100 U/ml) at 4°C overnight. Protein A/G plus agarose beads (0.2 ml/10 ml medium) were then added and incubated for 1 h followed by centrifugation at 10,000 × g for 30 min at 4°C. The supernatants were recovered and kept at −70°C until use.

448

B.M. Bany, J.C. Cross / Developmental Biology 294 (2006) 445–456

Lectin histochemistry

samples from pregnant animals. For genes whose expression was detected by the arrays, the hybridization signals for six of the genes were significantly greater in decidua compared to artificially-induced deciduomas (Table 1). Additional 14 genes were detected above background only in decidual tissue and not in deciduoma (Table 1). Semi-quantitative Northern blotting was used to confirm differences in expression and expression of Angpt2, G1p2, Dtprp and Prlpa were all found to be significantly (P < 0.01) greater in normal decidua compared to artificially-induced deciduoma (Fig. 2). Chorionic somatomammotropin hormone 1 (Csh1)/placental lactogen I was not detected in the pregnant uterine samples verifying a lack of placental contamination of the samples from pregnant mice) (data not shown). These data imply that while there are significant similarities between the decidua and deciduoma, the implanted conceptus plays an important signaling role in altering the development and/or function of the decidua during normal pregnancy.

Histochemical staining of uterine histological sections with Dolichos biflorous agglutinin (DBA) lectin was carried out as previously described (Paffaro et al., 2003) to localize uterine natural killer cells. Sections were counterstained with hematoxylin.

Results To directly compare the development and function of decidua and deciduoma, we used gene expression arrays for analysis of gene expression in three independent samples collected 3 days after implantation or artificial stimulation, respectively (Fig. 1). We prepared arrays containing 83 mouse cDNAs encoding transcription factors, signaling molecules and hormones that are known to be expressed in the early embryo and uterus. To ensure that the conceptus had been removed completely from the decidual samples, the arrays included several trophoblast-specific genes that are expressed at very high levels in trophoblast giant cells and ectoplacental cone (e.g., Csh1, Plf, Tpbp), the fetal cells in most direct contact with the deciduas (Carney et al., 1993; Faria et al., 1991; Lee et al., 1988; Nieder and Jennes, 1990). Expression was not detected for any of these genes thereby confirming the lack of contaminating trophoblast tissues in our dissected uterine

Reduced G1p2 expression in the anti-mesometrial region of deciduoma G1p2 encodes a ubiquitin-like protein and was first described as an interferon (IFN)-stimulated gene (Farrell et al., 1979; Reich et al., 1987). It has previously been shown to be expressed

Table 1 Membrane gene expression array results of decidua relative to artificial deciduoma Gene name (symbol)

Fold-Difference (mean ± SEM)

Decidua vs. Deciduoma Wingless-related MMTV integration site 2 (Wnt2) Adrenomedullin (Adm) Prolactin-like protein I (Prlpi) Twist gene homolog 1 (Twist1) Wingless-related MMTV integration site 1 (Wnt1) Heart and neural crest derivatives expressed transcript 2 (Hand2) Vascular endothelial growth factor (Vegfa) Cyclin D3 (Ccnd3) Follistatin (Fst) Eukaryotic transcription factor initiation factor 4, gamma 2 (Eif4g2) Inhibitor of DNA binding (Idb2) Interferon alpha- inducible (G1p2) Decidual/trophoblast prolactin-related protein (Dtprp) Fibroblast growth factor receptor 1 (Fgfr1) Angiopoietin 2 (Angpt2)

0.5 ± 0.2 0.7 ± 0.1 0.7 ± 0.4 0.9 ± 0.3 1.5 ± 0.2 1.6 ± 0.7 1.8 ± 0.7 1.9 ± 0.2 2.1 ± 0.1 2.2 ± 0.1 P < 0.05 2.6 ± 0.1 P < 0.05 2.9 ± 0.2 P < 0.05 4.7 ± 1.2 P < 0.05 7.2 ± 2.5 P < 0.05 10.5 ± 3.3 P < 0.05

Detected in decidua but not artificial deciduoma Bmp4 Ccnd1

Ccnd2 Cdc5l

Chx10 Egfl7

Hes1 Ifit1

Irx1 Irx3

Notch4 Pem

Fgf1 Fgf10 Fgf2 Fgf3 Fgf4 Fgf5

Flt1 Foxm1 Gata2 Gata3 Gata4 Gata6 Fgf7 Hes2 Hes3 Idb1

Idb3 Idb4 Ifit2 Irx4 Irx5 Jun Kdr Lamb1-1 Lfng

Limk1 Mfng Nek6 Nek7 Nog Nos2 Nos3 Plf Prlpm

Ramp2 Ramp3 Rfng Shh Tbx5 Tpbpa Vim Wnt6 Wnt7a

Not detected in either Ada Angpt1 Bmp1 Bmp8a Bmp8b Bysl Csh1 Elf4 Erf

Fgf8 Fgfr2

Prlpa Wnt7b

B.M. Bany, J.C. Cross / Developmental Biology 294 (2006) 445–456

449

Fig. 2. Northern blot analysis of Dtprp, G1p2, Angpt2, Prlpa and Hand2 mRNA levels in total RNA isolated from sections of uteri undergoing normal (decidua) and artificially-induced (deciduoma) decidualization at 3 days after implantation and induction, respectively. (A) Representative autoradiograms and (B) summary of results showing the mean (±SEM, N = 3) mRNA levels in the deciduoma and decidua with the amount (mRNA/18S rRNA) for the deciduoma normalized to 1. *Denotes a significant (P < 0.01) difference between the decidua and deciduoma.

in the decidua on day 7.5 of pregnancy (Austin et al., 2003). We confirmed by Northern blotting that G1p2 mRNA levels are higher in implantation segments of the uterus at 3 days after the onset of implantation compared to uterine horns at 3 days after artificially-inducing decidualization (Fig. 2). G1p2 mRNA was not detected in non-stimulated or non-implantation segments of the uterus (Fig. 3A). Within implantation sites, G1p2 mRNAwas detected beginning 1 day after the onset of implantation with levels becoming quite high by the third day onwards (Fig. 3A). On the other hand, G1p2 mRNA was only barely detectable by 3 days after artificial induction of decidualization and the level at day 4 was comparable to that found in decidua at day 1 (Fig. 3A). G1p2 mRNA localized specifically to the anti-mesometrial decidua (Fig. 3B) but was essentially undetectable in deciduomal tissue (data not shown). To ensure the specificity of the effect, mRNA levels for two other genes were evaluated in detail. Expression of Prlpi and Hand2 did not appear to be different based on the array analysis (Table 1) and this was verified by Northern blotting (Figs 2 and 3). Similar to G1p2, Prlpi (Fig. 3) mRNAs were localized to the anti-mesometrial decidua and the pattern was not different in deciduoma. Similar results have been reported about the localization of Hand2 mRNA in the anti-mesometrial decidua (Austin et al., 2003). Reduced Dtprp expression in mesometrial, but not anti-mesometrial, deciduoma Dtprp is a member of the prolactin/placental lactogen superfamily of genes and has previously been shown to be expressed in the anti-mesometrial mouse decidua starting 3 days after the onset of implantation (Lin et al., 1997). Dtprp mRNA was not detected in inter-implantation segments of the uterus or in non-stimulated uterine horns (Fig. 4), but was detected in implantation sites of the uterus within 2 days after the onset of implantation (Fig. 4A). By contrast, Dtprp mRNA was barely detected by 3 days after artificial induction of decidualization,

though the levels dramatically increased to a level found in the pregnant uterus by 4 days (Fig. 4A). In situ hybridization revealed that at 3 days, Dtprp mRNA was localized in decidual tissue immediately surrounding the conceptus (Fig. 4B), including both a broad band of anti-mesometrial expression and a narrower layer of decidual cells located lateral and mesometrial to the ectoplacental cone of the conceptus. In contrast, Dtprp mRNA was localized in the deciduoma but only in a narrow band of the anti-mesometrial region at day 3 after the onset of induction (Fig. 4B). Unlike the decidua at day 3, no mRNA was localized to the mesometrial decidua. This difference in the pattern of localization of Dtprp mRNA between the decidua and deciduoma was not different at day 4 (data not shown) implying that expression in the mesometrial deciduoma was not simply delayed. Expression Prlpa is dependent on the conceptus Prlpa is another member of the prolactin/placental lactogen gene family implicated in regulation of natural killer (NK) cell function (Ain et al., 2003; Muller et al., 1999). Prlpa mRNA was not detected in inter-implantation segments of the pregnant uterus and in non-simulated uterine horns at any time (Fig. 5A). Within implantation segments of the uterus, Prlpa mRNA was detected beginning 2 days after the onset of implantation and reached high levels by day 4. Prlpa expression had only previously been described in trophoblast giant cells (Lin et al., 1997; Muller et al., 1998) which explains the strong localization of Prlpa mRNA to the developing placenta (Fig. 5B). However, our in situ hybridization results also showed expression in a sub-population of cells in the decidua clustered in the mesometrial region (Fig. 5B). Since they are found in the mesometrial decidua during pregnancy (Paffaro et al., 2003), and Prlpa might influence their function (Ain et al., 2003; Muller et al., 1999), we localized uterine NK cells on serial sections using DBA lectin staining. We found that they localized to the same region as the Prlpa expressing cells

450

B.M. Bany, J.C. Cross / Developmental Biology 294 (2006) 445–456

Fig. 3. G1p2 transcript levels and localization in pregnant uteri (decidua) and those undergoing artificially-induced decidualization (deciduoma). (A) Northern blot analyses of G1p2 and Hand2 transcript levels. Representative autoradiograms from one of three blots (N = 3) subjected to Northern blot analyses. The blot contained samples from non-implantation and implantation sites, and from non-stimulated and stimulated horns on various days after the onset of implantation and induction, respectively. (B) In situ hybridization analysis for G1p2 mRNA in the decidua at 3 days after the onset of implantation. (C) In situ hybridization analysis for Prlpi mRNA in the decidua and artificial deciduoma at 3 days after the onset of implantation (day 7.5 of pregnancy) and the induction of decidualization, respectively. All sections oriented with the mesometrial (MM) decidua up and anti-mesometrial (AM) down.

Fig. 4. Dtprp transcript levels and localization in pregnant uteri (decidua) and those undergoing artificially-induced decidualization (artificial deciduoma). (A) Representative autoradiograms from one of three blots (N = 3) subjected to Northern blot analyses. The blot contained samples from non-implantation and implantation sites, and from non-stimulated and stimulated horns on various days after the onset of implantation and induction, respectively. (B) In situ hybridization analysis for Dtprp mRNA in the decidua and deciduoma at 3 days after the onset of implantation and the induction of decidualization, respectively. All sections oriented with the mesometrial (MM) decidua up and anti-mesometrial (AM) down; EPC, ectoplacental cone.

B.M. Bany, J.C. Cross / Developmental Biology 294 (2006) 445–456

451

Fig. 5. Prlpa transcript levels and localization in pregnant uteri (decidua) and those undergoing artificially-induced decidualization (artificial deciduoma). (A) Representative autoradiograms from one of three blots (N = 3) subjected to Northern blot analyses. The blot contained samples from non-implantation plus implantation sites and from non-stimulated plus stimulated horns on various days after the onset of implantation and induction, respectively. (B) In situ hybridization analysis for Prlpa mRNA in the decidua and deciduoma at 3 days after the onset of implantation (day 7.5 of pregnancy) and the induction of decidualization, respectively. All sections oriented with the mesometrial (MM) decidua; AM, anti-mesometrial; EPC, ectoplacental cone. (C) Histochemical localization of uterine natural killer (uNK) cells in the mesometrial deciduas by DBA lectin histochemistry at day 7.5 of pregnancy. All sections oriented with the mesometrial (MM) decidua up and anti-mesometrial (AM) down.

(Fig. 5C). Analysis of samples from deciduomas showed that expression of Prlpa was essentially undetectable either by Northern blotting or in situ hybridization (Fig. 5). The conceptus regulates mesometrial expression of Angpt2 and Angpt1 Angiopoietin-2 plays a critical role in the stability of blood vessels during angiogenesis and vascular remodeling (Thurston, 2003). During the peri-implantation period, Angpt2 mRNA was detected in the pregnant uterus within 1 day after the onset of implantation, though the greatest levels were detected at days 3– 5 (Fig. 6A). Angpt2 mRNA was not detected in the preimplantation uterus or in inter-implantation segments of the pregnant uterus. In contrast to the pregnant uterus, Angpt2 mRNAwas not detected in artificially- induced deciduomas until day 3 and was at much lower levels (Fig. 6A). Angpt2 mRNA localized throughout both the mesometrial and anti-mesometrial decidual regions (Fig. 6B). In artificially-induced deciduomas by contrast, Angpt2 mRNA was restricted to the antimesometrial region (Fig. 6B). Since Angiopoietin-1 (Angpt1) and vascular endothelial growth factor A (Vegfa) function together with Angpt2 to regulate angiogenesis and blood vessel stability (Thurston, 2003), we also examined their expression in the uterus during the peri-implantation period (Fig. 6A). Interestingly, the highest Angpt1 expression was observed in the uterus just prior to the onset of decidualization with levels being very low on days 1–5. In contrast, Angpt1 mRNA levels in artificial deciduomas remained high 1–3 days after induction and then decreased to a

low level by day 4. Vegfa mRNA levels were detected at similar levels in decidua versus deciduoma. Angpt1 and Vegfa mRNA levels remained low both in non-implantation segments and nonstimulated uterine horns. These data show that while Vegfa is induced by any implantation stimulus, the normal switch from high Angpt1/low Angpt2 to low Angpt1/high Angpt2 requires an implanting conceptus. Interferon-α induces G1p2, but not Dtprp, Prlpa or Angpt2, in cultured deciduomal cells The observed differences in expression of G1p2, Dtprp, Prlpa and Angpt2 between decidua and deciduoma could have been due to either cellular differences in the tissue related to altered differentiation of the endometrial stromal cells or migration of inflammatory cells into the tissue, or because of acute differences in the cytokine or growth factor environment that alter gene expression within the existing cells. To distinguish these two possibilities, we isolated deciduomal cells that were already differentiated in vivo after artificial induction and exposed them in culture to conceptusderived factors to determine if this would provoke a change in gene expression to mimic decidual-like cells. G1p2 is a known interferon (IFN)-regulated gene (reviewed in Dao and Zhang, 2005) and IFNs are produced by peri-implantation conceptuses in ruminant species (Bazer and Johnson, 1991; Roberts et al., 1990) and the mouse (Cross et al., 1990). As shown previously (Austin et al., 2003), IFNα could stimulate G1p2 mRNA expression in undifferentiated endometrial stromal cells that had been isolated from ovariectomized

452

B.M. Bany, J.C. Cross / Developmental Biology 294 (2006) 445–456

Fig. 6. Analysis of Angpt1, Angpt2 and Vegfa mRNA in pregnant uteri and those undergoing artificially-induced decidualization. (A) Representative autoradiograms from one of three blots (N = 3) subjected to Northern blot analyses. The blot contained samples from non-implantation and implantation sites, and from non-stimulated and stimulated horns on various days after the onset of implantation and induction, respectively. (B) In situ hybridization analysis localizing Angpt2 mRNA in the decidua and deciduoma at 3 days after the onset of implantation or induction of decidualization. All sections oriented with the mesometrial (MM) decidua up and antimesometrial (AM) down.

mice sensitized for a deciduogenic stimulus (Fig. 7A). This effect was specific since several other cytokines had no effect. Importantly, cells isolated from artificially-induced deciduomas could also be induced by IFNα to increase G1p2 expression (Fig. 7A). While IFNα-stimulated G1p2 expression, no effect was seen on Dtprp, Prlpa or Angpt2 in either the undifferentiated endometrial stromal cells or deciduomal cells (data not shown). Trophoblast giant cells release an IFN that regulates decidual G1p2 expression Because trophoblast giant cells lie in direct contact with decidual cells, we carried out cell culture studies to determine if trophoblast giant cells secrete factors that can regulate decidual cell gene expression. Endometrial stromal cells incubated with trophoblast giant cell-conditioned medium caused a dramatic increase (P < 0.05) in G1p2 mRNA expression compared to cells incubated with control medium that had been conditioned by undifferentiated trophoblast stem cells (Fig. 7B). There was no differential effect of trophoblast giant cell conditioned medium on Dtprp, Prlpa or Angpt2 expression (data not shown). To determine if the effect of the trophoblast giant cell conditioned medium was mediated by an IFN, the endometrial

stromal cells were co-incubated with an anti-IFNα antiserum. We found that the antiserum blocked the ability of the conditioned medium to stimulate G1p2 expression (Fig. 7C), though the experiment did not indicate whether the IFN was directly released by the trophoblast giant cells or, by contrast, whether some other factor was produced by the trophoblast giant cells which in turn stimulated IFN production by the endometrial stromal cells themselves. To distinguish these possibilities, the conditioned medium was subjected to immunodepletion using the IFNα antiserum (Fig. 7D). We first showed that the antiserum was effective for immunodepletion because it was able to deplete IFNα activity from medium that had been spiked with purified IFNα (Fig. 7D). Immunodepletion reversed the effect of trophoblast giant cell-conditioned medium on G1p2 expression in the endometrial cells, indicating that trophoblast giant cells directly release an IFNαlike factor (Fig. 7D). Reduced Prlpa expression in decidua surrounding Hand1 and Ets2 mutant conceptuses To further analyze the role of trophoblast giant cells in regulating decidual cell differentiation and/or function we assessed the expression of G1p2, Dtprp, Prlpa and Angpt2 in

B.M. Bany, J.C. Cross / Developmental Biology 294 (2006) 445–456

453

Fig. 7. Role of giant cell-derived IFNα on mouse endometrial stromal cell and deciduomal cell G1p2 expression. (A) Representative effects (N = 3) of recombinant mouse interferon alpha (IFNα) on G1p2 transcript levels in cultured mouse endometrial stromal cells isolated from uteri sensitized for a deciduogenic stimulus and deciduomal cell isolated from deciduomas on day 3. Treatments were for 24 h and included vehicle, IFNα, IGF1, TGFβ1, IGF2, LIFα or LIF for stromal cells and vehicle or IFNα for deciduomal cells. (B) Representative autoradiograms and summary graph of Northern blot analysis of mouse endometrial stromal cells incubated with stem or giant cell-conditioned medium for 72 h. Bars represent the mean (± SEM, N = 3) (C) Representative autoradiograms of Northern blot analysis (N = 3) if the effect of co-incubating the conditioned medium with non-immune or anti-IFNα serum for 72 h. (D) Representative autoradiograms of Northern blot analysis (N = 3) of the effect of prior immunodepletion of conditioned media using a non-immune or anti-IFNα antiserum. Additionally, the conditioned medium spiked with IFNα was incubated with or without anti-IFNα antiserum to ensure the specificity of the antibody. Bars in graphs with different letters (a vs. b) are significantly (P < 0.05) different.

the decidua surrounding mouse embryos with known trophoblast giant cell defects. Hand1 and Ets2 are transcription factor genes that are required for the development and normal function of trophoblast giant cells, respectively (Riley et al., 1998; Yamamoto et al., 1998). The implantation sites containing mutant embryos still underwent decidualization on cue and the decidual tissues appeared morphologically similar to those containing wildtype embryos at day 7.5 of pregnancy (data not shown). However, Prlpa expression was very low in decidual tissue surrounding both Hand1- and Ets2-deficient conceptuses compared to the adjacent to normal (wildtype and heterozygous) conceptuses (Fig. 8). No differences were found for G1p2, Dtprp or Angpt2. Finally, Csh1 mRNA was not detected in any samples indicating the samples were not contaminated with trophoblast tissue.

and 2 are known to play a major role in vascular remodeling and angiogenesis in other tissues (Thurston, 2003). Such vascular remodeling and angiogenesis occurs in the uterus during implantation (Sherer and Abulafia, 2001; Smith, 2000). On the other hand, Dtprp and Prlpa are believed to target immune/

Discussion In the present study, we have shown that the postimplantation mouse conceptus, at least in part, regulates the expression of specific genes (Angpt1/2, Dtprp, G1p2, and Prlpa) in the endometrium undergoing decidualization. Angpt1

Fig. 8. Representative autoradiograms of Northern blot analysis (N = 3) of Prlpa, Dtprp, G1p2 and Angpt2 transcript levels in uterine tissue segments surrounding (A) Hand1- or (B) Ets2-deficient (−/−) compared to expressing (+/+, +/−) embryos at 3 days after the onset of implantation.

454

B.M. Bany, J.C. Cross / Developmental Biology 294 (2006) 445–456

inflammatory cells that are found in the uterus during implantation (Ain et al., 2003; Wang et al., 2000). Finally, G1p2 is involved in intracellular protein modification and is a member of the ubiquitin-like protein family (Dao and Zhang, 2005; Kim and Zhang, 2003). Therefore, the results of this study provide strong evidence that the post-implantation mouse conceptus plays an active role in directing a variety of functions in the endometrium after implantation. A key to understanding how the post-implantation conceptus regulates gene expression in the uterus during decidualization is to define the specific signaling mechanisms involved. From the present study, we can conclude that more than one signaling mechanism is involved (Fig. 9). Although our results show that type I interferons from the conceptus regulates G1p2 expression in the decidua, the conceptus-derived factor or factors responsible for regulating the other genes is unclear at present. Another surprising feature of these conceptus effects on decidual gene expression is that they are spatially different. G1p2 and Prlpa expression occur only in the anti-mesometrial and mesometrial decidua, respectively. It remains to be determined whether this is due to directional secretion of factors by the conceptus or to a differential responsiveness of decidual cells to such factors between the anti-mesometrial and mesometrial decidua. In the case of Angpt2 and Dtprp expression in the decidua, regulation appears to be more complex because they are both expressed in the antimesometrial and mesometrial decidua but the conceptus only influences their expression in the mesometrial region. Trophoblast giant cells are the cells of the conceptus that are in closest contact to the endometrium during decidualization. They behave as both endocrine and paracrine regulators of the maternal system. Csh1 (placental lactogen I) is specifically secreted by trophoblast giant cells and has effects on target organs outside of the uterus including the mammary gland and corpus luteum of the ovary (Faria et al., 1991; Nieder and Jennes, 1990). On the other hand, several other proteins secreted by trophoblast giant cells are believed to act as paracrine regulators within the endometrium and include factors such as proliferin and proliferin-related proteins (Lee et al., 1988; Linzer and Fisher, 1999). The results of the present study indicate that a type I interferon should be added to the list of paracrine regulators secreted by mouse trophoblast giant cells that has an effect on uterine gene expression in the uterus during decidualization.

Fig. 9. Schematic diagram of proposed conceptus-induced effects on endometrial gene expression. Cartoons of uterine sections oriented with the mesometrial (MM) decidua up and anti-mesometrial (AM) down. C, conceptus.

Given the results of the present paper, the traditional view of an antagonistic relationship between the conceptus and the uterus must be revised, and instead a cooperative-type model is perhaps more appropriate. Indeed our data indicate that a relay mechanism exists between the conceptus and target cells within the decidua that involves decidual stromal cells as an intermediate (Fig. 9). For example, the conceptus induces expression of Angpt2, Dtprp and Prlpa in decidual cells but in turn these factors then act on vascular endothelial cells (Tsigkos et al., 2003), eosinophils (Wang et al., 2000) and uterine natural killer cells (Ain et al., 2003; Muller et al., 1999), respectively, presumably to exact changes that are required for pregnancy success. Perhaps, such a relay mechanism has an adaptive significance in that both the conceptus and decidua must be developing synchronously and correctly in order to convey the correct signal to the target cells. In such a cooperative model, if either the conceptus is not developing well or the uterine is not prepared for pregnancy, the signal is not relayed and pregnancy failure would be expected. The roles of the conceptus-directed changes in gene expression found in the present study are likely related to immune and vascular changes that normally occur during pregnancy. As the conceptus implants, it invariably expresses paternally derived antigens which are exposed to the maternal immune system and elicit an immune response (Beer and Billingham, 1971). Although such a maternal immune response would intuitively be detrimental for the conceptus, it currently it is believed that maternal immune cells present in the endometrium during implantation play a key supportive role in implantation (Croy et al., 2003; Hunt et al., 2000). Finally, angiogenesis/vascular remodeling is a hallmark feature of implantation (Sherer and Abulafia, 2001; Smith, 2000) and the functions of the angiopoietins in this process is well known (Tsigkos et al., 2003). Taken together, the findings of the present paper with regards to endometrial expression of Prlpa, Dtprp, and angiopoietins shows that the mouse conceptus plays a role in modulating uterine leukocyte function and angiogenesis/vascular remodeling that occurs in the uterus during implantation. In other species with rather different modes of implantation compared to rodents, there are well-documented effects of conceptus-derived factors on endometrial gene expression and function during early pregnancy. In pregnant women, conceptusderived chorionic gonadotrophin has been shown to modulate the expression of several uterine genes (Cameo et al., 2004; Licht et al., 2001a,b). Also, in ruminants interferon-τ is secreted by the conceptus and induces uterine expression of G1p2 (Austin et al., 1999; Hansen et al., 1999; Johnson et al., 1999; Parent et al., 2002), as well as other genes (Austin et al., 1999; Emond et al., 2000; Johnson et al., 2001; Nagaoka et al., 2003). The present study shows that that the mouse conceptus too can influence endometrial gene expression during implantation. Clearly more research is required to determine the identity of more target genes and how they are controlled by the conceptus. The discovery of genes whose expression in the uterus are dependent on the presence of a healthy conceptus has very important clinical implications. Early post-implantation embryo failure is very common in humans (Wilcox et al., 1988) and

B.M. Bany, J.C. Cross / Developmental Biology 294 (2006) 445–456

implantation failure after embryo transfer remaining the major barrier to IVF success in humans (Spandorfer and Rosenwaks, 1999). Knowing what genes reflect a normal decidual response and healthy conceptus should allow the development of markers than can be used to identify pregnancies at risk of spontaneous loss. In addition, by identifying the signals that promote normal decidual function, it may be possible to mimic or alter their function to alter the course of pregnancy. Acknowledgments This work was supported by grants from the Canadian Institute for Health Research (CIHR) (JCC) and a CIHR postdoctoral fellowship (BMB). References Abrahamsohn, P.A., Zorn, T.M., 1993. Implantation and decidualization in rodents. J. Exp. Zool. 266, 603–628. Afonso, S., Tovar, C., Romagnano, L., Babiarz, B., 2002. Control and expression of cystatin C by mouse decidual cultures. Mol. Reprod. Dev. 61, 155–163. 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. Armant, D.R., Wang, J., Liu, Z., 2000. Intracellular signaling in the developing blastocyst as a consequence of the maternal–embryonic dialogue. Semin. Reprod. Med. 18, 273–287. Austin, K.J., King, C.P., Vierk, J.E., Sasser, R.G., Hansen, T.R., 1999. Pregnancy-specific protein B induces release of an alpha chemokine in bovine endometrium. Endocrinology 140, 542–545. Austin, K.J., Bany, B.M., Belden, E.L., Rempel, L.A., Cross, J.C., Hansen, T.R., 2003. Interferon-stimulated gene-15 (Isg15) expression is up-regulated in the mouse uterus in response to the implanting conceptus. Endocrinology 144, 3107–3113. Bany, B.M., Schultz, G.A., 2001a. Increased expression of a novel heat shock protein transcript in the mouse uterus during decidualization and in response to progesterone. Biol. Reprod. 64, 284–292. Bany, B.M., Schultz, G.A., 2001b. Tissue inhibitor of matrix metalloproteinase3 expression in the mouse uterus during implantation and artificially induced decidualization. Mol. Reprod. Dev. 59, 159–167. Bazer, F.W., Johnson, H.M., 1991. Type I conceptus interferons: maternal recognition of pregnancy signals and potential therapeutic agents. Am. J. Reprod. Immunol. 26, 19–22. Beer, A.E., Billingham, R.E., 1971. Immunobiology of mammalian reproduction. Adv. Immunol. 14, 1–84. Cameo, P., Srisuparp, S., Strakova, Z., Fazleabas, A.T., 2004. Chorionic gonadotropin and uterine dialogue in the primate. Reprod. Biol. Endocrinol. 2, 50. Carlone, D.L., Rider, V., 1993. Embryonic modulation of basic fibroblast growth factor in the rat uterus. Biol. Reprod. 49, 653–665. 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. Cross, J.C., Farin, C.E., Sharif, S.F., Roberts, R.M., 1990. Characterization of the antiviral activity constitutively produced by murine conceptuses: absence of placental mRNAs for interferon alpha and beta. Mol. Reprod. Dev. 26, 122–128. 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. Croy, B.A., He, H., Esadeg, S., Wei, Q., McCartney, D., Zhang, J.,

455

Borzychowski, A., Ashkar, A.A., Black, G.P., Evans, S.S., Chantakru, S., van den Heuvel, M., Paffaro Jr., V.A., Yamada, A.T., 2003. Uterine natural killer cells: insights into their cellular and molecular biology from mouse modelling. Reproduction 126, 149–160. 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. Dao, C.T., Zhang, D.E., 2005. ISG15: a ubiquitin-like enigma. Front. Biosci. 10, 2701–2722. Deanesly, R., 1971. The differentiation of the decidua at ovo-implantation in the guinea-pig contrasted with that of the traumatic deciduoma. J. Reprod. Fertil. 26, 91–97. DeFeo, V.J., 1967. Decidualization. In: Wynn, R.M. (Ed.), Cellular Biology if the Uterus. Appleton-Century-Crofts, New York, pp. 191–290. Emond, V., Asselin, E., Fortier, M.A., Murphy, B.D., Lambert, R.D., 2000. Interferon-tau stimulates granulocyte-macrophage colony-stimulating factor gene expression in bovine lymphocytes and endometrial stromal cells. Biol. Reprod. 62, 1728–1737. 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. Farrell, P.J., Broeze, R.J., Lengyel, P., 1979. Accumulation of an mRNA and protein in interferon-treated Ehrlich ascites tumour cells. Nature 279, 523–525. Finn, C.A., 1996. Why do women menstruate? Historical and evolutionary review. Eur. J. Obstet. Gynecol. Reprod. Biol. 70, 3–8. Finn, C.A., 1998. Menstruation: a nonadaptive consequence of uterine evolution. Q. Rev. Biol. 73, 163–173. Hansen, T.R., Austin, K.J., Perry, D.J., Pru, J.K., Teixeira, M.G., Johnson, G.A., 1999. Mechanism of action of interferon-tau in the uterus during early pregnancy. J. Reprod. Fertil., Suppl. 54, 329–339. Hunt, J.S., Petroff, M.G., Burnett, T.G., 2000. Uterine leukocytes: key players in pregnancy. Semin. Cell Dev. Biol. 11, 127–137. Johnson, G.A., Austin, K.J., Collins, A.M., Murdoch, W.J., Hansen, T.R., 1999. Endometrial ISG17 mRNA and a related mRNA are induced by interferontau and localized to glandular epithelial and stromal cells from pregnant cows. Endocrine 10, 243–252. Johnson, G.A., Stewart, M.D., Gray, C.A., Choi, Y., Burghardt, R.C., Yu-Lee, L.Y., Bazer, F.W., Spencer, T.E., 2001. Effects of the estrous cycle, pregnancy, and interferon tau on 2′,5′-oligoadenylate synthetase expression in the ovine uterus. Biol. Reprod. 64, 1392–1399. Kayisli, U.A., Selam, B., Guzeloglu-Kayisli, O., Demir, R., Arici, A., 2003. Human chorionic gonadotropin contributes to maternal immunotolerance and endometrial apoptosis by regulating Fas–Fas ligand system. J. Immunol. 171, 2305–2313. Kim, K.I., Zhang, D.E., 2003. ISG15, not just another ubiquitin-like protein. Biochem. Biophys. Res. Commun. 307, 431–434. Kimber, S.J., 2000. Molecular interactions at the maternal–embryonic interface during the early phase of implantation. Semin. Reprod. Med. 18, 237–253. 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. Licht, P., Russu, V., Lehmeyer, S., Wildt, L., 2001a. Molecular aspects of direct LH/hCG effects on human endometrium—Lessons from intrauterine microdialysis in the human female in vivo. Reprod. Biol. 1, 10–19. Licht, P., Russu, V., Wildt, L., 2001b. On the role of human chorionic gonadotropin (hCG) in the embryo-endometrial microenvironment: implications for differentiation and implantation. Semin. Reprod. Med. 19, 37–47. Lin, J., Poole, J., Linzer, D.I., 1997. Three new members of the mouse prolactin/ growth hormone family are homologous to proteins expressed in the rat. Endocrinology 138, 5541–5549. 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. Loeb, L., 1907. Wounds of the pregnant uterus. Proc. Soc. Exp. Biol. Med. 4, 93–96. Loeb, L., 1908. The production of the deciduomata and the relation between the ovaries and formation of the decidua. JAMA 50, 1897–1901.

456

B.M. Bany, J.C. Cross / Developmental Biology 294 (2006) 445–456

Lundkvist, O., Nilsson, B.O., 1982. Endometrial ultrastructure in the early uterine response to blastocysts and artificial deciduogenic stimuli in rats. Cell Tissue Res. 225, 355–364. Muller, H., Ishimura, R., Orwig, K.E., Liu, B., Soares, M.J., 1998. Homologues for prolactin-like proteins A and B are present in the mouse. Biol. Reprod. 58, 45–51. 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. Nagaoka, K., Sakai, A., Nojima, H., Suda, Y., Yokomizo, Y., Imakawa, K., Sakai, S., Christenson, R.K., 2003. A chemokine, interferon (IFN)-gammainducible protein 10 kDa, is stimulated by IFN-tau and recruits immune cells in the ovine endometrium. Biol. Reprod. 68, 1413–1421. Nieder, G.L., Jennes, L., 1990. Production of mouse placental lactogen-I by trophoblast giant cells in utero and in vitro. Endocrinology 126, 2809–2814. Paffaro Jr., V.A., Bizinotto, M.C., Joazeiro, P.P., Yamada, A.T., 2003. Subset classification of mouse uterine natural killer cells by DBA lectin reactivity. Placenta 24, 479–488. Parent, J., Chapdelaine, P., Sirois, J., Fortier, M.A., 2002. Expression of microsomal prostaglandin E synthase in bovine endometrium: coexpression with cyclooxygenase type 2 and regulation by interferon-tau. Endocrinology 143, 2936–2943. Paria, B.C., Reese, J., Das, S.K., Dey, S.K., 2002. Deciphering the cross-talk of implantation: advances and challenges. Science 296, 2185–2188. Psychoyos, A., 1960. La reaction deciduale est preccdee de modifications prococes de la permeabilite capillarie de l'uterus. C. R. Soc. Biol. 154, 1384–1387. Reich, N., Evans, B., Levy, D., Fahey, D., Knight Jr., E., Darnell Jr., J.E., 1987. Interferon-induced transcription of a gene encoding a 15-kDa protein depends on an upstream enhancer element. Proc. Natl. Acad. Sci. U. S. A. 84, 6394–6398. 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. Roberts, R.M., Cross, J.C., Farin, C.E., Hansen, T.R., Klemann, S.W., Imakawa,

K., 1990. Interferons at the placental interface. J. Reprod. Fertil., Suppl. 41, 63–74. Sherer, D.M., Abulafia, O., 2001. Angiogenesis during implantation, and placental and early embryonic development. Placenta 22, 1–13. Smith, S.K., 2000. Angiogenesis and implantation. Hum. Reprod. 15 (Suppl. 6), 59–66. Spandorfer, S., Rosenwaks, Z., 1999. Rhe impact of maternal age and ovarian age on implantation efficacy. In: Carson, D.D. (Ed.), Embryo implantation: molecular, cellular and clinical aspects. Springer-Verlag, New York, pp. 12–19. Tanaka, S., Kunath, T., Hadjantonakis, A.K., Nagy, A., Rossant, J., 1998. Promotion of trophoblast stem cell proliferation by FGF4. Science 282, 2072–2075. Thurston, G., 2003. Role of angiopoietins and Tie receptor tyrosine kinases in angiogenesis and lymphangiogenesis. Cell Tissue Res. 314, 61–68. Toft, D.J., Linzer, D.I., 1999. Prolactin (PRL)-like protein J, a novel member of the PRL/growth hormone family, is exclusively expressed in maternal decidua. Endocrinology 140, 5095–5101. Tsigkos, S., Koutsilieris, M., Papapetropoulos, A., 2003. Angiopoietins in angiogenesis and beyond. Expert Opin. Invest. Drugs 12, 933–941. Wang, D., Ishimura, R., Walia, D.S., Muller, H., Dai, G., Hunt, J.S., Lee, N.A., Lee, J.J., Soares, M.J., 2000. Eosinophils are cellular targets of the novel uteroplacental heparin-binding cytokine decidual/trophoblast prolactinrelated protein. J. Endocrinol. 167, 15–28. Weitlauf, H.M., 1988. Biology of implantation. In: Knobil, E., Neill, J.D. (Eds.), The Physiology of Reproduction. Raven Press, New York, pp. 231–262. Welsh, A.O., Enders, A.C., 1985. Light and electron microscopic examination of the mature decidual cells of the rat with emphasis on the antimesometrial decidua and its degeneration. Am. J. Anat. 172, 1–29. Wilcox, A.J., Weinberg, C.R., O'Connor, J.F., Baird, D.D., Schlatterer, J.P., Canfield, R.E., Armstrong, E.G., Nisula, B.C., 1988. Incidence of early loss of pregnancy. N. Engl. J. Med. 319, 189–194. Yamamoto, H., Flannery, M.L., Kupriyanov, S., Pearce, J., McKercher, S.R., Henkel, G.W., Maki, R.A., Werb, Z., Oshima, R.G., 1998. Defective trophoblast function in mice with a targeted mutation of Ets2. Genes Dev. 12, 1315–1326.