The placenta is a transitory but indispensable structure for harmonious gestation in mammalian species (Sapin et al., 2001). It is the unique zone of exchanges between fetal and maternal compartments, and it is also implicated in fundamental physiological functions during gestation such as immunological tolerance between mother and fetus, or hormonal production necessary for the maintenance of pregnancy or fetal defense against environmental attacks (Kaufmann and Burton, 1994). From implantation to parturition, the placental development and physiology need molecular coordination of several biological events such as protein-protein interactions, extracellular matrix degradation, adhesion, differentiation, apoptosis, and angiogenesis.
The cellular lineages and mechanisms involved during placentation are well described, but the related placental molecular pathways are still poorly understood (Cross et al., 2003). Among them, transcriptional coordination by nuclear receptors (Mangelsdorf and Evans, 1995) seems to be strongly involved. Indeed, we previously established that the retinoid X receptor (RXR), a ligand-inducible nuclear receptor activated by 9-cis retinoic acid, is indispensable for a harmonious mouse placental development (Sapin et al., 1997a), regulating several specific target genes (Sapin et al., 2000). In human specie, one of the three RXR isoforms, RXRα, coordinates important functions of placental physiology such as hormone production, as well as trophoblastic invasion and proliferation (Tarrade et al., 2001). Data described the signaling based on the association of RXRα with peroxisome proliferator-activated receptor (PPAR) γ as the first heterodimeric pathway implicated in human and mouse placental development and physiology (Barak et al., 1999).
Among the other nuclear receptors known to form obligate heterodimers with RXR, liver X receptors (LXRs) α (NR1H3) and β (NR1H2) regulate lipid absorption, storage, and utilization. During the last few years, LXRs have been shown to act as major sensors of intracellular concentration of sterols (Repa and Mangelsdorf, 2002). Use of lxr-deficient mice has helped to elucidate the role of LXRs in various tissues. So far, no role for these receptors has been assigned in placenta despite their presence (Bookout and Mangelsdorf, 2003). In addition, hydroxycholesterols, natural LXR ligands, show a high concentration in human placenta (Schroepfer, 2000).
To define the physiologic role of LXRs during placentation, we characterized the expression pattern of LXRα and LXRβ in human and mouse placenta, two models presenting strong affinity in term of placental evolution and physiology (Cross et al., 2003). Northern blot, in situ hybridization, and immunohistochemistry analyses during various steps of gestation indicated that both LXRs presented specific placental expressions. Hence, it could be hypothesized that these nuclear receptors for oxysterols could have some physiological roles during early human placentation, and also in the nutrition and the delivery due to the strong LXRα expression in the mouse yolk sac through fetal development.
MATERIALS AND METHODS
Mouse and human placental tissues were frozen at −80°C for RT-PCR and Northern blot experiments or placed in molds with embedding medium and frozen on the surface of dry ice for in situ hybridization and immunohistochemistry assays. Cultures of JAR, JEG-3, BeWo, and Wish cell lines were conducted using ATCC recommendations. Primary cultures of amniotic cells were realized as previously described (Neeper et al., 1990).
Sectioning and immunohistochemistry were performed as already described (Blanchon et al., 2001). In mouse, LXRα was studied using specific polyclonal antibodies raised against the 72 first amino-terminal amino acids of the protein (Volle et al., 2004). As previously reported, this antiserum does not recognize mouse LXRβ. Cryosections or fixed cells, previously cultured on Lab-Tek culture chambers (MC2, Clermont-Ferrand, France), were incubated overnight at 4°C with the mLXRα antiserum RXRα (Santacruz Biotechnology sc-553) and ABCA1 (Santacruz Biotechnology sc-20794) at 1/150, followed by incubation with antirabbit IgG FITC-conjugated secondary antibody for 1 hr at room temperature. Histological examination was performed after DAPI/4′,6′-diamino-2-phenylindole (nuclear) staining (1 min, dilution in PBS: 1/500) using a Zeiss Axiophot microscope.
RT-PCR assays were performed on human mRNA obtained with the Quickprep Micro mRNA purification kit (Amersham Biosciences, Saclay, France) using Superscript First-Strand Synthesis System (Invitrogen Life Technologies, Cergy Pontoise, France). The oligonucleotide primers to amplify human LXRα and LXRβ and β-actin (Table 1) were designated using the Primer 3 Web program. The PCR amplification was carried out using 2 μl of cDNA according to the following program: initial denaturation at 95°C for 10 min, followed by denaturation at 95°C for 45 sec, annealing at 59°C for 45 sec, extension at 72°C for 1 min, and a final extension of 72°C for 7 min in a Mastercycler (Eppendorf, Le Pecq, France). PCR products were analyzed by electrophoresis on a 2% agarose gel. PCR products were sequenced on both strands, with the same primers used for amplification and the DNA dye terminator cycle sequencing kit (Applied Biosystems, Courtaboeuf, France).
Table 1. Sequences of synthetic oligonucleotides used for the RT-PCR analysis
Sequence (5′ to 3′)
Size of the PCR product
For the Northern blots, mouse total RNA was extracted from pooled samples of three to four embryos (7 and 9 days postcoitum, or dpc) and three to four chorioallantoic placenta, yolk sac, and decidua (12 and 17 dpc) with TRIzol (Invitrogen Life Technologies) according the manufacturer's instructions. At stage 7 and 9 dpc, mRNA extraction was realized on mouse total conceptus including embryonic, extraembryonic, and maternal tissues. At stage 12 and 17 dpc, mRNA yolk sac represented the extraction of amnios, visceral yolk sac, and Reichert membranes; 20 μg of total RNA were loaded in the gel. Northern blots were hybridized with [32P]-labeled cDNA probes for mouse lxrα and lxrβ as previously described (Volle et al., 2004). Results were quantified by scanner densitometry analysis and graphed relative to β-actin expression.
In situ hybridization was performed to determine the cellular localization of LXRs as previously described (Slavin et al., 1999). The riboprobes for the human LXRs genes were labeled with digoxigenin-11-UTP (Roche Diagnostics France, Meylan, France). The cryosections or cells, previously cultured on Lab-Tek culture chambers (MC2, Clermont-Ferrand, France), were postfixed at 4°C with 4% paraformaldehyde in buffered PBS treated with diethyl pyrocarbonate for 5 min. The slides were hybridized at 60°C overnight with digoxigenin-labeled sense and antisense riboprobes. The tissue sections were further incubated with antidigoxigenin antibody overnight at room temperature. The signal generated by the antidigoxigenin antibody was further amplified using the Tyramide Signal Amplification system (NEN Life Sciences) or NBT/BCIP (Roche Diagnostics France, Meylan, France). Histological examination was performed after DAPI nuclear staining (1 min, dilution in PBS: 1/500) using a Zeiss Axiophot microscope.
RESULTS AND DISCUSSION
Lxrα and lxrβ transcripts were detected at the first stages of choriovitelline placentation (first type of functional placenta between 7.5 and 11 dpc) in mouse (Figs. 1 and 2). Their expressions were still detectable through all the gestation in mouse until parturition (Fig. 1). The detection of LXRα and LXRβ at 7 and 9 dpc during mouse development completed previous studies reporting the presence of both LXRs only after 11.5 and 13.5 dpc (Annicotte et al., 2004). During mouse development, LXRs did not show the same placental patterns of expression (Fig. 1) as reported for embryos by Annicotte et al. (2004). While lxrβ seemed to be ubiquitous in the yolk sac, the chorioallantoic placenta (second type of mouse functional placenta between 11 dpc to parturition), and decidua with a homogeneous level of expression, a higher levels of lxrα was shown in the yolk sac membranes (at least 2.5-fold compared to the placenta and decidua), suggesting a specific role of this receptor in the fetal membranes development and physiology (source of nutrients, gas, and waste exchange, for example). For that reason, analysis of the cellular localization was focused on LXRα. The presence of LXRα in the mouse yolk sac was confirmed at the protein level by immunohistochemistry (Fig. 2B) and no signal was seen in this structure from LXRα-deficient mouse used as negative control.
In mouse chorioallantoic placentation, Lxrα and Lxrβ (Fig. 1) expression could also be detected in term of transcripts (Fig. 1) and of proteins in the labyrinthine zone (Fig. 3F). They are also present in early human placentation (6th week of gestation; Fig. 4). Their expressions were still detectable through all the gestation in human placenta (Fig. 4) until parturition. LXRs are thus two new nuclear receptors shown to be expressed in placental structures during mammalian development. As for other nuclear receptors RXR, PPAR (Wang et al., 2002), retinoic acid receptor (Sapin et al., 1997b), and thyroid receptor (Kilby et al., 1998; Leonard et al., 2001), the choriovitelline and chorioallantoic placental expressions of LXRα and LXRβ were described in two different mammalian species. LXRα and LXRβ were also detected by RT-PCR in the placental cell lines JAR, JEG-3, and BeWo (Fig. 4), establishing these cellular models as useful to study molecular and metabolic implications of LXRs in placenta. Due to the strong expression of LXRα in the mouse yolk sac, the human fetal membranes of the chorioallantoic placenta were also checked. LXRα was also shown to be expressed in human amniotic membranes as revealed by in situ hybridization (Fig. 5F). The same assay performed on primary cultures of amniotic cells and established amniotic Wish cell line confirmed this expression (Fig. 5B and D). We also confirmed that LXRα protein was also present in the Wish cell line (Fig. 3J).
Although LXRs were detected very early during placentation and their expression present during human and mouse fetal development suggesting some important physiologic roles, LXRα (Peet et al., 1998) and/or LXRβ (data not shown) null mice are viable and fertile, with no apparent major defects in placentogenesis. It could be suggested that the lack of each nuclear receptor could be easily overpassed in the single knockout mice through a functional redundancy between the two LXRs during placental development. While deletion of RXR leads to death early during mouse development probably by the alteration of numerous signaling pathways involving various nuclear receptors (Sapin et al., 1997a), the knockout of some LXRs target gene was associated with placental abnormalities. Hence, abca1 null mice develop severe embryo growth retardation, fetal loss, and neonatal death due to a markedly reduced female ability to bear young (Christiansen-Weber et al., 2000). The basis for these defects appeared to be an altered steroidogenesis, a direct result of the lack of HDL-C. This gene encodes the ATP-binding cassette protein A1 involved in the regulation of the intracellular concentration of cholesterol. Interestingly, this gene was also shown to be regulated by RARγ in macrophages (Costet et al., 2003). Lastly, mouse placental expression of abca1 matches those of LXRs: lining of decidual maternal blood vessels and labyrinthine trophoblast layer [Fig. 4H in Christiansen-Weber et al. (2000)]. Indeed, we established the coexpression of RXRα, LXRα, and ABCA1 in mouse yolk sac membranes, in mouse placental labyrinth, and in human Wish cell line (Fig. 3), strongly suggesting the potential regulation of ABCA1 gene by LXRα/RXRα heterodimer in the placental environment. As described for embryo (Annicote et al., 2004), LXRs could thus have some role in lipid metabolism and/or transport (specially cholesterol) in placenta. In human, we showed that their localization correlated well with the placental part where fixation, internalization, and degradation of acetylated or oxidized low-density lipoprotein (LDL) was previously clearly established (Bonet et al., 1995). Interestingly, these lipids are known to activate LXRs in cell cultures (Venkateswaran et al., 2000). In addition, expression of the scavenger receptor of oxidized LDL was shown to mediate 40–50% of the oxidized LDL uptake in the choriocarcinoma cell line JAR (Halvorsen et al., 2001).
LXR localization at the fetomaternal interface suggests a putative role in the regulation of the immune response system between fetal and maternal immunities and in the control of the trophoblastic invasion. Indeed, oxidized LDLs were reported to inhibit trophoblastic invasion and RXR/LXR could be hypothesized to modulate this phenomenon. The detection of LXRβ was also recently reported in human extravillous cytotrophoblast cells (Pavan et al., 2004). Among the LXR target genes described so far, a number of them are expressed in the placenta, such as those encoding for SREBP-1c (Repa et al., 2000), lipoprotein lipase (Zhang et al., 2001), tumor necrosis factor-α (Landis et al., 2002), and apolipoprotein E (Laffitte et al., 2001). Nevertheless, no major alteration of the placenta was reported so far in the lxrα/β null mice. Precise analysis of their placentae could be of interest to understand the signals mediated by oxysterols and retinoids during gestation.
In summary, early detection of LXRα and LXRβ in human and mouse suggested their implications during choriovitelline and chorioallantoic placentation. The higher expression levels of LXRα in mouse yolk sac and human amniotic membranes led us to suspect its crucial role in several mechanisms occurring in placenta such as hormone production, fetomaternal tolerance, lipid metabolism, and parturition. Undoubtedly, identification of new target genes of RXRs/LXRs in placenta will enhance the discovery of their physiologic roles and could open a new field of investigation to understand obstetrical pathologies such as preeclempsia.
The authors thank Mrs. Sandrine Plantade and Christine Puchol for expert technical assistance in breeding the mice. Supported by grants from the Ministère de l'Education de la Recherche et de la Technologie (to D.H.V.) and the Institut National de la Santé et la Recherche Médicale (to G.M.).