Placental expression of estrogen-related receptor gamma is reduced in fetal growth restriction pregnancies and is mediated by hypoxia

Abstract Fetal growth restriction (FGR) describes a fetus which has not achieved its genetic growth potential; it is closely linked to placental dysfunction and uteroplacental hypoxia. Estrogen-related receptor gamma (ESRRG) is regulated by hypoxia and is highly expressed in the placenta. We hypothesized ESRRG is a regulator of hypoxia-mediated placental dysfunction in FGR pregnancies. Placentas were collected from women delivering appropriate for gestational age (AGA; n = 14) or FGR (n = 14) infants. Placental explants (n = 15) from uncomplicated pregnancies were cultured for up to 4 days in 21% or 1% O2, or with 200 μM cobalt chloride (CoCl2), or treated with the ESRRG agonists DY131 under different oxygen concentrations. RT-PCR, Western blotting, and immunochemistry were used to assess mRNA and protein levels of ESRRG and its localization in placental tissue from FGR or AGA pregnancies, and in cultured placental explants. ESRRG mRNA and protein expression were significantly reduced in FGR placentas, as was mRNA expression of the downstream targets of ESRRG, hydroxysteroid 11-beta dehydrogenase 2 (HSD11B2), and cytochrome P-450 (CYP19A1.1). Hypoxia-inducible factor 1-alpha protein localized to the nuclei of the cytotrophoblasts and stromal cells in the explants exposed to CoCl2 or 1% O2. Both hypoxia and CoCl2 treatment decreased ESRRG and its downstream genes’ mRNA expression, but not ESRRG protein expression. DY131 increased the expression of ESRRG signaling pathways and prevented abnormal cell turnover induced by hypoxia. These data show that placental ESRRG is hypoxia-sensitive and altered ESRRG-mediated signaling may contribute to hypoxia-induced placental dysfunction in FGR. Furthermore, DY131 could be used as a novel therapeutic approach for the treatment of placental dysfunction.


Introduction
Fetal growth restriction (FGR) describes a fetus which does not reach its genetic growth potential; it affects between 5 and 10% of pregnancies. FGR is important as it has both short-and long-term consequences including stillbirth, neonatal death, neurodevelopmental delay in childhood, and increased cardiovascular or metabolic disorders in adulthood [1][2][3]. The majority of cases of FGR are mediated by abnormal placental structure and function [4]. In normal placentas, terminal villi are covered by syncytiotrophoblast and subjacent cytotrophoblast; the proliferation and fusion of which continuously renews the overlying syncytiotrophoblast to support normal placental function [5]. Reduced remodeling of maternal uterine spiral arteries into low-resistance vessels in early pregnancy is hypothesized to cause poor oxygen (O 2 ) delivery to the intervillous space and leads to placental hypoxia. This is thought to contribute to the pathogenesis of FGR, particularly in early-onset and severe cases, which are characterized by aberrant villous trophoblast turnover and placental dysfunction [4,6,7]. Indeed, we have previously shown that apoptosis is increased and proliferation is decreased in placental explants cultured in hypoxic conditions (1% O 2 ), which is reminiscent of the disordered cell turnover observed in FGR placentas [8,9].
Estrogen-related receptor gamma (ESRRG) is a gene that encodes the orphan nuclear receptor, ESRRG. It is highly expressed in metabolically active organs, including kidney, heart, liver, and placenta [10,11]. In the human placenta, the ESRRG protein is mainly located in the villous trophoblast, with the highest expression in syncytiotrophoblast [10,12]. Luo et al. [13] reported that both mRNA and protein expression of ESRRG are increased in placentas from pregnancies complicated by preeclampsia and may be involved in regulating blood pressure homeostasis in mice. Conversely, both mRNA and protein expression of ESRRG were decreased in placentas from pregnancies complicated by FGR [14,15]. ESRRG is also reported to regulate trophoblast proliferation, differentiation, and invasion [12,15]. siRNA-mediated silencing of ESRRG expression inhibited proliferation and invasion of the trophoblast cell line, HTR-8/SVneo (CVCL_7162) [15], and decreased differentiation of primary cytotrophoblast into syncytiotrophoblast [12]. Based on this evidence, it is plausible that ESRRG may play a role in regulating trophoblast function, and dysregulation of ESRRG may contribute to the placental dysfunction observed in FGR. Presently, there is no effective treatment of FGR and a better understanding of the pathophysiological mechanism(s) which drive placental dysfunction in FGR placentas, is key for therapeutic advances to be made. Therefore, understanding the mechanisms regulating ESRRG in FGR may help to identify new therapeutic targets.
There is some evidence that ESRRG expression is regulated by hypoxia, when second-trimester primary cytotrophoblasts were cultured in 2% O 2 , both the mRNA and protein level of ESRRG was decreased and the rate of differentiation into syncytiotrophoblast was reduced [12]. Furthermore, siRNA-mediated knockdown of hypoxia-inducible factor 1alpha (HIF-1alpha encoded by the gene, HIF-1A) induced the expression of ESRRG and its downstream gene CYP191.1 (cytochrome P-450) [12]. Other downstream genes of ESRRG in placental cells include hydroxysteroid 17-beta dehydrogenase 1 (HSD17B1), placenta-specific protein 1 (PLAC1), and hydroxysteroid 11-beta dehydrogenase 2 (HSD11B2) [15,16]. These products are also altered in the placenta in response to hypoxia supporting a role for ESRRG signaling in the placental response to hypoxia [17][18][19].
The ESRRG pathway can also be manipulated pharmacologically. GSK5182, a 4-hydroxytamoxifen analog, is a highly selective inverse agonist of ESRRG, due to its additional noncovalent interactions with Y326 and N346 at the active site of ESRRG [20,21]. DY131 is an agonist for ESRRG which has been used in primary cytotrophoblast and animal models to activate the expression of ESRRG [12,13,22,23]. Although past studies have shown reduced expression of ESRRG in FGR placentas [14,15], there is little knowledge as to how dysregulated ESRRG signaling is linked to the placental dysfunction observed in FGR [24]. Moreover, the potential for modulation of ESRRG pathway has not been explored in the placental explant model. In this study, we hypothesized that the reduced expression of ESRRG observed in FGR is mediated by hypoxia, and that agonists of ESRRG could rescue hypoxic changes in cultured placental explants.

Materials and methods
All the reagents were obtained from Sigma (Sigma, UK) unless specified.

Placental collection
This study was approved by the Research Ethics Committee (08/H1010/55+5) and informed consent was obtained from all participants in Saint Mary's Hospital, Manchester. All births included in this study were by Caesarean section to eliminate the effects of oxidative stress which occur during normal labor.
In our study, appropriate for gestational age (AGA) (n = 14) was defined as an individualized birth weight ratio (IBR) between the 10th to 90th centile and FGR (n = 14) was defined as an IBR below the 5th centile. Participant characteristics are shown in Table 1. Women with pregnancy complications such as chronic hypertension, renal disease, preeclampsia, diabetes, gestational diabetes, collagen vascular disease, premature rupture of membranes, and pregnancies complicated with fetal anomalies or chromosomal abnormalities were excluded. Although preeclampsia can also be associated with uteroplacental hypoxia [25,26], there are additional (and potentially) confounding maternal pathologies involved, thus, placentas from pregnancies complicated by preeclampsia were not included in this study. Term placentas (n = 15) from uncomplicated pregnancies were collected for experiments with cultured villous explants (participant characteristics are shown in Table 2). All placentas were collected within 30 min of delivery.

Treatment with GSK5182 and DY131
Villous explants from uncomplicated pregnancies (n = 7) were maintained in 1.5 mL culture medium, composed 1:1 of Dulbecco's modified Eagle medium (Gibco, UK)/Ham's F12 (Gibco, UK) supplemented with penicillin (0.6 mg/L), streptomycin (100 μg/mL), L-glutamine (0.292 g/L), and 10% fetal calf serum (Gibco, UK) at 21% O 2 (5% CO 2 , 95% air at 37 • C) for 24 h of culture, as previously described [30]. Culture media was replenished and half of the explants were transferred to the hypoxic incubator (1% O 2 , 5%CO 2 at 37 • C; 24 h), whilst the other half remained at 21% O 2 for a further 24 h. Media was then replaced with vehicle (0.75% (v/v) Dimethyl Sulfoxide (DMSO)), GSK5182 (20 or 50 μM) or DY131 (20 or 50 μM) and cultured at 21% O 2 or 1% O 2 for an additional 48 h without changing culture medium. In each experiment, three villous explants randomly sampled from different sites across the placenta were cultured in the same netwell. Each treatment was replicated in four netwells. The villous explants and conditioned culture medium were harvested at day 4 of culture. At this point, all explants were pooled in one tube, then were randomly selected for RNA extraction, protein extraction or processing for immunohistochemistry.
RNA extraction and reverse transcription polymerase chain reaction (RT-PCR)  a handhold homogenizer (SHM1, UK)) and a miRNeasy mini kit (QIAGEN, Germany) was used to extract the total RNA, according to manufacturer's instructions. An Affini-tyScript cDNA synthesis kit was used for reverse transcription of mRNA to cDNA, following the manufacturer's instructions (Agilent Technologies, UK). The PCR primer sequences (Eurofins, UK) for HIF-1A, VEGF, ESRRG, its downstream genes, and RPLP0 (60S acidic ribosomal protein P0, housekeeping gene) are listed in Supplementary Table 1. Because the mRNA expression of RPLP0 was stable in human placental tissue [15], it was used as an internal control. The generated cDNA was amplified by PCR by using a powerup SYBR Green (a dsDNA-binding dye) kit (Thermo Fisher Scientific, USA) in the Applied Biosystems Step-one system (Thermo Fisher Scientific, USA). The fold expression was calculated by the 2 − CT method.

Protein preparation and western blotting
Protein from villous explants or placental tissue were extracted using Radioimmunoprecipitation assay buffer (RIPA buffer) supplemented with a protease inhibitor complex (1% v/v), phosphatase inhibitor II (1% v/v), and phosphatase inhibitor III (1% v/v). A BCA protein assay (Thermo Fisher Scientific, USA) was used to quantify the protein concentration in the placental lysates.
Around 30 μg of protein from fresh placental tissue supernatants or 120 μg from placental explant supernatants was loaded into a 10% sodium-dodecyl sulfate polyacrylamide gel. The gel was run at 120 V for 80 min and then transferred to a polyvinylidene difluoride (PVDF) membrane (Millipore, Sigma, UK) at 120 V for 80 min. After blocking in 3% (w/v) milk (Marvel, UK) in PBS for 1 h at room temperature, the membrane was incubated with an anti-ESRRG (ab128930, 0.12 μg/mL Abcam, UK) primary antibody at 4 • C overnight. Beta-tubulin (ab6046, 0.1 μg/mL Abcam, UK) or Beta-actin (20536-1-AP, 0.09 μg/mL, Proteintech, UK) was used as the housekeeping protein. After washing, the PVDF membrane was incubated with a fluorescent-conjugated secondary antibody (IRDye green, 0.05 μg/mL, LI-COR Bioscience, USA) at room temperature for 1 h. Antibody binding and band intensity was quantified using a Licor Odyssey (LI-COR Bioscience, USA) and Image studio lite software (Image studio lite version 5.2, LI-COR Bioscience, USA). The target protein expression was normalized to house-keeping protein.

Analysis of human chorionic gonadotropin secretion and lactate dehydrogenase
Human chorionic gonadotropin (hCG) was used as a marker of cytotrophoblast differentiation. hCG in the explantconditioned culture medium was measured by ELISA (DRG Diagnostics, Marburg, Germany). Villous explants collected at day 4 were lysed in 0.3 M NaOH to provide a value for total protein content, measured using a BioRad protein assay (Bio-Rad Laboratories, Hempstead, UK). hCG secretion was expressed as mIU/mL/h/mg protein. Lactate dehydrogenase (LDH) is an enzyme for conversion of lactate to pyruvate in live cells, it is released from necrotic cells, and is considered a marker for cell viability [31]. LDH levels in explantconditioned culture medium were measured by a cytotoxicity detection kit (Roche Diagnostics, Mannheim, Germany) based on the manufacturer's instructions. LDH release was expressed as absorbance units/mg protein/h.

Immunohistochemical staining
5 μm sections from cultured villous explants or freshly collected placental tissues were transferred onto the slides precoated with poly-l-lysine. After dewaxing and dehydration, the slides were treated for antigen retrieval by microwave boiling for 5 min twice at full power (800 W) in 0.01 M citrate buffer (pH 6.0) or Tris/EDTA buffer (pH 9.0) and further incubated with 3% (v/v) hydrogen peroxide for 10 min to block endogenous peroxidase activity. Slides were incubated with non-immune block (10% goat serum and 2% human serum in 0.1% TBST (Tris-Buffered Saline-Tween-20)) for 30 min at room temperature. Sections were incubated with a monoclonal antibody against HIF-1alpha (Abcam, ab51608, 10 μg/mL), ESRRG (Abcam 215947, 10 μg/mL), Ki67 (Dako, 0.174 μg/mL), or M30 (Roche, 0.132 μg/mL) overnight at 4 • C. Matched concentrations of isotype-specific non-immune rabbit or mouse IgG were used as a negative control. After washing, biotin-conjugated goat anti-mouse or anti-rabbit antibodies (Dako-Cytomation, 3.85 μg/mL) were applied and incubated for 30 min at room temperature as a secondary antibody. After the incubation with avidinperoxidase (5 μg/mL) for 30 min, chromogenic substrate diaminobenzidine (DAB; Sigma-Aldrich, UK) was applied to the sections for between 2-10 min. Color development was monitored under a microscope. Harris's hematoxylin (Sigma-Aldrich, UK) was used to counterstain all slides for 5 min, then was further differentiated with acid alcohol for 2 s. All villous explants and placental tissues were stained in the same batch for comparison; no immunoactivity was observed on the negative controls.

Statistical analysis
All data are presented as the mean ± standard deviation (SD) (normally distributed) or median ± interquartile range (IQR) (non-normally distributed); a Shapiro-Wilk normality test was used to determine whether the data were normally distributed. GraphPad Prism version 8.0.1 (GraphPad Software, USA) was used to undertake the statistical analysis. Data were assessed using an unpaired t test or one-way analysis of variance (ANOVA) for normally distributed data or using a Mann-Whitney U test or Kruskal-Wallis, followed by a Friedman multiple comparison test for non-parametric data. QuPath (version 0.2.3, developed by the University of Edinburgh [32]), was used to quantify the immunostaining. DAB staining was quantified to measure the extent of HIF-1alpha, Ki67, and M30 staining; this was expressed as a percentage of the total number of cells. A P value <0.05 was considered indicative of statistical significance.

Reduced expression of ESRRG and its downstream genes in FGR placentas
The mean ESRRG mRNA expression level in the FGR group was 34.1% lower than the AGA group ( Figure 1A). The mean ESRRG protein level in FGR placentas was reduced, by almost 53.5%, compared with AGA placentas (Figure 1B and C). Immunostaining for protein expression of ESRRG ( Figure 1D) in AGA placentas (b) and FGR placentas (c) suggest the expression of ESRRG protein is mainly localized within the syncytiotrophoblast and stromal cells. The mRNA expression level of four genes downstream of ESRRG, HSD17B1, HSD11B2, CYP19A1.1, and PLAC1 were also assessed in FGR and AGA placentas ( Figure 1E). The median mRNA level of HSD17B1 and PLAC1 was not reduced in FGR compared to that observed in AGA placentas ( Figure 1E(a) and (d)); however, the median mRNA levels of HSD11B2 and CYP191.1 were significantly decreased by 29 and 25% in FGR placentas, respectively, when compared to AGA placentas ( Figure 1E(b) and (c), P < 0.05).

Hypoxia reduces ESRRG signaling pathway in cultured villous explants
The mRNA and protein expression and localization of ESRRG was comparable between explants cultured in 21% O 2 and 6% O 2 (Supplementary Figure 1A-D). mRNA expression of ESRRG's downstream genes was comparable between explants cultured in 21% O 2 and 6% O 2 (Supplementary Figure 1E-H). Compared to the explants cultured in 21% O 2 , mRNA expression of HIF-1A was increased in the explants treated with CoCl 2 (P = 0.01), but was unchanged in the explants cultured in 1% O 2 (Figure 2A). However, the immunostaining results indicated HIF-1alpha protein expression was increased in explants cultured in 1% O 2 or treated with CoCl 2 ( Figure 2B). HIF-1alpha immunostaining was rarely observed in explants cultured in 21% O 2 ( Figure 2C(b)) but was present in the nuclei of the cytotrophoblasts and stromal cells in explants treated with CoCl 2 (Figure 2C(c)) or cultured in 1% O 2 ( Figure 2C(d)). VEGF mRNA expression was also increased in the explants cultured in 1% O 2 or treated with CoCl 2 (Supplementary Figure 2), confirming activation of hypoxiaresponsive signaling pathways. Median ESRRG mRNA expression was significantly decreased by 51.6% in the explants maintained in 1% O 2 ( Figure 2D) and decreased by 53.7% in placental villous explants exposed to CoCl 2 ( Figure 2D). Western blotting showed that total tissue ESRRG protein expression was unchanged across treatment groups ( Figure 2E and F), immunohistochemical staining highlighted that compared to explants cultured at 21% O 2 ( Figure 2G(b)), ESRRG protein expression was reduced in the stroma and trophoblast of explants exposed to CoCl 2 ( Figure 2G(c)) or 1% O 2 ( Figure 2G(d)).

GSK5182 and DY131 rescued hypoxia-mediated alterations in ESRRG and its downstream genes
Under hypoxic conditions, GSK5182 and DY131 reversed the low expression of ESRRG; GSK5182 and DY131 did not alter ESRRG mRNA expression in hypoxia ( Figure 3A), but GSK5182 (50 μM) and DY131 (50 μM) increased the protein expression of ESRRG by 1.9-fold and 1.6-fold, respectively ( Figure 3B). For ESRRG's downstream genes, compared to the explants cultured in 1% O 2 DMSO, their mRNA expression was increased in explants cultured in 21%

DY131 increases the number of cells in cycle and reduces apoptosis induced by hypoxia
There was no difference in the percentage of cells in cycle, or apoptotic cells, between the explants cultured at 21% O 2 and 6% O 2 (Supplementary Figure 3A and B). Compared to villous explants cultured in 21% O 2 , the number of cells in cycle was significantly decreased, and apoptotic cells were significantly increased in villous explants cultured in 1% O 2 or treated with CoCl 2 (Supplementary Figure 4A

DY131 reduces hypoxia-induced necrosis of villous explants
Compared to explants cultured in 21% O 2 , hCG secretion was slightly decreased in the explants cultured in 6% O 2 (Supplementary Figure 7), and LDH levels were similar between the 21% O 2 group and 6% O 2 group (Supplementary Figure 7). The mean hCG levels at day 4 of culture were 55.3% or 50% lower following exposure to 1% O 2 or treatment with CoCl 2 (Supplementary Figure 8A and B). Compared to the explants cultured in 1% O 2 with DMSO, neither GSK5182 nor DY131 altered hCG secretion ( Figure 6A).
From day 2 to day 4 of culture, there was limited LDH release from the explants cultured in 21% O 2 (Supplementary Figure 8C); however, at day 4 of culture in 1% O 2 , the mean LDH release was modestly increased (Supplementary Figure 8D) showing a reduction in tissue integrity in this group. This increase in LDH levels normalized in tissue treated with DY131 (20 μM) at 1% O 2 ( Figure 6B). A reduction in LDH levels was also seen in explants cultured at 21% O 2 that were treated with DY131 ( Figure 6B).

Discussion
This study shows that both mRNA and protein expression of ESRRG and the mRNA expression of its downstream genes are decreased in placentas from pregnancies complicated by   FGR, and this can be reproduced by culturing healthy placental explants in 1% O 2 or exposing them to CoCl 2 . An agonist of ESRRG, DY131, rescued the abnormal cell turnover induced by hypoxia, by modulating ESRRG signaling. These findings suggest that the ESRRG pathway is dysregulated in FGR and may mediate some of the downstream effects of placental hypoxia, a key contributor to placental dysfunction (Figure 7).

Reduced ESRRG signaling in human placentas is related to FGR
Our assessment of AGA and FGR placentas confirms that the mRNA and protein level of ESRRG is reduced in FGR placentas, which is consistent with previous studies [14,15]. These studies used placental samples from southern Chinese and French populations which defined FGR as estimated fetal weight or birth weight less than the 10 th percentile. Most of  our samples were from Caucasians in the UK, and a more stringent definition of FGR (IBR less than 5th percentile) was applied. The consistent nature of this finding in three different populations suggests that a reduction in ESRRG expression is observed in FGR.
To further investigate the ESRRG signaling pathway, we quantified the expression of four genes downstream of ESRRG. We found the mRNA expression of HSD11B2 and CYP19A1.1 were significantly decreased in FGR placentas. CYP191.1 encodes the aromatase P450 which plays a role in the conversion of C19 steroid precursors into estrogen. Data regarding CYP191.1 expression in FGR placentas are inconsistent. One previous study of FGR placentas (IBR below 10th centile) found increased expression of CYP191.1 [33]. It is possible that the reduced expression level of CYP19A1.1 observed in our study might be related to the severity of FGR. HSD11B2 encodes an enzyme related to the conversion of active cortisol to inactive cortisone, which is expressed in villous syncytiotrophoblast; placental HSD11B2 level is correlated with fetal weight and postnatal growth velocity [34][35][36]. Similar to a previous study [36], our results also revealed reduced mRNA expression of HSD11B2 in FGR placentas.
HSD17B1 encodes a steroidogenic enzyme responsible for converting estrone to 17 beta-oestradiol. Previous reports have found that HSD17B1 is decreased in FGR placentas (IBR less than 10th centile) and activated by ESRRG in the HTR-8/SVneo cell line [15]. PLAC1 was mainly expressed in syncytiotrophoblast and increased in FGR placentas. However, our results did not show a statistical difference in the mRNA expression of HSD17B1 and PLAC1 in FGR placentas and this may be due to the difference in race and the definition of FGR placentas among these studies.

The ESRRG signaling pathway is hypoxia-responsive
We used a four-day explant culture at 1% O 2 or treatment with CoCl 2 to mimic the placental hypoxia observed in FGR. This model has previously been well characterized and reproduces aspects of FGR, including reduced trophoblast proliferation and increased apoptosis [8,9,37]. Both hypoxia and CoCl 2 (a chemical hypoxia mimic) can stabilize HIF-1alpha (encoded by HIF-1A) [38,39], which is also elevated in FGR placentas [40,41]. HIF-1 is a transcription factor that is specifically activated by hypoxia; it is a heterodimer composed of the HIF-1alpha and the HIF-1beta subunits [42]. HIF-1alpha is degraded under normoxic conditions, but rapidly accumulates in hypoxic conditions, where it can combine with HIF-1beta to transcriptionally modulate the expression of HIF-1 responsive downstream genes [42,43]. CoCl 2 treatment mimics a hypoxic microenvironment by inhibiting HIF-1alpha degradation, and 200 μM CoCl 2 has effectively induced HIF-1alpha in cultured placental explants [27]. As expected, mRNA expression of HIF-1A was significantly increased in explants treated with CoCl 2 . We also observed increased immunostaining of HIF-1alpha protein in explants exposed to CoCl 2 or 1% O 2 , but not in explants cultured in 21% O 2 . Although a previous study reported 200 μM CoCl 2 effectively mimicked a hypoxic environment in the placental explants [27], to the best of our knowledge it is unclear whether 200 μM CoCl 2 is equivalent to culture in 1% O 2 , which might explain the differences observed between the two culture conditions in this study. Kumar et al. showed high levels of HIF-1alpha protein expression can downregulate both ESRRG mRNA and protein expression in primary second-trimester cytotrophoblast cultured in 2% O 2 [12]. The current study used an in-vitro placental explant model and observed similar findings following treatment with CoCl 2 or culture in 1% O 2 , specifically, increased protein expression of HIF-1alpha which correlated with reduced expression of ESRRG expression.
Both treatments (CoCl 2 and 1% O 2 ) reduced hCG secretion and modestly increased LDH release, compared to control explants cultured at 21% O 2, suggesting these pathways might be regulated by the O 2 -sensitive transcription factor, HIF-1alpha. In common with previously reported findings in second trimester placental tissue, there was lower ESRRG expression at both the mRNA and protein level [12]. As ESRRG localizes to the syncytiotrophoblast layer and stromal cells in placental explants, one could hypothesize that it is involved in regulating trophoblast function and cell turnover. However, the levels of ESRRG protein measured by Western blotting after four days of culture were much lower than in fresh tissue and were not significantly decreased at 1% O 2 . This suggests that hypoxic culture may exert compensatory effects on the translation of the ESRRG pathway in the villous explant model, and there also could be other regulators of the signaling pathways involving ESRRG in hypoxia; to further explore this pathway, the identification of ESRRG's downstream effectors is needed.
All of the genes downstream of ESRRG measured here were reduced following culture in 1% O 2 or following treatment with CoCl 2 . Previously CYP191.1 has been validated as a hypoxia-responsive downstream effector of ESRRG in primary second-trimester cytotrophoblast [12]. Our findings provide further evidence for the regulation of CYP191.1 by ESRRG; however, to the best of our knowledge, our data are the first to suggest that ESRRG might mediate the expression of HSD11B2 in hypoxia. The reduced expression of HSD11B2 in hypoxia has been established previously in vitro and in vivo studies of pregnancy [44,45]. Homan et al. found a low promoter activity in HSD11B2 in a primary cytotrophoblast model under hypoxic conditions (1% O 2 ) [18], suggesting that this may be in response to a lack of an upstream stimulus. As we report low mRNA expression of HSD11B2 in our explant model in both 1% O 2 and CoCl 2 groups, it is possible that low levels of HSD11B2 are mediated by hypoxia-induced reductions in the ESRRG signaling pathway. To determine whether these effects are HIF-dependent would require further experiments that reduce HIF-1alpha, for example, using siRNA, to determine whether the changes in ESRRG signaling persist.

GSK5182 and DY131 rescue hypoxia-induced alterations in the ESRRG signaling pathway
In this study, GSK5182 and DY131 rescued the reduction in ESRRG expression, and its downstream genes induced by hypoxia. Interestingly, we observed that GSK5182, an inverse agonist for ESRRG, can induce ESRRG protein expression, as well as that of its downstream genes, in hypoxia. In previous studies using cell lines, GSK5182 was shown to inhibit the transcription of ESRRG [46,47]. However, the cultured villous explant model in the current study had very low protein expression of ESRRG. Therefore, it is possible that GSK5182 could not further reduce the gene expression of ESRRG, but instead induced its expression in hypoxia by activating coactivators or other regulators of ESRRG. Further studies are needed to uncover this unexpected effect of GSK5182 in placental villous explants. DY131, an agonist of ESRRG, can induce the expression of ESRRG in animal models or primary cytotrophoblast [13,22,23,48]. Our study suggests that DY131 not only increases the mRNA expression of ESRRG and its downstream genes (HSD11B2 and CYP191.1) in 21% O 2, but can also restore protein expression of ESRRG and its downstream genes (HSD11B2 and CYP191.1) in hypoxia. This further supports that both HSD11B2 and CYP191.1 are downstream genes of ESRRG in the placenta. This restoration of effect supports the hypothesis that ESRRG can regulate some of the effects of hypoxia in trophoblast, by mediating the expression of its downstream genes, HSD11B2 and CYP191. 1. Potential therapeutic efficacy of the ESRRG's agonist DY131 in the placental dysfunction The reduction in hCG secretion on day 4 of culture in 1% O 2 is consistent with previous reports of this explant model or primary trophoblast [9,37,49,50], which suggests an impaired differentiation of syncytiotrophoblast in hypoxia. Like our previous study [9], hypoxia reduced the number of cells in cycle and increased apoptosis. DY131 treatment reduced the level of cell necrosis, increased the number of cells in cycle, and reduced apoptosis in explants cultured under hypoxic conditions, but did not have any effect on hCG. These results suggest that ESRRG signaling mediates some of the downstream effects of hypoxia, which are implicated in the pathophysiology of FGR. Two studies have already identified the possibility of using DY131 as a therapeutic intervention in mice [13,51], and our observations suggest that DY131 could be investigated as a potential therapeutic agent to treat hypoxia-induced placental dysfunction. However, further studies are required to assess the safety and efficacy of DY131 in pregnant animal models by targeting ESRRG signaling in specific pregnancy complications, such as FGR. To avoid problems with reduced efficacy and offtarget effects in other organs that express ESRRG, approaches that facilitate placental-specific delivery of DY131 should be considered [52,53].

Strengths and limitations
This study has further confirmed that ESRRG signaling is reduced in placentas from FGR pregnancies, by applying a more stringent definition of FGR in a population not previously studied. Unlike the previous study [12], we firstly explored the relationship between hypoxia-mediated ESRRG signaling pathways and cell turnover in a cultured third-trimester villous explant model, which may be more physiologically relevant than isolated second-trimester primary cytotrophoblast, as cell-cell interactions between stromal cells, cytotrophoblast and syncytiotrophoblast are maintained. In addition, this is the first study to explore whether ESRRG's agonist, DY131, can restore impaired ESRRG signaling and rescue the aberrant cell turnover observed in villous explants exposed to hypoxia. But whether DY131 could target the placental dysfunction underlying FGR placentas by mediating ESRRG signals would be further investigated in the future.
Although our results are consistent with previous reports in other populations, further studies are needed to confirm whether ESRRG is reduced in other populations with FGR and whether there is a relationship with disease severity. To explore whether ESRRG mediated the placental phenotype seen in FGR, this study used a short-term culture of placental villous explants, but the time period selected was appropriate to study the effects of longer-term culture, which should also be explored.

Conclusion
This study shows that ESRRG signaling is dysregulated in FGR. The molecular mechanism underlying the regulatory role of ESRRG in FGR appears to be mediated in part through the hypoxia-sensitive ESRRG signaling pathway. An ESRRG agonist, DY131, can increase ESRRG signaling and rescue the abnormal cell turnover observed in placental explants cultured under hypoxia. Modulation of this signaling pathway offers a novel therapeutic option for the treatment of FGR.

Data availability
The data underlying this article will be shared on reasonable request to the corresponding author.