Progesterone receptor antagonism by mifepristone impairs trophoblast stemness and promotes apoptosis through upregulation of PDCD4

Abstract

Background

Trophoblast cells are a critical component of retained products of conception (RPOC). While mifepristone is widely used as a non-invasive treatment for RPOC, its precise molecular mechanisms in trophoblast regulation remain poorly defined.

Results

Through integrated in vitro and in vivo approaches using human trophoblast stem cells (TSCs), HTR8/SVneo cells, and placental villus explants, we demonstrated that mifepristone exerts its effects predominantly via progesterone receptor (PGR) antagonism rather than glucocorticoid receptor (GR) inhibition. PGR knockdown in TSCs and trophoblast organoids impaired trophoblast stemness. RNA-sequencing of PGR-knockdown TSCs revealed upregulated apoptosis and reduced self-renewal and differentiation abilility, identifying PDCD4 as a key downstream target. Functional experiments showed that PDCD4 overexpression recapitulated the mifepristone-induced trophoblast dysfunction, including diminished proliferation, migration, invasion, and stemness, as well as increased apoptosis. In vivo, mifepristone administration in pregnant mice elevated PDCD4 expression, enhanced placental apoptosis, and facilitated clearance of conception products.

Conclusions

Our findings reveal that mifepristone impairs trophoblast function by antagonizing PGR and inducing PDCD4, thereby impairing stemness and promoting apoptosis. This mechanistic insight not only advances our understanding of mifepristone’s action in RPOC treatment but also suggests broader clinical implications for targeting trophoblast function.

Supplementary Information

The online version contains supplementary material available at 10.1186/s13578-025-01508-5.

Background

Retained products of conception (RPOC) refer to persistent intrauterine placental or fetal tissue that remains after various types of gestational events, including miscarriage, mid-trimester termination, and term delivery [1–3]. RPOC can lead to a range of reproductive complications, including endometritis, intrauterine adhesions, uterine perforation, severe hemorrhage, and infection, all of which may seriously impair reproductive health and fertility [4–8].

Clinically, RPOC is typically managed through surgical evacuation or pharmacological intervention. As a non-invasive option, mifepristone has demonstrated clinical benefits in RPOC management by enhancing the spontaneous expulsion of retained tissue, reducing the risk of hemorrhage, and decreasing the need for surgical procedures [9, 10].

As a synthetic steroid, mifepristone primarily exerts its pharmacological effects through antagonism of progesterone (PGR) and glucocorticoid receptors (GR) [11]. PGR plays a pivotal role in regulating the female reproductive cycle, maintaining pregnancy, and directing trophoblast differentiation and function [12], while GR is broadly involved in cellular stress responses, metabolism, and immune regulation [13]. By competitively inhibiting PGR, mifepristone disrupts progesterone signaling, thereby impairing endometrial receptivity and decidualization, and promoting cervical dilation—mechanisms that underpin its clinical use in early pregnancy termination [14–17]. Moreover, mifepristone suppresses progesterone-driven stromal activation and extracellular matrix remodeling, which contributes to symptom relief in conditions such as adenomyosis and endometriosis [18]. Through GR antagonism, mifepristone also modulates the hypothalamic–pituitary–adrenal axis, forming the basis for its application in hypercortisolemic disorders such as Cushing’s syndrome [19]. Although trophoblast cells possess strong self-renewal, proliferative, and invasive capacities that are tightly regulated by steroid hormone signaling, particularly via progesterone pathways [20, 21], it remains unclear which receptor pathway, PGR or GR, is primarily responsible for mediating the effects of mifepristone, a known antagonist of both receptors, on trophoblast cells, as this has not been explicitly investigated.

To date, the effects of mifepristone on trophoblasts remain poorly understood. A few studies have reported histological alterations in placental villi following mifepristone-induced miscarriage, suggesting potential trophoblast apoptosis, but the molecular mechanisms involved remain unclear [22, 23]. Thus, elucidating whether and how mifepristone disrupts PGR signaling to influence trophoblast fate will help clarify its mechanism of action in RPOC treatment and support its potential application in other trophoblast-related pathologies.

In this study, we hypothesized that mifepristone might induce trophoblast dysfunction through dysregulation of the PGR-PDCD4 signaling axis. To explore this possibility, we employed a suite of advanced, relevant trophoblast models, including human placental villous tissue, villous explants, multiple trophoblast cell models, and an in vivo mouse model. Our results demonstrate that mifepristone antagonizes PGR signaling to upregulate PDCD4 expression, thereby impairing trophoblast stemness and differentiation potential, and promoting apoptosis. These findings reveal a critical role for PGR signaling in maintaining trophoblast homeostasis and establish PDCD4 as a downstream effector mediating mifepristone-induced changes in trophoblast fate, thereby advancing our understanding of its therapeutic mechanism and the broader biology of trophoblast regulation.

Methods

Patient samples

Placental villous tissue samples were collected at the International Peace Maternity and Child Health Hospital (IPMCH), Shanghai, China, between June 1, 2020, and December 31, 2024. Baseline clinical characteristics of the study patients are summarized (Tables 1 and 2), with no significant differences observed between groups. All patients had normal intrauterine pregnancies at 5–8 weeks of gestation, and intrauterine pregnancy was confirmed by ultrasound examination combined with blood or urine pregnancy tests. Villous tissues from surgical abortion were obtained from patients at 5–8 weeks of gestation who underwent vacuum aspiration under sterile conditions performed by experienced clinicians. For medical abortion, patients received oral mifepristone at a dose of 50 mg once daily for two consecutive days (total dose: 100 mg), starting in the morning of the first day at the outpatient clinic. To minimize variation in tissue expulsion time, only villous tissues expelled on the morning of the third day were collected for subsequent analysis. The tissues were immediately placed in sterile cold PBS and transported to the laboratory within 20 min to preserve viability and integrity. All samples were then processed according to experimental requirements: a portion was fixed in 10% neutral formalin for histological and immunological analyses; another portion was preserved in RNAlater (AM7021, Thermo Fisher Scientific, Waltham, MA, USA) for RNA extraction and real-time qPCR analysis; some were snap-frozen in liquid nitrogen for protein analysis (Western blotting); and the remaining tissues were dissected under a stereomicroscope into uniform explants of approximately 2–3 mm in size, which were subsequently cultured in appropriate media for explant experiments.

Table 1.

Details of baseline characteristics of all clinical samples

Sample	Age(years)	Gestational age(days)	Gravidity	Parity	HCG( ±)	Viable Fetus	Usage
Mifepristone1	37	38	2	1	+	YES	IHC/IF
Mifepristone2	39	45	5	2	+	YES	IHC/IF/HE
Mifepristone3	28	42	2	0	+	YES	IHC/IF/HE
Mifepristone4	36	37	4	2	+	YES	IHC/IF/HE
Mifepristone5	31	51	2	1	+	YES	IHC/IF/HE
Mifepristone6	34	41	3	1	+	YES	RT-qPCR/WB
Mifepristone7	34	40	2	1	+	YES	RT-qPCR/WB
Mifepristone8	33	50	2	2	+	YES	RT-qPCR/WB
Mifepristone9	23	54	1	0	+	YES	RT-qPCR/WB
Mifepristone10	26	40	1	0	+	YES	RT-qPCR/WB
Mifepristone11	39	42	3	2	+	YES	RT-qPCR/WB
Mifepristone12	35	46	2	1	+	YES	RT-qPCR/WB
Mifepristone13	32	42	2	1	+	YES	RT-qPCR/WB
Mifepristone14	36	46	2	1	+	YES	RT-qPCR/WB
Mifepristone15	26	43	1	0	+	YES	RT-qPCR/WB
Mifepristone16	35	39	3	2	+	YES	RT-qPCR/WB
Mifepristone17	41	42	4	2	+	YES	HE
Mifepristone18	34	49	2	1	+	YES	HE
Mifepristone19	25	39	1	0	+	YES	HE
Mifepristone20	36	42	2	1	+	YES	HE
Control1	30	52	1	0	+	YES	IHC/IF
Control2	33	51	3	1	+	YES	IHC/IF
Control3	36	41	1	0	+	YES	IHC/IF
Control4	36	40	2	1	+	YES	IHC/IF/HE
Control5	28	46	2	1	+	YES	IHC/IF/HE
Control6	42	40	2	1	+	YES	Explant/TSC/TO/RNA-seq
Control7	31	51	2	1	+	YES	Explant/TSC/TO/RNA-seq
Control8	31	34	1	0	+	YES	Explant/TSC/TO/RNA-seq
Control9	32	35	1	0	+	YES	Explant/TSC/TO
Control10	33	47	2	1	+	YES	Explant/TSC/TO
Control11	37	56	2	0	+	YES	Explant/TSC/TO
Control12	33	40	2	1	+	YES	Explant/TSC/TO
Control13	39	35	2	1	+	YES	Explant/TSC/TO
Control14	43	54	2	1	+	YES	RT-qPCR/WB
Control15	26	47	1	0	+	YES	RT-qPCR/WB
Control16	35	39	2	1	+	YES	RT-qPCR/WB
Control17	34	54	2	1	+	YES	RT-qPCR/WB
Control18	35	42	3	2	+	YES	RT-qPCR/WB
Control19	29	43	1	0	+	YES	RT-qPCR/WB
Control20	41	39	3	2	+	YES	RT-qPCR/WB
Control21	24	42	1	0	+	YES	RT-qPCR/WB
Control21	24	42	1	0	+	YES	RT-qPCR/WB
Control22	40	42	1	0	+	YES	HE
Control23	28	54	0	0	+	YES	HE
Control24	32	44	1	1	+	YES	HE
Control25	31	54	2	2	+	YES	HE
Control26	36	47	3	2	+	YES	HE
Control27	37	45	5	2	+	YES	HE

Abbreviations: IF, immunofluorescence; HCG, human chorionic gonadotropin; IHC, immunohistochemistry; H&E: hematoxylin and eosin staining; RT-qPCR, quantitative real-time polymerase chain reaction; WB, western blot; TSC, trophoblast stem cell; TO, trophoblast organoid

Table 2.

Baseline characteristics of study participants (Mean ± SEM)

Variable	Control group (n = 27)	Mifepristone group (n = 20)
Age (years)	33.43 ± 0.96	33.00 ± 1.13
Gestational days	44.86 ± 1.21	43.90 ± 1.02
Gravidity	1.82 ± 0.19	2.30 ± 0.24
Parity	0.79 ± 0.14	1.05 ± 0.17

Age, gestational days, gravidity, and parity are presented as mean ± SEM for the control and mifepristone groups. No significant differences were observed between the two groups

Animal care and use

The design of the animal experiments was based on previously published literature and was modified [24]. BALB/c mice (average body weight, ~ 22 g) were obtained from Gempharmatech (Gempharmatech Co.,Ltd, Jiangsu, China). All animals were housed in a temperature- and light-controlled environment (12 h light/12 h dark cycle) with free access to pathogen-free food and water. All experimental procedures were approved by the Experimental Animal Welfare and Ethics Committee of the International Peace Maternity and Child Health Hospital.

To establish pregnancy, female mice were housed overnight with proven fertile males, and the presence of a vaginal plug the following morning was designated as gestational day 1 (GD1). To investigate the effects of mifepristone on placental function during pregnancy, pregnant mice were randomly assigned to five groups: negative control (no treatment, n = 5), vehicle control (corn oil with DMSO, intraperitoneal injection, n = 20), low-dose mifepristone (0.3 mg/kg, n = 20), middle-dose mifepristone (3 mg/kg, n = 10), and high-dose mifepristone (20 mg/kg, n = 10). To further evaluate the potential rescue effect of progesterone, pregnant mice were additionally divided into four groups: mife^low (0.3 mg/kg, n = 5), mife^lowP4^low (0.3 mg/kg mifepristone + 20 ng/mouse progesterone, n = 5), mife^lowP4^middle (0.3 mg/kg mifepristone + 200 ng/mouse progesterone, n = 5), and mife^lowP4^high (0.3 mg/kg mifepristone + 2 mg/mouse progesterone, n = 5). In addition, to investigate the direct effects of mifepristone on the uterus and to serve as a non-pregnant control, 20 non-pregnant female mice were included and divided into five groups: blank control (n = 5), vehicle control (n = 5), low-dose mifepristone (0.3 mg/kg, n = 5), middle-dose mifepristone (3 mg/kg, n = 5), and high-dose mifepristone (20 mg/kg, n = 5).

In the pregnant control system, mice in the vehicle group received a single intraperitoneal injection of 200 µL corn oil on GD12, while the mifepristone-treated groups received a single intraperitoneal injection of mifepristone dissolved in 200 µL corn oil at doses of 0.3, 3, or 20 mg/kg. For the progesterone rescue experiments, mice additionally received a single subcutaneous injection at the neck of progesterone dissolved in corn oil at doses of 0.02, 0.2, or 2 mg per mouse, administered in parallel with mifepristone on GD12. The negative control (NC) group did not receive any treatment and was included only as a technical control to provide baseline reference values, not as a primary comparator for drug effects. In the non-pregnant control system, blank controls similarly received no treatment, vehicle controls received a single intraperitoneal injection of 200 µL corn oil, and mifepristone-treated mice received a single intraperitoneal injection of mifepristone at the indicated doses, following the same schedule. After 72 h of treatment, mice were sacrificed. Placental tissues were collected from pregnant mice for RNA and protein extraction or fixed in 10% neutral-buffered formalin for histological analysis, while uterine tissues were collected from non-pregnant mice and processed in the same way.

RNA sequencing and data processing

Total RNA of TSCs was extracted using TRIzol reagent (Invitrogen) and assessed with an Agilent 2200 system. Samples with RIN > 7.0 were used for library preparation with the VAHTS Universal V6 RNA-seq Library Prep Kit (Vazyme). Poly(A) + mRNA was isolated, fragmented, and reverse transcribed into strand-specific cDNA. Libraries were PCR-amplified, quality-checked, and sequenced on the Illumina NovaSeq 6000 platform (150 bp paired-end).

Clean reads were generated by trimming adaptors and low-quality bases, then aligned to the human genome (GRCh38, Ensembl 104) using STAR. Gene counts were obtained with HTSeq, and expression levels were calculated using the FPKM method.

RNA-seq based differential expression and pathway analysis

Differentially expressed genes (DEGs) were defined as those with fold change > 1.5 or < 0.667 (|log₂FC| >0.585) and FDR < 0.05. Putative mRNA target genes were predicted using the limma package in R. Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis was performed using the ClusterProfiler package. Gene Set Variation Analysis (GSVA) was conducted with the GSVA package (v1.36.2) in R, using the C5 gene set from the Molecular Signatures Database (MSigDB). GSVA scores were calculated using the kernel estimation method based on normalized RNA-seq expression data.

Isolation and culturing of trophoblast stem cells (TSCs)

The methods employed in this study were similar to those described in previous research [25]. To isolate human trophoblast stem cells from placental villi, each villus was finely minced using scalpels and then digested in a 20 mL solution consisting of equal parts ACCUMAX (7921, STEMCELL Technologies, Vancouver, BC, Canada) and TrypLE Express (12604021, Life Technologies, Carlsbad, CA, USA) with gentle shaking at 37°C for three rounds of 20 minutes each. The resulting cell suspensions were sequentially filtered through a 70 µm cell strainer. Cytotrophoblast cells were subsequently separated using immunomagnetic purification with the EasySep PE positive selection kit (18551, STEMCELL Technologies, Vancouver, BC, Canada) and a PE-conjugated anti-ITGA6 antibody (130-097-246, Miltenyi Biotec, Bergisch Gladbach, Germany). The purified cells were then seeded onto a plate coated with 5 mg/mL collagen IV (354233, Corning, NY, USA) and cultured in trophoblast stem cell (TSC) medium, which consisted of DMEM/F12 medium supplemented with 0.1 mM β-mercaptoethanol, 0.2% FBS, 0.5% penicillin/streptomycin, 0.3% BSA, 1% ITS-X, 1.5 mg/mL L-ascorbic acid, 50 ng/mL EGF, 2 mM CHIR99021, 0.5 mM A83-01, 1 mM SB431542, 0.8 mM VPA, and 5 mM Y-27632. The RNA interference was performed by infecting TSCs with lentiviruses expressing the shRNA targeting firefly luciferase (Luc) (5’-CACTTACGCTGAGTACTTCGA-3’) or human progesterone receptor(PGR)(5’-CCGGGCTGCACAATTACCCAAGATACTCGAGTATCTTGGGTAATTGTGCAGCTTTTTT-3’). Lentiviruses were produced by transfection HEK293FT cells with the pLKO.1 packaging vector and psPAX2/pMD2.G packaging vectors.

Generation of trophoblast organoids (TOs)

TSCs at 80% confluence were detached using TrypLE and rinsed twice with Advanced DMEM/F12. Approximately 3000 cells were embedded in 25 µL Matrigel and plated per well of a 48-well plate. Trophoblast organoids (TOs) were cultured in 250 µL trophoblast organoid medium (TOM), formulated with Advanced DMEM/F12, N2/B27 supplements, 2 mM Glutamax, 1% penicillin/streptomycin, 1.25 mM N-Acetyl-L-cysteine, 500 nM A83-01, 1.5 mM CHIR99021, and growth factors/regulators (50 ng/mL recombinant human EGF, 80 ng/mL recombinant human R-spondin1, 100 ng/mL recombinant human FGF2, 50 ng/mL recombinant human HGF, 10 mM nicotinamide, 2 mM Y-27632, and 2.5 mM PGE2).

Differentiation of extravillus trophoblast cells (EVT) from trophoblast organoids

Trophoblast organoids were passaged and seeded into 48-well plates. Differentiation followed a version of a previously described protocol [26]. After passaging, organoids were cultured in TOM for 3–4 days, then switched to extravillus trophoblast cell differentiation medium (EVTM), consisting of advanced DMEM/F12, 0.1 mM 2-mercaptoethanol, 0.5% penicillin–streptomycin, 0.3% BSA, 1% ITS-X, 100 ng/ml NRG1, 7.5 µM A83-01, and 4% knockout serum replacement. Once cellular outgrowth was observed (typically by day 7–10), NRG1 was removed from the medium, and culture continued for an additional 7–10 days. Both trophoblast organoids (TOs) and EVT-differentiated TOs (TO#EVT) were established from trophoblast stem cells (TSCs) derived from same donors, ensuring consistent timing for proliferation and differentiation.

Cell culture

HTR8/SVneo cell lines were obtained from Procell (Wuhan, China) and cultured in DMEM/F12 (Gibco) supplemented with 10% FBS (Sigma-Aldrich) and 1% penicillin-streptomycin (Life Technologies). HEK293FT cells (Chinese Academy of Sciences, Shanghai, China) were maintained in high-glucose DMEM (Gibco) with 10% FBS and 1% penicillin-streptomycin.

To generate PDCD4-overexpressing cells, the PDCD4 gene was cloned into a lentiviral vector (Tsingke, Beijing, China) and co-transfected with packaging plasmids into HEK293FT cells using PEI. Viral supernatants were collected after 48–72 h and used to infect HTR8/SVneo cells and TSCs. Stable cells were selected with puromycin and validated by real-time qPCR.

Explant culture

The methods employed in this study were similar to those described in previous research [27]. Placental villous tips (2–3 mm fragments) were collected and placed into a 96-well plate pre-coated with phenol-free Matrigel (356237, Corning, NY, USA). The plate was incubated at 37 °C for 30 min to allow gel solidification. Explants were then cultured in DMEM/F12 (8120254, Gibco, Thermo Fisher Scientific, Waltham, MA, USA) supplemented with 10% FBS (F8687, Sigma-Aldrich, St. Louis, MO, USA) and 1% penicillin-streptomycin. Once the villi adhered to the Matrigel and began to grow outward, they were considered fixed in place and imaged using an inverted microscope (Nikon). Cell migration (of control%) was calculated from bright-field images as the net outgrowth area (area at 24–48 h minus area at 0 h) normalized to the initial core area (A_o). Values for experimental groups were normalized to the vehicle control. For each condition, 3–5 fields were analyzed and averaged as one biological replicate.

Quantitative real-time polymerase chain reaction

Total RNA was extracted from each sample using the RNA Isolator Total RNA Extraction Reagent (R401-01, Vazyme, Jiangsu, China). cDNA was synthesized using HiScript II Q RT SuperMix (R223-01, Vazyme, Jiangsu, China) according to the manufacturer’s instructions. Quantitative real-time PCR was conducted using ChamQ Universal SYBR qPCR Master Mix (Q711-02, Vazyme, Jiangsu, China) on a QuantStudio 7 Flex system (Life Technologies). Relative mRNA expression levels were calculated using the 2^–ΔΔCt method, with GAPDH as the internal control. All reactions were performed in triplicate. All primer information in this article would be presented in supplementary information (Table 3).

Table 3.

Primers used in quantitative real-time Polymerase Chain Reaction

Primer for RT-qPCR
Gene name	Species	Forward (5'- > 3')	Reverse (5'- > 3')
GAPDH	Human	GTCTCCTCTGACTTCAACAGCG	CCACCCTGTTGCTGTAGCCAA
PGR	Human	TACCTGAGGCCGGATTCAGA	GCTCCCACAGGTAAGGACAC
MKI67	Human	AACTATGGAACTGGGATGGAGAGG	AGTGTGGTCTGGTGTCTGGAAG
TEAD4	Human	GGCAGACCTCAACACCAACATC	AGCAGACCTTCGTGGAGCAG
HLA-G	Human	CCACGCACAGACTGACAGAATG	GGTCGCAGCCAATCATCCAC
MMP2	Human	CCGTCGCCCATCATCAAGTTC	CAGCCATAGAAGGTGTTCAG
PDCD4	Human	GGGAGTGACGCCCTTAGAAG	ACCTTTCTTTGGTAGTCCCCTT
FOXC1	Human	GGCGAGCAGAGCTACTAC	TGCGAGTACACGCTCATGG
PAGE4	Human	CTCTCTGCTGACTCAAGTTCTT	CACGAATGCAACCACATCGG
NEDD4	Human	CAGGCCCTCAATCACAAGC	AGGCCCTAGATCATTGGAAGT
WNT7B	Human	GCAGGAAGGTTCTAGAGG	GTTGTACTTCTCCTTCAGC
GAPDH	Mouse	TGATGGGTGTGAACCACGAG	AGTGATGGCATGGACTGTGG
PDCD4	Mouse	CAAAGAAAGGTGGTGCAGGC	GGTCTCATCCAGGGGCAAAA
IL-1b	Mouse	TCGCAGCAGCACATCAACAAG	CCAGCAGGTTATCATCATCATCCC
IL-6	Mouse	GAGAGGAGACTTCACAGAGGATACC	TCATTTCCACGATTTCCCAGAGAAC
COX-2	Mouse	GTGCCTGGTCTGATGATGTATGC	TGAGTCTGCTGGTTTGGAATAGTTG
TNF-α	Mouse	ACGTGGAACTGGCAGAAGAGG	TGAGAAGAGGCTGAGACATAGGC

Western blot analysis

Tissue samples were washed with 0.01 M phosphate-buffered saline (PBS, pH 7.4) and stored in liquid nitrogen until use. Chorionic villi were homogenized in RIPA lysis buffer (Thermo Fisher Scientific, Waltham, MA, USA; #89900) and centrifuged at 12,000 × g for 15 min at 4 °C. Protein concentrations were determined using a bicinchoninic acid (BCA) assay, and equal amounts of protein (30 µg) were separated by 10% SDS–polyacrylamide gel electrophoresis (SDS-PAGE). Pre-stained protein molecular weight markers (Epizyme, Shanghai, China; Servicebio, Wuhan, China) were loaded in parallel. Proteins were transferred to polyvinylidene fluoride (PVDF) membranes (0.22 μm; Millipore, Burlington, MA, USA).

Membranes were blocked with 5% non-fat milk and incubated overnight at 4 °C with the following primary antibodies: rabbit anti-PDCD4 (1:1000, Abcam, Cambridge, UK), rabbit anti-caspase-3 (1:5000, Abcam; ab32351), rabbit anti-caspase-9 (1:2000, Abcam; ab202068), rabbit anti-cleaved caspase-3 (1:1000, Cell Signaling Technology, Danvers, MA, USA), rabbit anti-TEAD4 (1:1000; Genuine Biotech, Hefei, China), rabbit anti-MMP2 (1:1000; Genuine Biotech, Hefei, China), and rabbit anti-GAPDH (1:2000; Abcam, Cambridge, UK).

After washing, membranes were incubated with horseradish peroxidase (HRP)-conjugated goat anti-rabbit IgG secondary antibody (1:50000; Jackson ImmunoResearch, West Grove, PA, USA) at room temperature for 1 h. Protein bands were visualized using an enhanced chemiluminescent (ECL) detection kit (YEASEN Biotech, Shanghai, China; #36222ES60) and quantified using Fiji software (NIH, Bethesda, MD, USA). GAPDH was used as a loading control.

Histology and immunostaining

Paraffin-embedded human and mouse placental tissue samples were sectioned at 5 μm using a Leica RM2235 microtome. Hematoxylin and eosin (H&E) staining was performed according to standard protocols. For Prussian blue staining, sections were deparaffinized, rehydrated, and incubated with freshly prepared 2% potassium ferrocyanide solution containing 2% hydrochloric acid for 30 min, followed by counterstaining with nuclear fast red.

For immunofluorescence staining, sections were processed using the Double Homologous Antibody Fluorescence Labeling Kit (RecordBio) following the manufacturer’s instructions. A rabbit anti-PDCD4 antibody (1:200; Cell Signaling Technology, Danvers, MA, USA) was used as the primary antibody.

For immunohistochemistry (IHC), sections were deparaffinized in xylene and rehydrated through a graded ethanol series. Antigen retrieval was performed by microwave heating in pH 8.0 retrieval solution (RecordBio, #RC016). Endogenous peroxidase activity was blocked with 3% hydrogen peroxide for 25 min. Sections were then incubated with 0.2% Triton X-100 and 3% BSA for 1 h at room temperature, followed by overnight incubation at 4 °C with rabbit anti-PDCD4 antibody (1:200; CST). After PBS washes, sections were incubated with HRP-conjugated secondary antibodies. Signal was developed using DAB (RecordBio), and nuclei were counterstained with hematoxylin.

Organoid whole-mount Immunofluorescence staining

Trophoblast organoids (TOs) were fixed in 4% paraformaldehyde for 30 min at room temperature, permeabilized with 0.5% Triton X-100 for 30 min, and blocked with 3% BSA in PBS for 1 h. Organoids were incubated overnight at 4 °C with primary antibodies, including mouse anti-EPCAM (1:200; Abcam, Cambridge, UK) and mouse anti-HLA-G (1:200; Abcam, Cambridge, UK). After PBS washes, samples were incubated with Alexa Fluor–conjugated secondary antibodies (1:500; Invitrogen, Carlsbad, CA, USA) for 1 h at room temperature. Nuclei were counterstained with DAPI. Organoids were mounted in antifade reagent and imaged using a Nikon Eclipse Ti2-E spinning disk confocal microscope (Nikon, Japan).

Apoptosis assays

Apoptosis was assessed by both flow cytometry and TUNEL staining. For flow cytometry, cells were harvested, washed twice with cold PBS, and stained with Annexin V-FITC and DAPI using an Annexin V-FITC Apoptosis Detection Kit (Elabscience, Wuhan, China) according to the manufacturer’s instructions. Flow cytometry was performed using a Beckman Coulter CytoFLEX flow cytometer (Beckman Coulter, Brea, CA, USA), and the percentages of early and late apoptotic cells were quantified using FlowJo software.

TUNEL staining was performed using an In Situ Cell Death Detection Kit (Roche, Mannheim, Germany) following the manufacturer’s protocol. Briefly, cells or paraffin-embedded placental tissue sections were fixed in 4% paraformaldehyde, permeabilized with 0.1% Triton X-100, and incubated with the TUNEL reaction mixture for 1 h at 37 °C in the dark. Nuclei were counterstained with DAPI, and TUNEL-positive cells were visualized using a Nikon Eclipse Ti2-E spinning disk confocal microscope (Nikon, Japan). The apoptotic index was calculated as the percentage of TUNEL-positive cells.

Cell proliferation assay

Cell viability was assessed using the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) (Sigma-Aldrich, St. Louis, MO, USA) assay and the Cell Counting Kit-8 (CCK-8; A311, Vazyme, Jiangsu, China). To evaluate the effects of mifepristone, AL082D06 and progesterone, human trophoblast stem cells (TSCs) were seeded into 96-well plates at a density of 3500 cells per well, and HTR8/SVneo cells were seeded at 5000 cells per well. Once the cells reached approximately 95% confluence, various concentrations of mifepristone and AL082D06 were added. MTT and CCK8 reagent were applied to each well at 0, 24, and 48 h.

To assess the effect of PDCD4 overexpression on cell proliferation, HTR8/SVneo cells overexpressing PDCD4 (HTR8/SVneo-oe-PDCD4) and empty vector control cells were seeded into 96-well plates at a density of 1500 cells per well. After cell attachment, CCK-8 reagent was added at 0, 24 and 48 h.

Following the addition of MTT or CCK-8 reagents, cells were incubated at 37 °C for an additional hour. Absorbance was then measured at 450 nm using a microplate spectrophotometer (Infinite 200 PRO, Tecan, Männedorf, Switzerland).

Wound healing assay

The wound healing assay is a conventional in vitro method utilized to investigate collective cell migration in two dimensions. HTR-8/SVneo cells were cultured following the previously described method until reaching 90–95% confluence. A straight scratch was made on the cell surface using a 100 µl pipette tip, followed by washing with DPBS and replacing the medium with low serum medium (1% FBS + 1% P/S in DMEM/F12). Images were captured at 0 and 18 h, using an inverted microscope (Nikon).

Transwell invasion assay

The invasive capability of HTR-8/SVneo cells was assessed by their ability to traverse polycarbonate membranes with 8-µm pores (3422, Corning, NY, USA). Each well was seeded with 1 × 10⁵ cells in 100 µL of low-serum medium (1% FBS and 1% penicillin-streptomycin in DMEM/F12). For invasion assays, the upper chamber membranes were pre-coated with Matrigel. The lower chambers were filled with 500 µL of DMEM containing 15% FBS. After 24 h of incubation, cells that had migrated to the underside of the membrane were stained with crystal violet for 30 min. Stained cells were counted under an inverted microscope (Nikon).

Statistical analysis

For continuous variables, Student’s t-test was utilized when the data exhibited normal distribution and equal variances, while Welch’s t-test was employed for normally distributed data with unequal variances. Continuous variables are presented as mean ± SD/SEM. Statistical analyses were conducted using GraphPad Prism 10 (GraphPad Software Inc., La Jolla, CA, USA), with P-values < 0.05 considered statistically significant.

Results

Impaired trophoblast behavior following PGR, but not GR, antagonism by mifepristone

To investigate potential histopathological changes in trophoblasts following mifepristone exposure, we analyzed placental villous tissue from patients who underwent voluntary termination of intrauterine pregnancy at 5–8 weeks of gestation via mifepristone-treated medical abortion or surgical evacuation. Hematoxylin and eosin (H&E) staining revealed morphological differences between placental samples with and without mifepristone treatment (Fig. 1A). In placental samples from the control group (without mifepristone treatment), the trophoblast cell columns, representing the niche of trophoblast stem cells, were tightly arranged and exhibit small and hyperchromatic nuclei. The bilayered trophoblast structure was continuous and compact, and the outer syncytiotrophoblast layer exhibited intact and uniform apical microvilli. In contrast, placental samples from the mifepristone-treated group showed vacuolation within the trophoblast columns and enlarged, euchromatic nuclei. The syncytiotrophoblast layer appeared discontinuous, with with eosinophilic, vacuolated cytoplasm and loss of apical microvilli. The cytotrophoblast layer was markedly thinned.We also extracted trophoblast stem cells (TSCs) from sterile villi obtained from 5-8 - week elective abortions and cultured them in vitro to further investigate the mechanism underlying these effects. As mifepristone functions as a dual antagonist of both PGR and GR, we included a selective GR antagonist, AL082D06, in parallel functional assays to distinguish the specific receptor pathway involved (Fig. 1B). MTT assays revealed that TSCs treated with mifepristone exhibited a dose- and time-dependent decrease in viability compared to the DMSO-treated vehicle controls (Fig. 1C, left). Similarly, HTR8/SVneo cells treated with mifepristone also showed decreased viability relative to controls (Fig. 1C, right), but to a lesser extent than TSCs, possibly due to the well-established environmental sensitivity of stem cells compared to differentiated trophoblasts. In contrast, AL082D06 treatment showed no significant effect on proliferation in either TSCs or trophoblast cell lines across all tested conditions (Fig. 1B). Consistently, Cell Counting Kit-8 assays demonstrated similar reductions in cell viability, thereby providing additional validation of the inhibitory effect of mifepristone (Fig. S1A, B). We speculate that mifepristone may suppress trophoblast viability by blocking PGR-mediated signaling, which is required for decidualization and maternal–fetal immune tolerance. Moreover, we conducted a rescue experiment using the natural PGR agonist progesterone (P4). Across a range of P4 concentrations, most treatments failed to rescue the inhibitory effects of mifepristone, with only partial rescue observed at certain doses (Fig. S1C). This limited rescue is likely due to the substantially higher binding affinity of mifepristone to PGR compared with P4 [28], as well as the growth-inhibitory effects of supraphysiological P4 concentrations reported in previous studies [29]. These results further support that mifepristone primarily exerts its effects through PGR blockade. Subsequently, we cultured placental villi explants for 48 h in the presence or absence of mifepristone or AL082D06. Brightfield microscopy with image analysis showed that extravillous trophoblasts treated with mifepristone displayed significantly less outgrowth compared to vehicle controls, while AL082D06 had no obvious impact on invasion capacity (Fig. 1D, E). Together, these findings suggested that mifepristone primarily inhibits trophoblast proliferation and invasion.

Fig. 1.

Mifepristone influences trophoblast behavior through PGR rather than GR antagonism. A Representative H&E staining of placental villi in Control (n = 8) and mifepristone-treated (n = 8) tissues. Trophoblast cell column, yellow arrow; cytotrophoblast cell, red arrow; syncytiotrophoblast cell, green arrow. Scale bar, 100 µm. B MTT assays of trophoblast stem cells (TSCs, left) and HTR8/SVneo trophoblasts (right) over 48 h treatment with 1.25–20 µM AL082D06 or DMSO(Control) in vitro. C MTT assays of TSCs (left)and HTR8/SVneo trophoblasts (right) over 48 h treatment with 2.5-40 µM mifepristone or DMSO in vitro. D Representative brightfield images of cell migration in explants treated with mifepristone, AL082D06, or DMSO (Control) at 24 and 48 h of culture. Scale bar, 250 µm.E Quantitative analysis of cell migration for treatment of AL082D06 (top) and Mifepristone (Bottom) groups at 24 h and 48 h. B–E Control groups were treated with an equivalent concentration of DMSO (vehicle control).Data represent means ± SEM. **p < 0.01; ***p < 0.001; ****p < 0.0001. Abbreviations: PGR, progesterone receptor; GR, glucocorticoid receptor; H&E, hematoxylin and eosin; TSCs, trophoblast stem cells; DMSO, dimethyl sulfoxide; SEM, standard error of the mean

PGR knockdown impairs self-renewal and differentiation in trophoblast stem cells and organoids

Given that mifepristone impairs trophoblast proliferation and invasion through PGR antagonism in vitro, we next sought to mimic this antagonistic effect and investigate downstream mechanisms by performing PGR knockdown in trophoblast stem cells (TSCs) and trophoblast organoids (TOs). TSCs and TOs were derived from placental tissues collected from patients who underwent elective termination of intrauterine pregnancy at 5–8 weeks of gestation via surgical evacuation (Fig. 2A, B). PGR-knockdown (shPGR) and its control (shLUC) TSCs and TOs were subsequently generated (Fig. 2C, D), and knockdown efficiency was confirmed by real-time qPCR. Real-time qPCR analysis of the self-renewal markers, MKI67 and TEAD4, showed that PGR-knockdown led to significant downregulation compared to expression in control TSCs and TOs (TSC-shLUC, TO-shPGR) (Fig. 2A, B, E, F), indicating an impaired capacity for self-renewal capacity. Consistently, Western blot analysis further demonstrated a significant reduction of TEAD4 protein levels in TO-shPGR compared to TO-shLUC (Fig. 2H). To assess the effects of PGR knockdown on differentiation potential, we induced extravillous trophoblast (EVT) differentiation in both TO-shLUC and TO-shPGR. On day 12 post-induction, real-time qPCR revealed that PGR-knockdown organoids exhibited significantly reduced expression of EVT markers HLA-G and MMP2 compared to shLUC controls (Fig. 2G), suggesting defects in differentiation into the EVT lineage. Western blot confirmed a significant reduction of MMP2 protein expression in EVT-differentiated TO-shPGR compared with controls (Fig. 2I). Consistent with these observations, whole-mount immunostaining showed significantly reduced HLA-G fluorescence in TO-shPGR compared to TO-shLUC controls (Fig. 2J, K). These results demonstrated that PGR downregulation could disrupt the capacity for both self-renewal and differentiation, two key aspects of trophoblast stemness [30]. This may explain the morphological abnormalities observed in trophoblast cell columns, regions enriched for trophoblast stem cells, in placental tissue sections following mifepristone treatment.

Fig. 2.

PGR knockdown impairs self-renewal and differentiation in trophoblast stem cells and organoids. A Schematic workflow for deriving primary TSCs and induced TOs from clinical human placental villi samples. B Representative images of primary TSCs and induced TOs, with immunofluorescence staining showing syncytiotrophoblast marker CGB (red, inner layer), epithelial marker EPCAM (green), and nuclear marker DAPI (blue) in TOs derived from primary TSCs. C Experimental workflow for establishing PGR-knockdown (shPGR) and control (shLUC) TSCs and TOs, followed by extravillous trophoblast (EVT) differentiation. D Morphological comparison of primary TSCs, TOs, and EVT-differentiated TOs between shPGR and shLUC groups. E–F Real-time qPCR analysis of self-renewal markers (MKI67 and TEAD4) in TSCs and TOs with PGR knockdown (TSC-shPGR, TO-shPGR) versus controls (TSC-shLUC, TO-shLUC). G Real-time qPCR analysis of EVT differentiation marker (HLA-G and MMP2) expression in EVT-differentiated TOs (TO#EVT-shPGR vs TO-EVT#shLUC). H Western blot analysis of TEAD4 protein expression in TO-shPGR (n = 3) and TO-shLUC (n = 3) samples (left), with corresponding quantitative densitometry (right). I Western blot analysis of MMP2 protein expression in TO#EVT-shPGR (n = 3) and TO#EVT-shLUC (n = 3) samples (left), with corresponding quantitative densitometry (right). J-K Immunofluorescence detection of EVT marker HLA-G (green), epithelial marker EPCAM (green), and nuclear staining (DAPI, blue) in EVT-differentiated TOs, with quantitative analysis of HLA-G mean fluorescence intensity(K). Data represent means ± SEM. *p < 0.05, **p < 0.01. ***p < 0.001. ****p < 0.0001. Scale bar, 200 µm. Abbreviations: TSC, trophoblast stem cell; TO, trophoblast organoid; EVT, extravillous trophoblast; CGB, chorionic gonadotropin beta; EPCAM, epithelial cell adhesion molecule; DAPI, 4′,6-diamidino-2-phenylindole

Transcriptomic profiling reveals PDCD4-mediated apoptosis and stemness suppression upon PGR loss

To explore the mechanism through which PGR antagonism by mifepristone impairs trophoblast function, we performed RNA-seq analysis on PGR-knockdown and control TSCs. Heatmap visualization showed distinct expression profiles between PGR-knockdown and control TSCs, indicating that PGR knockdown has a broad impact on trophoblast gene regulation(Fig. 3A). A total of 730 differentially expressed genes (DEGs) were identified (FDR < 0.05 and |log₂FC| >0.585), including 365 upregulated and 365 downregulated genes in the PGR-knockdown group compared to controls (Fig. 3B). Subsequent gene set enrichment analysis (GSEA) revealed enrichment of pathways associated with pluripotency, WNT signaling, and apoptosis in PGR-knockdown TSCs (Fig. 3C–E; representative core genes are indicated in the figure legends). Given the critical role of WNT signaling in maintaining trophoblast stemness, these findings are consistent with the impaired self-renewal and differentiation phenotype observed in PGR-knockdown TSCs [31]. In parallel, apoptosis-related pathways were enriched, suggesting that loss of PGR function may also promote elevated apoptotic activity in TSCs (Fig. 3E). Apoptosis and stemness are typically inversely correlated. Apoptosis can precede and contribute to the loss of stemness by disrupting key regulatory pathways [30]. Stem cells typically resist apoptosis through enhanced DNA repair and activation of survival pathways such as WNT and Notch to preserve their self-renewal capacity [32]. Consistently, KEGG pathway analysis confirmed the downregulation of pluripotency- and WNT-related pathways and the upregulation of apoptosis-related pathways. (Fig. 3F). Based on DEG screening in combination with KEGG enrichment and GSEA results, we selected five representative candidate genes (PDCD4, NEDD4, PAGE4, WNT7B, and FOXC1) that map to key pathways associated with apoptosis, proliferation, and stemness (Fig. 3G; Table 4). PDCD4 is involved in apoptosis and p53 signaling and negatively regulates PI3K–AKT/mTOR activity [33]. NEDD4, an E3 ubiquitin ligase, modulates PI3K–AKT and Hippo–YAP pathways, thereby influencing proliferation and stemness [34, 35]. PAGE4, a cancer/testis antigen, has been linked to stress-response signaling, suppressing apoptosis [36]. WNT7B, a canonical Wnt ligand, sustains stemness and proliferation [37]. FOXC1, a transcription factor, participates in cell cycle regulation and modulates PI3K–AKT as well as Notch/Hedgehog pathways, thereby affecting proliferation and differentiation [38]. These functional associations support their identification as representative downstream genes potentially mediating the effects of PGR antagonism. Since RNA-seq was performed in TSCs that model the molecular effect of mifepristone, we next examined the expression of these candidate genes in human placental villous tissues from mifepristone-treated patients (Fig. 3H). Among the identified genes, PDCD4 exhibited the most robust upregulation at the mRNA level (Fig. 3H), suggesting its potential role as a downstream effector of PGR signaling.

Fig. 3.

Transcriptomic profiling reveals PDCD4-mediated apoptosis and stemness suppression upon PGR loss. A Heatmap of differentially expressed genes (DEGs) between shLUC and shPGR TSCs (FDR ≤ 0.05, |log2FoldChange|> 0.585 (≈ log2(1.5)), FDR < 0.05). n = 3 biological replicates per group. B Volcano plot of DEGs from RNA-seq analysis (P ≤ 0.05; |log2FoldChange|> 0.585 (≈ log2(1.5)); *n* = 3 per group). C Gene set enrichment analysis (GSEA) revealed downregulation of stem cell pluripotency pathways in PGR-knockdown TSCs, with core genes including SOX2, NANOG, and POU5F1. D GSEA showing downregulation of WNT signaling pathways in PGR-knockdown TSCs, with core genes including WNT2B, WNT3, and AXIN2. E GSEA showing enrichment of apoptosis-related pathways in PGR-knockdown TSCs, with representative core genes PMAIP1, BAK1, and FADD. F KEGG pathway analysis of DEGs, with dot size indicating gene count and color representing P-value. G Heatmap of candidate downstream regulated genes in shPGR and shLUC TSCs (n = 3 per group). Rows represent genes; columns represent individual samples. Expression values are z-score normalized. Color indicates relative expression (red: high, blue: low). H Real-time qPCR analysis of mRNA levels of candidate downstream regulated genes in placental samples from mifepristone-treated patients (n = 7) and controls (n = 5). Data represent means ± SEM. *p < 0.05, **p < 0.01. ***p < 0.001. ****p < 0.0001. Abbreviations: DEG, differentially expressed gene; TSC, trophoblast stem cell; RNA-seq, RNA sequencing; GSEA, gene set enrichment analysis; PGR, progesterone receptor; WNT, Wingless/Integrated; KEGG, Kyoto Encyclopedia of Genes and Genomes; FC, fold change; qPCR, quantitative polymerase chain reaction; SEM, standard error of the mean

Table 4.

Functional annotation of representative candidate genes identified from DEG, KEGG, and GSEA analyses

Candidate Gene	Regulation (shPGR vs. Control)	Major pathway(s) (KEGG/GSEA)	Biological process
PDCD4	Up	Apoptosis, p53 signaling	Promotes apoptosis and inhibits trophoblast proliferation
NEDD4	Up	WNT signaling, Ubiquitin-mediated proteolysis	Regulates WNT receptor turnover and cell growth
PAGE4	Down	Stem cell pluripotency, MAPK signaling	Marker of trophoblast stemness and stress response
WNT7B	Down	WNT signaling, Pluripotency-related pathways	Mediates trophoblast proliferation and differentiation
FOXC1	Down	Hippo signaling, Cell cycle regulation	Controls stem-like properties and EMT balance

Summary of representative candidate genes (PDCD4, NEDD4, PAGE4, WNT7B, and FOXC1) selected from DEG screening combined with KEGG and GSEA enrichment analyses. The table lists their regulation patterns (shPGR vs. control), major enriched pathways, and relevant biological processes in trophoblast function

Overall, these transcriptomic alterations suggest that PGR knockdown, which mimics the antagonistic effect of mifepristone on PGR, broadly disrupts multiple functional programs in trophoblasts.

Apoptosis and apoptosis-related PDCD4 expression were validated in placental villous tissues from mifepristone-treated patients

RNA-seq analysis of PGR-knockdown TSCs revealed downregulation of stemness-associated pathways and upregulation of pro-apoptotic programs. Among the candidate downstream regulators, PDCD4 was notably elevated. PDCD4 is known to promote apoptosis and inhibit proliferation and invasion [39], which is consistent with both our RNA-seq data and phenotypic observations from in vitro experiments. These findings prompted further investigation of PDCD4 expression and apoptotic activity in placental villous tissues from mifepristone-treated patients. We found that extrinsic apoptotic proteins were significantly elevated in placental samples from patients who received mifepristone treatment (Fig. 4A). We then assessed PDCD4 protein expression levels (Fig. 4B) and found a significant increase in placental villus tissue from the mifepristone group. Immunofluorescence (IF) (Fig. 4C, D) and immunohistochemistry (IHC) (Fig. 4F, G) further revealed stronger PDCD4 signal intensity and higher expression levels in villous sections from the mifepristone group compared to controls. Subcellular localization analysis showed that in the mifepristone group, PDCD4 signals were observed in both the nucleus and cytoplasm of placental villi, while in control samples, the signals were predominantly nuclear with minimal cytoplasmic distribution. Quantitative analysis confirmed that the proportion of cytoplasmic-localized PDCD4 positive cells was significantly higher in the mifepristone group compared to controls (Fig. 4E).

Fig. 4.

PDCD4 expression and Apoptosis were validated in placental villous tissues from mifepristone-treated patients. A Western blot analysis of extrinsic apoptosis-related protein(caspase-9, caspase-3, cleaved caspase-3) levels in mifepristone-treated (n = 3) and control (n = 3) placental samples (left), with quantitative analysis (right). B Western blot analysis of PDCD4 protein levels in placental tissues from the mifepristone group (n = 5) and control group (n = 5) (top), with quantitive analysis (bottom). C,D Representative images of immunofluorescence staining (IF) for PDCD4 (red) and DAPI (blue) showing cytoplasmic (green arrows) and nuclear (red arrows) PDCD4 localization in the Mifepristone Group, with quantitative analysis. E Quantitative analysis of cytoplasmic-localized positive cells (%) in Control and Mifepristone groups.F,G Representative immunohistochemistry (IHC) images of PDCD4 staining in placental villous tissue from the mifepristone and control groups, with corresponding quantification. Data represent means ± SEM. *p < 0.05, **p < 0.01. ***p < 0.001. ****p < 0.0001. Scale bar, 100 µm. Abbreviations: TSC, trophoblast stem cell; qPCR, quantitative polymerase chain reaction; IF, immunofluorescence; IHC, immunohistochemistry; PDCD4, programmed cell death protein 4; DAPI, 4′,6-diamidino-2-phenylindole; SEM, standard error of the mean

In conclusion, our results demonstrate that mifepristone administration leads to marked upregulation of PDCD4 and apoptosis-related markers in villous tissue, suggesting that PDCD4 may contribute to mifepristone-induced trophoblast dysfunction.

Overexpression of PDCD4 results in increased trophoblast apoptosis and decreased stemness, proliferation, migration, and invasion

We then established PDCD4-overexpressing (oe-PDCD4) models in the HTR8/SVneo trophoblast cell line, primary trophoblast stem cells (TSCs), and trophoblast organoids (TOs) to investigate the functional role of PDCD4 in trophoblasts. Overexpression of PDCD4 significantly increased both early and late apoptotic cell fractions compared with the empty vector control, as evidenced by Annexin V/DAPI flow cytometry (Fig. 5A). Moreover, TUNEL staining revealed a marked rise in the number of apoptotic nuclei in the oePDCD4 group (Fig. 5B). We also examined the expression of apoptosis-associated markers, including caspase-9, caspase-3 and cleaved caspase-3. Western blot analysis confirmed that PDCD4 overexpression significantly promoted apoptotic signaling, supporting its role in driving cell death (Fig. 5C).

Fig. 5.

PDCD4 overexpression promotes apoptosis and suppresses stemness, proliferation, migration, and invasion in trophoblasts. A Flow cytometric analysis of apoptosis by Annexin V–FITC/DAPI staining in HTR8/SVneo cells infected with empty vector or oe-PDCD4 lentivirus, with representative dot plots shown.B Representative TUNEL staining (Red) with DAPI counterstaining (blue) in HTR8/SVneo cells infected with empty vector or oe-PDCD4 lentivirus (left), and quantification of apoptotic nuclei (right) C Western blot analysis showing increased expression of extrinsic apoptosis-related proteins (caspase-9, caspase-3, cleaved caspase-3) in PDCD4-overexpressing HTR8/SVneo cells, indicating enhanced apoptotic activity. D Cell Counting Kit-8 (CCK-8) assay revealed significantly reduced proliferation in PDCD4-overexpressing HTR8/SVneo cells compared to vector controls. E Wound healing assay demonstrated that PDCD4 overexpression decreased the migration rate of HTR8/SVneo cells. F Transwell invasion assays showed significantly reduced invasive capacity in PDCD4-overexpressing HTR8/SVneo cells relative to controls. G, H Real-time qPCR (RT-qPCR) analysis of proliferation and stemness markers MKI67 and TEAD4 in PDCD4-overexpressing trophoblast stem cells (TSCs, G) and trophoblast organoids (TOs, H), compared to controls. I RT-qPCR analysis of extravillous trophoblast (EVT) differentiation markers HLA-G and MMP2 in EVT-induced TOs derived from empty vector controls (TO#EVT-empty vector) and PDCD4-overexpressing TOs (TO#EVT-oe-PDCD4). J Western blot analysis of TEAD4 protein expression in TO empty vector (n = 3) and TO oe-PDCD4 (n = 3) samples (left), with corresponding quantitative densitometry (right). K Western blot analysis of MMP2 protein expression in TO#EVT empty vector (n = 3) and TO#EVT oe-PDCD4 (n = 3) samples (left), with corresponding quantitative densitometry (right).L, M Whole-mount immunofluorescence staining of EVT-induced TOs to assess differentiation potential. HLA-G (red), EPCAM (green), and DAPI (blue, nuclear marker) were used to label EVT cells, epithelial cell membranes, and nuclei, respectively. Quantification of HLA-G mean fluorescence intensity is shown in (M). Data are represented as mean ± SEM. *p < 0.05, **p < 0.01. ***p < 0.001. ****p < 0.0001. Abbreviations: PDCD4, Programmed Cell Death 4; CCK-8, Cell Counting Kit-8; RT-qPCR, Real-Time Quantitative Polymerase Chain Reaction; TSC, Trophoblast Stem Cell; TO, Trophoblast Organoid; EVT, Extravillous Trophoblast; HLA-G, Human Leukocyte Antigen-G; MMP2, Matrix Metalloproteinase-2; EPCAM, Epithelial Cell Adhesion Molecule; DAPI, 4′,6-Diamidino-2-Phenylindole; SEM, Standard Error of the Mean.

CCK-8 assays showed a significant decrease in cell proliferation in the oe-PDCD4 group (Fig. 5D). Wound healing and transwell assays further demonstrated that migration and invasion capacities were both reduced in oe-PDCD4 cells compared to vector controls (Fig. 5E, F), suggesting that PDCD4 negatively regulates these trophoblast functions.

To evaluate trophoblast stemness, we assessed the self-renewal and differentiation potential of TOs derived from oe-PDCD4 TSCs. MKI67 and TEAD4, markers of self-renewal, were decreased in oe-PDCD4 TSCs and their organoids. (Figure 5F, G and I), indicating impaired self-renewal capacity.Upon induction of EVT differentiation, real-time qPCR (Fig. 5H), western blotting(Fig. 5J) and whole-mount immunofluorescence staining (Fig. 5K, L) showed significantly reduced HLA-G expression in oe-PDCD4 TSC-derived organoids, indicating a loss of differentiation potential.In our model, PDCD4 overexpression activated apoptotic signaling, which may, in turn, compromise trophoblast stemness, as stemness is often inversely associated with apoptosis [40]. These findings support the hypothesis that PDCD4 functions as a key effector regulating trophoblast fate in response to mifepristone exposure.

Mifepristone facilitates placental expulsion, activates PDCD4-mediated apoptosis, and induces uterine inflammation in mice

In order to test whether the above in vitro results were generalizable to the effects of mifepristone on trophoblasts in vivo, we mated 6–8 week-old female mice. After confirming pregnancy, pregnant mice were administered mifepristone by i.p. injection at low (0.3 mg/kg, n = 20), middle (3 mg/kg, n = 10), or high doses (20 mg/kg, n = 10) on gestational day 12 (GD12). At 72 h post-injection, placentas were collected for analysis.

Relative to the normal pregnancy group, we observed that increasing concentrations of mifepristone were associated with progressively smaller retained placental tissues(Fig. 6A, B). In addition, the number of retained placentas was significantly decreased in the mifepristone-treated groups (Fig. 6C), indicating that mifepristone can facilitate the expulsion of pregnancy products.With increasing doses of mifepristone, the low-dose group displayed evident hemorrhage at the maternal-fetal interface, whereas the middle- and high-dose groups exhibited pronounced placental resorption (Fig. 6D). We examined extrinsic apoptotic protein levels in retained placental tissues from the control and low-dose mifepristone groups. The mifepristone-treated group showed significantly higher levels of caspase-9 and cleaved caspase-3 (Fig. 6E), indicating that mifepristone promotes activation of the placental extrinsic apoptotic pathway in vivo. demonstrating that the administration of mifepristone facilitated the extrinsic apoptotic process of the placenta in mice. We further examined PDCD4 expression at both transcriptional and translational levels in mouse placental tissues. Real-time qPCR and Western blot analyses demonstrated that mifepristone treatment significantly upregulated PDCD4 expression (Fig. 6F, G), while immunohistochemical staining revealed a higher proportion of PDCD4-positive cells in mifepristone-treated roups(Fig. 6H), collectively verifying PDCD4 as a downstream effector of mifepristone at the in vivo level.

Fig. 6.

Effect of intraperitoneal mifepristone administration on placental tissues in mice. A Representative appearance of pregnant uteri from the Mifepristone group (n = 40) and the control group (n = 20) (Top). The Mifepristone group includes Low (0.3 mg/kg, n = 20), Middle (3 mg/kg, n = 10), and High (20 mg/kg, n = 10) subgroups. The control group received corn oil containing the same concentration of DMSO as vehicle(n = 5). NC (negative control, pregnant mice at the same gestational day that did not receive any drug or vehicle injection) was included as a technical control and not subjected to statistical analysis. Scale bar 1 cm.B Representative images of retained placentas under different treatments. Scale bar 1 cm.C Quantitative analysis of the number of retained placentas in the Mifepristone and control groups. Statistical analysis: Welch’s t-test. D Representative hematoxylin and eosin (H&E) staining images of pregnant uteri from control and mifepristone-treated groups (Low, Middle, and High doses). Scale bar, 1 mm. E Western blot analysis of extrinsic apoptosis-related proteins (caspase-9, caspase-3,cleaved caspase-3) in placental tissues from low-dose (Low, n = 3) and control groups (Control, n = 3), with quantitative analysis. F Real-time quantitative PCR (RT-qPCR) analysis of PDCD4 mRNA expression in low-dose (Low, n = 5) and control placental samples (Control, n = 5). G Western blot analysis of PDCD4 protein levels in low-dose (Low, n = 3) and control placental samples (Control, n = 3), with corresponding quantification. H Representative immunohistochemistry (IHC) staining images of PDCD4 in placental tissues from low-dose and control groups, with quantification of PDCD4-positive cell ratio. Data are represented as mean ± SEM. *p < 0.05, **p < 0.01. ***p < 0.001. ****p < 0.0001. Abbreviations: RT-qPCR, real-time quantitative polymerase chain reaction; IHC, immunohistochemistry; SEM, standard error of the mean; PDCD4, programmed cell death 4

To further test whether progesterone supplementation could rescue the effects of mifepristone, we established a P4 rescue model (Fig. S2A). We found that only high-dose progesterone (2 mg/mouse) was able to partially alleviate mifepristone-induced placental expulsion, while lower doses showed minimal effect (Fig. S2B) .Notably, this dose of progesterone has been reported to reverse LPS-induced miscarriage in mice, a model primarily used to validate inflammation-driven pregnancy loss [41, 42]. This suggests that progesterone exerts a strong protective effect against inflammatory abortion, whereas its ability to reverse apoptosis appears limited. Therefore, the observed protection may mainly result from suppression of inflammation rather than direct inhibition of apoptosis.At the molecular level, caspase-9 levels showed no significant changes between groups, whereas cleaved caspase-3 levels displayed partial restoration in the rescue group, indicating that progesterone supplementation may act mainly at the execution phase of apoptosis rather than the initiation stage (Fig. S2C).

Besides, injecting mifepristone on the non-pregnant uterus revealed that high-dose mifepristone markedly reduced endometrial thickness (Fig. S3A), but did not induce overt uterine bleeding (Fig. S3B). However, real-time qPCR analysis demonstrated significant upregulation of inflammation-related markers, including COX-2, TNF-α, IL-2, and IL-6, in a dose-dependent manner (Fig. S3C). Hence, mifepristone exerts a strong pro-inflammatory effect on the uterus, independent of its action on placental tissues. Therefore, the clinical use of mifepristone should also take into account its uterine inflammatory impact, and appropriate dosing strategies are essential to balance efficacy and safety.

Discussion

Retained products of conception (RPOC) may follow any type of pregnancy, with relatively high incidence across different gestational stages. If not managed in a timely manner, RPOC can lead to severe complications such as intrauterine infection, adhesions, hemorrhage, and potentially infertility [4–6]. Given the severity of these complications, investigating the mechanistic targets of mifepristone may provide a potentially effective, non-invasive intervention for RPOC. We identify mifepristone-induced antagonism of PGR and upregulation of its downstream effector PDCD4 as a key mechanism driving trophoblast dysfunction. This pathway promotes apoptosis, reduces proliferation and invasiveness, impairs stemness, and may facilitate the clearance of products of conception.

Using clinical samples, explants, trophoblast stem cells, trophoblast organoids, cell lines, and mouse models, we provide the first comprehensive evidence that mifepristone alters trophoblast biology at both molecular and organismal levels. The integration of multiple human-derived models, particularly TSCs and TOs, enables a more physiologically relevant simulation of the in vivo characteristics of human placental development and pathology.

To date, few studies have systematically examined specific impact of mifepristone on placental trophoblasts, particularly in the context of its dual antagonism of PGR and GR. Our findings demonstrate that mifepristone impairs trophoblast function predominantly through PGR signaling. Progesterone receptor (PGR), a ligand-dependent nuclear transcription factor, mediates the cellular effects of progesterone by regulating gene expression programs critical for female reproduction [43]. Upon binding to progesterone, PGR undergoes conformational changes, dimerizes, and translocates into the nucleus, where it binds to progesterone response elements (PREs) within the promoters of target genes [44]. Through recruitment of transcriptional co-regulators, PGR modulates diverse signaling pathways—including PI3K/AKT, MAPK/ERK, and Wnt/β-catenin—that control cellular proliferation, apoptosis, and differentiation [45–47]. In reproductive tissues such as the endometrium, ovary, and placenta, these molecular actions of PGR are essential for ovulation, embryo implantation, decidualization, and maintenance of pregnancy. Dysregulation of PGR signaling—via altered receptor expression, isoform imbalance (PGR-A vs. PGR-B), or aberrant co-regulator activity—has been implicated in several reproductive disorders, including infertility, endometriosis, and recurrent pregnancy loss [48]. Notably, recent studies have also linked PGR loss to endometrial aging, which impairs tissue receptivity, disrupts progesterone responsiveness, and compromises implantation potential [49]. The present study builds on this understanding by demonstrating that mifepristone exerts its effects on human trophoblasts via antagonism of the progesterone receptor (PGR). To substantiate this mechanism, we generated PGR-knockdown trophoblast stem cells (TSCs) and trophoblast organoids (TOs), which mirrored the cellular outcomes of mifepristone treatment. Our findings establish PGR as a key functional target in sustaining trophoblast stemness and regulating trophoblast behavior.

To further elucidate the downstream effector of PGR antagonism, we investigated the involvement of PDCD4, a molecule previously characterized for its tumor-suppressive roles—particularly through inhibition of AP-1, NF-κB, and PI3K/AKT signaling—and regarded as a potential target in cancer therapy [50]. However, its connection with progesterone signaling has been less frequently reported, with only one study to date demonstrating that progesterone exerts cyclic regulation of PDCD4 in endometrial epithelium [51], and another study showing that progesterone suppresses PDCD4 via the PI3K/AKT pathway in endometrial carcinoma cells [52]. Notably, emerging evidence has linked PDCD4 to the regulation of trophoblast survival and differentiation, suggesting its potential involvement in placental pathologies. Downregulation of PDCD4 via miR-21 has also been implicated in the progression of choriocarcinoma [53], while its suppression by circRNA/miRNA networks promoting trophoblast apoptosis has been identified as a factor in intrahepatic cholestasis of pregnancy [54]. Moreover, reduced PDCD4 SUMOylation has been reported as a mechanism in gestational trophoblast disease. In addition, the nuclear-cytoplasmic translocation of PDCD4 has been identified as a hallmark of cellular stress, and this mechanism is likely involved in the regulation of apoptosis [55, 56]. This mechanism may help explain why trophoblasts in mifepristone-treated placental villi predominantly display cytoplasmic PDCD4 localization, consistent with enhanced apoptotic signaling.

Our findings suggest that PGR–PDCD4 may constitute a potential regulatory axis in trophoblasts. Although no evidence currently supports direct binding of PGR to the PDCD4 promoter, the observed PDCD4 upregulation upon mifepristone treatment implies that PGR loss may indirectly relieve inhibitory signals such as PI3K/AKT/mTOR [20, 33], with additional contributions from TGF-β and NF-κB/AP-1 pathways [57]. Previous studies reporting progesterone-mediated cyclic regulation of PDCD4 in endometrium and its suppression via PI3K/AKT signaling in carcinoma cells further support this possibility [58]. Taken together, while a direct promoter-level regulation has not yet been established, these findings highlight a plausible PGR–PDCD4 axis that warrants further investigation.

Consistently, both PGR knockdown and PDCD4 overexpression in trophoblast stem cells (TSCs) resulted in impaired stemness and increased apoptosis, suggesting a convergent effect of perturbing the PGR–PDCD4 regulatory axis. Beyond indicating this functional convergence, these findings also prompted us to further examine the relationship between stemness and apoptosis, which are two interconnected processes that are essential for maintaining tissue homeostasis. Stem cells depend on self-renewal and controlled differentiation to sustain normal tissue function [59]. Disruption of this balance—either through excessive apoptosis leading to stem cell depletion, or insufficient apoptosis allowing the persistence of dysfunctional cells—can have detrimental consequences. Notably, recent studies suggest that apoptotic signaling can also influence stem cell fate decisions by triggering differentiation in specific contexts [60]. In this regard, our data indicate that mifepristone, through PGR antagonism, induces overexpression of PDCD4, a pro-apoptotic factor, which in turn activates apoptotic pathways and suppresses trophoblast stem cell self-renewal and differentiation capacity. These findings reveal a previously unrecognized mechanism of mifepristone action and provide novel insights into the interplay between apoptosis and stemness regulation in trophoblast biology.

The current study establishes a molecular connection between mifepristone, PGR, and PDCD4, thus offering a potential mechanistic basis for mifepristone administration in RPOC management. Conservative pharmacological management with mifepristone may be particularly advantageous in RPOC patients with placenta accreta spectrum, where retained tissue is often refractory to curettage and poses a risk of hysterectomy [8]. We further propose its prophylactic application following delivery or evacuation to promote timely trophoblast regression and reduce the need for invasive reintervention. Given the asymptomatic nature of early-stage RPOC, molecularly targeted preventive strategies may offer clinical benefit. PDCD4 enhances apoptosis, and antagonizes trophoblast stemness. As a progesterone receptor antagonist, mifepristone may induce PDCD4 expression, facilitating both apoptotic clearance and stemness attenuation in residual trophoblast populations. This mechanism provides a rationale for prophylactic mifepristone to mitigate persistent placental tissue and associated complications.

While this study reveals key roles of the mifepristone–PGR–PDCD4 axis in trophoblast regulation, several limitations should be acknowledged. First, our findings indicate a potential link between PGR signaling and PDCD4 expression. However, it remains unclear whether PGR directly regulates PDCD4 or whether the effect is mediated indirectly through signaling networks such as TGF-β and NF-κB/AP-1. Second, although mifepristone shows therapeutic potential, its safety as a progesterone receptor (PGR) antagonist requires careful attention. In our study, high-dose mifepristone reduced uterine thickness and increased inflammatory indicators, consistent with reports that excessive PGR inhibition may impair endometrial repair and disturb hormonal balance [61, 62]. Clinically, prolonged or repeated exposure has been associated with menstrual irregularities, thin endometrium, and even secondary infertility, underscoring the importance of evaluating uterine safety alongside efficacy [63]. Third, dose selection is therefore a critical consideration: while higher doses enhance efficacy in trophoblast inhibition and tissue expulsion, they also increase the risk of adverse uterine responses. Appropriately low-dose regimens may represent a safer strategy for RPOC management, but this requires confirmation in pharmacokinetic studies and well-designed clinical trials.

In summary, this study reveals that mifepristone impairs trophoblast function through PGR antagonism and PDCD4 upregulation, leading to increased apoptosis and reduced proliferation, invasiveness, and stemness. Using multiple models, we establish the PGR–PDCD4 axis as a key regulator of trophoblast cell fate, providing a mechanistic basis for the therapeutic potential of mifepristone in RPOC and related disorders.

Conclusions

The mifepristone-PGR-PDCD4 signaling axis offers a mechanistic basis for mifepristone’s efficacy in treating RPOC and supports its broader potential for clinical applications in which modulating trophoblast activity may be beneficial.

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviations

RPOC

Retained products of conception

PGR

Progesterone receptor

GR

Glucocorticoid receptor

TSCs

Trophoblast stem cells

TOs

Trophoblast organoids

EVT

Extravillous trophoblast

GD

Gestational day

MKI67

Marker of Proliferation Ki-67

TEAD4

TEA Domain Transcription Factor 4

HLA-G

Human Leukocyte Antigen G

MMP2

Matrix Metalloproteinase 2

PDCD4

Programmed Cell Death Protein 4

MTT Assay

A colorimetric assay for assessing cell viability

CCK-8 Assay

Cell Counting Kit-8 assay, used for evaluating cell proliferation

GSEA

Gene set enrichment analysis

KEGG

Kyoto Encyclopedia of Genes and Genomes

H&E

Hematoxylin and Eosin staining

IHC

Immunohistochemistry

IF

Immunofluorescence

RT-qPCR

Quantitative real-time polymerase chain reaction

WB

Western blot

Author contributions

Jie Zhou performed the experiments, performed RNA-seq computational analysis and wrote the manuscript. Xiaoya Zhao and Jie Zhou cultured the trophoblast stem cells. Qian Zhu cultured the villous explants. Li Yan, Yamei Li, Duo Zhang, and Qicheng Lan collected clinical samples. Suming Huang performed pathological reviews. Yamei Li, Xiaoya Zhao, and Jian Zhang provided funding. Qian Zhu, Xiaoya Zhao, and Jian Zhang designed the research and revised the manuscript.

Funding

National Natural Science Foundation of China, Grant Number: 82171667, 82371691. Shanghai Key Laboratory of Embryo Original Diseases Open Project Fund, Grant Number: shelab2023ZD01, shelab2024ZD01. National Natural Science Foundation of China (NSFC) Young Scientists Fund, Grant Number: 8240060467. Shanghai Sailing Program, Grant Number: 24YF2750400.

Data availability

All RNA sequencing data were deposited in the GEO datasets(https://www.ncbi.nlm.nih.gov/) and the GEO number is GSE265928.

Declarations

Ethics approval and consent to participate

All procedures performed in this study adhered to the ethical guidelines of the Institutional Ethics Committee of the IPMCH in Shanghai, China (approval number GKLW2019-09, date of approval: 2019.10.16). Written informed consent was obtained from all participants prior to the collection of samples.

Consent for publication

The authors affirm that human research participants provided informed consent for publication of the images in Figs. 1, 2, 3, 4 and 5.

Competing interests

The authors have declared that no conflict of interest exists.