The first-trimester human placenta tissues of RPL patients and healthy controls (HCs) were collected from women undergoing dilatation and curettage at the sixth to eighth week of gestation in the Department of Obstetrics and Gynecology of the Sir Run Run Shaw Hospital affiliated to the School of Medicine, Zhejiang University. The study was reviewed and approved by the Ethics Committee of Sir Run Run Shaw Hospital affiliated to the School of Medicine, Zhejiang University. All participants were provided with written informed consent to participate in this study.

The HCs were included women who already had a healthy pregnancy at term and opted for an elective abortion at 6-8 weeks in a subsequent apparently healthy pregnancy. Patients suffering from RPL were included if they had an empty gestational sac or embryo stoppage confirmed by type-B ultrasound scans at least twice, containing the current pregnancy. The exclusion criteria for RPL were (i) genital malformation; (ii) abnormal karyotype of parents or abortuses; (iii) endocrine or metabolic disorders; (iv) autoimmune diseases; (v) other major diseases such as hypertension and cancers; and (vi) improper drug treatment, exposure to chemicals or radiation (55). A total of 16 HC and 16 RPL first-trimester placentas were collected. The clinical characteristics of the recruited patients are provided in Table 1 .

The isolation and culture of human TS^CT cells were performed following a published protocol (18). Three first-trimester placental villi from healthy mothers were collected. The sex and developmental stage of each human sample used for the derivation of TS^CT cells are summarized in Supplementary Table SI . In brief, the first-trimester placental villi were collected and cut into small pieces. All tissues were enzymatically digested thrice in a mixture containing equal amounts of TrypLE (Thermo Fisher Scientific, USA) and Accumax (Innovative Cell Tech, USA) for 20 min at 37°C. Cell suspensions were then filtered through a 70 μm mesh filter (352350, BD Falcon, USA). ITGA6 is a widely used cell lineage marker for CTBs (56). CTBs were immunomagnetically purified using an EasySep phycoerythrin (PE)-positive selection kit (Stemcell Technologies, Canada) and a PE-conjugated anti-ITGA6 antibody (130-097-246, Miltenyi, Germany). The selected cells were seeded in a 6-well plate (Corning, USA) coated with 5 μg/mL collagen IV (354233, Corning, USA) at a density of 0.5-1×10⁶ cells per well and cultured in 2 mL of TS medium. CTBs were dissociated with TrypLE for 10-15 min at 37°C. The cells at passages 10-30 were used for analyses and differentiation assays. Information regarding the TS medium is presented in Supplementary Table SII .

For ST(2D)-TS^CT cell differentiation, TS^CT cells were collected and then seeded into a 6-well plate precoated with 2.5 mg/mL collagen IV at a density of 1×10⁵ cells per well. The ST(2D) medium was replaced at day 3, and the culture was continued for 3 days.

To assess ST(3D)-TS^CT cell differentiation, TS^CT cells were seeded in 6-cm Petri dishes and cultured in 3 mL of ST(3D) medium. An equal amount of fresh ST(3D) medium was added at day 3, and the cells were collected at day 6 after they were passed through a 40-μm mesh filter (352340, BD Falcon, USA) to remove dead cells and debris. The cells remaining on the 40-μm mesh filter were used for further research.

For the induction of EVT-TS^CT cells, TS^CT cells were seeded in 6-well plates precoated with 1 mg/mL collagen IV at a density of 0.75×10⁵ cells per well. Matrigel (354234, Corning, USA) was added to a final concentration of 2% immediately after suspending TS^CT cells in the precooled medium. The medium was replaced with EVT medium without NRG1 on day 3, and Matrigel was added to a final concentration of 0.5%. After the cells reached 80% confluence at day 6, they were dissociated with TrypLE for 15-20 min at 37°C and passaged to a new collagen IV precoated 6-well plate at a 1:2 split ratio. The cells were suspended in the EVT medium without NRG1 and KSR, and Matrigel was added to a final concentration of 0.5%. EVT-TS^CT cells were collected and analysed after two additional days. Information regarding ST(2D), ST(3D) and EVT media is presented in Supplementary Table SII .

The morphologies of TS^CT, ST(2D)-TS^CT and EVT-TS^CT cells were captured by an inverted microscope Primovert with a 10x objective using bright field microscopy (Zeiss, Germany).

The supernatant of TS cells was collected as a control. The ST(2D)-TS^CT cell medium was replaced on day 3 and collected on day 3 and 6. All supernatants were centrifuged at 10000 × g for 10 minutes to remove impurities before further research. The levels of secreted β-hCG and secreted S100P were measured using an hCG enzyme-linked immunosorbent assay (ELISA) kit (ABN-KA4005, Abnova, Japan) and a S100P ELISA kit (ABN-KA0093, Abnova, Japan), respectively. Total protein was extracted from cells using Radio Immunoprecipitation Assay (RIPA) buffer (P0013B, Beyotime, China) at 4°C for 30 minutes and measured with a Pierce Bicinchoninic Acid (BCA) Protein Assay Kit (23225, Thermo Scientific, USA). The final concentrations of secreted β-hCG and secreted S100P were normalized to the total protein concentration. Serum secreted S100P levels were measured with a S100P ELISA kit (ABN-KA0093, Abnova, Japan) following the instructions.

Total RNA samples were isolated from TS^CT and ST(2D)-TS^CT cells using an RNA-Quick Purification Kit (RN001, ES Science, China) in accordance with the manufacturer’s protocol. Total RNA samples (1 μg) were reverse-transcribed with HiScript II Reverse Transcriptase (R201-1, Vazyme, China). SYBR Green Supermix (D7260, Beyotime, China) and the CFX96 system (Bio-Rad, USA) were used for real-time PCR. The relative gene expression levels in the experimental samples were compared with those in the control samples, and all experiments were performed in triplicate. The primer sequences are listed in Supplementary Table SIII .

Protein samples were extracted from tissues and cells using ice-cold RIPA lysis buffer containing 1% phenylmethanesulfonyl fluoride (PMSF) (ST506, Beyotime, China). Total prepared protein samples were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) (PG113 and PG114, Epizyme Biotech, China) and then transferred to polyvinylidene difluoride (PVDF) membranes (Millipore, USA). After blocking with 5% skim milk (232100, BD Difco, USA) at room temperature for at least 1 hour, the membranes were incubated with primary antibodies overnight at 4°C. The following primary antibodies were used in the experiments: S100P (ab133554, 1:1000, Abcam, USA), TEAD4 (ab58310, 1:1000, Abcam, USA), β-hCG (sc271062, 1:300, Santa Cruz, USA), YAP1 (sc101199, 1:500, Santa Cruz, USA), 11β-HSD2 (sc365529, 1:500, Santa Cruz, USA), E-cadherin (3195, 1:1000, CST, USA), ZO-1 (13663, 1:1000, CST, USA), and GAPDH (60004-1-Ig, 1:1000, Proteintech, China). After washing in Tris-buffered saline with Tween 20 (TBST) buffer, membranes were incubated with horseradish peroxidase-conjugated anti-rabbit (7074, 1:3000, CST, USA) or anti-mouse IgG secondary antibodies (7076, 1:3000, CST, USA), and visualized by enhanced chemiluminescence (WBKLS0500, Millipore, USA). The images were captured using a ChemiDoc Touch Imaging System (Bio-Rad, USA). In addition, other membranes were incubated with preabsorbed goat anti-rabbit IgG H&L IRDye 800 (ab21677, 1:5000, Abcam, USA) or donkey anti-mouse IgG H&L Alexa Fluor 680 (ab175774, 1:5000, Abcam, USA) secondary antibodies and then visualized using an Odyssey Infrared Laser Imaging System (Licor Bioscience, USA). ImageJ software was used to quantify the intensities of the Western blotting bands.

Cells were fixed with 4% paraformaldehyde (MA0192, Meilunbio, China) for at least 30 min and then permeabilized with 0.3% Triton X-100 (T8787, Sigma-Aldrich, USA) for 5 min at room temperature. The frozen sections were permeabilized with 0.3% Triton X-100 for 10 min at room temperature. After blocking in 2% goat serum for 1 hour at room temperature, cells or sections were incubated with primary antibodies overnight at 4°C. The following primary antibodies were used: S100P (ab133554, 1:100, Abcam, USA), β-hCG (ab9582, 1:50, Abcam, USA), YAP1 (sc101199, 1:50, Santa Cruz, USA), TEAD4 (ab58310, 1:50, Abcam, USA), HLA-G (ab7758, 1:50, Abcam, USA), E-cadherin (3195, 1:50, CST, USA), and ZO-1 (13663, 1:50, CST, USA). After washing with TBST buffer, cells or sections were incubated with Alexa Fluor 488- or Alexa Fluor 568-conjugated goat secondary antibody (A11001; A10037; A11034; A11036, 1:500, Invitrogen, USA) for 1 hour at room temperature. The nuclei were stained with Hoechst 33342 Stain Solution (H3570, 1:1000, Invitrogen, USA) for an additional 10 min, and images were captured with a laser scanning confocal microscope (Zeiss, LSM800, Germany).

The fusion index is calculated as the ratio of the number of nuclei in the syncytia divided by the total number of nuclei, and the number of nuclei in the syncytia, defined as nuclei at least three are surrounded by a cell membrane (11). Cell boundaries were marked by ZO-1 (13663, 1:50, CST, USA). The total number of nuclei and the number of nuclei in the syncytia were counted in each slide. At least ten images from each group were captured and analyzed. All slides were counted by two independent individuals for comparison.

Paraffin slides were dewaxed with xylene for 20 min and 100%, 95%, and 75% ethanol in sequence for a total of 30 min and boiled in Antigen Retrieval Buffer (citrate buffer, pH 6.0, or Tris-EDTA buffer, pH 9.0) (ab93678 and ab93684, Abcam, USA) at 100°C for 15 minutes. The slides were blocked with 2% goat serum for 1 hour at room temperature and incubated with primary antibodies at 4°C overnight. The following primary antibodies were used: S100P (ab133554, 1:100, Abcam, USA) and β-hCG (ab9582, 1:50, Abcam, USA). After washing with TBST buffer, the slides were incubated with horseradish peroxidase (HRP)-conjugated secondary antibodies for 1 hour at room temperature (GK500710, Gene Tech, China). HRP-conjugated antibodies were detected with diaminobenzidine (DAB) (GK500710, Gene Tech, China) for 10 min. The nuclei were then stained with hematoxylin solution (ab220365, Abcam, USA) for an addition 5 min. Images were captured with an Axio Scope A1 (Zeiss, Germany).

To knockdown S100P expression, TS^CT and ST(2D)-TS^CT cells were cultured in 12-well plates and transfected with small interfering RNAs (siRNAs) targeting S100P (siS100P) or negative control siRNA (siNC) with Lipofectamine™ 3000 Transfection Reagent (L3000015, Thermo Fisher Scientific, USA) following the manufacturer’s instructions. The siRNAs were purchased from Ribo Bio Co., Ltd. (Guangzhou, China). The siS100P sequence was 5-CAAGGATGCCGTGGATAAA-3, and the negative control sequence was 5-GATCATACGTGCGATCAGA-3.

Cell lines overexpressing S100P were generated by infection with a lentiviral vector (CMV-MCS-3FLAG-Ubi-ZSGreen-IRES-Puro) carrying the S100P cDNA sequence; the vector was purchased from Guannan Bio Co., Ltd. (Hangzhou, China). TS^CT cells were seeded in six-well plates at 80-90% confluence the day before infection. The next day, TS^CT cells were infected with lentivirus in culture medium for 24 hours. Then, the medium was replaced before treatment. Approximately 80% of the cells were infected with overexpression lentivirus. Successful infection was confirmed by qPCR and Western blotting assessments of S100P levels. For long-term culture, clones were maintained in the presence of 2 µg/mL puromycin.

The recombinant human S100P protein (12635-HNAE, Sino Biological, China) consisted of 95 amino acids and had a calculated molecular mass of 10.4 kDa. The recombinant protein was first diluted with ultrapure water at room temperature and then added to TS^CT cell culture medium at concentrations of 5, 10, 15, and 20 μM for 24 hours.

TS^CT cells were seeded into 10-cm cell culture dishes and differentiated into ST(2D)-TS^CT cells at a 1:1 ratio. On day 1, the ST(2D)-TS^CT cells were treated with siRNA targeting S100P or negative control as described above. After 4 days, the cells were harvested and total RNA from each group was extracted with TRIzol reagent (Vazyme, China). Three replicate samples were collected for each group. The samples were then processed for reverse transcription, cDNA purification and library construction by Applied Protein Technology company (Shanghai, China). The results were analyzed on the Illumina NovaSeq 6000/MGISEQ-T7 platform. Differentially expressed genes were analyzed with significance criteria of log2 (fold change) ≥1 and multi-test adjusted p ≤ 0.05 by Cufflinks (version 2.1.1) and R (version 3.5.1). Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analyses were performed using clusterProfiler (version 3.18.1). The processed RNA-seq data with fragments per kilobase per million (FPKM) values of all genes are presented in Supplementary Table SIV .

All the experiments in this study were performed on three TS^CT cell lines at least in triplicate. GraphPad Prism 6.0 software was used for statistical analysis. Statistical comparisons between two groups were carried out using unpaired Student’s t test after confirming the normal distribution of the data by one-sample Kolmogorov-Smirnov test for Gaussian distribution or using Mann-Whitney U test for comparisons of data that did not show Gaussian distribution. The comparison of continuous variables among groups was carried out by one-way analysis of variance (ANOVA) followed by least significant difference (LSD) tests. All the data are shown as the mean values ± standard error of the means (SEMs).

We collected villi from HC and RPL women in the sixth to eighth week of gestation and performed the experiments laid out in the flowchart ( Figure 1 ). We found that S100P expression was exclusively localized with β-hCG and 11β-HSD2 expression on the syncytial layer of villi from healthy human placentas ( Figure 2A ). To further determine S100P expression levels in placenta tissues, we collected villi from healthy mothers and patients suffering from RPL during early gestation. The immunohistochemistry results showed low S100P and β-hCG expression in the villi of RPL patients ( Figure 2B ), in line with the Western blotting results, which indicated low expression of S100P ( Figure 2C ). As S100P is a secreted calcium-binding protein that can be detected in both human serum and plasma, we measured the serum levels of S100P in patients with RPL and compared them with those in HCs. Interestingly, the serum S100P levels were significantly decreased in patients with RPL ( Figure 2D ).

To detect the role of S100P in syncytialization regulation, we first isolated and identified TS^CT cells, which had the ability to differentiate into ST- and EVT- like cells ( Figure S1A ). TEAD4 is a marker for TS^CT cells, β-hCG is a marker for ST(2D)-TS^CT cells, and human leucocyte antigen-G (HLA-G) is a marker for EVT-TS^CT cells ( Figure S1B ). After the successful establishment of the TS^CT cell model, we treated human TS^CT cells with forskolin (FSK) as previously described (18). Under the presence of FSK for 6 days to fully induce TS^CT syncytialization, S100P expression was progressively increased at both the mRNA and protein levels in ST(2D)-TS^CT cells, as measured at days 3 and 6 ( Figures 3A , B ). The increased expression of 11β-HSD2 indicated successful syncytialization ( Figure 3B ). In addition, S100P was expressed both in the cytoplasm and nucleus of ST(2D)-TS^CT and ST(3D)-TS^CT cells ( Figure 3C ). Interestingly, the level of secreted S100P was significantly increased during FSK-induced syncytialization in vitro ( Figure 3D ), and the level of secreted β-hCG, which is a marker for syncytialization, was also increased ( Figure 3E ). These results indicated that S100P was progressively and exclusively expressed intracellularly and secreted outside the cells during TS^CT cell syncytialization. In addition, the localization of S100P in ST(2D)-TS^CT cells indicated that it may be involved in different biological processes, such as cellular signaling transduction and gene transcription assistance.

To determine the biological significance of S100P in human trophoblast cells, we used small interfering RNAs to interfere with S100P expression and a lentiviral vector to overexpress S100P in human trophoblast stem cells. After transfection followed by differentiation for 6 days, we found a substantial decrease in S100P expression at both the mRNA and protein levels in ST(2D)-TS^CT cells ( Figures 4A, B ). Cells transfected with siS100P exhibited a reduction in 11β-HSD2 expression levels by Western blotting ( Figure 4B ), intracellular β-hCG expression by Western blotting and immunofluorescence ( Figures 4C, D ), and secreted β-hCG concentration by ELISA ( Figure 4E ) compared with cells transfected with the control siRNA. ZO-1 expression was significantly increased in ST(2D)-TS^CT cells with S100P knockdown ( Figure 4C , representative images are shown in Figure 4F ), while the fusion index was decreased ( Figure 4G ). Conversely, neither overexpression of the S100P gene nor exogenous addition of recombinant human S100P protein triggered syncytialization of TS^CT cells, whereas E-cadherin protein expression was decreased ( Figures 4H, I ). These results demonstrated that S100P was involved in regulating the trophoblast syncytialization process, but failed to directly trigger TS^CT cell syncytialization even if the expression of the cell-cell adhesion protein E-cadherin was significantly decreased after S100P was overexpressed.

YAP1 colocalized with E-cadherin in the cytotrophoblast layer of human healthy placentas (six-to-eight weeks of gestation) ( Figure 5A ). YAP1 mRNA and protein expression gradually decreased with syncytialization, as measured on day 3 and 6 ( Figures 5B, C ). The expression of E-cadherin and TEAD4, another key component of the YAP1/TEAD4 transcriptional complex that is crucial for the self-renewal of TS^CT cells and proliferation of progenitors, also decreased with syncytialization ( Figure 5C ). In addition, YAP1 expression showed a significant decrease on day 6 after syncytialization ( Figure 5D ). These results demonstrated that YAP1 decreased with trophoblast syncytialization.

We reasoned that S100P affects syncytialization by repressing CTB progenitor gene expression or inducing the expression of genes that promote STB formation. First, we detected the mRNA expression levels of a few syncytialization-related genes; interestingly, none of these genes were altered after S100P knockdown or overexpression ( Figures S2A, 2B ). Therefore, to identify the potential repressed CTB progenitor gene that mediated S100P syncytialization, we examined YAP1 and TEAD4 mRNA and protein expression levels after transfection with siRNA targeting S100P during TS^CT cell syncytialization. Through this analysis, we found that the mRNA levels of YAP1 and TEAD4 were not significantly upregulated ( Figure 6A ), whereas their protein levels were both increased ( Figure 6B ). The immunofluorescence results further confirmed the upregulated expression of YAP1 and ZO-1 in ST(2D)-TS^CT cells transfected with siS100P ( Figure 6C ). Consistently, YAP1 and TEAD4 mRNA levels showed no significant change, whereas their protein levels were obviously decreased in S100Poverexpressing human TS^CT cells ( Figures 6D, E ). Immunofluorescence staining also confirmed the decreased YAP1 protein levels ( Figure 6F ). Moreover, TS^CT cells treated with the recombinant human S100P protein showed similar results regarding YAP1 and TEAD4 protein levels compared to those with S100P gene overexpression ( Figure 6G ). Together, these data suggested that S100P may be involved in regulating trophoblast syncytialization via YAP1, which is vital for maintaining the self-renewal capacity of TS^CT cells.

To further explore the potential mechanism of S100P regulated trophoblast syncytialization, we performed transcriptome sequencing (RNA-seq) analysis to detect the differentially expressed genes in ST(2D)-TS^CT cells with S100P knockdown. Data analysis identified 33 significantly differentially expressed genes, 11 of which were upregulated and 22 of which were downregulated in the S100P knockdown group compared to the negative control (p<0.05 and log2-fold change>1) ( Figure 7A ). We validated the expression of some genes that are known to affect syncytialization, migration and invasion of trophoblast cells by RT-qPCR. We found that COL1A2, LRP1, NOTCH3, IGFBP3 and LBP1 gene expression was downregulated, which was consistent with the RNA-seq results ( Figure 7B ). Moreover, the enriched differentially upregulated- and downregulated genes between the two groups of ST(2D)-TS^CT cells are shown in the heatmap ( Figure 7C ). We performed GO analysis of the downregulated genes in the S100P siRNA interference group, and the results showed specific enrichment of growth factor binding and extracellular matrix organization ( Figure 7D ). Overall, these results demonstrated that the syncytialization of ST(2D)-TS^CT cells was negatively regulated after S100P knockdown, and extracellular matrix components may play a vital role in the syncytialization process.