Arginine methyltransferase PRMT1 equipoises trophoblast development to prevent early pregnancy loss

All studies with mouse models were approved by IACUC at the University of Kansas Medical Center (KUMC). Human placental tissues were obtained after written consents from patients at the University of Kansas Medical Center (KUMC) and from the Research Center for Women’s and Infants’ Health (RCWIH) Biobank at the Mount Sinai Hospital, Toronto, Canada. The Institutional Review Boards at the KUMC and at the Mount Sinai Hospital approved the utilization of human placental tissues and all experimental procedures. All animal procedures were performed in accordance with protocols approved by the Institutional Animal Care and Use Committee (IACUC) at the University of Kansas Medical Center (KUMC). Animals were maintained with a regular diet in a 14-h light/10-h dark cycle and housed at 65–75 °F (~18–23 °C) with 40–60% humidity. The Prmt1 tm1a (EUCOMM)Wtsi mouse line ( Prmt1 -Tm1a) was obtained from the European Conditional Mouse Mutagenesis Program (EUCOMM). Heterozygous mice were intercrossed to generate experimental litters, and embryos were collected at the indicated gestational stages. The presence of a vaginal plug was designated as embryonic day (E)0.5. Pregnant females were euthanized at the desired gestational time points, and embryos were harvested accordingly. Genomic DNA for genotyping was prepared using the Extract-N-Amp Tissue PCR Kit (Sigma-Aldrich; XNAT) according to the manufacturer’s instructions. The sex of the fetus was not considered as a variable, as PRMT1 expression patterns and the effect of PRMT1-depletion did not show any sexual dimorphism. To generate mice with conditional alleles of Prmt1 ( Prmt1 fl/fl ), we used a Cre/loxP recombination strategy targeting exons 4 and 5 of the Prmt1 gene, which were flanked by loxP sites. PRMT1 Tm1a mice were initially crossed with FLP1 recombinase-expressing mice (Jackson Laboratory, strain #009086) to remove the β-galactosidase/neomycin resistance (β-gal/neo) selection cassette via FLP1-mediated recombination at FRT sites, thereby generating the conditional Prmt1 fl allele. The resulting Prmt1 fl /+ mice were intercrossed to obtain homozygous Prmt1 fl /fl animals. To induce trophoblast progenitor–specific deletion of Prmt1 , Prmt1 fl /fl mice were crossed with Eomes CreER mice. Prmt1 fl /+ ; Eomes CreER offspring were subsequently crossed with Prmt1 fl /fl mice to establish Prmt1 fl /fl ; Eomes CreER line. For experimental induction of Cre recombinase activity, Prmt1 fl /fl females were mated with Prmt1 fl /fl ; Eomes CreER males. The presence of a vaginal plug was designated as embryonic day (E)0.5. Pregnant females received intraperitoneal injections of tamoxifen (0.2 mg/g body weight; Sigma, T5648) at E5.5 and E6.5 to induce Cre-mediated recombination in trophoblast progenitors. Embryos were harvested at the indicated gestational stages, and genotyping and cryosectioning were performed according to previously published protocols 53 , 72 . For identifying the genotypes of control and Prmt1-cKO embryos, which were used for histological analyses, tissues from cryosections were collected and used for genotyping analyses. All genotyping primers are mentioned in Supplementary Table  2 . Prmt1 fl /fl females were mated with Prmt1 fl /fl ; Eomes CreER males and pregnant female mice were sacrificed at E7.5. Uterine horns were immediately dissected in sterile cold PBS containing 10% FBS (Gibco, 16000-044). ExE/EPCs regions were carefully dissected, washed twice in 10% FBS in PBS and cultured with or without 4OHT under proliferative condition for 72–96 h with mouse embryonic fibroblast conditioned medium (MEF-CM), which contained 70% MEF-CM, 30% mouse TSC medium 73 , in the presence of 25 ng/ml FGF4 (Invitrogen, PHG0154) and 1 ug/ml Heparin (Sigma, H3149). Explant outgrowths were measured under a microscope and used for further experimental analysis. The CT27 hTSC line was provided by Dr. Hiroaki Okae (Tohoku University, Japan) and maintained, passaged, and differentiated as previously described 34 . Cells were routinely tested for mycoplasma contamination using a detection kit (ATCC, 30-1012 K). To assess spontaneous STB differentiation in PRMT1KD hTSCs, cells were cultured under standard conditions (5% CO 2 , atmospheric O 2 ) for 3 passages, with ~96 h culture per passage and passaging at a 1:3 ratio. Fresh first-trimester [5( + 4) weeks–9( + 2) weeks of gestation] human placental tissues were obtained from legal pregnancy terminations via the service of Research Center for Women’s and Infants’ Health (RCWIH) BioBank at Mount Sinai Hospital, Toronto, Ontario, Canada. A total of 15 placental samples from the RCWIH biobank were used for isolation of CTBs and ex vivo cultures for EVT development. In addition, formalin-fixed paraffin-embedded sections of 6 first-trimester placental samples were also obtained for testing PRMT1 protein expression in floating and anchoring villi. Placental villous samples from patients with RPL were obtained from the Reproductive Endocrinology and Infertility Clinic at the University of Kansas Medical Center (KUMC) after written informed consent from all participants. The Institutional Review Boards at the KUMC approved all procedures involving human samples. To ensure RPL sample quality, all pregnancies were stringently monitored by serial β-hCG, progesterone measurements, and transvaginal ultrasound to confirm viability at 6–7 weeks of gestation. Tissue collection was performed within 1–2 days of miscarriage confirmation, and samples were transported to the laboratory within 1 h and processed immediately. Importantly, only tissues that yielded viable cytotrophoblasts capable of establishing hTSC lines were included, providing functional validation of sample integrity. Clinical characteristics of the patients included in this study are summarized in Supplementary Table  S1 . For primary CTB isolation, placental villous tissues were finely minced, and portions were fixed in 10% paraformaldehyde for immunohistochemical analyses, while the remaining tissue was snap-frozen in liquid nitrogen for protein and RNA extraction. Dissected villous tissues were processed using immunomagnetic separation according to established protocols 34 . Briefly, cells were isolated using the EasySep™ Human PE Selection Kit (Stemcell Technologies, 17664) with a PE-conjugated anti-ITGA6 (CD49f) antibody (Miltenyi Biotec, 130-127-209) following the manufacturer’s instructions. For hTSC derivation, purified ITGA6-positive CTBs were seeded onto collagen IV–coated (5 µg/ml) 6-well plates at a density of 0.5–1 × 10 6 cells per well and cultured in 2 ml of hTSC medium. Maternal age between 18 and 42 years. Exclusion of women with uterine anomalies, coagulopathy, balanced translocation, and/or autoimmune conditions. Specific criteria include: 1) 42 years of age; 2) morbid obesity (BMI greater than 40); 3) tobacco use within 3 months of conception; 4) history of gonadotoxic therapy; 5) endometrial hyperplasia or cancerous condition of the female reproductive tract; 6) submucosal fibroid(s); 7) endometrial polyp(s) at time of conception; 8) pathology confirmed acute or chronic endometritis; 9) hydrosalpinx communicating with endometrial cavity at time of conception; 10) Müllerian anomaly; 11) untreated endometriosis or adenomyosis 12) intrauterine adhesions; 13)poorly controlled endocrinopathy (diabetes mellitus with HgbA1c > 6.5%, hypothyroidism with TSH ≥ 5 mU/L); 16) hereditary or acquired coagulopathy; 17) active rheumatologic disease; and/or 18) human immunodeficiency virus (HIV). As mentioned above, patients with known balanced chromosomal translocations were excluded. The chromosomal compositions of the collected placental tissues were tested via chromosomal microarray. For this study, only euploid placental tissues from patients with at least two prior miscarriages were used. Lentiviral-mediated knockdown of PRMT1 in hTSCs was performed using a short hairpin RNA (shRNA) targeting PRMT1 (target sequence: TGAGCGTTCCTAGGCGGTTTC; TRCN0000310243). A scrambled shRNA construct (Addgene 1864; sequence: CCTAAGGTTAAGTCGCCCTCGC) was used as a control. Lentiviral particles were generated by transfecting the shRNA plasmids into HEK293T packaging cells using standard lentiviral production protocols. Viral supernatants were collected and concentrated using Lenti-X Concentrator (Takara Bio, 631231) according to the manufacturer’s instructions. hTSCs were transduced with concentrated lentiviral particles at approximately 30–40% confluency. Following transduction, cells were subjected to gradually increasing concentrations of puromycin (1–1.5 µg/mL; Sigma-Aldrich, P8833), starting at 1 µM and increasing to 1.5 µM over a 72-h period. After selection, transduced hTSCs were passaged and subsequently maintained in medium containing 1.5 µM puromycin. The same selection strategy was applied to control CT27 hTSCs transduced with a scrambled shRNA construct. As an additional control, non-transduced CT27 hTSCs were cultured in the presence of puromycin; these cells did not survive at puromycin concentrations ≥1 µM, confirming the effectiveness of the selection strategy. PRMT1 knockdown efficiency was verified by downstream analyses prior to use in subsequent experiments. Cell proliferation in control and PRMT1KD hTSCs was assessed by cell counting and BrdU incorporation assay. Cells were seeded in 6-well plates and counted at 24, 48, and 72 h post-seeding; cell growth was monitored by brightfield microscopy (Olympus IX71). For BrdU assays, 1 × 10 4 cells were seeded in 12-well plates and labeled using the BrdU Labeling and Detection Kit I (Roche, 12296736001) according to the manufacturer’s instructions. After 24, 48, and 72 h, nuclei were counterstained with DAPI (BD Biosciences, 564907) and mounted in SlowFade (Invitrogen, S36936). BrdU-positive nuclei were visualized and quantified using a fluorescence microscope (Nikon Eclipse 80i). Cell death was assessed using the Annexin V Apoptosis Detection Kit (Thermo Scientific, 88-8007-72). Control and PRMT1KD hTSC were used to generate self-renewing organoids as previously described 33 . Briefly, hTSCs were harvested and resuspended in ice-cold trophoblast medium (b-TOM: 10 mM HEPES (Sigma H3537), B27(Gibco 17504-044), N2 (Gibco 17502-048) and 2 mM glutamine (Gibco 25030081). Growth factor-reduced Matrigel (Corning 354230) was added to reach 60% final concentration. Drops of 25–30 µl containing 40,000 cells were placed in the center of 24-well plates, inverted, and incubated at 37 °C for 20 min to form hanging drops. Plates were then turned upright, and domes were overlaid with 500 µl room-temperature a-TOM (b-TOM plus 100 ng/ml R-spondin, 1 µM A83-01 (PeproTech 120-38), 100 ng/ml rhEGF (Sigma E9644), 50 ng/ml rmHGF (PeproTech 315-23), 2.5 µM Prostaglandin E2 (R&D System, 2296/10), 3 µM CHIR99021 (Sigma, SML1046), and 100 ng/ml Noggin (Invitrogen PHC1506)). Organoids were cultured for 8 days with medium changed every 3 days and appropriate inhibitors added. Brightfield images were taken to monitor growth. Placental villous explant cultures for testing EVT emergence were established as described previously 27 . Placental tissues from 5 + 4 to 7 + 5 weeks of gestation were washed in ice-cold 1 × PBS, finely dissected, and fragments containing branching villous structures were selected. Explants were washed in PBS containing 10% FBS and incubated with lentiviral particles expressing scrambled or PRMT1 shRNA for 12 h at 37 °C in a humidified atmosphere of 5% CO2. After transduction, explants were embedded in growth factor-reduced Matrigel (Corning) diluted in DMEM/F12 (Gibco, 11320-033) on ice. A 50 μL Matrigel suspension was dispensed into each well of a 24-well plate, and explants were placed within the matrix. Following gel solidification at 37 °C, 700 μL of EVT day 1 medium was added to each well. Cultures were maintained with medium replacement on days 3 and 6. The cells were fixed with 4% paraformaldehyde/PBS at 4 °C for 10 min, permeabilized with PBS-0.25%TritonX at room temperature for 15 min and blocked in PBS-0.1%TritonX-10%FBS at room temperature for an hour. Similarly, the cryosection of the human placenta and mouse embryos were also treated in the same way. Then they were incubated with primary antibodies in PBS-0.1%TritonX-10%FBS at 4 °C overnight. The next day, they were stained with secondary antibody in PBS-0.1%TritonX-10%FBS at room temperature for 2 h. DAPI (BD Biosciences, 564907) counterstain was done, and then antifade mounting media (Invitrogen, S36936) were used. Detailed information on antibodies is provided in Supplementary Table  S3 . For immunohistochemistry, after deparaffinization and rehydration, antigen retrieval with citrate buffer was done, followed by endogenous peroxidase (3% H₂O₂). DAB counterstaining with Mayer’s Hematoxylin was done and imaged in Nikon Eclipse 80i. RNA was extracted using the RNeasy Mini Kit (Qiagen, 74104) according to the manufacturer’s instructions. RNA concentration and purity were assessed using a NanoDrop spectrophotometer (Invitrogen). For cDNA synthesis, 1 μg of total RNA was reverse-transcribed using random hexamer primers (Thermo Scientific, S0142) and reverse transcriptase (Invitrogen, 28025-013) following standard protocols. Quantitative real-time PCR was performed using SYBR Green Master Mix (Applied Biosystems, 4367659), with 18S rRNA used as an internal control for normalization. Data analysis and visualization were conducted using GraphPad Prism. Statistical significance was determined using a two-tailed unpaired Student’s t -test. Primer sequences used for RT-qPCR are listed in Supplementary Table  S4 . Whole-cell lysates were prepared using RIPA buffer (Thermo Scientific, P189900), followed by sonication and protein quantification using the BCA assay (Thermo Scientific, 23227). Equal amounts of protein were denatured in 4× Laemmli buffer (Bio-Rad, 1610747) and resolved by SDS–PAGE at 80–100 V. Proteins were transferred onto Immobilon-P PVDF membranes (Thermo Scientific, 88518) at 300 mA for 1 h on ice. Membranes were blocked and incubated overnight at 4 °C with primary antibodies diluted in 5% skim milk in PBS-T (BD Biosciences, 232100), followed by incubation with appropriate secondary antibodies at room temperature. Protein bands were visualized using enhanced chemiluminescence (ECL) substrate (Sigma, WBLURO500) and imaged with a ChemiDoc system. Details of antibodies are provided in Supplementary Table  S3 , and uncropped, unprocessed blot images are shown in Supplementary Fig.  S12 . CUT&RUN was performed using the Cell Signaling Technology CUT&RUN Kit (86652) following the manufacturer’s instructions. For each reaction, 5 × 10 5 proliferating hTSCs or primary cytotrophoblasts (CTBs) isolated from first-trimester placenta were used. For mouse experiments, E7.5 ExE/EPC regions were collected from multiple litters to obtain sufficient cells. Cells were immobilized on Concanavalin A–coated magnetic beads, permeabilized with buffers containing 0.5% digitonin, and incubated with primary antibodies. Targeted DNA cleavage was performed using Protein A/G–fused micrococcal nuclease (pAG-MNase) activated by CaCl 2 . Reactions were terminated with Stop Buffer, and DNA fragments were purified by phenol–chloroform extraction followed by ethanol precipitation. DNA was quantified, and fragment size distribution was assessed using a 4200 TapeStation system (Agilent). Samples from three independent CUT&RUN experiments per condition were sequenced. Libraries were prepared using the Accel-NGS 1S Plus DNA Library Kit and 1S Plus Combinatorial Dual Indexing Kit (Swift Biosciences). Sequencing reads were processed using the nf-core/chipseq v2.1.0 pipeline. Peaks were called using MACS2 with a significance threshold of p  < 1e −5 . Genomic peak distribution was determined using ChIPseeker v1.42.1, and HOMER v4.11 was used for peak annotation and motif analysis. For scRNA-seq analyses of trophoblast cells in developing mouse placentas, CD-1 females were paired with CD-1 males, and the presence of a vaginal plug was designated as embryonic day (E)0.5. The ExE/EPC region was collected at ~E7.0, and developing placentas were dissected at E8.5, E9.5, E10.5 and E12.5, carefully avoiding decidual tissue contamination. For each gestational time point, tissues from at least three pregnant females were used for single-cell preparation. Placental tissues were enzymatically digested using collagenase type IV for 30 min at 37 °C, followed by mechanical dissociation through serial syringe passage. Cell suspensions were filtered through 40-µm cell strainers, and cellular debris and dead cells were removed using Debris Removal Solution and Dead Cell Removal Kit (Miltenyi Biotec). Red blood cells were depleted using an anti-mouse Ter-119 antibody (BD Biosciences, 558536). Cell viability was assessed using trypan blue (Corning, 25-900-Cl) staining with an automated cell counter (Countess 3, Thermo Fisher Scientific). Single-cell libraries were prepared at the University of Kansas Medical Center (KUMC) Genomics Core Facility using the 10× Genomics Chromium Single Cell 3′ v3 platform, following the manufacturer’s instructions. scRNA-seq datasets from developing first-trimester human placentas were generated previously 31 . Briefly, single-cell suspensions were prepared from placental villi of two first-trimester placentas (6 and 7.6 weeks of gestation) and sequenced using the 10× Genomics platform. In the present study, these datasets were reanalyzed to assess PRMT1 expression patterns across trophoblast cell populations. For scRNA-seq analysis of control and PRMT1 knockdown ( PRMT1KD ) human trophoblast stem cells (hTSCs), cells were dissociated using TrypLE (Gibco, 12604-021) and resuspended in hTSC culture medium. Cell viability was determined by trypan blue staining. Single-cell libraries were prepared at the KUMC Genomics Core Facility using the 10× Genomics Chromium Single Cell 3′ v3 platform according to the manufacturer’s protocol. Single-cell RNA sequencing data were processed using popsicleR (v0.2.1) and Seurat (v4.0.6). Quality control filtering removed cells with fewer than 500 detected molecules or greater than 5% mitochondrial gene content. Potential doublets were identified and removed using the CalculateDoublets() function. For mouse placental datasets, samples from different developmental stages were integrated using SelectIntegrationFeatures(), FindIntegrationAnchors(), and IntegrateData() functions in Seurat. Dimensionality reduction was performed using principal component analysis (PCA) followed by Uniform Manifold Approximation and Projection (UMAP). Cell clustering was conducted using FindNeighbors() and FindClusters(), and cluster-specific marker genes were identified using the FindAllMarkers() function of Surat. Visualization of gene expression was performed using the Featureplot_scCustom() function of the scCustomize R package. Additional sequencing metrics, including cell numbers, sequencing depth, and reads per cell for each scRNA-seq dataset, are summarized in Supplementary Table  S5 . To rescue PRMT1, Lentiviral particles of human PRMT1 (Origene, RC214074L3V) were used to transduce PRMT1 Low RPL-hTSCs . Overexpression constructs for TEAD4 and MYBL2 were generated by cloning the full-length human cDNAs of TEAD4 ( NM_003213.4 ) and MYBL2 ( NM_002466.4 ) into the pLV[Exp]-Puro-EF1A > EGFP vector (VectorBuilder; VB010000-9495wwg). An empty PLKO.1 vector was used as a control. PRMT1 Low RPL-hTSCs were transduced with the concentrated lentiviral particles at 30–40% confluency. Transduced cells were selected using puromycin, subsequently assessed for PRMT1, TEAD4 or MYBL2 overexpression efficiency and used for downstream experiments. RNA-seq analysis was performed according to the published protocol 27 . Total RNA from control and PRMT1KD hTSCs were isolated using the RNeasy Mini Kit following the manufacturer’s protocol. Integrity of the total RNA samples was evaluated using an Agilent Technologies 2100 Bioanalyzer. The total RNA fraction was processed by oligo dT bead capture of mRNA, fragmentation, and reverse transcription into cDNA. After ligation with the appropriate Unique Dual Index (UDI) adapters, the cDNA library was prepared using the Universal Plus mRNA-seq +UDI library preparation kit (NuGEN; Marion, SD). Adapter and low-quality reads were removed using the Rfastp R package. Reads were mapped to the reference genome hg38 using the Rsubread mapping tool. Mapping of reads was done using the featureCount tool of the subread package. The downstream analysis was done using the DESeq2 tool. The raw data for RNA-seq analyses have been submitted to the GEO database with accession No. GSE295904 . A single preparation of RNA from each PRMT1 High and PRMT1 Low RPL-placentas and PRMT1 High and PRMT1 Low RPL-hTSCs was sequenced to compare global gene expression. Because modest variability in PRMT1 knockdown efficiency was observed across independently transduced batches, RNA from three independent PRMT1KD hTSC batches was pooled to generate each cDNA library. Two such libraries were prepared and sequenced for the PRMT1KD condition. The same strategy was applied to control hTSCs, expressing a scrambled shRNA. Initial analyses of RNA-seq datasets from RPL placentas, RPL-hTSCs, and control versus PRMT1KD CT27 hTSCs showed genes with extreme fold changes, which were driven by very low transcript abundance. We therefore reanalyzed the RPL placenta and RPL-hTSC RNA-seq datasets using a raw-count cutoff of ≥10, which produced more robust fold-change estimates. We further confirmed, using first-trimester placental scRNA-seq datasets, that genes showing high fold changes between PRMT1 High and PRMT1 Low RPL placentas/hTSCs are expressed within trophoblast populations. For the control CT27 versus PRMT1KD -hTSC comparison, we similarly reanalyzed genes with raw counts ≥10 in control CT27 hTSCs. For all histology, immunofluorescence, western blot and RT-qPCR experiments, we performed at least three independent biological replicates. All uncropped and unprocessed scans of the western blots were supplied in Supplementary Fig.  12 in the Supplementary Information. Statistical significance was determined for RT-qPCR analyses for mRNA expression, for hTSC proliferation and organoid formation analyses. We performed at least n  = 3 experimental replicates for all these experiments. Independent datasets were analyzed using GraphPad Prism software (version 11). To determine statistical significance for comparisons between two experimental groups, we performed a two-tailed, unpaired t-test with Welch’s correction. Statistical significance for more than two experimental groups was determined using one-way ANOVA. Significantly altered values ( P  ≤ 0.05) are highlighted in the figures by asterisk/s. As mentioned above, RNA-seq data from control and PRMT1KD hTSCs were generated with n  = 3 experimental replicates per group, and two such groups were used for sequencing. For RNA-seq analyses, we used the Wald test to get the p -values while comparing the groups. We have used log 2 fold ≥ 1.5 and p -value ≤ 0.05 to get the statistically significantly differentially expressed genes. For Cut and Run-seq peaks were called using macs2 callpeak, which uses the Poisson distribution to calculate p -values. We have selected a p -value cutoff of 1 e-5 for experiments in hTSCs and 1 e-3 for experiments in CTBs to detect statistically significant peaks. Further information on research design is available in the  Nature Portfolio Reporting Summary linked to this article.

To examine whether PRMT1 is associated with trophoblast lineage specification, we profiled Prmt1 expression across trophoblast cell types using single-cell RNA sequencing (scRNA-seq) of mouse placentas spanning embryonic day (E)7.0 to E12.5. Clustering of trophoblast populations based on established marker genes revealed that Prmt1 expression is highest in trophoblast stem and progenitor cells (TSPCs) located within the extraembryonic ectoderm/ectoplacental cone (ExE/EPC) region of the E7.0 placental primordium (Figs.  1A , S1A,B ). Prmt1 expression was also detected in junctional zone trophoblast populations (Figs.  1A , S1A, B ). Sub-clustering of E7.0 trophoblast progenitors further identified two populations distinguished by high ( Prmt1 High ) and low ( Prmt1 Low ) Prmt1 expression (Fig.  1B ). The Prmt1 High population expressed elevated levels of canonical trophoblast stem cell regulators, including Eomes, Elf5, Cdx2, Id2 , and Esrrb , whereas Prmt1 Low progenitors showed increased expression of markers associated with differentiated trophoblast lineages, such as Dlx3, Nr6a1, Prl3b1, Tpbpa , and Plac1 (Fig.  1C ). Although matured STBs are underrepresented in our scRNA-seq dataset, our analyses indicate that Prmt1 expression correlates with the trophoblast stem-state transcriptional program during mouse placentation. Fig. 1 PRMT1 expression is conserved in rodent and human trophoblast progenitors. A scRNA-seq cluster of developing mouse placenta from E7.0 to E12.5 showing different cell clusters with Prmt1 expression. Note high Prmt1 expression in the cell cluster representing TSPCs from the E7.0 developing placenta primordium. B Further clustering showing Prmt1 High and Prmt1 Low cells within the TSPC population. C Dot plot showing average expression and percent of cells in Prmt1 High and Prmt1 Low TSPC clusters for genes representing undifferentiated TSC and differentiated trophoblast cell markers. D Immunofluorescence images showing PRMT1 expression in trophectoderm of an E3.5 mouse blastocyst. E , F ExE/ EPC regions in an E7.0 mouse placenta primordium and E9.0 rat placenta primordium, respectively, showing PRMT1 (red) and Cytokeratin 7 (PanCK, green) expression in TSPCs. G Histological section of a human first-trimester (week 7) placenta showing a high level of PRMT1 expression (red) in CTBs of a floating villous (left) and in both proximal cCTB and distal EVT precursors within an anchoring villous (middle). Cytokeratin 7 (KRT7) was used to mark trophoblast cells. H Histological section of a term human placenta showing PRMT1 expression in trophoblast cells (right). I RT-qPCR analyses of PRMT1 mRNA expression in CTBs, isolated from first-trimester (6–9 weeks) and normal term (38–39 weeks) placentas. Data are representative of six independent experiments ( n  = 6), and the values are normalized to 18S rRNA and indicated as mean ± SEM. Two-tailed unpaired t -test with Welch’s correction, ** p  = 0.0015). Source data are provided as a source data file. A scRNA-seq cluster of developing mouse placenta from E7.0 to E12.5 showing different cell clusters with Prmt1 expression. Note high Prmt1 expression in the cell cluster representing TSPCs from the E7.0 developing placenta primordium. B Further clustering showing Prmt1 High and Prmt1 Low cells within the TSPC population. C Dot plot showing average expression and percent of cells in Prmt1 High and Prmt1 Low TSPC clusters for genes representing undifferentiated TSC and differentiated trophoblast cell markers. D Immunofluorescence images showing PRMT1 expression in trophectoderm of an E3.5 mouse blastocyst. E , F ExE/ EPC regions in an E7.0 mouse placenta primordium and E9.0 rat placenta primordium, respectively, showing PRMT1 (red) and Cytokeratin 7 (PanCK, green) expression in TSPCs. G Histological section of a human first-trimester (week 7) placenta showing a high level of PRMT1 expression (red) in CTBs of a floating villous (left) and in both proximal cCTB and distal EVT precursors within an anchoring villous (middle). Cytokeratin 7 (KRT7) was used to mark trophoblast cells. H Histological section of a term human placenta showing PRMT1 expression in trophoblast cells (right). I RT-qPCR analyses of PRMT1 mRNA expression in CTBs, isolated from first-trimester (6–9 weeks) and normal term (38–39 weeks) placentas. Data are representative of six independent experiments ( n  = 6), and the values are normalized to 18S rRNA and indicated as mean ± SEM. Two-tailed unpaired t -test with Welch’s correction, ** p  = 0.0015). Source data are provided as a source data file. To validate these observations at the protein level, we performed immunofluorescence analyses in mouse blastocysts and early post-implantation mouse (E6.0–E7.0) and rat (E9.0) conceptuses. PRMT1 was highly expressed in both polar and mural trophectoderm of the blastocyst (Fig.  1D ) and in trophoblast stem and progenitor populations within mouse and rat placental primordia (Figs.  1E, F , S1C ). Consistent with the transcriptional data, PRMT1 expression was markedly reduced in the developing labyrinth zone of the E9.5 mouse placenta, which contains STBs (Fig.  S2A ). In contrast, PRMT1 expression persisted in the junctional zone of the E9.5 mouse placenta (Fig.  S2A ), a region enriched for invasive trophoblast progenitors, and in invasive trophoblast cells at the uterine–placental interface of E18.5 rat conceptuses (Fig.  S2B ). Together, these findings demonstrate that PRMT1 expression is conserved in trophoblast progenitors and invasive trophoblast populations during mammalian placental development. To determine whether this expression pattern is conserved in humans, we examined PRMT1 distribution in first-trimester placental tissues. In the early human placenta, proliferating villous CTBs form the trophoblast stem/progenitor compartment and share conserved transcriptional regulators with mouse trophoblast stem cells, including CDX2, ELF5, TEAD4, GATA2, and GATA3 12 , 25 , 26 . Immunofluorescence analysis revealed that PRMT1 is highly expressed in proliferating CTBs within floating villi but is markedly reduced in differentiated STBs (Fig.  1G , left). In anchoring villi, PRMT1 expression was maintained in cCTBs and in nascent extravillous trophoblasts (EVTs) emerging from the distal end of the trophoblast column (Fig.  1G , right), indicating that PRMT1 expression is associated with trophoblast progenitor and invasive trophoblast populations during early human placental development. Reanalysis of our published single-cell RNA-sequencing datasets from first-trimester human placentas 12 , 27 further confirmed this pattern, revealing that PRMT1 mRNA is highly enriched in undifferentiated CTB and mature EVT clusters but markedly reduced in the committed STB cluster (Fig.  S2C, D ). In contrast to first-trimester placentas, PRMT1 expression was substantially diminished in CTBs of term placentas (Fig.  1H, I ), suggesting that PRMT1 function is primarily associated with early trophoblast development. Collectively, these findings demonstrate a conserved PRMT1 expression pattern during mammalian placentation, in which PRMT1 is enriched in trophoblast stem and progenitor populations, maintained during invasive trophoblast/EVT differentiation, and suppressed during STB formation. An earlier report 16 using a gene-trap allele showed that global loss of Prmt1 in the mouse embryo results in embryonic death ~E6.5–E7.0, a stage when the extraembryonic placenta primordium, containing the ExE/EPC, develops. Because PRMT1 is highly expressed in trophoblast stem/progenitor cells (TSPCs), we investigated whether PRMT1 is required for TSPC development during early post-implantation stages. To address this, we first examined Prmt1 tm1a(EUCOMM)Wtsi mouse, which harbors the Knock-out first (Tm1a) allele ( Prmt1 tm1a mice) , developed by the Knockout Mouse Project 28 . Consistent with previous reports, homozygous Prmt1 tm1a (PRMT1-KO) embryos failed to develop beyond E7.0 (Figs.  2A, B , S3A ). Notably, the placental primordium of ~E7.0 Prmt1-null conceptuses was markedly smaller and exhibited a dramatic reduction in ESRRB-expressing TSPCs (Fig.  2C ), suggesting an early defect in trophoblast progenitor maintenance. Fig. 2 Loss-of PRMT1 in TSPCs of an early post-implantation mouse embryo impairs placentation leading to embryonic death. A Genotyping confirming Prmt1 knockout allele in E7.0 Prmt1 - Tm1a homozygous embryos. B Representative images of E7.0 Prmt1- heterozygous and Prmt1- KO conceptuses are shown. Embryonic and placentation defects were prominent in the Prmt1- KO conceptus. C Immunostaining images of E7.0 Prmt1-Tm1a heterozygous (control, Left) and homozygous Prmt1-KO (right) showing ESRRB (green, TSPC marker) and PRMT1 (red) expression. The drastic loss of ESRRB-expressing TSPCs at the ExE/EPC region of the Prmt1-KO conceptus is shown in the inset. D Conditional Prmt1-alleles were deleted in TSPCs of a developing mouse placenta, and embryonic and placental developments were analyzed at E7.5 in Control (Prmt1 fl/fl ) and Prmt1-cKO embryos. Representative images of control and Prmt1-c KO conceptuses are shown. Embryonic and placentation defects were prominent in Prmt1-c KO conceptus. E TSPCs in the developing placenta primordium of control and Prmt1-c KO implantation sites were analyzed at ~E7.5 via immunostaining with TEAD4 (green, TSPC marker) and pan-cytokeratin antibody (red, pan trophoblast marker). The developing Prmt1-c KO placenta primordium is smaller and mainly contains TGCs, which can be identified via larger nuclei (Blue). A Genotyping confirming Prmt1 knockout allele in E7.0 Prmt1 - Tm1a homozygous embryos. B Representative images of E7.0 Prmt1- heterozygous and Prmt1- KO conceptuses are shown. Embryonic and placentation defects were prominent in the Prmt1- KO conceptus. C Immunostaining images of E7.0 Prmt1-Tm1a heterozygous (control, Left) and homozygous Prmt1-KO (right) showing ESRRB (green, TSPC marker) and PRMT1 (red) expression. The drastic loss of ESRRB-expressing TSPCs at the ExE/EPC region of the Prmt1-KO conceptus is shown in the inset. D Conditional Prmt1-alleles were deleted in TSPCs of a developing mouse placenta, and embryonic and placental developments were analyzed at E7.5 in Control (Prmt1 fl/fl ) and Prmt1-cKO embryos. Representative images of control and Prmt1-c KO conceptuses are shown. Embryonic and placentation defects were prominent in Prmt1-c KO conceptus. E TSPCs in the developing placenta primordium of control and Prmt1-c KO implantation sites were analyzed at ~E7.5 via immunostaining with TEAD4 (green, TSPC marker) and pan-cytokeratin antibody (red, pan trophoblast marker). The developing Prmt1-c KO placenta primordium is smaller and mainly contains TGCs, which can be identified via larger nuclei (Blue). To directly test the trophoblast-specific requirement for PRMT1, we performed conditional loss-of-function studies using Prmt1 fl/f mice. We generated Prmt1 fl/fl :Eomes CreER conceptuses by crossing Prmt1 fl/fl with Eomes CreER mice, in which a tamoxifen-inducible Cre-recombinase ( CreER ) is expressed under the control of the Eomes genomic locus 29 . The Eomes CreER driver has been reported to be active in the ExE/EPC region and primitive streak at E6.5–E7.5 29 . We independently validated Cre activity by crossing Eomes CreER mice with ROSAmT/mG reporter mice 30 and analyzing embryos at ~E7.5. Robust EGFP expression was detected in the placenta primordium, with additional activity in the primitive streak but not in other regions of the embryo proper (Fig.  S3C ). To conditionally delete Prmt1 in trophoblast progenitors, Prmt1 fl/fl : Eomes CreER males were crossed with Prmt1 fl/fl females, thereby restricting CreER expression to the developing conceptus (Fig.  2D ). We induced Prmt1 deletion (Fig.  S3B ) at ~E5.5 by administration of 4-hydroxytamoxifen (4OHT) and analyzed resulting embryos at ~E7.5. Conditional knockout of Prmt1 resulted in a markedly reduced ExE/EPC region and impaired embryonic development (Figs.  2D , S3D ). Histological analysis further revealed that the placental primordium of Prmt1 -cKO conceptuses contained a severely depleted TSPC population and was predominantly composed of trophoblast giant cells (TGCs) (Fig.  2E ). As the Eomes CreER driver is also active in the primitive streak, embryonic development in Prmt1-cKO conceptuses could potentially be influenced by Eomes CreER -mediated deletion of Prmt1 in the primitive streak. To determine whether ExE/EPC defects arise from a trophoblast-autonomous requirement for PRMT1, we performed ex vivo culture of isolated ExE/EPC regions from control ( Prmt1 fl/fl ) and Prmt1 fl/fl : Eomes CreER conceptuses in the presence of 4OHT. Conditional deletion of Prmt1 resulted in reduced proliferation and loss of Tead4 expression specifically in mutant ExE/EPC cultures, whereas control cultures remained unaffected (Fig.  S3E–G ). Together, these findings demonstrate that PRMT1 is essential for maintaining the TSPC population in the early post-implantation placenta, preventing premature differentiation and enabling normal progression of placental and embryonic development. Idiopathic RPL is frequently associated with impaired trophoblast development and defective placentation. Consistent with this, previous studies, including our own, have shown that idiopathic RPL is linked to reduced expression of key regulators of cytotrophoblast (CTB) self-renewal, including TEAD4 and MYBL2 13 , 31 . In mice, loss of Prmt1 causes embryonic lethality around embryonic day (E) 7.0, a stage corresponding to Carnegie stage 7 (days 15–17 post-conception) in humans, when rapid expansion of CTB progenitors drives villous placental development 32 . These findings led us to hypothesize that defects in PRMT1 in CTB progenitors may contribute to early pregnancy loss in humans. Therefore, we collected placentas from patients with idiopathic RPLs (RPL-Placenta). Although the ~E7.5 developmental time point in the mouse corresponds to Carnegie stage 7 (~day 15–17) of human development, it is inherently challenging to obtain very early human placental tissues from pregnancies associated with RPL. Nevertheless, to address our hypothesis, we analyzed 11 first-trimester and one early second-trimester (14 + 5 weeks) euploid placental tissues from patients with idiopathic RPL (Supplementary Table  S1 ). We excluded cases with uterine anomalies, coagulopathy, balanced chromosomal translocations, or autoimmune disorders. Among 12 euploid idiopathic RPL placentas examined, PRMT1 expression was preserved in a subset (PRMT1 High RPL placentas, Supplementary Fig.  S4 ). However, a subset of RPL placentas displaying overt defects in villous development (Fig.  3A , Supplementary Table  S1 ) showed markedly reduced PRMT1 expression in CTBs (PRMT1 Low RPL placentas, Fig.  3B, C ). PRMT1 expression was also strongly repressed within CTB columns of anchoring villi in these samples (Fig.  3D, E ). Fig. 3 Idiopathic RPLs is associated with reduced PRMT1 expression in CTB progenitors and EVT Precursors. A Micrographs show defective villi formation in a first-trimester (9-week) placenta from a patient suffering from idiopathic RPL. An 8.5-week placenta from a sporadic miscarriage without an apparent defect in placental villi formation was used as a control. B Immunostaining images confirming a strong reduction of PRMT1 protein expression in CTBs in floating villi of idiopathic RPL placentas. C Image J quantitation of PRMT1 protein expression in CTBs of control and Idiopathic RPL placentas. Data are representative of 5 independent experiments ( n  = 5) and indicated as mean ± SEM. Two-tailed unpaired t -test with Welch’s correction, ** p  = 0.0072). Source data are provided as a source data file. D Immunostaining images confirming reduced PRMT1 protein expression in proximal cCTBs and distal EVT precursors in anchoring villi of idiopathic RPL placentas. E Image J quantitation of PRMT1 protein expression in EVT precursors of control and idiopathic RPL placentas. Data are representative of 5 independent experiments ( n  = 5) and indicated as mean ± SEM. Two-tailed unpaired t-test with Welch’s correction, **** p  < 0.0001). Source data are provided as a source data file. F Volcano plot showing global gene expression changes in PRMT1 High RPL and PRMT1 Low RPL placentas. Unbiased RNA-seq analyses were performed, and ≥1.5-fold gene-expression changes in RPL placentas with a false discovery rate of p  < 0.05 were indicated with colored dots (blue: up-regulated, red: down-regulated). Significant down-regulation in expression of PRMT1, TEAD4, MYBL2 , and up-regulation of CSH2, SDC1, PSG9 and PSG11 (STB-specific genes) are indicated. A Micrographs show defective villi formation in a first-trimester (9-week) placenta from a patient suffering from idiopathic RPL. An 8.5-week placenta from a sporadic miscarriage without an apparent defect in placental villi formation was used as a control. B Immunostaining images confirming a strong reduction of PRMT1 protein expression in CTBs in floating villi of idiopathic RPL placentas. C Image J quantitation of PRMT1 protein expression in CTBs of control and Idiopathic RPL placentas. Data are representative of 5 independent experiments ( n  = 5) and indicated as mean ± SEM. Two-tailed unpaired t -test with Welch’s correction, ** p  = 0.0072). Source data are provided as a source data file. D Immunostaining images confirming reduced PRMT1 protein expression in proximal cCTBs and distal EVT precursors in anchoring villi of idiopathic RPL placentas. E Image J quantitation of PRMT1 protein expression in EVT precursors of control and idiopathic RPL placentas. Data are representative of 5 independent experiments ( n  = 5) and indicated as mean ± SEM. Two-tailed unpaired t-test with Welch’s correction, **** p  < 0.0001). Source data are provided as a source data file. F Volcano plot showing global gene expression changes in PRMT1 High RPL and PRMT1 Low RPL placentas. Unbiased RNA-seq analyses were performed, and ≥1.5-fold gene-expression changes in RPL placentas with a false discovery rate of p  < 0.05 were indicated with colored dots (blue: up-regulated, red: down-regulated). Significant down-regulation in expression of PRMT1, TEAD4, MYBL2 , and up-regulation of CSH2, SDC1, PSG9 and PSG11 (STB-specific genes) are indicated. To define the associated transcriptional changes, we performed global RNA sequencing on PRMT1 Low RPL placentas and compared them with four control PRMT1 High RPL placentas in which PRMT1 expression was maintained in CTBs and developing EVTs. RNA-seq also confirmed a pronounced reduction in PRMT1 transcripts in the PRMT1 Low RPL placentas (Fig.  3F , Supplementary Dataset  S1 ). Consistent with previous reports 13 , 31 , 33 , transcripts encoding the CTB self-renewal regulators T EAD4 and MYBL2 were also significantly reduced in PRMT1 Low RPL placentas. In contrast, genes associated with syncytiotrophoblast differentiation, including placenta-specific glycoproteins ( PSG9, PSG11 ) and chorionic somatomammotropin hormone 2 ( CSH2 ) and Syndecan-1 ( SDC1 ), were upregulated. Together, these findings identify loss of PRMT1 expression in CTB progenitors and emerging EVT precursors as a feature of a subset of idiopathic RPLs, suggesting that disruption of PRMT1-dependent trophoblast progenitor programs may contribute to early pregnancy failure. Because a subset of idiopathic RPLs is associated with reduced PRMT1 expression in CTBs and EVT precursors, we next investigated the functional consequences of PRMT1 loss in human trophoblast development. As an experimental model, we used human trophoblast stem cells (hTSCs), which are derived from first-trimester CTBs 34 and maintain long-term self-renewal without genomic instability 31 while retaining the capacity to differentiate into both STBs and extravillous trophoblasts EVTs. Consistent with our in vivo observations, PRMT1 expression in hTSCs recapitulated the developmental expression pattern observed in placental tissues. RT-qPCR and Immunofluorescence analyses showed that PRMT1 mRNA and protein are highly expressed in self-renewing hTSCs and are largely maintained during EVT differentiation, whereas STB differentiation is associated with a marked reduction in PRMT1 expression (Fig.  S5A, B ). To determine the functional role of PRMT1, we initially attempted CRISPR-mediated gene deletion in hTSCs. However, PRMT1 knockout cells failed to expand in culture, suggesting that PRMT1 is essential for hTSC survival. Because complete PRMT1 loss is therefore unlikely to occur in vivo, we modeled the partial reduction observed in pathological pregnancies by performing lentiviral shRNA-mediated knockdown of PRMT1 ( PRMT1KD ) in hTSCs (Fig.  4A, B ). Notably, PRMT1 depletion resulted in a pronounced reduction of global ADMA modifications (Fig.  4C ), confirming that PRMT1 is the predominant enzyme responsible for ADMA formation in human trophoblast cells. Functionally, PRMT1KD hTSCs exhibited a severe defect in cell proliferation (Fig.  4D, E ), which was further confirmed by reduced BrdU incorporation (Fig.  4F, G ), without a corresponding increase in cell death (Fig.  S5C ). Fig. 4 Loss of PRMT1 in hTSCs impairs self-renewal. A RT-qPCR analyses of PRMT1 mRNA expression in control and PRMT1KD hTSCs. Data are representative of 10 independent experiments ( n  = 10), and the values are normalized to 18S rRNA and indicated as mean ± SEM. Two-tailed unpaired t-test with Welch’s correction, **** p  < 0.0001). Source data are provided as a source data file. B Representative western blots, respectively, confirming depletion of PRMT1 expression in PRMT1KD hTSCs . C Representative western blots confirming loss of Global ADMA in PRMT1KD hTSCs . D Representative micrographs showing colony morphology of control (CT27) and PRMT1KD hTSCs after 72 h in culture. Note the smaller colony sizes of the PRMT1KD hTSCs. E Growth kinetics of control and PRMT1KD hTSCs. Cells were collected and counted at the indicated time points, and the data are representative of three independent experiments ( n  = 3). Two-tailed unpaired t -test with Welch’s correction, * p  = 0.0176 for 72 h. Source data are provided as a source data file. F Representative immunostained colonies showing BrdU incorporation in control and PRMT1KD hTSC when cultured in stem-state culture conditions for 72 h. G Relative intensity of BrdU-positive nuclei in control and PRMT1KD hTSCs. Data are representative of 6 independent experiments ( n  = 6) and indicated as mean ± SEM. Two-tailed unpaired t -test with Welch’s correction, **** p  < 0.0001). Source data are provided as a source data file. H Micrographs show inefficient organoid formation by P RMT1KD hTSCs. I Quantitation of primary organoid diameters formed with control and PRMT1KD hTSCs. Data were generated from three independent experiments. Each dot in the column represents one organoid ( n  = 24, control organoid, n  = 27, PRMT1KD organoid). Two-tailed unpaired t -test with Welch’s correction, **** p  < 0.0001). Source data are provided as a source data file. A RT-qPCR analyses of PRMT1 mRNA expression in control and PRMT1KD hTSCs. Data are representative of 10 independent experiments ( n  = 10), and the values are normalized to 18S rRNA and indicated as mean ± SEM. Two-tailed unpaired t-test with Welch’s correction, **** p  < 0.0001). Source data are provided as a source data file. B Representative western blots, respectively, confirming depletion of PRMT1 expression in PRMT1KD hTSCs . C Representative western blots confirming loss of Global ADMA in PRMT1KD hTSCs . D Representative micrographs showing colony morphology of control (CT27) and PRMT1KD hTSCs after 72 h in culture. Note the smaller colony sizes of the PRMT1KD hTSCs. E Growth kinetics of control and PRMT1KD hTSCs. Cells were collected and counted at the indicated time points, and the data are representative of three independent experiments ( n  = 3). Two-tailed unpaired t -test with Welch’s correction, * p  = 0.0176 for 72 h. Source data are provided as a source data file. F Representative immunostained colonies showing BrdU incorporation in control and PRMT1KD hTSC when cultured in stem-state culture conditions for 72 h. G Relative intensity of BrdU-positive nuclei in control and PRMT1KD hTSCs. Data are representative of 6 independent experiments ( n  = 6) and indicated as mean ± SEM. Two-tailed unpaired t -test with Welch’s correction, **** p  < 0.0001). Source data are provided as a source data file. H Micrographs show inefficient organoid formation by P RMT1KD hTSCs. I Quantitation of primary organoid diameters formed with control and PRMT1KD hTSCs. Data were generated from three independent experiments. Each dot in the column represents one organoid ( n  = 24, control organoid, n  = 27, PRMT1KD organoid). Two-tailed unpaired t -test with Welch’s correction, **** p  < 0.0001). Source data are provided as a source data file. To directly assess the impact of PRMT1 loss on hTSC self-renewal, we evaluated the ability of hTSCs to form three-dimensional trophoblast organoids 31 , 35 , 36 . Control hTSCs efficiently generated large organoids that could be dissociated and passaged to form secondary organoids, demonstrating robust self-renewal capacity (Fig.  4H ). In contrast, PRMT1KD hTSCs either failed to organize into organoid structures or formed markedly smaller organoids that could not generate secondary organoids upon passaging (Fig.  4H, I ). These results indicate that PRMT1 is required to maintain the proliferative and self-renewing capacity of human trophoblast stem cells. To determine whether the impaired self-renewal of PRMT1KD hTSCs reflects altered transcriptional programs at single-cell resolution, we performed single-cell RNA sequencing (scRNA-seq) on control and PRMT1KD hTSCs maintained under stem-state culture conditions. Unsupervised Seurat clustering identified seven transcriptionally distinct cell clusters (Fig.  5A ). The majority of control hTSCs localized to clusters 1, 2, and 3, which were enriched for trophoblast stem-state regulators, including TEAD4, TP63 , and BCAM 31 , 37 – 39 (Fig.  5B, C ). Notably, PRMT1 expression was also enriched in these undifferentiated clusters (Fig.  5B ). In contrast, PRMT1KD hTSCs predominantly populated clusters 0, 4, and 5, which exhibited reduced expression of stem-state genes and increased expression of STB markers, such as CGA, CGB3 and KISS1 (Fig.  5B, C ). Cells in clusters 0 and 4, also showed enrichment of ITGA2 and NOTCH1 , which are induced in cCTBs of a developing human placenta 40 , 41 (Fig.  5B ). Interestingly, although a large number of PRMT1KD hTSCs are populated in cluster 0, identity of cells within cluster 0 is difficult to define as they simultaneously express markers associated with both column CTBs ( ITGA2 and NOTCH1 ) and STB lineage ( KISS1 ). Nevertheless, these findings indicate that PRMT1 depletion causes a widespread shift in trophoblast transcriptional identity, characterized by loss of the stem-state program and induction of differentiation-associated gene expression. Fig. 5 PRMT1 maintains transcriptome fidelity in hTSCs. A UMAP clustering of the scRNA-Seq data from control and PRMT1KD hTSCs. Clusters were generated according to transcriptome similarity. Note near complete loss of cells in clusters 1,2 and 3 and gain in cell numbers in clusters 0, 4, 5 and 6 for PRMT1KD hTSCs. B Dot plot showing average expression and percent of cells in each cluster expressing PRMT1 along with pan trophoblast (Pan TB) markers ( GATA2, GATA3 ), hTSC stem state markers ( TEAD4, TP63 , BCAM ), STB markers ( CGA, CGB and KISS1 ) and cCTB markers ( ITGA2 and NOTCH1 ) and novel marker genes identified for each cluster. Genes listed on the x-axis and clusters on the y-axis. C Single-cell feature plots showing expression of trophoblast marker genes in specific clusters, the scale represents the level of expression. A UMAP clustering of the scRNA-Seq data from control and PRMT1KD hTSCs. Clusters were generated according to transcriptome similarity. Note near complete loss of cells in clusters 1,2 and 3 and gain in cell numbers in clusters 0, 4, 5 and 6 for PRMT1KD hTSCs. B Dot plot showing average expression and percent of cells in each cluster expressing PRMT1 along with pan trophoblast (Pan TB) markers ( GATA2, GATA3 ), hTSC stem state markers ( TEAD4, TP63 , BCAM ), STB markers ( CGA, CGB and KISS1 ) and cCTB markers ( ITGA2 and NOTCH1 ) and novel marker genes identified for each cluster. Genes listed on the x-axis and clusters on the y-axis. C Single-cell feature plots showing expression of trophoblast marker genes in specific clusters, the scale represents the level of expression. To further understand how PRMT1 shapes the global gene expression program during human trophoblast development, we performed bulk RNA-seq in control and PRMT1KD hTSCs. Differential expression analysis identified 2753 upregulated and 3205 downregulated genes ( ≥ 1.5-fold) in PRMT1KD cells (Fig.  6A , Dataset  S2 ). Notably, key stem-state regulators, including TEAD4 and MYBL2 , both implicated in CTB progenitor maintenance and recurrent pregnancy loss (RPL), were significantly repressed (Figs.  6A S6A and Dataset  S2 ). RT-qPCR and immunofluorescence analyses confirmed loss of TEAD4 and MYBL2 expression (Figs.  6B , S6B ). In addition, mRNA expression of TIMP1/3, MMP14, TFPI, ITGB1, FSTL1, IGF2 42 , ISG15 and IFI27 , which are induced in invasive EVT lineage or are implicated in EVT development 34 , 36 , 43 – 47 , were markedly downregulated in PRMT1KD hTSCs (Figs.  6A , S6A and Dataset  S2 ). Conversely, genes that are associated with STB development, such as GCM1 48 , 49 , OVOL1 50 , HSD11B2 51 and KISS1 52 , along with CGB1/7, ERVFRD-1, SDC1, PRKCZ and multiple PSG family members were strongly induced (Figs.  6A , S6A and Dataset  S2 ). Interestingly, GATA2 and GATA3 , regulators implicated in both hTSC maintenance and differentiation 53 , were also reduced. Together, the scRNA-seq and bulk RNA-seq analyses indicate that PRMT1 sustains transcription of trophoblast stem-state regulators while restraining activation of the STB gene program. Fig. 6 PRMT1 function in hTSCs prevents spontaneous STB development and promotes EVT development. A Volcano plot showing global gene expression changes in PRMT1KD hTSCs. Unbiased RNA-seq analyses were performed, and ≥1.5-fold gene-expression changes in PRMT1KD hTSCs with a false discovery rate of P  < 0.05 were indicated with colored dots (blue: up-regulated, red: down-regulated). Downregulation of TEAD4 ( a marker of undifferentiated CTBs ), MYBL2 ( important for CTB and EVT development ), GATA2 (a master regulator for trophoblast lineage) and TFPI (Important for EVT development) and Upregulation of CGB1, SDC1 and ERVFRD-1 (STB-specific genes) are indicated. B Immunostaining images confirming loss of TEAD4 and MYBL2 protein expression in PRMT1KD hTSCs. C Representative micrographs showing colony morphology of control and PRMT1KD hTSCs after they were cultured for two passages in stem state culture conditions. Note the presence of cells with differentiated morphology (white dotted area) in the PRMT1KD hTSC colony. D Representative hTSC colonies, maintained at stem state culture condition, showing hCGβ induction upon loss of PRMT1. E Representative micrographs showing STB differentiation of control and PRMT1KD hTSCs after they were cultured in STB culture conditions. F Immunostaining images confirming strong hCGβ induction in both control and PRMT1KD hTSCs. G Representative micrographs showing EVT differentiation of control and PRMT1KD hTSCs, showing EVT differentiation was inhibited in PRMT1KD hTSCs. H Representative immunostained images of HLA-G (EVT marker) showing inefficient EVT differentiation in PRMT1KD hTSCs. I Representative micrographs showing that EVT development is impaired from first-trimester human placental explants, when treated with shRNAs against PRMT1 . A Volcano plot showing global gene expression changes in PRMT1KD hTSCs. Unbiased RNA-seq analyses were performed, and ≥1.5-fold gene-expression changes in PRMT1KD hTSCs with a false discovery rate of P  < 0.05 were indicated with colored dots (blue: up-regulated, red: down-regulated). Downregulation of TEAD4 ( a marker of undifferentiated CTBs ), MYBL2 ( important for CTB and EVT development ), GATA2 (a master regulator for trophoblast lineage) and TFPI (Important for EVT development) and Upregulation of CGB1, SDC1 and ERVFRD-1 (STB-specific genes) are indicated. B Immunostaining images confirming loss of TEAD4 and MYBL2 protein expression in PRMT1KD hTSCs. C Representative micrographs showing colony morphology of control and PRMT1KD hTSCs after they were cultured for two passages in stem state culture conditions. Note the presence of cells with differentiated morphology (white dotted area) in the PRMT1KD hTSC colony. D Representative hTSC colonies, maintained at stem state culture condition, showing hCGβ induction upon loss of PRMT1. E Representative micrographs showing STB differentiation of control and PRMT1KD hTSCs after they were cultured in STB culture conditions. F Immunostaining images confirming strong hCGβ induction in both control and PRMT1KD hTSCs. G Representative micrographs showing EVT differentiation of control and PRMT1KD hTSCs, showing EVT differentiation was inhibited in PRMT1KD hTSCs. H Representative immunostained images of HLA-G (EVT marker) showing inefficient EVT differentiation in PRMT1KD hTSCs. I Representative micrographs showing that EVT development is impaired from first-trimester human placental explants, when treated with shRNAs against PRMT1 . Consistent with these transcriptional changes, PRMT1KD hTSCs exhibited a spontaneous tendency to adopt STB-like features during prolonged culture in stem-state conditions, as evidenced by loss of stem-state colony morphology and induction of hCG-β (Figs.  6C, D , S6C ). Under STB differentiation conditions, PRMT1KD cells efficiently formed multinucleated STBs (Fig.  6E, F ), indicating that PRMT1 is not required for STB differentiation but rather prevents premature acquisition of STB fate in undifferentiated trophoblast progenitors In contrast, EVT differentiation, assessed by morphological features and marker gene expression, was markedly impaired in PRMT1KD hTSCs cultured under EVT differentiation conditions (Figs.  6G, H , S6D ). Consistent with this observation, RNAi-mediated depletion of PRMT1 in first-trimester placental explants strongly inhibited EVT outgrowth (Fig.  6I ). Together, these findings identify PRMT1 as a key regulator of trophoblast lineage balance, preventing premature STB differentiation while promoting EVT development. To investigate whether defective PRMT1 expression in cytotrophoblast (CTB) progenitors contributes to adverse pregnancy outcomes, we established patient-specific human trophoblast stem cell (hTSC) lines from CTBs isolated from placentas associated with idiopathic RPL. We generated hTSC lines from both PRMT1 High RPL-placentas (PRMT1 High RPL-hTSCs) and PRMT1 Low RPL-placentas (PRMT1 Low RPL-hTSCs). All lines retained key trophoblast stem cell characteristics, including expression of GATA3 and microRNAs from the chromosome 19 microRNA cluster, confirming their identity as bona fide hTSCs (Fig.  S7A, B ). Consistent with their parental CTBs, PRMT1 Low RPL-hTSCs maintained markedly reduced PRMT1 expression (Figs.  7A , S8A, B ). Functionally, these cells displayed profound defects in proliferation, organoid formation, and EVT differentiation capacity (Figs.  7B–E , S8C–E ). Notably, Ectopic expression of wild-type PRMT1 using a lentiviral vector restored these defects, significantly enhancing proliferation and EVT differentiation capacity (Figs.  7F–H , S8F, G ), indicating that PRMT1 is required for trophoblast progenitor maintenance and lineage competence. Fig. 7 Dysregulated PRMT1 expression in RPL patient-derived hTSCs (RPL-hTSCs) is associated with defective self-renewal ability and inefficient EVT development. A Representative immunofluorescence images depicting reduced PRMT1 protein expression in PRMT1 Low RPL-hTSCs. CT27 hTSC line was used as a control. B Representative micrographs showing colony morphology of control and one of the PRMT1 Low RPL-hTSC lines, when cultured in stem state culture condition. C Micrographs show organoid formation by control CT27 hTSCs and PRMT1 Low RPL-hTSCs. D Quantitation of organoid diameters developed from control CT27 and two PRMT1 Low RPL-hTSC lines. Each dot in the column represents one organoid ( n  = 14, control organoid, n  = 13, PRMT1 Low RPL-hTSC1 organoid and n  = 11, PRMT1 Low RPL-hTSC2 organoid). Two-tailed unpaired t-test with Welch’s correction, **** p  < 0.0001 for analyses between control and both of the PRMT1 Low RPL-hTSC organoids). Source data are provided as a source data file. E Immunofluorescence images showing defective EVT development and HLA-G induction in PRMT1 Low RPL-hTSCs, when cultured in EVT differentiation conditions. CT27 hTSCs was used as a control. F Representative immunofluorescence images showing rescue of PRMT1 protein expression in PRMT1 LOW RPL-hTSCs. G Representative colony morphology of PRMT1 Low RPL-hTSCs with or without the rescue of PRMT1 expression. An equal number of cells were plated and cultured for 72 h. H Rescue of PRMT1 expression promotes EVT differentiation in PRMT1 Low RPL-hTSCs. Representative micrographs show cell morphology after 6 days of culture in the EVT differentiation condition. I RT-qPCR plots show induction of TEAD4 and MYBL2 mRNA expression in PRMT1 Low RPL-hTSCs upon rescue of PRMT1 expression. Data are representative of 5 independent experiments ( n  = 5) and the values are normalized to 18S rRNA and indicated as mean ± SEM. Two-tailed unpaired t -test with Welch’s correction, **** p  < 0.0001 for both TEAD4 and MYBL2 ). Source data are provided as a source data file. A Representative immunofluorescence images depicting reduced PRMT1 protein expression in PRMT1 Low RPL-hTSCs. CT27 hTSC line was used as a control. B Representative micrographs showing colony morphology of control and one of the PRMT1 Low RPL-hTSC lines, when cultured in stem state culture condition. C Micrographs show organoid formation by control CT27 hTSCs and PRMT1 Low RPL-hTSCs. D Quantitation of organoid diameters developed from control CT27 and two PRMT1 Low RPL-hTSC lines. Each dot in the column represents one organoid ( n  = 14, control organoid, n  = 13, PRMT1 Low RPL-hTSC1 organoid and n  = 11, PRMT1 Low RPL-hTSC2 organoid). Two-tailed unpaired t-test with Welch’s correction, **** p  < 0.0001 for analyses between control and both of the PRMT1 Low RPL-hTSC organoids). Source data are provided as a source data file. E Immunofluorescence images showing defective EVT development and HLA-G induction in PRMT1 Low RPL-hTSCs, when cultured in EVT differentiation conditions. CT27 hTSCs was used as a control. F Representative immunofluorescence images showing rescue of PRMT1 protein expression in PRMT1 LOW RPL-hTSCs. G Representative colony morphology of PRMT1 Low RPL-hTSCs with or without the rescue of PRMT1 expression. An equal number of cells were plated and cultured for 72 h. H Rescue of PRMT1 expression promotes EVT differentiation in PRMT1 Low RPL-hTSCs. Representative micrographs show cell morphology after 6 days of culture in the EVT differentiation condition. I RT-qPCR plots show induction of TEAD4 and MYBL2 mRNA expression in PRMT1 Low RPL-hTSCs upon rescue of PRMT1 expression. Data are representative of 5 independent experiments ( n  = 5) and the values are normalized to 18S rRNA and indicated as mean ± SEM. Two-tailed unpaired t -test with Welch’s correction, **** p  < 0.0001 for both TEAD4 and MYBL2 ). Source data are provided as a source data file. To define the molecular basis of these defects, we performed RNA-seq on four PRMT1 High and four PRMT1 Low RPL-hTSCs lines maintained in stem-state conditions. Principal component analysis showed that PRMT1 High RPL-hTSCs clustered with control CT27 hTSCs (Fig.  S9A ), whereas PRMT1 Low RPL-hTSCs, except one line, formed a distinct transcriptional cluster. PRMT1 Low RPL-hTSCs exhibited downregulation of trophoblast stem-state regulators, including TEAD4 and MYBL2 , together with upregulation of STB-associated genes (Fig.  S9B , Dataset  S3 ). Furthermore, comparative analyses revealed substantial overlap with transcriptional changes observed in PRMT1 KD hTSCs, with 790 of 1708 upregulated genes and 830 of 1323 downregulated genes shared between the two datasets (Fig.  S9C, D , Datasets  2 , 3 ). Substantial overlaps of differentially expressed genes are also observed between PRMT1 Low RPL-Placentas and PRMT1 Low RPL-hTSCs (Fig.  S9E, F , Datasets 1 , 3 ). Given the established roles of TEAD4 and MYBL2 in trophoblast progenitor maintenance and their reported association with RPL, we further examined their expression in RPL-hTSCs. Transcripts of both genes were significantly reduced in PRMT1 Low RPL-hTSCs compared to control CT27 hTSCs (Fig.  S10A ), whereas PRMT1 restoration strongly induced TEAD4 and MYBL2 expression (Fig.  7I ). To test whether TEAD4 and MYBL2 function downstream of PRMT1, we ectopically expressed each factor in PRMT1 Low RPL-hTSCs. Forced expression of TEAD4 or MYBL2 restored proliferative capacity in PRMT1-deficient hTSCs (Fig.  S10A–D ). Notably, MYBL2 overexpression also partially rescued EVT differentiation potential (Fig.  S10E, F ), indicating a role for MYBL2, downstream to PRMT1, in promoting EVT differentiation. Together, these findings identify PRMT1 as a likely upstream regulator of TEAD4–MYBL2–dependent transcriptional programs in human trophoblast progenitors. Loss of PRMT1 disrupts trophoblast stem-state maintenance and EVT differentiation competence, providing a mechanistic link between defective trophoblast development and idiopathic RPL. To investigate how PRMT1 regulates transcription of key trophoblast stem-state genes such as TEAD4 and MYBL2 , we examined PRMT1-mediated histone arginine methylation in trophoblast cells. Earlier studies showed that PRMT1-mediated H4R3Me2a modification at chromatin loci establishes transcriptionally active chromatin 21 , 24 . Consistent with this mechanism, we observed strong enrichment of H4R3Me2a in CTBs and EVTs of first-trimester human placentas as well as in self-renewing hTSCs (Fig.  8A–C , S11A ). In contrast, global H4R3Me2a levels were markedly reduced in PRMT1KD hTSCs (Fig.  8B ) and in PRMT1 Low RPL-hTSCs (Fig.  8C , D). These observations suggested that PRMT1-dependent H4R3Me2a may function as an epigenetic mechanism that promotes transcription of key trophoblast regulatory genes. Fig. 8 A PRMT1-H4R3Me2a epigenetic axis promotes transcription of key regulators in human trophoblast progenitors. A Immunofluorescence analysis showing a high level of H4R3Me2a mark in CTBs and EVT precursors in a first-trimester (week 6 + 4) human placenta. B Representative western blots showing loss of H4R3Me2a modification in PRMT1KD hTSCs. Total Histone H4 and β-Actin levels were tested for loading control. C Immunofluorescence images showing loss of H4R3Me2a modification in an RPL-TSC line with respect to the CT27 control hTSCs. D Image J quantitation of H4R3Me2a modification in control and RPL hTSCs. Data are representative of 7 independent experiments ( n  = 7) and indicated as mean ± SEM. Two-tailed unpaired t -test with Welch’s correction, **** p  = 0.0001). Source data are provided as a source data file. E Integrative Genome Viewer (IGV) tracks showcasing PRMT1, H4R3Me2a and RNA POL-II CUT&RUN-seq peaks near the TEAD4 and MYBL2 promoter regions in control and PRMT1KD hTSCs. Circles indicate statistically significant PRMT1 peaks (red), H4R3Me2a peaks (green) and RNA POL-II peaks (violet) in control TSCs, which are reduced in PRMT1KD hTSCs. F IGV tracks showcasing PRMT1 CUT&RUN Seq peaks near the TEAD4 and MYBL2 promoter regions in primary CTBs isolated from first-trimester human placentas. A Immunofluorescence analysis showing a high level of H4R3Me2a mark in CTBs and EVT precursors in a first-trimester (week 6 + 4) human placenta. B Representative western blots showing loss of H4R3Me2a modification in PRMT1KD hTSCs. Total Histone H4 and β-Actin levels were tested for loading control. C Immunofluorescence images showing loss of H4R3Me2a modification in an RPL-TSC line with respect to the CT27 control hTSCs. D Image J quantitation of H4R3Me2a modification in control and RPL hTSCs. Data are representative of 7 independent experiments ( n  = 7) and indicated as mean ± SEM. Two-tailed unpaired t -test with Welch’s correction, **** p  = 0.0001). Source data are provided as a source data file. E Integrative Genome Viewer (IGV) tracks showcasing PRMT1, H4R3Me2a and RNA POL-II CUT&RUN-seq peaks near the TEAD4 and MYBL2 promoter regions in control and PRMT1KD hTSCs. Circles indicate statistically significant PRMT1 peaks (red), H4R3Me2a peaks (green) and RNA POL-II peaks (violet) in control TSCs, which are reduced in PRMT1KD hTSCs. F IGV tracks showcasing PRMT1 CUT&RUN Seq peaks near the TEAD4 and MYBL2 promoter regions in primary CTBs isolated from first-trimester human placentas. To test this hypothesis, we performed CUT&RUN sequencing to examine three chromatin features in control and PRMT1KD hTSCs: PRMT1 occupancy, H4R3Me2a enrichment, and RNA polymerase II (POL-II) recruitment (Fig.  8E ). Genome-wide analyses identified 1547 PRMT1 target genes in hTSCs, including TEAD4 and MYBL2 (Dataset  S4 ). PRMT1 binding sites were predominantly located within promoter regions of the trophoblast genome (Fig.  S11B ), including the promoters of TEAD4 and MYBL2 (Fig.  8E ). Motif enrichment analysis using HOMER revealed significant enrichment of TEAD and STAT transcription factor motifs within PRMT1 binding regions (Fig.  S11C ), suggesting potential transcriptional partnerships at these loci. Importantly, CUT&RUN analyses in primary CTB progenitors isolated from first-trimester human placentas confirmed PRMT1 recruitment near the TEAD4 and MYBL2 promoter regions (Fig.  8F ), demonstrating that this regulatory interaction occurs in vivo. Consistent with PRMT1-mediated transcriptional activation, CUT&RUN analyses revealed strong enrichment of H4R3Me2a and robust POL-II recruitment at PRMT1 target loci, including TEAD4 and MYBL2 , in control hTSCs (Fig.  8E ). Notably, RNAi-mediated depletion of PRMT1 markedly reduced PRMT1 occupancy at these promoter regions and was accompanied by a strong reduction in both H4R3Me2a enrichment and POL-II recruitment (Fig.  8E ). To determine whether this regulatory mechanism is conserved across species, we performed CUT&RUN analyses on TSPCs of the ExE/EPC region of E7.0 mouse conceptuses. Remarkably, both Tead4 and Mybl2 loci were also bound by PRMT1 in mouse trophoblast progenitors and showed enrichment of H4R3Me2a at these sites (Fig.  S11D ). Together, these findings identify a conserved PRMT1–H4R3Me2a epigenetic axis that promotes transcription of key trophoblast stem-state regulators such as TEAD4 and MYBL2 .

Successful progression of early human pregnancy depends on precise transcriptional regulation in CTB progenitors, which must balance self-renewal with differentiation into STB and EVT lineages. Our findings identify PRMT1 as a central regulator of this balance in human trophoblast progenitors. We demonstrate a conserved epigenetic mechanism in which a PRMT1–H4R3Me2a regulatory axis shapes transcriptional programs in both human and mouse trophoblast stem/progenitor cells. Importantly, disruption of this axis is associated with idiopathic RPL, highlighting its potential physiological significance during early placental development. Based on our data, we propose a model in which PRMT1-mediated deposition of the activating histone mark H4R3Me2a promotes recruitment of RNA polymerase II at chromatin loci of key trophoblast regulatory genes, thereby sustaining transcriptional programs required for trophoblast progenitor maintenance and lineage specification. Loss of PRMT1 disrupts this epigenetic regulatory mechanism, resulting in impaired gene expression, defective trophoblast development, and compromised placentation. To our knowledge, this study provides the first evidence that dysregulation of histone arginine methylation contributes to early pregnancy failure. Our findings establish PRMT1 as a central regulator of trophoblast progenitor identity and lineage specification during placental development. A previous study by Pawlak et al. 16 using a gene-trap allele to globally knockout Prmt1 in mouse, suggested that extraembryonic tissues develop normally in homozygous embryos recovered at E6.5. However, that study did not directly examine TSPC development. Using a conditional knockout approach, our study demonstrates that PRMT1 function in trophoblast progenitors is essential for the progression of pregnancy. Notably, PRMT1 expression is markedly repressed in STBs within the developing labyrinth zone during mouse placentation, and similarly, loss of PRMT1 in hTSCs disrupts stem-state maintenance without impairing STB differentiation. These observations indicate that the essential role of PRMT1 in self-renewing trophoblast progenitors, together with its dispensability during STB development, represents a conserved feature of trophoblast lineage regulation. Our findings suggest that PRMT1 is a key regulator of trophoblast lineage balance, restraining premature STB differentiation while enabling EVT development. In hTSCs, loss of PRMT1 markedly impaired EVT differentiation, accompanied by reduced expression of EVT regulators and induction of STB-associated genes, consistent with a shift in lineage commitment. Notably, a subset of idiopathic RPL placentas exhibited reduced PRMT1 expression in CTBs and EVT precursors within anchoring villi, together with defects in villous architecture, indicative of disrupted trophoblast differentiation. Although direct functional evidence linking PRMT1 deficiency in EVT precursors to RPL pathogenesis remains limited, these observations support a model in which reduced PRMT1 biases trophoblast progenitors toward premature STB differentiation at the expense of EVT formation, potentially compromising placental invasion and function. We therefore propose that PRMT1 dysregulation represents a plausible contributor to trophoblast lineage imbalance in RPL, while recognizing that this association is hypothesis-consistent rather than causal. Interestingly, an earlier study 54 that conditionally deleted PRMT1 in mouse junctional zone trophoblasts suggested that PRMT1 function in these cells is dispensable for placental and embryonic development. These differences may reflect fundamental species-specific aspects of placentation. Unlike human and rat placentation, which involve extensive trophoblast invasion and spiral artery remodeling 55 , trophoblast invasion during mouse placentation is relatively shallow. Consequently, the mouse may not fully recapitulate the biology of invasive trophoblast development. Given recent advances enabling invasive trophoblast–specific conditional gene targeting in the rat 56 , future studies using rat models may provide a more informative system to examine the in vivo role of PRMT1 in invasive trophoblast development. In normal human placental development, PRMT1 is highly expressed in CTB progenitors and EVT precursors but is markedly suppressed in differentiated STBs. Notably, idiopathic RPL is frequently associated with reduced PRMT1 expression in CTBs and EVT precursors. This differential expression pattern raises several important questions: what mechanisms maintain PRMT1 expression in undifferentiated CTBs and EVT precursors, why is PRMT1 repressed during STB differentiation, and what leads to defective PRMT1 expression in idiopathic RPL placentas? Several signaling pathways—including epidermal growth factor (EGF) and WNT signaling—are known to support self-renewal of CTBs and hTSCs 34 , 57 , 58 , whereas TGFβ signaling promotes EVT differentiation 59 . It is therefore plausible that these pathways contribute to the induction and maintenance of PRMT1 expression in trophoblast progenitors, and that disruption of these pathways may contribute to reduced PRMT1 expression in RPL. Alternatively, given that the incidence of idiopathic RPL increases with maternal age 60 , 61 , altered PRMT1 expression could arise from age-associated genetic or epigenetic alterations affecting regulatory elements controlling PRMT1 transcription. Our findings provide a framework for future studies aimed at identifying the upstream regulatory mechanisms governing PRMT1 expression during early placental development. PRMT1 catalyzes ADMA on multiple substrates, including histone H4. In this study, we focused on the H4R3Me2a modification to establish a mechanistic link between PRMT1 activity and transcriptional regulation of key trophoblast genes such as TEAD4 and MYBL2. In addition to promoting transcription of stem-state regulators, our findings indicate that PRMT1 helps maintain trophoblast progenitor identity by suppressing transcription of STB-associated genes, including CGB, PSG, and SDC1. However, transcriptomic analyses revealed that hundreds of genes were either induced or repressed following PRMT1 depletion, raising the possibility that multiple PRMT1-dependent regulatory mechanisms contribute to gene expression control. Indeed, studies in other cellular systems have shown that PRMT1 can activate gene expression through methylation and stabilization of MLL2 62 , while also repressing transcription through modification of transcriptional regulators such as SP1/3 or through modulation of STAT signaling via methylation of PIAS1 63 , 64 . Thus, PRMT1-mediated arginine methylation of non-histone substrates may also contribute to the regulation of trophoblast gene expression programs. Our CUT&RUN analyses identified numerous PRMT1 target genes in hTSCs and CTBs. Because PRMT1 lacks a canonical DNA-binding domain, its recruitment to chromatin is likely mediated through interactions with sequence-specific transcription factors. Previous studies have shown that PRMT1 can be recruited to chromatin through interactions with transcription factors such as HNF4 or FOXO1 65 , 66 .; however, these factors are either absent or expressed at very low levels in CTBs and hTSCs. Although USF1 has been implicated in PRMT1 recruitment at other genomic loci 67 , motif analyses of PRMT1 target regions in trophoblast cells did not reveal significant enrichment of USF1 binding sites. Instead, PRMT1-bound regions showed strong enrichment for TEAD and STAT motifs. Given that TEAD family transcription factors—including TEAD1 and TEAD3—are highly expressed in trophoblast cells and play important roles in CTB self-renewal and EVT development 68 , 69 , and that STAT family members regulate trophoblast proliferation and invasion 70 , 71 . Thus, our findings indicate potential recruitment of PRMT1 at TEAD/STAT-associated chromatin regions. Because the majority of PRMT1 binding sites are located near promoter regions, future studies investigating the interactions between PRMT1 and these transcription factors will be important for understanding the mechanisms that govern trophoblast gene regulation. The PRMT family comprises nine enzymes, and transcriptomic analyses indicate that PRMT1 is the most abundantly expressed PRMT in CTBs 34 . Consistent with this observation, our findings demonstrate that PRMT1 is the major enzyme responsible for ADMA modification in human trophoblast cells. Nevertheless, other PRMT family members—except PRMT8—are also expressed in CTBs 34 and may contribute to distinct aspects of trophoblast biology. Thus, our study establishes a new conceptual framework in which PRMT1 and potentially other PRMT family members regulate trophoblast lineage specification and placental development through protein arginine methylation–dependent epigenetic mechanisms.

RPL, defined as the loss of two or more consecutive pregnancies, affects ~1–2% of implantation-confirmed pregnancies. In many cases, the underlying cause remains unknown and is classified as idiopathic RPL. This condition represents a major reproductive health challenge and is often devastating for affected patients. Increasing evidence suggests that abnormal placentation resulting from defective trophoblast lineage development is a major contributor to idiopathic RPL 1 . During the first trimester of human pregnancy, the placenta develops an extensive villous structure in which CTB progenitors establish a self-renewing stem/progenitor compartment. CTB progenitors differentiate along two major pathways: villous and extravillous. Villous differentiation occurs in floating villi exposed to maternal blood, where CTBs exit the cell cycle and fuse to form multinucleated STBs that mediate maternal–fetal exchange. In contrast, extravillous differentiation occurs in anchoring villi, where proliferative column cytotrophoblasts (cCTBs) form trophoblast columns that give rise to invasive extravillous trophoblasts (EVTs). EVTs migrate into the maternal uterus and remodel spiral arteries to increase blood flow at the uteroplacental interface. Endovascular EVTs directly replace endothelial cells within spiral arteries, whereas interstitial EVTs invade the uterine stroma and interact with maternal immune and stromal cells 2 – 4 . Proper placental development therefore requires tightly coordinated regulation of CTB self-renewal and differentiation into STBs and EVTs during early pregnancy 5 – 11 . Defects in CTB progenitor maintenance or EVT development have been implicated in RPL. We previously demonstrated that a subset of idiopathic RPL is associated with impaired CTB self-renewal resulting from reduced expression of TEA Domain Transcription Factor 4 (TEAD4) 12 . More recently, MYB Proto-Oncogene Like 2 (MYBL2) has been shown to be essential for CTB self-renewal and EVT development, and reduced MYBL2 expression has also been linked to RPL 13 . Consistent with these observations, global transcriptomic analyses of RPL placentas have revealed widespread dysregulation of genes involved in trophoblast progenitor maintenance and EVT differentiation. However, the molecular mechanisms responsible for altered transcription of key regulators such as TEAD4 and MYBL2 remain unclear. Protein arginine methyltransferases (PRMTs) regulate diverse cellular processes, including gene transcription, RNA splicing, and DNA damage repair 14 . The PRMT family consists of nine members that catalyze mono- or dimethylation of arginine residues on target proteins. Among them, PRMT1 is the predominant type I enzyme responsible for asymmetric dimethylarginine (ADMA) formation and is essential for embryonic development 15 – 18 . PRMT1-mediated asymmetric dimethylation of histone H4 at arginine 3 (H4R3Me2a) promotes transcriptional activation by facilitating recruitment of chromatin remodeling complexes and histone acetyltransferases such as CBP/p300, ultimately enhancing RNA polymerase II recruitment 19 – 24 . Despite the fundamental role of PRMT1 in regulating transcriptional programs during development, its function in human trophoblast and placental development has not been investigated. Here, we demonstrate that PRMT1 expression is conserved in trophoblast progenitors and invasive trophoblast populations in both human and rodent placentas, whereas its expression declines during STB differentiation. Importantly, reduced PRMT1 expression is associated with a subset of idiopathic RPL cases. Functional studies reveal that PRMT1 is required for maintaining self-renewal of human trophoblast progenitors and for proper EVT development while preventing premature STB differentiation. Mechanistically, PRMT1 regulates transcription of trophoblast progenitor genes by modulating H4R3Me2a enrichment at their chromatin loci. Together, our findings identify a conserved PRMT1-dependent epigenetic mechanism that maintains trophoblast progenitor identity and promotes normal placental development, thereby providing new insight into the molecular basis of idiopathic RPL.

Source Data Source Data