Progesterone receptor isoform modulation via enhancer activation regulates progesterone signaling in endometrial stromal cells

To identify enhancers of PGR, we utilized the CRISPR activation system (CRISPRa) to target candidate enhancers around the PGR locus and measured PGR levels via RT-qPCR and western blot. Candidate PGR enhancers were identified through integrative analysis of published genome-wide data for H3K27ac enrichment ( GSE244729 ) and chromatin looping (HiC GSE244731 and HiChIP GSE295512 ) from both endometrial stromal cells and myometrial biopsies ( 31 , 32 ), two compartments of the uterus that arise from the Müllerian duct mesenchyme and may have conserved regulatory elements ( Figure 1A ). Nine candidate enhancers around the PGR transcription start site (TSS) were identified and labelled D1-D3 (if located downstream of the PGR TSS) or U1-U6 (if located upstream of the PGR TSS) ( Figure 1A ). 4–5 gRNAs targeting each region were designed, pooled together, and introduced into THESC that were engineered to express dCas9-VPR ( 31 ). We also included a gRNA targeting the PGR promoter (P gRNA) ( 32 ). By RT-qPCR, four regions (D1, Promoter, U1, U2) showed upregulation of PGR mRNA levels in comparison to cells transduced with a control gRNA (NT gRNA) ( Figure 1B ). The D1 region is located at the PGR 3’UTR and the U1 and U2 regions are located 60 and 220 kb upstream the PGR TSS, respectively. Via western blot, we show that targeting the promoter dominantly activated PGR-B (average PGR-A:PGR-B ratio of 0.24), while targeting the U1 and U2 enhancers activated PGR isoforms to relatively equal levels (average PGR-A:PGR-B ratio of 0.95 and 0.94, respectively) ( Figure 1C , 1D ; Figure S1 ). The results were replicated in a second endometrial stromal cell (H1644) for the promoter and U2 enhancer ( Figure S2 ). Published H3K27ac data in primary hESC ( GSE295512 ) shows that upon in vitro decidualization, there is increased contact between the U2 enhancer and the PGR promoter ( Figure 1E ) ( 31 ). These results identify two distal PGR enhancers that regulate the PGR isoform ratio and may be involved in endometrial stromal cell biology. To investigate how progesterone signaling is influenced by changes in the PGR isoform ratio in THESC, we altered the PGR ratio to be PGR-B dominant by targeting the PGR promoter (termed PGRB-dominant cells) or to have equal levels of PGR isoforms by targeting the U1 and U2 enhancers simultaneously (termed PGRA/B-equivalent cells) ( Figure 2A ; Figure S3A , B ). Both cell-lines expressed relatively equal levels of total PGR expression ( Figure 2B ). We investigated how altering the PGR isoform ratio impacts transcriptomic regulation by treating cells with 1 uM medroxyprogesterone acetate (MPA) or vehicle for 72 h and conducting bulk RNA-sequencing. The optimal concentration and timepoint for MPA treatment were assessed by measuring FKBP5 levels via RT-qPCR in parental THESC ( Figure S3C ). Comparing gene expression in MPA and vehicle treated cells, we identified 956 and 1,232 differentially expressed genes (DEGs) in PGRB-dominant and PGRA/B-equivalent cell-lines, respectively ( Figure 2C ; Dataset S1 , S2 ; Figure S3D , E ). 87–88% of these genes were active (FPKM > 1) in the midsecretory endometrium ( Figure 2D ; Dataset S3 ) ( GSE132713 ) ( 36 ). IPA pathway analysis of DEGs revealed both shared and distinct pathways regulated by progesterone under different PGR isoform ratios ( Figure 2E – F ; Dataset S1 , S2 ). Shared pathways included cytoskeleton and ECM remodeling, inflammatory pathway involved in decidualization (prolactin, TGF-B signaling), maternal-fetal immune tolerance (STAT3, IL-17A, IL-2 signaling) and placental development (HIF1a, VEGF signaling), myometrial contractility pathways (gap junction, oxytocin signaling), and metabolic pathways (insulin receptor and type II diabetes signaling) ( Figure 2D ). In contrast, pro-inflammatory pathways (IL-15, IL-17), labor inducing pathways (IL-8, IL-6, eicosanoid signaling), and estrogen signaling showed opposing z-scores between cell lines ( Figure 2E ). Furthermore, we also identified several pathways involved in cell cycle regulation (cell cycle checkpoints, cyclins, and synthesis of DNA) that were uniquely enriched in DEGs in PGRB-dominant cells only ( Figure 2E ). These results suggest that shifting the PGR isoform ratio alters progesterone signaling in endometrial stromal cells. To further investigate how progesterone signaling is influenced by changes in the PGR isoform ratio, we overlapped DEGs in PGRB-dominant and PGRA/B-equivalent cells. We identified 701 overlapping genes ( Figure 3A ; Dataset S4 ), with all of them changing in the same direction ( Figure 3B ). There were 531 and 255 and genes that were only regulated by MPA in the PGRB-dominant and PGRA/B-equivalent cells, respectively ( Figure 3C , 3E ; Dataset S4 ). IPA pathway analysis of the 531 DEGs in PGRB-dominant cells revealed a significant downregulation of cell cycle and DNA synthesis pathways, alongside an upregulation of senescence and GADD45 signaling, a mediator of cell-cycle arrest ( Figure 3D ). Pro-inflammatory pathways such as IL-8 and cytokine storm signaling, as well as the fibrosis pathway, were also downregulated ( Figure 4D ). IPA pathway analysis of the 255 DEGs in PGRA/B-equivalent cells showed a distinct pro-inflammatory gene expression profile with positive enrichment for pathways including IL-8, JAK/STAT, and NF-kB signaling, as well as fibrosis and arthritis pathways ( Figure 3F ). There was also a strong enrichment of pathways regulating the extracellular matrix and integrin interactions ( Figure 3F ). Upstream regulator analysis revealed opposing z-scores for the cell cycle regulator and proto-oncogene MYC between the two cell lines ( Figure 3G ). In PGRB-dominant cells, pro-inflammatory regulators such as TNF, IL-6, and IL1 were strongly downregulated ( Figure 3G ). Furthermore, SP600125 – a molecule commonly used to study inflammation as it inhibits the key inflammatory molecule JNK – was inferred to be upregulated in PGRB-dominant cells only ( Figure 3G ) ( 40 ). These results suggest that in endometrial stromal cells, progesterone’s regulation of inflammatory genes may be influenced by a shift in the PGR isoform ratio, with PGRA/B-equivalent cells showing a pro-inflammatory profile and PGRB-dominant cells showing an anti-inflammatory profile. We further investigated the cell cycle-related DEGs that were uniquely regulated in PGRB-dominant cells and found these genes to be strongly upregulated in PGRB-dominant cells in the absence of MPA in comparison to PGRA/B-equivalent cells ( Figure 4B ; Dataset S5 ). Comparing the transcriptomic profile of vehicle treated PGRB-dominant and PGRA/B-equiv cells revealed 54 DEGs enriched in cell cycle and inflammatory pathways ( Figure 4B ). RT-qPCR confirmed significant upregulation of CDK2, CDC6, and CCNB2 in PGRB-dominant cells in the absence of MPA ( Figure 4C ). Consistent with this, vehicle-treated PGRB-dominant cells showed a significant increase in the % EdU+ cells and increased cell viability, in comparison to PGRA/B-equivalent cells, suggesting increased proliferation rates ( Figure 4D , E ). All together, these results suggest that an unliganded PGRB-dominant state may promote stromal cell proliferation. We next investigated the PGR cistrome via Cut&Run assays in PGRB-dominant and PGRA/B-equivalent cells treated with MPA or vehicle for 3 h. In both cell lines, MPA treatment increased PGR genomic binding, with 3,793 and 3,509 peaks in PGRB-dominant and PGRA/B-equivalent cells, respectively ( Figure 5A ; Dataset S6 ). Comparative analysis in MPA-treated cells revealed 2,517 shared binding sites, with PGRB-dominant cells showing increased occupancy at 1,276 sites and PGRA/B-equivalent cells at 1,014 sites ( Figure 5A ). In both cell-lines, roughly 28% of peaks were located at the promoter or the 5’UTR of a gene ( Figure 5B ). Motif analysis identified PRE as the top enriched binding motif, along with motifs for bZIP proteins (Fra1/2), TEAD, RUNX, and SOX factors ( Figure 5C ). Forkhead motifs (FOXA1, FOXH1) were specific to PGRA/B-equiv cells, while PGRB-dom cells showed enrichment for ZBTB factors (ZBTB12) and the chromatin remodeler Arid5a ( Figure 5C ). Next, we identified genes with a PGR peak at the promoter or one chromatin loop away from the promoter using published H3K27ac HiChIP in decidualized hESC ( GSE295512 ) using the same protocol as cited in Montague Redecke et al. ( 31 ). To identify more direct targets of PGR, we overlapped these genes with the corresponding DEGs from the same cell-lines in the RNA-seq dataset ( Figure 5D , E ). In PGRB-dominant cells, we identified 3,104 peak-associated genes, of which 296 were differentially expressed in response to MPA ( Figure 5D ; Dataset S7 ). In PGRA/B-equivalent cells, we identified 2,922 peak-associated genes, of which 234 were differentially expressed in response to MPA ( Figure 5E ; Dataset S7 ). Overlapping the peak-associated DEGs in each cell-line showed that 160 were shared, while 126 and 74 DEGs were unique to PGRB-dominant and PGRA/B-equivalent cells. Pathway analysis of the 126 DEGs in PGRB-dominant cells showed downregulation of arthritis and fibrosis pathways, wound healing, IL-17 signaling, and cholesterol pathways (cholesterol biosynthesis, DHCR24 signaling) ( Figure 5F ). Conversely, pathway analysis of the 74 DEGs in PGRA/B-equivalent cells showed upregulation of fibrosis and arthritis pathways, wound healing, IL-4 signaling, cytokine storm signaling, and HIF1a signaling ( Figure 5G ). Altogether, these results indicate that the direct regulation of progesterone-responsive pathways in hESC is influenced by the relative ratio of PGR isoforms. Given the physiological importance of balanced PGR isoform expression in endometrial receptivity, decidualization, and pregnancy maintenance, we identified potential transcription factors (TFs) that bind to the U1 and U2 enhancers using public ChIP-seq datasets (ReMAP Atlas) ( 41 ) and motif enrichment (JASPAR Transcription Factor Binding Site Database). We refined the list of candidates by filtering for (1) TFs with evidence of binding or predicted binding motifs within these enhancers, (2) genes that are expressed in the endometrium at the protein level using Human Protein Atlas ( proteinatlas.org ) ( 42 ), and (3) genes with known roles in reproductive or embryonic phenotypes using the Mouse Genome Informatics database ( https://www.informatics.jax.org/ ) ( Figure 6A ). We identified potential regulators of PGR, including chromatin remodelers like ARID1A, BRD4, and SMARCB1, as well as TFs like ESR1, STAT3, FOXO1, and SOX4. Among these, ESR1 stood out as a known hormonal regulator of PGR during the proliferative phase ( 31 ). Using THESC engineered for ESR1 activation ( 31 ), we found that E2 treatment activated PGR in a PGR-A:PGR-B ratio > 1 ( Figure 6B ; Figure S4 ). Since U1 and U2 enhancers lack the canonical ESR1 binding motif ERE, we hypothesized that ESR1 regulation of PGR-A may occur indirectly via protein-protein tethering. Overlapping ESR1/E2 regulated genes from published RNA-seq ( 31 ) with the candidate PGR regulators revealed FOXO1 as a putative mediator. Knockdown of FOXO1 via siRNA attenuated ESR1/E2’s ability to activate PGR ( Figure 6C ; Figure S5 ), but FOXO1 was unable to activate PGR on its own ( Figure S6 ). FOXO1 Cut&Run-PCR showed that activation of ESR1 in THESC led to FOXO1 binding at the U2 enhancer in both the presence and absence of E2 ( Figure 6D ). Importantly, FOXO1 binding at the U2 enhancer overlapped with the ESR1 binding sites in proliferative endometrial biopsies ( GSE200807 ) ( 39 ) ( Figure 6E ). Altogether, this data suggests that FOXO1 may be involved in ESR1’s regulation of PGR through binding at the U2 enhancer.

Telomerase-transformed human endometrial stromal cells, THESC (ATCC, CRL-4003) ( 29 ), were cultured in DMEM/F-12 (Gibco Thermo Fisher Scientific, Waltham, MA, USA) supplemented with 10% Fetal Bovine Serum (FBS, Gibco Thermo Fisher Scientific) and antibiotics (10 000 IU/mL penicillin, 10 000 IU/ mL streptomycin; Life Technologies, Grand Island, NY, USA), termed hESC media. Telomerase transformed human endometrial stromal cells, H1644, were additionally supplemented with 2 mM sodium pyruvate (Gibco Thermo Fisher Scientific) as described in Song et al. ( 30 ). Cells were cultured and grown in a 5% CO 2 and 37°C incubator. THESC and H1644 were engineered to express dCas9-VPR (CRISPRmod CRISPRa lentiviral dCas9-VPR, Dharmacon/Horizon Discovery, Lafayette, CO, USA) as cited in Montague Redecke et al. ( 31 ). THESC dCas9-VPR and H1644 dCas9-VPR were maintained in hESC media supplemented with 4 ug/mL blasticidin and without the penicillin-streptomycin. ESR1 activated THESC were engineered as cited in Montague Redecke et al. ( 31 ). Candidate cis-regulatory elements of PGR were identified by integrative analysis of published datasets: H3K27ac Cut&Run in THESC (GSExxx), H3K27ac ChIP-seq in myometrial biopsies ( GSE244729 ) ( 32 ), H3K27ac HiChIP in primary hESC ( GSE295512 ) ( 31 ), and HiC in myometrial biopsies ( GSE244731 ) ( 32 ). Nine regions were selected that had enrichment for H3K27ac and looped to the PGR promoter in either HiChIP or Hi-C data (regions D1 – D3; U1 – U6). 4–5 guide RNAs (gRNA) targeting each region were designed using the CHOPCHOP ( 33 , 34 ) and CRISPick tools ( 35 ) ( Table 1 ). The PGR promoter targeting gRNA was obtained from Wu et al. ( 32 ). All gRNA expression vectors were synthesized by and acquired from VectorBuilder ( VectorBuilder.com , Chicago, IL, USA). gRNAs targeting each region were pooled together in equal ratios and lentivirus was generated by transfection of the constructs and packaging vector into HEK293T cells with the support of the NIEHS Viral Vector Core Facility. PGRB-dominant cells were engineered by transduction of dCas9-VPR engineered hESC with Promoter targeting gRNA at multiplicity of infection (MOI) 12. PGRA/B-equivalent cells were generated by co-transducing dCas9-VPR–engineered hESCs with pooled U1 and U2 gRNAs, each at an MOI of 6. RNA was isolated using Qiagen RNeasy RNA mini prep kit (Qiagen, Hilden, Germany) and reverse transcribed into cDNA using Moloney Murine Leukemia Virus reverse transcriptase (Thermo Fisher Scientific) according to manufacturer protocol. Quantitative real time PCR was performed using SsoAdvanced Universal SYBR Green Supermix (Bio-Rad, Hercules, CA, USA). Each reaction was performed in technical duplicates and relative mRNA levels were determined in comparison to 18S control. Primers used are available in Table 2 . Protein isolation and western blot was performed as cited in Montague Redecke et al. ( 31 ). Briefly, protein was isolated using RIPA Lysis and Extraction Buffer (#89900, Thermo Fisher Scientific) with cOmplete EDTA-free Protease Inhibitor Cocktail (Roche, Basel, Switzerland) and Phosphatase inhibitor (PhosSTOP, Roche). Protein was separated on 4–20% Criterion TGX Precast Gels (Bio-Rad, #5678094) with the Precision Plus Protein Dual Color Standards Ladder (Bio-Rad, #1610374). Membranes were blocked with antibodies for PGR (Dako, Cat. No. GA09061–2, clone PgR 1294, Glostrup, Denmark; 1:1000 dilution), ESR1 (Cell Signaling Technology, Cat. No. 13258, clone D6R2W, Danvers, MA, USA; 1:500 dilution), FOXO1 (Cell Signaling Technologies, Cat. No. 2880, clone C29H4; 1:1000 dilution), or GAPDH (Santa Cruz Biotech, sc-25778, Dallas, TX, USA; 1:1000 dilution). Blots were subsequently incubated with either anti-Rabbit IgG or anti-Mouse IgG secondary antibodies labeled with IR 800 (LI-COR Biosciences, Lincoln, NE, USA). Membranes were imaged using Odyssey Fc Imager (LI-COR Biosciences) using 800nm channel for 10 min, and 700nm channel for 30 sec. Band intensity was quantified using Image Studio (Version 5.5). Cells were seeded in hESC media and allowed to grow for 24 h. Cells were pre-treated with reduced FBS hESC media containing 2% FBS in the place of 10% for 24 h. Media was then changed to OptiMEM (Gibco Thermo Fisher Scientific) with 2% Charcoal Stripped Serum (csFBS, Gibco Thermo Fisher Scientific) and 1% Penicillin-Streptomycin (Life Technologies), termed low serum media. Low serum media was supplemented with 1 uM Medroxyprogesterone Acetate (Sigma-Aldrich, St. Louis, MO, USA) or 0.01% ethanol, termed MPA media or vehicle media. Cells were incubated for varying amounts of time (indicated under each experiment) prior to isolating RNA (RT-qPCR and RNA-seq), or conducting MTT assay, DNA synthesis assay, or Cut&Run. Plates seeded with PGRA/B-equivalent and PGRB-dominant THESC were washed with PBS and media was changed to 3 mL per well MPA media or vehicle media for 72 h. RNA was isolated as mentioned above, and libraries were prepared and sequenced as 75 bp paired-end reads by the NIEHS Sequencing Core using the Illumina Tru-seq Stranded mRNA library kit and Illumina NextSeq 500 instrument (Illumina, New England Biolabs, Ipswich, MA, USA). For RNA-seq analysis, reads were processed by trimming adapters and performing optical and PCR deduplication using BBMap (Version 39.01). Reads were aligned to the hg38 genome (GCF_000001405.40) using STAR (Version 2.6.0c) and differential expression analysis was performed using edgeR (Version 4.2.2) in R (Version 2024.12.1). Differentially expressed genes (DEGs) were defined as genes with |FC| > 1.5 and adjusted p-value < 0.05. Ingenuity Pathway Analysis software (IPA; www.ingenuity.com ; Version 01.22.01) and Metascape ( http://metascape.org ; Version X0200) were used for pathway analysis. For published RNA-seq ( GSE132713 ) ( 36 ), normalized counts were downloaded from GEO and filtered to include genes with average FPKM > 1 in the Human endometrium Midsecretory samples only. 1,000 cells were seeded in 96-well plates. 24 h later, media was changed to 100 uL per well vehicle media. Media was changed every two days and cells were incubated for up to 5 days. Cell viability was assessed using the Cell Proliferation Kit I (MTT) (Roche) according to the manufacturer’s protocol. Relative cell viability was calculated by determining the absorbance at 570 nm and subtracting the reference at 670 nm. 100,000 cells per well were seeded in 10 cm plates. Cells were pre-treated with reduced FBS hESC media containing 2% FBS in the place of 10% for 24 h. Media was then changed to vehicle media for 72 h. Cells were treated with 10 μM EdU for 3 h and EdU was detected using the Click-iT EdU Alexa Fluor 488 Flow Cytometry Assay Kit (Thermo Fisher Scientific) following the manufacturers protocol. % EdU positive cells were counted using the LSR Fortessa cell counter (BD Biosciences, Franklin Lakes, NJ, USA) thanks to the help of the NIEHS Flow Cytometry Center. ESR1-activated THESC were engineered as described in Montague Redecke et al. ( 31 ). Cells were seeded in hESC media and allowed to grow for 24 h. Media was then changed to low serum media supplemented with 10nM 17β-estradiol (Sigma-Aldrich), termed E2 media, or 0.01% ethanol (vehicle media) and cells were incubated for 24 h before collecting for Cut&Run or protein isolation. The PGR and FOXO1 Cut&Run samples were prepared using the CUTANA ™ ChIC/CUT&RUN Kit (Version 3.5; EpiCypher, Durham, NC, USA) as cited in Montague Redecke et al. ( 31 ). For PGR Cut&Run, PGRA/B-equivalent and PGRB-dominant cells were pre-treated with reduced FBS media and subsequently treated with MPA or vehicle media for 3 h to assess rapid PGR binding. PGR Cut&Run was conducted using a 1:25 dilution of the PGR antibody (Santa Cruz Biotechnology, Cat. No. sc-7208) incubated in antibody incubation buffer overnight at 4°C. For FOXO1 Cut&Run, ESR1-activated or control THESC were treated with E2 or vehicle media for 24 h and Cut&Run was performed using a 1:50 dilution of the FOXO1 antibody (Cell Signaling Technologies, Cat. No. 2880, clone C29H4). Targeted DNA amplicons were purified using the Qiagen QIAquick PCR Purification Kit (Qiagen). FOXO1 Cut&Run-PCR was conducted as mentioned above with primers targeting the U2 enhancer and negative control gene RPL30 ( Table 2 ). Sequencing libraries were prepared using the NEBNext Ultra II DNA library preparation kit for Illumina (New England Biolabs), with 13 cycles of PCR amplification. The libraries were sequenced by the NIEHS sequencing core using the Illumina NextSeq 500 instrument Reads from CUT&RUN were trimmed of adapter sequences using Trim Galore! ( https://www.bioinformatics.babraham.ac.uk/projects/trim_galore/ ; Version 0.6.10) with default settings except for specifying paired-end validation. Trimmed sequences were aligned to hg38 and E. coli K-12 MG1655 genomes using bowtie2 (Version 2.5.2) with maximum fragment length 2,000, forward-reverse mate orientation, local read alignment, and no mixed or discordant read alignments ( 37 ). Alignment files were deduplicated using Picard ( http://broadinstitute.github.io/picard ; Version 3.2.0) and filtered for reads with MAPQ 5 using samtools (Version 1.18) ( 38 ). Processed reads from replicate samples were concatenated and converted from paired-end format to fragment representation in BED format, then peak calls were performed via MACS2 (Version 2.2.9.1) with the callpeaks function and parameters “-f BED -g hs -q 0.001 --keep-dup all --nomodel --extsize 100”, using IgG samples as background control. Prior to downstream analysis, peaks overlapping blacklist regions were discarded. Motif enrichment analysis was performed using HOMER findMotifsGenome.pI function, using the “-size given -p 10” parameters. Peaks were annotated for overlaps with HiChIP loop anchors in EPC treated hESC from ( GSE295512 ), as cited in Montague Redecke et al. ( 31 ), and only peaks overlapping a gene’s promoter or a HiChIP anchor that loops to a genes promoter were used for integration with RNAseq. For published H3K27ac HiChIP, raw Fastq reads for EPC treated samples ( GSM8951706 and GSM8951708 ) were downloaded from GEO ( GSE295511 ) and processed and integrated with Cut&Run data as cited in Montague Redecke et al. ( 31 ). For H3K27ac ChIP-seq in human myometrial biopsies, processed bigwig files for sample GSM7825096 were downloaded from GEO ( GSE244729 ) and visualized on the UCSC genome browser ( 32 ). For HiC in human myometrial biopsies, raw reads for sample GSM7825116 were downloaded from GEO ( GSE244731 ) and processed and visualized as cited in Wu et al. ( 32 ). For published ESR1 ChIP-seq in proliferative endometrial biopsies, processed bigwig files for sample GSM6043914 were downloaded from GEO ( GSE200807 ) and visualized on the UCSC genome browser ( 39 ). ESR1-activated or control THESC seeded in 6-well plates were washed with PBS and transfected with 60 nM nontargeting siRNA (siNT) or FOXO1-targeting siRNA (siFOXO1) (ON-TARGETplus siRNA Smartpool, Dharmacon/Horizon Discovery) using Lipofectamine RNAiMax lipid (Life Technologies) in 1 mL transfection media (OptiMEM (Gibco Thermo Fisher) with 2% Charcoal Stripped Serum (Gibco Thermo Fisher). After 48 h of transfection exposure, wells were washed with PBS and treated with E2 or vehicle media for 24 h before protein was collected for western blot. Statistical data was expressed in a mean ± SEM of the 2–6 biological replicates. Statistical analysis was performed using GraphPad Prism (Version 10.4.0; GraphPad Software Inc.). Distribution of datapoints was assessed by Shapiro-Wilk test. Statistics were performed by parametric unpaired t test for two comparisons and One-Way ANOVA Turkey test for more than two comparisons. Significance was defined as p-value < 0.05. Institutional review board (IRB) approval was not required for this study because it did not involve human participants, human data, or animal subjects.

Progesterone signaling is mediated by a balance between its two receptor isoforms, with changes in the PGR isoform ratio occurring across the menstrual cycle ( 13 ), pregnancy ( 43 ), and in the case of gynecological pathologies like endometrial cancer and endometriosis ( 15 , 25 ). While studies have shown that PGR isoform-specific expression is regulated at the transcriptional level through alternate promoters ( 7 , 8 ), its regulation by distal cis-regulatory elements and transcriptional regulators remains poorly understood. In this study, we employed a CRISPR activation screen in hESC to identify two PGR enhancers located 60 and 220 kb upstream of the PGR TSS. Activation of these enhancers with CRISPRa upregulated both PGR-A and PGR-B expression, whereas activation of the PGR promoter primarily upregulated PGR-B, altering progesterone’s regulation of inflammatory pathways, as summarized in the graphical working model in Figure 7 . These findings extend current models of PGR isoform regulation beyond promoter-level control and provide new insight into how changes in the PGR isoform ratio shape progesterone’s regulation of endometrial stromal cells. As our experiments were limited to stromal cells, it remains an open question whether these enhancers also regulate PGR expression in other reproductive tissues where the isoform balance is critical. Future studies should explore their activity in other reproductive tissues, such as the endometrial epithelium and placenta, and test their functional relevance in vivo using mouse models. Transcriptomic profiling revealed that altering the PGR isoform ratio in hESCs significantly reshaped inflammatory gene expression in response to progesterone. PGR-B dominant cells suppressed pregnancy-associated inflammatory pathways, whereas increased PGR-A expression led to their activation. These include HMGB1, IL-8 (CXCL8), IL-6, and eicosanoid signaling – pro-inflammatory pathways regulated by progestin in hESCs and endometrial tissue ( 44 ) ( 45 ) ( 46 , 47 ) that play essential roles in implantation, decidualization, and labor ( 48 ) ( 49 ) ( 50 , 51 ). For example, excess HMGB1 secretion by decidual stromal cells has been detected at the maternal-fetal interface in spontaneous abortion patients ( 48 ) ( 52 ). Elevated IL-8 and prostaglandins are found in the decidua during preterm birth ( 53 ) ( 54 ), whereas reduced IL-8 levels are associated with prolonged pregnancy ( 55 ). Notably, PGR-A dominance has been linked to pro-inflammatory profiles in the myometrium ( 21 ) and proposed as a mechanism for functional progesterone withdrawal at labor onset ( 43 ). Our findings suggest a similar paradigm in endometrial stromal cells, where shifts towards PGR-A may sensitize cells to inflammatory activation. Dysregulation of this isoform balance could therefore contribute to aberrant immune responses, leading to obstetrical complications such as implantation failure, recurrent pregnancy loss, and preterm birth. To further explore how the PGR isoform balance affects disease-relevant pathways, we next examined direct transcriptional targets of PGR by integrating DEGs with PGR Cut&Run data following acute 3 h MPA treatment. In PGRA/B-equivalent and PGRB-dominant cells, we observed inferred opposite regulation of fibrosis- and inflammatory-related signaling pathways – many of which have been implicated in endometriosis, a condition characterized by the ectopic growth of endometrial-like tissue outside the uterus ( 56 ). Notably, elevated PGR-A expression has been reported in both endometriotic lesions and the eutopic endometrium of affected patients ( 15 ). Pathways implicated in endometriosis and activated by increased PGR-A expression in our engineered cells included HIF1A signaling ( 57 ), cytokine storm signaling ( 58 ), wound healing ( 59 ), extracellular matrix organization, and IL-4 signaling ( 60 ). IL-4 - a Th2 cytokine - is elevated in the serum and ectopic stromal cells of patients with endometriosis ( 61 , 62 ), where in vitro studies suggest it promotes endometriotic stromal cell proliferation and enhances eotaxin secretion ( 62 ) ( 63 ). Collectively, these findings point towards a model in which elevated PGR-A to B levels promotes pro-fibrotic inflammatory pathways that may be implicated in disease progression. We also found that decreasing the PGR-A to B isoform ratio in hESC led to the upregulation of cell cycle genes and cell viability in the absence of progesterone. Differences in PGR isoform abundance has been observed in gynecological cancers. In endometrial cancer, studies have reported a decrease in the PGR-A to B ratio in both the epithelial and stromal compartments ( 16 , 25 ), with PGR-A downregulation associated with reduced disease-free survival of patients ( 64 ). While endometrial cancer arises from the epithelium, studies have shown that stromal PGR expression is necessary to induce the anti-tumor effects of progesterone treatment, suggesting that paracrine signaling from stromal PGR expression may be involved in the progression of endometrial cancer ( 65 ). We previously found that constitutive expression of PGR-B in PGR-positive tissues differentially regulated cell cycle related genes and led to the development of hormone-dependent ovarian neoplasms ( 66 ). Our study suggests that a decreased PGR-A to B isoform ratio in the stroma may promote cell proliferation and viability, particularly in the absence of progesterone, highlighting the need for further investigation into the underlying mechanism. Lastly, we found that ESR1 drives a PGR-A dominant expression profile in hESCs, despite the absence of canonical ERE motifs at the enhancers, suggesting potential indirect regulation via co-factors. One such factor is FOXO1, a downstream ESR1/E2-regulated gene that we identified as a mediator of ESR1-induced PGR activation. Cut&Run experiments further confirmed FOXO1 binding at the PGR U2 enhancer. Given that ESR1 and FOXO1 are critical mediators of estrogen and progesterone signaling in the endometrium ( 67 ), our findings support a model in which they co-regulate this enhancer region to fine-tune PGR isoform expression to coordinate the PGR transcriptional network. This enhancer-based regulatory mechanism may be vulnerable to disruption by environmental endocrine-disrupting chemicals (EDCs), such as genistein, bisphenol A (BPA), and phthalates ( 68 ), which interfere with ESR1 signaling and downstream transcriptional responses ( 69 ). Such disruptions could impair distal enhancer activity, skew PGR isoform expression, and contribute to progesterone resistance and reproductive dysfunction. In this study, we identified a novel enhancer-mediated mechanism regulating PGR-isoform expression in hESC. We found that in contrast to PGR promoter activation, which preferentially induced PGR-B expression, enhancer activation promoted equal expression of both PGR-A and PGR-B. CRISPR activation of these enhancers altered the progesterone-responsive transcriptome, with an increased PGR-A:B ratio promoting a proinflammatory gene expression profile. ESR1 emerged as a key modulator of the PGR-A isoform, potentially through FOXO1-mediated activation at the enhancer. These findings reveal the critical role of enhancer activity in fine-tuning the PGR isoform balance and its broader implications for endometrial function in health and disease. Therapeutically, targeting PGR isoform enhancers or upstream regulators such as ESR1 and FOXO1 may offer promising avenues to restore normal progesterone responsiveness in disorders marked by PGR isoform imbalance, including endometriosis and obstetric complications. Emerging strategies such as CRISPR activation or interference (CRISPRa/i)-mediated enhancer targeting and small molecule epigenetic modulators could be leveraged to rebalance PGR-A/B expression via targeted enhancer regulation ( 70 , 71 ). In parallel, emerging transcription factor-targeting therapeutics, such as protein–protein interaction inhibitors and DNA-binding antagonists, aimed at the ESR1-FOXO1 regulatory axis or other upstream regulators of PGR, may inform the development of isoform-specific progesterone receptor modulators (PRMs) ( 72 ). Future studies should focus on delineating the precise molecular mechanisms that govern isoform-specific enhancer activity and identifying additional isoform-specific regulators that could be leveraged for therapeutic strategies. Collectively, the results of this study advance our understanding of progesterone receptor regulation and lay the groundwork for innovative therapies to address reproductive disorders marked by progesterone resistance.

The ovarian hormone progesterone is a critical regulator of the female reproductive tract, playing a key role in the regulation of endometrial function during pregnancy and female health. Alteration in progesterone signaling can result in diseases such endometriosis and endometrial cancer, as well as infertility. The endometrium consists of a luminal epithelial layer with a connective tissue matrix comprised of endometrial stromal cells, endothelial cells, and immune cells ( 1 ). The endometrium undergoes dynamic changes in response to waves of estrogen and progesterone signaling to prepare for an implanting embryo and early pregnancy ( 2 ). During the first half of the menstrual cycle, termed the proliferative phase, estrogen promotes endometrial proliferation and primes the endometrium to become responsive to progesterone by activating the progesterone receptor (PGR) gene ( 3 ). A rise in progesterone secretion from the corpus luteum during the secretory phase regulates the differentiation of the endometrial epithelium and the proliferation and differentiation of the endometrial stromal cells. Expression of PGR in the stromal cells is critical for controlling epithelial cell proliferation, triggering stromal cell decidualization, and creating a receptive environment for embryo implantation ( 4 , 5 ). Progesterone signaling primarily exerts its effects in the uterus through PGR, with PGR null mice being infertile due to ovulatory and endometrial dysfunction ( 6 ). PGR is expressed as two unique isoforms – a full-length PGR-B and truncated PGR-A isoform - transcribed from the same PGR gene via alternate promoters ( 7 , 8 ). PGR-B and PGR-A share an identical DNA binding domain, ligand binding domain, and two transactivation domains, while PGR-B contains an additional 164 amino acids at the N-terminal domain that encodes for a third transactivation domain ( 9 ) that drives isoform-specific functions ( 10 – 12 ). The expression of PGR isoforms is dynamic in a tissue-specific and temporal manner in the uterus. The cellular response to progesterone is governed by the relative ratio of PGR-A to PGR-B ( 11 ). This ratio shifts throughout the menstrual cycle ( 13 ) and pregnancy ( 14 ), and has been found to be perturbed in the uterus of patients with gynecological pathologies like endometrial cancer and endometriosis ( 15 , 16 ). In the endometrial stroma, both PGR-A and PGR-B are expressed, with PGR-A being the dominant isoform ( 13 , 17 , 18 ). In the myometrium, PGR-B is the dominant isoform throughout gestation and mediates progesterone’s anti-contractile and anti-inflammatory effect ( 19 – 21 ). As term approaches, a switch in the PGR isoform ratio to become PGR-A dominant facilitates the functional withdrawal of progesterone to induce a pro-inflammatory contractile state for labor ( 14 , 21 , 22 ). Altered progesterone signaling through the dysregulation of PGR isoform expression is linked to impaired myometrial contractility and preterm birth ( 23 ). The PGR isoform ratio has also been found to be misregulated in gynecologic pathologies, with PGR-A overexpressed in the endometrial stroma of patients with endometriosis ( 15 ) and polycystic ovarian syndrome ( 24 ), and downregulated in both the stroma and epithelium of patients with endometrial cancer ( 16 , 25 ). Understanding how the PGR isoform ratio is regulated, and how shifts in this ratio influence uterine function, is critical to elucidating progesterone’s role in reproductive physiology and gynecological pathologies. Studies have largely focused on promoter-mediated regulation of PGR isoforms ( 23 , 26 – 28 ), and the mechanisms governing isoform-specific expression at the enhancer level remain unknown. In this study, we demonstrate an additional layer of PGR-isoform transcriptional regulation at two distal enhancers in immortalized human endometrial stromal cells (THECS). Targeting these enhancers with CRISPR activation upregulated both PGR-A and PGR-B, while targeting the promoter favored PGR-B expression. Bulk RNA-sequencing and Cut&Run assays revealed altered progesterone regulation of inflammatory and cell cycle pathways in response to changes in PGR isoform ratios. Integrative analysis of public ChIP-seq datasets and motif enrichment at the enhancers identified candidate upstream regulators of PGR, including estrogen receptor alpha (ESR1) and Forkhead box protein O1 (FOXO1). ESR1 dominantly activated PGR-A, potentially in cooperation with FOXO1 at the U2 enhancer. These findings extend current models of PGR isoform regulation beyond promoter-level control and provide new insight into how changes in the PGR isoform ratio shape progesterone’s regulation of endometrial stromal cells.