https://orcid.org/0000-0003-2815-766X
Abstract
Uterine glands and, by inference, their secretions impact uterine receptivity, blastocyst implantation, stromal cell decidualization, and placental development. Changes in gland function across the menstrual cycle are primarily governed by the steroid hormones estrogen (E2) and progesterone (P4) but can also be influenced by extrinsic factors from the stroma. Using a human endometrial epithelial organoid system, transcriptome and proteome analyses identified distinct responses of the organoids to steroid hormones and prostaglandin E2 (PGE2). Notably, P4 and PGE2 modulated the basolateral secretion of organoid proteins, particularly cystatin C (CST3), serpin family A member 3 (SERPINA3), and stanniocalcin 1 (STC1). CST3, but not SERPINA3 or STC1, attenuated the in vitro stromal decidualization response to steroid hormones and PGE2. These findings provide evidence that uterine gland-derived factors impact stromal cell decidualization, which has implications for pregnancy establishment and fertility in women.
Introduction
The inner lining of the uterus, termed the endometrium, is replete with glands that are hypothesized to be essential for the establishment of pregnancy through interactions with the embryo and other endometrial cells (Burton et al., 2002; Hempstock et al., 2004; Filant and Spencer, 2013; Kelleher et al., 2017, 2018). In women, the endometrium must undergo cyclic changes in response to preovulatory and postovulatory steroid hormones to prepare for embryo implantation and pregnancy establishment. Predominantly, hormones from the ovary [estrogen (E2) and progesterone (P4)] co-ordinate the acquisition of endometrial receptivity that involves secretory transformation of the luminal and glandular epithelium (GE), differentiation of the stroma into decidual cells, recruitment of specialized immune cells, and remodeling of the vasculature (Evans et al., 2016). Perturbations in any of those early pregnancy events can lead to adverse ripple effects that compromise pregnancy outcomes (Lédée-Bataille et al., 2002; Klemmt et al., 2006; Al-Sabbagh et al., 2012; Lucas et al., 2016; Peter Durairaj et al., 2017; Fitzgerald et al., 2018).
Differentiation of the endometrial stromal cells into decidual cells (DC) is critical for implantation and pregnancy establishment in women (Cooke et al., 1997; Li et al., 2011; Chen et al., 2013). Stromal cells undergo spontaneous decidualization during the mid-secretory phase of the menstrual cycle in response to rising levels of P4 from the ovary. Progesterone is required for stromal cell decidualization and increased intracellular cAMP hastens that process in vitro (Gellersen and Brosens, 2003). Factors, such as prostaglandin E2 (PGE2) that utilize the cAMP/protein kinase A pathway, can also activate the core gene regulatory network of DCs in concert with P4 (Stadtmauer and Wagner, 2022) and this was previously found to enhance stromal cell decidualization (Frank et al., 1994). PGE2 is produced by both endometrial epithelial and stromal cells (Smith and Kelly, 1988; Alecozay et al., 1991); however, PGE2 regulation of endometrial gland function is not well investigated. Disordered decidualization is linked to reproductive pathologies such as recurrent pregnancy loss (Klemmt et al., 2006; Velarde et al., 2009; Aghajanova et al., 2010; Sherwin et al., 2010; Al-Sabbagh et al., 2012; Macklon and Brosens, 2014; Lucas et al., 2016), implantation failure (Peter Durairaj et al., 2017), and pre-eclampsia (Garrido-Gomez et al., 2017). Uterine glands and their products have long been implicated in uterine receptivity and blastocyst implantation in domestic animals, rodents, and humans (Kelleher et al., 2019c); however, the concept that uterine glands secrete paracrine-acting factors into the stroma that impact decidualization is relatively new and primarily based on studies of glandless mice (Kelleher et al., 2017, 2018). Collective evidence supports the concept that epithelial-stromal crosstalk is essential for uterine function and pregnancy establishment.
During the past few years, 3D endometrial epithelial organoid culture models have been established that recapitulate in vivo epithelial responses to steroid hormones and are viable models to study human uterine gland function (Boretto et al., 2017; Turco et al., 2017; Fitzgerald et al., 2019; Rawlings et al., 2021). The objectives in this study were to: modify the endometrial organoid culture system to remove wingless-type MMTV integration site family, member (WNT)-activating factors determine the effects of steroid hormones and PGE2 on uterine gland function; and test the hypothesis that uterine gland-derived factors impact stromal cell decidualization. These studies provide new evidence that uterine glands secrete proteins in response to P4 and PGE2 that directly impact stromal cell decidualization.
Materials and methods
Establishment and maintenance of human organoid and primary endometrial stromal cultures
All experiments involving human subjects were approved by the Institutional Review Board of the University of Missouri, and written informed consent was obtained from each donor. Donor 1 was 25 years old, undergoing a bilateral tubal ligation, and was menstrual cycle day 7 (mid-proliferative phase). Donor 2 was 24 years old, having a diagnostic laparoscopy, and left salpingectomy and oophorectomy, and was cycle day 13. Donor 3 was 25 years old, undergoing permanent sterilization and cycle day 12. All donors were not taking any steroid-modulating medications.
Upon collection, endometrial tissue biopsies were immediately placed in Dulbecco’s modified Eagle’s medium: nutrient mixture F-12 medium (DMEM/F-12) (Gibco, 11320-033, ThermoFisher Scientific, Waltham, MA, USA) supplemented with 10% fetal bovine serum (FBS) (Sigma-Aldrich, St. Louis, MO, USA) and 1% antibiotic-antimycotic (Anti-Anti, Gibco, 15240-062) at 4°C. Isolation of endometrial epithelial cells and establishment of organoids was performed using previously described methods (Fitzgerald et al., 2019). To collect stromal cells, filtrate from the 100-micron cell strainer was sequentially passed through 40-micron and 10-micron cell strainers. The resulting filtrate was pelleted by centrifugation (3000×g for 5 min at 4°C), resuspended in DMEM/F-12 supplemented with 1% Anti-Anti and 10% charcoal stripped FBS (CSFBS), placed in a flask, and incubated at 37°C. Sequential seeding was performed to ensure the purity of stromal cells in this first flask; the medium from the first flask and any unattached cells were then transferred to a second flask, and medium replaced in the first flask.
Hormone treatment of organoids
To examine the steroid hormone response, organoids were passaged, resuspended in Matrigel, and plated in 12-well plates (three droplets per well). Organoids were established for 4 days in expansion medium (ExM) (Supplementary Table SI). Organoids were then treated with either vehicle as a control (100% ethanol) or 10 nM estradiol (E2, Sigma, E1024) for 2 days in ExM without N-2 supplement (Fig. 1A). Next, organoids were treated with either vehicle control, 10 nM E2, 10 nM E2 and 1 µM medroxyprogesterone acetate (MPA, Sigma, PHR1589) (E2 + MPA), 1 µM MPA, or 1 µM MPA and 1 µM prostaglandin E2 (PGE2, Sigma, P0409) in base medium alone (Supplementary Table SI details base media content) for 6 days with media changed every 2 days. Each treatment was performed in triplicate wells, and organoids derived from three individual donors were used. MPA, a synthetic progestin and progesterone receptor (PGR) agonist, and PGE2 were utilized based on their requirement for, or enhancement of, human stromal cell decidualization in vitro (Frank et al., 1994; Stadtmauer and Wagner, 2022). The rationale for conducting hormone treatments in base media starting on Day 6 is to avoid compromising determination of the basolateral secretome, because the ExM contains proteins [epidermal growth factor (EGF), Noggin, Respondin-1, fibroblast growth factor (FGF)10, hepatocyte growth factor (HGF)] that would interfere with protein analysis by mass spectrometry.
Effects of steroid hormone and PGE2 treatment on human endometrial epithelial organoids. (A) Organoids were seeded in Matrigel and grown for 4 days in expansion media (ExM), followed by treatment with vehicle as a control or estrogen (E2) for 2 days. Next, the ExM was replaced with base media, and organoids were treated for 6 days with either vehicle as a control, E2, E2 + medroxyprogesterone acetate (MPA), MPA alone, or MPA + prostaglandin E2 (PGE2) for 6 days. (B) Gross morphology of organoids on Days 1 and 12. Scale bar = 1000 µm. (C) Immunofluorescent localization of forkhead box A2 (FOXA2), acetylated tubulin (AcTUB), and ectonucleotide pyrophosphatase/phosphodiesterase 3 (ENPP3) proteins in Day 12 organoids. Nuclei were visualized with Hoechst stain. The white arrows denote AcTUB-positive cilia in organoid cells. Scale bar = 75 µm.
Immunofluorescence analyses
Medium was removed and organoids were fixed using zinc formalin for 15 min. Organoids were washed with PBS, embedded in Tissue-Tek O.C.T. Compound (Sakura Finetek USA, Torrance, CA, USA) frozen at −80°C, and sectioned at 7 µm. Sections were air dried for 10 min at 37°C then post-fixed in zinc formalin for 5 min and washed in PBS. Permeabilization was performed with 0.1% Tween 20 (PBST) for 20 min at room temperature and washed using PBS. Antigen retrieval was performed by incubating organoid sections in boiling Reveal Decloaker (RV1000M, Biocare Medica, Pacheco, CA, USA) for 30 min. After cooling for 45 min, organoid sections were briefly washed in Milli-Q H2O, followed by PBST for 20 min and two washes of PBS (10 min each). Sections were blocked with 2% normal goat serum (ThermoFisher Scientific, 01-6201) in PBS for 30 min at room temperature, followed by overnight incubation with primary antibody at 4°C (Supplementary Table SII). Sections were washed twice with PBS (10 min each) and incubated with either Alexa Fluor 488-conjugated secondary antibody (112-545-143, Jackson ImmunoResearch, West Grove, PA, USA) or Alexa Fluor 594-conjugated secondary antibody (111-585-144, Jackson ImmunoResearch) in 2% normal goat serum in PBS for 1 h at room temperature. Sections were then incubated with Hoechst (H3570, Invitrogen) to visualize nuclei and imaged with a Leica DM5500 B upright microscope using the Leica Application Suite X (LAS X).
Real-time quantitative PCR analysis
Total RNA was extracted from stromal cells using a standard TRIzol-based protocol and Direct-zol RNA MiniPrep Plus isolation kit (R2070, Zymo Research, Irvine, CA, USA). To eliminate genomic DNA contamination, isolated RNA was treated with DNase I (Rnase-Free Dnase Set, 79256, Qiagen, Germantown, MD, USA). Total RNA (500 ng) from each sample was reverse transcribed using iScript RT Supermix (1708841, Bio-Rad, Hercules, CA, USA). Real-time PCR was performed using SsoAdvanced Universal SYBR Green Supermix and either Bio-Rad PrimePCR primers or Integrated DNA Technologies primers (Supplementary Table SIII). HPRT1 (hypoxanthine phosphoribosyltransferase 1) and RPLP0 (ribosomal protein lateral stalk subunit P0) were used as housekeeping genes for normalization of data. A one-way ANOVA was used to compare mRNA levels between treatment groups with a P-value of ≤0.05 considered significant.
RNA-seq transcriptome analysis
Quality and concentration of RNA extracted from hormone-treated organoids from each donor were determined using a Fragment Analyzer (Advanced Analytical Technologies, Ankeny, IA, USA). Libraries were prepared by the University of Missouri Genomics Technology Core Facility using an Illumina TruSeq mRNA kit (Illumina, San Diego, CA, USA) and sequenced (2 × 100 bp paired-end to a depth of 40–50 million reads) using an Illumina NextSeq 500. FastQC (v 0.11.9) and MultiQC (v 1.10.1) were used to evaluate the quality of the sequence. Trimmomatic (v 1.10.1) was used to remove adapters and low-quality bases from the reads. Trimmed reads were aligned to the human genome (GRCh38) with STAR aligner (v 2.7.9a) (average unique mapping rate of 89.21% and read length of 200 bp). Gene quantification was calculated by using featureCounts from Subread package (v 2.0.3). Counts extracted from featureCounts were used to perform differential gene analysis in R (version 4.0.2) using DESeq2 (v 1.30.1). ClusterProfiler software (version 4.0.5) was used to perform Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) Pathway analysis on differentially expressed genes (DEGs) with absolute log2 fold change >1 and adjusted P-value ≤ 0.05. Design formula (donor – condition) was used as a factor to account for variation in the data. Increased and decreased genes were each subjected to a separate KEGG pathway analysis. Transcriptome data are available in the Gene Expression Omnibus (GSE205084).
Mass spectrometry analysis of basolateral secretion of organoid proteins
Mass spectrometry analysis was conducted by the Charles W. Gehrke Proteomics Center at the University of Missouri. To analyze the secreted proteome, media were collected from four wells per treatment group and pooled (three donors per treatment). Samples were mixed with four volumes of 5% trichloroacetic acid in acetone and incubated at −20°C overnight. Following centrifugation (20 000×g for 10 min at 4°C), samples were washed twice with 80% acetone and pellets were resuspended in 6 M urea, 2 M thiourea, and 100 mM ammonium bicarbonate. Protein was quantified using the Pierce 660 nM Protein Assay method (Pierce Chemical, Dallas, TX, USA). Thirty micrograms of protein from each sample was reduced and alkylated, digested with lysC at an enzyme to protein ratio of 1:70 and incubated for 3 h at 37°C. Trypsin was added to the samples at a trypsin to protein w/w ratio of 1:50 and the samples were digested overnight at 37°C. Digested peptides were purified using Pierce C18 tips. To generate a spectra library, 5 µg purified peptides from each sample were combined, lyophilized, and resuspended in 5% acetonitrile, 0.1% formic acid. Samples were acquired on the Bruker timsTOFPRO with the DIA acquisition method using the 30 min liquid chromatography gradient. The spectral library was built by using data-dependent acquisition data by Pulsar in Spectronaut, filtered with a PSM/peptide/protein false discovery rate (FDR) of 1%. The Uniprot-Human database was used (20 381 items). Data were searched with tryspin as enzyme, two missed cleavages allowed; carbamidomethyl cysteine as a fixed modification; acetylation on protein N-terminus, oxidation (M) as variable modifications. Data independent acquisition (DIA) data were searched using Spectronaut with the spectra library generated by Pulsar. The precursor/protein Q value cutoff was 0.01. Quantitation type of MS2 area was chosen and data filtering was set to Q value sparse. The DIA data were filtered with precursor FDR and protein group at 1%. A protein was considered present if identified in two samples per treatment group and a protein intensity of >6 in at least one sample. One-way ANOVA with a post hoc Tukey’s honestly significant difference test was used to compare treatment groups. A P-value of ≤0.05 was considered significant. The data were deposited in the proteomics identifications database (PRIDE, www.ebi.ac.uk/pride/ submission reference 1-20220602-7305).
Culture and treatment of primary human endometrial stromal cells
Primary endometrial stromal cells were grown to 80% confluence in 10% CSFBS DMEM/F-12 and then treated with either vehicle as a control or 10 nM E2 and 1 µM MPA (E2 + MPA) in 0.5% CSFBS DMEM/F-12 for 6 days. Next, stromal cells were treated with either vehicle control, E2 and MPA, E2 + MPA + 0.5 mM cAMP (2′-O-dibutyryladenosine 3′, 5′-cyclic monophosphate sodium salt; Sigma, D0627), E2 + MPA + 1 µM PGE2 (Prostaglandin E2, Sigma, P0409), and E2 + MPA with either cystatin-C (CST3, 1196-PI-010, R&D Systems, Minneapolis, MN, USA), stanniocalcin-1 (STC1, R&D Systems, 9400-SO-050), or alpha-1-antichymotrypsin (SERPINA3, R&D Systems, 1295-PI-010) at 10, 100, and 1000 ng/ml for an additional 4 days (three wells/treatment). To determine whether CST3 influences PGE2-induced stromal cell decidualization, primary endometrial stromal cells were treated with either vehicle control or E2 + MPA in 0.5% CSFBS DMEM/F-12 for 6 days. Stromal cells were then treated with either vehicle control, E2 + MPA, E2 + MPA + PGE2, or E2 + MPA + PGE2 and CST3 at 10, 100, and 1000 ng/ml for an additional 4 days (three wells/treatment).
Results
Development of human organoids in base media without WNT-activating factors
During the 12-day experiment (Fig. 1A), all human endometrial epithelial organoids grew substantially larger in diameter, with no obvious differences in growth or morphology between the treatment groups (Fig. 1B). Forkhead box A2 (FOXA2) is a transcription factor that is expressed specifically in the GE of the endometrium of the human uterus (Kelleher et al., 2019a). The immunofluorescence signal intensity for FOXA2 was similar in the nuclei of most cells in the organoid regardless of hormone treatment (Fig. 1C). The organoids contained a mixed population of GE cells, including non-ciliated and ciliated cells based on immunofluorescence analysis of acetylated tubulin (AcTUB), a marker of ciliated cells (Fig. 1C). No differences in growth, FOXA2 or ciliated cells were observed among organoids from the three donors (data not shown).
Steroid hormones and PGE2 alter the organoid transcriptome
Bulk RNA-seq analysis of hormone-treated organoids derived from three donors revealed significant changes in their transcriptomes in response to hormone treatment (Fig. 2A, Supplementary Data SI). Changes in mRNA abundance between control and hormone treatment groups were assessed using several parameters (log2 fold change ≥ 1.0, P ≤ 0.05, and fragments per kilobase of exon per million mapped fragments ≥ 1.0 in one group). The effects of hormone treatment and PGE2 on the organoid transcriptome are summarized in Fig. 2.
Transcriptome analysis of hormone- and PGE2-treated human endometrial epithelial organoids. Organoids were seeded in Matrigel and grown for 4 days in expansion media (ExM), followed by treatment with vehicle as a control or estrogen (E2) for 2 days. Next, the ExM was replaced with base media, and organoids were treated for 6 days with either vehicle as a control, E2, E2 + medroxyprogesterone acetate (MPA), MPA alone, or MPA + prostaglandin E2 (PGE2) for 6 days. On Day 12, RNA was extracted, and the transcriptome determined by RNA sequencing. (A) Volcano plots of differentially expressed genes based on organoid response to hormone treatments. (B) Venn diagrams comparing genes that were differentially increased or decreased in all treatments compared to control. (C) Venn diagrams comparing genes that were differentially increased or decreased in E2 and E2 + MPA treatments compared to control. (D) Venn diagrams comparing genes that were differentially increased or decreased in MPA and MPA + PGE2 treatments compared to control.
COL1A2 (collagen type I alpha 2 Chain) showed the greatest increase in mRNA level, and PGR was among the top 10 genes impacted by E2. Treatment with MPA alone or in combination with E2 or PGE2 altered expression of the greatest number of genes compared to the control (Fig. 2A and B). ENPP3 (ectonucleotide pyrophosphatase/phosphodiesterase family member 3) and LEFTY1 (left-right determination factor 1) were among the top 10 genes stimulated by MPA compared to control. Treatment with MPA + PGE2 compared to the control or MPA alone revealed several PGE2-regulated genes, particularly an increase in metallothionein genes (MT1G, MT1H, MT1M). MMP10 (matrix meallopeptidase 10) showed the greatest decrease in expression in E2 + MPA and MPA treatments compared to control.
GO enrichment analysis of DEGs in hormone-treated organoids found enrichment (P < 0.05) for specific biological processes based on hormone treatment (Supplementary Fig. S1). Biological processes enriched by E2 treatment of organoids included nucleus localization, and cilia assembly and movement. Genes decreased by E2 treatment were enriched for biological processes related to immune and stress responses. Genes increased by E2 + MPA treatment compared to control were also enriched for cilium assembly and movement as well as extracellular transport. Indeed, an increase in ciliated cells in the organoids was observed with E2 and MPA treatment (Fig. 1C). Processes decreased by E2 + MPA compared to control included angiogenesis and tube morphogenesis. Biological processes enriched in genes increased by MPA + PGE2 treatment included responses to and regulation of metal ions copper and zinc. Genes decreased by MPA + PGE2 were enriched for cell motility, migration, and locomotion. No biological processes were identified as enriched by treatment with MPA alone.
ENPP3 was one of the top five genes increased by MPA, E2 + MPA, and MPA + PGE2, indicating P4 responsivity (Fig. 2A–C). Immunostaining revealed that ENPP3 was substantially increased in the organoids by treatment with MPA, MPA + E2, or MPA + PGE2 as compared to E2 alone or the control (Fig. 1C). ENPP3 was observed at the apical surface of the organoid cells, which is reflective of the in vivo secretion pattern of ENPP3 and presence in the uterine lumen (Boggavarapu et al., 2016; Chen et al., 2018).
Identification of proteins secreted by the organoids
The secreted proteome of organoids treated with hormones and PGE2 was determined by mass spectrometry. This analysis identified ∼124 proteins in the basolateral organoid secretome (Supplementary Data SII). Biological processes enriched for all proteins identified in the organoid secretome included extracellular matrix organization, regulation of peptidase activity, wound healing, secretion, and immune response (Fig. 3A).
Mass spectrometry analysis of proteins secreted basolaterally by hormone- and PGE2-treated human endometrial epithelial organoids. Organoids were seeded in Matrigel and grown for 4 days in expansion media, followed by treatment with vehicle as a control or estrogen (E2) for 2 days. Next, the ExM was replaced with base media, and organoids were treated for 6 days with either vehicle as a control, E2, E2 + medroxyprogesterone acetate (MPA), MPA alone, or MPA + prostaglandin E2 (PGE2) for 6 days. On Day 12, proteins in the base media surrounding the organoid were determined by mass spectrometry analysis. (A) ShinyGO visualization of enriched biological processes for genes identified in organoid conditioned media. (B) Integration of proteins that were increased or decreased for all treatments compared to control. (C) REVIGO visualization of enriched biological processes for proteins increased or decreased by E2 and E2 + MPA treatment compared to control.
Changes in secreted protein abundance between control and hormone treatment groups were assessed, according to statistical differences (P < 0.05) and fold change (≥1.5-fold) between the control and treatment groups (Fig. 3B, Tables I, II, III, and IV, Supplementary Data SII). Compared to the control, E2 alone increased 22 proteins including SERPINA3, COL1A2, and CST3, and decreased 43 proteins including ADAMTS1, PROM1, and MMP10. Treatment with E2 + MPA increased 23 proteins including COL1A2 and SERPINA3, and decreased 39 proteins including IGFBP3, ADAMTS1, and MMP10. MPA alone increased 18 proteins including CST3, COL1A2, and SERPINA3, and decreased 33 proteins including APOA1, ADAMTS1, and MMP10. Finally, treatment with MPA + PGE2 increased the abundance of 19 proteins including CST3, SERPINA3, and B2M, and decreased 71 proteins including ADAMTS1, APOA2, and FN1. Of note, five proteins were uniquely increased by E2 + MPA treatment, while all other increased proteins were common amongst two or more treatment groups (Fig. 3B). Moreover, 27 proteins were uniquely decreased by MPA + PGE2 treatment, while all other decreased proteins were common among two or more treatment groups (Fig. 3B). Integration of the organoid transcriptome and basolateral secreted proteome data is presented in Supplementary Fig. S2. Note that relatively few genes and encoded proteins were coordinately increased or decreased by hormone treatment. One caveat of the present study is that the intraorganoid, apical secretome (i.e. proteins present within the lumen of the organoid itself) was not determined. Indeed, substantial differences exist in the apical (intraorganoid) as compared to basolateral (extraorganoid) secretome (Simintiras et al., 2021).
Top 10 proteins in the human organoid secretome responding to E2 treatment.a
| Increasedb | Decreasedb | ||||
|---|---|---|---|---|---|
| Symbol | Name | Fold change | Symbol | Name | Fold change |
| SERPINA3 | Alpha-1-antichymotrypsin | 9.1 | MMP10 | Stromelysin-2 | −6.5 |
| COL1A2 | Collagen alpha-2(I) chain | 5.5 | ADAMTS1 | A disintegrin and metalloproteinase with thrombospondin motifs 1 | −4.6 |
| B2M | Beta-2-microglobulin | 4.9 | PROM1 | Prominin-1 | −3.6 |
| TFPI2 | Tissue factor pathway inhibitor 2 | 4.0 | CLIC1 | Chloride intracellular channel protein 1 | −3.3 |
| CST3 | Cystatin-C | 3.3 | APOA1 | Apolipoprotein A-I | −2.8 |
| ALPP | Alkaline phosphatase, placental type | 2.5 | SERPINF2 | Alpha-2-antiplasmin | −2.8 |
| RBP4 | Retinol-binding protein 4 | 2.3 | IGFBP3 | Insulin-like growth factor-binding protein 3 | −2.5 |
| SERPINA5 | Plasma serine protease inhibitor | 2.1 | HABP2 | Hyaluronan-binding protein 2 | −2.4 |
| LEFTY2 | Left-right determination factor 2 | 2.1 | CPXM2 | Inactive carboxypeptidase-like protein X2 | −2.3 |
| SEMA3B | Semaphorin-3B | 2.1 | CFB | Complement factor B | −2.3 |
| Increasedb | Decreasedb | ||||
|---|---|---|---|---|---|
| Symbol | Name | Fold change | Symbol | Name | Fold change |
| SERPINA3 | Alpha-1-antichymotrypsin | 9.1 | MMP10 | Stromelysin-2 | −6.5 |
| COL1A2 | Collagen alpha-2(I) chain | 5.5 | ADAMTS1 | A disintegrin and metalloproteinase with thrombospondin motifs 1 | −4.6 |
| B2M | Beta-2-microglobulin | 4.9 | PROM1 | Prominin-1 | −3.6 |
| TFPI2 | Tissue factor pathway inhibitor 2 | 4.0 | CLIC1 | Chloride intracellular channel protein 1 | −3.3 |
| CST3 | Cystatin-C | 3.3 | APOA1 | Apolipoprotein A-I | −2.8 |
| ALPP | Alkaline phosphatase, placental type | 2.5 | SERPINF2 | Alpha-2-antiplasmin | −2.8 |
| RBP4 | Retinol-binding protein 4 | 2.3 | IGFBP3 | Insulin-like growth factor-binding protein 3 | −2.5 |
| SERPINA5 | Plasma serine protease inhibitor | 2.1 | HABP2 | Hyaluronan-binding protein 2 | −2.4 |
| LEFTY2 | Left-right determination factor 2 | 2.1 | CPXM2 | Inactive carboxypeptidase-like protein X2 | −2.3 |
| SEMA3B | Semaphorin-3B | 2.1 | CFB | Complement factor B | −2.3 |
E2 alone versus control.
Differences (P ≤ 0.05) between the estrogen (E2) alone and control treatment groups was determined by one-way ANOVA with a post hoc Tukey’s honestly significant difference test.
Top 10 proteins in the human organoid secretome responding to E2 treatment.a
| Increasedb | Decreasedb | ||||
|---|---|---|---|---|---|
| Symbol | Name | Fold change | Symbol | Name | Fold change |
| SERPINA3 | Alpha-1-antichymotrypsin | 9.1 | MMP10 | Stromelysin-2 | −6.5 |
| COL1A2 | Collagen alpha-2(I) chain | 5.5 | ADAMTS1 | A disintegrin and metalloproteinase with thrombospondin motifs 1 | −4.6 |
| B2M | Beta-2-microglobulin | 4.9 | PROM1 | Prominin-1 | −3.6 |
| TFPI2 | Tissue factor pathway inhibitor 2 | 4.0 | CLIC1 | Chloride intracellular channel protein 1 | −3.3 |
| CST3 | Cystatin-C | 3.3 | APOA1 | Apolipoprotein A-I | −2.8 |
| ALPP | Alkaline phosphatase, placental type | 2.5 | SERPINF2 | Alpha-2-antiplasmin | −2.8 |
| RBP4 | Retinol-binding protein 4 | 2.3 | IGFBP3 | Insulin-like growth factor-binding protein 3 | −2.5 |
| SERPINA5 | Plasma serine protease inhibitor | 2.1 | HABP2 | Hyaluronan-binding protein 2 | −2.4 |
| LEFTY2 | Left-right determination factor 2 | 2.1 | CPXM2 | Inactive carboxypeptidase-like protein X2 | −2.3 |
| SEMA3B | Semaphorin-3B | 2.1 | CFB | Complement factor B | −2.3 |
| Increasedb | Decreasedb | ||||
|---|---|---|---|---|---|
| Symbol | Name | Fold change | Symbol | Name | Fold change |
| SERPINA3 | Alpha-1-antichymotrypsin | 9.1 | MMP10 | Stromelysin-2 | −6.5 |
| COL1A2 | Collagen alpha-2(I) chain | 5.5 | ADAMTS1 | A disintegrin and metalloproteinase with thrombospondin motifs 1 | −4.6 |
| B2M | Beta-2-microglobulin | 4.9 | PROM1 | Prominin-1 | −3.6 |
| TFPI2 | Tissue factor pathway inhibitor 2 | 4.0 | CLIC1 | Chloride intracellular channel protein 1 | −3.3 |
| CST3 | Cystatin-C | 3.3 | APOA1 | Apolipoprotein A-I | −2.8 |
| ALPP | Alkaline phosphatase, placental type | 2.5 | SERPINF2 | Alpha-2-antiplasmin | −2.8 |
| RBP4 | Retinol-binding protein 4 | 2.3 | IGFBP3 | Insulin-like growth factor-binding protein 3 | −2.5 |
| SERPINA5 | Plasma serine protease inhibitor | 2.1 | HABP2 | Hyaluronan-binding protein 2 | −2.4 |
| LEFTY2 | Left-right determination factor 2 | 2.1 | CPXM2 | Inactive carboxypeptidase-like protein X2 | −2.3 |
| SEMA3B | Semaphorin-3B | 2.1 | CFB | Complement factor B | −2.3 |
E2 alone versus control.
Differences (P ≤ 0.05) between the estrogen (E2) alone and control treatment groups was determined by one-way ANOVA with a post hoc Tukey’s honestly significant difference test.
Top 10 proteins in the human organoid secretome responding to treatment with E2+MPA.a
| Increasedb | Decreasedb | ||||
|---|---|---|---|---|---|
| Symbol | Name | Fold change | Symbol | Name | Fold change |
| SERPINA3 | Alpha-1-antichymotrypsin | 17.2 | MMP10 | Stromelysin-2 | −14.8 |
| COL1A2 | Collagen alpha-2(I) chain | 10.4 | ADAMTS1 | A disintegrin and metalloproteinase with thrombospondin motifs 1 | −6.7 |
| TFPI2 | Tissue factor pathway inhibitor 2 | 5.6 | IGFBP3 | Insulin-like growth factor-binding protein 3 | −4.5 |
| B2M | Beta-2-microglobulin | 3.7 | PROM1 | Prominin-1 | −4.2 |
| LEFTY2 | Left-right determination factor 2 | 3.4 | SERPINF2 | Alpha-2-antiplasmin | −4.1 |
| SERPINA5 | Plasma serine protease inhibitor | 3.2 | APOA1 | Apolipoprotein A-I | −3.7 |
| MIF | Macrophage migration inhibitory factor | 2.7 | HABP2 | Hyaluronan-binding protein 2 | −2.9 |
| ALPP | Alkaline phosphatase, placental type | 2.6 | CPXM2 | Inactive carboxypeptidase-like protein X2 | −2.7 |
| CST3 | Cystatin-C | 2.5 | MUC5B | Mucin-5B | −2.5 |
| SEMA3B | Semaphorin-3B | 2.4 | CFB | Complement factor B | −2.4 |
| Increasedb | Decreasedb | ||||
|---|---|---|---|---|---|
| Symbol | Name | Fold change | Symbol | Name | Fold change |
| SERPINA3 | Alpha-1-antichymotrypsin | 17.2 | MMP10 | Stromelysin-2 | −14.8 |
| COL1A2 | Collagen alpha-2(I) chain | 10.4 | ADAMTS1 | A disintegrin and metalloproteinase with thrombospondin motifs 1 | −6.7 |
| TFPI2 | Tissue factor pathway inhibitor 2 | 5.6 | IGFBP3 | Insulin-like growth factor-binding protein 3 | −4.5 |
| B2M | Beta-2-microglobulin | 3.7 | PROM1 | Prominin-1 | −4.2 |
| LEFTY2 | Left-right determination factor 2 | 3.4 | SERPINF2 | Alpha-2-antiplasmin | −4.1 |
| SERPINA5 | Plasma serine protease inhibitor | 3.2 | APOA1 | Apolipoprotein A-I | −3.7 |
| MIF | Macrophage migration inhibitory factor | 2.7 | HABP2 | Hyaluronan-binding protein 2 | −2.9 |
| ALPP | Alkaline phosphatase, placental type | 2.6 | CPXM2 | Inactive carboxypeptidase-like protein X2 | −2.7 |
| CST3 | Cystatin-C | 2.5 | MUC5B | Mucin-5B | −2.5 |
| SEMA3B | Semaphorin-3B | 2.4 | CFB | Complement factor B | −2.4 |
E2 + MPA (estrogen + medroxyprogesterone acetate) alone versus control.
Differences (P ≤ 0.05) between the E2 + MPA and control treatment groups was determined by one-way ANOVA with a post hoc Tukey’s honestly significant difference test.
Top 10 proteins in the human organoid secretome responding to treatment with E2+MPA.a
| Increasedb | Decreasedb | ||||
|---|---|---|---|---|---|
| Symbol | Name | Fold change | Symbol | Name | Fold change |
| SERPINA3 | Alpha-1-antichymotrypsin | 17.2 | MMP10 | Stromelysin-2 | −14.8 |
| COL1A2 | Collagen alpha-2(I) chain | 10.4 | ADAMTS1 | A disintegrin and metalloproteinase with thrombospondin motifs 1 | −6.7 |
| TFPI2 | Tissue factor pathway inhibitor 2 | 5.6 | IGFBP3 | Insulin-like growth factor-binding protein 3 | −4.5 |
| B2M | Beta-2-microglobulin | 3.7 | PROM1 | Prominin-1 | −4.2 |
| LEFTY2 | Left-right determination factor 2 | 3.4 | SERPINF2 | Alpha-2-antiplasmin | −4.1 |
| SERPINA5 | Plasma serine protease inhibitor | 3.2 | APOA1 | Apolipoprotein A-I | −3.7 |
| MIF | Macrophage migration inhibitory factor | 2.7 | HABP2 | Hyaluronan-binding protein 2 | −2.9 |
| ALPP | Alkaline phosphatase, placental type | 2.6 | CPXM2 | Inactive carboxypeptidase-like protein X2 | −2.7 |
| CST3 | Cystatin-C | 2.5 | MUC5B | Mucin-5B | −2.5 |
| SEMA3B | Semaphorin-3B | 2.4 | CFB | Complement factor B | −2.4 |
| Increasedb | Decreasedb | ||||
|---|---|---|---|---|---|
| Symbol | Name | Fold change | Symbol | Name | Fold change |
| SERPINA3 | Alpha-1-antichymotrypsin | 17.2 | MMP10 | Stromelysin-2 | −14.8 |
| COL1A2 | Collagen alpha-2(I) chain | 10.4 | ADAMTS1 | A disintegrin and metalloproteinase with thrombospondin motifs 1 | −6.7 |
| TFPI2 | Tissue factor pathway inhibitor 2 | 5.6 | IGFBP3 | Insulin-like growth factor-binding protein 3 | −4.5 |
| B2M | Beta-2-microglobulin | 3.7 | PROM1 | Prominin-1 | −4.2 |
| LEFTY2 | Left-right determination factor 2 | 3.4 | SERPINF2 | Alpha-2-antiplasmin | −4.1 |
| SERPINA5 | Plasma serine protease inhibitor | 3.2 | APOA1 | Apolipoprotein A-I | −3.7 |
| MIF | Macrophage migration inhibitory factor | 2.7 | HABP2 | Hyaluronan-binding protein 2 | −2.9 |
| ALPP | Alkaline phosphatase, placental type | 2.6 | CPXM2 | Inactive carboxypeptidase-like protein X2 | −2.7 |
| CST3 | Cystatin-C | 2.5 | MUC5B | Mucin-5B | −2.5 |
| SEMA3B | Semaphorin-3B | 2.4 | CFB | Complement factor B | −2.4 |
E2 + MPA (estrogen + medroxyprogesterone acetate) alone versus control.
Differences (P ≤ 0.05) between the E2 + MPA and control treatment groups was determined by one-way ANOVA with a post hoc Tukey’s honestly significant difference test.
Top 10 proteins in the human organoid secretome responding to treatment with MPA alone.a
| Increasedb | Decreasedb | ||||
|---|---|---|---|---|---|
| Symbol | Name | Fold change | Symbol | Name | Fold change |
| SERPINA3 | Alpha-1-antichymotrypsin | 15.7 | MMP10 | Stromelysin-2 | −5.7 |
| COL1A2 | Collagen alpha-2(I) chain | 6.7 | ADAMTS1 | A disintegrin and metalloproteinase with thrombospondin motifs 1 | −4.3 |
| B2M | Beta-2-microglobulin | 4.5 | APOA1 | Apolipoprotein A-I | −3.6 |
| ALPP | Alkaline phosphatase, placental type | 4.2 | PROM1 | Prominin-1 | −2.8 |
| TFPI2 | Tissue factor pathway inhibitor 2 | 3.9 | FN1 | Fibronectin | −2.6 |
| SERPINA5 | Plasma serine protease inhibitor | 3.3 | CLIC1 | Chloride intracellular channel protein 1 | −2.4 |
| CST3 | Cystatin-C | 3.1 | CPXM2 | Inactive carboxypeptidase-like protein X2 | −2.3 |
| LEFTY2 | Left-right determination factor 2 | 2.8 | TIMP1 | Metalloproteinase inhibitor 1 | −2.2 |
| LTF | Lactotransferrin | 2.5 | IGFBP3 | Insulin-like growth factor-binding protein 3 | −2.2 |
| RBP4 | Retinol-binding protein 4 | 2.1 | NPC2 | NPC intracellular cholesterol transporter 2 | −2.2 |
| Increasedb | Decreasedb | ||||
|---|---|---|---|---|---|
| Symbol | Name | Fold change | Symbol | Name | Fold change |
| SERPINA3 | Alpha-1-antichymotrypsin | 15.7 | MMP10 | Stromelysin-2 | −5.7 |
| COL1A2 | Collagen alpha-2(I) chain | 6.7 | ADAMTS1 | A disintegrin and metalloproteinase with thrombospondin motifs 1 | −4.3 |
| B2M | Beta-2-microglobulin | 4.5 | APOA1 | Apolipoprotein A-I | −3.6 |
| ALPP | Alkaline phosphatase, placental type | 4.2 | PROM1 | Prominin-1 | −2.8 |
| TFPI2 | Tissue factor pathway inhibitor 2 | 3.9 | FN1 | Fibronectin | −2.6 |
| SERPINA5 | Plasma serine protease inhibitor | 3.3 | CLIC1 | Chloride intracellular channel protein 1 | −2.4 |
| CST3 | Cystatin-C | 3.1 | CPXM2 | Inactive carboxypeptidase-like protein X2 | −2.3 |
| LEFTY2 | Left-right determination factor 2 | 2.8 | TIMP1 | Metalloproteinase inhibitor 1 | −2.2 |
| LTF | Lactotransferrin | 2.5 | IGFBP3 | Insulin-like growth factor-binding protein 3 | −2.2 |
| RBP4 | Retinol-binding protein 4 | 2.1 | NPC2 | NPC intracellular cholesterol transporter 2 | −2.2 |
MPA (medroxyprogesterone acetate) alone versus control.
Differences (P ≤ 0.05) between the MPA and control treatment groups was determined by one-way ANOVA with a post hoc Tukey’s honestly significant difference test.
Top 10 proteins in the human organoid secretome responding to treatment with MPA alone.a
| Increasedb | Decreasedb | ||||
|---|---|---|---|---|---|
| Symbol | Name | Fold change | Symbol | Name | Fold change |
| SERPINA3 | Alpha-1-antichymotrypsin | 15.7 | MMP10 | Stromelysin-2 | −5.7 |
| COL1A2 | Collagen alpha-2(I) chain | 6.7 | ADAMTS1 | A disintegrin and metalloproteinase with thrombospondin motifs 1 | −4.3 |
| B2M | Beta-2-microglobulin | 4.5 | APOA1 | Apolipoprotein A-I | −3.6 |
| ALPP | Alkaline phosphatase, placental type | 4.2 | PROM1 | Prominin-1 | −2.8 |
| TFPI2 | Tissue factor pathway inhibitor 2 | 3.9 | FN1 | Fibronectin | −2.6 |
| SERPINA5 | Plasma serine protease inhibitor | 3.3 | CLIC1 | Chloride intracellular channel protein 1 | −2.4 |
| CST3 | Cystatin-C | 3.1 | CPXM2 | Inactive carboxypeptidase-like protein X2 | −2.3 |
| LEFTY2 | Left-right determination factor 2 | 2.8 | TIMP1 | Metalloproteinase inhibitor 1 | −2.2 |
| LTF | Lactotransferrin | 2.5 | IGFBP3 | Insulin-like growth factor-binding protein 3 | −2.2 |
| RBP4 | Retinol-binding protein 4 | 2.1 | NPC2 | NPC intracellular cholesterol transporter 2 | −2.2 |
| Increasedb | Decreasedb | ||||
|---|---|---|---|---|---|
| Symbol | Name | Fold change | Symbol | Name | Fold change |
| SERPINA3 | Alpha-1-antichymotrypsin | 15.7 | MMP10 | Stromelysin-2 | −5.7 |
| COL1A2 | Collagen alpha-2(I) chain | 6.7 | ADAMTS1 | A disintegrin and metalloproteinase with thrombospondin motifs 1 | −4.3 |
| B2M | Beta-2-microglobulin | 4.5 | APOA1 | Apolipoprotein A-I | −3.6 |
| ALPP | Alkaline phosphatase, placental type | 4.2 | PROM1 | Prominin-1 | −2.8 |
| TFPI2 | Tissue factor pathway inhibitor 2 | 3.9 | FN1 | Fibronectin | −2.6 |
| SERPINA5 | Plasma serine protease inhibitor | 3.3 | CLIC1 | Chloride intracellular channel protein 1 | −2.4 |
| CST3 | Cystatin-C | 3.1 | CPXM2 | Inactive carboxypeptidase-like protein X2 | −2.3 |
| LEFTY2 | Left-right determination factor 2 | 2.8 | TIMP1 | Metalloproteinase inhibitor 1 | −2.2 |
| LTF | Lactotransferrin | 2.5 | IGFBP3 | Insulin-like growth factor-binding protein 3 | −2.2 |
| RBP4 | Retinol-binding protein 4 | 2.1 | NPC2 | NPC intracellular cholesterol transporter 2 | −2.2 |
MPA (medroxyprogesterone acetate) alone versus control.
Differences (P ≤ 0.05) between the MPA and control treatment groups was determined by one-way ANOVA with a post hoc Tukey’s honestly significant difference test.
Top 10 proteins in the human organoid secretome responding to treatment with MPA+PGE2.a
| Increasedb | Decreasedb | ||||
|---|---|---|---|---|---|
| Symbol | Name | Fold change | Symbol | Name | Fold change |
| B2M | Beta-2-microglobulin | 7.0 | KRT8 | Keratin, type II cytoskeletal 8 | −67.5 |
| SERPINA3 | Alpha-1-antichymotrypsin | 5.9 | FN1 | Fibronectin | −4.8 |
| CST3 | Cystatin-C | 5.0 | ITIH5 | Inter-alpha-trypsin inhibitor heavy chain H5 | −4.4 |
| RBP4 | Retinol-binding protein 4 | 2.8 | APOA1 | Apolipoprotein A-I | −3.8 |
| COL1A2 | Collagen alpha-2(I) chain | 2.6 | ADAMTS1 | A disintegrin and metalloproteinase with thrombospondin motifs 1 | −3.4 |
| LTF | Lactotransferrin | 2.3 | B4GALT1 | Beta-1,4-galactosyltransferase 1 | −3.3 |
| APOC3 | Apolipoprotein C-III | 2.3 | MMP10 | Stromelysin-2 | −3.3 |
| SULT1C2 | Sulfotransferase 1C2 | 2.3 | CLIC1 | Chloride intracellular channel protein 1 | −3.1 |
| SERPINA5 | Plasma serine protease inhibitor | 2.2 | PROM1 | Prominin-1 | −3.1 |
| ALPP | Alkaline phosphatase, placental type | 2.1 | NPC2 | NPC intracellular cholesterol transporter 2 | −3.0 |
| Increasedb | Decreasedb | ||||
|---|---|---|---|---|---|
| Symbol | Name | Fold change | Symbol | Name | Fold change |
| B2M | Beta-2-microglobulin | 7.0 | KRT8 | Keratin, type II cytoskeletal 8 | −67.5 |
| SERPINA3 | Alpha-1-antichymotrypsin | 5.9 | FN1 | Fibronectin | −4.8 |
| CST3 | Cystatin-C | 5.0 | ITIH5 | Inter-alpha-trypsin inhibitor heavy chain H5 | −4.4 |
| RBP4 | Retinol-binding protein 4 | 2.8 | APOA1 | Apolipoprotein A-I | −3.8 |
| COL1A2 | Collagen alpha-2(I) chain | 2.6 | ADAMTS1 | A disintegrin and metalloproteinase with thrombospondin motifs 1 | −3.4 |
| LTF | Lactotransferrin | 2.3 | B4GALT1 | Beta-1,4-galactosyltransferase 1 | −3.3 |
| APOC3 | Apolipoprotein C-III | 2.3 | MMP10 | Stromelysin-2 | −3.3 |
| SULT1C2 | Sulfotransferase 1C2 | 2.3 | CLIC1 | Chloride intracellular channel protein 1 | −3.1 |
| SERPINA5 | Plasma serine protease inhibitor | 2.2 | PROM1 | Prominin-1 | −3.1 |
| ALPP | Alkaline phosphatase, placental type | 2.1 | NPC2 | NPC intracellular cholesterol transporter 2 | −3.0 |
MPA + PGE2 (medroxyprogesterone acetate + prostaglandin E2) versus control.
Differences (P ≤ 0.05) between the MPA+PGE2 and control treatment groups was determined by one-way ANOVA with a post hoc Tukey’s honestly significant difference test.
Top 10 proteins in the human organoid secretome responding to treatment with MPA+PGE2.a
| Increasedb | Decreasedb | ||||
|---|---|---|---|---|---|
| Symbol | Name | Fold change | Symbol | Name | Fold change |
| B2M | Beta-2-microglobulin | 7.0 | KRT8 | Keratin, type II cytoskeletal 8 | −67.5 |
| SERPINA3 | Alpha-1-antichymotrypsin | 5.9 | FN1 | Fibronectin | −4.8 |
| CST3 | Cystatin-C | 5.0 | ITIH5 | Inter-alpha-trypsin inhibitor heavy chain H5 | −4.4 |
| RBP4 | Retinol-binding protein 4 | 2.8 | APOA1 | Apolipoprotein A-I | −3.8 |
| COL1A2 | Collagen alpha-2(I) chain | 2.6 | ADAMTS1 | A disintegrin and metalloproteinase with thrombospondin motifs 1 | −3.4 |
| LTF | Lactotransferrin | 2.3 | B4GALT1 | Beta-1,4-galactosyltransferase 1 | −3.3 |
| APOC3 | Apolipoprotein C-III | 2.3 | MMP10 | Stromelysin-2 | −3.3 |
| SULT1C2 | Sulfotransferase 1C2 | 2.3 | CLIC1 | Chloride intracellular channel protein 1 | −3.1 |
| SERPINA5 | Plasma serine protease inhibitor | 2.2 | PROM1 | Prominin-1 | −3.1 |
| ALPP | Alkaline phosphatase, placental type | 2.1 | NPC2 | NPC intracellular cholesterol transporter 2 | −3.0 |
| Increasedb | Decreasedb | ||||
|---|---|---|---|---|---|
| Symbol | Name | Fold change | Symbol | Name | Fold change |
| B2M | Beta-2-microglobulin | 7.0 | KRT8 | Keratin, type II cytoskeletal 8 | −67.5 |
| SERPINA3 | Alpha-1-antichymotrypsin | 5.9 | FN1 | Fibronectin | −4.8 |
| CST3 | Cystatin-C | 5.0 | ITIH5 | Inter-alpha-trypsin inhibitor heavy chain H5 | −4.4 |
| RBP4 | Retinol-binding protein 4 | 2.8 | APOA1 | Apolipoprotein A-I | −3.8 |
| COL1A2 | Collagen alpha-2(I) chain | 2.6 | ADAMTS1 | A disintegrin and metalloproteinase with thrombospondin motifs 1 | −3.4 |
| LTF | Lactotransferrin | 2.3 | B4GALT1 | Beta-1,4-galactosyltransferase 1 | −3.3 |
| APOC3 | Apolipoprotein C-III | 2.3 | MMP10 | Stromelysin-2 | −3.3 |
| SULT1C2 | Sulfotransferase 1C2 | 2.3 | CLIC1 | Chloride intracellular channel protein 1 | −3.1 |
| SERPINA5 | Plasma serine protease inhibitor | 2.2 | PROM1 | Prominin-1 | −3.1 |
| ALPP | Alkaline phosphatase, placental type | 2.1 | NPC2 | NPC intracellular cholesterol transporter 2 | −3.0 |
MPA + PGE2 (medroxyprogesterone acetate + prostaglandin E2) versus control.
Differences (P ≤ 0.05) between the MPA+PGE2 and control treatment groups was determined by one-way ANOVA with a post hoc Tukey’s honestly significant difference test.
Several biological processes were enriched in secreted proteins that increased in response to E2 alone, including acute inflammatory response, secretion, cell proliferation, and regulation of proteolysis (Fig. 3C). Similarly, secreted proteins increased by E2 + MPA treatment were enriched for biological processes such as acute inflammatory response, regulation of hydrolase activity, neutrophil activation, and negative regulation of developmental processes. Collectively, these results support the idea that E2, MPA, and PGE2 have distinct and combinatorial effects on proteins secreted from the basolateral aspect of the organoid, with potential paracrine effects on surrounding endometrial stromal cells.
Impact of selected organoid secreted proteins on stromal cell decidualization
To determine the impact of organoid secreted proteins on stromal cell decidualization (Fig. 4A), three proteins were selected (CST3, STC1, SERPINA3) as they were increased in the media of organoids treated with MPA, MPA + E2, and MPA + PGE2 as compared to the control (Fig. 4, Tables I, II, III, and IV, Supplementary Data SII).
Effects of PGE2 and proteins secreted by human endometrial epithelial organoid on stromal cell decidualization. (A) Primary endometrial stromal cells were seeded, grown to 80% confluence in 10% charcoal stripped FBS (CSFBS) for 2 days, changed to 0.5% CSFBS media, and then treated with either vehicle as a control or estrogen and medroxyprogesterone acetate (E2 + MPA) for 6 days. Next, stromal cells were treated with vehicle as a control (CTRL), and E2 + MPA, E2 + MPA with cyclic adenosine monophosphate (E2 + MPA + cAMP), E2 + MPA with prostaglandin E2 (E2 + MPA + PGE2), or E2 + MPA with increasing amounts (10, 100, or 1000 ng/ml) of recombinant cystatin C (CST3), stanniocalcin 1 (STC1), or serpin family A member 3 (SERPINA3) for 4 days. On Day 10, RNA was harvested from cells, and mRNA levels assessed by quantitative RT-PCR. Each treatment was performed in triplicate, and the entire study contained three biological replicates. (B) Established markers of stromal cell decidualization including prolactin (PRL), insulin-like growth factor binding protein 1 (IGFBP1), interleukin 15 (IL15), and somatostatin (SMST). Solid dots show the means for the individual biological replicates. (C) Markers of active decidual cells including crystallin alpha B (CRYAB), hydroxysteroid 11-beta dehydrogenase 1 (HSD11B1), and scavenger receptor class A member 5 (SCARA5). (D) Markers of senescent decidual cells including ABI family member 3 binding protein (ABI3BP), clusterin (CLU), and iodothyronine deiodinase 2 (DIO2). Data are presented as mean ± SEM for each of the n = 3 biological replicates. One-way ANOVA was performed to compare treatment of cells with cAMP, PGE2 or recombinant proteins to E2 + MPA only. Significant effects of treatment (P ≤ 0.05) are denoted by an asterisk (*).
As illustrated in Fig. 4A, primary cultures of human endometrial stromal cells were seeded and grown to 80% confluence, treated with E2 + MPA, E2 + MPA + cAMP, or E2 + MPA + PGE2, and increasing amounts of recombinant CST3, STC1, or SERPINA3 in the presence of E2 + MPA alone (Fig. 4A). Of note, cAMP has been used for decades to induce endometrial stromal cells to decidualize in vitro, but PGE2 may be a more natural and conserved inducer of decidualization (Stadtmauer and Wagner, 2022). Treatment of stromal cells with cAMP increased (P ≤ 0.01) expression of genes that are established markers of decidualization including PRL (prolactin), IGFBP1 (insulin-like growth factor binding protein 1), IL15 (interleukin 15), and SMST (somatostatin) as compared to E2 + MPA only (Fig. 4B). Similarly, PGE2 also elicited decidualization based on increased (P < 0.01) levels of PRL, IGFBP1, IL15, and SMST mRNA compared to E2 + MPA only (Fig. 4B). In the presence of E2 + MPA, treatment of stromal cells with CST3, SERPINA3, and STC1 did not (P > 0.10) stimulate decidualization (Fig. 4B). Compared to E2 + MPA only, CST3 at the 1000 ng per ml concentration decreased (P < 0.05) PRL and IGFBP1 mRNA levels compared to E2 + MPA only but had no effect (P > 0.10) on IL15 or SMST mRNA levels (Fig. 4B).
A recent single-cell transcriptome analysis of human mid-secretory phase endometrium and in vitro endometrial stroma cultures defined genes involved in differentiation of stromal cells into active or senescent DCs (Lucas et al., 2020). CRYAB (crystallin alpha B), HSD11B1 (hydroxysteroid 11-beta dehydrogenase 1), and SCARA5 (scavenger receptor class A member 5) marked active DCs, while ABI3BP (ABI gene family member 3-binding protein), CLU (clusterin), and DIO2 (iodothyronine deiodinase 2) marked senescent DCs. In the present study, cAMP treatment induced the differentiation of endometrial stromal cells into active DCs (Fig. 4C) as well as senescent DCs (Fig. 4D) based on marker gene expression. In contrast, treatment of stromal cells with PGE2 increased (P < 0.01) expression of genes marking active DCs (Fig. 4C), but not (P > 0.10) senescent DCs (Fig. 4D). Addition of increasing amounts of CST3, SERPINA3, or STC1 in the presence of E2 + MPA only had no effects on expression of those marker genes (Fig. 4C and D).
In the next experiment, the effect of CST3 on in vitro stromal cell decidualization was tested by adding increasing amounts of CST3 in the presence of E2 + MPA with PGE2 (Fig. 5A). Inclusion of CST3 at a concentration of 1000 ng per ml decreased PGE2-induced stromal decidualization based on PRL and IGFBP1 mRNA levels (Fig. 5B).
Effects of CST3 on PGE2-induced stromal cell decidualization. (A) Primary endometrial stromal cells were seeded, grown to 80% confluence in 10% charcoal stripped FBS (CSFBS) for 2 days, changed to 0.5% CSFBS media, and then treated with either vehicle as a control or estrogen and medroxyprogesterone acetate (E2 + MPA) for 6 days. Next, stromal cells were treated with vehicle as a control (CTRL), E2 + MPA with prostaglandin E2 (E2 + MPA + PGE2), or E2 + MPA + PGE2 with increasing amounts (10, 100, or 1000 ng/ml) of recombinant cystatin C (CST3), stanniocalcin 1 (STC1), or serpin family A member 3 (SERPINA3) for 4 days. On Day 10, RNA was harvested from cells, and mRNA levels assessed by quantitative RT-PCR. Each treatment was performed in triplicate, and the entire study contained three biological replicates. (B) Established markers of stromal cell decidualization including prolactin (PRL) and insulin-like growth factor binding protein 1 (IGFBP1). Data are presented as mean ± SEM for each of the n = 3 biological replicates. One-way ANOVA was performed to compare treatment of cells with recombinant proteins to E2 + MPA + PGE2 only. Solid dots show the means for the individual biological replicates. Significant effects of treatment (P ≤ 0.05) are denoted by an asterisk (*).
Discussion
This study highlights the steroid hormone responsiveness of human endometrial epithelial organoids and the influence of those hormones on their transcriptome and secreted proteome. Culture conditions that mimic the steroidal milieu of the menstrual cycle resulted in transcriptional changes in organoids that are reflective of uterine glands in vivo, making them an attractive model for investigating P4-regulated gland products and their influence on stromal cell decidualization and early pregnancy. A comprehensive analysis of the basolateral organoid secretome revealed changes in protein abundance in response to P4. Moreover, one of the secreted proteins (CST3) impacted stromal cell decidualization by attenuating differentiation, supporting the role of gland-stromal crosstalk in regulating biological events important for the establishment of pregnancy.
The development of human organoids (Boretto et al., 2017; Turco et al., 2017) and comprehensive analysis of their cell composition and transcriptome (Fitzgerald et al., 2019) highlights the potential for this culture system as a model of uterine epithelial glands. Commonly, organoids are cultured in a defined medium, established from the study of other tissues and organs such as the intestine (Boretto et al., 2017; Turco et al., 2017). While the media components are necessary for the initial establishment of organoids and to maintain the stem progenitor-like nature of the cells during long-term culture, specific components in the ExM may inhibit the organoid from fully differentiating in response to E2 and P4. In the current study, several factors [N-2 supplement, Noggin, RSPO1 (roof plate-specific spondin-1), HGF, FGF10, EGF, N-acetyl-L-cysteine, Nicotinamide, A83-01] that were in the ExM were removed at the onset of hormone treatment to investigate hormone responsiveness of endometrial organoids. In these conditions, organoids maintained their morphology and exhibited sustained expression of the GE marker FOXA2 and upregulation of a known P4-regulated gene ENPP3. Of note, RSPO1 is a ligand for leucine-rich repeat-containing G-protein coupled receptors (LGR proteins) and positively regulates the WNT signaling pathway (Binnerts et al., 2007). Physiologically, the receptive mid-secretory phase endometrium in women is associated with decreased WNT signaling, as Dickkopf-1 that is a potent inhibitor of WNT signaling, is substantially upregulated in the stroma surrounding the endometrial glands in response to P4 during the transition from the proliferative to secretory phase of the cycle (Tulac et al., 2003, 2006). Thus, removal of WNT activating factors from the organoid culture medium before MPA treatment is likely more physiological.
In the present study, treatment of organoids with E2 and MPA induced substantial changes in the transcriptome, including both increased and decreased expression of genes. By the mid-secretory phase, the PGR is downregulated in and lost from the endometrial epithelial cells but not stromal cells (Lessey et al., 1996), and the loss of PGR is associated with the induction or upregulation of many genes in the endometrial glands. One of the most P4-responsive genes encodes SCGB1D4, a secretoglobin (SCGB) family member (Jackson et al., 2011). SCGB1D4 is stimulated by interferon-gamma and is postulated to have roles in the immune response and regulation of migration and invasion (Choi et al., 2004). OLFM4 (olfactomedin 4) was another P4-increased gene that is mainly expressed in the digestive system, prostate, and bone marrow, with aberrant expression in cancers. In endometrial adenocarcinoma, OLFM4 is positively associated with estrogen receptor 1 (ESR1) and PGR expression and correlated with tumor development (Duan et al., 2014). The relation of OLFM4 to ESR1 and regulation by E2 and EGF suggest a role for OLFM4 in tissue remodeling prior to the secretory phase (Dassen et al., 2010). Our previous study with human organoids identified ENPP3 as a P4-regulated gene (Fitzgerald et al., 2019). In the present study, ENPP3 was increased by E2 + MPA, MPA, and MPA + PGE2, but not E2 alone at the transcript and protein level. In the uterine lumen of women, ENPP3 was enriched in the mid-secretory as compared to the early secretory phase (Boggavarapu et al., 2016; Chen et al., 2018). Reflective of in vivo conditions, ENPP3 was localized apically in steroid hormone-treated organoids, suggesting secretion into the organoid lumen (Boggavarapu et al., 2016). The increase of ENPP3 during the mid-secretory phase of the menstrual cycle, its secretion into the uterine cavity, and regulation by P4 suggest a role for GE-derived ENPP3 in embryo implantation. In fact, overexpression of ENPP3 in an endometrial epithelial cell line increased rates of embryo attachment and expression of endometrial receptivity-related factors, leukemia inhibitory factor, and beta3-integrin (Chen et al., 2018). Of note, endometrial glands and their secretions have an unequivocal role in embryo growth and implantation in sheep (Spencer et al., 2019; Kelleher et al., 2019b). Future studies can leverage organoids as an in vitro model to understand how P4 acts on uterine glands to stimulate genes involved in endometrial receptivity, and the role of uterine glands in post-implantation development of the decidua and the placenta (Fitzgerald et al., 2021; Alzamil et al., 2021; Rawlings et al., 2021).
PGE2 was recently identified as a conserved factor inducing the cyclic AMP/protein kinase A (cAMP/PKA) pathway during stromal cell decidualization (Stadtmauer and Wagner, 2022). Expression of PGE synthase and synthesis of PGE2 was localized to GE and endothelial cells in both basalis and functionalis regions of the human endometrium (Milne et al., 2001). By contrast, stromal staining was predominantly localized in the functionalis layer. Given that PGE2 has a biological role in stromal cell decidualization, PGE2 could also have a role in secretory transformation of the endometrial glands. In the present study, PGE2 treatment had synergistic effects with MPA on the organoid to modulate the organoid transcriptome, and induced decidualization of stromal cells in vitro toward active DC. The effects of PGE2 are via cAMP signaling in rat endometrial stromal cells (Yee and Kennedy, 1991).
In the present study, PGE2 specifically increased expression of several metallothionein genes in the organoids, including MT1G, MT1M, MT1H, MT1X, and MT1E. A recent single-cell transcriptome study of the human endometrium found upregulation of MT genes in unciliated epithelia in the early secretory phase as compared to the proliferative phase (Wang et al., 2020). Metallothioneins are involved in zinc binding and detoxication and are upregulated in human endometrium during the window of implantation (Kao et al., 2002). They play a key role in defending against oxidative stress by scavenging free radicals and protecting against oxidative damage by heavy metals (Cai et al., 1999; Habeebu et al., 2000; Liang et al., 2015; Milnerowicz et al., 2015). Mitigation of oxidative stress during embryo implantation may be important, as oxidative stress is implicated in subfertility and infertility in humans (Agarwal et al., 2012).
The importance of uterine glands and their influence on stromal cell decidualization in animal models suggests a key reciprocal relationship between endometrial epithelia and stromal cells (Kelleher et al., 2017; 2018). While this has been inferred, a direct relationship between them has yet to be identified. Indeed, most studies have focused on proteins secreted apically, either from uterine glands or luminal epithelium, that would directly regulate embryo implantation as well as development of the placenta (Dhakal et al., 2020). In contrast, proteins secreted basolaterally have the potential to regulate the underlying stroma (Fahey et al., 2005). As such, the present study identified P4-regulated proteins secreted in a basolateral manner by the organoids and determined the influence of three proteins using an in vitro stromal cell decidualization assay.
Of the proteins selected for analysis (SERPINA3, STC1, CST3), CST3 was the only protein that impacted stromal cell decidualization. CST3 is a small secretory protein that inhibits cysteine proteases, such as cathepsins, and is present in the luminal and glandular epithelial cells of human endometrium (Abrahamson et al., 1990; Lee et al., 2018). Cathepsins CTSB and CTSL1 and their inhibitor CST3 are increased in the porcine endometrial epithelia during pregnancy and in response to P4 (Song et al., 2010). In the rat endometrium, CST3 peaks just prior to embryo implantation in the GE (Quinn et al., 2006). In the mouse, Cst3 mRNA is initially increased in the peripheral decidualizing cells and the protein localized to the primary decidual zone (Afonso et al., 1997). Of note, cathepsin B and CST3 are upregulated after the completion of decidualization on gestational day 9.5 in mice (Afonso et al., 1997). Therefore, it is possible that CST3 is involved in regulating the degree of decidual breakdown, apoptosis, and breakdown by cysteine proteinases. This idea is supported here by our findings that increasing concentrations of CST3 attenuated stromal cell differentiation into active DC. As such, CST3 may be secreted by endometrial epithelia to govern stromal cell decidualization during pregnancy establishment in women. Of note, autocrine PRL inhibits human endometrial stromal cell decidualization (Eyal et al., 2007). Thus, factors from the glands as well as the differentiated and developing DC may act in concert to locally regulate the complex process of endometrial decidualization during embryo implantation and pregnancy establishment in women. Indeed, early or late decidualization relative to embryo implantation can cause a number of issues including recurrent pregnancy loss and later pregnancy complications (Salker et al., 2010; Lucas et al., 2016; Garrido-Gomez et al., 2017; Sebastian-Leon et al., 2018). In summary, increased knowledge of how uterine glands secrete or produce factors in response to P4 is important to understand the paracrine regulation of stromal cell decidualization, embryo implantation, and placental development in women. Thus, human endometrial epithelial organoid models are useful to interrogate peri-implantation endometrial function, as this area of early pregnancy is difficult to impossible to obtain tissues for study in women.
Supplementary data
Supplementary data are available at Molecular Human Reproduction online.
Data availability
The transcriptome data underlying this article are available in the Gene Expression Omnibus (GSE205084) at https://www.ncbi.nlm.nih.gov/geo/. The proteomic data underlying this article are available in PRIDE (submission reference 1-20220602-7305) at www.ebi.ac.uk/pride/.
Acknowledgements
The authors are grateful to the Research Success Center of the Department of Obstetrics, Gynecology and Women’s Health at the University of Missouri for aid with procuring endometrial tissue. Further, they appreciate the able assistance of Hongyu Liu in transcriptome analysis, and Advanced Technology Core facilities at the University of Missouri.
Authors’ roles
H.C.F., A.M.K., D.J.S., and T.E.S. designed research; H.C.F. and C.R. performed research; H.C.F. contributed new reagents/analytic tools; H.C.F. analyzed data; and H.C.F., A.M.K., and T.E.S. wrote the paper. All authors reviewed the article.
Funding
This work was supported by the Eunice Kennedy Shriver National Institute of Child Health and Development (NIH R01 HD096266 and NIH U01 HD104482) and a Tier 2 award from the University of Missouri System Research and Creative Works Strategic Investment Program. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Conflict of interest
All authors declare no conflict of interests.
