Human Decidual RUNX1 Promotes Angiogenesis and Trophoblast Differentiation by Regulating Extracellular Vesicle Signaling

Abstract

During early pregnancy, human endometrial stromal cells differentiate into secretory decidual cells via a process regulated by ovarian steroid hormones. Decidual cells play a crucial role by secreting various factors that support essential events in forming a functional placenta, including uterine angiogenesis and the differentiation and development of trophoblasts. We previously reported that the conditional ablation of the transcription factor runt-related transcription factor 1 (RUNX1) in the mouse uterus leads to subfertility due to insufficient maternal angiogenesis and impaired trophoblast differentiation. In this study, we examined the role of RUNX1 in facilitating communication mechanisms among human decidual cells and other cell types present in the pregnant uterus. We demonstrate that RUNX1 regulates the conserved hypoxia-inducible factor 2 α-RAB27B pathway in primary human endometrial stromal cells (HESCs) during decidualization, which promotes the secretion of extracellular vesicles (EVs) by these cells. Consequently, the depletion of RUNX1 in HESC led to reduced EV secretion. Mass spectrometry identified several cargo proteins in decidual EVs, including angiopoietin-related protein 2 (ANGPTL2) and IGF2, which could regulate angiogenesis or trophoblast differentiation. We found that RUNX1 directly regulates their expression, resulting in partial changes to these cargoes when it is absent. We observed that delivering EVs lacking ANGPTL2 or IGF2 to human endothelial cells significantly decreased the formation of vascular networks compared to introducing control EVs carrying these factors. Furthermore, adding IGF2-depleted EVs to human trophoblast cells inhibited their differentiation into the extravillous trophoblast lineage. These findings collectively highlight the crucial role of decidual RUNX1 in promoting essential cell-cell interactions for angiogenesis and trophoblast differentiation during placenta formation.

The uterine endometrium plays a major role during early pregnancy. In response to ovarian steroid hormones, estrogen and progesterone, the endometrium undergoes a remarkable transformation into the secretory decidua in a process known as decidualization. By relaying secreted factors to the fetus and other nearby maternal tissues, the decidua orchestrates critical processes within the pregnant uterus that are necessary for reproductive success. These processes include embryonic implantation, where the embryo attaches to the endometrial luminal epithelium and begins to invade deeper into the decidua (1, 2), as well as maternal angiogenesis, which is essential for providing an increased blood supply to the uterus to meet the demands of the growing fetus (3, 4). Both events are crucial for the development of the placenta, the transient organ responsible for supporting fetal growth throughout gestation.

Additionally, for proper placentation, the decidua acts in a paracrine manner to regulate the differentiation of the trophoblast precursor cells into appropriate subtypes (5, 6). In humans, trophoblast stem cells can differentiate into 2 main lineages: the multinucleated syncytiotrophoblast and the invasive extravillous trophoblast (EVT). Syncytiotrophoblasts play a crucial role in the exchange of nutrients and gases between maternal and fetal blood (7), while EVTs are vital for their capability to invade deeper into maternal tissue, remodeling spiral arteries to redirect maternal blood flow toward the fetus (8).

Our understanding of the decidual factors that help coordinate the complex process of placentation is constantly evolving. Uterine-specific conditional knockout (cKO) mouse models have revealed several proteins whose expression in the decidua is crucial for proper placentation. These include activin receptor-like kinase 5 (9), basic helix-loop-helix ARNT-like protein 1 (10), nodal (11), Ras-related C3 botulinum toxin substrate 1 (12), and bone morphogenic protein receptor type 2 (13), among others. A recent report from our lab revealed that the conditional ablation of runt-related transcription factor 1 (RUNX1) in the uterus, through progesterone receptor-mediated Cre expression, also resulted in abnormal placentation and severe subfertility observed in the cKO female mice. This appeared to result from 2 main phenotypes: deficient maternal angiogenesis and impaired trophoblast differentiation and invasion (14).

Recent studies have highlighted the significance of extracellular vesicle (EV) signaling from the endometrium during pregnancy, particularly as a mechanism for initiating and supporting placentation by the maternal decidua (15). EV signaling involves the transport of molecules such as proteins or miRNAs via EVs from a donor cell to a recipient cell, triggering a functional change in the recipient. Endometrial EVs have been shown to play a significant role taken up by trophoblast cells facilitating trophoblast differentiation and invasion, which are critical components of placenta formation. Differentiation of cytotrophoblasts into invasive EVTs has been shown to be mediated by endometrial EVs (16), and increased invasive capacity of trophoblasts via upregulation of N-cadherin expression has also been reported after treatment of trophoblast cells with endometrial EVs (17). Additionally, endometrial EVs have been described to enhance the formation of an extensive uterine vasculature that provides blood flow to the placenta (16, 18).

In mice and humans, a conserved regulatory pathway for EV secretion exists in endometrial stromal cells (16, 19) Specifically, the transcription factor hypoxia-inducible factor 2 α (HIF2α) is expressed under the hypoxic conditions of early pregnancy and enhances the expression of the protein RAB27B, which is among the proteins responsible for transporting the multivesicular body from the cytosol to the cell membrane, facilitating the release of EVs into the extracellular space (16, 19). This conserved pathway allows the endometrium to coordinate events during early pregnancy, including maternal angiogenesis and trophoblast differentiation invasion.

In this study, we tested the hypothesis that endometrial stromal RUNX1 plays a similar role in coordinating maternal angiogenesis and trophoblast differentiation in a human in vitro model as it does in mice. We identified how RUNX1 coordinates these functions by establishing a relationship between RUNX1 and EV secretion, as well as protein cargo composition. Additionally, we utilized human in vitro culture systems to examine the mechanisms through which RUNX1-controlled EV secretion promotes angiogenesis and influences trophoblast differentiation.

Methods

HESC Culture and Differentiation

Deidentified primary HESCs were collected in accordance with the guidelines for the protection of human subjects participating in clinical research, as approved by the institutional review board of Wake Forest School of Medicine. Samples were acquired from fertile women aged 28 to 42 years, with a parity of 1 to 2, during the proliferative stage of the menstrual cycle via biopsy. Donors provided written informed consent and were confirmed to show no signs of endometrial pathologies. HESCs were isolated as described previously (20, 21). Cells from at least 2 different donors were utilized in most experiments.

HESCs were passaged under normoxic conditions (20% O₂) and cultured in DMEM/F-12 (Gibco #21041025) supplemented with 5% charcoal dextran-stripped fetal bovine serum (FBS; Gibco #12676029) and 50 μg/mL penicillin and streptomycin (Gibco #15140122). At ~70% confluence in a 6-well plate, HESCs were treated to induce decidualization with a differentiation cocktail consisting of 1 μM progesterone (Sigma-Aldrich # P0130), 10 nM 17-β-estradiol (Sigma-Aldrich # E1024), and 0.5 mM 8-bromo-adenosine-3′,5′-cyclic monophosphate (Sigma-Aldrich # B7880) in DMEM/F-12 medium (Gibco) supplemented with 2% exosome-depleted FBS (Gibco #A2720803) and cultured in a hypoxia incubator (3% O₂) for 72 hours before sample collection. Media and treatments were replenished every 48 hours.

Small Interfering RNA-mediated Knockdown

When primary HESCs reached ~70% confluence in a 6-well plate, they were transfected by small interfering RNA (siRNA) targeting RUNX1 (Dharmacon #J-003926-05-0002), ANGPTL2 (Dharmacon #J-007805-06-0002), or IGF2 (Dharmacon #J-004093-05-0002) gene transcripts or a scrambled siRNA control (Dharmacon #D-001810-01-05) following the manufacturer’s protocol (siLentFect; Bio-Rad). In short, a final concentration of 20 nM siRNA was mixed with siLentFect lipid reagent to transfect cells. After incubation overnight with a siRNA-lipid mixture the media was replaced with media containing differentiation cocktail as described earlier.

RNA Isolation and Quantitative PCR Analysis

After 72 hours of differentiation, RNA was extracted from HESCS using TRIzol (Invitrogen) according to the manufacturer’s protocol. AffinityScript Mulitple Temperature Reverse Transcriptase kit (Agilent) was used following the manufacturer’s instructions to convert 1 μg RNA to cDNA. Quantitative PCR analysis was performed on the cDNA using Power Sybr Green PCR master mix (Applied Biosystems) and gene-specific primers (IDT) with 36B4 being used as a housekeeping gene. A primer table is supplied in Table 1. For each treatment condition, the mean cycle threshold (Ct) was calculated from 3 replicates of each sample. ΔCt was calculated as the mean Ct of the gene of interest subtracted by the mean Ct of the housekeeping gene. ΔΔCt was calculated as the difference of ΔCt of the experimental and control groups. Fold change of gene expression relative to control was computed as 2^−ΔΔCt. The data were then displayed as mean fold induction and SEM calculated from 2 to 6 independent experiments.

Table 1.

Primer table

Primer	Strand	Experiment	Location	Sequence
Hu_36B4_F_qPCR	Forward	qPCR		GTGTTCGACAATGGCAGCAT
Hu_36B4_R_qPCR	Reverse	qPCR		GAGACCCTCCAGGAAGCGA
Hu_Runx1_F_qPCR-1	Forward	qPCR		CTGAGCCCAGGCAAGATGAGCG
Hu_Runx1_R_qPCR-1	Reverse	qPCR		CCCCTAGGGCCACCACCTTGAA
Hu_HIF2a_F_qPCR-1	Forward	qPCR		ACGAGTCCGAAGCCGAAGCTGA
Hu_HIF2a_R_qPCR-1	Reverse	qPCR		TCCTCATGGTCGCAGGGATGAGT
Hu_RAB27b_F_qPCR-1	Forward	qPCR		GGGGATTCAGGGGTGGGGAAGA
Hu_RAB27b_R_qPCR-1	Reverse	qPCR		GGTGAGACTCCGGAACCGCTCT
Hu_IGF2_F_qPCR-1	Forward	qPCR		TTACCGCCCCAGTGAGACCCTG
Hu_IGF2_R_qPCR-1	Reverse	qPCR		GGGGAAGTTGTCCGGAAGCACG
Hu_ANGPTL2_F_qPCR-1	Forward	qPCR		CTGGTCCGGCCGCAAAGTCTTT
Hu_ANGPTL2_R_qPCR-1	Reverse	qPCR		GCCTCCCTTCTGGTAGTGGGCA
Hu_Decorin_F_qPCR-1	Forward	qPCR		TGGCCAACACGCCTCATCTGAG
Hu_Decorin_R_qPCR-1	Reverse	qPCR		GGTGTTGTGTCCAGGTGGGCAG
Hu_ANGPT1_F_qPCR-1	Forward	qPCR		ACGCTCTGCAGAGAGATGCTCCA
Hu_ANGPT1_R_qPCR-1	Reverse	qPCR		AGCCGTGTGGTTCTGAACTGCA
Hu_ANGPT2_F_qPCR-1	Forward	qPCR		ACGATGACTCGGTGCAGAGGCT
Hu_ANGPT2_R_qPCR-1	Reverse	qPCR		TTCCGCGTTTGCTCCGCTGTTT
Hu_ChIP_IGF2_Runx1-1_F	Forward	ChIP	−7.5 kb of IGF2 TSS	CACACAGCCTTCTGTCTGGGCG
Hu_ChIP_IGF2_Runx1-1_R	Reverse	ChIP	−7.5 kb of IGF2 TSS	CAATCTAGGGGGCCAGCCTGGA
Hu_ChIP_IGF2_Runx1-2_F	Forward	ChIP	−4.2 kb of IGF2 TSS	ACCCGTGAGCTGCTTTCTGTGC
Hu_ChIP_IGF2_Runx1-2_R	Reverse	ChIP	−4.2 kb of IGF2 TSS	AGCTGCTCCTCCCTCCAAGGTC
Hu_ChIP_IGF2_Runx1-3_F	Forward	ChIP	−2.4 kb of IGF2 TSS	CTGCCCCGGGCCTctcagtaat
Hu_ChIP_IGF2_Runx1-3_R	Reverse	ChIP	−2.4 kb of IGF2 TSS	GCTGTGGCCTGGTGAGCATCCT
Hu_ChIP_IGF2_Runx1-4_F	Forward	ChIP	−0.4 kb of IGF2 TSS	AGGCTGCGAACCTGTCCAATCG
Hu_ChIP_IGF2_Runx1-4_R	Reverse	ChIP	−0.4 kb of IGF2 TSS	TGAGCAAGGTACTCAGCTGGGGG
Hu_ChIP_MIP1alpha_F	Forward	ChIP	−0.25 kb of MIP1A TSS	CTCTTCACACTCACAGGAGA
Hu_ChIP_MIP1alpha_R	Reverse	ChIP	−0.25 kb of MIP1A TSS	TAGGCAGCCCTGGCGGAT

Sequences of oligonucleotide primers used in qPCR and ChIP assays.

Abbreviations: ChIP, chromatin immunoprecipitation; qPCR, quantitative PCR.

EV Isolation, Quantitation, and Protein Cargo Identification

As described previously (21), conditioned media were collected from HESCs after 72 hours of differentiation. Conditioned media were centrifuged at 3000 g for 10 minutes and 16 500 g for 20 minutes at 4 °C. Following centrifugation, supernatant was used to extract EVs using the precipitation-based miRCURY Exosome Cell/Urine/CSF Kit (Qiagen #76743) as described in the manufacturer’s protocol. Before liquid chromatography/mass spectrometry (LC/MS) analysis, harvested EVs were further purified by ultracentrifugation at 120 000 g for 90 minutes at 4 °C in a Sorvall MX 120 Plus Micro-Ultracentrifuge.

EV quantitation was performed using microfluidic resistive pulse sensing on a Spectradyne nCS1^™ instrument (Spectradyne LLC, USA) according to the manufacturer’s instructions. EV pellets were resuspended in filtered PBS and further diluted with PBS supplemented with 1% Tween20. EV suspension was loaded on polydimethylsiloxane cartridges that were factory calibrated to quantify particles only within specific size ranges. We utilized the C-400 cartridge to quantify particles from 65 to 400 nm in size as previously described (16, 19). Data are presented as average particle concentrations (p/mL) of repeated measurements of independent samples.

After EV isolation and purification, EV pellets were submitted to the Mass Spectrometry Laboratory at the University of Illinois at Urbana-Champaign. LC/MS proteomics data were analyzed by Mascot (Matrix Science) to identify EV cargo proteins, and label-free protein quantitation was performed as described previously (12). Briefly, EVs were lysed in a buffer containing 6 M guanidinium chloride, 0.1% sodium deoxycholate, and 100 mM triethy-lammonium bicarbonate. Protein concentration was measured using a bicinchoninic acid assay (Pierce #23225), and remaining samples were heated at 95 °C for 15 minutes to denature proteins and reduce/alkylate disulfide bonds. Proteins were digested sequentially with 500 ng LysC protease (3 hours at 30 °C) and 500 ng trypsin (overnight at 37 °C). An equal amount of protein from treatment or control EV samples was then acidified, desalted, dried, and injected into a Thermo Scientific UltiMate 3000 RSLCnano system at 300 nL/min.

Peptides were separated, washed, and analyzed by mass spectrometry. MS1 scans (350–1500 m/z) were collected at 120k resolution, followed by MS2 scans (15 k resolution) of the top 15 ions after higher energy collisional dissociation fragmentation. Isotope exclusion and a 60-second dynamic exclusion window were applied to reduce oversampling. Data were analyzed using Mascot v2.8.0. Label-free protein quantification was performed using Mascot Distiller v2.8.0 based on extracted ion chromatograms, requiring at least 2 peptides per protein.

Chromatin Immunoprecipitation

Chromatin immunoprecipitation (ChIP) assays were performed on HESC collected after 72 hours of differentiation using a ChIP Kit (Abcam Cat# ab500, RRID:AB_2924347) following the manufacturer’s protocol. For immunoprecipitation, fragmented DNA was incubated overnight with 5 μg of anti-RUNX1 (Abcam Cat# ab272456, RRID:AB_3675503) or rabbit IgG (Abcam Cat# ab172730, RRID:AB_2687931). Real-time PCR was performed after immunoprecipitation and DNA purification steps. Primers were designed spanning putative RUNX1 response elements within the regulatory region of the IGF2 gene and a positive control within the MIP1α regulatory region (22). Resulting signals were normalized to input DNA, and data were displayed as percent input and SEM calculated from 3 independent experiments using HESCs from the same donor.

Endothelial Tube Formation Assay

To assess capillary-like tube formation ability, 33 750 human umbilical vein endothelial cells (HUVECs) (PromoCell #C12203) were seeded onto Matrigel basement membrane (Corning) that had solidified for 30 minutes at 37 °C. Cells were incubated for 6 hours in the presence of EVs (~2 × 10⁸ particles) from either HESCs treated with RUNX1, IGF2, ANGPTL2, or control siRNA. We used Calcein-AM (Thermo Fisher #C1430) at a final concentration of 2 μg/mL to stain the cells. Random microscopic images were captured on a Zeiss 710 fluorescent confocal microscope, and, to compare the tube formation ability across treatment groups, the number of junctions, segments, and branches were calculated in ImageJ as described previously (23). Representative images are shown, and quantitation data is from 3 to 5 biological replicates.

Trophoblast Culture and Differentiation

Human trophoblast stem cells were acquired from Dr. Michael Soares’ laboratory at the University of Kansas Medical Center with the permission of Dr. Hiroaki Okae of Tohoku University who developed the line (24, 25). Cells were cultured on culture dishes coated with 5 μg/mL mouse collagen IV (Corning #CB40233). Stem cells remain in a highly proliferative state when cultured in complete medium consisting of DMEM/F-12, 0.2% FBS, 0.3% fatty acid–free bovine serum albumin (Thermo Fisher #BP704100), 1% ITS-X (Thermo Fisher #51500056), 0.8 mM valproic acid (Sigma #P4543), 50 ng/mL recombinant human epidermal growth factor (Sigma #E9644), 1.5 mg/mL L-ascorbic acid (Sigma #A8960), 5 μM Y27632 (Reprocell #04-0012-02), 0.1 mM 2-mercaptoethanol (Sigma #21985-023), 1 μM SB431542 (Reprocell #04-0010), 2 μM CHIR99021 (Reprocell #04-0004), 0.5 μM A83-01 (Reprocell #04-0014), and penicillin–streptomycin.

To induce trophoblast stem cells differentiation into EVT cells, cells were seeded in chamber slides (Ibidi) with polymer coverslips coated with 1 μg/mL collagen IV and grown on days 1 and 2 of differentiation in a differentiation media containing DMEM/F-12, 0.3% bovine serum albumin, 1% ITS-X, 0.1 mM 2-mercaptoethanol, 2.5 μM Y27632, 7.5 μM A83–01, 4% KnockOut Serum Replacement (Thermo Fisher #10828010), 100 ng/mL human neuregulin-1 (NRG1) (Cell Signaling Technology #5218SC), and penicillin–streptomycin supplemented with 2% Matrigel (Corning #CB40234). On day 3, the media was replaced with the same media, excluding NRG1, and supplemented with 0.5% Matrigel. On day 6, media was replaced with the same media, excluding NRG1 and KnockOut Serum Replacement, and supplemented with 0.5% Matrigel. Media throughout the differentiation process were supplemented with EVs (~2 × 10⁸ particles) derived from HESCs treated with RUNX1, IGF2, or control siRNA. On day 8, cells were analyzed for markers of EVT differentiation.

Immunocytochemistry

EVTs were fixed with 3.7% formaldehyde and permeabilized with 0.3% Triton-X 100 in PBS for 5 minutes, followed by blocking with PBS + 2% FBS. Cells were incubated overnight with primary antibodies (1:1000 dilution) against human leukocyte antigen-G (Novus Cat# NBP1-43123-0.025 mg, RRID:AB_10006482), matrix metalloproteinase 2 (MMP2; Cell Signaling Technology Cat# 40994, RRID:AB_2799191) or Integrin αV (Cell Signaling Technology Cat# 4711, RRID: AB_2128178). The next day, cells were washed with PBS and then incubated with the appropriate DyLight-conjugated or Cy3-conjugated secondary antibodies (1:500 dilution) (Jackson Laboratories; RRID:AB_2340616, AB_2315777, AB_2307443). Cells were mounted with Prolong GOLD mounting medium (Thermo Fisher #P36941) containing 4,6-diamidino-2- phenylindole nuclear counterstain. Imaging was performed on a Zeiss 710 confocal microscope, and relative fluorescent quantitation was done in ImageJ.

Statistical Analysis

Statistical analysis was performed in all experiments using an unpaired Student’s t-test to determine if treatments were statistically different from the control. GraphPad Prism Version 9 (GraphPad Inc.) was utilized for statistical analysis. A P-value of ≤ .05 was needed to be considered statistically significant.

Results

RUNX1 Acts Upstream of HIF2α/RAB27B to Control EV Secretion by HESCs

In RUNX1 cKO mice, ablation of RUNX1 from endometrial stromal cells induced abnormal responses from other cell types within the uterus. Namely, mesometrial endothelial cells exhibited deficient vascular network formation, and trophoblast cells displayed reduced invasion into the uterine tissue (14). This phenomenon suggests the possibility that the effects of endometrial stromal RUNX1 may be related to paracrine signaling, which would explain how the absence of endometrial stromal RUNX1 impairs the function of other tissues. RUNX1 has been previously shown to regulate the expression of secreted proteins in the skin as well as the muscle (26, 27), which may suggest a similar role for RUNX1 in uterine tissue. Also, RUNX1 signaling pathways in mesenchymal stem cells have been suggested to be mediated by EV secretion to promote angiogenesis (26). EV signaling from endometrial stromal cells is particularly critical for supporting a successful pregnancy (15). For these reasons, we first investigated how RUNX1 may influence EV secretion via a signaling pathway that is vital for EV secretion and is conserved in both mice and human endometrial stromal cells during decidualization (16, 19). Within decidualizing endometrial stromal cells, transcription factor HIF2α is known to regulate the expression of RAB27B gene, which codes for a vesicular trafficking protein crucial to the secretion of EVs (16, 19). We aimed to determine if this EV secretion pathway is dysregulated by the absence of RUNX1 in decidualizing HESCs. To do this, we used a well-established system to induce decidualization in HESCs (28). A differentiation cocktail consisting of 1 μM progesterone, 10 nM 17-β-estradiol, and 0.5 mM 8-bromo-adenosine-3′,5′-cyclic monophosphate was added to HESC media, and after 72 hours of differentiation, relevant gene transcripts, such as RUNX1, HIF2α, and RAB27B, were quantified.

Before induction of the decidualization program, HESCs were treated with 20 nM of scrambled control siRNA or siRNA specifically targeted to RUNX1 transcripts for 24 hours. After 72 hours of decidualization, treatment with 20 nM RUNX1 siRNA led to a robust ~80% knockdown of RUNX1 gene expression as quantified by quantitative PCR (Fig. 1A). Additionally, we found that RUNX1 knockdown caused a reduction in HIF2α expression (Fig. 1B), which led to a concomitant decrease in RAB27B gene expression (Fig. 1C).

Figure 1.

RUNX1 mediates EV secretion in differentiating HESCs by regulating HIF2α/RAB27B gene expression. HESCs were treated with 20 nM control or RUNX1-specific siRNA for 24 hours, then grown in culture under hypoxic conditions with decidualization cocktail for 72 hours. RNA was extracted and gene expression analysis was performed using primers specific for RUNX1 (A) and members of an EV secretion pathway in differentiating HESCs, transcription factor HIF2α (B) and vesicular trafficking protein RAB27B (C). EVs were also isolated from conditioned media of RUNX1-specific or control siRNA using the miRCURY (Qiagen) kit and analyzed by microfluidic resistive pulse sensing (Spectradyne, LLC). Concentration of EVs ranging from 60 to 400 nm was quantified (D) Data shown as mean fold change ± SEM (n = 3–6 independent experiments). *P < .05, **P < .01 relative to control siRNA treatment.

Abbreviations: EV, extracellular vesicle; HESC, human endometrial stromal cell; RUNX1, runt-related transcription factor 1; siRNA, small interfering RNA.

After 72 hours of differentiation, we also collected the conditioned media from the cells. We isolated EVs from the conditioned media using a precipitation-based kit (miRCURY) and, after resuspension in PBS, quantified the EV concentration through microfluidic resistive pulse sensing. We found that treatment with RUNX1 siRNA caused decidualizing HESCs to secrete about half as many EVs as control siRNA-treated HESCs (Fig. 1D). Together, these data suggest an important mechanism through which human decidual RUNX1 may enhance uterine angiogenesis and trophoblast differentiation, regulating the secretion of EVs.

Absence of RUNX1 in HESC Alters Specific EV Protein Cargoes

EVs contain a variety of cargoes, including proteins, RNA, and lipids, which can induce a functional change in the recipient cells (29, 30). EVs isolated from endometrial cells have been shown to contain many proteins that are known to play roles in key processes during early pregnancy, such as angiogenesis, trophoblast differentiation, extracellular matrix remodeling, and decidualization. The endometrial EV cargo proteins that regulate these processes are diverse, but they include angiopoeitin-1 (angiogenesis), IGF2 (trophoblast differentiation), MMP2 (extracellular matrix remodeling), and pyruvate kinase mutase (decidualization), among many others (16, 31). Alteration to the EV protein cargo profile can result in dysfunction in the processes typically guided by endometrial EV secretion, potentially leading to adverse reproductive outcomes (15, 32).

Therefore, we aimed to determine if, in addition to the regulation of EV secretion, RUNX1 also plays a role in determining the protein cargoes of EVs secreted by decidualizing HESCs. After treatment with RUNX1 or control siRNA, followed by the addition of differentiation cocktail, as described earlier, conditioned media were collected from the treated HESCs. We proceeded to isolate EVs from the conditioned media using a precipitation-based miRCURY Exosome Cell/Urine/CSF Kit (Qiagen #76743). After isolation, we further purified the EVs by subjecting them to ultracentrifugation. The resultant EV pellet was then analyzed by LC/MS to identify the protein cargoes of the EVs. Relative abundance of cargo proteins was assessed based on label-free quantitative LC/MS.

Comparison of protein cargoes isolated from EVs secreted by cells treated with control siRNA or RUNX1 siRNA [data deposited at (33)] revealed a number of differential protein cargoes between the 2 treatment groups. Eight hundred seventy-five proteins were found to be common EV cargoes regardless of the expression level of RUNX1 in the HESCs. In contrast, 199 cargoes were unique to EVs from control siRNA-treated cells, and 298 cargoes were unique to EVs derived from HESCs lacking RUNX1 [Fig. 2A and data deposited at (33)].

Figure 2.

RUNX1 partially influences HESC EV protein cargo composition via gene expression regulation. HESCs were treated with 20 nM control or RUNX1-specific siRNA for 24 hours then grown in culture under hypoxic conditions with decidualization cocktail for 72 hours. EVs were isolated from conditioned media of Runx1-specific or control siRNA using ultracentrifugation. Proteomic analysis identified cargo proteins present in either or both samples (A). RNA was extracted and gene expression analysis was performed using primers specific for genes encoding select differential EV cargoes, ANGPT1 and ANGPT2 (B), as well as DCN, ANGPTL2, and IGF2 (C). 36B4 was used to normalize gene expression. Data shown as mean fold change ± SEM (n = 3–6 independent experiments). *P < .05, relative to control siRNA treatment.

Abbreviations: EV, extracellular vesicle; HESC, human endometrial stromal cell; RUNX1, runt-related transcription factor 1; siRNA, small interfering RNA.

Among the shared EV cargo proteins, we found several proteins involved with EV biogenesis, including various annexins, tetraspanins, as well as classical markers of EVs, CD9 and CD81. Also included in the shared proteins were angiogenic modulators, metabolic regulators, growth factors, and binding proteins, all of which have been previously identified as EV cargoes in decidualizing HESCs (16) and appear to be unaffected by the absence of RUNX1 (Table 2, center).

Table 2.

Partial list of relevant EV cargoes isolated from HESCs

EV cargoes from control siRNA-treated HESCs	EV cargoes found in both	EV cargoes from RUNX1 siRNA-treated HESCs
Angiogenic modulators	EV biogenesis	Angiogenic modulators
Angiopoietin-related protein 2	Annexins A1, A2, A6, A11	Angiopoietin-1
Decorin	CD81	Angiopoietin-2
Growth regulation	CD9
IGF2	HSP-90β
	PDCD6IP
	Tetraspanin-14
	Tetraspanin-9
	VAT-1
	Angiogenic modulators
	ADAM-10
	ADAMTS-1
	ADAMTS-2
	GJA1
	ITGB1
	LOXL2
	MMP1
	TIMP2
	TIMP3
	Growth regulation
	IGF2R
	IGFBP7
	TGFBI
	Metabolic regulation
	PKM
	SCL1A5

A partial list of EV cargoes with potential for angiogenic, growth regulation, EV biogenesis, or metabolic regulation secreted by RUNX1-specific siRNA HESCs or control siRNA-treated HESCs, as identified by mass spectrometry after 72 hours of differentiation. The complete list of accession numbers and gene names is provided at Harvard Dataverse (33).

Abbreviations: EV, extracellular vesicle; HESC, human endometrial stromal cell; siRNA, small interfering RNA.

We also found that, while some proteins were present as EV cargoes in both control and RUNX1 siRNA-treated groups, there were some changes to the amount of each protein detected within EVs after RUNX1 siRNA treatment. Interestingly, when examining the fold change in the EV cargo load of proteins known to be angiogenic modulators (Table 3), we found that several proteins traditionally associated with promotion of angiogenesis, ANGPTL2, decorin (DCN), IGF2, and MMP3 and 14, were found to be absent after RUNX1 siRNA treatment, while traditional negative regulators of angiogenesis, metalloproteinase inhibitor 2 and 3 were found to be increased in EVs after siRNA treatment. However, surprisingly, a couple of positive promoters of angiogenesis, angiopoietin 1 and 2, were found in HESC EVs after RUNX1 siRNA treatment. Similarly, when examining the fold change in EV cargo load of proteins known to be influencers of trophoblast differentiation (Table 4), we found that Erbb2 receptor tyrosine kinase 4 and IGF2, factors vital for trophoblast differentiation, were absent in HESC EVs after RUNX1 siRNA treatment. In contrast, inhibitors of IGF2 signaling, cation-independent mannose-6-phosphate receptor, and IGF binding protein 5 were increased in EVs after RUNX1 siRNA treatment.

Table 3.

RUNX1 induces changes in EV cargo load of potentially angiogenic proteins in differentiating HESCs

Gene name	Protein description	Fold change
Angiogenic modulators
ANGPT1	Angiopoietin-1	inf
ANGPT2	Angiopoietin-2	inf
TIMP2	Metalloproteinase inhibitor 2	4
TIMP3	Metalloproteinase inhibitor 3	2.5
LOXL2	Lysyl oxidase homolog 2	1.928571429
MMP2	72 kDa type IV collagenase	1.454545455
ADAMTS1	A disintegrin and metalloproteinase with thrombospondin motifs 1	1.428571429
ITGB1	Integrin beta	1.333333333
MMP1	Interstitial collagenase	1.333333333
ADAM10	Disintegrin and metalloproteinase domain-containing protein 10	1
GJA1	Gap junction a-1 protein	1
ADAMTS5	A disintegrin and metalloproteinase with thrombospondin motifs 5	0.25
ANGPTL2	Angiopoietin-related protein 2	0
DCN	Decorin	0
IGF2	Insulin-like growth factor II	0
MMP3	Stromelysin-1	0
MMP14	Matrix metalloproteinase-14	0

A partial list of EV cargoes with potential for angiogenic function secreted by RUNX1-specific siRNA HESCs compared to control siRNA-treated HESCs, as identified by mass spectrometry after 72 hours of differentiation. Fold change is shown as the number of protein matches in EVs isolated from RUNX1-specific siRNA HESCs divided by the number of protein matches in EVs isolated from control siRNA HESCs. Fold change is provided. Inf refers to proteins that were only present in HESC EV cargo after treatment with RUNX1-specific siRNA, while 0 refers to proteins that were totally absent in HESC EV cargo after treatment with control siRNA. The complete list of accession numbers and gene names is provided at Harvard Dataverse (33).

Abbreviations: EV, extracellular vesicle; HESC, human endometrial stromal cell; siRNA, small interfering RNA.

Table 4.

RUNX1 induces changes in EV cargo load of proteins with the potential to modulate trophoblast differentiation in differentiating HESCs

Gene name	Protein description	Fold change
Regulators of trophoblast differentiation
IGFBP5	Insulin-like growth factor-binding protein 5	inf
IGF2R	Cation-independent mannose-6-phosphate receptor	3.5
IGFBP7	Insulin-like growth factor-binding protein 7	1.5
CDC42	Cell division control protein 42 homolog	1.333333333
TGFBI	Transforming growth factor-β-induced protein ig-h3	1
EIF2S3	Eukaryotic translation initiation factor 2 subunit 3	1
EIF4A1	Eukaryotic initiation factor 4A-I	1
ERBB4	Erb-b2 receptor tyrosine kinase 4	0
IGF2	Insulin-like growth factor II	0

A partial list of EV cargoes with potential to modulate trophoblast differentiation secreted by RUNX1-specific siRNA HESCs compared to control siRNA-treated HESCs, as identified by mass spectrometry after 72 hours of differentiation. Fold change is shown as the number of protein matches in EVs isolated from RUNX1-specific siRNA HESCs divided by the number of protein matches in EVs isolated from control siRNA HESCs. Fold change is provided. Inf refers to proteins that were only present in HESC EV cargo after treatment with RUNX1-specific siRNA, while 0 refers to proteins that were totally absent in HESC EV cargo after treatment with control siRNA. The complete list of accession numbers and gene names is provided at Harvard Dataverse (33).

Abbreviations: EV, extracellular vesicle; HESC, human endometrial stromal cell; siRNA, small interfering RNA.

We specifically further investigated the angiogenic modulators ANGPTL2 and DCN, as well as the growth factor IGF2, which may also play a role in trophoblast differentiation, in addition to angiogenesis. These proteins were found in the control-treated EVs but were no longer present in the EV cargoes from RUNX1 siRNA-treated HESCs (Table 2, left). This implies that the presence of RUNX1 is required for these particular proteins to be included in the HESC EV cargoes. We hypothesized that this might be due to the role of RUNX1 as a transcription factor in regulating the gene expression of proteins. As a result of the ablation of RUNX1, there would be a decrease in the cytoplasmic accumulation of proteins whose expression is regulated by RUNX1. This would make it less likely that these proteins are available to be enclosed within the multivesicular body and eventually secreted within EVs.

To test this hypothesis, we analyzed the gene expression of several relevant proteins that were found to be differentially present in the EV cargoes of our 2 treatment groups. Surprisingly, we had identified 2 angiogenic modulators, angiopoietin-1 and angiopoietin-2, that were only present in EVs from RUNX1 siRNA-treated HESCs (Table 2, right). However, upon further investigation, we found that the absence of RUNX1 did not induce an increase in gene expression of these factors (Fig. 2B). Similarly, the gene expression of the angiogenic modulator DCN was unchanged after RUNX1 siRNA treatment, despite DCN only being present in EVs isolated from control treated HESCs (Fig. 2C). Interestingly, the other 2 relevant factors that had been identified to be present as EV cargoes in control treated HESCs but absent in EVs from HESCs treated with RUNX1 siRNA, ANGPTL2 and IGF2, both had decreased gene expression in the absence of RUNX1 (Fig. 2C). This suggests that during decidualization, RUNX1 normally promotes ANGPTL2 and IGF2 gene expression. ANGPTL2 is a key angiogenic factor (34), and IGF2, in addition to being an angiogenic factor, is a crucial regulator of trophoblast differentiation (35, 36).

Together, these data suggest that the absence of RUNX1 induces changes to the EV protein cargo profile of HESCs. Specific proteins, such as ANGPTL2 and IGF2, are now absent from the HESC EV cargo due to RUNX1 no longer being present to promote their gene expression. As both proteins are likely to play roles in angiogenesis and trophoblast differentiation, this could be a cause of the phenotypes displayed in the previously reported RUNX1 cKO mouse model (14).

RUNX1 Expression in HESCs Enhances Proangiogenic Activity of HESC EVs

Endometrial stromal regulation of uterine angiogenesis is crucial during early pregnancy. The development of an extensive vascular architecture within the uterus is necessary to meet the demands of the growing embryo. Maternal vasculature plays a crucial role in supporting the embryo before placentation and acts as the maternal blood supply during the placentation process (37, 38). Several angiogenic factors produced by the decidua have been previously reported to regulate uterine angiogenesis (26–28). Here, we observed that RUNX1 promotes the gene expression of the angiogenic factors ANGPTL2 and IGF2, which are secreted from HESCs as EV protein cargoes. Additionally, we have shown that depletion of RUNX1 induces a change to the EV cargo protein profile, including the exclusion of these angiogenic factors from the cargo proteins. To assess whether the angiogenic promotion capability of HESC EVs is compromised due to the altered protein cargoes resulting from HESC RUNX1 depletion, we conducted a functional assay involving the knockdown of transcripts corresponding to specific angiogenic cargo proteins.

We previously reported that primary HUVECs gradually form capillary-like structures when seeded onto a matrix membrane, and the addition of HESC-derived EVs enhances the tube formation process (16). We found that the addition of EVs derived from HESCs depleted of RUNX1 displayed impaired tube formation compared to the addition of an equivalent number of control EVs (Fig. 3A and 3B), as assessed by quantitation of junctions, segments, and branches formed by the HUVECs (Fig. 3E). Also, EVs derived from HESCs depleted of ANGPTL2 (Fig. 3C and 3E) or IGF2 (Fig. 3D and 3E) when added to the HUVEC culture exhibited reduced tube formation compared to the addition of control EVs.

Figure 3.

RUNX1 expression in HESCs enhances HESC EVs’ proangiogenic activity by promoting the expression of angiogenic factors. HESCs were treated with 20 nM control or RUNX1-specific siRNA for 24 hours, then grown in culture under hypoxic conditions with decidualization cocktail for 72 hours. (A-D) EVs were isolated from conditioned media of HESCs treated with control (A), RUNX1- (B), ANGPTL2- (C), or IGF2- (D) specific siRNA using the miRCURY (Qiagen) kit. EVs were added at a concentration of 3 × 10¹¹ p/mL to HUVECs plated with basement membrane matrix and cultured for 72 hours. Representative images were taken of tube formation. (E) Tube formation capability was quantified by counting the number of junctions, nodes, branches, and segments in ImageJ from 4 random microscopic images per replicate. Data shown as mean fold change ± SEM. *P < .05, **P < .01.

Abbreviations: EV, extracellular vesicle; HESC, human endometrial stromal cell; HUVEC, human umbilical vein endothelial cell; RUNX1, runt-related transcription factor 1; siRNA, small interfering RNA.

These data reveal the mechanism by which RUNX1 and its downstream targets expressed in endometrial stromal cells regulate the function of endothelial cells. This study identifies specific angiogenic factors, such as ANGPTL2 and IGF2, as critical RUNX1-regulated cargoes carried by EVs secreted from HESCs. The presence of these angiogenic factors is crucial for HESC EVs’ ability to enhance angiogenesis, as demonstrated by our in vitro angiogenesis model.

RUNX1 Regulates IGF2 Cargo in EVs Secreted by HESCs to Modulate Trophoblast Differentiation into the EVT Lineage

Throughout placentation, the maternal blood supply remains a source of nutrition for the growing embryo. In addition to the extensive vasculature architecture developed within the decidua, the fetal trophoblast cells must also invade deep into the decidua to remodel maternal spiral arteries and redirect blood flow toward the fetus. Specialized trophoblast cells called EVTs are primarily responsible for this invasion and remodeling (39). EVT differentiation from trophoblast progenitor cells is a tightly controlled process, and our lab has previously described that maternal-derived EVs markedly enhance the differentiation of trophoblast stem cells into EVTs (16).

The differentiation of trophoblast stem cells to EVT is characterized by an increased expression of IGF2 by EVTs, and autocrine secretion of IGF2 has long been thought to be a driver of human trophoblast invasiveness (40). However, our findings suggest a potential role for the contribution of maternal IGF2, packaged within HESC EVs, on EVT differentiation. We hypothesized that RUNX1-dependent expression and secretion of HESC IGF2 could influence EVT differentiation via EV-mediated cell-cell communication.

IGF2 gene expression is regulated by RUNX1 in decidualizing HESCs (Fig. 2D). To determine if RUNX1 control of IGF2 expression is a direct effect, we tested several putative RUNX1 binding sites in the IGF2 gene regulatory region and found through ChIP studies that RUNX1 significantly occupies DNA within the IGF2 regulatory region (Fig. 4A). These data suggest that RUNX1 is a direct regulator of IGF2 gene expression in decidualizing HESCs.

Figure 4.

RUNX1 expression in HESCs modulates trophoblast differentiation into EVT lineage by regulating IGF2 expression. (A) ChIP was performed on HESCs that had been treated with decidualization cocktail for 72 hours. Chromatin enrichment was quantified by quantitative PCR. Primers flanking RUNX1 consensus sequences within the IGF2 promoter or the positive control MIP1α regulatory region were used to quantify chromatin enrichment. Representative data shown as percent input ± SEM (n = 3 independent experiments). * P < .05, relative to IgG. (B-C) Primary trophoblast stem cells were induced to differentiate into EVTs and were incubated with EVs isolated from HESCs treated with control, RUNX1-specific siRNA (B), or IGF2-specific siRNA (C). Expression of EVT markers MMP2, human leukocyte antigen-G, and ITGAV was quantified by immunocytochemistry at the halfway point (day 4) and completion (day 8) of the differentiation process. The relative fluorescence was quantified by ImageJ and normalized to 4,6-diamidino-2- phenylindole. Data is shown as mean fold change ± SEM relative to treatment with control siRNA EVs. **P < .01, ***P < .001.

Abbreviations: ChIP, chromatin immunoprecipitation; EVT, extravillous trophoblast; HESC, human endometrial stromal cell; RUNX1, runt-related transcription factor 1; siRNA, small interfering RNA.

To test whether the expression of RUNX1 in HESCs altered the ability of HESC EVs to influence trophoblast function, we utilized a primary human trophoblast stem cell line derived by Drs. Hiroaki Okae and Takahiro Arima. These cells are maintained in a self-renewing stem state and can be differentiated into EVTs by an 8-day differentiation process (24).

Over the course of differentiation, cells exhibit morphological changes as well as an increase in the expression of markers of EVT differentiation, HLAG, MMP2, and ITGAV. We compared the expression of these markers in response to supplementation with EVs derived from control, RUNX1 siRNA-treated (Fig. 4B), or IGF2 siRNA-treated (Fig. 4C) HESCs. As shown in Fig. 4B, EVs from RUNX1 siRNA-treated HESCs were moderately less efficient in promoting EVT differentiation compared to control EVs. At the midpoint of differentiation (day 4), only the HLAG marker was differentially expressed between the treatments; however, by the end of differentiation (day 8), this difference was not apparent. In contrast, the control EVs appear to be efficiently promoting the expression of both MMP2 and ITGAV markers at later stages of differentiation compared to those secreted by RUNX1 siRNA-treated HESCs.

Remarkably, as seen in Fig. 4C, EVs derived from IGF2 siRNA-treated HESCs displayed a more prominent deficiency in enhancing EVT differentiation. Throughout the differentiation process, all markers of EVT differentiation were reduced in the cells treated with EVs derived from IGF2 siRNA-treated HESCs compared to control EVs, suggesting a significant role for IGF2 as a maternal EV cargo in promoting trophoblast to EVT differentiation. Collectively, these data point to a pathway where endometrial stromal RUNX1-regulated expression and secretion of IGF2 as EV cargo enhances the differentiation of trophoblast stem cells to the EVT lineage during early pregnancy.

Discussion

Endometrial stromal cells play a central role during early pregnancy by secreting a variety of proteins that mediate various processes essential for reproductive success. Therefore, it is crucial to understand the function of proteins within endometrial stromal cells and to determine how their role can impact the function of other cell types in the uterus. In vivo rodent studies can provide valuable insights into these interactions; however, not all processes are conserved across species, and validation in a human in vitro system remains an important step in advancing our understanding of the cellular mechanisms that operate during early pregnancy in humans.

Here, we built upon the exciting recent findings of Kannan et al (8), who reported that the conditional knockout of transcription factor RUNX1 in the murine uterus led to severe subfertility due to deficient maternal vascularization and impaired trophoblast differentiation and development. Using a well-established human in vitro model of endometrial decidualization, we discovered that the role of decidual RUNX1 is indeed conserved between mice and humans, with downregulation of RUNX1 expression in decidualizing HESCs resulting in a muted capacity of HESCs to promote angiogenesis and differentiation in human endothelial and trophoblast cells, respectively. Additionally, we were able to provide unique insights into the mechanism by which HESC RUNX1 coordinates these processes.

A major finding of this study is that RUNX1 plays a role in regulating EV secretion in HESCs. We found that in decidualizing HESCs, RUNX1 acts upstream of the HIF2α/RAB27B pathway, which is conserved in decidual cells of both mice and humans, to promote EV secretion. This finding enhances our understanding of how the ablation of decidual RUNX1 had such profound consequences for maternal angiogenesis and trophoblast differentiation and development in the RUNX1 cKO mouse model described by Kannan et al (8). EV secretion is a crucial method of cell-to-cell communication during early pregnancy. Its dysregulation can lead to dire consequences for the health of a pregnancy (31). To our knowledge, we are the first to report the involvement of RUNX1 in endometrial stromal cell EV secretion; however, mutations in the RUNX1 gene have been previously linked to defects in dense granule secretion in platelets (32). Interestingly, platelet dense granule secretion is also regulated by RAB27B (33), providing another example of RUNX1 promoting RAB27B-controlled secretion pathways.

In addition to a reduced quantity of EVs secreted into the extracellular space, we also found that downregulation of RUNX1 transcripts led to an altered EV cargo profile. Specific modulators of angiogenesis, such as ANGPTL2 and IGF2, were no longer present as EV cargoes in HESCs treated with RUNX1 siRNA. ANGPTL2 has been previously reported to promote tube formation in endothelial cells (34, 35). IGF2 has also been linked to the growth of vasculature (22) and has been shown to facilitate the transition of trophoblast cells into an invasive lineage (30), making it a crucial protein that likely regulates multiple processes during early pregnancy. We found that the gene expression of ANGPTL2 and IGF2 is directly related to RUNX1 expression. However, a few other differential EV protein cargoes, such as DCN, exhibited gene expression that was unaffected by RUNX1, suggesting that RUNX1 is only indirectly involved in the inclusion of these proteins as EV cargoes through another unidentified mechanism.

One of the most important aspects of the changing EV protein cargo profile that we have observed in HESCs treated with RUNX1 siRNA is that these alterations to EV cargoes lead to a muted response in the recipient cells. Using an in vitro tube formation assay, we showed that EVs derived from HESCs treated with RUNX1 siRNA failed to promote tube formation to the same extent as control HESC EVs. In various biological contexts, RUNX1 has been shown to promote angiogenesis (13, 40–42). In this study, we have discovered that part of RUNX1’s ability to promote angiogenesis in HESCs is mediated by EV signaling. We report that 2 significant angiogenic modulators, ANGPTL2 and IGF2, were found in control HESC EVs but not EVs derived from HESCs treated with RUNX1 siRNA. After further investigation, we found that EVs from ANGPTL2- or IGF2-depleted HESCs failed to promote tube formation by HUVECs to the same extent as control EVs. These data suggest that RUNX1 facilitates uterine angiogenesis by promoting the expression of specific angiogenic factors in HESCs. The inclusion of at least 2 of these factors, ANGPTL2 and IGF2, as EV cargoes is essential for the maximal capacity of HESC EVs to promote uterine angiogenesis.

HESC EVs are also known to enhance the differentiation of human trophoblast stem cells (16) Consistent with the phenotypic defects observed by Kannan et al in mice lacking uterine Runx1 (13), we found that RUNX1 depletion in HESCs resulted in a reduced enhancement of human trophoblast differentiation into EVTs by HESC EVs. Notably, we found that while treating trophoblast stem cells with EVs derived from HESCs treated with RUNX1 siRNA somewhat diminished the enhancement of trophoblast differentiation by HESC EVs, EVs from IGF2-depleted HESCs dramatically reduced trophoblast stem cell differentiation into the EVT lineage. From these data, we postulate that the majority of RUNX1’s influence on trophoblast differentiation comes directly from its regulation of IGF2 expression and secretion. Incomplete knockdown of IGF2 gene expression in HESCs, which occurs after treatment with RUNX1 siRNA, appears to be sufficient to partially mute the influence of HESC EVs on the EVT differentiation process. However, a more complete knockdown of IGF2 gene expression occurs after HESC treatment with IGF2 siRNA, and that directly corresponds to a more dramatic reduction in the ability of HESC EVs to promote EVT differentiation. Traditionally, it has been suggested that only autocrine secretion of IGFs by the embryo is responsible for driving trophoblast proliferation and differentiation (41–43). However, our discovery that RUNX1-regulated IGF2 inclusion as an EV cargo is crucial for HESC EV enhancement of EVT differentiation suggests that the role of maternal paracrine IGF2 signaling in trophoblast differentiation may be more significant than previously believed.

In conclusion, our study has revealed an important role for human decidual RUNX1 in regulating essential processes during early pregnancy, demonstrating conservation of decidual RUNX1’s profound impact on angiogenesis, trophoblast differentiation, and development that was previously uncovered in murine studies. We have also further elucidated the molecular mechanisms by which RUNX1 influences these processes; namely, RUNX1’s regulation of EV secretion and protein cargo composition, particularly the inclusion of key factors ANGPTL2 and IGF2 as EV cargoes. We have also provided intriguing evidence highlighting the significant contribution of maternal IGF2 to trophoblast differentiation, challenging existing paradigms and inviting future investigation. Overall, this study has established that human decidual RUNX1 is a key factor during early pregnancy, influencing uterine angiogenesis and trophoblast differentiation through the regulation of EV-mediated communication. This research provides valuable insights into our growing understanding of the regulation of EV-mediated communication, which is essential for orchestrating the complex series of events crucial for a successful pregnancy.

Data Availability

Original data generated and analyzed during this study are included in this published article or in the data repositories listed in References.