https://orcid.org/0000-0002-8254-863X
Abstract
Are the extracellular vesicles (EVs) secreted by the maternal endometrium uptaken by human embryos and is their miRNA cargo involved in implantation and embryo development?
Data suggest that EVs secreted by human endometrial epithelial cells are internalized by human blastocysts, and transport miRNAs to modulate biological processes related to implantation events and early embryo development.
Successful implantation is dependent on coordination between maternal endometrium and embryo, and EVs role in the required cell-to-cell crosstalk has recently been established. In this regard, our group previously showed that protein cargo of EVs secreted by primary human endometrial epithelial cells (pHEECs) is implicated in biological processes related to endometrial receptivity, embryo implantation, and early embryo development. However, little is known about the regulation of these biological processes through EVs secreted by the endometrium at a transcriptomic level.
A prospective descriptive study was performed. Endometrial biopsies were collected from healthy oocyte donors with confirmed fertility on the day of oocyte retrieval, 36 h after the LH surge. pHEECs were isolated from endometrial biopsies (n = 8 in each pool) and cultured in vitro. Subsequently, conditioned medium was collected and EVs were isolated and characterized. Uptake of EVs by human blastocysts and miRNA cargo of these EVs (n = 3 pools) was analyzed.
EVs were isolated from the conditioned culture media using ultracentrifugation, and characterization was performed using western blotting, nanoparticle tracking analysis, and transmission electron microscopy. EVs were fluorescently labeled with Bodipy-TR ceramide, and their uptake by human blastocysts was analyzed using confocal microscopy. Analysis of the miRNA cargo of EVs was performed using miRNA sequencing, target genes of the most expressed miRNA were annotated, and functional enrichment analysis was performed.
EVs measured 100–300 nm in diameter, a concentration of 1.78 × 1011 ± 4.12 × 1010 (SD) particles/ml and expressed intraluminal protein markers Heat shock protein 70 (HSP70) and Tumor Susceptibility Gene 101 (TSG101), in addition to CD9 and CD81 transmembrane proteins. Human blastocysts efficiently internalized fluorescent EVs within 1–2 h, and more pronounced internalization was observed in the hatched pole of the embryos. miRNA-seq analysis featured 149 annotated miRNAs, of which 37 were deemed most relevant. The latter had 6592 reported gene targets, that in turn, have functional implications in several processes related to embryo development, oxygen metabolism, cell cycle, cell differentiation, apoptosis, metabolism, cellular organization, and gene expression. Among the relevant miRNAs contained in these EVs, we highlight hsa-miR-92a-3p, hsa-let-7b-5p, hsa-miR-30a-5p, hsa-miR-24-3p, hsa-miR-21-5p, and hsa-let-7a-5p as master regulators of the biological processes.
This is an in vitro study in which conditions of endometrial cell culture could not mimic the intrauterine environment.
This study defines potential biomarkers of endometrial receptivity and embryo competence that could be useful diagnostic and therapeutic targets for implantation success, as well as open insight further investigations to elucidate the molecular mechanisms implicated in a successful implantation.
This study was supported by the Spanish Ministry of Education through FPU awarded to M.S.-B. (FPU18/03735), the Health Institute Carlos III awarded to E.J.-B. (FI19/00110) and awarded to H.F. by the Miguel Servet Program ‘Fondo Social Europeo «El FSE invierte en tu futuro»’ (CP20/00120), and Generalitat Valenciana through VALi+d Programme awarded to M.C.C.-G. (ACIF/2019/139). The authors have no conflicts of interest to disclose.
N/A.
Introduction
Embryo implantation is a complex but essential process in pregnancy establishment, that occurs when a competent embryo migrates into the uterus, attaches to, and invades, the receptive endometrium (Cuman et al., 2014). To this end, the endometrium undergoes dynamic changes throughout the menstrual cycle (principally under the regulation of ovarian-derived hormones) (Lessey and Young, 2019) and leading to the proper endometrial receptive phenotype (Aplin and Ruane, 2017), known as the ‘window of implantation’, which comprises a short period of time (4–5 days) (Cuman et al., 2014; Ojosnegros et al., 2021). The probability of achieving pregnancy during this period in a given menstrual cycle is only 30% (Norwitz et al., 2001). This low rate is likely due to the difficulty of synchronizing the embryo-endometrium crosstalk during implantation (Machtinger et al., 2016; Tan et al., 2020b). In this regard, recent studies have reported the role of extracellular vesicles (EVs) in this communication system (Kurian and Modi, 2019; Aleksejeva et al., 2022; Chen et al., 2022).
EVs are membrane-enclosed compartments secreted by all types of cells into the extracellular environment that transport proteins, DNA, mRNA, micro RNAs (miRNAs), and other noncoding RNAs and participate in intercellular communication (Raposo and Stoorvogel, 2013; Colombo et al., 2014; Abels and Breakefield, 2016). Numerous groups have postulated the potential role of EVs in embryo implantation, based on their presence in uterine fluid (UF) (Ng et al., 2013; Luddi et al., 2019) and embryo-conditioned culture medium (Abu-Halima et al., 2017; Pallinger et al., 2017). Indeed, the EVs secreted by embryos can be internalized by the endometrial cells (Giacomini et al., 2017), improving endometrial receptivity (Ashary et al., 2018) and inversely, the EVs secreted by endometrial cells can be taken up by trophoblast cell-derived spheroids, altogether enhancing embryonic adhesive and invasive capacities (Ruiz-González et al., 2015; Greening et al., 2016; Gurung et al., 2020). Further, a recent study found that EVs secreted by endometrial cells from patients with recurrent implantation failure decreased murine embryo competency (Liu et al., 2020). Accordingly, the proteomic profile of trophoblast cells also changed following internalization of the EVs secreted by endometrial cells (Evans et al., 2019). Although human endometrial epithelial cells secrete EVs containing proteins that are implicated in biological processes related to endometrial receptivity, embryo implantation, and early embryo development (Segura-Benítez et al., 2022), little is known about the endometrial EV-mediated transcriptomic regulation of these biological processes.
miRNAs are emerging as transcriptional regulators for many biological processes throughout the human body. These small noncoding RNAs (∼22 nucleotides long) are synthesized in the nucleus and subsequently processed in the cytoplasm to post-transcriptionally regulate gene expression (Liang et al., 2017; de Sousa et al., 2019). Although it is widely accepted that miRNAs suppress mRNA expression though mRNA cleavage, deadenylation, and translational repression, new evidence suggests they may also promote gene expression by targeting promoter elements (Xiao et al., 2017; Tan et al., 2019). To date, there are 2654 annotated human miRNAs in miRBase (Kozomara et al., 2019), and approximately one-third of human genes appear to be conserved miRNAs targets. Since each miRNA can regulate hundreds of targets, miRNAs may collectively modulate up to 60% of protein-coding genes (Tan et al., 2019). The roles of miRNAs are tissue-specific, and within female reproductive tissues, they have been reported in the regulation of oogenesis, fertilization, endometrium-embryo communication, and placentation (Liang et al., 2017).
The EVs present in the UF of different organisms transport miRNA cargo (Campoy et al., 2016; Burns et al., 2018; Nakamura et al., 2019; Tan et al., 2020b; Hua et al., 2021; Xie et al., 2021; Hu et al., 2022) whose composition changes throughout the menstrual cycle, and thus contributes to the implantation process during the window of implantation (Ng et al., 2013; Vilella et al., 2015). In fact, the miRNA cargo of uterine EVs was differentially expressed in fertile and infertile patients (Li et al., 2021; Liu et al., 2021). Since the miRNAs have only been analyzed in endometrial cell lines (Tan et al., 2020a), a descriptive analysis of the miRNA cargo carried by the EVs secreted by the fertile human maternal endometrium is required. Thus, the aim of this study was to characterize the miRNA cargo of EVs secreted by primary human endometrial epithelial cells (pHEECs), corroborate the uptake of these EVs by human embryos and elucidate their potential role in the regulation of embryo competence and implantation success.
Materials and methods
Ethical approval, endometrial biopsy collection, and primary cell culture
Endometrial biopsies n = 24 (8 biopsies × 3 pools) were collected 36 h following the luteinizing hormone surge (day of oocyte retrieval), from healthy oocyte donors (18–35 years old, body mass index <30 kg/m2) with confirmed fertility, at the Instituto Valenciano de Infertilidad (IVI) Clinics (Spain). Donors’ characteristics are presented in Supplementary Table S1. This study was approved by the Clinical Ethics Committee of the IVI Clinics (1902-FIVI-032-HF). All participants provided written informed consent.
The pHEECs were isolated and cultured as previously described (Simón et al., 1993; Segura-Benítez et al., 2022), with a starting concentration of 85 000 cells/ml. Once 70% confluency was reached, cells were treated with 10−8M β-estradiol (E2) and 10−7M progesterone (P4) (Sigma-Aldrich, USA) in culture medium with exosome-depleted fetal bovine serum (Thermo Fisher Scientific, OR, USA) at 37°C and 5% CO2, to mimic the secretory phase of the menstrual cycle. After 48 h, conditioned culture medium was collected for EV isolation. A small volume portion of conditioned culture medium from all 24 patients was pooled into a single pool and then split into two different replicates for EV characterization.
Immunofluorescence staining
To verify the purity of pHEECs isolated, a double immunofluorescence staining was performed. After pHEEC isolation and culture until 60% confluence was reached, cells were fixed with 4% paraformaldehyde for 20 min at room temperature. Samples were blocked with 3% bovine serum albumin (BSA) in PBS and 0.05% Tween 20 (Sigma-Aldrich) for 30 min at room temperature. Subsequently, cells were first incubated overnight with anti-pancytokeratin (epithelial marker) antibody (1:100; Abcam, UK), followed by incubation with its corresponding secondary antibody, AlexaFluor 488 (1:500; Invitrogen, USA). Then, cells were blocked again in the same way as before and incubated with anti-vimentin (stromal component) antibody (1:100; Agilent, USA), followed by incubation with AlexaFluor 568 (1:500; Invitrogen). Finally, cell nuclei were counterstained with 4′,6-diamidino-2-phenylindole dihydrochloride (DAPI) (Life Technologies). Stained cells were visualized and analyzed using a fluorescence microscope (Carl Zeiss™ Axio Vert.A1).
RNA extraction, reverse transcription, and real-time PCR
To determine whether the hormone treatment was effective at mimicking secretory phase, Progesterone Receptor Membrane Component 1 and 2 (PGRMC1, PGRMC2) gene expression levels were analyzed in pHEECs (n = 3 endometrial biopsies) after hormonal treatment. Once confluence was reached, these cells were divided into two groups: (i) E2P4 group (10−8M E2 and 10−7M P4 (Sigma-Aldrich)) and (ii) control group (only with 10−8M E2) for 48 h, and cells from each individual patient and group were collected. Total cellular RNA was isolated using an RNeasy Mini Kit (Qiagen, USA) according to manufacturer’s instructions. RNA was reverse transcribed to cDNA using the PrimeScript RT reagent kit (Takara Bio, USA). Quantitative real-time PCR (qRT-PCR) was performed using Power-Up SYBR Green (Thermo Fisher Scientific). Each sample was amplified in duplicate for 40 cycles and run on a StepOnePlus Real-Time PCR System (Applied Biosystems, USA). Relative expression levels were determined by the ΔΔCt method and normalized to housekeeping gene expression of GAPDH.
EVs isolation via ultracentrifugation
A pool of 80 ml conditioned culture medium was first centrifuged at 300g for 10 min at 4°C, followed by 2000g for 20 min at 4°C to pellet cells. The supernatant was transferred to polycarbonate tubes and centrifuged at 10 000g for 30 min at 4°C to remove cellular debris. Finally, EVs were isolated by double ultracentrifugation at 100 000g for 70 min at 4°C, using a Beckman-Coulter JA-30.50 Ti Rotor.
Nanoparticle tracking analysis
To characterize the size distribution and concentration of the EVs, as well as confirm EV fluorescent labeling, nanoparticle tracking analysis (NTA) was performed using a NanoSight NS300 instrument (Malvern, Spain). Samples were diluted in PBS (ratio 1:8), and two replicates were analyzed. For analysis of labeled EVs, a 532-nm laser diode source with a 565-nm filter was used to detect fluorescence.
Protein extraction and western blot
Western blot (WB) was performed to determine the presence of EV protein markers. A negative control was also analyzed, using fresh culture medium to which the same EV isolation protocol by ultracentrifugation as for the conditioned culture medium was applied. Proteins were extracted from EV pellets using radioimmunoprecipitation assay lysis buffer containing protease inhibitors (Sigma-Aldrich). Protein concentration was measured using a Bradford protein assay (Bio-Rad, USA) following the manufacturer’s protocol. Samples were separated by sodium dodecyl-sulfate polyacrylamide gel electrophoresis, and Heat shock protein 70 (HSP70) (1:200; Santa Cruz Biotechnology), Tumor Susceptibility Gene 101 (TSG101) (1:500; Abcam), CD9 (1:200; Santa Cruz Biotechnology, USA), and CD81 (1:100; Santa Cruz Biotechnology) protein markers were detected using chemiluminescence substrates (Thermo Fisher Scientific), and visualized on an Amersham Imager 680 system (GE Healthcare Bio-Sciences AB, Sweden). β-Actin (1:2000; Santa Cruz Biotechnology) was evaluated for normalization.
Transmission electron microscopy
The morphology of EVs was evaluated by transmission electron microscopy (TEM) using a FEI Tecnai G2 Spirit BioTwin (Thermo Fisher Scientific). EVs were resuspended in PBS, and 6 μl was placed on a carbon-coated grid and contrasted with 2% uranyl acetate.
Embryo devitrification and in vitro culture
To verify whether the EVs secreted by pHEECs are taken up by human blastocysts, we thawed vitrified Day 5 hatching human blastocysts (n = 15 blastocysts) and subsequently cultured them in vitro. All the devitrification material/reagents were obtained from Kitazato (Kitazato Corporation, Japan). Briefly, the Cryotube was retrieved from the liquid nitrogen tank, and immediately placed in prewarmed (37°C) TCM199 medium supplemented with 20% synthetic serum substitute (SSS) and 1M sucrose. After 1 min, the embryos were transferred to TCM199 medium supplemented with 20% SSS and 0.5 M sucrose for 3 min at room temperature. Finally, embryos were washed with TCM199 supplemented with 20% SSS for 5 min followed by 1 min, and cultured in Blastocyst Medium (Genea Biomedx, Australia) at 37°C, with 6% CO2 and 5% O2.
Blastocyst survival and morphology was evaluated 2 h after thawing, based on re-expansion of the blastocele, and appearance of the inner cell mass (ICM) and trophectoderm (TE), in accordance with ASEBIR criteria stablished by the Spanish Association for the Study of Reproductive Biology (Cuevas Saiz et al., 2018), and all blastocysts were grade ‘B’ or ‘C’; any degenerated blastocysts were not used in this study. The use of human blastocysts was approved by the National Commission of Human Assisted Reproduction (CNRHA), the General direction of Research, Innovation, Technology, and Quality (20-04_PI-VAL), and Ethics Committee of Clinical Research of the IVIRMA Valencia Clinic (1902-FIVI-032-HF). Written informed consent was obtained from couples who donated their embryos to research.
EVs labeling and uptake by blastocysts
EVs were resuspended in PBS (300 µg/ml) and labeled with 10 µM BODIPY-TR ceramide (Thermo Fisher Scientific) for 20 min at 37°C. The excess of unincorporated dye was removed by ultracentrifugation at 100 000g for 70 min at 4°C and labeled EVs were resuspended in Blastocyst Medium. Blastocysts were cultured in Blastocyst Medium in the absence or presence of labeled EVs (145 µg/ml in each drop) for 1 or 2 h (n = 5 blastocysts per condition), and subsequently, fixed with 4% paraformaldehyde in PBS (Electron Microscopy Sciences, USA) for 20 min at room temperature. Embryonic cells were stained with 150 µM DAPI (Sigma-Aldrich).
A subset of embryos from each group (n = 3 per condition) was further labeled with immunofluorescence to confirm the presence of healthy TE and epiblast lineages. Cell membranes were permeabilized with 0.5% Triton-X100 in PBS (Sigma-Aldrich), following blocking with 5% BSA in PBS (Sigma-Aldrich) and incubation with the anti-GATA-3 (1:80) and anti-NANOG antibodies (1:100) (R&D-Systems) overnight at 4°C. Then, embryos were incubated with secondary antibodies conjugated with Alexa Fluor 488 or Alexa Fluor 647 (1:500, Abcam) against GATA-3 and NANOG antibodies, respectively, for 1 h at room temperature. Fluorescent images were captured with a Leica TCS-SP8 confocal microscope.
Library construction and miRNA sequencing
Total RNA from the EVs (n = 3 pools) was extracted with the miRNeasy Mini Kit (Qiagen, USA). miRNA libraries were prepared using NEXTFLEX® Small RNA-Seq v3 for Illumina Platforms (Bio Scientific® Corporation, USA). For quality control, the concentrations and RNA integrity were validated using High Sensitivity DNA chips on an Agilent 2100 Bioanalyzer (Agilent Technologies, USA). Samples were sequenced on a NextSeq™ 550 system (Illumina, USA) with a single read design.
miRNA-seq data processing
miRNA-seq data were processed in R computing environment (version 4.1.3). Quality of the raw data was assessed by FastQC (Andrews, 2010). Raw data were processed to eliminate adapters and clean lecture sequences based on their size using the BBDuk package (Bushnell et al., 2017). miRNAs present in these samples were aligned using the Subread package (Liao et al., 2013), and annotated with miRbase, using the reference human genome HG38 (Kozomara et al., 2019). The number of reads per sample was normalized using the Trimmed mean of M-values (TMM) method in edgeR (Robinson et al., 2010), generating the counts per million (CPM). Sequencing of medium-length alignments (51 bp) yielded an average of 11.7 million reads per sample after preprocessing. All raw sequencing data are available through the Gene Expression Omnibus (GEO) under accession number GSE209589.
Selection of most expressed miRNAs
miRNAs detected in the three pools were selected to analyze miRNA cargo in EVs secreted by pHEECs. Venn diagram was constructed to illustrate the miRNAs in common between the three pools of EVs. Box plots were constructed with GraphPad Prism 6.0 to qualitatively assess the distribution and skewness of miRNAs based on the CPMs, by displaying the data quartiles and averages. The miRNAs common to all three pools with a mean CPM > 3702 (75th percentile), which showed a low coefficient of variation between the three pools, were considered relevant for subsequent analysis (Supplementary Table S2).
Target prediction and gene ontology analysis
The gene targets of the miRNAs with the highest expression (75–100 percentile), were predicted using the experimentally validated miRNA-target interactions database miRTarBase (Huang et al., 2020a). Gene Ontology (GO) enrichment analysis was carried out with the resulting targets via ShinyGO (version 0.76) (Ge et al., 2020). The 100 most significant biological processes with false discovery rate (FDR) < 0.05 were considered enriched functions regulated by the EVs’ miRNA cargo. Chord plot showing the relationship between GO functional groups and miRNAs was plotted by a bioinformatics online tool (www.bioinformatics.com.cn).
Integration with external datasets
miRNAs in common between the three EV pools were compared to previously published UF miRNA data (von Grothusen et al., 2022). von Grothusen et al. (2022) sequenced miRNAs present in the UF of healthy women as a control group, and patients with recurrent implantation failure, during the secretory phase of the menstrual cycle (LH + 7–9 days). Sequencing data were downloaded from GEO under accession number GSE173289. Only control group data from this study were used (n = 15), and miRNAs were considered if they were present in at least 80% of the samples. Venn diagram was constructed to illustrate the miRNAs in common between the two datasets.
Gene targets of the most expressed miRNAs (75–100 percentile) were compared to previously published data from single-cell RNA-seq analysis of TE and epiblast (EPI) cells from human embryos (Blakeley et al., 2015), which were downloaded from GEO under accession number GEO66507. Genes were considered if they were present in at least 75% of the samples. GO enrichment analysis was carried out with the miRNA targets in common with differential expressed genes (DEG) in TE and EPI via ShinyGO (version 0.76) (Ge et al., 2020). The 20 most significant biological processes with FDR < 0.05 were analyzed. Venn diagram was constructed to illustrate genes in common between the two datasets.
Statistical analysis
GraphPad Prism 6.0 was used for statistical analyses and graphics generation of qRT-PCR data, applying Student’s t-test. A P value of <0.05 was considered to be statistically significant. Bioinformatic analysis was performed using R computing environment (version 4.1.3). A descriptive statistical analysis of the common miRNAs found among all three pools was carried out with GraphPad Prism 6.0, to calculate the mean and standard deviation, minimum and maximum range, and quartiles of the CPMs.
Results
Isolation of primary human endometrial epithelial cells and their hormonal response in vitro
Pancytokeratin (epithelium) and vimentin (stroma) immunostaining confirmed successful isolation of pHEECs from endometrial biopsies. Pancytokeratin was expressed in cell cytoplasmic compartment, while vimentin was not expressed, corroborating the presence of epithelial cells and absence of stromal cells (Supplementary Fig. S1A). Results obtained from gene expression levels of epithelial hormone response markers in pHEECs by qRT-PCR showed the downregulation of PGRMC1 (fold change = 0.32; P = 0.0046) and upregulation of PGRMC2 (fold change = 1.86; P = 0.018), corroborating the acquisition of the secretory phenotype of these cells after hormonal treatment (Supplementary Fig. S1B).
Characterization of the EVs secreted by primary human endometrial epithelial cells
The EVs isolated from the culture media of pHEECs derived from fertile women were characterized by NTA, WB, and TEM (Fig. 1). The NTA revealed a size distribution between 100 and 300 nm, with a mean diameter of 213 ± 0.4 (SD) nm, and a concentration of 1.78 × 1011 ± 4.12 × 1010 (SD) particles/ml (Fig. 1A and B). Subsequent analysis by WB verified the expression of the intraluminal protein markers HSP70 and TSG101, as well as expression of the transmembrane protein markers CD9 and CD81, which were not detected in ultracentrifuged fresh culture medium (negative control), confirming the particles were indeed EVs (Fig. 1C). Finally, examination of the particles’ ultrastructure with TEM revealed their typical cup-shaped morphology, and structures within the expected EV-size distribution, corroborating the presence of EVs (Fig. 1D).
Characterization of the extracellular vesicles secreted by primary human endometrial epithelial cells. (A) Representative concentration/size graph obtained by nanoparticle tracking analysis (NTA) conducted on particles isolated from culture media of primary human endometrial epithelial cells. (B) NTA results summary table. Data are presented as mean of two biological replicates ± standard deviation (SD). (C) Representative image of western blot results of evaluated extracellular vesicle (EV) markers revealing protein expression of Heat shock protein 70 (HSP70) (70 kDa), Tumor Susceptibility Gene 101 (TSG101) (47 kDa), CD9 (24 kDa), and CD81 (22 kDa), compared to β-actin (43 kDa). No signal was detected in ultracentrifuged fresh culture medium. (D) Representative transmission electron microscopy image showing the morphology and size range of EVs. Scale bar set to 200 nm.
Human embryos efficiently internalize endometrial EVs
To verify whether the EVs are taken up by human embryos in vitro, EVs were fluorescently labeled using Bodipy-TR ceramide. The NTA showed that out of total isolated particles concentration (1.96 × 1010 ± 3.1 × 108 (SD) particles/ml), 1.51 × 109 ± 1.6 × 108 particles/ml were fluorescently labeled, 7.7% of total isolated particle concentration (Fig. 2A and B). When considering both labeled and nonlabeled EVs detected with light scattering, the most frequent size observed was 151 ± 3.8 (SD) nm, whereas when exclusively focusing on labeled EVs, most frequent size observed was 289.9 ± 20.6 nm.
Extracellular vesicle labeling and uptake by human embryos. The graphs in (A) represent the averaged particle concentration/size obtained by nanoparticle tracking analysis (NTA), using light scattering for all nanoparticles (top) and a 532-nm laser for labeled nanoparticles (bottom). (B) Summary table of the size distribution and concentration data. Data are presented as mean of two replicates ± standard deviation (SD). (C) Confocal images of human blastocysts (whose cells were fluorescently labeled with 4′,6-diamidino-2-phenylindole (DAPI, blue)) that were cultured with extracellular vesicles (fluorescently labeled with Bodipy-TR Ceramide (red)) for 1 or 2 h. Human blastocysts cultured with PBS labeled with Bodipy-TR, or simple culture medium (without labeled extracellular vesicles) served as technical control groups. BF, bright field; EVs, extracellular vesicles. Scale bars set to 58 µm.
Human blastocysts were all hatching and B or C quality based on ICM and TE characteristics (Supplementary Fig. S2), and efficiently internalized the fluorescently labeled EVs present in the culture medium within the first 1–2 h of in vitro culture, as observed by fluorescent microscopy (Fig. 2C, Supplementary Fig. S3, and Supplementary Videos S1, S2, S3, and S4). During the first hour, EVs were predominantly internalized by the cells from the hatched portion of the embryo, probably because these cells were more exposed to the extracellular medium, while a more dispersed internalization was achieved throughout the whole embryo within 2 h. No fluorescence was observed when the blastocysts were cultured with labeled PBS or the basic culture medium, validating that the fluorescence we detected indeed represented the internalized EVs.
Immunofluorescence labeling of GATA-3 and NANOG in blastocysts from each condition corroborated the presence of both epiblast and trophoblast healthy cell lineages in all embryos, validating the integrity of these blastocysts in the absence or presence of endometrial EVs (Supplementary Fig. S4).
Functional implications of the miRNA cargo of endometrial EVs
miRNA-seq analysis was carried out to characterize the miRNA cargo of the EVs secreted by pHEECs; quality data of the sequencing is shown in Supplementary Fig. S5. Deep sequencing respectively, revealed 371, 231, and 201 annotated miRNAs in the three pools of EVs, with 149 miRNAs in common between them (Fig. 3A). Further, constant expression levels were observed among the three groups of miRNAs (Fig. 3B). Finally, a descriptive statistical analysis was carried out on the CPM values to establish the most relevant miRNAs based on their expression levels (Fig. 3C). A list of the shared 149 miRNAs and their mean CPM is presented in Supplementary Table S3. Target prediction analysis and functional enrichment was performed using the 37 miRNAs with the highest expression levels that correspond to the 75–100 percentiles. These miRNAs are known to target 6592 genes and functional enrichment analysis indicated that these participate in several processes related to embryo development and oxygen metabolism (Fig. 4A), cell cycle, differentiation, and apoptosis (Fig. 4B), cellular metabolism (Fig. 4C), cellular organization, and gene expression (Fig. 4D). Other affected biological processes included cell growth, histone modification, and response to hormones (Supplementary Fig. S6), which are implicated in implantation process and early embryo development.
Analysis of the miRNA cargo carried by the extracellular vesicles secreted by primary human endometrial epithelial cells. (A) Venn diagram depicting the relations of the miRNAs detected in each pool of extracellular vesicle (EV) isolation and microRNA (miRNAs) in common. (B) Boxplot analysis of the 149 shared miRNAs. Data is presented as log2 transformed counts per million (CPM). Box is drawn from quartile 1 to quartile 3 with median in the middle and whisker ends are minimum and maximum values of the data set. (C) Descriptive statistics of the miRNA expression (in CPM) for each of the pools and mean values for each measure.
Enrichment analysis of the predicted gene targets of the extracellular vesicles’ miRNA cargo. Enriched gene ontology (GO) biological processes were categorized under (A) embryo development and oxygen metabolism, (B) cell cycle, cell differentiation, and apoptosis, (C) cellular metabolism, and (D) cellular organization and gene expression functions. Processes are ordered according to their significance. The color intensity represents the number of predicted target genes (related with the 37 most expressed miRNAs) associated with each process. (E) Master miRNA regulators related to each category of biological processes. Horizontal axis shows FDR: false discovery rate; count, number of target genes enriched in each biological process.
The top miRNAs regulators of these biological processes are depicted in Fig. 4E. We highlight hsa-miR-30c-5p, involved in embryo development, cell differentiation, apoptosis, and cellular organization; hsa-let-7c-5p, related with cell differentiation and oxygen metabolism; hsa-miR-483-5p, related with embryo development; hsa-miR-26a-5p, related with apoptosis; hsa-miR-25-3p, related with cellular metabolism; and hsa-miR-23a-3p, related with gene expression. Moreover, we highlight hsa-miR-92a-3p, hsa-let-7b-5p, hsa-miR-30a-5p, hsa-miR-24-3p, hsa-miR-21-5p, and hsa-let-7a-5p as master regulators of all these functions, while hsa-let-7e-5p and hsa-miR-320a were related to all the aforementioned processes except embryo development and apoptosis, respectively.
Endometrial epithelial EV miRNAs are found in the UF during secretory phase
To assess the in vivo relevance of EVs’ miRNA cargo described, this was compared with miRNAs previously identified in the UF. Out of the 210 miRNAs detected in at least the 80% of the UF samples of the study, 102 were present in EVs secreted by pHEECs, this being 48.6% of total miRNAs in the UF. Interestingly, 47 miRNAs were exclusively found in these vesicles and 108 exclusively found in the UF (Fig. 5A and Supplementary Table S4). Among the 102 miRNAs in common between the two different datasets, 27 of them were found within the 37 most expressed miRNAs in pHEECs’ EVs, being this a 72.9% (Fig. 5B).
Integration of primary human endometrial epithelial cells extracellular vesicle microRNA data with external datasets. (A) Venn diagram showing miRNAs in common between EVs secreted by primary human endometrial epithelial cells and miRNAs present in the uterine fluid (UF), both from healthy women, and (B) list of miRNAs in common from the top 37 miRNAs present in these EVs. Venn diagrams showing genes targeted by pHEECs’ EV miRNAs which are enriched in (C) trophectoderm cells and (D) epiblast cells of human blastocysts, and functional groups enriched by these targeted genes in each cell lineage. EVs, extracellular vesicles; miRNA, microRNA; pHEECs, primary human endometrial epithelial cells; UF, uterine fluid.
In silico validation of target genes reveals endometrial epithelial EV miRNAs role in human blastocysts
To further assess the in vivo relevance of endometrial epithelial EV uptake by human blastocysts, 6592 target genes from the top 37 miRNAs previously described were compared with genes previously identified in TE and EPI cells. Out of the 9359 total genes identified in at least 75% of the TE cells analyzed, 3970 genes were targeted by miRNA cargo of endometrial EVs; regarding EPI cells, out of 9370 total genes identified, 3818 were targeted by these miRNAs. Further analysis comparing target genes from the top 37 miRNAs with DEG in TE, 137 DEG in TE cells were targeted by endometrial EV miRNAs, this being 40.7% of total TE DEG identified; these genes were related to biological functions assigned to different functional groups, such as development, morphogenesis, cell migration and motility, cellular secretion, cell adhesion and cytoskeleton organization (Fig. 5C and Supplementary Table S5). Comparison of target genes from the top 37 miRNAs with EPI DEG identified 169 DEG in epiblast cells which were targeted by endometrial EV miRNAs, being this a 33.8% of total EPI DEG; these genes were related to biological functions assigned to different functional groups, such as cell differentiation, development, morphogenesis, proliferation, cellular response and signaling, and cell migration (Fig. 5D and Supplementary Table S5).
Discussion
The potential role of EVs in embryo implantation has been recently reported, suggesting the existence of a communication system between the maternal endometrium and the embryo via these vesicles (Bridi et al., 2020; Mishra et al., 2021). This communication may be crucial for establishing a successful pregnancy and is presumably altered in some reproductive pathologies (Machtinger et al., 2016; Simon et al., 2018). Indeed, EVs secreted by human endometrial epithelial cells contain proteins related to endometrial receptivity, embryo implantation, and early embryo development (Segura-Benítez et al., 2022), which may alter the phenotype of human embryos during implantation. However, little is known about the endometrial EV-mediated transcriptomic regulation of these biological processes. In this study, we demonstrated that EVs secreted by the maternal endometrium are efficiently and rapidly taken up by human embryos, and characterized the miRNA cargo of these EVs, most of which had been previously related with implantation and/or early embryo development.
First, pHEECs were successfully isolated from endometrial biopsies, as showed by the double immunofluorescence staining that was positive for pancytokeratin and negative for vimentin; cells were hormonally treated to achieve a responsive cell-stage, as corroborated by the downregulation of PGRMC1 and upregulation of PGRMC2 (Medina-Laver et al., 2021). We found that the nanoparticles isolated from the pHEEC culture medium expressed the transmembrane protein markers CD9 and CD81, corroborating the lipid bilayer structure of EVs; and intraluminal protein markers HSP70 and TSG101, substantiating their integrity (Théry et al., 2018), confirmed that pHEECs secrete EVs. Size distribution analysis showed that pHEECs secrete both small (<200 nm) and medium/large EVs (>200 nm), in accordance with the classification by the International Society for Extracellular Vesicles (Théry et al., 2018), and these EVs showed the typical cup-shaped morphology. The fluorescent labeling of the EVs corresponded to 7.7% of total isolated particle concentration, which was more efficient in medium/large EVs, which could be due to the differences in the origin and biogenesis of exosomes (30–150 nm) versus microvesicles (100–1000 nm) (Abels and Breakefield, 2016). This percentage of fluorescent labeling could be a technical limitation that needs to be overcome for future studies in which a high number of labeled EVs is required.
Subsequently, we demonstrated that these EVs are efficiently internalized by human blastocysts within 1–2 h of in vitro culture, being the internalization achieved by the whole embryo after 2 h. This suggests that EVs could be able to go through the zona pellucida (ZP), as other studies have previously suggested (Vyas et al., 2019), although further studies would be needed to corroborate it. Notably, this is the first study to track the active internalization of endometrial EVs by human blastocysts, as has previously been described in trophoblastic cell lines (Ruiz-González et al., 2015; Greening et al., 2016; Evans et al., 2019), and murine embryos (Vilella et al., 2015; Liu et al., 2020). Fluorescent labeling of GATA-3 and NANOG confirmed the presence of healthy trophoblast and epiblast lineages in blastocysts cultured in the presence or absence of vesicles, corroborating the integrity of these embryos in the presence of endometrial EVs. Based on these findings, further studies to evidence whether EV uptake is specific to one type of embryonic lineage would extend the knowledge to assess embryonic quality.
The implications of embryos internalizing EVs depends on their cargo, which when released within the embryonic cells, can modify gene expression that may affect adhesive and invasive capacities during implantation (Greening et al., 2016; Liu et al., 2020). As miRNAs have previously been described as mediators of endometrial receptivity and embryo implantation (Tan et al., 2020a), we aimed to study the miRNA content of the EVs to evaluate their potential role in EV-mediated communication between the maternal endometrium and human blastocyst. This study described 149 miRNAs in endometrial epithelial EVs, most of them previously reported in UF of healthy women during secretory phase of menstrual cycle (von Grothusen et al., 2022). Remarkably, the miRNAs identified in this study were enriched in functions related to embryo development and oxygen metabolism, which supports the transition to aerobic metabolism that occurs during early embryo development (Ottosen et al., 2007), cell cycle, differentiation, or apoptosis, which are crucial during early embryo development, in addition to cellular metabolism, cellular organization, and gene expression. These miRNAs target several genes expressed in human blastocysts, and more than one third of the genes differentially expressed between epiblast and trophoblast, highlighting the role of endometrial EVs in the different cell lineages of the early embryo and the importance of their uptake by the whole blastocyst (Blakeley et al., 2015).
Deep sequencing of the miRNAs carried by the pHEEC-secreted EVs, revealed miRNAs that were previously described in the endometrium and endometrial fluid, as mediators of embryo implantation. Specifically, we identified members of the let-7 family who have previously been associated to embryo implantation and promote the endometrial epithelium’s acquisition of an adhesion-competent state during the window of implantation by altering the expression of anti-adhesive components (Inyawilert et al., 2015). In fact, these adhesion-associated functions might explain why downregulation of some let-7 family members such as let-7a, let-7f-5p, let-7g-5p, let-7e-5p, and let-7d-5p can lead to spontaneous miscarriages (Wang et al., 2016). Further, expression of the let-7 family is enhanced in the early pregnancy decidua compared to menstrual endometrium (Wang et al., 2016). A recent study suggested that let-7a and let-7g enhance endometrial receptivity via the Wnt/β-catenin signaling pathway, which increases uterine receptivity along with adhesion of the embryo and endometrial cells (Li et al., 2020; Shekibi et al., 2022). While let-7a influences the implantation competency of activated blastocysts and mediates blastocyst development by regulating Toll Like Receptor 4 (TLR4) gene expression in mice (Hosseini et al., 2020), the specific roles played by the let-7 family in human embryo communication with the endometrium require further investigation.
We also identified miR-30 family members contained in EVs, such as miR-30a-5p, miR-30a-3p, miR-30b-5p, miR-30c-3p, miR-30c-5p, miR30d-5p, miR-30e-3p, and miR-30e-5p. Similar to the let-7 family, the miR-30 family members have been associated with endometrial receptivity. In particular, miR-30d expression is suppressed in the prereceptive endometrium of fertile women compared to receptive endometrium, decidua of endometrium following abortion, and endometrium of recurrent implantation failure patients (Altmäe et al., 2013; Grasso et al., 2020; Zhao et al., 2021). In addition to its role in remodeling the endometrium, miR-30d was also associated to embryo development. Indeed, following internalization by mouse blastocysts in vitro, miR-30d modified the embryonic transcriptome and phenotype (Vilella et al., 2015). Similarly, the miR-30d contained in EVs secreted by human endometrium might regulate genes related to cell adhesion, integrin-mediated signaling pathways, and developmental maturation in the embryo (Balaguer et al., 2019).
Other miRNAs carried by the EVs secreted by human endometrium, such as miR-21-5p, miR-29a-3p, miR-10a-5p, miR-10b-5p, miR-27b-3p, and miR-320a-3p, are also involved in endometrial receptivity. Particularly, miR-21-5p is highly expressed in the murine uterine lumen during their receptive window, and its knockdown results in implantation failure (Hua et al., 2020). Alternatively, miR-29a is scarce during the pre- and postimplantation periods in rats, but markedly increased during implantation (Xia et al., 2014). In addition, the presence of miR-10a, miR-10b, miR-92a-3p, and miR-27b has been described as especially relevant during the receptive phase of the human endometrium (Nikolova et al., 2021). Moreover, miR-320a has been proposed as a driver of endometrial response during implantation (by stimulating the migration of decidualized stromal cells) and is exclusively secreted by high-quality embryos (Berkhout et al., 2020). These results suggest that endometrial epithelial EVs can be taken up by the endometrium itself, regulating endometrial receptivity, supporting other studies that demonstrated EVs secreted by the endometrium can be internalized by nearby endometrial cells to affect the uterine microenvironment in an autocrine/paracrine fashion (Harp et al., 2016; Chen et al., 2020). Therefore, these findings open insights for further research to demonstrate an autocrine and/or juxtracrine effect in the endometrium of the EVs secreted by the maternal endometrium that could be involved in endometrial receptivity regulation, required for a successful implantation.
Regarding miRNAs that are directly related with embryo attachment, we identified miR-183-5p and miR-182-5p in the EVs, which promote migration and proliferation of endometrial cells. Indeed, experiments with JAr (choriocarcinoma cell line) spheroids showed that attachment rates were significantly decreased following inhibition of miR-183-5p and miR-182-5p, and alternatively, treatment with these miRNAs enhanced attachment via repression of CTNNA2 and consequent upregulation of the Wnt pathway (Akbar et al., 2020). We additionally identified miR-23a-3p and miR-221-3p in EVs, corroborating that the overexpression of miR-23a-3p promotes JAr spheroid attachments, and reinforcing the benefits of miR-23a-3p in endometrial receptivity and embryo implantation via downregulation of CUL3 gene and consequent upregulation of Wnt pathway (Huang et al., 2020b). On the other hand, direct transfection of miR-221-3p into trophoblasts (i.e. HTR-8/SVneo cells) has promoted trophoblast cell growth, invasion, and migration (Yang et al., 2019), laying foundations for early embryo development and subsequent implantation.
Finally, some of the microRNAs that were broadly expressed in our EVs have been linked directly to early embryo development. For example, miR-200c-3p and the aforementioned miR-23a-3p, let-7b-5p, and miR-92a-3p, were significantly upregulated in murine outgrowth embryos compared to nonoutgrowth blastocysts, implying these miRNAs could be involved in embryo attachment and embryonic-endometrial interaction during implantation (Kim et al., 2019). Furthermore, miR-24 overexpression has been associated with improved embryo development and reduced abortion rates in mice through reduces of decidual tissue apoptosis and increases vascular endothelial grown factor expression by inhibiting Caudal Type Homeobox 1 (CDX1) (Wang et al., 2020).
Although this study has analyzed EVs secreted by endometrium that has not been directly co-cultured with human embryos, this approach allowed us to demonstrate that the EVs secreted by the endometrial epithelium are internalized by human embryos, as well as analyze their miRNA cargo, which could be related to the implantation process by regulating functions related to endometrial receptivity, embryo attachment, and early embryo development. This study opens insights for further research to describe potential biomarkers of endometrial receptivity and embryo competence, that may be useful diagnostic and therapeutic targets for implantation success. In this regard, further investigation that elucidates epithelial EVs roles in the regulation of molecular mechanisms involved in the embryo-endometrium crosstalk necessary for a successful implantation would be essential.
Data availability
The data underlying this article are available in the Gene Expression Omnibus (GEO) at www.ncbi.nlm.nih.gov/geo/, and can be accessed with accession number GSE209589.
Acknowledgements
The authors are grateful to the participants of this study, who made this work possible, and to all the medical staff of the IVI Valencia Clinic for their assistance in obtaining samples.
Authors’ roles
M.S.-B. was involved in study design, executed experiments, and wrote and edited the manuscript. A.B.-R was involved in experimental execution and wrote the manuscript. E.J.-B. and M.C.C.-G. participated in EVs isolation and analyzed results from miRNA-seq. A.F. and M.J.D.L.S. was involved in sample collection of endometrial biopsies and embryos. A.P. devised and supervised the study, contributed to data interpretation, and drafted the manuscript. H.F. coordinated the study design, contributed to data interpretation, and edited the manuscript. All authors reviewed the manuscript and provided critical feedback and discussion.
Funding
This study was supported by the Spanish Ministry of Education through FPU awarded to M.S.-B. (FPU18/03735), the Health Institute Carlos III awarded to E.J.-B. (FI19/00110) and awarded to H.F. by the Miguel Servet Program ‘Fondo Social Europeo «El FSE invierte en tu futuro»’ (CP20/00120), and Generalitat Valenciana through VALi+d Programme awarded to M.C.C.-G. (ACIF/2019/139).
Conflict of interest
The authors have no conflicts of interest to disclose.
References
Author notes
Marina Segura-Benítez and Alba Bas-Rivas consider that the first two authors should be regarded as joint First Authors.
