Extracellular Vesicles, Lipid Droplets and AhR Ligands in Early Implantation: The Dynamics of Embryo‐Maternal Crosstalk

ABSTRACT

Intercellular communication is essential for healthy embryo development, yet the role and dynamics of extracellular space in the maternal‐embryonic dialogue remain unclear. Furthermore, little is known of maternal and embryonic metabolic states during early endometrial preparation and after the embryo enters the uterine cavity. Using a human in vitro co‐culture model and extracellular vesicle (EV)‐specific tools, we dynamically tracked EV secretion, uptake and processing between embryonic and endometrial cells at early stages of cell‐to‐cell communication. Hormonal stimulation altered endometrial secretory output, producing distinct EV populations. Stimulated EVs (St‐EVs) differed from non‐stimulated EVs (NSt‐EVs) in size, secretion dynamics, uptake efficiency and metabolic cargo, selectively packaging energy‐related metabolites and aryl hydrocarbon receptor (AhR) ligands. AhR inhibition increased spheroid attachment, suggesting that AhR signalling regulates implantation by modulating the endometrial environment. Additionally, lipid droplets (LDs) affected by endometrial‐ and embryo‐derived EVs were actively secreted and taken up by embryonic cells, highlighting their role in implantation. EVs were not only exchanged between the embryo and endometrium but were also rapidly internalized, influencing mitochondrial activity, lipid metabolism and extracellular matrix remodelling. Translation of EV‐derived mRNA occurred within 1 h of uptake, driving cellular changes and enhancing embryo attachment. These findings suggest EVs, extracellular metabolites and LDs mobilized between the endometrium and embryo coordinate to promote embryo attachment and implantation. This study advances our understanding of embryo‐maternal EV‐mediated communication and provides a valuable model for investigating EV‐mediated simultaneous intercellular bidirectional crosstalk in other biological contexts.

Hormonal stimulation induces lipid droplet (LD) production, shifts extracellular vesicle (EV) trafficking dynamics and alters metabolic cargo, selectively loading them with aryl hydrocarbon receptor (AhR) ligands and energy‐related metabolites. Simultaneous bidirectional EV exchange and utilization between endometrial cells and spheroids induces changes in mitochondrial function, LD secretion and uptake, as well as extracellular matrix (ECM) remodelling. These processes culminate in the initiation of a close endometrial‐spheroid interaction, a key step in establishing the foundation for implantation.

1. Introduction

The development of healthy offspring involves a sequence of molecular dialogues and relies upon communication among different cell types, tissue‐level circuits and organs working together. Processes such as the trafficking of signals (e.g., RNA, DNA, proteins, metabolites, hormones, growth factors) in and out of the extracellular space, where many physiological events unfold simultaneously, are crucial for successful conception. The ability of cells to sense the extracellular environment and communicate with adjacent and distant cells plays a significant role in health and disease (Van Niel et al. 2018; Simon et al. 2018; Andronico et al. 2019; Valls and Esposito 2022), yet we do not fully understand the dynamics and role of extracellular space in intercellular communication or, specifically, the maternal‐embryonic dialogue.

In humans, the probability of natural conception is relatively low (∼30%); two‐thirds of conception failures can be attributed to a failure of embryo implantation (ASRM 2022). The extracellular environment of the female reproductive system, where the embryo develops and later implants, plays a significant role in successful pregnancy (Greening et al. 2016; Kaczmarek et al. 2021; Poh et al. 2021; Sehring et al. 2022). Moreover, embryos are sensitive to the composition of the extracellular environment, and changes in that environment during early embryonic development can affect future growth and developmental potential, both pre‐ and postnatally (Lamberto et al. 2021; Fleming et al. 2004). Studies have shown that the extracellular space/environment contains extracellular vesicles (EVs), critical determinants of cell function in all tissues. EVs, a heterogeneous population of secreted vesicles, modulate the response of recipient cells by delivering genetic material and other molecules such as metabolites, proteins and lipids (Van Niel et al. 2018; El Andaloussi et al. 2013). They affect both adjacent and distant cells and are involved in intercellular regulation of metabolism, transcriptional gene expression and other processes essential for cell function (Ando et al. 2017; Buzas 2023; Dixson et al. 2023; Giacomini et al. 2017; Raposo and Stahl 2023; Song and Testa 2018).

Once the embryo has reached the uterine cavity, successful implantation requires that the endometrium and a properly developed embryo be receptive to bidirectional signals. Only a well‐orchestrated molecular dialogue between the two sides will lead to attachment, adhesion, invasion of the endometrium and formation of the placenta. Any alterations in the development of the embryo and endometrium and/or their secretory output can lead to desynchronization of embryo‐maternal communication, impeding implantation (Carson et al. 2000; Coughlan et al. 2014; Simon and Laufer 2012).

Embryo‐maternal communication has been extensively studied using various mammalian and human models; however, most studies rely on static time‐point sampling, limiting insights into real‐time, dynamic interactions. Fatmous et al. 2022 and Segura‐Benítez et al. 2023 have shown that human trophoblast stem cell lines and human blastocysts take up small EVs derived from endometrial epithelial cells within 2 h of exposure. Additionally, JAR trophoblast‐derived EVs have been shown to induce unique transcriptomic alterations in RL95 endometrial cells within 30 min post‐exposure to spheroid EVs (Godakumara et al. 2021), indicating early initiation of the dialogue and its associated molecular responses; these plausibly begin immediately upon the arrival of the embryo in the uterine cavity. However, little is known of early embryo‐maternal communication once the embryo enters the uterine cavity, including the dynamics of the early bi‐directional communication mediated by EVs. Moreover, it is not known whether hormonal stimulation changes the combination of metabolites carried by endometrial EVs or mobilized endometrial lipid droplets (LDs) to the embryo, if metabolites carried by EVs in the extracellular uterine space provide energy needed by the embryo for specific functions supporting implantation, or whether they are also necessary for biosynthesis of cellular constituents or processes essential for initiating implantation.

Using a human in vitro cell line model and EV‐specific tools, we show that the hormonally stimulated endometrium‐derived EV population (St‐EVs) differs in size, dynamics of secretion, uptake and metabolic cargo from the non‐stimulated population of EVs (NSt‐EVs). Interestingly, St‐EVs are enriched with aryl hydrocarbon receptor (AhR) ligands, particularly indoles; our findings show that St‐EVs deliver metabolic cargo enriched with metabolites related to energy production, hormonal regulation and cell‐to‐cell communication. Examination of the extracellular environment outside EVs allowed us to assess metabolites that are secreted from cells, not necessarily incorporated into EVs, but essential for intercellular communication, metabolic support and the modulation of recipient cell function. Additionally, LDs, which are influenced by both endometrial‐ and embryo‐derived EVs, were actively secreted and taken up by embryonic cells, highlighting their role in early implantation.

These results provide a deeper understanding of the molecular events underlying successful implantation and will guide research on in vivo enhancement of implantation and promotion of successful pregnancies.

2. Methods

See Additional Information for detailed material and methods, and descriptions of all procedures.

3. Results

3.1. Hormonal Stimulation and Early Spheroid‐Endometrial Crosstalk Alter Endometrial Metabolism

The endometrium undergoes cyclic changes governed by ovarian steroids, estrogen and progesterone. In humans, endometrial receptivity lasts about 5 days (window of implantation, WOI, Figure 1a) (Greening et al. 2016; Sehring et al. 2022; Carson et al. 2000; Bagchi et al. 2001; Martín et al. 2002). Implantation involves structural and molecular remodelling of the embryo and the endometrium; these processes require energy to maintain intracellular homeostasis, deliver metabolites for biosynthesis of cellular components, regulate cell signalling and to traffic components outside the cells (Sehring et al. 2022; Leese 2012). Little is known, however, about the changes in metabolism essential for progress towards successful implantation that occur in the embryo and endometrium once the embryo reaches the uterine cavity during WOI.

FIGURE 1.

Hormonal stimulation modulates changes in metabolism in endometrial cells early post‐progesterone addition. (a) Scheme illustrating the hormonal and physiological transitions in the human endometrium during the menstrual cycle in vivo, alongside the in vitro hormonal stimulation protocol; estrogen (E2; purple), progesterone (P4; blue). The in vitro chart details the E2 and P4 stimulation protocol we established to mimic the stimulated state (St) versus non‐stimulated state (NSt) of endometrial cells (RL95‐2) in culture. (b) Differences between NSt and St endometrial cells as a function of lipid droplets (LDs) number per cell; n = 3 for NSt and St; (c) Representative confocal images of the NSt and St cells stained with LipidSpot (Red) to visualize LDs and Hoechst 33342 for nuclei. Scale bar = 30 and 10 µm in insets; arrowhead, LD spots. (d) Schematic representation of LD spots analyzed within the spheroids. (e) Quantification of the spot number and (f) volume in co‐culture with NSt or St cells during the first 2 h of incubation; p < 0.001, linear GLM; ranking by time is shown in Supplementary Data 1; n = 3 biological replicates. (g) Representative images of LD transfer (Red) during 2 h co‐culture of Palm‐GFP tagged spheroid and St and NSt endometrial cells, illustrate the transfer of lipid droplets (LDs) to the spheroid at all‐time points examined, arrowhead, LD spot. The inserts show the top layers of the spheroid. Scale bars = 10 µm. *=p < 0.001, linear GEE; n = 3 biological replicates. (h) Schematic representation of the fluorescent NRF1mitoGFP sensor; (i) quantification of NRF1mitoGFP sensor fluorescence intensity in St and NSt cells, and (j) in co‐culture with spheroids, during 4 h after P4 or spheroid addition; *=p < 0.001, linear GEE; n = 3 biological replicates. AU—arbitrary units. Results presented as mean ± standard error*1.96. Latin letters in b, e, i, f, j presents NSt versus St significant mean difference, “a” for the lowest difference. Ranking by time and statistics, see Supplementary Data 1. (a), (d) and (h) were created with Biorender.com.

To investigate changes in metabolism during the early molecular dialogue between embryo and endometrium, we established a hormonal stimulation protocol modified from Stern‐Tal et al. 2020 and Singh et al. 2010. RL95‐2 endometrial epithelial cells were stimulated with estrogen for three days, after which progesterone was added to the culture to simulate the secretory phase (hereafter, stimulated [St] endometrial cells). Examining the expression of specific secretory phase markers (progesterone receptor membrane components 1 and 2 [PGRMC1, PGRMC2]) following hormonal treatment showed that the hormonal stimulation regime used in this study effectively mimics the secretory phase (Figure S1a). To study embryo‐endometrial interactions, we created multicellular‐3D spheroids from trophoblast JAR cells (hereafter, spheroids, Sph) and co‐cultured them with endometrial epithelial cells (Figure 1a). We first focused on lipid droplets (LDs), which have been shown to play a key role in lipid and energy homeostasis. LD production has been found to change during the menstrual cycle, suggesting that LDs may be involved in endometrial receptivity (Aizawa et al. 2019; Mau et al. 2022). To examine whether endometrial LD production undergoes changes following addition of progesterone, we stained the endometrial cells with LipidSpot, a specific LD dye (Figure 1a–c), and analyzed the number and size of LDs in non‐stimulated (NSt) and St endometrial cells. Significantly more St endometrial cells were found to contain 2‐8 LDs per cell, as compared with NSt cells. Furthermore, the LD area in St was significantly larger than that of NSt (p < 0.001, Figure 1b and Figure S1a,b, respectively). These results indicate that hormonal stimulation induces changes in LD production and/or storage in endometrial cells.

To determine whether spheroid presence can further affect the LD population in endometrial cells and whether spheroids can take up endometrial LDs, we generated Palm‐GFP‐tagged spheroids (Zaborowski et al. 2015; Lai et al. 2015) and used them to label spheroid membranes. Endometrial cells were stained with LipidSpot and co‐cultured with GFP spheroids for up to 2 h (Figure 1a,d–g). We detected a significant increase in LD fluorescence intensity in the spheroid during the first hour after spheroid addition to St endometrial cells, followed by a decrease in fluorescence at 2 h compared to NSt cells (p < 0.05, Figure 1e). This suggests rapid secretion from endometrial cells, uptake and fast processing of St‐derived LDs by the spheroid. At 1 h, LD volume is significantly greater in spheroids exposed to St cells than at 2 h; this contrasts with the volume of LDs in spheroids exposed to NSt cells, the size of which increased during the 2 h following exposure (p < 0.05, Figure 1f). It is possible that as the St‐derived LDs are processed, resulting in a decrease in the observed fluorescence and volume of LD, the NSt‐derived LDs are not rapidly used and thus accumulate; this is expressed as an increase in LD volume. To verify the localization of LD spots inside the spheroid, we measured the distance from the spheroid centre to its surface. St‐derived LDs are situated significantly further from the spheroid centre and closer to the surface than are NSt‐derived LDs (p < 0.05, Figure S2a). St‐derived LDs may be processed in spheroid layers nearest to the endometrium, potentially mediating signalling and energy demands for attachment and implantation. Our findings suggest early LD‐mediated communication between St endometrial cells and the embryo during implantation. In recent studies, LD secretion and uptake by recipient cells (Amarasinghe et al. 2023) have been reported. This supports the role of LDs in endometrium‐embryo intercellular communication.

In light of the role of LD‐mitochondria interactions in energy metabolism (Veliova et al. 2020), we generated endometrial cells expressing nuclear respiratory factor 1 (NRF1)‐mito‐GFP sensor, a reporter construct designed to monitor the transcriptional activity of NRF1, which is a key regulator of mitochondrial biogenesis (Figure 1h). This sensor expresses GFP regulated by NRF1‐responsive elements, coupled with a mitochondrial targeting sequence that localizes GFP within mitochondria when NRF1 is active (Hodneland Nilsson et al. 2015). GFP intensity in NSt and St endometrial cells expressing the NRF1mitoGFP sensor was monitored for 4 h following the addition of progesterone; GFP fluorescence in St increased during the first 45 min following the addition of progesterone, to a greater extent than in NSt cells, suggesting a rapid mitochondrial response to hormonal stimulation. Elevated GFP fluorescence in St endometrial cells continued for up to 2 h (p < 0.05, Figure 1i; Supplementary Data 1); subsequently, up to 24 h after exposure, the differences between NSt and St cells were minimal (Figure S1c). Spheroid presence had no effect on NRF1mitoGFP levels in endometrial cells during the 4 h of observation (Figure 1j).

Mitochondria rely on changes in membrane potential to produce energy. We therefore assessed changes in mitochondrial activity by measuring mitochondrial polarization using JC‐1 dye, which enables visualization of changes in mitochondrial membrane potential (Sivandzade et al. 2019). NSt and St endometrial cells were stained and monitored for 4 h following the addition of progesterone. No differences in mitochondrial polarization were observed between the NSt and St cells (Figure S1d–f), suggesting that hormonal stimulation affects mitochondrial biogenesis but does not immediately impact mitochondrial activity. This further highlights the complexity of mitochondrial regulation, in which changes in gene expression can occur without immediate effects on membrane polarization.

Hormonal stimulation and the presence of the embryo also modulate early specific structural changes in endometrial cells (Supplementary Text; Figure S3). As a whole, these results show the subtle and intricate metabolic and morphological processes required for initiation of implantation.

3.2. Endometrial and Spheroid Cells Alter Their EV Secretory and Internalization Patterns in Response to Mutual Interactions

Our findings suggest that post‐hormonal stimulation and early contact between the endometrial epithelium and trophoblast spheroids modulate the early critical implantation dialogue. The dynamics of interactions during embryo attachment remain, however, poorly understood. To investigate the effects of early mutual interactions on secretory patterns, we employed a pH‐sensitive EV secretion reporter (pHluo_M153R‐CD63‐mScarlet; Figure 2a,b) for real‐time EV secretion monitoring. This sensor allows tracking of small EVs (80–160 nm) and intracellular multivesicular bodies (MVBs) (Sung et al. 2020). This sensor enables detection of the transition from an acidic to a neutral environment (i.e., flashes) that occurs when late endosomes or MVBs fuse with the plasma membrane and EVs are released to the extracellular environment (Figure 2b) (Sung et al. 2020; Verweij et al. 2018).

FIGURE 2.

Endometrium‐spheroid interactions induce changes in EV secretory and internalisation patterns. (a) Schematic representation of the experimental timeline and (b) pHluo_M153R‐CD63‐mScarlet reporter. (c) Quantification of pHluo‐CD63‐mScarlet fluorescent puncta inside and outside (secreted) St endometrial cells in the presence or absence of spheroid (blue and yellow dashed lines for 95% confidence intervals inside and outside the cells, respectively, as reference; blue spots outside the cells, without spheroid versus yellow spots with spheroid); GFP in the cells, p < 0.0001, in medium p = 0.086. (d) Quantification of pHluo‐CD63‐mScarlet fluorescent puncta inside and outside (secreted) spheroid in the presence or absence of St endometrial cells (blue and yellow dashed lines for 95% confidence intervals inside and outside the cells, respectively, as reference; blue spots, without St cells versus yellow spots with St cells); GFP in the medium and cells, p < 0.0001, Two independent sample t‐test; n = 3; still confocal images from Movies S1 (upper panel) and S2 (lower panel) showing pHluo‐CD63‐GFP and mScarlet puncta inside and outside endometrial cells in the presence of spheroids, and inside and outside the spheroid, respectively. Dashed white line presents the localization of the spheroid (Sph); scale bars = 5 µm. (e) Schematics presents the PKH experimental timeline. (f) Mean fluorescence intensity level of Spheroid‐derived EVs stained with PKH in NSt and St endometrial cell, 24 h post P4 stimulation; p < 0.006. (g) Volume of PKH spheroid puncta detected inside NSt and St cells during the 4 h examined; p < 0.001, Two‐way ANOVA; n = 4 biological replicates. (h) Representative confocal images of PKH stained spheroid EVs (red puncta) inside endometrial cells, and PKH stained St‐EVs (red) inside spheroid; nuclei stained with Hoechst (blue); scale bars = 5 µm. (i) Mean fluorescence intensity level and (j) volume of NSt‐ and St‐derived EVs, stained with PKH in spheroid, 24 h post P4 stimulation. p < 0.001 Weighted GLM; n = 4 biological replicates for each time. Results presented as mean ± standard error*1.96. (a), (b) and (e) were created with Biorender.com. Latin letters in f, g, i, j presents NSt versus St significant mean difference, “a” for the lowest, “ns” for insignificant difference. Ranking by time and statistics, see Supplementary Data 1.

Interestingly, the presence of spheroids significantly decreased the fluorescence flashes of pHluo‐CD63 puncta associated with the St endometrial cell membranes (p < 0.0001, Figure 2c, Movies S1–S3). In contrast, the presence of St endometrial cells significantly decreased the fluorescence flashes of pHluo‐CD63 puncta associated both with the spheroid cell membranes and with the extra‐spheroid space (i.e., GFP‐EVs in the medium) during the 40 min observation period (p < 0.0001, Figure 2d; Movies S4 and S5). Although by using a pHluo‐CD63 sensor we focus on the fusion of MVB to the plasma membrane for CD63‐EV secretion rather than the entire EV population, these results suggest that the spheroid changes its secretory output shortly after sensing the endometrium. Additionally, the decrease in pHluo‐CD63 fluorescent flashes inside the spheroid and endometrium cell membranes may indicate a reduction in MVB fusion or in vesicles tethering to the cell surface; this could lead to changes in the dynamics of secretion and uptake. Changes in fluorescent flashes of pHluo‐CD63 that correspond to changes in MVB fusion at the plasma membrane, or in vesicles tethered to the cell surface, were also detected in HeLa and other cancer cells (Sung et al. 2020; Verweij et al. 2018).

We next examined uptake patterns on both sides of the endometrium‐spheroid dialogue. The uptake of EVs was evaluated using PKH26 lipophilic membrane dye, which labels EV membranes. To examine the uptake of spheroid‐derived EVs into endometrial cells, isolated PKH26‐labelled EVs were added to the endometrial cells, and PKH26 fluorescence intensity was measured every hour for 4 h (Figure 2a,e,h). During the initial 2 h, no difference was detected in the level of spheroid‐derived EV uptake into St versus NSt endometrial cells. However, more spheroid‐EVs were taken up by NSt endometrium cells than by St cells at the 3 and 4 h intervals. Moreover, spheroid EV uptake level increased in NSt‐cells from 2 to 4 h, while the uptake level by St‐cells was constant from 1 to 4 h (p < 0.006, Figure 1f, Supplementary Data 1). Interestingly, at 2 h, a larger volume of spheroid EVs was taken up by St cells and a smaller volume by NSt cells; this difference was observable for up to 4 h (2 to 4 h; p < 0.001, Figure 2g; Supplementary Data 1). To determine whether spheroid‐EV puncta enter endometrial cells, we measured the distance of the spheroid‐EV puncta from the cell nucleus. Analysis of distances from the nuclei showed that ∼ 40% of the total spheroid‐derived EVs detected are closer to the nuclei in NSt and St cells, indicating that they were within the endometrial cells, rather than at the plasma membrane (Figure 2h, Supplementary Data 1).

Examination of the uptake of endometrial‐derived EVs by spheroid cells (Figure 2e–j) revealed that the level of fluorescence intensity of NSt endometrial EVs was significantly higher at 1 h in the spheroids than in St‐EVs. St‐derived EV uptake by the spheroid cells was significantly lower than that of NSt‐EVs at all observation times, except for 2 and 4 h (p < 0.001, Figure 2i, Supplementary Data 1). Additionally, St endometrial EVs taken up by spheroids are significantly smaller than the NSt‐EVs at all observation times, excluding the 2 h point (p < 0.001, Figure 2j). Also, NSt endometrial EVs are significantly closer to the spheroid centre than are St‐EVs, up to 2 h (p < 0.001, Figure S2b); this suggests that NSt‐EVs are internalized into the spheroid faster than St‐EVs, possibly owing to differences in internalization mechanisms (Feng et al. 2010; Van Niel et al. 2018; Tian et al. 2013).

To assess reciprocal uptake and EV exchange dynamics between endometrial cells and spheroids, we created and used fluorescent EV reporters PalmGFP and PalmtdTomato, enabling simultaneous labelling of various EV sizes and populations (Zaborowski et al. 2015; Lai et al. 2015) (Figure 3a–d). To simulate embryo arrival in a uterine cavity containing endometrial‐derived EVs, spheroids were added to endometrial cells that were either depleted or non‐depleted of pre‐dialogue EVs (i.e., EVs secreted prior to spheroid addition; Figure 3b). Using lightsheet 3D microscopy, which enabled visualization of the embryo‐maternal interaction in a fluorinated ethylene propylene (FEP) tube (e.g., simulating the uterine cavity;), we observed PalmGFP‐spheroid‐derived puncta and PalmtdTomato‐endometrial‐derived puncta (hereafter, Palm‐Sph‐puncta and Palm‐endo‐puncta, respectively). Palm puncta were either found inside the cells associated with the plasma membranes or secreted to the extracellular environment from both sides. Palm‐derived NSt and St EVs also differed in dynamics and patterns of secretion (Figure 3d, Movies S7, S8). Upon depletion of pre‐dialogue EVs, the fluorescence intensity of Palm St endo‐puncta inside the spheroid is significantly higher than that of NSt puncta after adding the spheroid and significantly lower at 2 h (p < 0.001, Figure 3e and Supplementary Data 1). Surprisingly, in the presence of EVs pre‐dialogue, mean spot fluorescence indicated a significant decrease in Palm St endo‐puncta inside the spheroid at 1 h, while NSt puncta levels significantly decreased at 2 h after addition of the spheroid. However, comparison between NSt and St puncta fluorescence levels in pre‐dialogue depleted and non‐depleted environments revealed NSt fluorescence to be significantly higher in both cases, beginning 1 h after adding the spheroid (p < 0.001, Figure 3e, Supplementary Data 1), suggesting that NSt endometrial EV uptake is faster, similarly to the uptake pattern we observed using PKH26‐labelled EVs (Figure 2i).

FIGURE 3.

Dynamics of EV secretion and uptake by the spheroid and endometrial cells is influenced by hormonal stimulation and the presence of EVs pre‐dialog. (a) Schematic presents the experimental timeline and (b) experiments with PalmGFP and PalmtdTomato reporters. (c) Representative confocal images showing PalmGFP (green, spheroid) and PalmtdTomato (magenta, NSt, and St endometrial cells) immediately and at 2 h after adding the spheroid to endometrial cells, in the presence of EVs pre‐dialog (NSt/St) or absence (NSt/St EV‐dep) where EVs were depleted before adding the spheroid; scale bar = 10 µm. (d) Live Lightsheet imaging of endometrial cells expressing PalmtdTomato (magenta) grew in FEP tube, and spheroid expressing PalmGFP (green); scale bar = 10 µm; see also Movies S3 and S4. (e) Quantification of PalmtdTomato‐derived fluorescence puncta inside the spheroid in the presence of EVs pre‐dialog (dashed blue‐NSt and dashed red‐St lines) or when EVs were depleted (bold blue‐NSt and red‐St lines; EV‐dep); p < 0.001, weighted Linear GLM; n = 3. (f) Quantification of fluorescence level of PalmGFP puncta inside the spheroid in the presence or absence of EVs pre‐dialog when the spheroid was exposed to NSt and St endometrial cells (dashed blue‐NSt and red‐St lines vs. bold blue and red lines) and (g) volume of PalmGFP puncta inside the spheroid; p < 0.001, weighted linear GLM, n = 3. (h) and (i) Representative transmitted electron microscopy (TEM) micrographs of spheroid and endometrial cells apical surfaces, respectively. Dashed white square indicates the location of the inset in the spheroid; note the multivesicular body (MVB, black asterisk) close to the surface with intraluminal vesicles (ILVs), elongated mitochondria (M) and vesicles at the endometrium surface (white arrowheads); scale bars = 1 µm. Arrowheads in i highlight the interphase; note the endometrial‐derived lipid droplets close to the surface (white asterisk) and mitochondria distribution at the surface. Scale bars = 1 µm. See also Figure S5. (j) Schematic presents the experiments with GlucB‐EV reporter. (k) The level of GlucB signal detected in the medium and cells of NSt and St endometrium, and (l) the spheroid examined during 2 h; p < 0.001; Linear GLM; n = 4. (m) Representative confocal image of St‐EVs stained with PKH26 (red) and SytoRNA (green), showing RNA localized inside and on the surface of the EVs. Scale bar = 1 µm. Results presented as mean ± standard error*1.96. Latin letters in e‐g, k, l, presents NSt versus St significant mean difference, “a” for the lowest difference. Ranking by time and statistics, see Supplementary Data 1. (a), (b) and (j) were created with Biorender.com.

While confocal and lightsheet microscopy detected Palm‐Sph puncta inside endometrial cells, accurate quantification was not possible. However, based on significant changes in spheroid puncta observed with the pHluorin tag in the presence of St endometrial cells (Figure 2d), we hypothesized that quantifying Palm‐Sph‐puncta within the spheroid could reveal its response (Figure 3c,f–g). In the presence of EVs pre‐dialogue, the fluorescence level of Palm‐Sph‐puncta inside the spheroid significantly increased over time when exposed to NSt and St endometrial cells (p < 0.001, Figure 3f, Supplementary Data 1). In addition, once pre‐dialogue EVs were depleted before adding the spheroid, the effect of the presence of NSt and St endometrial cells on Palm‐Sph‐puncta inside the spheroid was significantly different: while the fluorescence level of Palm‐Sph‐puncta inside the spheroid was significantly elevated up to 1 h and significantly decreased at 2 h in the presence of St endometrial cells, the reverse is true for the level of Palm‐Sph‐puncta in the presence of NSt endometrial cells (p < 0.001, Figure 3f, Supplementary Data 1). Additionally, the volume of Palm‐Sph‐puncta increased for up to 2 h inside spheroids exposed to NSt and St endometrial cells in the presence of pre‐dialogue EVs (p < 0.001, Figure 3g, Supplementary Data 1); once pre‐dialogue EVs were depleted, the volume of Palm‐Sph‐puncta inside the spheroid was significantly higher in the presence of St cells at all observation times (Figure 3g, p < 0.001). To test whether Palm‐Sph‐puncta are trafficked towards the surface of the spheroid (i.e., directed for secretion), we measured the distance of the Palm‐Sph‐puncta from the centre of the spheroid to its surface. We found that Palm‐Sph‐puncta inside spheroids exposed to St cells, with or without pre‐dialogue EVs, were significantly further from the centre than in the case where the spheroids are exposed to NSt cells, beginning 1 h after exposure (Figure S2c, p < 0.001, Supplementary Data 1). This indicates that the presence of St endometrial cells induces faster trafficking of spheroid EVs towards the spheroid surface than does the presence of NSt cells and further suggests that the uptake of endometrial EVs by the spheroid may modulate trafficking of spheroid EVs, possibly for secretion.

Electron micrographs of endometrial cells, spheroids and the interphase between the endometrial cells and attached spheroids show trafficking of EVs towards the apical side (Figure 3h,i and Figure S4a,b,f) and trafficking of endometrial LDs towards the spheroid attachment surface. In addition, we detected mitochondria and EVs at the spheroid interphase during exocytosis and endocytosis (Figure 3i and Figure S4c–e).

While evidence exists for endometrial‐ or embryo‐derived EVs in other organisms (Guzewska et al. 2023; Guzewska et al. 2023; Mazzarella et al. 2024; Fereshteh et al. 2019; Rudolf Vegas et al. 2022; Sheller‐Miller et al. 2016; Szuszkiewicz et al. 2022), the role of the extracellular space in EV exchange remains unclear. Hormonal stimulation influences EV uptake on both sides of the dialogue, and our results suggest that EV secretion and uptake dynamics are shaped by hormonal stimulation and pre‐dialogue EV presence in the extracellular space. This may indicate that pre‐dialogue extracellular components play a crucial role in initiating communication, while those secreted post‐dialogue are vital for sustaining and fine‐tuning the interaction between endometrium and embryo.

To examine whether the internalized endometrial and spheroid‐derived EVs are functional and whether they have an early effect on the recipient cells, we first used a bioluminescent (GlucB‐EV) EV membrane reporter (Lai et al. 2015; Lai et al. 2014). The GlucB‐EV detects translation of EV‐delivered GlucB mRNA by measuring luciferase activity. GlucB‐EVs were isolated from the culture media of GlucB‐expressing spheroid cells and endometrial cells. NSt and St endometrial cells were treated with GlucB‐EVs for 1 and 2 h, in the presence or absence of cycloheximide (CHX), a protein synthesis inhibitor. The GlucB bioluminescent signal of GlucB‐EV‐treated cells and conditioned media of spheroids and endometrium was measured (Figure 3a,j–m).

The level of GlucB signal detected in the medium of NSt and St endometrial cells increased significantly 1 h post‐exposure to GlucB‐spheroid EVs (Figure 3k, p < 0.001, Supplementary Data 1), indicating that translation of EV‐delivered GlucB mRNA began soon after EV uptake by the cells. This also suggests that 1 h after exposure, GlucB mRNA is already packed in EVs and secreted outside the cell. An elevated GlucB signal was also detected in NSt and St endometrial cells up to 1 h post‐EV exposure; it was, however, significantly lower in St than in NSt cells. Significant elevation of the GlucB signal was detected in St cells later, at 2 h post‐exposure (p < 0.001). At all the time points, samples treated with CHX, which prevents GlucB mRNA translation, consistently showed a significantly lower GlucB signal than non‐CHX‐treated medium and cells (p < 0.001). Also, in both NSt and St cells, the level of GlucB signal was lower than that detected in the media collected from these cells post‐exposure to EVs, suggesting that more EVs are released to the medium at an earlier time, and GlucB mRNA translation is ongoing. Exposure of spheroids to NSt and St GlucB‐endometrial EVs showed an elevation pattern similar to the one detected in the endometrium (p < 0.001, Figure 3l, Supplementary Data 1). However, the GlucB signal detected in the medium of spheroids exposed to NSt and St endometrial GlucB‐EVs was significantly stronger at 1 h post‐exposure and decreased at 2 h (p < 0.001). These translation patterns are correlated with levels of detected EV uptake patterns, where uptake of St endometrial‐derived EVs into the spheroid is immediately high and lower at 1 h post exposure to EVs (Figures 3e and 2i, respectively).

To further characterize NSt, St, spheroid and EV populations, we isolated the EVs and used nanoparticle tracking analysis (NTA). Although each EV population showed a characteristic size distribution, all four populations peaked at 100 nm. St‐derived EVs showed a wider size distribution and significantly larger mean particle size than NSt and media (p < 0.05, Figure S3e, Supplementary Data 1). Furthermore, endometrial EV populations are enriched with tetraspanin CD81 and CD9 and show various sizes of typical lipid bilayer spherical cup‐shaped morphology (Figure S4).

3.3. Hormonal Stimulation Increases the Levels of AhR Ligands and Alters the Levels of Energy Production‐Related Metabolites in Endometrial EVs

To understand how hormonal stimulation affects the metabolites exchanged by the endometrial cells and spheroids and their implications for implantation, we analyzed the metabolic profiles of EVs and the extracellular medium (EM) surrounding the EVs in both St and NSt endometrial cells. EV cargo analysis highlights specific metabolite changes packaged and transported between cells, shedding light on localized metabolic shifts and signalling mechanisms. Concurrently, examining the EM‐metabolites secreted from cells but not incorporated into EVs provides insights into broader extracellular environmental changes and the exclusiveness of EV cargo (Figures 4 and 5; Figures S6 and S7). The natural environment of cells in vivo is enriched with metabolites, both free and carried by EVs; we mimicked this state by culturing cells in non‐FBS‐depleted medium. This approach simulates the in vivo environment by showing how cells interact with and utilize available nutrients, offering a clearer understanding of metabolic dynamics during hormonal stimulation and the window of implantation. The basic growth medium was examined in the metabolomics experiments in parallel with medium collected after incubation with both St and NSt endometrial cells.

FIGURE 4.

AhR ligands are enriched in stimulated endometrium‐derived EVs. (a) Schematic presents the experimental timeline of the metabolomic experiment. (b) Metabolic diagram of metabolites in the tryptophan pathway. Metabolites that are uniquely found in EVs are marked in red. Undetected metabolites in this pathway are marked in gray. (c) Relative levels of tryptophan metabolites enriched or unique to EVs. (d) Mean fold change (log2 normalized to media) of metabolites related to tryptophan‐indole pathway. (e) Mean fold change (log2 normalized to media) of metabolites related to tryptophan‐kynurenine pathway, and acetylpyrazine. Mean ± SEM, n = 3 biological replicates. (c), (d), (e) Wilcoxon signed ranks test; Benjamini–Hochberg expected p value. Red asterisks show differences between St‐EVs and medium‐EVs, and St‐EM and medium‐EM; blue asterisks show differences between NSt‐EVs and medium‐EVs, and NSt‐EM and medium‐EM; black asterisks show differences between St‐ and NSt‐EVs, and St‐ and NSt‐EM. EM, extracellular medium without EVs See Supplementary Data 1 and 2. (a) and (b) were created with Biorender.com.

FIGURE 5.

Energetic shift‐related metabolites are enriched in endometrium‐derived EVs. (a) Mean fold change (log2 normalized to media) of TCA cycle‐related metabolites. (b) Metabolic diagram with indicators (red arrows) for specific metabolite enrichment in endometrium‐derived EVs. Presented are metabolites related to TCA cycle, glutamine‐glutamate, and glycolysis/gluconeogenesis pathways. Undetected metabolites in this pathway are marked in gray. (c) Mean fold change (log2 normalized to media) of metabolites related to glutamine‐glutamate pathway. (d) Mean fold change (log2 normalized to media) of metabolites related to glycolysis/gluconeogenesis pathways. Mean ± SEM, n = 3 biological replicates. Black asterisks show differences between St‐ and NSt‐EVs, and St‐ and NSt‐EM. EM, extracellular medium without EVs. Wilcoxon signed ranks test; Benjamini‐Hochberg expected p value. See Supplementary Data 1. (b) was created with Biorender.com.

To analyze a wide range of metabolites and lipids, we used targeted LC‐MS to profile 221 and 218 polar metabolites and 349 and 189 lipids in EVs and EM, respectively (Supplementary Data 2 and 3, Figures S6a and S7a). Global clustering of peak intensities revealed a wider range of enriched metabolites in EVs than in EM (Figure S6b,c). The metabolomic analysis of EVs from both St and NSt endometrial cells, as well as from the media, showed them to be significantly enriched with tryptophan and its metabolites, including those involved in the kynurenine and indole pathways (Figure 4b,c). Interestingly, four indoles (indole‐3‐pyruvate [IPyA], indoxyl sulphate [IS], trans‐3‐indoleacrilic acid [IacA] and 3‐[1H‐indol‐3‐yl]‐propionic acid [IPA]) were found exclusively within the EVs and not in the EM, suggesting that EVs might be crucial for the localized delivery of these metabolites within the endometrial environment. Many of these metabolites are endogenous ligands of the aryl hydrocarbon receptor (AhR); they are present in human endometrial cells (Küchenhoff et al. 1999), where they have been shown to inhibit estrogen receptors. AhR is a ligand‐activated transcription factor that responds to environmental toxins and endogenous metabolites, including those derived from tryptophan metabolism. Upon activation, AhR can modulate immune responses, cellular differentiation and the expression of genes involved in. Interestingly, the metabolite acetylpyrazine (APZ) was also found to be enriched in EVs compared to the EM (p < 0.004, Figure 4c). APZ is a pyrazine, an important class of heterocyclic compounds that can be obtained naturally or synthesized (Miniyar et al. 2013) and are produced naturally by all living organisms (Dolezal and Zitko 2015). Some pyrazines are AhR antagonists (Sun 2021) and have been demonstrated to have anti‐inflammatory activities (Hou et al. 2023). Next, to explore more precise differences between St and NSt endometrial cells unreferenced to extracellular medium‐origin metabolites, we examined fold changes of the AhR ligands (St and NSt normalized to medium; Figure S6c, Figure 4d,e, Supplementary Data 1). The metabolites indole‐3‐pyruvate (IPyA), trans‐3‐indoleacrilic acid (IacA), 3‐(1H‐indol‐3‐yl)‐propionic acid (IPA), acetylpyrazine (APZ), kynurenate (KynA) (p < 0.004) and indole‐3‐lactate (ILA; p < 0.04) were higher in St than in NSt endometrium. In St, but not in NSt endometrium‐derived EVs, we found 8 metabolites that were differentially enriched in EVs compared to the medium. These results suggest selective packaging of specific metabolites, in particular AhR ligands, into EVs following hormonal stimulation, as a part of preparation of the St endometrium for embryo implantation.

Analyzing pathway enrichment of differential metabolites in St compared to NSt endometrium‐derived EVs also revealed a significant increase in metabolites related to energy production pathways. (Figure S6d). Significant enrichment of tricarboxylic acid (TCA) cycle‐related metabolites α‐ketoglutarate (αKG; p < 0.04), isocitrate (p < 0.004), fumarate (p < 0.04), malate (p < 0.02) and lactate (p < 0.05) was observed in St endometrium‐derived EVs. Glutamate was significantly enriched in St endometrium‐derived EVs and EM (p < 0.04), while the opposite was observed regarding glutamine (p < 0.04) and oxidized glutathione (p < 0.04) (Figure 5c and Supplementary Data 1). With reference to glycolysis and gluconeogenesis, glucose was significantly enriched in NSt compared to St endometrium‐derived EVs (p < 0.04) and to the medium (p < 0.004) (Figure 5d and Supplementary Data 1). Glycerol was significantly reduced in St versus NSt endometrium‐derived EVs (p < 0.04), while glycerol‐3‐phosphate (G3P) was significantly enriched in St compared to NSt endometrium EM (p < 0.004). G3P is usually generated from glucose by glycolysis, but when the concentration of glucose drops in the cytosol, it is generated by the glyceroneogenesis pathway, using TCA cycle products (Nye et al. 2008). This pathway regulates cytosolic lipid levels and can regulate fatty acid concentration in the cytosol. It has also been shown to be involved in re‐esterification of fatty acids to generate triglycerides (Hanson and Reshef 2003). Our findings indicate that hormonal stimulation induces metabolic changes in both endometrial EVs and EM, playing a synergistic role in meeting the energy needs of implantation.

Analysis of lipid metabolites revealed that 169 of the 358 lipids detected were uniquely packaged into EVs (Figure S7). Specifically, triglycerides (TGs), sphingolipids (including ceramides, sphingomyelins and gangliosides), lysophosphatidylcholines, glycerophospholipids, and saturated and unsaturated fatty acids (including monosaturated and polysaturated fatty acids, MUFA and PUFA, respectively) were more highly enriched in EVs than in EM (Supplementary Data 1). Sphingolipids, a class of lipids located in the plasma membrane, in intracellular organelle membranes, and within EVs, function as signalling molecules and are known to be involved in implantation and decidualization in human and animal models (Kaneko‐Tarui et al. 2007; Matorras et al. 2020; O'Neil and Spencer 2021). This may indicate that they serve as a substrate for cellular signalling and organelle membrane biogenesis during endometrial receptivity acquirement. TGs as well as PUFA tended to be lower in St endometrial EVs than in extracellular media and NSt endometrium‐derived EVs. The detection of enriched G3P, together with decreased TG levels, suggests increased TG utilization in the St endometrium, probably due to high energetic demands.

3.4. EVs From Stimulated Endometrial Cells Modulate Their ECM Adhesion Profile and Increase Attachment of Spheroids

We next examined whether EVs and their cargo affect the attachment of the embryo and whether this effect varies with endometrial hormonal status. Initial assessment of the basic status of spheroid attachment to NSt and St endometrial cells (Figure 6a and b) showed a significantly higher attachment rate in St than in NSt cells (p < 0.001, Figure 6c and Supplementary Data 1), suggesting that hormonal stimulation is important for embryo attachment. Introduction of St endometrium‐derived EVs increased the attachment rate to the same level as that resulting from the hormonal stimulation itself (p < 0.002), indicating that EVs are actively involved in the attachment process. The introduction of EVs from NSt to St cells led to a decrease in embryo attachment (p < 0.003), suggesting that NSt‐EVs carry factors which might counteract the effects of hormonal stimulation, possibly by disrupting signalling pathways enhanced by estrogen and progesterone (Hart et al. 2023). Interestingly, adding EVs from St cells to other St cells significantly increased attachment beyond that observed in St (p < 0.05), suggesting that an St‐EV‐enriched environment enhances implantation. These findings highlight the distinct roles of EVs from St versus NSt cells in regulating embryo attachment, depending on the hormonal status of the endometrium.

FIGURE 6.

EVs and their cargo affect spheroid attachment to the endometrium and involved in remodelling of spheroid and endometrium ECM. (a) Schematic presents the experimental timeline and, (b) spheroid attachment experiments. (c) The effect of NSt and St‐EVs on spheroid attachment to endometrial cells (right graphs), also in the presence of aryl hydrocarbon receptor (AhR) specific inhibitor CH‐223191 (left graph), and (d) in the presence of indole‐3‐propionic acid (IPA), indole‐3‐acetate (IAA) and acetylpyrazine (APZ). *=p < 0.05, ANOVA with specific contrasts. (e) Schematic presents the experimental timeline and ECM remodelling experiments. (f) The effect of hormonal stimulation, spheroid (Sph)‐EVs, St‐EVs, and NSt‐EVs on endometrial and spheroid cells; p < 0.05, ANOVA with specific contrasts; see also Figure S8 and Supplementary Data 1. (a), (b) and (e) were created with Biorender.com.

Our metabolomics analysis revealed a large number of EV‐enriched metabolites known as AhR ligands. We thus assessed whether spheroid attachment may be affected by the inhibition of AhR by the specific inhibitor CH‐223191. AhR inhibition increased the attachment rate of spheroids to NSt endometrial cells (p < 0.002, Figure 6c and Supplementary Data 1). In St cells, inhibition of AhR following St‐EV enrichment led to increased attachment (p < 0.002). In general, AhR inhibition led to increased embryo attachment across all experimental conditions, whether the endometrial cells were hormonally stimulated or not. AhR activation can influence various cellular processes, including immune response and cell adhesion. By inhibiting AhR, we potentially alter the expression of adhesion molecules (Kung et al. 2009) and reduce the immune modulation (Kerkvliet 2009) that negatively impacts embryo attachment. This is consistent across all experimental groups, indicating a positive effect of AhR inhibition on embryo attachment. Given the presence of tryptophan metabolites, including indole derivatives, within the EVs, and their known ability to activate AhR, it is likely that these metabolites play a role in regulating endometrial receptivity via AhR signalling. To further examine the direct effect of AhR ligands on spheroid attachment, we added metabolites enriched in EVs (indole‐3‐acetic acid [IAA] and acetylpyrazine [APZ]) and unique to EVs (indole‐3‐propionic acid, IPA) to the endometrial culture environment.

Our results showed these metabolites to enhance embryo attachment in both St and NSt cells (p < 0.001; Figure 6d and Supplementary Data 1), suggesting their positive influence on embryo attachment via pathways related to AhR signalling (Jiang et al. 2023). This may involve direct influence on cell adhesion mechanisms or be governed by modulation of local immune responses (Wang et al. 2024).

To examine how endometrium‐ and spheroid‐derived EVs modulate spheroid attachment, we analysed the ECM of endometrial and spheroid cells before and after exposure to EVs. We tested the profiles of 5 components (fibronectin, vitronectin, laminin and collagen I and IV) using a Millicoat ECM screening kit. After 24 h of exposure to EVs, adhered cells were fixed and stained, and relative adherence to the ECM components was determined by absorbance readings (Figure 6a, e). We found that hormonal stimulation significantly increases the adherence of endometrial cells to Collagen I (p < 0.05; Figure S8 c, d, Supplementary Data 1). Exposure of NSt endometrial cells to NSt‐EVs increased adherence of NSt to fibronectin and collagen I and decreased adherence to vitronectin, while exposure of NSt to St‐EVs decreased only adherence to collagen IV (p < 0.05; Figures 6f, and S8c, respectively; Supplementary Data 1). Moreover, specific comparison between the effect of St‐EVs and NSt‐EVs on NSt endometrial cells revealed decreased adherence to fibronectin and collagen IV and increased adherence to vitronectin (p < 0.05; Figures 6f and S8c; Supplementary Data 1), highlighting the differential action of St‐ and NSt‐EVs on the endometrial cells. Exposure of NSt endometrial cells to spheroid‐derived EVs decreased adherence of NSt to vitronectin and collagen IV (p < 0.05; Figures 6f; Supplementary Data 1), while exposure of St to Sph‐derived EVs decreased adherence of St to laminin (p < 0.05; Figure 6f; Supplementary Data 1). These results suggest that St endometrium‐ and spheroid‐EVs, in addition to hormonal stimulation, may promote remodelling of the ECM, making it receptive to the spheroid upon its arrival to the uterine cavity (Figures 6f).

In comparison with St endometrial cells, spheroids showed greater adherence to fibronectin, vitronectin, laminin and collagen I and IV (p < 0.05; Figures 6f and S8 b, d; Supplementary Data 1). The effect of St‐derived EVs on spheroid cells is specific: exposure of Sph cells to St‐derived EVs significantly decreased their adherence to fibronectin, while exposure of Sph cells to NSt‐derived EVs led to increased adherence to fibronectn (p < 0.05; Figure 6f, Supplementary Data 1), possibly indicating that uptake of St‐EVs remodels the spheroid ECM by affecting its adherence capacity to fibronectin (Figure 6f).

Cells interact with ECM proteins through receptors such as integrins (Humphries et al. 2006, Whitby et al. 2020, Rüegg et al. 2023), which play a critical role in pregnancy. Collagen is particularly important, as abnormal collagen expression is linked to pregnancy failure (Shi et al. 2020). In porcines, trophoblast attachment to the endometrial epithelium involves integrin binding to the RGD sequence in fibronectin and vitronectin (Johnson et al. 2003, Bowen et al. 1997). Similarly, Schultz and Armant observed that fibronectin‐coated spheres bind to mouse blastocysts at specific initial attachment sites (Schultz and Armant 1995). Given the effects of St endometrium‐ and spheroid‐derived EVs on cell surfaces, EV uptake may activate reciprocal adhesion molecules such as integrins, altering adherence to ECM components such as fibronectin, vitronectin, laminin and collagen I and IV, thereby facilitating spheroid adhesion and attachment.

3.5. EVs Modulates Mitochondrial Polarisation and Lipid Droplet Dynamics in Endometrial and Spheroid Cells

To determine whether endometrium‐ and spheroid‐derived EVs affect metabolism, we first evaluated early changes in mitochondrial membrane potential following EV uptake. We exposed NSt and St endometrial cells to spheroid‐derived EVs and trophoblast cells to NSt and St endometrium‐derived EVs for 24 h, after which the cells were washed, stained with JC‐1 and monitored for 4 h (Figure 7a and b). The addition of spheroid‐derived EVs significantly altered endometrial mitochondrial polarization at all observation times (0 to 4 h; p < 0.001; Figure 7c, Supplementary Data 1), suggesting that uptake of spheroid‐EVs led to a decrease in endometrial mitochondrial activity at all observation times; we did not, however, detect a difference between the response of NSt and St to Sph‐EVs. In contrast, trophoblast cells exposed to St endometrial EVs showed high mitochondrial membrane potential immediately after the addition of JC‐1 dye, compared to trophoblast cells that were not exposed to endometrial EVs (i.e., 0 = 24 h post exposure to EVs; p < 0.05, Figure 7d, Supplementary Data 1). High mitochondrial membrane potential reflects an increase in energetic capacity and synthesis of ATP (Zorova et al. 2018), but exposure to St endometrial EVs significantly decreased membrane potential of trophoblast cells at 1 h and increased it from 1 to 3 h, followed by a significant decrease at 4 h (p < 0.05). Exposure to NSt‐EVs had no further effect on the trophoblast mitochondrial membrane potential after the decrease observed at 1 h (p < 0.05, Figure 7d, Supplementary Data 1). As a whole, these results demonstrate that endometrium‐ and spheroid‐derived EVs play a significant role in modulating mitochondrial activity and that internalization of St endometrial EVs has a unique effect on the mitochondria, with possible involvement in achieving the metabolic efficiency required during the process of implantation.

FIGURE 7.

EVs modulate mitochondrial activity and trafficking of LDs in both endometrial and spheroid cells. (a) Schematic presents the experimental timeline and, (b) mitochondrial activity experiments. (c) JC1 fluorescence presented as log transform of red/green ratio in NSt and St endometrial cells exposed to spheroid‐EVs, compared to NSt cells that were not exposed to EVs; Sph cells exposed to NSt‐EVs, St‐EVs, and Sph cells that were not exposed to EVs; p < 0.001, Linear GLM, Benjamini–Hochberg expected p value, n = 3. (e) Schematic presents the experimental timeline and lipid droplet (LD) examination experiments. (f) and (g) Label‐free LD quantification using NanoLive microscopy and Smart Lipid Droplet Assay (SDLA): number (f) and mass (g), LDs in St and NSt endometrial cells that were exposed to St‐EVs, NSt‐EVs and Sph‐EVs during 60 min; (f) p <  0.001, NBD GLM, Benjamini‐Hochberg expected P value; (g) p < 0.001; Linear GLM, Benjamini–Hochberg expected p value. n = 4. (h) Segmentation of LDs based on their unique refractive index in endometrial cells exposed to St‐ and NSt‐EVs and Sph‐EVs, using SDLA; scale bar = 5 µm. Latin letters in c, d, f, g, presents NSt versus St significant mean difference, “a” for the lowest difference. Ranking by time and statistics, see Supplementary Data 1. (b), (e) and (j) were created with Biorender.com.

To examine whether the addition of EVs has a functional effect on cellular LDs, we added spheroid‐derived EVs to St and NSt endometrial cells, St endometrium‐derived EVs to NSt cells and NSt endometrium‐derived EVs to St cells. We then monitored LD count, LD dry mass and LD distribution inside the cells for 60 min using NanoLive microscopy and Smart Lipid Droplet Assay (SDLA) (Figure 7a, e‐h). Addition of St endometrium‐derived EVs to NSt endometrial cells significantly reduced the number of LDs per cell after 15 min of incubation (p < 0.02, Figure 7f,m; Figure S9, Supplementary Data 1). Similarly, the addition of spheroid‐derived EVs to NSt endometrial cells significantly reduced the number of LDs per cell after 15 min of incubation (p < 0.008). Addition of spheroid‐derived EVs to St endometrial cells significantly reduced the number of LDs per cell as compared to NSt cells, but only at 60 min of incubation (p < 2 × 10⁻⁶). This may suggest higher LD utilization in the St endometrium following embryo introduction. The most significant reduction in LD number was observed 15 min after the addition of NSt endometrium‐derived EVs to St endometrial cells (reduction of 1.8, p < 1 × 10⁻⁹) as compared with the response of NSt cells to St‐EV (p < 0.05, Figure 7f, Supplementary Data 1).

We next measured the dry mass of LDs (pg) per cell based on their refractive index (Figure 7g); data regarding the dry mass of LDs can provide information on the quantity of lipids stored within cells. Addition of St endometrium‐derived EVs to NSt endometrial cells significantly increased the LD mean mass per cell after 15 min of incubation (p < 5 × 10⁻⁸); this was true to an even greater extent after 60 min (p < 6 × 10⁻¹⁰, Figure 7g and Supplementary Data 1). Such an increase in mass, accompanied by a reduction in the number of LDs, could occur if smaller LDs merge into larger ones, leading to fewer but larger droplets, and thus a higher overall mass. This process, known as LD fusion, occurs during metabolic shifts that favour storing lipids in fewer, larger droplets, often in response to changes in energy demands, cellular stress, or lipid metabolism. Smaller LDs, with their higher surface‐area‐to‐volume ratio, enable the rapid enzyme access for lipid mobilization that is crucial for processes with high energy demands, such as embryo implantation. Additionally, smaller LDs serve as reservoirs for membrane lipids, facilitating quick redistribution for vesicle formation, membrane repair and signalling, thus supporting processes requiring rapid membrane turnover or remodelling.

Addition of spheroid‐derived EVs to NSt endometrial cells significantly reduced LD mean mass per cell after 15 min of incubation (p < 0.05); a significant increase in mass was observed after 60 min (p < 2 × 10⁻¹⁰⁷). In contrast, addition of spheroid‐derived EVs to St endometrial cells significantly decreased the mean mass of LDs per cell after 60 min of incubation (p < 0.007, Figure 7g and Supplementary Data 1), hinting at LD utilization by the cells and altered dynamics of St endometrium following EV utilization. Lastly, we examined LD distance from the cell nucleus. Addition of St endometrium‐derived EVs to NSt endometrial cells did not affect LD mobilization inside the cells, while addition of NSt endometrium‐derived EVs to St endometrial cells significantly reduced LD distance from the nucleus after 15 min of incubation (p < 3 × 10⁻⁵). Addition of spheroid‐derived EVs to NSt endometrial cells gradually reduced the distance of LD from the nucleus beginning after 15 min of incubation (p < 0.03, Figure S2e, Supplementary Data 1). Addition of spheroid‐derived EVs to St endometrial cells significantly decreased LD distance from the nucleus only after 60 min of incubation (p < 3 × 10⁻¹³). These results suggest that exposure of the endometrium to spheroid‐derived EVs leads to mobilization of LD inside the cells towards the nucleus, hinting at utilization of the energetic content of LDs via signal transduction mechanisms.

4. Discussion

Implantation is a time‐sensitive event occurring within a narrow window of the menstrual cycle and requiring optimal endometrial preparation. Our study underscores the intricate interplay between hormonal stimulation and cellular and extracellular modifications. Hormonal cues drive significant changes in EV composition, function and trafficking dynamics. Endometrium‐derived EVs are enriched with specific metabolites, lipids and bioactive molecules and influence LD production and secretion. Together with spheroid‐derived EVs, they play a crucial role in maintaining the progress of the embryo‐maternal dialogue, modulating endometrial receptivity and facilitating early embryo attachment (Figure 8).

FIGURE 8.

A model illustrating the intricate interplay between hormonal stimulation and the cellular and extracellular modifications necessary for establishing endometrial receptivity during the implantation window.

The endometrium changes its secretory output in response to hormonal stimulation. The St endometrium‐derived EV population contains larger microvesicles than does the NSt population. Although not evaluated here, the fact that St endometrium‐derived EVs contain larger vesicles may suggest that the EVs are enriched with a higher and/or different content of specific bioactive molecules in response to hormonal stimulation (Kumar et al. 2024). Interestingly, metabolomic analysis revealed EVs to be highly enriched with AhR ligands, especially indoles (IAA, ICAld and ILA) and acetylpyrazine. Moreover, four of the indoles (IPA, IPyA, IS and IacA) were found exclusively within the EVs and not in the extracellular environment, suggesting that EVs may be crucial for the localized delivery of these metabolites within the endometrial milieu. The amino acid tryptophan is recognized as a source for generating several AhR ligands in the body via both microbial and host metabolism (Hubbard et al. 2015). Human in vitro decidualization has been shown to be enhanced by tryptophan and its metabolite kynurenine via the AhR pathway (Wang et al. 2020) in the mouse uterus; inhibition of AhR significantly improved decidualization under E2‐ and P4‐stimulated conditions (Chen et al. 2024). By inhibiting AhR, we obtained an increase in spheroid attachment to the NSt endometrial cells; this increase was even greater in St endometrium, suggesting that AhR may play a role in regulating embryo attachment. Furthermore, IPA, a specific microbial tryptophan metabolite entering the blood from the gut, was shown to improve glucose metabolism, suggesting anti‐inflammatory and antioxidant effects (Abildgaard et al. 2018). While IPA was negatively correlated with embryo implantation in human female serum (Zheng et al. 2023), our findings show that it is enriched in endometrium‐derived EVs and induces initial attachment of the spheroid to endometrial cells. This emphasizes the involvement of microbial metabolism in the implantation process. These findings point to potential therapeutic strategies for enhancing implantation by targeting AhR, particularly in environments rich in tryptophan metabolites (Gargaro et al. 2021).

Metabolomic analysis of EV cargo and metabolites from St and NSt endometrial cells in the extracellular environment also revealed significant differences in metabolites related to energy production, such as TCA cycle intermediates and G3P; this reflects the metabolic shift occurring in the endometrium during the implantation window (Chen and Dean 2023). The elevated levels of these metabolites, coupled with a glycolytic shift indicated by reduced glucose and pyruvate and increased lactate, reflect an adaptation towards enhanced energy production and metabolic flux. A metabolic shift in uterine cells, accompanied by a decrease in aerobic respiration at an early stage of embryo implantation, was also reported in mice (Yang et al. 2020). This shift, reminiscent of the Warburg effect, supports the increased energy and biomass requirements for preparing the endometrium for embryo implantation (Krisher and Prather 2012; Zhang et al. 2021; Zuo et al. 2015). The observed correlation between high G3P levels and reduced glucose levels, along with increased G3P shuttle pathway activity, suggests a metabolic adaptation for sustaining cellular energy production in a state of fluctuating nutrient availability. Our data show that populations of St endometrium‐derived EVs are more highly enriched than the extracellular medium with specific classes of lipids, including sphingolipids and triglycerides. By contrast, St‐EVs exhibit reduced sphingomyelin, ceramide and triglyceride levels compared with NSt‐EVs, suggesting enhanced lipid remodelling during vesicle formation. This may involve increased sphingomyelinase activity, which converts sphingomyelin into ceramide and downstream metabolites, altering membrane curvature and promoting EV biogenesis (Robciuc et al. 2012; Vallejo et al. 2012). The lower triglyceride levels in St endometrial EVs, despite increased G3P, suggest increased utilization of triglycerides to meet the high energy demands of the implantation process. Sphingolipids, which are involved in cellular signalling and membrane biogenesis, are particularly enriched in EVs and are crucial for implantation and decidualization (Kaneko‐Tarui et al. 2007; Yang et al. 2022). The altered lipid composition may result in formation of EVs with larger, more complex membranes (Llorente et al. 2013) necessary to support functions occurring during the implantation window. Ceramide itself is known to drive LD formation by facilitating neutral lipid synthesis. In trophoblasts, exposure to non‐esterified fatty acids (NEFA) or hypoxic stress similarly triggers formation of PLIN2‐marked triglyceride‐ and diacylglycerol‐rich LDs while suppressing β‐oxidation (Bildirici et al. 2018), indicating a buffering response to lipid overload. Parallel mechanisms may govern the way in which EV‐derived sphingolipids and neutral lipids are processed by embryo spheroids: they are initially stored in LDs and are then mobilized for fatty acid oxidation via CPT1/2 (Smolková and Gotvaldová 2025) or channelled into membrane phospholipids and signalling lipid pools (e.g., LPA, prostaglandins, S1P) (Li et al. 2020). Moreover, decidual very low‐density lipoprotein receptor (VLDLR) upregulation and trophoblast lipoprotein lipase (LPL) activity suggest co‐operative histotrophic lipid transfer from endometrium to embryo (Magnusson‐Olsson et al. 2006; Heerwagen et al. 2018; Gauster et al. 2007; Tamura et al. 2020). To delineate the relevant pathways, future studies should use labelled sphingomyelin, ceramide and triglycerides for uptake tracing, complemented by β‐oxidation flux assays, LD imaging and bioactive lipid profiling. Such an integrated approach will elucidate how reduced lipid content in St‐EVs influences trophoblast metabolism, signalling and, ultimately, implantation competence. Our results also show induction of LD production in endometrial cells regulated by St endometrium‐derived and spheroid‐derived EVs. Moreover, we report here that LDs are not only secreted from the endometrium during its preparation for implantation but are also taken up by embryonic cells in the very early stages of embryo implantation as a component of the soluble signals directed towards the embryo. We further demonstrate, for the first time, that EV uptake alters number, dry mass and intracellular localization of LDs during early endometrial preparation for implantation. LDs have been found to be involved in cell regulation and metabolic homeostasis via a physical interaction with mitochondria (Pu et al. 2011; Welte 2015). Although alterations in the endometrial metabolome have been documented (O'Neil and Spencer 2021; Zheng et al. 2023), our findings reveal selective packaging of energy shift‐related metabolites into EVs in response to hormonal stimulation; this may indicate a targeted mechanism for delivering bioactive molecules to recipient cells, which potentially influences endometrial receptivity, embryo development in preparation for implantation and implantation itself. We suggest that these rapid changes occur as early as the first stages of implantation, upon the embryo's initial arrival in the uterus, and not only at the decidualization step.

With the aim of bridging gaps in our knowledge of early embryo‐endometrium communication occurring upon the embryo's arrival in the uterine cavity prior to physical contact, we investigated the patterns of EV secretion, uptake and exchange in the presence of both sides of the dialogue. Our study shows EV uptake to be influenced by the hormonal status of endometrial cells, significantly shaping embryo‐endometrium dynamic interactions, aligning with findings by Vilella et al. 2015. St endometrial cells exhibited a constant uptake of spheroid‐derived EVs after 1 h of exposure, in contrast to NSt cells. St cells initially internalized larger volumes of spheroid‐derived EVs, consistent with the notion that hormonal priming alters endometrial cellular responses (Desrochers et al. 2016). Conversely, when examining the uptake and internalization of endometrium‐derived EVs by spheroid cells, we found that NSt endometrial EVs were more rapidly internalized by the spheroid than were St‐derived EVs. This rapid uptake could be linked to internalization of different EV subpopulations by distinct mechanisms, as previously noted (Feng et al. 2010; Tian et al. 2013). However, examination of the EV exchange pattern in the presence of both endometrium and spheroid showed that pre‐dialogue EVs found in the endometrial extracellular environment play a crucial role in modulating EV uptake and retention by the spheroid, particularly in a hormonally primed environment. The rapid decrease in St endometrial EV puncta within the spheroid in the presence of pre‐dialogue EVs aligns with the hypothesis that pre‐dialogue EVs may prime endometrial cells to respond differently to incoming EVs, facilitating faster EV processing and potential secretion (Vlassov et al. 2012). These findings highlight the critical role of pre‐dialogue EVs and the hormonal status of endometrial cells in modulating uptake, trafficking and secretion dynamics of EVs during the early embryo‐endometrium interaction. Our findings also indicate that the dialogue is initiated by components in the extracellular environment, while those secreted during embryo attachment further sustain and regulate the embryo‐maternal communication.

Importantly, our findings show that EVs are not only exchanged between the embryo and the endometrium during the early dialogue but are also rapidly utilized within cells, affecting cell functionality. Molecules such as miRs and EVs are known to act rapidly, affecting or internalized by their targets within only a few hours (Ando et al. 2017; Giacomini et al. 2017). Translation of EV‐derived mRNA occurred within 1 h of EV uptake by endometrial cells, as well as by spheroids. The utilized EVs rapidly influence a range of cellular processes, further regulating mitochondrial activity, LD production and secretion and extracellular matrix remodelling of the endometrium and the spheroids. All of these serve to enhance spheroid attachment to the endometrial cells. The distinct patterns of uptake, trafficking and utilization of EVs observed in our study highlight the importance of EV‐mediated communication in the early stages of embryo‐endometrium interaction. Additionally, the interplay among LDs, mitochondria and EVs suggests a coordinated mechanism supporting the metabolic demands of the endometrium and the developing embryo.

Taken together, while the use of specific cell lines rather than primary cultures limits the biological variability among samples (e.g., age, BMI), can significantly impact the implantation process and may not entirely replicate the in vivo environment, modelling embryo‐maternal interactions using cell lines is widely practised. Moreover, endometrial RL95‐2 cells and JAR trophoblasts serve as useful approximations of the cellular biology involved in embryo‐maternal dialogue during implantation. Thus, this study not only advances our understanding of embryo‐maternal EV‐mediated communication but also provides a valuable model for investigating EV‐mediated simultaneous intercellular bidirectional crosstalk in other biological contexts. Finally, the results presented here may also inform future therapeutic strategies aimed at improving fertility outcomes.

Author Contributions

Alisa Komsky‐Elbaz: conceptualization, investigation, writing – original draft, methodology, validation, visualization, writing – review and editing, supervision. Margarita Shuhmaher: conceptualization, investigation, writing – original draft, methodology, validation, visualization, writing – review and editing. Javier Arturo Sanchez‐Lopez: conceptualization, investigation, writing – original draft, methodology, validation, visualization, writing – review and editing. Oria Teena: visualization. Daniel Waiger: visualization, methodology. Einat Zelinger: methodology, visualization. Tally Kossovsky: methodology, visualization. Leilah Otikovs: methodology, visualization. Eshel Dilevsky: methodology, visualization. Yoav Soen: conceptualization, methodology. Amir Hefetz: investigation, writing – original draft, methodology, validation. Yael Heifetz: conceptualization, investigation, funding acquisition, writing – original draft, methodology, validation, visualization, writing – review and editing, supervision.

Conflicts of Interest

The author declares no conflicts of interest.

Reporting Summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

Supporting information

Supporting Information: jev270161‐sup‐0001‐movieS1

Supporting Information: jev270161‐sup‐0002‐movieS2

Supporting Information: jev270161‐sup‐0003‐movieS3

Supporting Information: jev270161‐sup‐0004‐movieS4

Supporting Information: jev270161‐sup‐0005‐movieS5

Supporting Information: jev270161‐sup‐0006‐movieS6

Supporting Information: jev270161‐sup‐0007‐movieS7

Supporting Information: jev270161‐sup‐0008‐movieS8

Supporting Information: jev270161‐sup‐0009‐movieS9

Supporting Information: jev270161‐sup‐0010‐figureS1

Supporting Information: jev270161‐sup‐0011‐figureS2

Supporting Information: jev270161‐sup‐0012‐figureS3

Supporting Information: jev270161‐sup‐0013‐figureS4

Supporting Information: jev270161‐sup‐0014‐figureS5

Supporting Information: jev270161‐sup‐0015‐figureS6

Supporting Information: jev270161‐sup‐0016‐figureS7

Supporting Information: jev270161‐sup‐0017‐figureS8

Supporting Information: jev270161‐sup‐0018‐figureS9

Supporting Information: jev270161‐sup‐0019‐SuppMat.pdf

Supplementary Data 1: jev270161‐sup‐0020‐SupplementaryData1.xlsx

Supplementary Data 2: jev270161‐sup‐0021‐SupplementaryData2.xlsx

Supplementary Data 3: jev270161‐sup‐0022‐SupplementaryData3.xlsx

Komsky‐Elbaz, A. , Shuhmaher M., Sanchez‐Lopez J. A., et al. 2025. “Extracellular Vesicles, Lipid Droplets and AhR Ligands in Early Implantation: The Dynamics of Embryo‐Maternal Crosstalk.” Journal of Extracellular Vesicles 14, no. 11: e70161. 10.1002/jev2.70161

Data Availability Statement

The authors declare that the data supporting the findings of this study are available within the paper and its supplementary information files. Source microscopy images are available from the corresponding author upon request.