Extracellular vesicles and female reproduction

Abstract

Extracellular vesicles (EVs) are nano-sized membrane bound complexes that have been identified as a mean for intercellular communication between cells and tissues both in physiological and pathological conditions. These vesicles contain numerous molecules involved in signal transduction including microRNAs, mRNAs, DNA, proteins, lipids, and cytokines and can affect the behavior of recipient cells. Female reproduction is dependent on extremely fine-tuned endocrine regulation, and EVs may represent an added layer that contributes to this regulation. This narrative review article provides an update on the research of the role of EVs in female reproduction including folliculogenesis, fertilization, embryo quality, and implantation. We also highlight potential pitfalls in typical EV studies and discuss gaps in the current literature.

Introduction

Extracellular vesicles (EVs) are heterogeneous nano-sized lipid bound organelles shed by cells into the extracellular space that can transfer proteins, lipids, cytokines, and RNAs between cells, both locally (autocrine and paracrine) and remotely [1, 2]. Growing evidence indicates that EV-packaged materials may induce functional modifications in the recipient cells, thereby positioning EV communication between cells and tissues as a previously underappreciated signaling system. The three main subtypes of EVs include exosomes, microvesicles, and apoptotic bodies. They are distinguished according to their biogenesis, release pathways, size, cargo, and functions [3–6]. Exosomes are intraluminal vesicles, typically 30–150 nm in diameter, formed by an endosomal route and secreted by all cell types. Exosomes are a source of cellular clearance but in addition participate in cell–cell communication, cell maintenance, and tumor progression. Microvesicles bud directly from the plasma membrane and are usually 100 nm to 1 μm in diameter. Similar to exosomes, microvesicles are also a source of cellular clearance and in addition are involved in cell–cell communication both between local and distant cells. Exosomes and microvesicles are characterized by protein markers as tetraspanins (CD9, CD63, and CD81) and adhesion integrins [7]. Apoptotic bodies are released into the extracellular space as blebs of cells undergoing apoptosis [5]. Their size varies from 50 to 5000 nm in diameter. In contrast to exosomes and microvesicles, apoptotic bodies contain nuclear fragments and cell debris. They do not exhibit surface markers characteristic for exosomes or microvesicles and do not play role in cellular communication [7–10]. EVs can be isolated from biological fluids. Hence, their bioactive content and possible functional roles make them ideal candidate biomarkers of health and disease, as well as potential targets for therapeutic interventions [11].

Female reproduction is dependent on extremely fine-tuned endocrine regulation, and EVs may represent an added layer that contributes to this regulation. This review summarizes the recent literature published on EVs and female reproduction including the role of EVs in folliculogenesis, oogenesis, fertilization, embryo quality, implantation (Fig. 1), and how infertility diagnoses and environmental factors might affect EV cargo. We will also discuss gaps and pitfalls in current studies.

Fig. 1.

Role of EVs in Female Reproduction

EVs in the ovarian follicle

Ovarian follicular growth is a harmonized process involving bidirectional interactions between granulosa cells, cumulus cells, and the oocyte [12, 13]. Several studies to date have isolated EVs from follicular fluid (FF) of various species including equine, bovine, and human [13–19]. Both in vitro and in vivo studies showed that fluorescent-labeled FF EVs can be taken up by their surrounding granulosa and cumulus cells [14, 20, 21]. While it is assumed that most FF EVs are secreted from cumulus and mural granulosa cells, some of these vesicles may originate from theca cells, oocytes, and non-ovarian tissues [22, 23] (Table 1).

Table 1.

Isolation of EVs based on the cell type or tissue that released them

Tissue/media	Species	Reference
Follicular fluid	Equine	Da Silveira et al. 2012 [14]
	Bovine	Hung et al., 2015 [17] Navakanitworakul et al., 2016 [19] Hung et al., 2017 [21] Morales Dalanezi et al., 2019 [18]
	Human	Machtinger et al., 2017 [24] Martinez et al., 2018 [25]
Granulosa/cumulus cells	Equine	Da Silveira et al. 2012 [14]
Oocyte	Mouse	Barraud-Lange et al., 2012 [23]
Embryo culture media	Bovine	Dissanayake et al., 2020 [26]
	Porcine	Saadeldin et al., 2014 [27]
Blastocoel fluid	Human	Battaglia et al., 2019 [28]
Uterine lumen/flushing	Sheep	Burns et al., 2014 [29] O'Neil et al., 2020 [30]
	Bovine	Ruiz-Gonzalez et al. 2015 [31] Kusama et al., 2018 [32] Qiao et al., 2018 [33] Nakamura et al., 2019 [34] O'neil et al., 2020 [30] Bridi et al., 2020 [35]

EV cargo profiles change throughout follicular development. In bovine, EV-miRNAs abundant in small follicles, representing early stage of antral follicle growth, were linked to genes relevant for cell proliferation pathways, and EV-miRNA rich in large follicles, representing late stages of antral follicle growth, were linked to genes in the inflammatory response pathways [19]. Furthermore, EVs isolated from small antral follicles were preferentially taken up by granulosa cells compared with EVs from larger follicles, thus advocating that EV surface markers and possibly their cargo can change during folliculogenesis [21]. Although more direct evidence is necessary to establish functional roles of EVs and their cargo in follicular growth, the changes in EV shows the potential for EV-mediated communication within the ovarian follicle.

Indeed, there is some suggestion that EVs may have a functional role. Bovine EVs released from FF promoted granulosa cell proliferation in a stage-specific manner where EVs from the highly proliferative small follicles induced greater granulosa cell proliferation compared to EVs from large antral follicles [21].

In bovine and murine models, FF EVs were able to promote cumulus expansion and alter the mRNA expression of genes such as prostaglandin-endoperoxide synthase 2 (PTGS2) and pentraxin 3 (PTX3) that are linked to cumulus expansion [17, 21]. The ability to promote partial cumulus expansion was replicated in a later study in porcine, but no changes in PTGS2 mRNA expression were observed [20]. Thus, while there is evidence that FF EVs can encourage cumulus expansion across species, it is unclear whether there are species-dependent differences and mechanisms or if the observed discrepancies arose from technical variability, such as different methods used to isolate the EVs.

There are some suggestions that FF EV-miRNAs act on mRNA targets that impact the wingless signaling pathway (WNT), transforming growth factor beta (TGF-β), mitogen-activated protein kinase (MAPK), neurotrophin, epidermal growth factor receptor (ErbB) pathways, and ubiquitin-mediated pathways. These pathways are known to be relevant for follicular development, granulosa cell proliferation and cumulus expansion, steroidogenesis, meiosis, and embryo mitosis in mammals [2, 14–16, 36–39]. Transcriptome analysis of porcine FF EVs shows that mRNAs isolated from FF EVs are highly enriched in genes annotated to PI3K-AKT, MAPK, and metabolic pathways such as lipid metabolism, glycolysis, and cholesterol biosynthesis, which are critical for follicular development and oocyte competence [22, 40–43], latency, activation and existence of primordial follicles [44], proliferation and differentiation of the somatic cells in the follicle [45], as well as meiosis [46, 47]. Ultimately, these speculative mechanisms need to be specifically investigated in future studies to clarify the functional impact of EVs on follicular growth.

Fertilization and embryo quality

Fertilization comprises of consecutive events including binding of the sperm to the zona pellucida, acrosome reaction, penetration of the sperm to the oocyte perivitelline space (PVS), and fusion of the sperm with the oocyte membrane [48–53]. EVs are found on the oocyte microvilli at the site of sperm attachment as well as on the surface of fertilizing sperm and can be required for the fusion of the oocyte and sperm [54]. In mice, EVs are transferred from the oocyte to the sperm in the PVS before direct interaction between the two gametes [55], possibly to induce sperm membrane reorganization and facilitate eventual fertilization, although this hypothesis has not been validated.

Juno, a critical receptor essential for sperm-oocyte fusion for the sperm-expressed Izumo1 protein, is rapidly shed from the oocyte immediately post-fertilization. Juno is not internalized after fertilization but found in EVs, apparently derived from the microvillus-rich oolemma that undergoes substantial architectural changes upon fertilization [56–58]. The loss of Juno from the oocyte surface leads to an effective membrane block and may even serve as “decoy eggs” that counteracts acrosome-reacted sperm [56, 58, 59].

EVs isolated from oviductal fluid can be incorporated into the oocytes, sperm, and embryos and thus improve fertilization and prevent polyspermy [60]. Oviductal EVs from cats contain proteins that are important for fertilization as in vitro culture of sperm with these EVs increase their fertilization ability [61]. A major limitation of IVF in porcine model is the excessive rates of polyspermy [62]. Supplementation of oviductal EVs during IVF significantly increased monospermy rates while maintaining an acceptable penetration rate compared with controls, highlighting a possible role of EVs in the prevention of polyspermy during fertilization [62]. The mechanism(s) for this phenomenon are still unexplored, but based on what was shown in mice model, it is possible that EVs play a role in the process of sperm-oocyte fusion.

In a human IVF setting, FF EV-miRNAs released from follicles that contained metaphase II oocytes were linked with fertilization potential and embryo quality. Using affinity-based commercial kit for EV isolation, several miRNAs, including miR-202-5p, miR-206, miR-16-1-3p, and miR-1244, were present at higher levels in FF from normally fertilized oocytes versus FF from oocytes that failed to fertilize [24]. In a follow-up study with a similar population but a different method of EV isolation (ultracentrifugation), miR-92a and miR-130b were over-expressed in FF samples from oocytes that failed to fertilize compared to FF from oocytes that were normally fertilized [25]. Of note, both miR-206 and miR-1244 were undetected in the follow-up study using ultracentrifugation, highlighting the technical differences underlying the discordant findings. The follow-up study also found that FF EV-miRNAs miR-766-3p, miR-663b, miR-132-3p, miR-16-5p, miR-888, miR-214, and miR-454 were differentially expressed in FF from oocytes that yielded top quality day 3 embryos versus those that did not yield top quality embryos [25]. Together, these two human studies suggest that FF EV-miRNAs may be related to fertilization and embryo quality, but the underlying causal mechanisms connecting them are unclear.

Embryonic EVs can affect early developmental potential

EVs were detected in the zona pellucida of human zygotes and in embryo conditioned culture media, but not in the zona pellucida of metaphase II oocytes, suggesting that EV secretion begins shortly after fertilization [7]. Bovine, porcine, murine, and human studies have all demonstrated that zygotes, embryos at cleavage, morula, and/or blastocyst stages secrete EVs into the extracellular medium [26, 27, 63–66]. In bovine, the size and concentration of EVs released by embryos differ based on embryo quality, which may hint at potential biological functions and their capability to serve as non-invasive markers for embryo quality [26, 66]. In bovine, fluorescent-labeled embryonic EVs were able to cross the intact zona pellucida and were internalized by other embryos [67]. Individual bovine embryo culture in medium enriched with EVs showed significantly higher blastocyst rates and significantly lower apoptotic cell ratio compared with individual embryo cultures that were not supplemented with EVs [67].

Also in bovine, supplementation of EVs extracted from pre-ovulatory follicles to the culture media during oocyte maturation and early embryo development induced changes in mRNA expression of metabolic and developmental related genes, blastocyst rates, global DNA methylation, and expression of miR-631, a bovine miRNA previously detected in EVs from cumulus-oocyte complexes, with no yet established functions in the context of female fertility [68, 69].

Prevailing data revealed that in vitro culture of embryos in groups can be more effective than embryos that were cultured individually [67, 70, 71]. A possible explanation for this phenomenon is that embryos release EVs that support the development of their co-cultured embryos by autocrine or paracrine manners [27, 72, 73]. In porcine, embryo-derived EVs were internalized by other embryos sharing the same culture medium and the mRNA expression of pluripotency genes as Oct4, Klf4, and Nanog was significantly higher among blastocysts that were cultured in groups compared with blastocysts that were cultured individually [27]. In mice, embryos co-cultured with EVs demonstrated higher blastocyst formation rate and reduced apoptotic rate of embryos in pregnant mice compared with embryos that were cultured without EVs [74].

While these findings can be attributed to delivery of proteins, miRNAs, mRNAs, or lipids by EVs that elicit direct or indirect effects on embryo quality, they can also result from originate from the serum itself [7, 64, 75] or from cumulus cells remaining in the culture drop or attached to the embryo [76, 77] or from other metabolites that were not separated from the EVs during the isolation process, thus emphasizing the shortcoming of most EV studies.

EV-miRNAs were recently isolated from human blastocoel fluid [28]. Most of the EV-miRNAs (87%) detected in the blastocoel fluid were also reported to be present in gametes (mainly in oocytes), including known embryonic miRNA clusters such as miR-302/367 cluster, miR-290/miR-371 cluster, miR17-92a-1 cluster (oncomiR cluster), and miR-106a-363 cluster. Interestingly, blastocoel fluid comprised also EV-miRNAs that belong to chromosome 19 miRNA cluster (C19MC). This cluster is known to be expressed nearly exclusively in the placenta [78]. The pathways of these blastocoel EV-miRNAs regulate key signals that control embryo development, such as pluripotency, cell reprogramming, epigenetic modifications, and intercellular communication [28].

Role of EVs during implantation

The interaction between the embryo and the endometrium is crucial for implantation. This complex procedure requires cross-talk between a receptive uterus and a competent blastocyst during limited time [79]. Asynchrony in any of these stages may lead to implantation failure. EVs were isolated from the uterine lumen fluid of several species including ovine, bovine mice, and human during various stages of the cycle [19, 29–35, 63, 80–83].

In ovine, the number of EVs in the uterine lumen varied over the estrous cycle [30]. In vitro, uterine lumen flush EVs were internalized by trophectoderm cells of the ovine conceptus and led to a decrease in trophectoderm cell proliferation rates [35]. In mouse, EVs collected from the endometrium on the pre-receptive phase, receptive phase, and during the time of implantation showed a different expression of miRNAs. One of them, miR-34c-5p, regulates growth arrest specific gene 1 (GAS1) which is likely involved in embryo implantation [82]. Incubation of mouse blastocysts with let-7 g–enriched EVs from endometrial epithelial cells prolonged their survival and led to the development of live pups after they were transferred to foster mothers [84].

In bovine, the addition of EVs collected from uterine flushing around the time of implantation led to downregulation of endometrial epithelial cell transcripts associated with the immune system [34]. MiRNA-sequencing of EVs taken from different pregnancy days showed changing miRNA profiles. MiR-98, in particular, was identified as a potential underlying regulator of the maternal immune system and mechanism behind the EV-induced gene expression change as incubation of bovine endometrial epithelial cells with synthetic miR-98 led to downregulation of cathepsin C (CTSC) and interleukin-6 (IL6) [34]. CTSC is involved in cell growth, activation of serine proteases, and capable of degrading intracellular protein [85, 86]. IL-6 is a pleiotropic cytokine that has various roles such as acute phase response, chronic inflammation, autoimmunity, and endothelial cell dysfunction [87].

In human, the field of blastocyst–endometrial interaction remains largely unexplored due to the difficulty of studying implantation [88]. Recent in vitro studies suggested a two-way communication between the conceptus and the endometrium via EVs. Hormonal treatment of endometrial cancer cell lines (ECC1) that represent the proliferative and secretory phases of the menstrual cycle lead to changes in EV protein load that was linked to major steps in implantation: adhesion, migration, invasion, and extracellular matrix remodeling [80]. These results were further validated in human primary uterine epithelial cell-derived EVs. Endometrial EVs internalized by trophoblast cells affected the adhesive ability of the trophectoderm through active focal adhesion kinase (FAK) signaling [80, 81]. In another study, embryo-derived EVs labeled with fluorescent dye were uptaken by primary human endometrial cells [63]. The intensity of the fluorescent signal of the endometrial cells was higher for those that were treated with EVs derived from blastocysts compared with those that were treated with EVs derived from day 3 embryos, suggesting a different cargo of according to the embryonic developmental stage [63].

This accumulating evidence supports the hypothesis that EVs may act as an emerging tool of communication between the endometrium and the implanting embryo. However, most of the published studies to date did not pinpoint a specific EV cargo or mechanism of action in which either embryo-derived EVs or endometrial EVs contribute to this process.

Possible factors affecting EV cargo

Several studies have indicated that gynecological diagnosis such as polycystic ovary syndrome (PCOS) and endometriosis, aging, body mass index (BMI), and exposure to environmental chemicals can alter the cargo of FF EVs.

FF EV-miRNA signature differs between the PCOS and healthy patients undergoing IVF, as the abundance of numerous miRNAs in FF EVs were found to be different between PCOS and healthy patients undergoing IVF treatment [89]. Similarly, the profile of EV-miRNAs isolated from lesions of endometriosis patients differed from EV-miRNAs derived either from the endometrium of the same women or endometrium collected from healthy controls. In addition, miRNAs from EVs collected from plasma of patients with endometriosis carried unique miRNAs such as miR-30d-5p, miR-16-5p, and miR-27a-3p which were not present in EVs isolated from plasma EV-miRNA from healthy controls [9]. A major limitation of this study is the inability discern of whether the observed differences in EV-miRNAs are causally related to PCOS or if they are biomarkers that reflect either PCOS or underlying mechanisms that cause PCOS.

The effect of age on FF EV-miRNA profile was demonstrated both in mares and human. In mares, FF EV-miRNA content differed between young (3–12 years) and old (20–26 years) animals, which was then predicted to target TGF-β signaling family, which affects follicle growth and development [90]. Among women undergoing IVF, EVs from FF of young (< 31 years old) and old (> 38 years old) patients also showed dissimilar miRNA profiles [91]. The predicted targets of these miRNAs were enriched in genes involved in p53 signaling, which is important to aging [92], and heparin-sulfate biosynthesis, which is important for cumulus oocyte complex expansion and oocyte maturation [93].

Patients’ BMI has been associated with FF EV-miRNAs cargo. Among women undergoing IVF, BMI is associated with altered expression of FF EV-miRNAs that are predicted to affect follicular and oocyte developmental pathways including PI3K-Akt signaling, extracellular matrix-receptor interaction, focal adhesion, FoxO signaling, and oocyte meiosis pathways. These findings provide a possible insight into a mechanism for the reduced fertility rates among patients with increased BMI [94].

Phenols and phthalates are potential endocrine disrupting chemicals (EDCs) that may alter female reproduction. IVF patients’ urinary concentrations of ethyl paraben and several metabolites of phthalate and phthalate alternatives were associated with differences in select FF EV-miRNA expression. KEGG enrichment analysis suggests that the target pathways of these FF EV-miRNAs include those related to follicular development and oocyte competence [95]. In another study, exposure of primary human granulosa cells to 20 mcg/ml bisphenol A resulted in decreased levels of EV encapsulated miR-27b-3p in the collected culture media compared with controls that were cultured without bisphenol A [96]. It is speculated that miR-27b-3p acts through its putative targets peroxisome proliferator-activated receptor gamma (PPARγ), insulin-like growth factor 1 (IGF-1), and Fas-associated via death domain (FADD) to affect follicular growth and other female reproductive processes [97–99]. Together, these findings suggest that exposure to select endocrine disruptor chemicals may lead to changes in EV-miRNAs that lead to downstream effects.

A prevalent limitation across all of the presented human studies is the lack of related health outcomes data. It is unclear whether the observed changes have any functional impact on health. Future studies should specifically aim to investigate whether these miRNA profiles associated with age, BMI, and environmental exposures are also associated with health outcomes. These studies should ideally be then integrated with experimental data to understand the functional role of these miRNAs, if any.

Challenges in current EV studies

Despite the increasing scientific interest in the field of EVs, studies of EVs in female reproductive health vary in quality and abundance of methodological details. Different protocols have been used for the isolation of EVs (ultrafiltration, differential centrifugation, density gradient centrifugation, chromatography, immune capture, and precipitation). These techniques vary according to their ability to discriminate soluble components from EVs and the size of EVs they can isolate. Using of various methods led to inconsistency in the size and component of EVs that were analyzed in different EV studies to date [24, 25]. Future studies should adhere to the revised 2018 minimal information of studies of extracellular vesicles (MISEV2018) guidelines, which presents terminology and technical suggestions for more standardized and rigorous reporting of EV science [100]. Collection, isolation, and storage of the fluids can influence the concentration, composition, and function of EVs and might have a significant impact on the downstream analysis of EVs [101]. The viscosity of these fluids, as well as their protein content, is mutable and therefore requires tailored protocols according to the biofluid or tissue analyzed [99]. While a variety of techniques are now available, balancing recovery and specificity, studies to date are hindered by the lack of a standardization of the techniques, which is crucial to minimize artefacts [102]. In studies that isolate EVs from culture media, there is possible contamination by EVs present in the serum used in this media. In order to decrease impurity, it is recommended to incubate the embryos for EV studies in serum-free media [103, 104]. In case that it is impossible, it is important also to separate EVs from the serum added to the culture media, as control.

One of the key features that affect the outcomes of EV studies is the storage of the vesicles. It is recommended to freeze the samples at − 80 °C to prevent formation of crystals, to reduce cryo-precipitation, and to avoid the degradation of the vesicles [105]. In addition, it is advocated to avoid repeated freeze-thaw of the aliquots before analysis [106–109]. In order to achieve reliable results, it is important to choose the EV isolation method according to the cargo planned to be assessed. Also, it is crucial to perform replications as well as correct to multiple comparisons [101, 110, 111]. Table 2 summarizes pitfalls in reproductive EV studies.

Table 2.

Pitfalls in EV studies

EV-associated processes	Pitfalls
Collection of EVs	A need for an optimal protocol is tailored to the type of (body) fluid, the type and/or cellular origin of the EVs of interest, and the downstream analysis
Different methods for EV isolation	A need for a consistency in the protocols for EV isolation to prevent variability in the concentration and size of EVs captured and analyzed
Storage	It is recommended to freeze the samples at 80 °C to reduce the degradation of the vesicles. In addition, it is advocated to avoid repeated freeze-thaw of the aliquots before analysis
Analyzing EV content	Sample replications and correction of the results for multiple comparisons are crucial

Future directions and human research

Much of the research to date has focused on in vitro and animal models, thus advancing our understanding of the role of EVs and their cargo in reproduction. However, in vitro studies cannot replicate the dynamic and complex human reproductive environment. More epidemiologic studies, particularly prospective ones guided by existing data, are needed to better understand relationships between external exposures, EV cargo, and reproductive outcomes relevant to human health. Additionally, although there is sufficient evidence that EVs and their cargo act as an intercellular communication system and play a major role in female reproductive health, the potential mechanistic pathways, including the specific EV cargo components that elicit the putative biological responses, are less well understood. There have been few studies using sensitive untargeted methods to assess the proteomic, metabolomics, and RNA cargoes. Many studies that identified differentially expressed EV-miRNAs use existing databases to identify potential mRNA targets and ontological approaches to surmise biological implications. Given our incomplete understanding of miRNAs, including their targets, such exploratory analyses are worthwhile, but carry uncertainty and need to be interpreted with great caution. Ontological and pathway analyses, in particular, are hampered by challenges such as incomplete or inaccurate annotations, inability to account for cell-types, often limited to canonical pathways [112, 113], and may produce numerous vague pathways that require substantial curation. Thus, there is a need for mechanistic studies to identify the specific biological responses in the target tissue in response to the uptake of EVs with certain cargo profiles.

Summary

EVs are an emerging tool to help understand different biological mechanisms in the female reproductive field, especially related to folliculogenesis, oogenesis, and embryo quality. While there have been great advances in our understanding, there are technical challenges such as a need for strict guidelines for the collection, isolation, storage, and analyses of the samples from different tissues and media in order to get reliable results. Human and mechanistic studies are needed, especially those that can identify both EV cargo changes as well as biological responses in the target tissue.