Extracellular vesicle-mediated delivery of molecular compounds into gametes and embryos: learning from nature

methods: AcomprehensiveelectronicsearchofPubMedandWebofSciencedatabaseswasperformedusingthefollowingkeywords:‘nano-particles’, ‘nanomaterials’, ‘cell-penetrating peptides’, ‘sperm’, ‘oocyte’, ‘egg’, ‘embryo’, ‘exosomes’, ‘microvesicles’, ‘extracellular vesicles’, ‘de-livery’,‘reproduction’,toidentifytherelevantresearchandreviewarticles,publishedinEnglishuptoJanuary2015.Thereferencelistsofidentiﬁed results: Biocompatibleengineered nanomaterials withhighloading capacity,stabilityand selectiveafﬁnityrepresenta potential versatile tool for the minimally invasive internalization of molecular cargo into gametes and embryos. However, it is becoming increasingly clear that the trans-lationoftheseexperimentaltoolsintoclinicalapplicationsislikelytobelimitedbytheirnon-biodegradablenature.Toallowthesubsequentuseof thesemethodologiesforclinicalART,studiesshouldutilizebiodegradabledeliveryplatforms,whichmimicnaturalmechanismsofmolecularcargo trafﬁckingascloselyaspossible.Currently,EVsrepresentthemostphysiologicalintracellulardeliverytoolsforreproductivescienceandmedicine. These natural mediators of cell communication combine the beneﬁts of engineered nanomaterials, such as the potential for in vitro production, targeting and loading, with the essential feature of biodegradability. conclusion: WeanticipatethatfutureinvestigationsintothepossibilityofapplyingEVsfortheintentionalintracellulardeliveryofmolecular compounds into gametes and embryos will open new horizons for reproductive science and clinical ART, ultimately leading to improvements in patient care.


Introduction
Background Assisted reproductive technology (ART) has revolutionized the field of infertility treatment, resulting in the birth of .5 million children worldwide ever since its first successful use in humans in 1978 (Adamson et al., 2013). Over the last four decades, pregnancy rates following ART have increased by nearly 6-fold, from 6% (Wang and Sauer, 2006) to 35% (ESHRE, 2014;HFEA, 2014). In these years, ART has expanded and improved, perhaps even more than anticipated in its early days, and transformed from a controversial experimental procedure to a routine medical treatment. However, its success rates, from the modern perspective, remain remarkably insufficient to consider this technique a reliable solution to the problem of infertility. According to recent estimations, the average live birth rate after ART globally still remains reasonably low, and does not exceed 30% per started cycle (ESHRE, 2014). At the same time, the demands for successful ART, especially in post-industrial economies, are continuously growing. The main driving forces for this trend are the increasing prevalence of age-related infertility due to the voluntary postponement of parenthood, and the expansion of assisted reproduction into 'non-infertility' indications, such as the preimplantation genetic diagnosis of hereditary diseases and fertility preservation for medical or social reasons (reviewed in Barkalina et al., 2014a). The sub-optimal success rates of ART are generally attributed to two key factors. Firstly, the conventional techniques for selection of embryos to be transferred back into the patient's uterus have inherent limitations, since they are based exclusively upon the morphological assessment of embryos, and not the evaluation of their chromosomal status and, therefore, developmental potential in the long-term (Fragouli et al., 2014). Secondly, in vitro handling of gametes and embryos, which forms an integral part of ART, has been reported to induce microstructural and functional damage in these delicate structures, with consequential reduction in developmental competency. There is mounting evidence that gamete processing in vitro during ART also promotes the fragmentation of sperm DNA (Toro et al., 2009;Matsuura et al., 2010;Rougier et al., 2013), reduces the levels of sperm-borne oocyte-activating factor phospholipase C zeta (PLCz: Kashir et al., 2011;Yelumalai et al., 2013), and facilitates oxidative stress in unfertilized oocytes (Martin-Romero et al., 2008;Otsuki et al., 2009); all of which, in the case of gametes with already compromised fertility, can have profound negative effects. Optimization of in vitro culture conditions, such as the supplementation of culture media with antioxidants, small molecules and growth factors (Kawamura et al., 2012;Yun et al., 2013;Tardif et al., 2014), or embryo incubation in time-lapse monitoring systems, which do not require repeated interruptions of culture for morphology assessment (Meseguer et al., 2012), has been reported to increase gamete/embryo survival and improve developmental potential. These observations elegantly indicate that the potential to improve ART success rates via the wider adoption of such approaches is both exciting and necessary. Nevertheless, substantial breakthroughs in the field of clinical ART can only be achieved via ongoing fundamental reproductive biology studies into the physiological mechanisms underlying reproduction, enabling the discovery of targeted molecular tools to investigate and manipulate these fine mechanisms at the cellular level.

Research into the mechanisms of gamete function: current challenges
The use of molecular research tools, including oligonucleotides, nucleic acids, peptides, antibodies, fluorescent markers and small molecules, forms the cornerstone of experimental studies in developmental and reproductive biology. These tools allow the precise mapping of specific cellular structures and molecular pathways, and tracking of their activity and fate at the different stages of gamete/embryo development. However, this seemingly straightforward approach, which is universally applied for experiments in biology, is associated with substantial challenges when used for studies of gamete structure and function in vitro. These highly specialized cells, especially in their mature form and after isolation from the natural microenvironment, acquire remarkable resistance towards the uptake of exogenous compounds. The specific molecular structure of the sperm membrane, characterized by an increased proportion of polyunsaturated fatty acids and the presence of rare ether-linked phospholipids, plasmalogens (Lenzi et al., 2000;Tapia et al., 2012), along with high structural and functional compartmentalization (James et al., 2004) and low activity of endocytotic processes , render sperm a particularly difficult target for the intracellular delivery of investigative compounds in vitro. Similarly, the oocyte, throughout its maturation in vivo, maintains an intimate relationship with the surrounding cumulus cells, which deliver essential nutrients into the oocyte via a system of gap junctions between the long cumulus cell processes penetrating the zona pellucida and the oocyte plasma membrane (Eppig et al., 2005). Studies of oocyte structure and function in vitro often require the mechanical removal of these surrounding nurturing cells to facilitate visualization of the female gamete, and, subsequently, compromise the physiological mechanisms of cargo internalization.
In its current form, the in vitro intracellular delivery of research compounds into gametes, and particularly into sperm, often requires the use of powerful membrane-disrupting agents, such as cholamidopropyldimethylammoniopropanesulfonate hydrate (CHAPS), Tween 20 and Triton X-100 (Jakop et al., 2009) with subsequent fixation, which renders the gametes entirely unsuitable for further use (Garcia-Vazquez et al., 2009;Yamauchi et al., 2012). Consequently, this approach does not allow for the evaluation of how the differences in gamete structure relate to their functionality, especially the ability to initiate and sustain normal embryo development. Improvement of intracellular delivery into gametes in a non-aggressive fashion and without effects upon developmental potential is, therefore, pre-exquisite for the improvement of our existing knowledge of reproductive biology, and, consequently, the advancement of ART.
From a rather more applied perspective, tools for efficient and nondamaging intra-gamete delivery could hold a therapeutic promise for patients with infertility caused by specific molecular deficiencies in gametes, for example deficiency of the sperm-borne oocyte-activating factor PLCz, resulting in oocyte activation failure post-fertilization, even following the intracytoplasmic injection of sperm into oocytes (ICSI) (Amdani et al., 2013). Similarly, these tools could be used in applied ART to supplement gametes with fertility-enhancing compounds, either promoting sperm motility or protecting gametes from deterioration during long-term culture in vitro (Kawamura et al., 2012;Yun et al., 2013;Tardif et al., 2014), especially for such indications as the in vitro maturation of oocytes or the in vitro culture of oocytes from primordial follicles for experimental fertility preservation programmes (Telfer and McLaughlin, 2012).
In this review, we outline the current state of nanomaterial-mediated delivery into gametes and embryos in vitro, and discuss the potential of a novel exciting drug delivery technology, based upon the use of targeted 'natural' nanoparticles, known as extracellular vesicles (EVs), for reproductive science and ART, given the promising emerging data from other fields.
Nanoparticle-mediated delivery for reproductive biology: a potential strategy to improve uptake into gametes Biomedical nanomaterials represent versatile small-scale platforms for targeted delivery Nanotechnology is a novel and rapidly developing field of science, positioned at the interface of physical, chemical, biological, materials and computer sciences, which investigates and manipulates physical matter at the nanoscale (1 -100 nm). From the biomedical perspective, the revolutionary nature of nanotechnology lies in its ability to design a customisable small-scale biocompatible delivery platform with large loading capacity, stability and highly specific affinity towards selected cell populations (Riehemann et al., 2009;Lehner et al., 2013;Tsai et al., 2014). Biomedical nanomaterials offer enormous targeting options, which can be achieved either via the modification of the physicochemical properties of nanomaterials, such as size, shape, surface charge and chemistry ('passive' targeting), or intentional functionalisation of the surface of nanomaterials with specific affinity moieties, for example peptides, antibodies and aptamers ('active' targeting), which selectively bind with the complimentary ligands on the surface of target cells (Riehemann et al., 2009;Petros and DeSimone, 2010;Albanese et al., 2012). Nanomaterials are universally characterized by their small size, comparable to the size of biological molecules and/or intracellular organelles, and their vast surface area, allowing the nanocarrier to be loaded with large amounts of almost any type of biological cargo, including a combination of contrast and therapeutic agents for the simultaneous detection and targeted treatment of pathologic lesions ('nanotheranostic') (Lammers et al., 2011). The small size enables straightforward internalization of nanomaterials inside the cells using the innate physiological mechanisms of uptake, with subsequent intracellular transport and metabolism (Petros and DeSimone, 2010;Kunzmann et al., 2011). Furthermore, nanomaterials are robust and, therefore, capable of carrying payloads to distant target locations following systemic administration. Collectively, these features, summarized in Table I, render biomedical nanomaterials strong candidates for the targeted delivery of diagnostic and therapeutic agents, including those with poor bioavailability after systemic use or high non-specific cytotoxicity.
Over the last decade, the use of nanomaterials for diagnostic imaging and drug delivery into pathological lesions has consistently proven advantageous in such fields as oncology, and infectious and chronic internal diseases (Ulbrich and Lamprecht, 2010;Brakmane et al., 2012;Psarros et al., 2012;Holmes, 2013;Tsai et al., 2014). This success has triggered an expansion of nanobiotechnological 'vision' to other scientific disciplines, including reproductive biology (Barkalina et al., 2014a). Indeed, universal evidence that nanomaterials improve the selectivity and efficacy of cargo delivery across a variety of cell types (Ryan and Brayden, 2014;Tsai et al., 2014) and do not compromise cell function, render them particularly attractive candidates for intracellular delivery into gametes and embryos.

Nanomaterial-mediated delivery into gametes is an encouraging, yet controversial field
The number of studies utilizing nanomaterials for the transfer of molecular compounds into gametes has been steadily growing since the mid-2000s, however the total number of publications still remains relatively low (Table II). Currently, the spectrum of nanomaterials with favourable biocompatibility with gametes/embryos includes polyvinylalcohol-functionalised iron oxide (Ben-David Makhluf et al., 2006;Makhluf et al., 2008), magnetic (Kim et al., 2010) and polystyrene (Fynewever et al., 2007) nanoparticles, mesoporous silica (Barkalina et al., 2014b), cerium dioxide (Falchi et al., 2014), perfluorocarbon (Jallouk et al., 2014), halloysite clay nanotubes and commercial polymeric nanotransfectants (Campos et al., 2011a, b), specialized CdSe/ZnS quantum dots (Feugang et al., 2012), and nanogold (Taylor et al., 2014b;Tiedemann et al., 2014). Most of these studies have consistently demonstrated that the use of nanomaterials improves the efficacy of research techniques, based upon the internalization of molecular compounds into gametes. These techniques primarily involved loading sperm with exogenous genetic constructs for subsequent sperm-mediated gene transfer into the oocyte at the time of fertilization (Kim et al., 2010;Campos et al., 2011a, b), proof-of-principle transfer of proteins into sperm (Makhluf et al., 2008), labelling of preimplantation embryos during in vitro culture (Fynewever et al., 2007), sperm bioimaging (Feugang et al., 2012), and sorting into subpopulations (Odhiambo et al., 2014;Barchanski et al., 2015)-all with positive outcomes.
However, even in view of these encouraging findings, substantial concerns associated with the use of nanomaterials for intra-gamete delivery remain. Thus far, most studies evaluating the potential effects of nanomaterials in gametes have focused specifically on sperm since this cell represents the main target for loading with exogenous compounds, either for sorting purposes or for sperm-mediated gene/protein delivery into the oocyte. Secondly, the nanomaterials, which have been tested for safety in sperm almost exclusively belong to the non-biodegradable category, which raises legitimate concerns about their potential long-term effects in the case of stable integration into embryonic cells. Although the studied nanomaterial-based intra-gamete delivery platforms have been reported to exert their transport function primarily via anchoring to the surface of the gamete plasma membrane or by intra-membrane sequestration, rather than cytoplasmic internalization, a small proportion of nanomaterial has been reported to reach the intracellular compartment in most cases (Kim et al., 2010;Feugang et Nanogold Proof-of-principle investigation of the potential to label the specific DNA sequences in viable sperm NPs, nanoparticles; SMGT, sperm-mediated gene transfer. of data regarding the long-term effects of non-biodegradable nanomaterials upon the embryo/fetal development, which arise from the methodologically diverse studies utilizing different protocols and animal models (Celá et al., 2014), form the main reasons for general caution towards the application of non-biodegradable nanomaterials for intra-gamete delivery outside the experimental setting. Therefore, discovery of an alternative small-scale versatile delivery platform, which would interact with gametes and transport molecular compounds inside these cells in a fashion similar to previously studied inorganic nanomaterials but at the same time undergo biodegradation, would be highly advantageous for reproductive biology and ART.
Cell-penetrating peptides act both as targeting tools for nanomaterials and independent delivery platforms In recent years, the search for a targeted biodegradable delivery tool, capable of transporting biological cargo into gametes and embryos, prompted an investigation into the potential benefits of cell-penetrating peptides (CPPs). CPPs are a specific class of short cationic/amphipathic peptides (,30 amino acids), capable of undergoing the energy-and receptor-independent translocation across the plasma membrane and transporting a considerably larger molecular cargo inside the cell, bypassing the traditional internalization pathways (Patel et al., 2007;Jones and Sayers, 2012). To date, a number of CPPs have been described as possessing affinity towards mammalian gametes and embryos, with several CPPs also demonstrating a promising delivery capacity (Table III). Apart from their innate delivery potential, these CPPs can be applied as functionalisation tools for active targeting of nanomaterials towards a particular cell population. For example, in a recent study, Dr Coward's group have identified that functionalisation of mesoporous silica nanoparticles with the CPP C105Y results in an increase of their binding rate with mammalian sperm in vitro and changes in binding profiles, which start to mimic those previously described for free C105Y Barkalina et al., 2015). Other authors have utilized the functionalisation with nona-arginine R9 for CdSe/ZnS quantum dots (Feugang et al., 2012) and deca-arginine (R10), transactivator of transcription and simian-virus 40 large T antigen nuclear localization signal peptide for gold nanoparticles (Barchanski et al., 2015), to target mammalian sperm, although the efficacy of these CPPs in improving the sperm-particle interaction has not been demonstrated consistently. Although CPPs are biodegradable and demonstrate a cargo delivery capacity, the high cost of production, need for specialized peptide synthesis equipment and dependence upon a third-party manufacturer, along with a lower loading capacity and limited potential for additional functionalisation, restrict their use as delivery tools primarily to a large research laboratory setting. An ideal delivery vehicle for reproductive science and medicine would represent a nano-sized multifunctional biodegradable cargo carrier, which can be produced in most laboratories in a straightforward fashion, targeted towards gametes/embryos and loaded with different types of cargo. In fact, a prototype of this platform exists in nature already, and is known to cell biologists as EVs. There is increasing understanding that these natural nanoparticles, universally secreted by pro-and eukaryotic cells, are involved in the crucial processes underlying gamete development, maturation and acquisition of fertilization potential. Moreover, these processes appear to be highly conserved across a variety of biological species (Sohel et al., 2013;Sullivan and Saez, 2013). These observations form another justification for studies into the feasibility of manipulating these key processes using similar 'recombinant' nanoplatforms.
Exosomes and microvesicles in reproductive biology: natural delivery vectors and mediators of cell function  Extracellular vesicle for intra-gamete delivery in ART by a variety of cell types into their microenvironment. According to the most recent views, the main difference between exosomes and MVs lies not in their size, from 40 nm to 100 nm for exosomes and up to 1 mm for MVs, as it was assumed previously, but in the mechanism of production (Raposo and Stoorvogel, 2013). Exosomes represent derivatives of multivesicular bodies (MVBs), and are first formed as intra-luminal vesicles (ILVs) inside these enclosed intracellular compartments via inward budding of the MVB membrane. These ILVs undergo release from cells upon the fusion of MVBs with the plasma membrane, and form exosomes. In contrast, MVs originate via direct budding from the plasma membrane (Akers et al., 2013). Ever since exosomes were first described in the 1980s as reticulocyte-secreted vesicles involved in the process of transferrin receptor recycling (Harding et al., 1983;Pan and Johnstone, 1983), our understanding of the fundamental role of these natural organic nanoparticles in cellular communication has evolved enormously. Exosomes have been demonstrated to have a complex tissue-specific organic content, including bioactive lipids (Record et al., 2014), proteins (Fontana et al., 2013), cytokines, growth factors, messenger RNAs (mRNAs) and non-coding transcripts, such as microRNAs (miRNAs) (Braicu et al., 2015). However, the composition of MVs has been studied to a far lesser extent (Raposo and Stoorvogel, 2013). Today, EVs are universally recognized as powerful paracrine and long-range mediators of cellular communication, playing an essential role in stem cell maintenance, cell proliferation and apoptosis, tissue repair, angiogenesis and immune response (El-Andaloussi et al., 2012). Interest in the potential role of EVs in the key processes underlying reproduction began with the elucidation of their function in mammalian pregnancy, both as mediators of maternal immunosuppression, preventing the rejection of the semi-allogenic fetus by the mother's immune system, and also as vasoactive messengers involved in endothelial dysfunction during pre-eclampsia (Knight et al., 1998;Taylor et al., 2006). Such seemingly conflicting roles were later attributed to the contrasting functions of two distinct pregnancy-specific EV subpopulations: placental exosomes with multiple immunomodulatory properties, favouring successful pregnancy, and pro-inflammatory syncytiotrophoblast-derived microvesicles/microparticles (STBMs), closely involved in pre-eclampsia (Redman et al., 2012;Mincheva-Nilsson and Baranov, 2014). Today, pregnancy-specific EVs are recognized as paramount mediators of fetomaternal crosstalk, responsible for the orchestration of a series of events leading to the establishment and maintenance of mammalian pregnancy. In particular, placental exosomes have been reported to promote vascular smooth muscle cell and endothelial migration, which is essential for the remodelling of uterine spiral arteries and the establishment of physiological feto-maternal placental circulation (Salomon et al., 2013(Salomon et al., , 2014, and triggering apoptosis in activated immune cells (Stenqvist et al., 2013). STBMs, in contrast, are characterized by their proinflammatory, anti-endothelial, and procoagulant effects, and the total concentration and levels of expression of associated 'endogenous danger molecules', such as heat shock protein 70 (HSP70) and high mobility group box 1 (HMGB1), have been reported to directly correlate with the severity of pre-eclampsia (Redman et al., 2012).

EVs regulate sperm maturation via the direct transfer of essential proteins to sperm
Over recent years, and in addition to their role in mammalian pregnancy, the significant contribution of EVs to the fine processes of gamete maturation and the acquisition of fertilization potential are beginning to be elucidated (Table IV). Furthermore, the fact that these processes are highly conserved across many species is being increasingly recognized (Corrigan et al., 2014). Although prostate-derived EVs ('prostasomes') were first discovered in human semen in 1978, the highly important contribution of EVs (produced in various portions of the male reproductive tract) to post-testicular sperm maturation across the variety of mammalian species was not recognized until the 1990s (reviewed in Saez et al., 2003;Sullivan et al., 2005). Post-testicular sperm maturation involves structural and functional reorganization of the sperm membrane during its passage through the epididymis, and is essential for the acquisition of motility and fertilization ability. At this stage, the sperm will have already lost the capacity for active protein synthesis. Therefore, the modification of sperm surface structure with protein targets for the recognition of zona pellucida is largely dependent upon the direct transfer of essential compounds from the epididymal microenvironment. The limited endocytotic activity of sperm at this stage necessitates the nonconventional molecular translocation mechanisms, which are currently considered to be mediated by the EVs of epididymal origin, also referred to as epididymosomes. Epididymosomes are involved in the enrichment of the sperm surface membrane with a wide range of proteins: glycosylphosphatidylinositol (GPI)-anchored proteins, including the proteins P26h, P25b, and P34H, essential for fertilization in hamster, bull, and human, respectively (Legare et al., 1999;Frenette et al., 2002); a disintegrin and metalloprotease (ADAM) 7 (Oh et al., 2009) and glioma pathogenesis-related 1-like protein 1 (GliPr1L1) (Caballero et al., 2012), both of which are involved in the interaction of sperm with zona pellucida; tyrosine kinase cSrc, playing a role in sperm capacitation (Krapf et al., 2012); epididymal sperm binding protein 1 (ELSPBP1), serving as a 'molecular tag' for dead epididymal sperm in certain animals ; plasma membrane Ca 2+ -ATPase (PMCA) with important functions for male fertility (Schwarz et al., 2013). In addition, recent observations show that epididymosomes are also involved in the regulation of post-transcriptional gene expression within the epididymal epithelium via the intercellular transport of miRNAs between the portions of male reproductive tract . The mechanisms of protein transfer between epididymosomes have not been yet elucidated in detail. However, it has been shown that epididymosomes, similarly to sperm, demonstrate the compartmentalization of proteins on the surface membrane and contain similar detergent-resistant domains, or lipid rafts, enriched in cholesterol and sphingomyelin, which can directly exchange proteins with the corresponding domains on the sperm surface (Girouard et al., 2009). Another study has shown that epididymosomes can fuse with the sperm membrane, and this fusion changes both the protein and lipid composition of sperm (Schwarz et al., 2013).
Prostasomes, secreted by the prostate gland epithelium and rich in sphingomyelin and cholesterol, have also been reported to 'supplement' sperm with complement-inhibitory molecules (CD59), shielding the sperm from immune recognition in the female genital tract (Rooney et al., 1993) via the pH-dependent fusion mechanism (Arienti et al., 1997), interact with polymorphonuclear and mononuclear leucocytes (Arienti et al., 1998) and reduce the overall reactive oxygen species production by polymorphonuclear neutrophils via the inhibition of NADPH oxidase activity by lipid transfer (Saez et al., 2000). Prostasomes enhance sperm motility (Fabiani et al., 1994a, b;Arienti et al., 1999;Wang et al., 2001) and influence sperm capacitation (Cross and Mahasreshti, 1997;Pons-Rejraji et al., 2011;Piehl et al., 2013) and acrosome reaction (Cross and Mahasreshti, 1997;Palmerini et al., 2003;Siciliano et al., 2008). These EVs play an essential role in supplying sperm with the essential Ca 2+ -signalling machinery required for the maintenance of sperm motility, capacitation and acquisition of its specific patterns associated with fertilization. The Ca 2+ -signalling tools which are currently considered to be provided to sperm by prostasomes, include the enzymes adenosine diphosphate (ADP) ribosyl cyclase (CD38), V-ATPase A1, secretory pathway Ca 2+ ATPase, progesterone receptors, ryanodine receptors (Park et al., 2011) and free Ca 2+ . This molecular machinery, acquired by sperm during fusion with prostasomes, has been shown to play an essential role in the promotion of sperm hyper-activation, a crucial motion pattern required for penetration of the zona pellucida, and the acrosome reaction (Park et al., 2011). The mechanism of prostasome-sperm transport appears to be similar to the epididymosome-sperm exchange of molecular compounds. Similarly to epididymosomes, the prostasomes have been shown to fuse with sperm at acidic or low pH (Arienti et al., 1997;Palmerini et al., 1999), and deliver the molecular cargo. Apart from the proteins, prostasomes have been recently demonstrated to contain fragments of DNA randomly selected from the genome and rapidly, within 15 min of co-incubation, transfer these fragments into the sperm head, neck and tail under the physiological vaginal pH-an important observation, highlighting yet another potential biological role of these mediators of cell communication (Ronquist et al., 2011).

Emerging evidence of the role of EVs in regulating female reproductive function
More recently, EVs have been identified in the follicular fluid of ovarian follicles where they have been shown to harbour a wide array of granulosa and cumulus cell-derived miRNAs and common exosomal and cell-type-specific proteins, suggesting a role in cell communication within mammalian ovarian follicles and in the regulation of follicular maturation (da Silveira et al., 2012(da Silveira et al., , 2014Sohel et al., 2013;Santonocito et al., 2014). Exosomes have also been isolated from mammalian oviducts (Al-Dossary et al., 2013;Alminana et al., 2014) and uterine cavity fluid (Ng et al., 2013;Ruiz-Gonzalez et al., 2015), and in all cases have been shown to transport molecular cargo (miRNAs/proteins) with important functions for fertilization, implantation and early pregnancy or the induction of specific sperm motility patterns, required for penetration of the zona pellucida ('hyperactivation'). In the same way as pregnancy-specific exosomes and STBMs, EVs in the male and female reproductive tract are structurally heterogenous and form multiple subpopulations (Poliakov et al., 2009;Aalberts et al., 2012;Brouwers et al., 2013;Caballero et al., 2013). It is hypothesized that different subpopulations of these EVs have different functional roles, however the exact mechanisms involved remain to be characterized.
Interestingly, the powerful effects of natural EV-mediated paracrine regulation were intentionally exploited by Saadeldin et al. (2014), who reported that the co-culture of cloned porcine embryos, produced  using the nuclear transfer technique, improved substantially during co-culture with parthenogenetic embryos releasing EVs containing multiple pluripotency gene mRNAs, which could be internalized into the cloned embryos. This promising observation further strengthens the hypothesis that EVs could represent an attractive candidate for intracellular delivery in reproductive biology and medicine, and improve the efficacy of existing investigative and therapeutic techniques.

EV-mediated delivery: bridging the gap between nature and nanotechnology
The concept of using EVs therapeutically is the focus of intense research in the fields of cancer treatment, regenerative medicine and infectious, inflammatory and neurodegenerative diseases (El-Andaloussi et al., 2013). The beneficial effects of EV-mediated paracrine signalling form the cornerstone of mesenchymal stromal cell-based therapies for the treatment of neural, cardiac or acute generalized tissue damage (Cashman et al., 2013;Monsel et al., 2014;Xin et al., 2014). These cells are characterized by strong paracrine activity rather than differentiation potential, and produce large amounts of EVs, potentiating cell proliferation, regenerative reprogramming, angiogenesis and immunomodulation in affected areas (El-Andaloussi et al., 2013). In addition to these approaches involving the indirect uses of EVs, an increasing number of studies describe the intentional production of EVs for subsequent use as drug delivery platforms. There is mounting evidence that EVs are naturally produced by a variety of cell types (El-Andaloussi et al., 2012), and can be purified from culture media using relatively straightforward ultracentrifugation protocols, labelled with fluorescent probes (Nazarenko et al., 2013;Takahashi et al., 2013) and loaded with molecular cargo via co-incubation (Sun et al., 2010) or electroporation (Alvarez-Erviti et al., 2011;El-Andaloussi et al., 2012;Tian et al., 2014). Furthermore, cells can be engineered through transfection to secrete modified EVs, either loaded with specific cargo (Akao et al., 2011;Mizrak et al., 2013) or expressing targeting moieties on their surface, which allows users to direct EVs towards a particular cell population in order to improve selectivity and range of action Exosomes and MVs can be isolated by centrifugation and loaded with molecular cargo (nucleic acids (NAs), proteins, small molecules, etc.) through co-incubation or electroporation. Loaded EVs and MVs can then be applied to sperm, oocytes and embryos in vitro to promote the internalization of molecular cargo. 'Loaded' sperm can be subsequently used for downstream applications, including sorting, fertilization, or sperm-mediated delivery of genetic constructs, proteins or small molecules into the oocyte at the time of gamete fusion. Addition of loaded EVs to in vitro culture systems could promote the maturation of oocytes, protect oocytes from degradation and improve the developmental parameters of early-stage embryos. IUI, intrauterine insemination; IVM, in vitro maturation of oocytes. El-Andaloussi et al., 2012;Ohno et al., 2013;Tian et al., 2014). The expression of targeting moieties on the EV surface could be extensively explored in reproductive science, especially considering emerging evidence that certain CPPs promote the internalization of compounds into gametes and embryos and improve the outcome of in vitro culture Kwon et al., 2013;Yang et al., 2014a;Barkalina et al., 2015). Interestingly, in response to growing interest in EV-mediated drug delivery, several research groups have proposed alternative approaches for EV production, involving the passage of source cells through a series of filters (Jang et al., 2013;Jo et al., 2014b) or microfluidic channels  to create artificial exosome-mimetic nanovesicles. These nanovesicles have been demonstrated to have a similar composition and delivery capacity to secreted EVs, and contain mRNAs and intracellular and plasma membrane proteins.
Thus far, purified and loaded EVs have been successfully applied for the delivery of anti-inflammatory compounds into activated myeloid cells and microglial brain cells as prototype treatments for autoimmune/inflammatory diseases (Sun et al., 2010;Zhuang et al., 2011), targeting of chemotherapeutics, suicide gene mRNAs/proteins, miRNAs and investigative therapeutic cancer vaccines towards malignant cells (Rountree et al., 2011;Mizrak et al., 2013;Ohno et al., 2013;Tian et al., 2014), and the targeted systemic delivery of small interfering RNAs into the brain as an experimental therapy for Alzheimer's and Parkinson's disease (Alvarez-Erviti et al., 2011;El-Andaloussi et al., 2012;Cooper et al., 2014). In these studies EVs have been consistently shown to have highly promising features for intracellular delivery, including high specificity and selectivity, and significant potential for the systemic delivery of experimental agents with otherwise unfavourable biodistribution profiles.

Figure 2
Translational aspects of EV-mediated transfer of molecular compounds into gametes and embryos in vitro: potential applications. EVs loaded with molecular cargo could be applied for (a) targeted delivery of sperm motility-activating compounds to facilitate fertilization during IVF; (b) loading of sperm, characterized by molecular deficiencies of oocyte-activating factors (SOAFs), with novel and more physiological exogenous compounds acting as oocyte activators (recombinant versions of SOAFs, small molecules with similar activity, etc.) for subsequent delivery into the oocyte at the time of fertilization and assisted oocyte activation (AOA) as an alternative to currently applied physical, mechanical or chemical AOA; (c) facilitation of spermmediated gene transfer-a phenomenon based upon the property of mammalian sperm to bind and incorporate exogenous DNA upon co-incubation in vitro and deliver into the oocyte at the time of fertilization to produce genetically-edited embryos; (d) direct supplementation of oocytes with molecular compounds, enhancing their developmental capacity, and increasing the chances to sustain early embryo development; (e, f) delivery of gene editing tools into the oocytes or embryos, respectively, as a tool to treat hereditary diseases. The use of EVs as versatile delivery tools could allow us to bypass invasive micromanipulation procedures, which are currently considered the gold standard for gamete manipulation in assisted reproduction. Extracellular vesicle for intra-gamete delivery in ART Collectively, the relative ease of production, biodegradability and the possibility for targeting and loading with a variety of compounds relevant for reproductive biology and science (including small molecules, nucleic acids and proteins) along with the enormous physiological role of EVs in cell communication processes underlying reproduction, as well as encouraging evidence from other fields, render these natural 'nanoplatforms' particularly attractive candidates for compound delivery into gametes and embryos (Figs 1 and 2).

Conclusions
Today, ART is viewed not only as a routine approach for infertility treatment but also, progressively, as a state-of-art guarantee of successful conception and childbirth at any chosen time in an individual's life. However, despite the ever-growing success of ART in the field of infertility treatment, this technique still only results in the live birth of healthy children in approximately 30% of couples starting treatment. The increasing reliance of medical practitioners and the general public upon ART justifies further extensive investigation into the fundamental mechanisms of reproduction to improve our existing levels of knowledge and facilitate the continuous improvement of current medical technology. However, it is highly unlikely that a substantial breakthrough in the field of reproductive science and medicine can be achieved without the discovery of novel research tools that allow us to systematically study and manipulate gamete and embryo function in a real-time setting, while fully preserving their viability. In recent years, biomedical nanotechnology has offered potent solutions to the problem of delivering molecular compounds into gametes. Nevertheless, the predominant use of non-biodegradable nanoparticles to promote the uptake of DNA and proteins into gametes continues to raise concerns, which limits the use of these techniques to a purely investigative platform. To allow the subsequent translation of these methodologies to clinical ART, studies should utilize biodegradable delivery platforms, which mimic natural mechanisms of molecular cargo trafficking as closely as possible. In this view, the field of reproductive science could be substantially advanced by actively developing EV-mediated delivery technology.
Currently, EVs represent the most physiological intracellular delivery tools available for reproductive science and medicine. These natural universal mediators of cell communication combine the benefits of engineered nanomaterials, such as the potential for in vitro production, targeting and loading, with the essential feature of biodegradability. Furthermore, the high degree of involvement of EVs in the essential processes underlying gamete maturation, the acquisition of fertilization potential and the establishment/maintenance of pregnancy, renders the use of similar 'modified' nanoplatforms particularly exciting. We anticipate that future investigations into the possibility of applying EVs for the intentional intracellular delivery of molecular compounds into gametes and embryos will open new horizons for reproductive science and clinical ART, ultimately leading to significant improvements in patient care.