The mammalian pregnancy, in particular the human hemochorial pregnancy, is the highest and most complicated mode of reproduction that puts a great strain to the maternal organism and comprises a biological challenge to create an immunologic tolerance to the fetal allograft and support its development and growth. The key for successful mammalian reproduction, the placenta, is a transient organ that mediates not only hormonal, nutritional and oxygen support to the fetus but also actively secretes signal substances and immunoregulatory factors that modulate/alter the maternal immune response during pregnancy. Recent reports1–6 have shown that the main and unique cell type in the human placenta, the syncytiotrophoblast, releases biogenetically and morphologically defined, nanometer-sized microvesicles called exosomes, which are used in the fetal–maternal cross-talk for adaptation of the maternal organism to the ongoing pregnancy.
Cell-to-cell communication is of crucial importance for all living organisms. Cells may communicate and exchange information by different mechanisms: (i) via soluble mediators such as hormones, cytokines, chemokines, bioactive ions and lipids that are released from cells and act in an autocrine or paracrine manner; (ii) via direct adhesion contacts between signaling and target cells, including trogocytosis; (iii) exchanging information through intercellular nanotubules.7,8 There is, however, emerging evidence that a fourth mechanism built on release and uptake of membrane-bound microvesicles (MV) exists. Cell–cell communication by MV comprises a combination of secretion of molecules on one hand and preservation of their membrane-association on the other hand. Secreted MV execute cell-to-cell contact ‘by proxy’ delivering signals and transferring packages of information from a donor cell to a target cell locally, and/or at a distance. The smallest members of the MV family, the exosomes, are nanometer-sized and actively secreted by most cell types throughout the body. Their functions are diverse and related to the physiological functions of the donor cells.9 Most published reports demonstrate their immunomodulatory features (reviewed in ref. 10,11).
Microvesicles’ secretion in the placenta occurs both by blebbing/shedding of syncytiotrophoblast microparticles (STBM) from the plasma membrane12–14 and by exosome release through the endosomal pathway.1–6 STBM are 0.2–2 μm in size and are present in a low plasma concentration in normal pregnancy. They are distinct from exosomes and comprise a heterogeneous population, involving shed microvilli and apoptotic fragments. Enhanced shedding of STBM promotes endothelial and immune cell dysfunction, inflammation and necrosis/apoptosis and is somehow associated with the placental pathophysiology in preeclampsia.12–14 Distinguishing between STBM and exosomes as separate bioactive entities is required to determine precisely their role in reproduction. In this review, we will focus on the placenta-derived exosomes and discuss their role in normal pregnancy. In the following text, a brief characterization of various microvesicles and a description of the biogenesis and nature of exosomes are given as a background.
Microvesicles are everywhere
Although various membrane-bound MV have been observed for years in the intercellular space by electron microscopy, the existing general dogma considered them as an inert cellular debris/dust that should be more or less ignored. Only recently has it become clear that the MV are not a uniform entity but differ in size, morphology and nature and can be either shed from the plasma membrane or secreted through the endosomal membrane compartment. The realization that various MV are found in health and disease, and that they might have different roles depending on their type and the cells from which they are derived, has opened new perspectives in understanding biological processes and their regulation. To state that MV are everywhere, in the blood and all bodily effusions, is not an overstatement. In fact, they are an integral part of the intercellular environment. MV are a heterogeneous population that vary in size (0.03–2 μm), shape and composition depending on the cellular sources and how they are generated. The MV can be divided by size into large (0.1–2 μm) and small (30–100 nm). Two main types comprise the large MV: (i) those produced by blebbing of the plasma membrane upon programmed cell death called apoptotic bodies/apoptotic blebs/apoptotic vesicles that are a product of dying cells;15,16 and (ii) those produced by direct budding from the plasma membrane of living cells called microvesicles (sometimes named microparticles or ectosomes). Examples of the latter are MV shed by plasma membrane of various epithelia17–19 including syncytiotrophoblast, such as shed microvilli from cell surface membranes; prostasomes,20 budded from the apical plasma membrane of prostate epithelium; and prominosomes,21 prominin-1-expressing plasma membrane microvesicles from brain stem cells. The small-sized MV (<100 nm) called exosomes, which represent an apparently distinct class of MV, arise from the endosomal membrane compartment, have a definite morphology and are more homogeneous in their size and biological composition. MV, including exosomes, are produced by many hematopoietic and non-hematopoietic cells, such as reticulocytes, mast cells, T and B cells, platelets, dendritic cells, neurons and microglia, various epithelial cells like enterocytes from small and large intestine, uroepithelia, bronchial epithelia, syncytiotrophoblast, hepatocytes as well as tumor cells.10,11,19 MV have been isolated from blood and various bodily fluids such as saliva, urine, amniotic fluid, malignant effusions, bronchial lavage fluid, synovial fluid, and breast milk (reviewed in ref. 22). The major part (about 80%) of MV, constitutively present at low levels in the peripheral blood of healthy subjects, are platelet derived.23 The rate of clearance of circulating MV in humans is presently not known. In rat experiments, 80% of labeled MV were cleared from the circulation within 5 min.24 Various populations of microvesicles and some of their characteristics regarding morphology, size and genesis are presented in Table I.
|Characteristics||Exosomes||Microvesicles/microparticles||Shed microvilli||Apoptotic bodies/vesicles|
|Size||30–100 nm||0.1–2 μm||>400 nm||100–600 to700 nm|
|Density in sucrose||1.13–1.19 g/mL||Undetermined||Undetermined||1.16–1.28 g/mL|
|Morphological shape||Cup-shaped, electron translucent||Various shapes, electron-dense and/or electron translucent||Various shapes, round, elongated and cylinder-like||Irregular and heterogeneous in shape|
|Membrane composition||Cholesterol-, sphingomyelin-, and ceramide-rich lipid rafts, expose phosphatidylserine||Expose phosphatidylserine, some enriched in cholesterol and diacylglycerol, some undetermined||Undetermined||Undetermined|
|Specific marker(s) for identification||Tetraspanins (CD63, CD9, CD83), ESCRT complex members (Alix, TSG101)||Integrins, selectins, CD40 and others, depending on the cell type||Various, depending on the cell type||Histones, DNA|
|Origin in the cell||Multivesicular bodies (MVB)||Plasma membrane||Plasma membrane||Fragments of dying cells, undetermined|
|Mode of release/secretion||Fusion of MVB with the plasma membrane||Plasma membrane blebbing||Plasma membrane blebbing||Plasma membrane blebbing and cellular fragmentation|
General description of the biogenesis, structural organization and biological properties of exosomes
History of exosome discovery
The first reports of exosome-like microvesicles produced by ovarian cancer and neoplasic cell lines appeared about 30 years ago.25,26 Independently, Johnston et al. discovered that normal reticulocytes used microvesicles to eliminate the transferrin receptor and made the important discovery that they were secreted from the endosomal compartment.9,27 Since then, for many years, exosomes remained in oblivion and only in the past decade were they ‘rediscovered’ as a new and exciting ‘fourth mechanism of cellular communication’ with a powerful influence and a biological role.
Definition and biogenesis
Exosomes are defined as secreted membrane-bound nanovesicles that are identified by the following characteristics: (i) cup-shaped form; (ii) 30- to 100-nm size; (iii) buoyant density of 1,13-1,19 g/mL on sucrose gradient; (iv) endosomal origin; (v) tetraspanins present in their lipid raft–rich membrane.10 The suggested biogenesis of exosomes, illustrated schematically in Fig. 1, separates them from all other MV so far known. They are formed in the late endosomal membrane compartment by inward budding of the limiting membrane of late multivesicular endosomes/multivesicular bodies (MVB) and contain cell surface–expressed proteins and cytosolic components. They are actively secreted into the extracellular space by fusion of the MVB with the plasma membrane.
The mechanism(s) of protein sorting during exosome formation is not fully understood. It is suggested that proteins, carried by exosomes, are sorted to the MVB in two ways: (i) by endocytosis and transport of plasma membrane proteins to the early recycling- and further to the late endosomal compartment; and (ii) by direct transportation from the Golgi complex to MVB and insertion in the MVB membrane to bud as exosomes. For many years, MVB were considered to be the late stage in the maturation of endosomes to lysosomes –‘garbage stations’ for proteins destined for destruction. However, approximately two decades ago, an alternative function for the MVB was described, namely their ability to move to the plasma membrane and fuse with it. After fusion, the MVB secrete their intraluminal vesicle cargo in the extracellular space as exosomes.10 Thus, the MVB are situated at a ‘cross-road’ in the endosomal pathway, where the fate of the proteins sorted to the MVB is decided – either secretion to the extracellular space as exosomes or degradation in the lysosomal ‘dustbin’.28 Accordingly, two classes of MVB are proposed – degradative MVB that evolve to lysosomes and exocytotic MVB that fuse with the plasma membrane. Recently, it was shown that different biochemical properties may govern the alternative fates of these two types of MVB.29 The mechanisms governing degradative MVB involve the well-characterized multiprotein network called endosomal-sorting complexes required for transport (ESCRT complexes) and an ubiquitinylation process resulting in ubiquitin tagging of both cell surface proteins and intracellular proteins targeted for lysosomal degradation.28,30–32 The mechanism underlying exocytic MVB trafficking is less clear. The transmembrane protein TSAP6 has been suggested to regulate exosome production,33 and Rab11, a member of the small GTPase family, together with calcium was shown to be important for the docking and fusion of MVB with the plasma membrane.34 It is, furthermore, suggested that sorting of proteins for exosome release might be ubiquitin independent and instead involve sphingomyelin metabolites, such as ceramide as well as the ESCRT multiprotein complex.29 Although considerable progress is made to reveal the key links between ubiquitin, phospholipids and the ESCRT proteins, more studies are needed to elucidate the biogenesis of exosomes.
Morphology and general biochemical composition
Electron microscopy (EM) is a superior methodological approach for studies of exosome morphology and biogenesis. Exosomes, visualized in situ by EM, are uniform spherical vesicles 50–90 nm in size situated within the lumen of MVB. The EM image of isolated exosomes is different – they are typically uniformly cup-shaped but heterogeneous in size, varying between 30–100 nm. The reason for this appearance (Fig. 3a), currently used as the morphological hallmark for isolated exosomes, is not known but might be a consequence of the isolation procedure. The exosomal membrane is detergent resistant with a lipid raft–rich bilayer, built up of cholesterol, sphingolipids and tetraspanins, where proteins with transmembrane or GPI linkage are inserted.35 As other MV, they expose phosphatidylserine, but at low level, on their surface and can be captured by phospatidylserine receptors expressed on surface of activated T lymphocytes and phagocytes.36 The molecular content of exosomes includes proteins and the ribonucleic acids mRNA and micro RNA (miRNA) and is dependent on the tissue/cell type from which the exosomes originate.37 Nearly all exosomes, regardless of their origin, carry a conserved set of proteins. Examples of commonly found exosomal proteins are cytoplasmic proteins such as tubulin, actin, actin-binding proteins, annexins and Rab proteins; the heat shock proteins hsp70 and hsp90; signal transduction kinases, heterotrimeric G-proteins; members of the ESCRT complex Alix and TSG101. Common surface–expressed proteins are MHC class I molecules; adhesion molecules such as β integrins, and ICAM-1, and the class of proteins that have become exosomal markers – the tetraspanins CD9, CD63, CD81 and CD82 (reviewed in ref. 37).
Advantages of exosome-mediated protein secretion and suggested functions of secreted exosomes
Exosome release seems to be a powerful way of intercellular communication. It has emerged as a ‘non-classical’ form of secretion compared to the ‘classical’ excretory pathway. The advantages of exosome-mediated protein secretion are many: (i) preservation of the three-dimensional structure of the transported proteins and thus their biological activity; (ii) independence of cell-to-cell contact for signal delivery; (iii) packages of carried molecules give lower mobility and higher concentration of the carried molecules; (iv) independence of de novo protein synthesis; (v) biological effect at a distance.
A variety of important biological functions like intercellular signaling, antigen presentation, immune regulation, pro- or antiapoptotic effects, delivery of protein molecules via binding to the target’s plasma membrane or internalization by endocytosis and transport of bioactive RNA between cells19,38–40 have been ascribed to exosomes. A major role of secreted exosomes is their powerful effect on the immune system that divides them into two groups – exosomes with immunoactivating properties and those with tolerogenic or even immunosuppressive effect. In general, exosomes produced by antigen-presenting cells such as dendritic cells (DC), macrophages and B cells are immune activating. They activate directly or indirectly, via DC or cytokine production of T helper cells, immune effector mechanisms such as cytotoxicity, antibody and cytokine production, and priming of T cells.19 Importantly, in activated T and NK cells, the cytotoxic molecules perforin, granzyme, the proapoptotic molecules FasL and TRAIL are carried by endosomal vesicles of specialized MVB, also named secretory lysosomes.2,41,42 On the other side, exosomes produced by epithelial cells and the great majority of tumors are immune inhibitory.43,44 Normal epithelial cell-derived exosomes exert an immunosuppressive effect promoting homeostasis and immune tolerance.43 Remarkably, many tumors ‘hijack’ features of the immune cells by expression and release of ligands to important immune receptors and/or key immune molecules and signal substances on exosomes that are used as decoys, disturbing and down regulating normal immune effector mechanisms or enhancing the suppressive function of T regulatory cells.44,45 Thus, the net effect of the function of tumor exosomes originating from mammary, lung, colon, prostate and ovarian cancer46 is a powerful inhibition of the host immune system promoting the establishment of the primary tumor and its metastases. Placenta-derived exosomes, like tumor-derived ones, are immunosuppressive.
Secretion of placental exosomes – a constitutive feature of human normal pregnancy
Biogenesis of placental exosomes
Taylor et al. did some pioneering work on describing circulating ‘shed placental fragments’47,48 that later they characterized as exosomes and studied their role in the down regulation of T-cell responses of pregnant women.3,4 Mor et al.1 reported that isolated trophoblast cells from first trimester pregnancy lacked plasma membrane–associated FasL but expressed cytoplasmic FasL that was secreted via microvesicles. We confirmed these findings at the ultrastructural level, and using immunoelectron microscopy (IEM), we provided the first data of the biogenesis of placental exosomes demonstrating that the intracellular FasL in human placenta was concentrated to MVB in the syncytiotrophoblast and expressed and secreted by microvesicles/exosomes of 60–100 nm size.2 Since then, we have studied the biogenesis, composition and role of syncytiotrophoblast-derived exosomes in situ and in vitro from placental explant cultures. We studied exosome biogenesis in the syncytiotrophoblast of human placenta by IEM expression analyses of FasL and the two families of the NKG2D receptor ligands, i.e. the MHC class I chain–related antigens A and B (MICA /B) and the human retinoic acid early transcript 1 (RAET1) proteins also called ULBP 1-5, as marker molecules of endosomal compartments.2,5,6 We were surprised to find that in case of FasL and ULBP, the protein expression was absent on the surface of the syncytiotrophoblast and instead found only in the late endosomal compartment, where the proteins were located on the MVB’s limiting membrane and on numerous intraluminal vesicles with the size of exosomes (Fig. 2a). MVB, filled with exosomes carrying a particular protein, could be found at different levels in the syncytioplasm and frequent fusion of MVB with the apical microvillous surface and release of their vesicles into the intervillous extracellular space could be observed as illustrated in Fig 2b. In contrast to FasL2 and ULBP/RAET1 proteins (Fig. 2a, and ref. 6), which were typically retained in late endosomes/MVB (Fig. 2a) and released as exosomes, the MICA and B molecules were expressed both on the apical cell surface and inside MVB (Fig 2b, ref. 5). At present, the mechanisms regulating the expression, distribution and turnover of proteins in the plasma membrane or MVB (exosomes) are not clear. It has been suggested that in some cell types, post-transcriptional mechanisms exist that regulate the levels of NKG2D ligands at the cell surface by retaining them in intracellular compartments, particularly in MVB.49 Indeed, it has very recently been shown in several cellular systems that MICB molecules have a very short half-life at the plasma membrane but are accumulated in late endosomal compartment.50 It is possible that sorting of MICB, FasL, ULBP4 and 5 to MVB is controlled by ubiquitination as they possess lysine residues within their cytoplasmic tail.51,52 In Fig. 2d, syncytiotrophoblastic MVB, stained with the ESCRT member TSG101, is shown, supporting this suggestion. In contrast, GPI-linked ULBP1-3 and the transmembrane MICA molecule from 008 locus, which is the most frequent allele in human populations and has truncated cytoplasmic tail, are preferentially expressed in lipid rafts at the cell surface.53,54 It has been suggested that these raft domains may support the sorting of proteins to MVB and formation of exosomes.54,55 From these studies can be concluded that both plasma membranal and exosomal protein expressions are used in the syncytiotrophoblast. The conspicuous endosomal membrane compartment, represented primarily by MVB, and the frequent observation of MVB fused with the apical plasma membrane suggested that exosome secretion might be a constitutive feature of the syncytiotrophoblast and seemed to be preferentially used. It is obvious that the choice/preference of exosomal release of FasL, ULBP and, partially, of MICA/B over plasma membranal expression is of crucial importance for the protection of the fetal allograft. Surface expression of FasL in the placenta would be easily involved in induction and promotion of inflammatory response at the fetal–maternal interface,56 while exosomal FasL is proapoptotic and promotes immunotolerance.1–4 Similarly, a strategy of releasing NKG2D ligand-carrying exosomes from placenta would be a decoy mechanism to downregulate the cognate cytotoxic receptor; in contrast to a plasma membranal ligand expression that would make the syncytiotrophoblast a target for attack by NKG2D receptor-bearing maternal lymphocytes. In conclusion, our IEM data show that the exosome biogenesis in the syncytiotrophoblast follow that of other cells and occurs within MVB. As a vigorous protein producer, the syncytiotrophoblast is an excellent model for studies of the biogenesis of exosomes.
Composition, structure and functional properties of secreted placental exosomes
Phenotypic and functional studies of isolated exosomes demand stringent purification procedures. The choice of isolation method is of crucial importance for obtaining ‘pure’ exosome populations, i.e. free from contaminating proteins, non-exosomal microvesicles/microparticles and apoptotic bodies. This is even more important when studies of exosomes in pregnancy are performed, as it is well known that the syncytiotrophoblast in vivo constitutively generates considerable amounts of microvesicles/microparticles.13 The most used method, consisting of a series of centrifugations to remove dead cells/debris followed by ultracentrifugation, will pellet the exosomes in a mixture with other MV. For enrichment and purification of the exosomes, additional ultrafiltration and continuous sucrose gradient (floating density 1,13-1,19) or ultracentrifugation with sucrose cushions57,58 should be performed. Another method is using adherence to magnetic beads (MACS, Miltenyi Biotec, or Dynabeads, Dynal A/S) coated with specific markers expressed on the exosome surface, followed by elution and ultracentrifugation.59,60 New methods improving the yield and purity of exosome isolation are under constant development.61 Only then can phenotypic and functional studies produce reliable results.
Another consideration is how fluorescence-activated cell sorting (FACS) is used for exosome phenotyping. In some reports, a direct FACS analysis of ultracentrifuged MV pellets is used for phenotypic characterization of exosomes. The discriminative capacity of FACS is down to 300 nm,62 i.e. a size three times the upper limit for the size of exosomes. Thus, a FACS analysis of directly stained ultracentrifuged pellets63,64 will exclude the exosome fraction. The use of FACS for exosome phenotyping must be proceeded by loading of exosomes, directly or via antibody capture on artificial beads.6,59 In conclusion, it is essential to elaborate on and choose appropriate methods in all work with exosomes. Most studies today focus on quantity, cell origin and biological activity of MV without specific analysis of the inherent organization and the purity of the microvesicle/microparticle populations studied. We strongly advise against work63–67 where an isolated crude mixture of all kind of MV, apoptotic bodies, exosomes and other particles is included in phenotypic or functional analyses.
Morphology and protein composition of placental exosomes
Taking the methodological consideration described earlier, only studies with reliable exosome-derived data will be reviewed here. There are few investigations into isolated pregnancy-specific exosomes so far. They comprise studies of exosomes isolated from the peripheral blood of pregnant women performed by the group of Taylor et al.3,4 and studies of exosomes isolated from ex vivo placental explant cultures from first trimester normal pregnancy performed by our group.2,5,6 An illustration of the morphology, size and some proteins present on the membrane of placental exosomes isolated from explant cultures is given in Fig. 3. As can be seen, they have the typical exosomal cup shape and a size of 40–90 nm. They express the exosome-specific marker tetraspanin CD63 and the placenta-specific enzyme placental-type alkaline phosphatase indicating their origin. In contrast to other exosomes, such as those derived from immune cells, placental exosomes lack MHC molecule expression. Instead, the MHC-related molecules MICA/B and RAET1/ULBP1-5, ligands of the activating NK cell receptor NKG2D, were expressed on their surface (Fig. 3). Furthermore, the placental exosomes express the proapoptotic molecules FasL1–4 and TRAIL. Interestingly, Western blot analyses revealed that these molecules were aggregated on the exosomal membrane in a functional trimeric form which is required for Fas signaling (LM-N, manuscript in preparation). In addition, we found that the membranal form of the regulatory cytokine TGFβ is also expressed on the placental exosome membrane (unpublished result). Based on these studies, a schematic drawing of a ‘typical’ placental exosome with some of its protein composition is presented in Fig. 4.
The entire protein composition of different exosomes is not yet known and that is also valid for placental exosomes. We have recently completed a preliminary proteomic analysis of exosomes isolated from supernatants of placental explant cultures, and in Table II are listed some of the proteins, commonly identified in exosomes, which were present in placental exosomes as well. The detailed presentation of the data is currently under processing (LM-N, manuscript in preparation). Largely, there were two groups of proteins (i) those present on the exosomal membrane and (ii) those entrapped in the lumen. The proteomic analysis revealed some membrane-associated proteins like the tetraspanins CD9 and CD63 and proteins involved in adhesion and targeting (integrins, CD47, transferrin receptor, epidermal growth factor receptor), but the great majority were cytosolic such as cytoskeleton and cytoskeleton-binding proteins (tubulin, actin, cofilin1, profilin1); proteins involved in protein biosynthesis and degradation such as ribosomal proteins, elongation factors and proteasomes; proteins involved in intracellular transport, fusion and signal transduction such as annexins, Rab- and Ras-related proteins, vesicle transport proteins, small GTP-as family members, clatrin, dysferlin, syntaxin; heat shock proteins and chaperons (HSP27, 70 and 90); ESCRT-associated proteins such as TSG101, ALIX, vascular sorting protein 29 and charged MVB protein 1B and 4B; and enzymes such as α-enolase, 5′nucleotidase and dipeptidyl peptidases.
|Heat shock proteins chaperons||HSP27|
|Vacuolar sorting protein 29 (ESCRT)|
|Charged MVB proteins 1B and 4B (ESCRT)|
|Apoptosis regulation||Programmed cell death proteins 6 and 10|
|Protein biosynthesis and degradation||60S ribosomal proteins|
|40S ribosomal proteins|
|Elongation factors 1-α1, α2, α3 and γ|
|Proteasome α4 subunit|
|Proteasome α5 subunit|
|Proteasome 26S non-ATPase subunit|
|Adhesion, targeting||Integrins α5, αV, β1, β3|
|Epidermal growth factor receptor|
|Signal transduction||14-3-3 proteins|
|Rab 1A, 1B, 35|
|Ras-related proteins 1B and R|
|Guanine nucleotide binding protein|
|Ras GTPase-activating protein|
|Transforming protein RhoA|
|Membrane transport and fusion||Annexins|
|Rab proteins: 2A, 5A, 5B, 5C, 6, 7, 10, 14|
|Clatrin heavy chain|
|Vesicle transport through interaction protein 1B|
|Multidrug resistance protein 1|
|Lysosomal membrane protein 2|
Identification of RNA in trophoblast-derived exosomes
One of the most intriguing and powerful role of exosomes is their transport of selected mRNAs and microRNAs (miRNA) and thus their ability to transfer genetic information to target cells and regulate their cellular metabolic pathways.7,39,40 These findings suggest that exosomes may represent a new mechanism of exchange of genetic information between cells. Moreover, the possibility to analyze mRNA and miRNA patterns in exosomes generated in health and diseases opens a great opportunity to refine disease diagnostics.40,68,69
Very recently, the first detailed miRNA profile of human placenta was reported.70 In this study, performed on human early and term placental tissue, was demonstrated that most placenta-specific miRNAs were linked to a miRNA cluster on chromosome 19. These miRNA cluster genes were upregulated in placental development. Six novel miRNAs were identified, and of those only four were expressed in placenta. Simultaneously, miRNAs were identified in maternal plasma, a finding supporting a previous report by Chim et al.71 on placental miRNA presence in plasma. The authors suggested that syncytiotrophoblast secretes miRNA in the maternal circulation by exosomes. To prove this notion, the trophoblast cell line BeWo was used as a model of villous trophoblast and exosomes were isolated from culture supernatant. Two placenta-specific miRNAs MIR517A and MIR21 were demonstrated in the exosome-enriched fraction. Proteome analysis suggested that MIR17A might be involved in the regulation of TNF signal transduction. It will be of interest to characterize the placental miRNAs enriched in placenta-derived exosomes in connection with normal and pathologic pregnancies, different fetal developmental stages or pregnancies with fetal abnormalities. Retrieval of such exosomal profiles from the maternal plasma will open new revenues for infertility and prenatal diagnostics. In summary, these results are promising first steps toward establishing the exosomal miRNA profiles in pregnancy. More investigations are needed to elucidate the role of the placenta-specific mRNA and miRNA and their capacity to enter and reprogram maternal cells in favor of fetal survival.
Functions associated with placenta-derived exosomes
Information about the function of placenta-derived exosomes is presented in few reports1–6 and is based on functional studies of placental exosomes isolated from (i) peripheral blood of pregnant women3,4 and (ii) supernatants of cultured trophoblast cells1 and placental explants.5,6 Taylor et al. has characterized placenta-derived exosomes in peripheral blood of pregnant women that completed normal-term pregnancies or had a preterm delivery. The exosomal concentration in plasma was elevated in women with normal pregnancies delivering at term in comparison with preterm delivery and non-pregnant women. Incubation of T cells (Jurkat cell line) with placental exosomes resulted in downregulation of the expression of CD3-ζ and Janus kinase 3 (JAK3) and activation of caspase 3. These responses affected the clonal T-cell selection and lead to impaired T-cell-mediated responses supporting the existing hypothesis of Th2 deviation of immune responses during pregnancy. The observed downregulation of the CD3-ζ chain correlated to FasL- and PD-L1 expression on the exosomes and could be explained by induction of apoptosis in target cells.4 These results are in line with the report of Mor et al.1 that trophoblast-derived FasL-expressing microvesicles have proapoptotic function. Furthermore, it was shown that exosomes from term-delivering mothers could inhibit IL-2 production by activated T cells, while the exosomes from preterm-delivering mothers could not.3,4
Our studies of exosomes from placental explant cultures included analyses of FasL2, TRAIL and the NKG2D ligands MICA/B and ULBP1-5.5,6 Initially, we suggested that exosome-associated MIC molecules were present in the serum of pregnant women and could suppress NKG2D receptor-mediated cytotoxicity.5 Thereafter, we demonstrated that isolated exosomes carried all NKG2D ligands and were able to downregulate the NKG2D receptor on NK-, CD8+- and γδT cells in a dose-dependent manner. The exosome-induced cellular internalization of NKG2D resulted in impairment of the receptor-mediated cytotoxicity without affecting the lytic potential of the cells measured by perforin mRNA expression and perforin protein content. Our results were indirectly confirmed in a very recent publication by Ashiru et al. who reported that cancer-derived exosomes expressed MICA and suppressed NK cell cytotoxicity in a similar way as placental exosomes, i.e. by downregulation of the NKG2D receptor.71 In our recent studies, we have found that placental exosomes display bioactive trimeric FasL and TRAIL and are proapoptotic. Using IEM, we have further proven that placental exosomes are indeed formed in the MVB/endosomal compartment by the syncytiotrophoblast of the explant cultures.2,5,6 Summarizing ours and others results, several functions for the placenta-derived exosomes can be proposed: (i) impairment of T-cell signaling; (ii) impairment of cytotoxicity by downregulation of the major activating NK cell receptor NKG2D and (iii) apoptotic activity through FasL-, TRAIL- and PD-L1-mediated pathways. Thus, it is clear that placental exosomes are involved in the control of critical immune mechanisms such as cytotoxicity, T-cell response and apoptosis in the local vicinity and/or at a distance from the fetal–maternal interface. These functions define the placental exosomes as inhibitory/immune suppressive, using in a redundant way a number of mechanisms that promote maternal immune tolerance of the fetal allograft.
Exosomes in amniotic fluid
Finally, we would like to briefly mention and comment on the fact that exosomes are present in and can be isolated from the amniotic fluid in human72–74 and murine74 pregnancies. Here again, one can find studies with insufficiently characterized amniotic microvesicles and speculative suggestions about their origin and functions.73 In the first report of amniotic exosomes,72 their connection with the fetal kidney was made, complying with the fact that the renal system of the fetus is the main contributor to the production of amniotic fluid. Moreover, the group of Keller et al.74 proved that these exosomes were released by the fetus, not the mother, and investigated their phenotype in human and murine pregnancies. The amniotic exosomes expressed CD24 as their specific address marker, annexin-1, and kidney markers such as aquaporin-2, and have a similar composition to exosomes from urine of newborn infants. At this stage, more studies are needed to evaluate if amniotic exosomes have a pregnancy-promoting role. We are inclined to think that their presence in the amniotic fluid is because of the fetal urine production rather than that they have a special role in the immunomodulation of the maternal immune system. However, we are open for convincingly proven suggestions otherwise. On the other side, the amniotic exosome discovery72,74 opens fantastic possibilities to monitor prenatal kidney development and develop prenatal diagnosis of kidney diseases and genetic malformations.
Final remarks and future directions
Intercellular communication by MV has opened new perspectives in understanding cross-talk mediated between donor/signaling and target cells. From this point of view, we can say that shedding MV from the plasma membrane or releasing exosomes from the endosomal compartment of the syncytiotrophoblast expands the boundaries of the maternal–fetal communication and affirms the role of the placenta as a powerful regulatory organ that uses MV-mediated signaling to target and reprogram maternal cells for the benefit of reproduction.
Among the developments we expect in the near future is the detailed differential characterization of the STBM and exosomes. New optimized and refined techniques for their separation are urgently needed. Applying stringency and accuracy in isolation procedures is the only way to ensure the possibility to reveal the typical protein and RNA pattern profiles of STBM and placental exosomes. Using isolated pure fractions of MV in functional experiments will clarify their functional differences and roles in normal and pathological pregnancies.
How do we envisage the placental exosomes and their role in pregnancy? From the accumulated knowledge so far is clear that the main cell type of the placenta, the syncytiotrophoblast, continuously and constitutively produces and releases exosomes that are used in the intercellular communication between the mother and the child throughout the pregnancy. Evidence summarized in this review shows that the placental exosomes are packages transporting important signaling molecules and genetic information to be delivered to specific targets locally or at a systemic level. They seem to be immunosuppressive in nature, and have the ability to shape the maternal immune system through different mechanisms, which makes them pluripotent. The exosomes directly secreted in the maternal blood are, probably, at the highest abundance in the intervillous space of the chorionic villi and decrease in concentration with increasing distance away from the placenta. Thus, the continuous release of exosomes by the syncytiotrophoblast creates an exosomal concentration gradient where the protection against maternal immune attack is strongest at the fetal–maternal interface, i.e. in the immediate vicinity of the chorionic villi. One can imagine that the fetus, together with the placenta, is surrounded by a ‘cloud of exosomes’ that modulate the maternal defense and create a protective and beneficial milieu for the fetus.
The exciting field of placental exosome research is still just at its beginning. An important progress will be the exact identification of the mechanisms that govern the exosomal biogenesis – molecular sorting, inward MVB membrane budding and mechanisms of placental exosome release. How is the fate of MVB decided: why and how do some MVB become degradative and others exocytotic? New molecules and additional exosomal functions will be elucidated and added. More such analyses will be performed to come to a consensus regarding the mRNAs’ and miRNAs’ role and contribution to the control of successful pregnancy.
Finally, the role of exosomes in the pathogenesis of pregnancy disturbances and related diseases, recurrent abortions and infertility waits to be evaluated.
Pregnancy is not the sole condition gaining from studies of placental exosomes. Understanding how the well-being of the fetal allograft is created can benefit transplantation. Moreover, many proliferative, invasive and immune tolerance mechanisms that support normal human pregnancy are also exploited by malignancies to establish a nutrient supply and evade the host immune response. Placenta-derived and tumor-derived exosomes share similar composition and properties. Thus, release of exosomes that can edit immune responses to promote survival and well-being of the fetus or the tumor, respectively, might be one common link between the physiological state of pregnancy and the pathological state of cancer. Knowledge, gained from exosome research in reproduction, could lead to identification of novel diagnostic and therapeutic approaches in both conditions.