2019
Extracellular vesicles are integral and functional components of the
extracellular matrix
Rilla, Kirsi
Elsevier BV
Tieteelliset aikakauslehtiartikkelit
© Elsevier B.V.
CC BY-NC-ND https://creativecommons.org/licenses/by-nc-nd/4.0/
http://dx.doi.org/10.1016/j.matbio.2017.10.003
https://erepo.uef.fi/handle/123456789/7548
Downloaded from University of Eastern Finland's eRepository
Kirsi Rilla, Anne-Mari Mustonen, Uma Thanigai Arasu, Kai H¨ark¨onen, Johanna Matilainen, Petteri Nieminen
PII: S0945-053X(17)30328-1
DOI: doi:10.1016/j.matbio.2017.10.003 Reference: MATBIO 1363
To appear in: Matrix Biology Received date: 22 August 2017 Revised date: 10 October 2017 Accepted date: 16 October 2017
Please cite this article as: Rilla, Kirsi, Mustonen, Anne-Mari, Arasu, Uma Thanigai, H¨ark¨onen, Kai, Matilainen, Johanna, Nieminen, Petteri, Extracellular vesicles are in- tegral and functional components of the extracellular matrix, Matrix Biology (2017), doi:10.1016/j.matbio.2017.10.003
This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
Extracellular vesicles are integral and functional components of the extracellular matrix
Kirsi Rilla, Anne-Mari Mustonen, Uma Thanigai Arasu, Kai Härkönen, Johanna Matilainen, Petteri Nieminen
Faculty of Health Sciences, School of Medicine, Institute of Biomedicine, University of Eastern Finland, P.O. Box 1627, FI-70211, Kuopio, Finland
Corresponding author: email: kirsi.rilla@uef.fi
Running title: Extracellular vesicles as matrix modulators
Abbreviations used
ACV, articular cartilage vesicle; CCN2, connective tissue growth factor; CD44, hyaluronan receptor, cluster of differentiation 44; COPD, chronic obstructive pulmonary disease; ECM, extracellular matrix; EV, extracellular vesicle; GEC, gingival epithelial cell; GFP, green fluorescent protein; GMC, glomerular mesangial cell; HA, hyaluronan; HAS, hyaluronan synthase; HGF, human gingival fibroblast; HSC, hepatic stellate cell; IL, interleukin; ILV, intraluminal vesicle;
MGP, matrix Gla protein; MMP, matrix metalloproteinase; MSC, mesenchymal stem cell; MV, matrix vesicle; MVB, multivesicular body; OA, osteoarthritis; PGE2, prostaglandin E2; Phospho1, phosphoethanolamine/phosphocholine phosphatase; PS, phosphatidylserine; RA, rheumatoid arthritis; S100A9, S100 calcium binding protein A9; TGF-β, transforming growth factor β; TIMP, tissue inhibitor of metalloproteinases; TNAP, tissue non-specific alkaline phosphatase; vSMC, vascular smooth muscle cell
ACCEPTED MANUSCRIPT
Abstract
Extracellular vesicles (EV) are small plasma membrane-derived particles released into the extracellular space by virtually all cell types. Recently, EV have received increased interest because of their capability to carry nucleic acids, proteins, lipids and signaling molecules and to transfer their cargo into the target cells. Less attention has been paid to their role in modifying the composition of the extracellular matrix (ECM), either directly or indirectly via regulating the ability of target cells to synthesize or degrade matrix molecules. Based on recent results, EV can be considered one of the structural and functional components of the ECM that participate in matrix organization, regulation of cells within it, and in determining the physical properties of soft connective tissues, bone, cartilage and dentin. This review addresses the relevance of EV as specific modulators of the ECM, such as during the assembly and disassembly of the molecular network, signaling through the ECM and formation of niches suitable for tissue regeneration, inflammation and tumor progression. Finally, we assess the potential of these aspects of EV biology to translational medicine.
Keywords
Cancer; connective tissue; exosome; extracellular vesicle; inflammation; microvesicle
Introduction to extracellular vesicles (EV)
EV as messengers between cells and their environment
EV biology is one of the most rapidly growing areas in biomedical research. The journey into the mysteries of these tiny phospholipid-bilayer-covered particles dates back to 1946 when Chargaff and West noticed that the coagulation time of blood plasma increased after centrifugation at 31,000 g [1]. For the first time, EV were documented in 1967 by Peter Wolf with transmission electron
ACCEPTED MANUSCRIPT
microscopy as small particles originating from platelets, which he called ‘platelet dust’ [2]. Later, Polasek described bulbous multivesicular structures arising from the plasma membrane extensions of collagen-activated platelets [3].
Since then, we have learned that this ‘dust’—first considered to be a mechanism of cell waste management—is not solely a key player in the plethora of physiological events of the healthy human body, but also associated with several pathological conditions, such as cancer [4-6]. EV have been detected in virtually all body fluids, such as bile [7, 8], breast milk [9], saliva [10], urine [11, 12], semen [13, 14], blood [15], ascites [16], cerebrospinal fluid [17] and synovial fluid [18].
EV can act as envelopes that protect their cargo (such as DNA, RNA, proteins) from degradation in the extracellular environment [19-21]. They interact with recipient cells via surface receptors, directly fuse with the plasma membrane or become ingested through endocytosis, resulting in the transfer of their contents to target cells [22].
Classification and biogenesis of different types of EV
EV nomenclature has been inconsistent and variable over the years. In many cases, vesicle terminology, such as argosome [23], cardiosome [24], deteriosome [25], ectosome [26], oncosome [4] and sebosome [27], has been based on their origin or biological function. At present, EV are generally classified into three main categories based on their mechanism of biogenesis: exosomes (30–250 nm in diameter), microvesicles (or microparticles, 100–1000 nm) and apoptotic bodies (1–
5 µm). However, these diameters overlap, which makes the simple separation of different subgroups challenging. Therefore, in this review article, EV is the general term used to cover all the categories mentioned above. A typical preparation isolated from conditioned medium of cultured epithelial cells shows the variable size of EV (Figure 1B).
A specific subgroup of EV, matrix vesicles (MV), were first detected in cartilage and bone by ultrastructural analysis [28-30] and isolated from epiphyseal cartilage [31]. The formation
ACCEPTED MANUSCRIPT
mechanism of MV was first expressed as ‘Verdämmerung der Zellen’ [32], which describes their budding from the plasma membranes of mineral-forming cells, such as chondrocytes, osteoblasts and odontoblasts [33]. Despite their crucial role in the initiation of bone and cartilage calcification, MV are often ignored in reports and reviews related to EV biology.
The biogenesis of exosomes occurs via the endocytic pathway. When an early endosome matures into a late endosome, intraluminal vesicles (ILV) bud inside the endosome forming a structure called multivesicular body (MVB) [34]. MVB can either be degraded in lysosomes or fuse with the plasma membrane to release their ILV to the extracellular space. After the release, ILV are called exosomes [35-37].
Other EV types, such as microvesicles, apoptotic bodies and oncosomes, are assumed to form via direct budding from the cell surface or its protrusions. The mechanism is related to the active modification of the phospholipid composition of the inner and outer leaflets of the plasma membrane by flippases and floppases resulting in phosphatidylserine (PS) being transferred to the outer leaflet [38]. Actin–myosin-based contraction of the cytoskeleton is crucial for the completion of the budding [39].
EV as integral components of the extracellular environment
Interactions of EV with the extracellular milieu are important factors in determining their biological effects. The role of MV in the mineralization of developing hard tissues, such as calcified cartilage, bone and dentin, has been recognized for 50 years [30], demonstrating the role of EV in matrix remodeling. Nevertheless, most in vitro experiments are performed in monolayer cell cultures without the presence of the natural 3D extracellular network of proteins, proteoglycans and glycosaminoglycans that absorb water and organize the ECM. However, in vivo, the presence of the ECM adds further complexity to the functions and activity of EV, regulating their biogenesis, penetration and destination. A recent finding by Huleihel et al. showed that biologic scaffold
ACCEPTED MANUSCRIPT
materials contain active EV bound by matrix components [40], which strengthens the general role of EV as structural and functional components of the ECM. Additionally, it must be noted that all extracellular fluids that are rich in EV can actually be considered components of the ECM or filtrates originating from plasma or interstitial fluid. This encourages us to speculate that all EV could actually be categorized as MV.
ECM molecules are carried by and interact with EV
Typically EV display a similar set of surface molecules as the original plasma membrane [41]. All EV contain a variety of surface molecules that allow them to be targeted to the recipient cells and to interact with the ECM [42]. The proteins, lipids and nucleic acids carried by EV have been widely studied, but carbohydrates on the surface of EV have received less attention. The polysaccharides and oligosaccharides on the EV surface assemble a glycocalyx [43], which acts as an important regulator of EV biogenesis and could be utilized in diagnostics and therapeutics. In addition, the glycocalyx may regulate the interactions of EV with the target cells and the ECM through plasma membrane receptors, adhesion proteins and other molecules.
Polysaccharides are important constituents of the interstitial fluid, constructing the ground substance of the ECM. The role of hyaluronan (HA) as a backbone for MV-associated proteoglycans was shown by Wu et al. [44], and recent reports demonstrate that EV carry both non- sulfated and sulfated glycosaminoglycans, such as HA [45-47], chondroitin sulfate [47] and heparan sulfate proteoglycans [48, 49]. Some studies also suggest that polysaccharides have an impact on EV functions. Heparan sulfate proteoglycans on the surface of target cells act as receptors for EV uptake [48] and heparanase, an enzyme that cleaves heparan sulfate, enhances secretion of EV that promote tumor progression [50]. Glycosaminoglycans, such as chondroitin sulfate and HA, are involved in the regulation of MV activity, their interactions with the ECM and bone mineralization [47]. Because high ECM HA content is the key determinant factor of drug barrier in many tumors,
ACCEPTED MANUSCRIPT
oligo-HA-loaded nanoparticles have been used to disrupt the HA-coat and to promote drug delivery into cancer cells [51].
The association of cartilage MV with ECM proteoglycans was indicated as early as in the 1980’s by electron microscopy [52, 53] and, later, the proteoglycan attachment to the MV membrane was demonstrated to be mediated by HA, stabilized by the link protein and HA-binding region in the growth plate cartilage [44]. One of the recently detected cell surface proteoglycans that is carried by EV is glypican-1, which is enriched on cancer cell-derived EV and is one of the most promising EV markers in the early diagnostics of pancreatic cancer [54]. An ECM glycoprotein tenascin C carried by MV regulates bone mineralization activity [55]. Enzymatically active lysyl oxidase-like 2, which is carried on the surface of EV derived from hypoxic endothelial cells, facilitates collagen crosslinking in the ECM [56]. Furthermore, it was reported that EV are closely associated with the collagen network in biological scaffolds [40]. These findings suggest that EV carry and interact with ECM molecules, but the significance of these interactions for the ECM structure and regulation will be unraveled in the future.
Receptor–ligand interactions between EV and ECM adhesion molecules
In concert with their multiple surface adhesion molecules and ligands, EV interact with several ECM components, such as laminin [57] or fibronectin [58] via integrins, and with HA via cluster of differentiation 44 (CD44) [57]. The presence of integrins on EV has been recognized for some time [59], and the latest studies suggest that their expression patterns could be utilized to predict the organotrophic metastasis of cancer cells to lungs, liver or brain [60]. Integrins dictate EV adhesion to specific cell types and ECM molecules, such as laminin and fibronectin, in those organs.
Furthermore, fibronectin carried by EV originating from the plasma samples of breast cancer patients is one of the potential biomarkers that could be detected from liquid biopsies [61], and it was shown to mediate interactions of myeloma cell-derived EV with their targets [62]. EV are able
ACCEPTED MANUSCRIPT
to exert autocrine control of directional cell migration in vivo by promoting cell polarization and the assembly of adhesions with the ECM components, such as collagen fibers via integrin–fibronectin interactions [58]. These findings suggest potential structural, functional and diagnostic roles for the adhesion molecules carried by EV that interact with ECM molecules.
The adhesion receptor CD44 on the membranes of EV (Figure 1G, H) has an important role in EV–ECM interactions via HA, which is its most important ligand [63]. EV that are positive for CD44 and its splice variants have been detected in samples of different origins, including pancreatic adenocarcinoma cells [57], ovarian cancer cells [64] and equine synovial fluid [65]. Interestingly, primary mesothelial cells secrete CD44-positive EV upon epithelial-to-mesenchymal transition induced by wounding or epidermal growth factor [66]. Because of its connection to stemness [72]
and malignant properties of cells, CD44 carried by EV is a potential biomarker and pivotal regulator of EV interactions with target cells and ECM molecules.
Penetration of EV in the extracellular network
Puzzling questions remain as follows: how efficiently and how far are EV able to penetrate inside the fibrous ECM, and what is the proportion of EV that are released into the circulation where EV are transported passively due to fluid flow? EV of up to several µm in diameter are not likely to be able to diffuse out of the dense network of the ECM, unless they are located next to lymphatics or capillaries or originate from the apical faces of epithelial cells lining the body cavities or vessels.
EV contain proteinases, such as matrix metalloproteinases (MMP), which degrade the basement membranes and other ECM components, assisting cell invasion during angiogenesis and cancer progression [67]. Another example is aggrecanase, which may contribute to the degradation and remodeling of the aggrecan-rich ECM, which are typical of brain and cartilage and may be related to diseases, such as brain tumors, neurodegenerative diseases and various types of arthritis [68].
ACCEPTED MANUSCRIPT
Interestingly, 3D cultures demonstrated the shedding of EV from both the apical and basal faces of epithelial cells (Figure 1I–K) into the surrounding collagen-rich matrix [46, 69].
Furthermore, a recent study showed that large EV can contain an active cytoskeleton and are independently capable of dynamic shape-changes and even motility [70]. However, the mechanisms by which the fibrous and dense network of the ECM regulates and restricts the EV penetration, are still waiting for detailed explanations.
Cellular protrusions invading the ECM enhance both shedding and uptake of EV
Long HA-coated protrusions and shedding of HA-coated EV are common features of cells that synthesize high levels of ECM components, such as HA [45, 46]. These cells include epithelial cells overexpressing HA synthase (Figure 1C–K), tumor fibroblasts [71], mesothelial cells [66], human mesenchymal stem cells [72] and synovial fibroblasts [18]. Interestingly, MV originate from cytoplasmic extensions of chondrocytes [73] and mineralizing osteoblasts [74], which suggests that a significant proportion of EV actually derive from various cellular extensions. These extensions and HA-coated EV originating from them may have a general role as matrix depositors (Figure 1A). In addition to their role as specific sites of EV biogenesis, filopodia act as putative organelles for EV uptake in the target cells [75]. By increasing the surface area of the plasma membrane and expanding the cellular dimensions in the extracellular environment, different cellular extensions potentially enhance both the shedding and uptake of EV.
Role of EV in the modeling of connective tissue ECM
Cartilage, bone and dental matrix
MV with a diameter of 20–200 nm are a specialized type of EV with critical roles in the mineralization processes of cartilage, bone and dentin [33]. In contrast to other EV, MV release
ACCEPTED MANUSCRIPT
their bioactive cargo to the ECM changing its properties, rather than to other cells [76]. Their function is not limited to mineralization, as MV are also present in the growth plate resting zone as well as in non-calcifying tissues, such as articular cartilage [77, 78]. MV interact with ECM macromolecules, such as cartilage-specific collagens and proteoglycan link protein [79], and carry proteins involved in cell–matrix interactions and in matrix degradation and remodeling [33, 80].
These include a collagenase and gelatinases (MMP-13, MMP-2 and MMP-9) that can degrade collagen to facilitate mineralization. MMP have been proposed to play roles in the release and activation of growth factors, in growth plate maturation and development and in osteoblastic differentiation. Furthermore, MV transport intraluminal miRNA and they can function in intercellular communication [81, 82]. In this way, MV could influence the proliferation and differentiation of recipient cells through modification of their gene expression.
In cartilage, MV derive from the plasma membrane of chondrocytes [33] during endochondral calcification [83]. They play a pivotal role in the initiation of the mineralization process in the growth plate, although the exact mechanisms and the participation of different enzymes in this process remain unclear [33, 84, 85]. During phase I of mineralization, Ca2+ and PO43–
are transported from the ECM into the interior of MV, where they interact with phospholipids to form a nucleational core complex. This results in the formation of the first hydroxyapatite crystals within the vesicle lumen. Once the accumulation of the intraluminal crystals becomes sufficient, they penetrate through the MV membrane to the extracellular fluid (phase II). These nanocrystals are postulated to seed additional crystals and to mineralize collagen fibrils in the ECM. MV are enriched with various proteins and lipids involved in the mineralization process, including tissue non-specific alkaline phosphatase (TNAP), annexins, nucleotide pyrophosphatase phosphodiesterase, phosphoethanolamine/phosphocholine phosphatase (Phospho1) and PS.
In bone, the plasma membrane of mineralizing osteoblasts releases MV with a diameter of 30–300 nm that are involved in matrix mineralization via the deposition of hydroxyapatite [33, 83].
ACCEPTED MANUSCRIPT
MV not only function in the initial calcification that occurs during embryonic ossification but also contribute to bone formation in adults [86]. EV also represent a newly found mechanism of communication between osteoblasts and osteoclasts, as osteoblast-derived EV were documented to facilitate osteoclast formation [87]. On the other hand, osteoclasts were observed to produce EV that inhibit osteoclastogenesis [88]. Osteocytes are also known to secrete EV that have still speculative roles in the reduction of osteocyte cytoplasmic volume, in the regulation of mineral deposition via communication with osteoblasts and osteoclasts and in the promotion of osteoclast formation [83, 89].
EV with a diameter of 50–250 nm, called articular cartilage vesicles (ACV), are present in normal and diseased articular cartilage [81, 90-92]. They are generated by budding from chondrocytes, and are firmly rooted in the ECM. ACV transport proteins and RNA and concentrate substrates, ions and enzymes required for mineral formation. At the moment, only little is known about the functions of ACV in healthy joints, but they may serve as communication shuttles between distant cells within the articular cartilage. In addition to matrix mineralization, they have been hypothesized to play roles in collagen fibril formation, matrix repair and in the neutralization of potentially toxic substances.
A special case of mineralization is tooth formation and MV have emerged as participants of also this process. At least 3 different types of mineralization can be identified during dentinogenesis, and MV-derived mineralization principally occurs in the mantle dentin [93, 94]. In this case, the plasma membrane of odontoblasts secretes MV [33] that are enriched with high levels of TNAP, Phospho1 and annexin V and accumulate Ca2+ from the extravesicular environment to produce hydroxyapatite within the lumen of MV [76]. Regarding the periodontal apparatus, Phospho1 plays a role in the mineralization of alveolar bone and cellular cementum, while the formation of acellular cementum does not seem to rely on the MV-mediated initiation of mineralization [95]. In addition to the primary mineralization, MV can be involved in the
ACCEPTED MANUSCRIPT
subsequent appositional mineralization in alveolar bone, mantle dentin and cellular cementum, but not in circumpulpal dentin [96]. MV do not participate in the mineralization of enamel in contrast to collagen-based mineralized tissues [84, 94]. It can be concluded that an increasing amount of evidence supports the role of EV in mineralization—both physiological and ectopic (as discussed below)—but future research is still required to unravel their functions especially in healthy synovial joints.
Skin and soft connective tissue
There is an increasing amount of evidence for the role of EV in tissue repair and wound healing.
Myofibroblasts isolated from skin granulation tissue produce high numbers of EV [97], and scratch wounding induces EV shedding in rat primary mesothelial cell cultures (Figure 2), especially in cells close to the wound edge [66]. EV derived from mesenchymal stem cells (MSC) and keratinocytes are able to promote the migration of fibroblasts as well as angiogenesis [98, 99].
Fibroblasts migrating into collagen matrix were documented to release EV that carry MMP-9 to act upon the collagen fibrils, which leads to a gradual transformation of the collagen matrix from a laminar to a fibrillar type of architecture [100]. Other effects of MSC-derived EV include the stimulated growth and proliferation of fibroblasts and the increased synthesis and secretion of collagen and elastin [99]. This eventually leads to re-epithelialization and the enhancement of cutaneous wound healing. EV from keratinocytes stimulate fibroblasts to produce, for example, MMP-1, MMP-3 and interleukins IL-6 and IL-8, which are also related to wound repair [98].
EV can also have a protective role against excessive fibrosis, possibly via specific miRNA [101]. This has been documented, for instance, regarding hepatic stellate cells (HSC), the principal fibrogenic cell type in liver [102]. A recent study using in vivo rat skin model demonstrated that EV originating from human amniotic epithelial cells accelerate scarless wound healing and enhance the reorganization of collagen fibers [103]. Moreover, fibroblasts were observed to transport proteins
ACCEPTED MANUSCRIPT
and lipids to their neighboring cells via ectosomes [104]. The conversion of normal fibroblasts to cancer-associated fibroblasts increased the protein transfer flux in ectosomes toward tumor cells to sustain their high proliferation rate. The pivotal role of EV in tumor progression is discussed in detail below. To conclude: while the exact mechanisms by which EV affect fibroblast migration and proliferation, fibrosis and angiogenesis are to be elucidated in the future, it seems plausible that EV play complex roles in tissue repair and wound healing and, as stated below, offer significant applications to enhance tissue healing.
Synovial fluid
Synovial fluid can be considered a specialized form of ECM in joints. EV are abundant in human synovial fluid, and their main populations have been reported to be ≤300 nm (with variability up to 2000 nm) [18] and between 80–400 nm in diameter [105]. One source of synovial fluid EV is presumably the long HA synthase-positive protrusions of synovial fibroblasts [18]. In addition to synoviocytes, other possible cells of origin include leukocytes and chondrocytes, but EV could also derive indirectly from blood plasma, of which synovial fluid is an ultrafiltrate [18, 81, 106].
EV have been proposed to function in joint development, regulation of joint homeostasis and articular tissue regeneration [106]. Their levels in synovial fluid have been noted to reflect particular joint diseases [105], and platelet- and leukocyte-derived EV are known to stimulate the production of inflammatory IL-6, IL-8 and prostaglandin E2 (PGE2) by synovial fibroblasts [107, 108]. The significance of EV in rheumatoid arthritis (RA) and osteoarthritis (OA) and their potential as biomarkers of cartilage pathology are discussed more thoroughly below.
Knowledge about the functions of EV in healthy human joints remains rudimentary. EV of variable size and shape have been isolated from healthy equine joints, and pretreatment with hyaluronidase results in an enhanced recovery of CD44-positive EV from the HA-rich synovial fluid [65]. As EV in human synovial fluid can be surrounded by a HA coat, they have been
ACCEPTED MANUSCRIPT
proposed to function as transport vehicles for HA [18]. These HA-coated EV were documented to originate from the tips of microvilli or to directly bleb from the plasma membrane. The presence of HA on the surface of EV has been speculated to mediate the binding and uptake of EV by lymphatic vessels via HA receptors, and to regulate the drainage of synovial fluid. Future studies with synoviocyte cultures are needed to assess the diverse factors that regulate the HA–EV synthesis and to design suitable applications for translational medicine.
Role of EV–ECM interactions in tumor progression and angiogenesis
Cancer-derived EV were first detected in the spleen nodules and lymph nodes of a patient with Hodgkin’s lymphoma in the late 1970’s [109]. These EV have the capacity to carry tumor- promoting factors and to initiate premetastatic niche formation [42]. The ECM can be considered pivotal for tumor cells survival, as it provides a niche for them in the tissue/organ of origin as well as in distant organs during metastasis [110, 111]. Metastatic cells incite a series of events to favor their metabolism and growth that interfere with the organization and composition of the ECM [112]. In this way, the tumor cells produce a favorable habitat for themselves by releasing various factors and EV that induce changes in their vicinity.
The acidic pH of the tumor microenvironment activates EV-associated proteinases, such as cathepsin B, which degrades several ECM proteins, including fibronectin, laminin, tenascin C and type IV collagen [113]. This could be responsible for the proinvasive capacity of tumor cells and the proangiogenic activities of endothelial cells but could also enhance the interaction of EV with the matrix. EV derived from ovarian cancer ascites contain high levels of MMP-2, MMP-9 and urokinase-type plasminogen activator proteinases that could degrade the ECM [114]. The ECM is also remodeled by EV released from human fibrosarcoma and melanoma cells that contain enzymatically active MMP-14 [115]. This enables the primary tumor cells to recondition the niche
ACCEPTED MANUSCRIPT
after their arrival. In addition, tumor-derived EV carry extracellular matrix metalloproteinase inducer and kallikreins (hK5, hK6, hK10 and hK14) that also contribute to matrix degradation [42, 114]. Proteinases carried by EV may promote invasion-stimulating properties and focal proteolysis at distant sites creating a path of least resistance for the metastatic cells. Loss of the tissue inhibitors of metalloproteinases (TIMP) gene family causes the release of EV rich in proteins, such as a disintegrin and metalloproteinase containing domain 10, from the fibroblasts that promote tumor by altering the microenvironment [116].
EV participate in preparing secondary sites of the body to metastasis [117]. EV also have specificity towards the organ they are targeting, supporting Stephen Paget’s ‘seed and soil’
hypothesis on organotropism [60, 118]. For instance, macrophage migration inhibitor factor, carried by EV from pancreatic ductal adenocarcinoma, induced the transforming growth factor β (TGF-β) production in Kupffer cells. This stimulated HSC to produce more fibronectin that creates a premetastatic niche to attract the metastasizing tumor cells [119].
In addition to the preparation of the metastatic site for invasion, the proliferation of tumor cells also requires a supply of oxygen and nutrients, which is obtained by establishing a vascular network. This process facilitates the metastatic niche formation, and recent findings show that EV are also involved [120, 121]. Umezu et al. have shown that exosomes derived from leukemia cells carry miRNA-92a, which inhibits the mRNA encoding integrin α5 in endothelial cells leading to an increase in angiogenesis [122]. Breast cancer cell-derived EV promote the differentiation and angiogenic behavior of adipose-derived stem cells via increased vascular endothelial growth factor and fibronectin levels [123]. Furthermore, MMP-8 and MMP-9, which are strong inducers of angiogenesis, are expressed in EV collected from patients with glioblastoma multiforme [124].
It is clear that cancer cell-derived EV are actively involved in creating a supportive microenvironment for tumor progression via their indirect or direct regulation of angiogenesis, invasion, metastasis and degradation of the ECM. However, tumor microenvironment may, in turn,
ACCEPTED MANUSCRIPT
facilitate the release and the activity of tumor cell-derived EV. This interplay is also emerging as an attractive target for novel treatment options.
Modulation of the ECM by EV during inflammation
By contributing to the mineralization and degradation of the ECM, EV have been shown to play a role in pathogenic, inflammatory events related to vascular calcification, rheumatic diseases, periodontal disease and respiratory diseases. Furthermore, EV participate in the initiation of hepatic and renal fibrosis, which are inflammatory responses characterized by the accumulation of ECM proteins.
Vascular calcification
Vascular calcification is a significant medical issue in several chronic inflammatory disorders, including chronic kidney disease, diabetes mellitus and atherosclerosis [125]. In addition to chronic inflammation and several other features, it is characterized by the mineralization of the ECM of blood vessels or heart valves. Vascular calcification is increasingly regarded to be an active process that includes several mechanisms [126]. Together with high extracellular levels of phosphate and calcium, EV derived from vascular smooth muscle cells (vSMC) and/or macrophages are considered major factors leading to pathological calcification [127]. These vesicles possess calcifying abilities and, thus, contribute to the ectopic mineralization by accumulating to the ECM in blood vessels [127]. By interacting with collagen matrix, EV act as a foci for the formation of small hydroxyapatite crystals, microcalcifications, which further culminates in extended calcification of the ECM in vasculature. The events occurring in vascular calcification highly resemble the ones involved in normal skeletal formation [128].
ACCEPTED MANUSCRIPT
Annexins are calcium-binding proteins that form ion channels and, thus, mediate the calcium influx into chondrocyte-derived MV during mineralization [129, 130]. EV derived from vSMC contain annexins II, V and VI, of which annexin V is also found in vesicles secreted by calcium- and phosphate-activated macrophages [131, 132]. Furthermore, a recent study demonstrated enrichment of annexin I–VII and XI in EV derived from valve interstitial cells [133] that are responsible for the matrix calcification of the aortic valve [134]. Similarly, annexins contribute to the formation of a nucleation complex on the EV membrane, initiating the mineralization of the ECM in blood vessels [135]. A study of vSMC-derived EV has demonstrated that the silencing of annexin VI reduces the calcification, which supports the importance of annexins in EV-mediated calcification [131]. The findings of Kapustin and colleagues have also unveiled the increased levels of PS, a phospholipid that binds annexin, on EV-membranes in calcifying conditions, which leads to the notion that the annexin–PS complex on the outer membrane of EV may act as a nucleation site for mineralization. In turn, the upregulated expression of S100 calcium-binding protein A9 (S100A9) and annexin V in macrophage-derived EV, and the co-localization of these components in calcified plaques suggest that PS, annexin V and S100A9 may also form a complex that promotes nucleation [132].
Several other proteins enriched in calcifying EV act at the vesicle–ECM interface and regulate the accumulation and mineralization competence of EV in the ECM [136]. Under normal physiological conditions, EV derived from vSMC are loaded with matrix Gla protein (MGP) and fetuin-A that prevent mineral precipitation inside vesicles [136]. In calcifying conditions, the activity of mineralization inhibitors is suppressed and the calcifying competence of vesicles enabled. Apart from MGP, Gla-rich protein has also been suggested to inhibit mineralization [137].
Furthermore, EV derived from human coronary arterial vSMC contain sortilin, which promotes the mineralization potential of these EV by increasing the activity of TNAP [138, 139]. The higher TNAP activity has also been connected to the enhanced ECM calcifying ability of EV [136, 140]
ACCEPTED MANUSCRIPT
and TNAP is a key regulator of physiological mineralization in MV [141]. However, the absence of TNAP activity in vSMC-derived EV in the presence of high calcium [131] has prompted the hypothesis that TNAP may act in vesicles mediating calcification in chronic kidney disease, in which high phosphate levels play a role, whereas vesicles without TNAP activity may be involved in atherosclerosis [131, 138].
Generally, it seems that the EV-mediated mechanisms of vascular calcification are multifactorial and mainly ascribed to EV-membrane proteins and other molecular mediators that enable the high mineralization competence of EV. Revealing the expression profiles of these novel proteins in EV derived from different cellular sources in varying calcifying conditions could clarify the precise mineralization mechanisms in various clinical conditions, which are characterized by vascular calcification. In particular, a thorough understanding of these processes can yield information about new potential targets to treat atherosclerosis, for which the existing medical therapies are mostly useful for prevention but not for the treatment of the actual calcification.
Rheumatoid arthritis and osteoarthritis
There is a growing evidence for the crucial role of EV in the progression of rheumatic diseases, RA and OA—inflammatory conditions characterized by the degradation of articular cartilage. In addition to the participation in pathogenesis, EV may also serve as potential biomarkers for rheumatic diseases. The interest in studying EV as biomarkers has especially emerged in the context of RA, as EV counts are notably increased in plasma and urine of patients with RA and the levels also correlate with disease activity [142, 143]. In addition, distinct changes in the levels of EV from synovial fluid have been observed in RA, OA and juvenile idiopathic arthritis, indicating that EV profiles could have potential as signatures of joint diseases [105].
In addition to antigen presentation, cell-to-cell communication and several other mechanisms, EV can contribute to RA pathogenesis via the degradation of the ECM [144]. EV-mediated
ACCEPTED MANUSCRIPT
mechanisms of ECM destruction rely on the initiation of MMP production or direct degradation enabled by the enrichment of proteolytic enzymes in EV. In RA, synovial fluid contains high levels of EV that originate from several sources and contribute to disease pathogenesis by various mechanisms. T cell- and monocyte-derived EV induce the production of several MMP, including MMP-1, MMP-3, MMP-9 and MMP-13 by synovial fibroblasts [145], indicating the participation of EV in the MMP-mediated ECM degradation. EV derived from RA synoviocytes and/or synovial fluid also contain aggrecanase, hexosaminidase D and β-glucuronidase, degradative enzymes that further augment the catabolic activity of EV [68, 146, 147]. EV originating from leukocytes and platelets in synovial fluid of RA patients have been shown to increase the production of IL-6, IL-8 and PGE2 by synovial fibroblasts [107, 108], supporting the contribution of EV to inflammation in RA. In addition to IL-6 and IL-8, monocyte- and granulocyte-derived EV induce the increased release of chemoattractant protein-1, intercellular adhesion molecule-1, vascular endothelial growth factor and chemokine C-C motif ligand 5 by synovial fibroblasts [148]. Furthermore, being a coagulant, leukocyte-derived EV in inflammatory synovial fluid may promote the hypercoagulation in joints in RA [142]. These data suggest that EV participate in the catabolic and inflammatory processes of RA via resident matrix-degrading enzymes and by inducing the production of several mediators related to inflammation and tissue injury by synovial fibroblasts. The complexity of the EV-mediated mechanisms is evident, as EV causing the destructive and inflammatory phenotype of synovial fibroblasts can originate from various cell types. Increasing our knowledge of these mechanisms and intricate communication pathways that they comprise would be a great step toward a more comprehensive understanding of the complex pathogenesis of RA.
In OA, MMP-13 has been indicated to be the main contributor in ECM degradation and the resulting disturbed articular homeostasis [149], although increased activities of other MMP have also been associated to OA [150]. Apart from inducing ECM degradation, EV play a role in OA pathogenesis by inducing the pathologic mineralization of articular cartilage [91]. As is the case of
ACCEPTED MANUSCRIPT
calcifying EV derived from vSMC, EV shed from OA chondrocytes are also enriched with proteins, including TNAP, annexin II and V, whose function at the level of EV–ECM interaction promotes the vesicle-mediated calcifying activity [151]. Furthermore, the fetuin-A–MGP complex is also absent in vesicles shed from OA cartilage, supporting its crucial function of inhibiting mineralization also in joints [150]. However, the EV levels in synovial fluid of OA patients are not significantly different from those of healthy individuals [144]. These vesicles were reported to possess increased expression of miR-200c and to suppress the inflammation after endocytosis by chondrocytes, suggesting that the transfer of miR-200c-containing EV acts in maintaining joint homeostasis [144]. It is intriguing that although EV derived from platelets are abundant in the synovial fluid of patients with RA, they have not been observed in the synovial fluid of OA patients [108]. The precise roles of EV in the phenomena of OA remain to be fully elucidated.
Other inflammatory pathologies with ECM degradation
Several types of EV related to different inflammatory pathologies contain MMP, supporting the direct ECM degradation capability of these vesicles. For instance, gingival epithelial cell (GEC)- derived EV affect the expression of several genes in human gingival fibroblasts (HGF) and have a crucial role in periodontal disease [152]. Bacterial biofilm extract induced the secretion of EV from GEC and these EV significantly upregulated MMP-1, MMP-3, IL-6 and IL-8 levels in HGF, while the expression levels of a MMP inhibitor, TIMP-4, was downregulated. Thus, these findings indicate that GEC-EV promote a HGF phenotype that can participate in the pathogenic events observed in periodontal disease, including increased matrix degradation and promoted inflammation [153, 154].
Chronic obstructive pulmonary disease (COPD) is characterized by the chronic inflammation of peripheral airways and lung parenchyma, with the participation of several mediators, e.g.,
ACCEPTED MANUSCRIPT
cytokines, chemokines and proteinases [155]. Proteinase activity, which comprises elastase, MMP- 1, MMP-2, MMP-7 and MMP-12, occurs in emphysema, resulting in the release of fragments of ECM components [156]. A recent study suggests that the cigarette smoking-induced ECM damage would be mediated by macrophage-derived EV that possess gelatinolytic and collagenolytic potential due to the increased expression of MMP-14 [157]. Furthermore, lung epithelial cells are also a source of EV in COPD and these EV are enriched with cysteine-rich protein 61, which can trigger MMP-1 and IL-8 production in the lung [158]. These data suggest a novel mechanism of endothelial cell- and macrophage-derived EV in the regulation of inflammation and tissue injury in COPD pathogenesis. An additional molecule that links EV to inflammation is prolyl endopeptidase, an extracellular proteinase associated with cystic fibrosis that is released within exosomes from lung endothelial cells upon toll-like receptor 4 activation [159]. Similar to rheumatic diseases, EV contribute to periodontal disease and COPD via the induction of inflammation and tissue injury and by triggering the production of MMP and inflammatory cytokines. While the exact roles of EV in the pathogenesis of, for instance, COPD are only beginning to be understood, previous findings suggest that the ECM injury and inflammation can be promoted by EV derived from various cell types, including lung epithelial cells and macrophages. Future studies are necessary to further elucidate our knowledge about these EV-mediated events that are likely to comprise complex pathological mechanisms.
Fibrosis
Liver fibrogenesis is a repair process defined by the pathological accumulation of ECM proteins in response to several chronic liver diseases, including alcoholic and non-alcoholic fatty liver disease and viral hepatitis [160]. HSC are mainly responsible for the excessive production of ECM proteins during fibrosis [161]. The activation of quiescent HSC to proliferative myofibroblast-like cells involves the up-regulation of several genes, e.g., TGF-β and MMP inhibitors TIMP-1–2, which leads to HSC migration at the sites of tissue repair and the increased production of ECM proteins
ACCEPTED MANUSCRIPT
[161, 162]. Studies by Povero et al. have expanded current paradigms of HSC activation mediated by EV [163]. The cellular uptake of EV was dependent on vanin-1 located on the EV-membrane, and they contained miR-128-3p that was shown to induce the activation of HSC. Sphingosine kinase 1-containing exosomes released by endothelial cells have also been reported to activate HSC by inducing AKT signaling and migration [164]. In addition to the activation of HSC, EV contribute to interactions between HSC, by getting transferred to neighboring quiescent or activated HSC and inducing events that can promote fibrosis or fibrolytic activities [165]. EV secreted by activated HSC have been observed to contain connective tissue growth factor (CCN2), a component that plays a role in, e.g., ECM production and has been shown to be up-regulated in liver fibrosis [165, 166]. On the other hand, quiescent HSC have been shown to release EV that reduce CCN2 production in adjacent HSC, resulting in activated miR-214 and suppressed fibrogenic signals [167]. Furthermore, Kornek et al. demonstrated that activated and apoptotic T cells in active hepatitis C release EV and, in HSC, these particles activate the upregulation of several MMP as well as diminish the fibrogenic responses triggered by TGF-β1 [168]. These fibrolytic changes could enable tumor invasion and metastasis in hepatocellular carcinoma [169], similar to the phenomena described above for other tumors.
The regulation of the ECM by EV has also been implicated in renal fibrosis, a feature of diabetic nephropathy [170]. The accumulation of ECM proteins observed in diabetic kidney diseases was suggested to be triggered by the activation and subsequent phenotypic change of glomerular mesangial cells (GMC) [171]. Exosomes derived from high glucose-treated glomerular endothelial cells induce the activation of GMC and the subsequent upregulation of type IV collagen and fibronectin expression in vitro [172]. The regulation of the ECM by glomerular endothelial cell- derived exosomes was further supported by in vivo experiments, where these exosomes induced significant mesangial expansion in mice.
ACCEPTED MANUSCRIPT
These findings have revealed that EV are responsible for the activation of the cells, which cause the pathological accumulation of ECM proteins both in liver and renal fibrosis. Furthermore, studies on liver fibrosis suggest that EV also play a crucial role in cell-to-cell communication, as they are competent to induce either fibrogenic or fibrolytic events, depending on the type and activity of cells, from which the EV are secreted. The understanding of this kind of sophisticated crosstalk and regulation between cells could reveal potential targets for limiting pathological fibrosis and diagnostic use of EV as early-stage disease markers.
Extracellular vesicles and matrix biology: Therapeutic applications
While the use of EV as routine therapy is still in its very early stages, there are several promising applications being developed. Generally, the use of EV-based therapies can be classified under three main categories. First, the EV particle could be used to carry various medical agents to the target tissues (drug delivery), for instance, in cancer therapy [173]. Second, EV could replace cellular therapies (whole cells) as carriers of signaling molecules, such as mRNA. In this case, EV could be either artificially produced or derived from whole organisms, tissues or cell cultures.
Finally, EV could offer novel tools for diagnostics and biomonitoring for diverse medical conditions. In this chapter, we especially concentrate on the potential use of EV in diseases related to matrix biology: those of connective tissue, inflammation and degenerative diseases.
As stated above, the current use of EV in therapies is quite limited [174]. Some therapeutic applications are in phase I–II clinical trials with quite a small number of participants (N ≤ 40), mostly involving cancers. Various miRNA carried by EV are promising tools to be used as biomarkers for cancers [175], EV could work as carriers for anticancer drugs [173], induced elimination of EV could prevent metastatic growth [176] and EV could be used as anticancer therapies per se [177]. There are also other diagnoses where it would be useful to control or
ACCEPTED MANUSCRIPT
eliminate the release of particular types of EV that may be responsible for the initiation or progression of tissue pathology, as is the case in some pulmonary diseases discussed below.
An application with high potential is a new EV-based approach to wound healing, especially regarding biological therapies to chronic wounds [99]. EV from human induced pluripotent stem cell-derived MSC have been examined in fibroblast cultures, where they stimulate proliferation and migration as well as the secretion of collagens and elastin. The effects are not restricted to fibroblasts, and experiments on human umbilical vein endothelial cells show increased tube formation, indicating angiogenesis, another useful phenomenon in wound healing. In vivo in rats, treatment with these EV promotes wound healing when measured with scar width and collagen deposition. It seems that these effects are mediated through Wnt family member 4 signaling, inducing β-catenin activation, as well as AKT signaling [178].
Other instances of tissue repair and regeneration could also benefit from EV therapies [101].
The pivotal step of tissue repair is the resolution of inflammation that can be enhanced by specific EV [179]. The origin of the EV is of importance as, depending on the source, EV can have either pro- or anti-inflammatory properties [101]. The beneficial immunosuppressive potential can be exemplified by EV from MSC cultures that can increase the levels of anti-inflammatory IL-10 while decreasing those of pro-inflammatory cytokines IL-1β, IL-6 and IL-12p40 [180]. Another potential application in the field of combatting inflammation could be to prevent acute graft-versus-host disease after stem cell transplantation, which has already proven to be efficient in a mouse model [181].
Regarding other practical applications in tissue healing, MSC-derived EV show great promise in osteochondral regeneration [182]. In rats with experimental osteochondral defects on distal femurs, intra-articular exosome treatment for 12 weeks not only induced the complete restoration of damaged cartilage with a regular surface, but also a return of subchondral bone to a histologically healthy pattern. This was in stark contrast to controls that exhibited only fibrous tissue repair. The
ACCEPTED MANUSCRIPT
authors did not specify the EV cargo responsible for the effect but hypothesized that the whole array of diverse components could be necessary for effective healing. Similarly in bone repair, MSC- derived EV could be useful to accelerate the healing of fractures. In a mouse model with reduced production of exosomes (CD9–/–), femoral fractures healed significantly faster after local injections of human MSC-derived exosomes [183]. Again, it is yet impossible to tell which EV components were responsible for the beneficial effects. Both the cartilage and bone repair potentials are very promising, but the regeneration of cartilage tissue merits even more attention, as the results of Zhang et al. indicated complete histological restoration of healthy cartilage [182] and the possible applications of this are enormous due to the osteoarthritic diseases of the aging population. To supplement the actual therapies, EV also have potential as biomarkers in joint diseases. Malda et al.
suggested that EV from immune cells could be used as signs of early-onset joint disease and the EV in synovial fluid could be applied to the monitoring of disease state and progression [106]. One potential cargo of significance in synovial fluid is the HA-coat of EV [18] that could become useful in both diagnostic and therapeutic applications.
A combination of mechanisms in tissue repair—not only targeted at matrix biology—comes into play when considering EV as therapeutic tools for diseases of the internal organs. Just to give a few examples, EV miRNA have efficacy in protecting the kidney from ischemia through endothelial and tubular epithelial effects [184]. The repair of hypoxic injury to kidney tubular cells seems to be mediated by TGF-β1 mRNA-containing exosomes through the activation of fibroblasts [185], offering potential targets for EV therapy intervention including both initiation of repair and inhibition of excessive fibrous tissue formation. COPD provides an opposite example of utilizing the suppression of EV production for treatment. Here the secretion of specific miRNA is dysregulated due to cigarette smoke, which can ultimately induce oncogene expression and myofibroblast differentiation, causing disease initiation and/or progression [186]. Thus, elimination of the EV carrying the miRNA responsible for disease pathogenesis could be useful in both
ACCEPTED MANUSCRIPT
prevention and therapy [155]. In the same manner, EV can enhance the formation of amyloid fibrils in neurodegenerative diseases, indicating potential use in biomonitoring or prevention [187].
Finally, in cardiovascular disease, engineered EV with sonic hedgehog cargo could enhance angiogenesis and preserve cardiac function during myocardial infarction as well as facilitate brain remodeling during or after stroke [188].
Future perspectives
Based on the discussion above, it is obvious that the use of EV in treatment or biomonitoring is still in its early stages. Lener et al. have assessed the challenges that have to be overcome to make the various EV applications eventually a part of evidence-based medicine [189]. These aspects are important when the safety, efficacy and cost-effectiveness of EV-based products are considered before any wider clinical use. The authors classify EV-based therapeutics as belonging to biological medicine—derived from humans or animals as opposed to products of chemical and herbal origin—
that is regulated as a class but without particular recommendations in the case of EV. The exact mode of action of EV-based therapies needs to be provided for each product to attain adequate safety and quality standards.
Even before phase I clinical trials, EV-based products could be characterized for safety based on the standards for tissues and cells that are used therapeutically [189]. As EV derive from a donor (e.g., cell culture, tissue or biological fluid), there is the potential risk for high variability of products, which should be taken into consideration. When production per se is considered, the manufacturing needs to have proper infrastructure and technology, and appropriate preclinical models for testing (in vivo animal models, etc.) should be in place. At present, there are no standards for the isolation or storage of EV, which would be pivotal when releasing these for established clinical use. New protocols may be necessary. Nor are there yet universally accepted criteria for quality control. Finally, the delivery route of EV should be taken into careful
ACCEPTED MANUSCRIPT
consideration as, for instance regarding osteoarthritic joint disease, both intravenous and intra- articular products have been considered [106].
In addition to these technical aspects of EV-based therapeutics, we also suggest the evaluation of existing treatments or diagnostic tools before examining the potential use of EV in medicine.
Based on our current review, the field is extremely wide and EV are being suggested as potentially useful tools in a diverse array of disorders. In these cases, it would be practical to assess if the EV- based solution is more useful than the existing ones, and if the EV-based approach has any potential of producing better outcomes than the established methods. For instance, alcoholic hepatitis increases EV numbers and the levels of particular miRNA in EV [190]. Considering the diagnosis of alcoholic hepatitis, existing standard laboratory tests when combined with patient history, can still be highly accurate and predictive [191]. In cases where available acute diagnostics are adequate, we propose that the potential focus of EV-based monitoring would be in refining the prognosis and pinpointing the time points when to intervene with various treatments, preferably as preventive measures before overt symptoms, as discussed above for COPD.
A good example of EV diagnostics supplementing conventional methods is their potential use in cardiovascular diseases. While the acute manifestations of myocardial infarction and stroke can be readily diagnosed with existing methods [192, 193], there are also marginal cases, where it is hard to distinguish critical or intervention-requiring conditions from stable patients [194]. For instance, EV may offer a tool to assess the risk of a carotid plaque being detached (causing cessation of blood flow distally)—i.e., discriminating high-risk patients for ischemic myocardial injury or stroke requiring rapid intervention. In addition, EV can help to evaluate the long-term prognosis of patients developing coronary artery disease, again, enabling the implementation of preventive measures before overt symptoms would develop. In contrast to acute emergencies, where the analytical tools of EV measurement and characterization are at the moment too complex and
ACCEPTED MANUSCRIPT
unstandardized for bed-side diagnostics, these prognostic tools could be used at leisure in an outpatient clinic.
Final conclusions
As discussed above, there is an active communication between EV and the ECM by several mechanisms (summarized in Figure 3): i) EV act as structural and functional key components of the ECM and with their surface adhesion proteins, ii) they are able to bind matrix molecules and iii) to carry and distribute the components of the ECM both in original and target tissue modulation and reorganization. Furthermore, iv) proteinases and signaling molecules carried by EV modulate the matrix composition. We are expecting to obtain more detailed data on the mechanisms of EV interplay with ECM molecules and how EV participate in the ECM architecture and remodeling.
Studies on EV in the context of the extracellular niche will be crucial to understanding the biological effects of EV and their potential in numerous diagnostic and therapeutic areas.
Figure legends
Figure 1. Cells secrete EV into the ECM. A human chondrosarcoma cell expressing green fluorescent protein (GFP)–hyaluronan (HA) synthase 3 (HAS3)-induced filopodia and EV (green, arrows) (A). The large green particles are fixed erythrocytes indicating the exclusion area of the HA coat and the red color indicates HA stained with the fluorescent HA-binding complex (HABC).
Transmission electron microscopy of EV preparation collected from the culture media of GFP–
HAS3 expressing MDCK cells (B). Live MDCK cells expressing GFP (C, D) and GFP-tagged HAS3 (E, F) were grown in 3D conditions in basement membrane extract gel and imaged with differential interference contrast (DIC) and confocal microscopy. The 3D cultures of GFP (G) and GFP–HAS3 expressing cells (H) were sectioned and stained for HA receptor CD44 (red) and HA
ACCEPTED MANUSCRIPT
with HABC (green) (F, G) showing CD44- and HA-positive particles embedded in the matrix (white arrows). Transmission electron microscopy of GFP–HAS3 expressing cultures (I, J, K) shows EV shedding on both apical and basal faces of the epithelium (black arrows). Figure represents a summary of original results published in [46, 69].
Figure 2. Budding of EV from apical surface of a mesothelial cell. A rat primary mesothelial cell (isolated from abdominal peritoneum) near the wound edge of scratch-wounded cell culture imaged by scanning electron microscopy. A higher magnification image from the area indicated by the black box in (A) is presented in (B). The arrows indicate large and the arrowheads smaller vesicles.
Original results were published in [66].
Figure 3. EV as functional components and modulators of the ECM. A schematic drawing on the relationship between EV and the ECM and the impact of EV on various cellular and tissue processes in health and disease. EV form by direct budding from the plasma membrane (1) or its protrusions (2), or by the fusion of multivesicular bodies with the plasma membrane (3). They interact with the target cells by direct fusion (4), receptor–ligand interactions (5) or endocytosis (6).
Direct and indirect communication between EV and ECM (black arrows) have multiple beneficial and harmful effects on target cells and tissues.
Acknowledgments
Financial support for the preparation of this review was provided by the Academy of Finland (grants #276426 & #284520), Jane and Aatos Erkko Foundation, Otto A. Malm Foundation, K.
Albin Johanssons Stiftelse Foundation, Kuopio University Foundation and Northern Savo Cancer Foundation.
ACCEPTED MANUSCRIPT
References
[1] E. Chargaff, R. West, The biological significance of the thromboplastic protein of blood, J. Biol.
Chem. 166 (1946) 189–197.
[2] P. Wolf, The nature and significance of platelet products in human plasma, Br. J. Haematol. 13 (1967) 269–288.
[3] J. Polasek, The appearance of multivesicular structures during platelet activation as observed by scanning electron microscopy, Thromb. Res. 28 (1982) 433–437.
[4] M. Morello, V.R. Minciacchi, P. de Candia, J. Yang, E. Posadas, H. Kim, D. Griffiths, N.
Bhowmick, L.W. Chung, P. Gandellini, M.R. Freeman, F. Demichelis, D. Di Vizio, Large oncosomes mediate intercellular transfer of functional microRNA, Cell. Cycle. 12 (2013) 3526–
3536.
[5] J. Skog, T. Wurdinger, S. van Rijn, D.H. Meijer, L. Gainche, M. Sena-Esteves, W.T. Curry Jr, B.S. Carter, A.M. Krichevsky, X.O. Breakefield, Glioblastoma microvesicles transport RNA and proteins that promote tumour growth and provide diagnostic biomarkers, Nat. Cell Biol. 10 (2008) 1470–1476.
[6] S. Bhatnagar, K. Shinagawa, F.J. Castellino, J.S. Schorey, Exosomes released from
macrophages infected with intracellular pathogens stimulate a proinflammatory response in vitro and in vivo, Blood. 110 (2007) 3234–3244.
[7] R.P. Witek, L. Yang, R. Liu, Y. Jung, A. Omenetti, W.K. Syn, S.S. Choi, Y. Cheong, C.M.
Fearing, K.M. Agboola, W. Chen, A.M. Diehl, Liver cell-derived microparticles activate hedgehog signaling and alter gene expression in hepatic endothelial cells, Gastroenterology. 136 (2009) 320–
330.
ACCEPTED MANUSCRIPT
[8] A.I. Masyuk, B.Q. Huang, C.J. Ward, S.A. Gradilone, J.M. Banales, T.V. Masyuk, B. Radtke, P.L. Splinter, N.F. LaRusso, Biliary exosomes influence cholangiocyte regulatory mechanisms and proliferation through interaction with primary cilia, Am. J. Physiol. Gastrointest. Liver Physiol. 299 (2010) G990–G999.
[9] C. Admyre, S.M. Johansson, K.R. Qazi, J.-J. Filén, R. Lahesmaa, M. Norman, E.P. Neve, A.
Scheynius, S. Gabrielsson, Exosomes with immune modulatory features are present in human breast milk, J. Immunol. 179 (2007) 1969–1978.
[10] Y. Ogawa, M. Kanai-Azuma, Y. Akimoto, H. Kawakami, R. Yanoshita, Exosome-like vesicles with dipeptidyl peptidase IV in human saliva, Biol. Pharm. Bull. 31 (2008) 1059–1062.
[11] T. Pisitkun, R.F. Shen, M.A. Knepper, Identification and proteomic profiling of exosomes in human urine, Proc. Natl. Acad. Sci. U. S. A. 101 (2004) 13368–13373.
[12] H. Zhou, P.S. Yuen, T. Pisitkun, P.A. Gonzales, H. Yasuda, J.W. Dear, P. Gross, M.A.
Knepper, R.A. Star, Collection, storage, preservation, and normalization of human urinary exosomes for biomarker discovery, Kidney Int. 69 (2006) 1471–1476.
[13] I. Brody, G. Ronquist, A. Gottfries, Ultrastructural localization of the prostasome - an organelle in human seminal plasma, Ups. J. Med. Sci. 88 (1983) 63–80.
[14] A. Poliakov, M. Spilman, T. Dokland, C.L. Amling, J.A. Mobley, Structural heterogeneity and protein composition of exosome-like vesicles (prostasomes) in human semen, Prostate. 69 (2009) 159–167.
[15] R.M. Johnstone, M. Adam, J.R. Hammond, L. Orr, C. Turbide, Vesicle formation during reticulocyte maturation. Association of plasma membrane activities with released vesicles (exosomes), J. Biol. Chem. 262 (1987) 9412–9420.
ACCEPTED MANUSCRIPT
[16] F. Andre, N.E. Schartz, M. Movassagh, C. Flament, P. Pautier, P. Morice, C. Pomel, C.
Lhomme, B. Escudier, T. Le Chevalier, T. Tursz, S. Amigorena, G. Raposo, E. Angevin, L.
Zitvogel, Malignant effusions and immunogenic tumour-derived exosomes, Lancet. 360 (2002) 295–305.
[17] L.J. Vella, D.L. Greenwood, R. Cappai, J.P. Scheerlinck, A.F. Hill, Enrichment of prion protein in exosomes derived from ovine cerebral spinal fluid, Vet. Immunol. Immunopathol. 124 (2008) 385–393.
[18] A.-M. Mustonen, P. Nieminen, A. Joukainen, A. Jaroma, T. Kääriäinen, H. Kröger, E. Lázaro- Ibáñez, P.R. Siljander, V. Kärjä, K. Härkönen, A. Koistinen, K. Rilla, First in vivo detection and characterization of hyaluronan-coated extracellular vesicles in human synovial fluid, J. Orthop. Res.
34 (2016) 1960–1968.
[19] C. Lässer, V.S. Alikhani, K. Ekström, M. Eldh, P.T. Paredes, A. Bossios, M. Sjöstrand, S.
Gabrielsson, J. Lötvall, H. Valadi, Human saliva, plasma and breast milk exosomes contain RNA:
Uptake by macrophages, J. Transl. Med. 9 (2011) 9.
[20] K.G. Ronquist, G. Ronquist, L. Carlsson, A. Larsson, Human prostasomes contain chromosomal DNA, Prostate. 69 (2009) 737–743.
[21] P.A. Gonzales, T. Pisitkun, J.D. Hoffert, D. Tchapyjnikov, R.A. Star, R. Kleta, N.S. Wang, M.A. Knepper, Large-scale proteomics and phosphoproteomics of urinary exosomes, J. Am. Soc.
Nephrol. 20 (2009) 363–379.
[22] L.A. Mulcahy, R.C. Pink, D.R. Carter, Routes and mechanisms of extracellular vesicle uptake, J. Extracell. Vesicles. 3 (2014) 24641.
ACCEPTED MANUSCRIPT
[23] V. Greco, M. Hannus, S. Eaton, Argosomes: A potential vehicle for the spread of morphogens through epithelia, Cell. 106 (2001) 633–645.
[24] A. Waldenström, N. Gennebäck, U. Hellman, G. Ronquist, Cardiomyocyte microvesicles contain DNA/RNA and convey biological messages to target cells, PLoS One. 7 (2012) e34653.
[25] K. Yao, X. Wu, J.E. Thompson, J.C. Carlson, Isolation and characterization of deteriosomes from rat liver, J. Cell. Biochem. 51 (1993) 488–494.
[26] J.M. Stein, J.P. Luzio, Ectocytosis caused by sublytic autologous complement attack on human neutrophils. The sorting of endogenous plasma-membrane proteins and lipids into shed vesicles, Biochem. J. 274 (1991) 381–386.
[27] A. Nagai, T. Sato, N. Akimoto, A. Ito, M. Sumida, Isolation and identification of histone H3 protein enriched in microvesicles secreted from cultured sebocytes, Endocrinology. 146 (2005) 2593–2601.
[28] E. Bonucci, Fine structure and histochemistry of "calcifying globules" in epiphyseal cartilage, Z. Zellforsch. Mikrosk. Anat. 103 (1970) 192–217.
[29] G.W. Bernard, D.C. Pease, An electron microscopic study of initial intramembranous osteogenesis, Am. J. Anat. 125 (1969) 271–290.
[30] H.C. Anderson, Electron microscopic studies of induced cartilage development and calcification, J. Cell Biol. 35 (1967) 81–101.
[31] S.Y. Ali, S.W. Sajdera, H.C. Anderson, Isolation and characterization of calcifying matrix vesicles from epiphyseal cartilage, Proc. Natl. Acad. Sci. U. S. A. 67 (1970) 1513–1520.
[32] E. Bonucci, L.C. Dearden, Matrix vesicles in aging cartilage, Fed. Proc. 35 (1976) 163–168.
