The Role of Extracellular Vesicles in De Novo Mineralization: An Additional Novel Mechanism of Cardiovascular Calcification

Abstract

Extracellular vesicles are membrane micro/nanovesicles secreted by many cell types into the circulation and the extracellular milieu in physiological and pathological conditions. Evidence suggests that extracellular vesicles, known as matrix vesicles, play a role in the mineralization of skeletal tissue, but emerging ultrastructural and in vitro studies have demonstrated their contribution to cardiovascular calcification as well. Cells involved in the progression of cardiovascular calcification release active vesicles capable of nucleating hydroxyapatite on their membranes. This review discusses the role of extracellular vesicles in cardiovascular calcification, and elaborates on this additional mechanism of calcification as an alternative pathway to the currently accepted mechanism of biomineralization via osteogenic differentiation.

Introduction

Cardiovascular calcification, a growing burden in Westernized countries, is not only a risk factor for cardiovascular events, but may itself contribute to cardiovascular risk (1). A growing number of studies have demonstrated that microcalcification in vulnerable plaques contribute to plaque destabilization and fatal plaque rupture (2–5). Calcification of the cardiovascular system — including the coronary arteries and heart valves — follows an active process in which smooth-muscle cells (SMCs) or valve interstitial cells undergo osteogenic transformation (6–8). In addition, it is now evident that calcification progresses through, and indeed may be initiated by, the release of calcifying extracellular vesicles by cells residing in the calcification niche (2, 9).

Extracellular vesicles possess a metabolically active outer membrane that protects the internal cargo — consisting of proteins, miRNA, and other components from the parental cell. They can be found throughout the body in various tissues and fluids, and they participate in both physiological and pathological processes. Their involvement in a broad range of pathological pathways has made them attractive diagnostic biomarkers (10), while their therapeutic use is an emerging field (11, 12). Extracellular vesicles appear to have advantages over existing drug delivery systems due to their size, lack of toxicity, and target specificity. A growing number of studies have contributed to the concept that cells implicated in the progression of cardiovascular calcification release active extracellular vesicles capable of nucleating hydroxyapatite (13–15). This emerging subset of the field provides additional mechanisms by which to therapeutically target cardiovascular calcification.

Discovery of Calcifying Extracellular Vesicles

The discovery of cell-derived extracellular vesicles followed the introduction of transmission electron microscopy in the mid-20th century. The groups of Anderson and Bonucci discovered that extracellular vesicles associate with the earliest sites of mineral formation in bone and cartilage mineralization (16, 17). These extracellular membrane-bound structures were later termed matrix vesicles (MVs). Physiological mineralization is now widely believed to be initiated in bone, dentin, and cartilage by vesicles released from specific regions of the outer membranes of osteoblasts, odontoblasts, and osteoblasts (18). Using ultrastructural, histological, and cytochemical techniques, Anderson and colleagues showed the presence of matrix vesicle-like structures that were believed to originate from smooth muscle cells (9, 19).

Classification Controversy

Much controversy exists in this field regarding the classification and nomenclature used for extracellular vesicles. Depending on size and type, extracellular vesicles are broadly classified as ectosomes (or shedding microvesicles), exosomes, and apoptotic bodies (20). Ectosomes, also known as microparticles, are large extracellular vesicles ranging from 50–1000 nm in diameter; exosomes are small membranous vesicles of endocytic origin ranging from 40–100 nm in diameter; and apoptotic bodies are released from fragmented apoptotic cells and are 50–5000 nm in diameter. MVs (the main focus of this review) are another category that should be added to this classification. MVs are small membranous structures (30–300 nm in diameter) surrounded by a lipid biolayer, are produced by blebbing of plasma membrane, and can calcify. The current criteria for the classification of extracellular vesicles includes size, density, morphology, lipid/protein composition, and subcellular origin (Table) (21). Several limitations exist in our current understanding of the field. The recently formed International Society of Extracellular Vesicles hopes to overcome these issues by producing guidelines to standardize the field (http://www.isevmeeting.org).

Table.

Classification of extracellular vesicles.

Extracellular vesicle	Size (nm)	Density in sucrose (g/ml)	Morphology	Lipid composition	Protein markers	Subcellular location
Exosome	40–100	1.10–1.21	Cup shape; multivesicular bodies	Enriched in cholesterol, sphingomyelin, ceramide; lipid rafts; exposed phosphatidylserine	Tetraspanins (CD63, CD9), Alix, TSG101	Endosomes (internal compartments)
Ectosome (microparticles)	50–1000		Bilamelar round and tubular structures	Enriched in cholesterol and diacylglycerol; exposed phosphatidyserine	Proteolytic enzymes; CR1	Plasma membrane
Apoptotic body	50–5000	1.16–1.28	Heterogeneous	Exposed phosphatidylserine	Histones
Matrix vesicle	30–300		Double membrane vesicles	Exposed phosphatidylserine	Enriched in annexins	Plasma membrane

Preparation of Extracellular Vesicles

The isolation of these entities is a major issue of dispute. Different groups use different protocols, which leads to differences between study results. The few groups that study the role of MVs in cardiovascular calcification follow a similar protocol, allowing for more direct comparison of results between groups. The main method for isolating calcifying extracellular vesicles from cell supernatant is differential centrifugation — short time periods of low centrifugal forces to remove cellular fragments (~800–1000 g) and apoptotic bodies (16,500 g), followed by a much higher centrifugal force (~100,000 g) for a longer duration (13, 14). The correctness of this method is debatable; some scientists in the realm of exosomes, for example, prefer to perform a sucrose gradient in which vesicles are separated by density. Future analyses and growth of the field should clarify whether the current protocol of MV isolation needs alteration.

Role of Extracellular Vesicles in Pathogenesis

Circulating microparticles contain large numbers of extracellular vesicles released from platelets, erythrocytes, leukocytes, and endothelial cells — first described as “cell dust”— and are key in the haemostatic response (22, 23). They participate in the pathogenesis of thromboembolic diseases, wherein the number of circulating procoagulant-microparticles increases greatly (23). In the field of cancer research, exosomes — extracellular counterparts of endosomes — may enable invasive growth of tumor cells (24). Circulating extracellular vesicles represent excellent biomarkers for determining the severity of various diseases. Early evidence suggests that increased numbers of endothelial extracellular vesicles in the pulmonary circulation demonstrate advancement of pulmonary arterial hypertension (25). In addition, extracellular vesicles contain varying cargoes of lipids, proteins, and miRNA, which have sparked interest in their potential as diagnostic markers of disease (26).

Pathological Calcification

Cardiovascular calcification is now recognized as regulated biomineralization that follows similar pathways to that of bone development. Nuclear magnetic resonance and X-ray diffraction techniques have shown that the matrix-mineral atomic interface in calcified plaque is similar to that in bone (27). Several studies have indicated that, in addition to bone-like mechanisms, other mechanisms - including cell death - also contribute to cardiovascular calcification (28, 29, 30). A combination of in vitro studies (31, 32) and molecular imaging have formed the concept of an inflammation-dependent calcification paradigm (30, 33), suggesting that macrophage infiltration and inflammation precede calcification, and activated pro-inflammatory pathways induce osteogenic transformation of SMCs or release of MVs.

Calcifying extracellular vesicles also have been identified in calcifying aortic valves (34), atherosclerotic intimal lesions (15), and medial arterial calcification (2, 34, 35) — similar to the MVs involved in physiological bone mineralization (18). Literature is lacking regarding the role of MVs in calcifying aortic valves, which could differ from the role of MVs in arterial calcification — although this concept requires further investigation. MVs may initiate the mineralization processes akin to those in bone, but also may lead to the rupture of vulnerable plaques (2). During the early stages of calcification, MVs and apoptotic bodies released from macrophages and SMCs may contribute to the calcification process (14, 15). Regions of “spotty” calcifications in atherosclerotic lesions that contain calcified vesicles have been predicted, via finite element simulations, to have increased stress levels and to be prone to rupture (5), which can lead to acute thrombosis and fatal myocardial infarction (36). Patients with chronic renal disease (CRD) have advanced atherosclerosis and a high cardiovascular mortality rate (37). The severity and complexity of CRD causes a perfect storm of metabolic dysfunction, leading to accelerated intimal and extensive medial calcification (37, 38). Concentrations of extracellular calcium and phosphate, similar to those found in serum of CRD patients on dialysis, induce the release of calcifying MVs from cells involved in cardiovascular calcification, such as vascular SMCs and macrophages (14, 15). The plaques of non-CRD patients also may contain MVs (39), suggesting that this phenomenon also occurs in an atherosclerotic or diabetic milieu.

Structure

Ultrastructural studies of components of the cardiovascular system have identified highly heterogeneous mineral-associated vesicles (MVs) ranging from 30–300 nm in diameter and displaying different structural appearances (2, 34, 35). They are typically double-membrane- bound bodies that are round or ovoid in shape (17), often associated with extracellular matrix components, particularly collagen, and displaying evidence of hydroxyapatite crystals on the inner membrane within the lumen, and/or on the outer membrane of the vesicle (Figure 1).

Figure 1.

Mechanism of pathological matrix vesicle mineralization: A mineral imbalance leads to an influx of phosphate and calcium into the vesicle via appropriate channels – ALP generates inorganic phosphate in the extravesicular space. Nucleation of hydroxyapatite (HA) is facilitated by an annexin (Anx)-PS complex, enabling generation of microcalcification/calcification.

Biogenesis

MVs arise from cells during physiological mineralization through a “budding” process (40), originating from specialized regions of the cell plasma membrane and released in the same orientation as that of the membrane (41). When the vesicle buds from the cell, it takes a subset of the cell’s “cargo” with it, safely packaged within a region of the membrane. Evidence suggests that the plasma membrane of the parenting cell contains components similar to those of the released vesicle, but in different quantities (42). MVs are rich in annexins and acidic phospholipids, such as phosphatidylserine (PS), compared to the corresponding plasma membranes, which enable them to calcify (14, 15, 42).

Matrix Vesicle Cargo

Extracellular vesicles mediate active communication between cells. They participate in the exchange of functional and genetic information, and in the mediation of adaptive immune responses. The cargo carried by extracellular vesicles dictates their function. Proteomics studies have provided evidence of the cargo contained within osteoblast-derived and chondrocyte-derived MVs (14, 43, 44). Numerous non-calcifying extracellular vesicles also appear within calcifying plaques. Studies by Shananan and colleagues (14, 29) have provided evidence suggesting that non-calcifying vesicles contain inhibitors of calcification such as fetuin- A and matrix Gla protein, which decrease in MVs released by cells in a CRD milieu, enabling them to calcify. These MVs may possess a physiological function under non-pathological conditions. In one study, proteomics analysis of vascular SMC-MVs detected signaling molecules (14). Calcifying vesicles therefore may be merely dysfunctional vesicles, released to act as intercellular communicators. On the other hand, cells may release these particular vesicles as a reaction to stress, as a means of releasing unwanted “cargo” — such as excess calcium.

Nucleation Core

PS forms a complex with calcium at sites of early mineralization in skeletal tissue (45), and inorganic phosphate is required to form the nucleation core (46). We now believe that in both physiological and pathological MVs, annexins also may contribute to the formation of nucleation complex (14, 15). Annexins — specifically, annexin-2, -5, and -6 — are major proteins within MVs (15, 45). They facilitate calcium influx and mineralization by binding to PS and forming ion channels in the MV membrane (45, 47). The exact importance of annexins in MV mineralization, however, is still under investigation.

Annexin-5 was previously shown to facilitate nucleation and growth of mineral (48). Despite this finding, a recent in vivo study produced evidence that a lack of annexin-5 and/or annexin-6 functionality does not affect skeletal development in mice (49). These results suggest that a compensatory mechanism is at play — another member of the annexin family may compensate for the loss of annexin-5 and/or annexin-6. Annexin-5 is seemingly, if controversially, important in the mineralization potential of macrophage-derived MVs involved in biomineralization (15). In contrast, annexin-6 seems to play a main role in SMC-MV mineralization (14). The nucleation core is required for nucleation — a physiochemical process through which ions accumulate with the correct orientation to mimic a crystal surface — to ensue. We recently demonstrated that pro-inflammatory and pro-thrombotic S100A9 might be involved in nucleation of MVs via formation of PS–annexin5–S100A9 complex (15).

Nucleation, Crystallization, and Microcalcification Formation

Whether crystal formation or transdifferentiation of vascular SMCs occurs first is debatable (50). Whether osteogenic transformation of SMCs is required for MV release and calcification is also unknown. But we do know that in skeletal tissue, the first crystals of hydroxyapatite are formed within extracellular vesicles. This phenomenon can be split into two phases: In conditions of mineral imbalance, the influx of calcium and phosphate into the extracellular vesicles, via their appropriate channels, leads to initial mineral accumulation in the form of amorphous calcium phosphate (51). For mineralization to proceed, the levels of phosphate and pyrophosphate must be in balance (52), achieved via alkaline phosphatase (ALP) within the membranes of calcifying extracellular vesicles in bone (53). Whether MVs involved in cardiovascular calcification contain ALP, however, is controversial — with conflicting reports on ALP within vascular SMC-derived MVs (13, 14). The second phase, mineral propagation, seems to ensue via the release of crystal through the MV membrane, exposing preformed hydroxyapatite nanocrystals to the extracellular fluid. Once exposed to this fluid, the nanocrystals can act as loci or templates for the formation of new crystals via homologous nucleation (9), and perhaps mineralize extracellular matrix (ECM) components. We presume that MVs associated with pathological ectopic calcification generate crystals in a similar way.

MVs bind to ECM components, such as collagens and glycosaminoglycans (GAGs). Vesicles appear to be the initial site of nucleation, and intravesicular mineral crystals seed these ECM components. Vesicle–GAG interaction is enhanced in mineralization-competent/calcifying chondrocyte-derived MVs (54). Atherosclerotic plaque contains a predominance of GAGs associated with mineralization similar to that of bone, further suggesting that GAGs play a direct role in initiating pathological and physiological mineralization (27). Crystalline extracellular vesicles and calcified collagen fibers that abound in Randall’s plaques (found in renal medulla) seem to follow this alternative mechanism of MV mineralization, akin to vascular calcification (55).

Molecular imaging has demonstrated that inflammation and microcalcification evolve within close proximity in intimal arterial calcification, and overlap at border regions that are prone to rupture (33). A positive feedback loop of calcification and inflammation drives disease progression (56). Calcified extracellular vesicles/microcalcification in vulnerable macrophage- rich plaques may contribute to plaque destabilization and rupture, as predicted by Vengrenyuk and colleagues (2–4). Components of calcifying extracellular vesicles, including pro-thrombotic tissue factor and S100A9 may further lead to acute cardiovascular events. A similar inflammation-driven mechanism appears to ensue in calcific aortic valve disease (33, 38). While our electron microscopy observations show that MVs closely associate with microcalcification, suggesting a role of calcifying MVs in the generation of microcalcification (Figure 2), the exact mechanism of pathological microcalcification formation by MVs remains to be elucidated.

Figure 2.

Ultrastructural images of a calcified human tissue: A, defined border of microcalcification; B, higher magnification shows extracellular vesicles (red arrowhead) within regions of microcalcification and nucleation of hydroxyapatite on the outer membrane of MV (red arrow). Inset image shows the nucleation of MV on the inner membrane.

Conclusion

Much of our knowledge of the role of extracellular vesicles in cardiovascular calcification relies heavily on previously established evidence of MVs involved in physiological bone mineralization. Although research in this field dates back to the 1970s, the role of extracellular vesicles in pathological calcification is still largely unknown. Progression of this field has been hindered by the lack of sophisticated modalities to visualize extracellular vesicles in vitro and in vivo. Novel modalities that can detect the size and numbers of extracellular vesicles and visualize them in vitro in real time — such as nanoparticle tracking analysis — are emerging. In addition, molecular imaging has proven useful in visualizing fluorescently probed calcifying vesicular structures in vivo in live animals (30). The development of innovative technologies would further advance this field of research. Further understanding of extracellular vesicle structure and function may pinpoint novel means of treating diseases associated with these entities.