Cardiovascular diseases (CVD) result in more than 19 million deaths annually and coronary heart disease (CHD) accounts for the majority of this death toll. In most cases, atherosclerosis and thrombosis associated with unstable plaques are the major cause for cardiovascular events (CVEs) including acute coronary syndromes (ACS) and stroke (van der Wal and Becker, 1999). Atherosclerosis is a systemic disease that starts early in life, asymptomatically progressing though adulthood, until clinically manifested. Indeed, large numbers of subjects who die suddenly of CHD are apparently healthy and had no previous symptoms (Mujica et al., 2010). Atherothrombosis is regulated by both genetic and environmental factors (e.g., dyslipidemia, hypertension, smoking, diabetes, and obesity; Marenberg et al., 1994; Palomo et al., 2006). The development of atherosclerotic lesions result from a complex interplay between circulating factors and various cell types in the vessel wall, which leads to the accumulation of lipids in the subendothelial space and a complex process of chronic inflammation, mainly characterized by endothelial dysfunction, leukocyte infiltration, and platelet activation (van der Wal et al., 1994; Nishijima et al., 2004). Indeed, platelet activation and subsequent platelet aggregation processes play an essential role in the development of atherosclerosis, possibly through a vast amount of molecules released upon platelet activation (Palomo et al., 2008). Interestingly, increasing evidence support the view that the role of platelets in atherosclerosis and its thrombotic complications may be mediated, in part, by local secretion of molecular effectors embedded or packed into microvesicles from the platelet surface.

Circulating microvesicles (cMVs) may participate in haemostatic and inflammatory responses, neovascularization, cell survival, and apoptosis, processes which are involved in atherothrombosis (Aatonen et al., 2012). Thus, increased levels of cMVs derived from platelets (pMVs), erythrocytes (ErMVs), leukocyte (LMVs), and endothelial cells (eMVs) are associated with individual metabolic abnormalities caused by metabolic syndrome and oxidative stress (Ueba et al., 2008; Helal et al., 2011). ErMVs, LMVs, and eMVs seem to be more abundant in human atherosclerotic plaques than in plasma (Leroyer et al., 2007). High amounts of these plaque MVs result from apoptotic leukocytes within vulnerable plaques. In contrast, MVs of platelet origin are the most abundant in blood (Rank et al., 2010). Circulating MVs are found in the plasma of healthy subjects (Horstman and Ahn, 1999; Berckmans et al., 2001; Caby et al., 2005; Toth et al., 2007; Grant et al., 2011; Herring et al., 2013) although their relative concentrations are determined by the pathophysiological context. Background levels of circulating pMVs in the absence of disease likely originate from aging platelets in the absence of activation (Cauwenberghs et al., 2006). The local release of MVs from platelets at the site of platelet plug formation indicates a possible role of pMVs in the haemostatic response in vivo (Lubsczyk et al., 2010). pMV membrane is a composite of the platelet plasma- and granule membranes (Biro et al., 2005) and have procoagulant properties which lead to thrombin generation. Such procoagulant activity (PCA) relies on the exposure of membrane anionic phospholipids that enable the assembly of coagulation complexes at the MV surface, and on the eventual thrombin formation (Sinauridze et al., 2007). Besides the well-known role of platelet-derived MVs in coagulation, thrombosis, and haemostasis, pMVs have been involved in a variety of processes such as wound healing, inflammation, CVD, diabetes, arthritis, tissue regeneration, and cancer.

This review summarizes and highlights the latest findings of the complex cross-talk between platelet-derived microvesicles, inflammatory cells, and vascular elements, and provides novel insight and understanding in the development of the atherosclerotic and thrombotic disease as well as potential clinical application of pMVs in diagnosis and therapy.

Platelet-derived microvesicles retain certain parental cell characteristics, such as surface receptors, although their content depends on platelet stimuli, giving differences in their phenotype (Sims et al., 1989; Montoro-Garcia et al., 2014). pMV protein content is highly dependent on the type of physiological agonists involved in platelet stimulation (Milioli et al., 2015). MVs were shown to differ in content between human samples and between MV size classes (Dean et al., 2009; Bastos-Amador et al., 2012). Regarding phospholipid MV content, the relative abundance of lyso-phosphatidylcholines among other phospholipid classes (phosphatidylcholines, lyso-phosphatidylethanolamines, PSs, phosphatidylinositols, sphingomyelins, and diacyl- and plasme(a)nyl-phosphatidylethanolamines) was found to be significantly higher in blood MVs than platelets and platelet-derived MVs (Losito et al., 2015). In terms of protein composition, pMVs from activated platelets are highly enriched in α-granule-derived factor Va and Xa (Monkovic and Tracy, 1990) and thus possess procoagulant properties (Connor et al., 2009). Despite their heterogeneity (Perez-Pujol et al., 2007) a high percentage of pMVs express surface activation markers such as glycoprotein (GP)-Ibα (CD42b), P-selectin (CD62P), β₃-integrin (GPIIIa, CD61), α_IIb-integrin (GPIIb, CD41), lysosomal-associated membrane protein-3 (LAMP3, CD63), and thrombospondin-1 (TSP-1). However, healthy individuals may also have CD41⁺/CD61⁺/PS⁺-MVs derived from megakaryocytes instead of activated platelets. Megakaryocyte-derived MVs differ from pMVs in that they do not express granule fusion markers (CD62P^-, LAMP-1^-) and only contain full-length filamin A (Flaumenhaft et al., 2009). pMVs carry as well other platelet proteins such as platelet activating factor (PAF), β-amyloid precursor, anticoagulant protein C/S, complement C56b-9, and the chemokines CXCR4, CXCL4, CXCL7, and CCL5, which may exert potent biological effects in recipient cells of the circulation (Garcia et al., 2005). Indeed, pMVs were able to transfer fully operational surface receptors, such as CSCR4 and CD41, onto the recipient cells (Rozmyslowicz et al., 2003). pMVs thereby might modulate platelet, leukocyte, and vascular endothelial cells (EC) function, by modulating monocyte-EC interactions (Barry et al., 1998) and inducing chemotaxis (Barry et al., 1999). Because pMVs circulate in the blood flow, they could serve as shuttle modules and signaling transducers not only to neighboring cells (local environment) but also to cells at relative distance from their cell or site of origin, triggering cell activation, phenotypic modification, and reprogrammation of cell function, complementing the well-known processes of intercellular communication. How MVs are selectively released and targeted to exert their various pathophysiological functions remains unknown.

Plasma levels of circulating MVs are known to increase with the presence of cardiovascular risk factors [hypercholesterolemia (Suades et al., 2014), obesity (Murakami et al., 2007), hypertension (Preston et al., 2003), diabetes mellitus (DM; Sabatier et al., 2002; Zhang et al., 2013), metabolic syndrome (Helal et al., 2011), sleep apnea (Trzepizur et al., 2014), and progression of atherosclerosis (Nomura et al., 1995; Ueba et al., 2008)]. Within DM, type-1 DM patients had twice tissue factor (TF)-positive pMVs than type-2 DM patients (Chiva-Blanch et al., 2016d). Likewise, elevated pMVs are associated with the 10-year CVD risk score (Ueba et al., 2010) and with age (van der Zee et al., 2006) in healthy men. Indeed, total number of pMVs was significantly greater in women than men in a population of healthy individuals (Gustafson et al., 2015). During past decades, the involvement and importance of the immune system and inflammation in the initiation and progression of atherothrombosis has become clear, also contributing to CVD risk. Raised levels of circulating pMVs have been detected in young women with polycystic ovary syndrome (Willis et al., 2014), in patients with rheumatoid arthritis in which pMVs express CLEC-2 (Gitz et al., 2014) and in patients with antiphospholipid antibodies in which pMVs correlate with anti-β2-GP (Chaturvedi et al., 2015). Additionally, elevated levels of circulating pMVs contribute to the increase PCA in patients with oral cancer (Ren et al., 2015).

High concentrations of circulating pMV (ranging from ∼3,000–11,000/μL) have been reported in patients with CHD including ACS (Michelsen et al., 2008; Bulut et al., 2009; Biasucci et al., 2012; Cui et al., 2013; George et al., 2015). In addition, circulating pMVs have been reported to reflect the size of myocardium at risk in patients with ST-elevation myocardial infarction (Jung et al., 2012) and correlated with the presence of atherothrombotic lesions in carotid plaques (Lukasik et al., 2013), intracranial atherosclerotic lesions (Kuriyama et al., 2010), early stage of coronary artery calcification in menopausal women (Jayachandran et al., 2008), cerebrovascular disease (Chiva-Blanch et al., 2016b) and peripheral arterial disease (van der Zee et al., 2006). Stenting in stable coronary atherosclerotic lesions is associated with a substantial release of pMVs (Horn et al., 2015). Interestingly, the link with increased circulating pMV levels is more consistent with coronary artery disease patients than in those with chronic kidney disease (Chen et al., 2014). Indeed, overproduction of pMVs and platelet activation with suppressed aggregation may be even implicated in the pathogenesis of coagulation abnormalities in children with congenital heart disease (Horigome et al., 2002; Ismail and Youssef, 2012). Recently, pMVs have been point out as biomarker of the vaso-occlusive phenotype-related severity in sickle cell anemia (Nebor et al., 2014). Furthermore, pMVs seem to be involved in the development of disseminated intravascular coagulopathy in critically ill patients but are not related to hospital mortality, while pMV/platelet ratio is independently associated to hospital mortality (Ohuchi et al., 2015). Finally, total pMVs were lower while activated platelet-derived MVs (P-Sel⁺) were higher in preeclampsia than in healthy women, being involved in the hypercoagulable intravascular reaction during pregnancies complicated by preeclampsia (Campello et al., 2015). Therefore, pMV may reflect the severity of the endothelial injury and platelet activation during thrombotic events (Tan et al., 2005; Jung et al., 2012).

Acquiring knowledge about the role of pMVs in CVD might also have implications for treatment. Caution must be taken when considering cMVs as pathological markers, since medical therapy with pharmacological agents can affect MV release during disease progression (Mobarrez et al., 2011; Nomura et al., 2011). Thus, pMVs are susceptible targets for pharmacological modulation and offer new options for therapies specifically focused on lowering MV levels. For instance, anti-platelets drugs such as GPIIb/IIIa inhibitors (Goto et al., 2003; Morel et al., 2004), acetilsalicilic acid (Bulut et al., 2011), and clopidogrel (Judge et al., 2010; Franca et al., 2012), and the anti-diabetic drug ticlopidine (Nomura et al., 2004c) have shown to reduce pMVs. In patients under antithrombotic treatment, pMVs exposing CD62P or CD142 are still elevated 6 months after initiation of the therapy (Skeppholm et al., 2012), possibly due to the fact that low-dose of acetilsalicilic acid might not be strong enough to suppress shedding of pMVs into the microcirculation (Lubsczyk et al., 2010). Our group has reported that aspirin intake in diabetic patients has no effect on pMVs (Chiva-Blanch et al., 2016d). However, statins as the cornerstone drug therapy for lipid-lowering decrease pMVs in patients with different cardiovascular risk factors. Thus, simvastatin and pravastatin decreased pMVs in patients with hypertension (Nomura et al., 2004a) and type-2 diabetes (Nomura et al., 2004b; Sommeijer et al., 2005). Similarly, atorvastatin reduces thrombin generation and expression of TF, GPIIIa, and CD62P on pMVs in patients with peripheral vascular disease (Mobarrez et al., 2011) and with type-1 diabetes and dyslipidemia (Tehrani et al., 2010). In a recent study aimed to evaluate the effects of lipid-lowering treatment on cMV generation in patients in primary prevention of atherosclerosis, we have demonstrated that, in hypercholesterolemic patients, statin treatment reduces not only the number of pMVs but markers of activated platelets, activated inflammatory cells and TF with respect to untreated subjects even when LDL levels were similar (Suades et al., 2013). In agreement, stroke patients with hyperlipidemia presented a significantly lower percentage of pMVs compared to control subjects, not only due to lipid-lowering but also because of a significant role in reduction of platelet activation and reactivity (Pawelczyk et al., 2015). Hence, in view of all these data statins may exert beneficial effects by inhibiting microvesicle generation and the triggering of MV-dependent mechanisms. Other cardioprotective agents that have an impact on pMV levels include calcium channel blockers (Nomura et al., 2005a,b), antioxidants such as vitamin C (Morel et al., 2003), and PPAR-pan agonists like bezafibrate (Kagawa et al., 2001). As several therapeutic drugs seem to influence the levels and composition of pMVs, the lowering of pMV load in the circulation might prove, at least in part, to be a novel therapeutic strategy for treatment. Nevertheless, whether the beneficial effect of a pharmacological approach is associated to pMV reduction and to a clinical improvement needs to be fully demonstrated.

In addition to pharmacological modulation, the therapeutic potential of progenitor cell-derived microvesicles is promising since they are naturally occurring, efficient, therapeutic delivery vehicle that might be used to deliver drugs to specific targets. Furthermore, the therapeutic potential of MVs has also been pointed out by the use of synthetic MVs, mimicking natural ones. MVs could have a broad potential in several conditions from inflammation to MI (Getts et al., 2014). Further characterization of the biological effects of these MVs is warranted.

Platelets are not only key mediators of thrombosis but also of inflammation by directly interacting with cells of the immune system (Fuentes et al., 2013). Reports in the last decade have described the secretion by platelets of proinflammatory molecules that exacerbate the inflammatory response in atherosclerotic lesions, during the initial injury to the endothelium as well as in the later stages when the atherosclerotic plaque is destabilized (von Hundelshausen and Weber, 2007). In addition to their pivotal role in CVD event presentation, platelet-derived MVs participate in inflammatory responses because they carry immune complexes, which are highly proinflammatory (Boilard et al., 2010; Cloutier et al., 2013). High-shear-stress-induced pMVs in a cone-plate viscometer device in vitro enhance expression of inflammatory cytokines either in ECs or in the human monocytic THP1-cell line (Nomura et al., 2001). Of note, pMVs may have a role in initial stages of atherosclerotic process, as they can facilitate cell-to-cell communication and adhesion processes between blood and vessel wall (Mause et al., 2005). pMVs could also play part in atherogenesis by enhancing proliferation of vascular smooth muscle cells (SMCs) as well as EC and chemotaxis and proliferation of hematopoietic cells (Miyazono et al., 1985; Weber et al., 2000; Baj-Krzyworzeka et al., 2002; Pakala, 2004). These results clearly suggest that pMVs contribute to atherosclerosis development and to vascular damage occurring in inflammatory disorders (Nomura et al., 2001; Inoue et al., 2006; Suades et al., 2015b). To this respect, Csongrádi et al. (2011) have described that blood pMV levels positively associate with abnormal carotid IMT and other risk factors in obesity suggesting a critical role of enhanced platelet reactivity in atherosclerotic wall alteration. In addition to pMVs, circulating lymphocyte-derived CD45⁺/CD3⁺-MVs have recently been shown to be biomarkers of asymptomatic subclinical lipid-rich atherosclerotic plaques in patients with familial hypercholesterolemia (Suades et al., 2014), and CD11b⁺-LMV of unstable plaques in asymptomatic patients with high-grade carotid stenosis (Sarlon-Bartoli et al., 2013). Indeed, CD45⁺/CD3⁺-lymphocyte-derived MVs were found increased in individuals at high cardiovascular risk (HCVR) who were about to develop a major CVE (Chiva-Blanch et al., 2016c).

In a similar fashion, levels of pMVs were significantly higher in patients with intermediate coronary lesions compared to subjects with normal coronary arteries (Chou et al., 2014). Progression of early atherosclerotic lesions to advanced plaques and their thrombotic complications are consequence of complex interactions between blood cells and arterial vessel wall components (Badimon and Vilahur, 2014). Erosion, fissure, or rupture of the atherosclerotic plaques are triggering events in ACS, being platelets a key player in these atherothrombotic processes (Fuster et al., 1990). We have found that HCVR patients have significantly high shedding of cMVs carrying epitopes of platelet activation and directly associate to lipid-rich subclinical atherosclerotic burden (Suades et al., 2015a). Besides, these prothrombotic pMVs have demonstrated an incremental prognostic value beyond the classical risk factor model for the prediction of cardiovascular risk, indicating that the state of activation of platelets in the blood-vascular interface may increase the release of pMVs that become markers of the high atherothrombotic risk (Suades et al., 2015a). Blood thrombogenicity can be partially explained by the fact that TF is not only present in the subendothelium, but also in the circulation (Sambola et al., 2003). TF is associated with monocytes, platelets, and even microvesicles, the latter represent an important source of the so-called blood-borne TF. We have also detected higher numbers of TF-positive MVs derived from platelets in HCVR patients; thus, pMV-associated TF might contribute to atherothrombosis. In the same line, our group has reported that clustering the information provided by TF⁺-pMVs, eMVs, and LMVs might predict CVEs in high-risk patients following a Mediterranean diet supplemented with nuts (Chiva-Blanch et al., 2016a).

Besides a potential relevance of MVs and pMVs as markers of subclinical atherosclerosis with a critical importance in reclassification of asymptomatic subjects, the follow-up study performed by Namba et al. (2007) in patients with a first ACS points to the fact that high pMV levels at discharge may also be an independent predictor for secondary thrombotic events and poorer clinical outcomes at 1 year. This study stresses the usefulness of pMV levels to differentiate patients who develop secondary atherothrombotic events from patients who develop a stable phenotype after a first ACS (Namba et al., 2007).

As pre- and analytical methodological procedures (Aatonen et al., 2014; Eckstein et al., 2014; out of the scope of this review) are still under assessment, care should be taken when expanding findings to prospective clinical studies. Prior to large-scaling, it would be desirable a complete international standardization of cMVs analysis.

Further to their potential as biomarkers of cell activation, increasing evidence support the concept that pMVs are causal inducers of atherosclerosis progression and thrombosis. MVs from human atherosclerotic plaques are highly abundant and more thrombogenic than plasma MVs, with differences between LMVs and pMVs (Leroyer et al., 2007). Ramacciotti et al. (2009) using a model of venous thrombosis, provided evidence that thrombus weight correlated negatively with LMVs and positively with pMVs. Besides, levels of pMVs have been found to be higher in culprit coronary arteries than in peripheral arteries of STEMI patients (Suades et al., 2016). In this microenvironment, pMVs released by activated platelets may provide a new prothrombotic interface for fibrin formation between the circulating blood and the growing thrombus (Ando et al., 2002), since they are able to interact with fibrin (Siljander et al., 1996). Thus, it has been proposed that pMVs might play a key role in vivo causing thrombotic events, even without the direct involvement of platelets (Oberle et al., 2007). The response to this hypothesis was provided by demonstrating that pMVs, beyond being biomarkers of cell activation, have functional effects on cardiovascular atherothrombotic disease because they enhance platelet and fibrin deposition on atherosclerotic arterial wall (Suades et al., 2012). This proof of principle study was performed perfusing atherosclerotic vessel wall with blood with/without exogenously added pMVs and showed that high pMVs concentration was able to induce more platelet and fibrin deposition (Figure 1) (Suades et al., 2012; Mause, 2013). We have also found a decrease of pMVs with surface markers of adhesion and activation in the post-thrombus blood after perfusing the exposed thrombogenic surfaces and in STEMI-patients (Suades et al., 2015c). Therefore, our data reinforce pMVs in blood promotes platelet adhesion due to a high tendency to adhere, as previously reported (Forlow et al., 2000) and support their clear implication in the atherothrombotic process. Altogether, this growing body of evidence support the view that pMV dissemination and exposure of their procoagulant membrane to the extracellular matrix (Merten et al., 1999) at the site of endothelial injury or onto the forming fibrin (Siljander et al., 1996) may serve as an adhesion surface on adhered platelets and within the thrombus to enable thrombin generation and to further support recruitment of platelets (Suades et al., 2012) and leukocytes (Mause et al., 2005) stimulating platelet aggregation (Berckmans et al., 2001; Nomura et al., 2001; Raturi et al., 2008).

As described above, pMVs are considered to be both biomarkers and effectors of cell signaling. Importantly, circulating MVs can mediate communication between vascular cells because they allow membrane interactions between cells at distance (Meziani et al., 2008). Elucidation of the molecular mechanisms by which MVs might evoke and promote vascular inflammation, atherosclerosis plaque progression, and thrombus formation (Figure 2) is crucial for improving our understanding of their role in health and disease. Hereby, we present a compilation of functional studies investigating the underlying pMV-driven molecular processes in atherothrombosis by using either in vitro, in vivo, or ex vivo models (Table 1):

Within EVs platelet-derived exosomes, first described by Heijnen et al. (1999), have received increasing attention due to their potential role as mediators of cell-to-cell communication. The presence of exosomes containing miRNAs in circulating blood, mainly derived from platelets, has emerged as a potential source of biomarkers of CVD. Additionally, they have been implicated in vascular and heart functions. Ago-miR-223 is delivered to EC via microvesicles from activated platelets (Laffont et al., 2013) and platelet-derived exosomes reduce ICAM-1 expression decreasing monocyte adhesion (Gidlof et al., 2013) and induce EC apoptosis (Janiszewski et al., 2004; Gambim et al., 2007). Thrombin-stimulated platelet-derived exosomes, enriched in miR-223, miR-339, and miR-21, inhibit PDGFR β expression in SMCs (Tan et al., 2016). Moreover, exosomes may serve as effectors by which damaged heart communicate for cardiac injury repair in the setting of MI (Ottaviani et al., 2016). It is not clear so far whether the intervention on exosome secretion processes during disease would have therapeutic effects. However, in the next upcoming years, study of exosomes molecular insights may help not only to detect novel biomarkers of disease but also to provide potential new therapeutic approaches for atherothrombotic cardiovascular disease.

Platelet-derived microvesicles may not only act as a biomarkers of cell activation but also as important functional effectors that linking inflammation, hypercoagulability and neovascularization may contribute to the exacerbation of atherosclerotic lesion growth and to ongoing thrombosis. Thus, pMVs are emerging as novel and specific pharmacological targets to manage atherothrombosis and, additionally as potential therapeutic tools for drug delivery, cardioprotection, and regenerative and personalized medicine. Although our understanding of pMVs has considerably been expanded in the last decade, we are far from completely understanding platelet microvesicle biology. In the upcoming years, platelets and platelet-derived MV research will likely advance and take a leading position in tackling down CVD.

Further studies are required to improve our knowledge on the mechanisms underlying generation of circulating pMV, from the characterization of the specific phenotype of released pMVs by distinct stimuli and pathophysiological context, to the identification of the complete composition and its influence on other cells, and to the disentangling of molecular and signaling processes involving pMVs as functional effectors in atherosclerosis and atherothrombosis.

LB, IP, and TP conceived and coordinated the design of the review. LB, RS, and EF wrote the paper. RS and EF made figures; LB and RS edited the paper. All authors wrote part of the manuscript, provided critical comments, revised the manuscript and approved the final version of the manuscript.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. The reviewer JZ and handling Editor declared their shared affiliation, and the handling Editor states that the process nevertheless met the standards of a fair and objective review.