Circulating microparticles as biomarkers of stroke: A focus on the value of endothelial‐ and platelet‐derived microparticles

Stroke is one of the leading causes of mortality and disability worldwide. Numerous pathophysiological mechanisms involving blood vessels, coagulation and inflammation contribute to the vascular occlusion. Perturbations in these pathways can be detected by numerous methods including changes in endoplasmic membrane remodeling and rearrangement leading to the shedding of microparticles (MPs) from various cellular origins in the blood. MPs are small membrane‐derived vesicles that are shed from nearly all cells in the body in resting state or upon stimulation. MPs act as biological messengers to transfer information to adjacent and distant cells thus regulating various biological processes. MPs may be important biomarkers and tools for the identification of the risk and diagnosis of cerebrovascular diseases. Endothelial activation and dysfunction and altered thrombotic responses are two of the main features predisposing to stroke. Endothelial MPs (EMPs) have been recognized as both biomarkers and effectors of endothelial cell activation and injury while platelet‐derived MPs (PMPs) carry a strong procoagulant potential and are activated in thrombotic states. Therefore, we reviewed here the role of EMPs and PMPs as biomarkers of stroke. Most studies reported high circulating levels of EMPs and PMPs in addition to other cell origins in stroke patients and have been linked to stroke severity, the size of infarction, and prognosis. The identification and quantification of EMPs and PMPs may thus be useful for the diagnosis and management of stroke.

within predefined populations (Sacco et al., 2013). The World Health Organization defined stroke in the 1970s as "rapidly developed clinical signs of focal (or global) disturbance of cerebral function, lasting more than 24 hr or leading to death, with no apparent cause other than of vascular origin" (Aho et al., 1980). However, since significant advances have been made in the clinical and radiographic diagnosis of stroke and its mimics, this definition is now considered outdated and does not reflect the heterogeneity and complexity of the disease. Consequently, in 2013, the AHA and the American Stroke Association developed an expert consensus document for an updated definition of stroke, taking into account the remarkable advances in brain imaging and its clinical implications on the diagnosis of stroke and its subtypes (Coupland, Thapar, Qureshi, Jenkins, & Davies, 2017;Sacco et al., 2013). This document summarized different stroke etiologies and enlisted them in a table for a better definition of stroke.
Stroke can be classified into two major types: ischemic (~80%) which is due to cerebral vessels' blockade and hemorrhagic stroke (~20%) which is due to brain hemorrhage (Ojaghihaghighi, Vahdati, Mikaeilpour, & Ramouz, 2017). Each type is further classified into subtypes according to their etiology. Ischemic stroke has been classified according to the Trial of ORG 10172 in Acute Stroke Treatment (TOAST) classification into five subtypes based on clinical features and other prognostic: large artery atherosclerosis, small vessel occlusion, cardioembolism, stroke of other determined etiology, and stroke of undetermined etiology (Adams et al., 1993).
Hemorrhagic stroke is divided into two subtypes based on bleeding location in the brain to; intracerebral hemorrhage (ICH) where bleeding is directly in brain parenchyma or subarachnoid hemorrhage (SAH) where bleeding takes place between arachnoid space and pia mater surrounding the brain (Ojaghihaghighi et al., 2017). However, other classifications are available for both types of stroke as well.
There is no single outcome that can accurately measure and predict postacute strokes disability and prognosis (Kasner, 2006). Many scales have been developed to measure stroke outcomes including the modified Rankin scale, the National Institutes of Health stroke scale (NIHSS), the Barthel index, the stroke impact scale, and the Glasgow outcome scale (Harrison, McArthur, & Quinn, 2013;Kasner, 2006). In addition, each scale has its advantages and pitfalls that must be taken into consideration before choosing a certain one (Kasner, 2006).
Early and accurate stroke diagnosis is crucial in patient care to adopt the right therapeutic strategies and prevent permanent disabilities (Ojaghihaghighi et al., 2017). For instance, intravenous thrombolysis with recombinant tissue plasminogen activator (rtPA) which is currently the first line drug used for acute ischemic stroke treatment, is believed to be most effective when administered within the first few hours of stroke onset (Gumbinger et al., 2014). Various clinical and radiographic tools are used to diagnose stroke (Sacco et al., 2013). Clinical manifestations and findings help clinicians to determine the type of stroke but usually, brain imaging is required (Ojaghihaghighi et al., 2017). Computed tomography and magnetic resonance imaging are used to recognize the brain lesion location, shape, size and extent and is helpful to determine the cause of stroke and exclude mimics (Sacco et al., 2013). Blood biomarkers, including S100 calcium binding protein, brain natriuretic peptides, and matrix metalloproteinase-9 have also been evaluated in the diagnosis of acute stroke (Jickling & Sharp, 2011;Sacco et al., 2013). However, these markers lack specificity since they are generated in other diseases that can also affect the blood-brain barrier (Kim, Moon, & Bang, 2013). Therefore, it is necessary to explore novel effective biomarkers to be used in the identification of high-risk patients, rapid diagnosis, determining stroke etiology and pathogenesis, predicting drug response and clinical outcomes, for novel drug development and as endpoints in clinical trials (Bang, 2017). Microparticles (MPs), which are small membrane vesicles released from activated cells, are being currently heavily investigated as promising biomarkers and vectors for multiple cerebrovascular disorders including stroke.
The primary purpose of this review is to highlight the emerging role of MPs in cerebrovascular disease and specifically focusing on MPs of endothelial and platelets origins because they reflect, respectively, endothelial activation and thrombotic activity, two major contributors to the pathophysiology of ischemic stroke. We will initially review the current understanding of the roles of MPs and then focus on their importance in the understanding of the cerebrovascular disease.

| CELL MEMBRANE REMODELING AND MPs RELEASE
Extracellular vesicles (EVs) are submicron, intact membrane-derived vesicles that are released from cells after activation or during apoptosis (Shet, 2008). These vesicles harbor membrane constituents such as phospholipids, integral membranes, and receptors as well as cytoplasmic cargo including various proteins, DNA and RNA adopted from their cells of origin (Hargett & Bauer, 2013;Hugel, Martinez, Kunzelmann, & Freyssinet, 2005). There are several types of EVs characterized based on their size and formation/release pathways including, MPs, exosomes, oncosomes, and apoptotic bodies (Zaborowski, Balaj, Breakefield, & Lai, 2015). MPs were first discovered in 1967 by Wolf during his ongoing coagulation studies when he observed subcellular coagulant vesicles in platelet-free plasma to which he referred to as "platelet dust" (Wolf, 1967).
Subsequently, numerous studies have been conducted to further investigate the structure, biogenesis, and functions of these vesicles gradually replacing it by the term MPs (Crawford, 1971;Hargett & Bauer, 2013;Webber & Johnson, 1970). Over the past 40 years, the understanding of MPs has tremendously advanced from being viewed as useless cell debris or cellular trash to one of the key cellular players regulating various physiological and pathological processes throughout the body (Hargett & Bauer, 2013). In 2005, a standard definition for MPs has been proposed by The International and Hemostasis Vascular Biology subcommittee. They described MPs as 0.1-1 μm cell-derived vesicles that lack a nucleus or synthetic capacity, may contain cytoskeletal proteins, and expose some quantity of phosphatidylserine on their surfaces (Hargett & Bauer, 2013).

| MECHANISMS OF MP FORMATION AND SHEDDING
The shedding of MPs from the plasma membrane is an essential part of cell membrane remodeling and is thought to be initiated by increased intracellular calcium (Shet, 2008). The phospholipid cell membrane bilayer is well-structured and each of the two leaflets has a specific lipid composition. Phospahtidylcholine and sphingomyelin are expressed on the outer leaflet while aminophospholipids (phosphatidylserine [PS] and phosphatidylethanolamine) are oriented inwards (Hugel et al., 2005). Membrane enzymes; floppase, scramblase, and aminophospholipid translocase (flippase) regulate phospholipid movements and maintain this dynamic asymmetry in a steady state. Upon cell stimulation, cytosolic calcium increases which in turn dysregulates membrane enzymatic balance and leads to the collapse of membrane asymmetry and subsequent externalization of PS. In addition, intracellular calcium activates calpain and Rho kinase which together causes cytoskeleton cleavage and rearrangement facilitating thus membrane blebbing as well as activating apoptosis via a caspase 3-mediated pathway. However, it must be noted that other stimuli that cause disruption of membrane integrity can also lead to the formation of MP (Hugel et al., 2005). Moreover, it has been shown that MPs can be released from inactivated platelets without intracellular calcium signaling and it was observed that this mechanism was dependent on αIIβ3 integrin and cytoskeleton turnover (Cauwenberghs et al., 2006). Therefore, other pathways are also most probably involved in the formation of MPs hence adding to the complexity of the biology of MP formation and release.
Further research is required to fully elucidate the mechanisms governing the formation and shedding of MPs from cell membranes.

| PROPERTIES AND CHARACTERISTICS OF MPs
MPs have been found to be present in bloodstream of healthy individuals as well as in patients (Berckmans et al., 2001;Martinez, Tesse, Zobairi, & Andriantsitohaina, 2005). This suggests that MPs can be beneficial, maintaining homeostasis under normal physiological conditions, or deleterious if overproduced or if they carry pathogenic constituents on their surface or in cargo (Hugel et al., 2005). The concept of MPs is interesting as they are considered a circulating storage pool of bioactive molecules that if interpreted correctly can provide valuable insight about the body state and can be used as biomarkers for various pathological states (Hugel et al., 2005). In addition, MPs act as a form of effective intercellular communication system regulating various fundamental biological processes (Hoyer, Nickenig, & Werner, 2010). Studies have shown that MPs convey biological information by either direct interaction with cell receptors/integrins initiating cell signaling cascades or by direct fusion with effector cells by endocytosis and the subsequent emptying of MPs content (Hargett & Bauer, 2013). In this sense, MPs can act as signaling molecules via the expression of different proteins and lipids on their surface such as tissue factor (TF) and PS (Hoyer et al., 2010). Alternatively, MPs have been shown to transfer complete receptor proteins (e.g., CCR5 receptor), mRNA, microRNAs, proteins, and even cellular organelles to recipient cells (Deregibus et al., 2007;Hoyer et al., 2010;Mack et al., 2000;Ratajczak et al., 2006). Nevertheless, more studies are required to confirm that molecular material transferred to recipient cells are biologically functional and significantly affecting cellular responses in target cells (Zaborowski et al., 2015). MPs are very heterogeneous and vary widely depending on the stimulus and cell of origin. For example, the release of MPs in vitro can be induced using a plethora of agents and the resulting MPs vary in structure and composition (Herring, McMichael, & Smith, 2013). For instance, previous studies have shown that in vitro stimulation of human lymphoid CEM T cell line using two different pharmacological stimuli generated antigenically and functionally different MP populations. When the CEM cell line was stimulated with phytohemagglutinin for 72 hr, followed by phorbol-12-myristate-13 and actinomycin D for further 24 hr, generated MPs harbored the morphogen Sonic hedgehog (Shh) and were found to improve endothelial function both in vitro and in vivo (Agouni et al., 2007). However, when the same cell line was stimulated with an apoptotic signal using actinomycin D for 24 hr, MPs generated failed to express the morphogen Shh and caused a deleterious effect on endothelial function both in vitro and in vivo (Mostefai et al., 2008). These studies highlight the importance of the stimulus at the origin of generating MPs in conditioning the antigenic content and biological messages carried out by the shed EVs.

| FUNCTIONS OF MPs
MPs are involved in various vital biological functions such as hemostasis, coagulation, inflammation, and angiogenesis (Herring et al., 2013). It has been shown that some MPs may have procoagulant activity due to the externalization of PS which acts as a catalytic surface for enzymatic coagulation complexes assembly (Mooberry & Key, 2016). In addition, some MPs harbor TF on their surface which is considered to be the main initiator of the coagulation cascade in the extrinsic pathway (Meziani, Tesse, & Andriantsitohaina, 2008). Remarkably, one study showed that platelet-derived MPs (PMPs) have 50-100 more procoagulant activity when compared to activated platelets (Sinauridze et al., 2007). In addition, MPs have been shown to play a role in hemostasis disorders such as Scott Syndrome, hemophilia A and Von Willebrand disease (Mooberry & Key, 2016). There is also evidence showing that MPs may have anticoagulant and fibrinolytic activity as well (Mooberry & Key, 2016). MPs have been shown to carry TF pathway inhibitor on their surface and to promote the activity of the anticoagulant protein C (Koshiar, Somajo, Norstrom, & Dahlback, 2014;Kushak et al., 2005;Steppich et al., 2005;Tans et al., 1991).  providing a source of aminophospholipids, and participating in the release of cytokines (Meziani et al., 2008;Puddu, Puddu, Cravero, Muscari, & Muscari, 2010). Furthermore, numerous studies have shown that MPs can regulate angiogenesis and modulate several steps in blood vessels' formation (Martinez & Andriantsitohaina, 2011). The effects of MPs on the vascular system have been investigated intensively as well. Although it has been shown that some MPs have positive effects on endothelial function, many studies demonstrated that MPs can induce endothelial dysfunction (Agouni et al., 2007;Mostefai et al., 2008). Some of the mechanisms by which MPs induce endothelial dysfunction include, reducing nitric oxide concentration, inducing inflammation, promoting coagulation, altering angiogenesis and apoptosis (Lovren & Verma, 2013). A cellular MP and its proposed functions on coagulation, angiogenesis, inflammation, and vascular function are illustrated in Figure 1.

| CHALLENGES WITH THE ANALYSIS OF MPs IN CLINICAL PRACTICE
Given the importance of MPs as potential biomarkers of disease and vectors of biological messages, they have attracted attention from the scientific community; however, various challenges related to the methods of detection and analysis of MPs are hindering the wide application of MPs in clinical use. Such challenges include: 1. Methods to isolate MPs are not fully developed and standardized yet. Most of the current isolation methods allow enrichment rather than separation of EV subtype populations (Zaborowski et al., 2015). Each of the methods available has its pitfalls and there is no standard protocol for MP isolation. For instance, a series of ultra-centrifugations yields are means of good separation of EVs; however, it is often contaminated with high molecular weight protein complexes and cannot differentiate between EVs subtypes and cells of origin. Immunoaffinity-based assays provide better enrichment of EV-specific subtypes and provide more homogenous EV profiles, but the overall yield is low. Many MPs subtypes share the same CD pattern of expression and therefore, combining multiple markers to exclude possible subpopulation contamination has been used to overcome this issue. For example, Simak, Gelderman, Yu, Wright, and Baird (2006) used cell-specific antibodies combinations to ensure the detection of endothelial MPs (EMPs) while making sure to eliminate MPs of platelet, leukocyte or erythrocyte origins.
2. MP quantification is highly challenging. Maas et al. (2015) conducted a thorough comparison between three of the most common methods used to quantify MPs namely; nanoparticle tracking analysis, tunable resistive pulse sensing and flow cytometry and found substantial differences in EVs counts among these techniques. The study also reiterated the importance of proper technical knowledge of the instruments and their settings and their effect on correct data interpretation (Maas et al., 2015).
3. Handling and processing of blood samples are very crucial and may introduce variability when comparing different studies.
Aspects of sample collection, including needle size, the use of a tourniquet, the use of syringe versus vacutainer, and the type of anticoagulant used. In addition, special attention should be given to speeds of centrifugation, number of centrifugations, freeze/thaw cycles and the duration of MPs storage (Shet, 2008).

4.
Proper identification of MPs is also very challenging. Arraud et al. (2014) conducted an interesting study to examine MPs morphology and size in which they found that MPs (which were mainly of platelet and erythrocyte origins) are spherical and tubular in shape. Although tubular MPs constituted a small percentage, the total membrane surface area was of the same order of magnitude as spherical MPs. In addition, they found that the majority of MPs in their study did not bind Annexin V which is in contrast with the theory that MPs expose PS when they are formed. They have concluded that these MPs might be formed via a mechanism that maintains the membrane phospholipid asymmetry such as direct cell fragmentation or breakdown from PS-negative tubular MPs (Arraud et al., 2014).
Since not all studies take these challenges into account, especially older ones, it is very hard to compare and analyze findings reported from various studies.
Several studies are now being conducted to investigate the potential role of MPs in the context of stroke. It was recently reported in a systematic review that circulating MPs from various cell origins increase in ischemic stroke patients (Wang et al., 2018). Since endothelial dysfunction and thrombotic dysregulation are two major contributors to the pathophysiology of stroke, in this review we focused on the role of endothelial-derived MPs (EMPs), as key biomarkers of endothelial activation and dysfunction, and PMPs, as biomarkers of thrombotic state, in the development of stroke. We also reviewed their value as biomarkers in the onset and progression of the disease. Surface antigens that characterize EMPs and PMPs and their functions are summarized in Table 1.

| EMPs AND STROKE
Endothelial dysfunction is evident early and is recognized as an important initiator of cerebrovascular disease (Lovren & Verma, 2013). MPs shed from endothelial cells are involved in inflammation, procoagulation, and angiogenesis and thus are considered as biomarkers of endothelial cell activation and injury . Profiling of EMPs in stroke may provide an understanding of the pathogenesis of the disease and reflect its prognosis. Distinct been associated with recent ischemic attacks and increased risk of cardiovascular events; however, no association has been found between CD62E + EMPs and stroke onset or recurrence (Chiva-Blanch et al., 2016;Jung et al., 2009;Lee et al., 2012). Since Eselectin is a member of the adhesion molecule family that facilitates inflammatory cells' adhesion and rolling, these findings suggest the following hypotheses. First, high levels of CD62E + EMPs without being correlated to the time of stroke onset demonstrate continuous endothelial cell activation. Second, endothelial dysfunction in ischemic stroke is more likely to be due to cell activation rather than apoptosis. Finally, the lack of correlation between CD62E + EMPs and stroke recurrence while its association with cardiovascular events and hospitalization may suggest general endothelial dysfunction rather than pathology of cerebral ischemic origin (Chiva-Blanch et al., 2016;Jung et al., 2009;Lee et al., 2012).
There is some suggestion that changes in EMPs may be used as a biomarker of stroke severity (Chiva-Blanch et al., 2016;Jung et al., 2009;Li & Qin, 2015;Simak et al., 2006). Higher levels of EMPs have been found in patients with moderate/severe stroke versus mild stroke patients. Stroke severity was assessed using NIHSS scores where a score greater than five defined moderate/severe stroke EL-GAMAL ET AL.

| 16743
while scores less than five indicated a mild stroke. Li and Qin (2015) have shown a positive correlation between CD144 + (VE-cadherin) EMPs and stroke severity. Another study using the same scale found an increase in some EMPs (including CD144 + ) in moderate to severe patients compared to mild stroke patients (Simak et al., 2006). Therefore, CD144 + EMPs have the potential to be used to assess stroke severity but further evidence is needed to confirm these findings. In addition, EMPs can also be used in stroke diagnosis and identification of its etiology and subtypes. Simak et al. (2006) have shown some correlation between EMPs counts (CD105 + and CD54 + ) and brain lesion which can be useful to identify the extent of brain damage. In addition, Li and Qin (2015) showed a mild correlation between CD62E + EMPs and OCSP classification which categorizes stroke subtypes according to ischemic lesion location. Therefore, CD62E + EMP values may predict stroke subtypes. Furthermore, the ratio of CD62E + /CD31 + (E-selectin/PECAM-1) EMPs has been shown to be able to distinguish between the cause of arterial stenosis and whether it is of intracranial or extracranial origin thus providing potential information about stroke etiology (Jung et al., 2009).
EMPs may also be useful in prognosis in patients with SAH. One study conducted on patients with a hemorrhagic stroke caused by SAH showed elevated levels of EMPs positive for CD54 + (ICAM-1), CD62E + (E-selectin), and CD106 + (VCAM-1). Severe SAH may be followed by complications including cerebral vasospasm (CVS) and cerebral infarction attributable to vasospasm (CIV). This study measured EMPs in patients with CVS and CIV and found that only CD105 + (endoglin) EMPs were upregulated. Since endoglin is considered to be a marker of apoptosis, this finding showed that apoptosis may be a cause of endothelial damage in the early onset of CVS following hemorrhagic strokes (Lackner et al., 2010). It is also worth noting that Simak et al. (2006)  Constitutively expressed at endothelial adherence junctions. It plays a role in controlling vascular permeability and leukocyte extravasation.
Vestweber (2008) CD31 Platelet and endothelial cell adhesion molecule (PECAM-1) It is expressed in most vascular compartment cells. It is found at cell junctions in endothelial cells and play various roles in inflammation and vascular biology.
Woodfin, Voisin, and Nourshargh (2007) CD54 Intercellular adhesion molecule (ICAM-1) It is an inducible cell adhesion protein that plays a role in leukocyte and endothelium interaction to regulate vascular permeability. It is also induced in inflammation and is expressed on a wide range of immune cells such as monocytes and macrophages. Roebuck and Finnegan (1999) CD146

Melanoma cell adhesion molecule (MCAM)
Adhesion molecule involved in cell signaling, vascular permeability, and immune response.

It is a transmembrane glycoprotein and is a marker of endothelial cell activation and induced in inflammation.
Ley and Huo (2001) Platelet-derived MPs (EMPs) CD41 αIIb chain Bind together to form the glycoprotein GPIIa/IIb (integrin α IIb β 3 ) which is a member of the integrin transmembrane family. It is expressed on platelets and is essential for platelets aggregation. It acts as a receptor for fibrinogen and von Willebrand factor (vWF) to facilitate platelets aggregation and cross-linking.

Fullard (2004) CD61
β3 integrin CD42a GP IX Two membrane glycoproteins that bind together to form (GP)Ib-IX-V complex. (GP)Ib-IX-V is expressed on platelets surface and is involved in thrombosis and acts as a receptor for vWF and other molecules such as thrombin.

Li and Emsley (2013) CD42b
GP1bα CD62P P-selectin Also known as Platelet Activation-Dependent Granule to External Membrane Protein (PADGEM) or Granule Membrane Protein 140 (GMP-140). It is a transmembrane glycoprotein that is expressed by activated platelets and plays a key role in immune cells adhesion and rolling. Koedam et al. (1992) Abbreviation: PMP, platelet-derived microparticle.

T A B L E 2 Summary of key studies that investigated EMPs and PMPs in stroke patients
Type of stroke  And after 4-weeks of antiplatelet therapy Measuring other platelet parameters along with PMPs.
AIS patients divided based on TOAST classification into: • No significant difference in PMP levels between LAA and SAO subtypes Weaknesses: Using only one marker for detecting PMPs (CD61) 1. Large artery atherosclerosis subtype (LAA) • PMP levels decreased in both subtypes after antiplatelet therapy (prognosis and evaluation of drug response)

2.
Small artery occlusion (SAO) • Weak correlation between PMP levels and infarct volume in LAA subtype • PMPs could be an independent risk factor for AIS patients Acute ischemic stroke (AIS) patients receiving recombinant tissue plasminogen activator (rt-PA) therapy divided into: After 24 hr of stroke onset Refs.
• Following-up using mRS to assess outcomes  • PMPs decreased after 6 months in S & L patients (more in acute than chronic phase)

Stroke of undetermined etiology (U)
• No direct relation between PMPs and infarct size (Continues) EMPs as a marker of endothelial activation and CD31 + (PECAM) EMPs as a marker of apoptosis in both groups. The ratio of CD62E + / CD31 + (E-selectin/PECAM) EMPs was more than four in both groups indicating endothelial injury. However, the study showed no difference in EMPs count between stroke and mimic groups questioning the reliability of using EMPs as a marker to discriminate between them. There were some limitations in this study including the low NIHSS score (small infarct volume) in stroke patients and the long sample storage time that may have affected the number of EMPs resulting in the underestimation of MP counts (Williams et al., 2007). Further studies may be required to confirm these findings and judge the effectiveness of EMPs as biomarkers in distinguishing between acute stroke and its mimics.
In conclusion, EMPs evaluation studies conducted on stroke patients present a number of potential benefits in clinical settings but require further research. EMPs may potentially be used as biomarkers, provide insights about stroke severity and infarct lesion, differentiate between stroke subtypes and identify etiologies thereby, helping clinicians to make more effective therapeutic decisions.
The ability to differentiate between stroke and mimics may also be very useful. The main studies reported in this section are summarized in Table 2.

| PMPs AND STROKE
PMPs are the most abundant MPs circulating in the blood (Italiano, Mairuhu, & Flaumenhaft, 2010). Collection of PMPs requires careful processing because all platelets may not be successfully depleted from samples before measuring and thus may cause misinterpretation (Burnouf, Goubran, Chou, Devos, & Radosevic, 2014). This may be addressed by adding more centrifugation cycles which can decrease the content of platelets to up to 0.02%. PMPs are measured by many ways including physical methods such as flow cytometry and dynamic light scattering (Lawrie, Albanyan, Cardigan, Mackie, & Harrison, 2009;Shantsila, Montoro-Garcia, Gallego, & Lip, 2014).
Other methods include assays that measure PMPs' procoagulant activity, ELISA assays, and proteomics (Burnouf et al., 2014;Capriotti et al., 2013;Osumi et al., 2001). In addition, cellular assays are conducted in vitro to measure the functional impact of PMPs on cell proliferation, angiogenesis, inflammation, and the release of cytokines (Burnouf et al., 2014). Flow cytometry and ELISA are the most commonly used methods to determine levels of PMPs in clinical trials.
PMPs carry the surface antigens CD41 + (αIIb chain) and CD61 + (β3 integrin) which bind together to form glycoprotein (GP)IIa/IIb Patients with chronic phase of cerebral infarction (more than 3 months) versus controls More than 3 months from stroke onset • ELISA • Higher PMP levels in patients with chronic cerebral infarction that have remained elevated after antiplatelet therapy (4-week course) Strengths: et al. (2008) • Comparing PMP levels before and after antiplatelet therapy Weaknesses:

Shirafuji
• A small number of patients EL-GAMAL ET AL.

| 16749
There are conflicting reports on PMPs following acute stroke or response to treatment. This may be attributed to differences in methods of detection. For instance, studies have shown elevated PMPs levels after an acute ischemic stroke that remain elevated after antiplatelet therapy. Other studies showed a significant decrease in PMPs following the initiation of antiplatelet medications (Chen et al., 2015;Chiva-Blanch et al., 2016;Kuriyama et al., 2010;Shirafuji, Hamaguchi, & Kanda, 2008). Similarly, some studies associated elevated PMPs with the acute phase of cerebral infarction while others linked it to the chronic phase (Kuriyama et al., 2010;Shirafuji et al., 2008). Nevertheless, most of the studies showed that there are higher levels of PMPs post-stroke (ischemic and hemorrhagic) which means that PMPs can be used as biomarkers for diagnosis; however, the benefits of measuring PMPs levels for assessing clinical outcomes after antiplatelet therapy remain to be elucidated (Chen et al., 2015;Chiva-Blanch et al., 2016;Kuriyama et al., 2010;Lackner et al., 2010;Pawelczyk, Baj, Chmielewski, Kaczorowska, & Klimek, 2009;Shirafuji et al., 2008). PMPs may be markers of successful treatment following an acute stroke. PMP levels were related to better reperfusion and recanalization outcomes in acute ischemic stroke patients receiving rt-PA making it a potential candidate to assess recanalization in stroke patients (Bivard et al., 2017). However, it was noted that PMP levels were not related to infarct size and only a weak correlation has been observed between PMP levels and infarct volume in LAA subtype (Chen et al., 2015;Kuriyama et al., 2010). Taken together, PMPs appear to be good biomarkers for stroke diagnosis and determining stroke etiology but subsequent measuring for follow-up purposes needs more standardization for better identification of the value of PMPs as biomarkers for stroke recurrence and prognosis. PMPs may also be of potential value in determining reperfusion in patients being treated with rt-PA. The major studies discussed in this section are summarized in Table 2.

| CLINICAL UTILITY OF MPs IN STROKE
Since MPs play a pivotal role in the pathophysiology of different diseases, its clinical applications have been widely investigated. One of the approaches is to use MPs to evaluate and monitor therapeutic efficacy (Chen, Li, & Liu, 2018). For example, some drugs such as statins, aspirin and anti-oxidants have been shown to reduce MPs levels in patients and thus can be used to assess treatment effectiveness (Baron, Boulanger, Staels, & Tailleux, 2012;Bulut, Becker, & Mugge, 2011;Morel et al., 2003;Suades, Padro, Alonso, Mata, & Badimon, 2013). Another approach is utilizing MPs characteristics as therapeutic agents. For instance, the ability of some MPs to stimulate the formation of new blood vessels has been shown to induce revascularization and improve mouse kidney function in an ischemia/reperfusion injury model (Ranghino et al., 2012). In addition, as stated earlier, MPs carry genetic material cargo such as DNA, mRNA, and microRNA which can be transferred to distal cells (Hugel et al., 2005). Therefore, MPs have the potential to be used in gene therapy or as nontoxic therapeutic delivery tools (Chen et al., 2018). MPs provide the advantage of being nontoxic and stable in blood; however, the effectiveness of the genetic material or drug delivery to target cells is still to be elucidated (Chen et al., 2018). Furthermore, due to MPs coagulant activity especially of PMPs, their use as prothrombotic agents has been explored to treat bleeding disorders like thrombocytopenia (Blajchman, 2003;Piccin, Murphy, & Smith, 2007). Therefore, various functions of MPs could be utilized to provide therapeutic effects. Also, one of the main promising clinical applications of MPs is their use as biomarkers and diagnostic tools. As mentioned earlier, MPs levels vary with many diseases and therefore could be used for diagnosis (Hugel et al., 2005). Researchers tried to investigate changes in circulating MPs patterns of expression that can be distinctive to certain phenotypes of disease (Berezin, 2016). For example, studies by Berezin et al. have widely investigated the imbalance between MPs released from apoptotic versus activated endothelial cells, to which has been referred to as "impaired phenotype", to provide predictive scores for heart failure assessment and predicting cardiovascular risks (Berezin, 2016;Berezin, Kremzer, Samura, Berezina, & Kruzliak, 2015). These scores have been shown to provide better predictive values when combined with other markers in heart failure patients . A similar score has also been used for stroke diagnosis where a ratio between CD62E/CD31 has been used rather than absolute numbers to discriminate between endothelial activation and apoptosis and to identify the site of arterial stenosis (Jimenez et al., 2003;Jung et al., 2009). In the context of stroke, the use of MPs as diagnostic tools could provide potential value given challenges in stroke's diagnosis. Imaging equipments are expensive, not easily available, time-consuming, and require professional interpretation (Saenger & Christenson, 2010).
Thus, there is a need for novel stroke biomarkers characterized by early and steady release after infarction, improved sensitivity, and specificity over current biomarkers that could differentiate between stroke types and stroke mimics (Saenger & Christenson, 2010). MPs are being investigated as a promising candidate, but their use is faced by some challenges with regard to sample handling and reliable measurement tools as discussed earlier in this review. Nevertheless, as referred in Figure 2, promising fields include the use of MPs to predict the severity of infarct, site of injury and differentiate between different phenotypes. In addition, MPs can also be used to monitor therapeutic efficacy and predict survival rates as shown by Bivard et al. (2017) relating PMPs with better recanalization outcomes.
More studies are warranted, however, to confirm the value of MPs in stroke treatment and diagnosis.

| CONCLUSION
Normal basal levels of MPs reflect a tightly controlled balance between cell survival, proliferation, and death. Based on the nature of the stimulus that causes their release, this balance is shifted to either direction with changes in the levels and composition of MPs and upon which MPs could be good or bad. Given the complexity of MPs functions and roles, it is still a matter of debate whether they are a cause or a consequence, friends or foes, active participants or passive observers. Nevertheless, MPs acting as remarkable biological sensors are very attractive targets for studying and exploring novel and personalized therapeutic strategies. In this review, we reviewed in more detail the roles of EMPs and PMPs in stroke patients (studies are summarized in Table 2). Despite the challenges, EMPs and PMPs could provide potential benefits as biomarkers, diagnostic tools and prediction of stroke severity (promising markers are illustrated in Figure 2).

However, standardization of MPs detection and quantification methods
is highly required to further confirm and generalize the results.

CONFLICT OF INTERESTS
The authors declare that there are no conflicts of interest.