WO2012090149A1

WO2012090149A1 - Methods for analyzing components of biological fluids

Info

Publication number: WO2012090149A1
Application number: PCT/IB2011/055955
Authority: WO
Inventors: Zouher Majd; Geneviève CORDONNIER
Original assignee: Genfit SA
Current assignee: Genfit SA
Priority date: 2010-12-28
Filing date: 2011-12-26
Publication date: 2012-07-05
Anticipated expiration: 2013-06-28

Abstract

Methods for isolating, characterizing, comparing and using specific components of biological fluids that are characterized on the basis of post-translational modifications of cell surface, and that can be used for the medical management of disorders associated to endothelial dysfunctions. More specifically, said method is for prognosing, diagnosing or monitoring an endothelial dysfunction in a subject, comprising: a) isolating Glyco-EMPs populations in a test sample of biological fluid of said subject by means of an agent binding a specific glycan epitope and at least one further binding agent recognizing a CD antigen from endothelial origin; b) characterizing said test sample Glyco-EMPs populations, and c) comparing said characterized test sample Glyco-EMPs populations with a control sample of Glyco-EMPs populations. Step a) may optionally includes the use of a further binding agent recognizing a phospholipid.

Description

METHODS FOR ANALYZING COMPONENTS OF BIOLOGICAL FLUIDS

TECHNICAL FIELD

The present invention relates to methods for isolating, characterizing, comparing and using specific components of biological fluids that are characterized on the basis of post- translational modifications of cell surface, and that can be used for the medical management of disorders associated to endothelial dysfunctions.

BACKGROUND OF THE INVENTION

Despite huge efforts made by the scientific community, there is still an urgent need in identifying novel biomarkers for a number of chronic conditions including metabolic and cardiovascular diseases, by using biological fluids that can be more easily obtained and analyzed in clinical settings. For example, selected plasma proteins are now considered as relevant biomarkers of various clinical conditions and the use of "omics" technologies should lead to the discovery of more (non-)protein biomarkers and other complex profiles such as biochemical and metabolic signatures (Bain J et al., 2009; Turer A et al., 2009).

Biomarker research activities suffer from a major technical problem related to the presence of extremely abundant proteins within biological fluids that mask the less abundant sub-proteome that can provide reliable biomarkers of medical interest. In the case of blood or cerebrospinal fluids, specific analytical and/or protein-depletion technologies are used to exclude proteins such as albumin, immunoglobulins, or transferrin and try circumventing this problem (Thouvenot E et al., 2008; Apweiler R et al., 2009a). However, major technical improvements are still needed for characterizing the more variable, but minor, fractions of circulating proteome that are difficult to separate such as proteins on cell surface membrane, which are of major interest as potential novel drug targets and biomarkers (Lai Z et al., 2009; Cordwell S and Thingholm T, 2009).

In the field of membrane proteomics within biological fluids, a growing number of preclinical and clinical studies indicate circulating microparticles (MPs) as potential biomarkers and/or additional mediators of protein-based signalling through cell surface receptors or transporter. As extensively reviewed (Miguet L et al., 2007; Doeuvre L et al., 2009; Burnier L et al., 2009), MPs are vesicles having a size comprised between 1 and 0.1 micrometer that are formed by the plasma membrane and shed from apoptotic or otherwise activated cell types in response to various conditions and stimuli (chemicals, growth factors, shear stress, apoptotic signals, etc.). MPs composition and surface antigens are dependent from cell origin and/or the stimulus that triggered their generation.

MPs formation is associated with the loss of membrane asymmetry and the exposure of specific phospholipids, such as phosphatidylserine, on the outer leaflet which, together with MPs surface antigen, are responsible of strong procoagulant activity that MPs normally exhibit. Phospholipid-binding agents, in particular proteins (Stace C and Ktistakis N. 2006; Lemmon M, 2008), are used for the affinity-based isolation of MPs, after being separated from cellular components within the biological fluid by centrifugation.

The MPs variable concentration and composition, release mechanism, and activities in biological fluids have been extensively studied in plasma or urine, as well as in cell culture conditions, but other biological fluids, such as cerebrospinal or synovial fluids, have been indicated as containing vesicles features that are similar to those found for plasma MPs. However, many of the studies have not been pursued in a standardized manner regarding the MPs purification from biological fluids (Piccin A et al., 2007). In fact, other types of vesicular particles, called exosomes, are also released after cell activation and can be co-purified with MPs, but they diverge from MPs in size (in general below 0.1 micrometer), surface antigens (given their intracellular origin), and absence of phosphatidylserine, leading to potential confusions and doubts on the actual relevance of some published data (Thery C et al., 2009).

MPs are present in the blood of healthy individuals (being produced in particular by endothelial cells, platelets and other circulating cells) but their absolute levels as well as the proportion of their different cellular origin may dramatically change under several pathological conditions. MPs appear as being released from a large variety of cell types within biological fluids, consequently shedding cell antigens that can be relevant for various biological functions or disorders that are associated to inflammation and apoptosis or cell activation. Due to the phospholipids, Tissue Factor, and other molecules present on their surface, MPs are generally believed to have noxious properties, related to their high procoagulant activity and capability of impairing endothelial activities, inhibiting Nitric Oxide production, inducing cytokine release, and activating cell proliferation and many other biological pathways (Morel O et al., 2009; Leroyer A et al., 2007).

The actual patho-physiological role of MPs is discussed in a number of reviews summarizing the evidences in the literature that are not always consistent and comparable, but there is a growing consensus in considering MPs as major elements in cardiovascular disorders (Tushuizen M et al., 201 1 ; Amabile N et al., 2010). In fact, amongst the different MPs classes that are defined according to their cell origin, MPs of endothelial origin (generally named as EMPs) have been the object of several studies on their role in endothelial dysfunctions and biology that are associated to cardiovascular disorders, including thrombosis, angiogenesis, cell survival, transfer of cell information, inflammation, vascular tone, and atherosclerosis (Leroyer A et al., 2010; Dignat-George F and Boulanger C, 201 1).

Morphological, immunological, and functional characteristics of EMPs have been studied in vitro, ex vivo and in vivo for evaluating how EMPs can be used as biomarkers of vascular risk and, more in general, of endothelium health. For instance, EMPs have been reported to reduce cell proliferation rate, to increase apoptosis rate, to have considerable impact on cytokine release, and thus they may be an important contributor to the pathogenesis of diseases that are accompanied by impaired angiogenesis (Mezentsev A et al., 2005).

The most common cell-based model for studying similar mechanisms is based on human umbilical vein endothelial cells (HUVECs) where the various effects of treatments were assessed in connection to EMPs production and effects: apoptosis, cytokine release, cell proliferation/activation, formation of reactive oxygen species, phospholipid composition, as well as prothrombotic, proinflammatory, procoagulant, and/or proadhesive effects. These treatments involve the exposure of HUVEC to different agents such as interleukin-1 alpha (Abid Hussein M et al., 2008; Abid Hussein M et al., 2003), staurosporin (Abid Hussein M et al., 2007), TNF-alpha (Szotowski B et al., 2007; Combes V et al., 1999), camptothecin (Simak J et al., 2002), daunorubicin (Fu Y et al., 2010), calpeptin (Abid Hussein M et al, 2007), fullerenes (Gelderman M et al., 2008), and simvastatin (Diamant M et al. 2008). Specific studies were also performed in specific locations such as pulmonary microvascular endothelium (Bauer N and Smith C, 2010) or bone marrow-derived endothelial progenitor cells (Benameur T et al., 2010).

In vivo, Berberine appeared to decrease the concentration of circulating EMPs with specific antigenic profile, improving at the same time endothelial functions in humans and altering HUVEC gene expression (Wang J et al., 2009). HUVEC functions themselves are altered by interaction and internalization of MPs of other cell origin or of endothelial origin (Terrisse A et al., 2010; Lester E and Babensee, 2003; Klinkner D et al., 2006; Barry O et al., 1998). Moreover, preliminary studies on cell type-specific MPs populations that are released from endothelial cells in patients suggest the possibility of implementing EMPs- based, multiple biomarkers strategy for detecting endothelial dysfunction and improve risk stratification for cardiovascular events such as coronary heart disease (Nozaki T et al., 2009), heart failure (Nozaki T et al., 2010) as well as in connection to other diseases such as Metabolic syndrome (Helal O et al., 2010), pulmonary hypertension (Amabile N et al., 2009) and diabetes (Tramontano A et al, 2010).

Qualitative and quantitative analysis of MPs in general, and of EMPs in particular, tried to elucidate if and how the composition and/or the concentration of MPs can be associated to activities of potential therapeutic or diagnostic interest using proteomic technologies, in addition to classical flow cytometry analysis.

For instance, the composition of MPs-associated proteomes have been studied in endothelial cell-derived MPs (Banfi, C et al, 2005; Peterson D et al, 2008), and atherosclerotic plaques-derived MPs (Mayr M et al., 2008). Even though it has been suggested that MPs-based proteomic studies can provide novel biomarkers, in particular for atherosclerosis and other vascular diseases (Merrick B, 2008; Leroyer A et al., 2010; Dignat-George F and Boulanger C, 201 1), MPs-based features have not been clinically validated yet as biomarkers.

An important aspect that has been so far not systematically compared and explored within MPs populations is the level and the importance of post-translationally modified (PTM) proteins on their surface. Altered levels of PTM proteins, even without global protein expression changes, are often linked to cellular responses and disease states. The complex and dynamic nature of the PTM proteome represents a major technical challenge but recent improvements have been made in the separation and the characterization of PTM proteins, leading to identification of more precisely defined post-translational modifications (Mirza S and Olivier M, 2008; Tate E, 2008; Krueger K and Srivastava S, 2006). Phosphorylation, glycosylation, ubiquitination, and prenylation are the most common and characterized categories of post-translational modifications but, by including specific variants, more than 300 modifications have been listed, and large repertoires of PTM proteins have been generated in databases such as dbPTM (Lee H et al., 2009) and HPRD (Keshava Prasad T et al., 2009).

There are isolated examples in the literature on the detection of apoptosis-/activation- related alterations and antigens of the surface of endothelial cells (Woywodt A et al., 2003; Kitazume S et al., 2010) or EMPs (Jy W et al., 2006). However, the lack of systematic analysis did not provide a more comprehensive and precise understanding on the possible relevance of EMPs in connection to PTM proteins in general, and in particular to glycosylated proteins, for defining biomarkers to be used in clinical proteomics, and more in general in clinical settings (Apweiler R et al., 2009a; Apweiler R et al., 2009b).

SUMMARY OF INVENTION

The present invention provides methods for isolating, characterizing, comparing, and using novel components of biological fluids that are characterized within MPs populations on the basis of post-translational modifications present on cell surface. In particular, Glycan-containing Endothelial Microparticles (Glyco-EMPs) populations also named Glycan-containing Endothelial cell-derived Microparticles populations have been characterized on the basis of their dimension (comprised between 100-1000 nm) and the presence of glycan epitopes (such as polysialic acid or those recognized by a specific lectin) on their surface.

The methods of the invention involve the isolation of biological fluids (in particular of human, primate, or rodent origin) and the separation of Glyco-EMPs from the acellular fraction of such biological fluids using at least a glycan-binding agent in a solid or a liquid phase. Glyco-EMPs can be further isolated, characterized, and/or compared using additional binding agents that are specific for cell surface components, such as phospholipids (and in particular phosphatidylserine) and/or cell specific antigens.

The Glyco-EMPs that are obtained by this method can be used to establish the concentration and/or other molecular components (such as cell-specific antigens, phospholipids, nucleic acids (miRNA, mRNA, DNA) or glycans) that differ between control and test subjects (e.g., normal or at risk of a disease, treated or untreated for a disease) and that can be used as biomarkers within biological fluids for diagnosing or monitoring diseases in a subject, and/or for evaluating the therapeutic efficacy of a medical treatment or a candidate drug.

Further objects of the invention, including kits and medical methods for isolating and characterizing Glyco-EMPs populations, as provided in the Detailed Description.

DESCRIPTION OF THE FIGURES

Figure 1 - Characterization of features that differentiate Glyco-EMPs (sub)populations in biological fluids: An exemplary process for the characterization of features that differentiate Glyco-EMPs (sub)populations in biological fluids (e.g. blood) involves the isolation of Control Blood Samples (e.g. from one or more healthy or untreated subjects) and of Test Blood Samples (e.g. from one or more disease-affected or treated subjects). The corresponding Glyco-EMPs populations are then isolated on the basis of their size and by means of at least a Glycan-binding agent that binds the surface of Glyco-EMPs in experimental conditions that allow maintaining their integrity. Binding agents that recognize other components of Glyco-EMPs surface (such as phospholipids, a lectin- binding epitope, and/or cell type-specific antigens) can be also used for isolating and characterizing Glyco-EMPs (sub) populations in the preferred format (in solid or liquid phase). The resulting materials are then compared by any of the known methods for analyzing materials of biological origin that can lead to the determination of features that are associated to a Glyco-EMPs subpopulation (or to an EMPs subpopulation that lacks the glycan epitope associated to a specific Glyco-EMPs subpopulation) in specific subjects, biological fluids, and/or patho-physiological conditions. Then, this signature can be used for evaluating the status of a given subject with (or without) the isolation of Glyco-EMPs populations.

Figure 2 - Quantitative analysis of Glyco-EMPs populations in human plasma: Blood samples were collected and Platelet Free Plasma fraction is obtained from two groups of subjects (Control and Coronarian) as indicated in Example 1. The concentration of Annexin V-positive MPs were compared with the concentration of MPs subpopulations that were detected and quantified by flow cytometry using the indicated combinations of labelled binding agents: Annexin V (AnnV), lectins (ConA, SNA, MAA), and an antibody directed to PECAM1 , an endothelial antigen (CD31) (also named endothelial cell antigen). Data are expressed as Mean ± SEM, indicating the statistical significance of these values when compared between the two groups of subjects.

Figure 3 - Characterization of features that allow distinguishing Glyco-EMPs (sub) populations that have specific cell origins: An exemplary process for the characterization of feature(s) that allow comparing Glyco-EMPs (sub)populations having specific cell origins involves the generation and the isolation of Glyco-EMPs from primary cells or cell lines of endothelial origin (such as HUVECs). Control and Test cell cultures (e.g. differentiated or undifferentiated, treated or untreated, healthy or in an apoptotic state), can be compared by any of the known methods for characterizing materials of biological origin, as described for Figure 1. This analysis can lead to the determination of normal (or disease-specific) features that, alone or combined, can be identified as being associated to Glyco-EMPs (sub) populations, such as cell surface antigens, that can be used for the analysis of blood and/or other biological fluids in different categories of subjects (e.g. being treated or not treated, affected or no affected by a disorder) for determining if any of them can be defined as an endothelium-associated biomarker for a normal, drug-related, metabolism-related, or a disorder-specific status.

Figure 4 - Quantitative analysis of Glyco-EMPs of Control Glyco-EMPs populations using different glycan binding agents: The analysis has been then performed by flow cytometry using HUVEC cultures that are maintained in standard culture conditions and that shed a basal level of MPs in the cell culture medium (at a concentration of approx. 7500 Annexin V-positive MPs / μΕ), by comparing the data that have been generated by flow cytometry using a series of fluorescently labelled binding agents either at the level of cell surface (as percentages of all cells) or at the level of the concentration of MPs populations (as Annexin V-positive only MPs population and as specific Glyco-EMPs subpopulations) in the cell culture supernatant.

DETAILED DESCRIPTION OF THE INVENTION

The present invention is based on unexpected findings of the presence and distribution of a specific post-translational modification on the surface of MPs populations of endothelial origin. These evidences suggest that MPs populations of endothelial origin that are isolated within biological fluids on the basis of at least a post-translational modification (such as phosphorylation, glycosylation, ubiquitination, and prenylation) can be used for obtaining information for the medical management of disorders, in particular for defining and comparing biomarkers present in biological fluids.

In particular, the glycan epitopes that are present on the surface of MPs populations of endothelial origin can be used for isolating and characterizing a novel and useful MPs population named as Glyco-EMPs. Glyco-EMPs populations can provide information about the state of endothelium in a subject through the analysis of biological fluids wherein MPs populations of endothelial origin are released, or wherein they are subsequently localized.

Changes in the regulation and processing of glycans have been correlated to a number of abnormal physiologic conditions. However, no evidences have been provided so far on the possibility to integrate the information on glycans (and in general on a specific post-translational modification) on the analysis of MPs populations of endothelial origin in order to provide novel and useful insight of their biological features (such as concentration, composition, and/or surface antigens).

As shown in the Examples, Glyco-EMPs populations do not simply represent MPs subpopulations having a specific cell origin and a specific type of glycosylation. In fact, Glyco-EMPs populations are formed by distinct MPs sub-populations having distinct profiles based on the glycan epitopes they present, independently (or not) from known endothelium-specific antigens. Then, phosphatidylserine-based MPs detection technologies can be usefully combined in methods that allow isolating Glyco-EMPs subpopulations potentially representing a more informative MPs subpopulation of endothelial origin for proteomic, immunologic, transcriptomic and/or glycomic analysis. In particular, the isolation, the comparison, and the use of Glyco-EMPs populations is intended to provide means for a more sensitive detection and characterization of biological features that can be used as biomarkers for an endofhelium-related-disorder or endothelial dysfunction.

In a first aspect, the present invention provides a method for identifying Glycan- containing Microparticles of endothelial origin (Glyco-EMPs) populations in a biological fluid comprising the following steps:

a) Obtaining a sample of a biological fluid from a subject;

b) Isolating the acellular fraction of said biological fluid;

c) Separating the Glyco-EMPs from the said acellular fraction by means of at least a binding agent preferably glycan-binding agent, more preferably a lectin.

Preferably said binding agent is selected from the group consisting of an antibody, a protein, a peptide, a lectin, a glycan, a nucleic acid, a lipid, a phospholipid and an inorganic compound.

Step (c) of the method involves separating the Glyco-EMPs populations from the said acellular fraction by means of an agent that bind a specific epitope such as an N-linked Glycan, an O-linked Glycan, an high-mannose containing, a fucose containing, or a PolySia-related epitope. Step (c) of the method may also involve the use of at least one further binding agent recognizing a phospholipid, a protein, a lipid, or a glycan other than the selected glycan epitope, preferably a CD antigen from endothelial origin. In a preferred embodiment of the invention, said in vitro or ex vivo method for identifying Glycan-containing Endothelial Microparticles (Glyco-EMPs) populations comprises the following steps:

a) Isolating the acellular fraction from a sample of a biological fluid from a subject; and

b) Separating the Glyco-EMPs populations from said acellular fraction by means of an agent binding a specific glycan epitope and at least one further binding agent recognizing a CD antigen from endothelial origin.

Prior to step a) the sample of said biological fluid is obtained by any means.

According to said preferred embodiment, in step b) the Glyco-EMPs populations are separated from said acellular fraction by means of a further binding agent that recognizes a phospholipid, preferably, phosphatidylserine, phosphatidylethanolamine and phosphatidylinositol; more preferably phosphatidylserine.

Preferably, the Glyco-EMPs populations having a size comprised between 100 and 1000 nanometres are separated from said acellular fraction.

Preferably, said CD antigen from endothelial cell origin is selected from the group consisting of CD31 , CD34, CD51, CD61, CD62E, CD105, CD106, CD141, CD142, CD144 and CD 146, more preferably selected from the group consisting of CD62E, CD31 and CD144.

Preferably said agent binding a specific glycan epitope is selected from the group consisting of antibodies, lectins, enzymes containing carbohydrate recognition domain, cytokines, chaperone and transport proteins, microbial carbohydrate-binding proteins, glycosaminoglycan-binding proteins, preferably a lectin.

Preferably, said agent binding a specific glycan epitope is a lectin selected from the group consisting of GNA (Snowdrop lectin), PNA (Peanut agglutinin), VVL (Hairy vetch lectin), WGA (Wheat Germ agglutinin), SNA (Sambuscus nigra lectin), MAL/MAA (Maackia amurensis leukoagglutinin), MAH (Maackia amurensis hemoagglutinin), LFA (Limax flavus agglutinin), BS-1 (Bandeiraea Simplicifolia-1), Hippeastrum hybrid (HHA), UEA-I (Ulex Europeaus Agglutinin-I), UEA-II (Ulex Europeaus Agglutinin-II) and ConA (Concanavalin A), more preferably selected from the group consisting of GNA, MAA and UEA-I.

Preferably, the selected glycan epitope is a Sia-related epitope, a PolySia-related epitope, a fucosylated group or a high mannose group.

Preferably said binding agent is immobilized on a solid phase and/or is labelled.

Preferably, the biological fluid is selected from the group consisting of plasma, blood, cerebrospinal fluid, proximal fluid and cell culture supernatants. Thus, preferably, said in vitro or ex vivo method for identifying Glycan-containing Endothelial cell derived Microparticles (Glyco-EMPs) populations comprises, in step c), the use of three binding agents: (i) an agent binding a specific glycan epitope, (ii) a binding agent recognizing a CD antigen from endothelial origin and (iii) a binding agent recognizing a phospholipid, preferably phosphatidylserine.

Preferably, step a) of isolation of the acellular fraction is performed by any method able to separate the cells from said biological fluid; preferably said isolation is performed by centrifugation and/or flow cytometry and/or affinity-based chromatography and/or sorting and/or microfiltration.

The expression "Glyco-EMPs" refers to phospholipid-containing vesicles of endothelial origin that have a submicron dimension, in particular comprised between 100 and 1000 nanometers, and that are separated from a biological fluid by means of at least a glycan-binding agent. Preferably, Glyco-EMPs expose phospholipids, in particular anionic phospholipids (such as phosphatidylserine) as well as glycans, on their surface.

The origin of the Glyco-EMPs can be established on the basis of the presence of molecules that are known to be associated to the endothelium, and in particular those localized on the surface of the endothelial cells that is exposed to blood circulation, or that is at the interface between the endothelial cells and other tissues (e. g. in the blood vessels, in the pulmonary or pancreatic microvascular endothelium, or in the blood-brain barrier). Glyco-EMPs populations can be also isolated using biological fluids obtained in cell culture conditions from laboratory cell lines and primary cells of endothelial origin that are isolated from biopsies or biological fluids.

In general any criteria for validating (identifying) a cell as being of endothelial origin, including endothelial progenitor cells (EPCs) and circulating endothelial cells (CECs) that are published in the literature (Goon P et al., 2006; Benameur T et al., 2010) may be considered applicable in the separation and/or detection of Glyco-EMPs. Such criteria include the presence/absence of specific (combinations of) CD antigens, as detected by means of CD-specific antibodies (as described in the Examples). A non- limiting list of such CD antigens includes CD31, CD34, CD51 , CD61, CD62E, CD105, CD106, CD141 , CD142, CD144, or CD146, preferably, CD62E, CD31 and CD144. Otherwise, the interaction/internalization of synthetic or natural compounds known to be specific for endothelial cells, such as Dil-Acetyl-LDL (l '-dioctadecyl-3,3,3',3'-tetramethyl- indocarbocyanine-labeled acetylated-LDL) can be measured (Benameur T et al., 2010).

Flow cytometry is a preferred technology for isolating and/or quantifying Glyco- EMPs populations (and MPs in general for in vitro/ex vivo studies) according to their size and it can be standardized using size-calibrated fluorescent beads (Robert S et al., 2009; Rukoyatkina N et al., 2009). Flow cytometry analysis of Glyco-EMPs populations can be performed by adapting technologies that have been developed for the measurement, the preparative sorting into distinct size fractions, and the image processing of artificial nanoparticles and liposomes having similar size (van Gaal E et al., 2010; Kunding A et al., 2008). Combinations of antibodies, labelling agents, and other technical elements for an improved detection of EMPs have been recently summarized (Orozco A and Lewis D, 2010; van der Heyde H et al., 201 1 ; Amabile N et al., 2010).

Glyco-EMPs populations can be also isolated and characterized using other methods previously reported for MPs populations, including atomic force microscopy (Yuana Y et al. 2010), dynamic light scattering (Lawrie A et al. 2009), ELISA (Abid Hussein M et al., 2008; Nomura S et al., 2009), functionalized beads (Shah M et al., 2007), or antibody arrays (Lai S et al., 2009).

The terms "separating" or "separation" refer to both the physical separation and isolation of Glyco-EMPs populations (and of MPs populations in general) from a biological fluid (e. g. by microfiltration or centrifugation), and the separation of Glyco- EMPs populations (and of MPs populations in general) that can be performed by technologies, such flow cytometry or microscopy, which provide means for detecting images and other quantifiable signals characterizing MPs within a sample. The terms "separating" or "separation" also imply a quantification of the isolated cells, especially when flow cytometry is carried out.

The expression "biological fluids" refers to any bodily fluid (and fraction thereof) from, excreted by or secreted by any living cell or organism, where MPs of endothelial origin can circulate and accumulate including but not limited to blood, cerebrospinal fluid, and proximal fluid (the fluid derived from the extracellular milieu of tissues). Databases and other repertoires of the proteomes from biological fluids, and in particular blood (Li S et al., 2009; Hu S et al., 2006) can be used to compare with results obtained by studying Glyco-EMPs populations. This expression preferably applies also to plasma, cell culture supernatant and any other fluid obtained from a preliminary fractionation, depletion, or any other purification of such biological fluids.

The term "sample" encompasses both an initial aliquot of the biological fluid as well as the product of any manipulation of the initial source of proteins, including but not limited to partial purification, fractionation, enrichment, enzyme digestion, or other treatment.

The expression "acellular fraction" refers to a fraction of a biological fluid in which cells are absent, for example following a centrifugation, a separation by flow cytometry, or an affinity-based chromatography or sorting. The methods of the invention involve the isolation of biological fluids from humans, primates, rodents, or any other animal presenting an interest for medical or veterinary research. The biological fluids can be obtained by puncture, involving the removal of a volume of at least 0.01 ml (e.g. in smaller animal) or at least 1 ml (e.g. in human or primates). It is of major importance to ascertain that the biological fluid is not mixed with cells that results from the rupture of tissues that contain the biological fluid (such as arterial or venous walls in the case of blood) during the puncture, thus excluding any contamination from undesired tissues and cell types.

The isolation of the acellular fraction of biological fluids can be performed by eliminating any cellular elements having a size superior to 1000 nanometers, as it is made possible by flow cytometry, microfiltration, or centrifugation. Consistently with the literature (Piccin A et al., 2007), the separation of the Glyco-EMPs populations from the acellular fraction of a biological fluid can be performed by applying technologies for isolating cell vesicles having a diameter comprised between 100 and 1000 nm, as well as a composition, that is typical of Glyco-EMPs populations. These two criteria can be applied in any order that is appropriate for eliminating undesired biological entities that may be present in the biological fluid (such as cell debris, soluble proteins, exosomes, cells, or virus). The centrifugation of the biological fluids can be performed at a speed comprised between l ,500g and 15,000g, at a temperature comprised between 15°C and 37°C, and for a time comprised between 1 minute and 60 minutes and should allow the separation of a fraction containing the Glyco-EMPs populations (the supernatants) from the cells (forming the pellet). Optionally, a further purification step can be performed to isolate Glyco-EMPs (but not exosomes) by centrifuging the supernatant obtained above at a speed comprised between 15,000g and 30,000g, at a temperature comprised between 15°C and 37°C, and for a time comprised between 1 and 60 minutes, in order to obtain a pellet formed by Glyco-EMPs populations.

The expression "binding agent" refers to any material that can bind to the desired molecule (that is, a component of Glyco-EMPs such as a protein, a protein variant, a phospholipid, a glycan, or a lipid) and consequently allow detecting, labelling, separating and/or quantifying the structures containing such molecule (i.e. Glyco-EMPs) in a sample (i.e. the acellular fraction of a biological fluid), preferably by interacting with components on the surface of Glyco-EMPs. The binding agent for the desired molecule can be a natural or recombinant protein (such as an antibody or a protein that binds a cell surface antigen), a peptide, a lectin, a glycan, a nucleic acid such as an aptamer, a lipid, a phospholipid, an inorganic compound, a nanomaterial, or a low molecular weight ligand.

The binding agent for the desired molecule can be labelled. There are numerous methods by which the label can produce a signal detectable by external means, for example, desirably by visual examination or by electromagnetic radiation, heat, and chemical reagents. The label or other signal producing system component can also be bound to a specific binding partner, another molecule or to a support such as beads, using any method known in the art, such as chemically cross-linking or using the biotin- streptavidin system. The label can directly produce a signal, and therefore, additional components are not required to produce a signal. Numerous organic molecules, for example fluorescers (such as FITC, PE, PC5, PC7, APC, or any other known to be compatible with flow cytometry-based MPs detection), absorb ultraviolet and visible light. Their use, alone or in appropriate combinations for the simultaneous detection of different antigens on MPs by flow cytometry, is well described in the literature (Orozco A and Lewis D, 2010; van der Heyde H et al., 201 1 ; Amabile N et al., 2010).

Other types of label directly produce a signal, such as radioactive isotopes and dyes. Alternatively, the label may need other components to produce a signal, and the signal producing system would then include all the components required to produce a measurable signal, which may include substrates, coenzymes, metal ions, or substances that react with enzymatic products.

The binding agent for the desired molecule (that is, a component of Glyco-EMPs surface) can be provided in a liquid phase or in a solid phase (for example, by the immobilization on a bead or a plate from which it can be or not separated) forming thus a complex with the Glyco-EMPs once that the acellular fraction of a biological fluid is contacted with such agent. Subsequently, depending on the further uses, such complex can be dissociated (for instance, by temperature or chemical-induced denaturation) or the binding agent for the desired molecule can be kept associated.

The term "Glycan" refers to chemical groups also named as sugars or carbohydrates. Glycans can be monomers or polymers of sugar residues, but typically contain at least three sugars, and can be linear or branched. A glycan may include natural sugar residues (e.g., glucose, N- acetylglucosamine, N-acetyl neuraminic acid, galactose, mannose, fucose, hexose, arabinose, ribose, xylose, sialic acids (Sia for the monomer and PolySia for the polymers)) and/or modified sugars (e.g., 2'-fluororibose, 2'-deoxyribose, phosphomannose) at low or high density.

Glycans can be distinguished in N-linked, O-linked, or C-linked glycans on the basis of the linkage to a glycoconjugate via nitrogen, oxygen, or carbon linkage. The glycoconjugate can be also in the form of an advanced glycation end products, resulting from the non-enzymatic modification of glycans. The relevance of protein glycation has been demonstrated in several pathological conditions, including chronic complications associated to diabetes mellitus and renal failure as well as degenerative changes, and more sensitive and selective methods are now available for identification and quantification of such glycated proteins (Priego Capote F and Sanchez J, 2009; Thornalley P, 2005).

In the present disclosure, the term "glycan" refers in general to the carbohydrate portion of a glycoconjugate, includes, but is not limited to, glycoproteins, glycolipids, proteoglycans and glycophosphosphingolipids or any other known glycoconjugate that present a glycan epitope. Glycoconjugates are found predominantly on the external surface of the plasma membrane and in secreted fluids usually consist of O- or N-glycosidic linkages of oligosaccharides (a polymer containing a small number, typically three to ten saccharides) to compatible amino acid side chains in proteins or to lipid moieties. Some glycans also have modifications such as acetylation and sulfation. Glycoconjugates contain cell-surface glycans that have been shown to be important in cell-cell interactions due to the presence on the cell surface of various glycan binding receptors, in addition to the glycoconjugates themselves.

The expression "cell surface glycan" refers to a glycan that is present on the exterior surface of a cell that is, in general, covalently linked to a polypeptide (as part of a cell- surface glycoprotein) or a lipid (as part of a cell membrane glycolipid) and that can be exposed as well on the surface of Glyco-EMPs populations. A cell surface glycan is formed by homo- and/or heteropolymers of sugar residues which form specific glycan epitopes (also called glycoepitopes) that are specifically recognized by glycan-binding agent.

The expression "glycan-binding agent" refers to binding agents that specifically bind a glycoepitope. Recognition systems of glycoepitopes include, but is not limited to, antibodies, lectins (of animal, plant, or pathogen origin), enzymes containing carbohydrate recognition domain (CRD), antibodies against glycans, cytokines, chaperone and transport proteins, microbial carbohydrate-binding proteins, glycosaminoglycan-binding proteins, or any other known recognition system for glycan epitopes, without limitation.

Several properties of the cell surface glycans present on Glyco-EMPs surface can be determined, with or without using glycan-binding agents, including the mass of part or all of the saccharide structure, the charges of the chemical units of the saccharide, identities of the chemical units of the saccharide, total charge of the saccharide, or total number of sulfates or acetates. The methods that are applicable for identifying such properties include, but are not limited to, capillary electrophoresis, NMR, mass spectrometry (both MALDI and ESI), and HPLC with fluorescence detection. Glycans can be detected and/or analyzed on Glyco-EMPs, either in the absence or in presence of proteases or glycosidase. In particular, the structural and functional diversity within glycan epitopes that can be used for isolating and characterized within Glyco-EMPs populations may be explored using technologies for comparative glycan analysis (Krishnamoorthy L and Mahal L, 2009), lectin- or glycan-based arrays (Taylor M and Drickamer K, 2009; Gupta G et al., 2010), means for interfering protein-glycan interactions (Rek A et al., 2009), bioinformatics methods (Mahal L, 2008), and chemical tools for binding and/or modifying glycans. Extensive lists of glycans and of the corresponding glycan-binding agents (such as monoclonal antibodies, lectins of plant or animal origin, and chemicals) and related technologies have been published, indicating the specificity of the interaction, the compatibility with the integrity and viability of cells, and their potential use for identifying specific cell types and/or antigens of medical interest, (Varki A et al., 2009).

A general overview of glycan epitopes of potential interest and of the corresponding glycan binding agents is provided in Table 1.

Among the glycan-binding agents, lectins are of particular interest given their specificity for different sugar moieties that are either onto a soluble carbohydrate or onto a carbohydrate moiety that is a part of a glycoprotein or glycolipid (Gemeiner P et al., 2009). Consequently, distinct glycoprofiles can be generated and compared according to the choice and/or the combination of lectins according to their glycan epitope. A list of lectin that can be used in the methods of the invention includes: GNA (Snowdrop lectin), PNA (Peanut agglutinin), VVL (Hairy vetch lectin), WGA (Wheat Germ agglutinin), SNA (Sambuscus nigra lectin), MAL/MAA (Maackia amurensis leukoagglutinin), MAH (Maackia amurensis hemoagglutinin), LFA (Limax flavus agglutinin), BS-1 (Bandeiraea Simplicifolia-1), Hippeastrum hybrid (HHA), UEA-I/-II (Ulex Europeaus Agglutinin- I/-II), and ConA (Concanavalin A), preferably GNA, MAA and UEA-I.

The examples show that a specific glycan known as sialic acid, alone (Sia) or in polymerized forms (PolySia) of this glycan known as polysialic acid, as well as any glycan recognized on the basis of the presence of a sialic acid (hereafter collectively defined as "PolySia-related epitopes") can be preferably used in methods for isolating Glyco-EMPs populations of interest. Alternatively, lectins that recognize fucosylated groups, such as UEA-I (Wu et al., 2009) can be used. The expression "sialic acid," as used herein, is a generic term for the N- or O- substituted derivatives of neuraminic acid, a nine-carbon monosaccharide (Varki A, 2008). The amino group of neuraminic acid typically bears either an acetyl or a glycolyl group in a sialic acid. It is also the name for the most common member of this group, N- acetylneuraminic acid (Neu5Ac or NANA) and 2-Keto-3-deoxynononic acid (Kdn). Other members of sialic acid include, but are not limited to, N-Acetylglucosamine, N- Acetylgalactosamine (GalNAc), N-Acetylmannosamine (ManNAc), and N-Glycolylneur- aminic acid (Neu5Gc).

Sialic acids are found widely distributed in animal tissues and in bacteria, especially in glycoproteins and gangliosides. The amino group bears either an acetyl or a glycolyl group. The hydroxyl substituents may vary considerably: acetyl, lactyl, methyl, sulfate and phosphate groups have been found. Hydroxyl substituents present on the sialic acid may be modified by acetylation, methylation, sulfation, and phosphorylation. The predominant sialic acid is N-acetylneuraminic acid (Neu5Ac). Sialic acids impart a negative charge to glycans, because the carboxyl group tends to dissociate a proton at physiological pH.

The Examples below show that evidences generated by studying the glycan epitopes present on the surface of cell lines in presence of apoptotic agents cannot be used for interpreting or anticipating the presence of specific glycan epitopes on the surface of Glyco-EMPs populations, which rather appear as biological entities providing distinctive (and possibly more relevant) biological information. When compared on literature referring to EMPs antigens that may be glycosylated (Abid Hussein M et al., 2003; Jy W et al., 2006), these findings show the relevance of performing a more detailed and comparative analysis of glycan epitopes on EMPs in order to assess their pathophysiological relevance.

Antibodies to variants of PolySia-related epitopes (in terms of chemical derivativization, linkage and/or degree of polymerization) have been generated and compared in terms of binding specificity, epitope presentation, and degree of polymerization (Hayrinen J et al., 1995; Hayrinen J et al., 2002; Sato C et al., 1995; Sato C et al., 2000), establishing the specificity of these antibodies for PolySia related epitopes having defined lengths and degrees of polymerization ranging from 2 up to 200 sialyl residues.

A series of human and rodent proteins have been showed as presenting one or more PolySia-related epitopes and/or degree of polymerization of the PolySia-related epitopes, the main one being PSA-NCAM has been used for identifying, studying the activities, and/or sorting specific cell types such as neuronal cells (Bonfanti L, 2006) and pancreatic beta-cell subpopulations (Banerjee M and Otonkoski T, 2009). Other proteins presenting PolySia-related epitopes have been identified such as betal integrin (Bartik P et al., 2008), DPP-4 (Cuchacovich M et al. 2001), sodium channels (Zuber C et al, 1992), CD31 (Kitazume S et al, 2010), and CD36 (Yabe U et al., 2003).

Moreover, panels of peptides and proteins that are linked with PolySia-related epitopes were identified and quantified in blood using (glyco)proteomic technologies, providing additional candidates of proteins that can be associated to this type of glycosylation (Alley W and Novotny M, 2010; Kurogochi M et al.,2010; Ghesquiere B et al.,2007).

A list of preferred PolySia-related epitope and related PolySia-binding agents that are known in the literature and then can be used in methods for identifying Glyco-EMPs populations is provided in Table 2.

The interaction between Glyco-EMPs populations and glycan-binding agents can be studied by means of competing soluble glycans such as those generated using chemical technologies and characterized by mass spectrometry (Galuska S et al., 2007; Patane J et al., 2009). Moreover, technologies for the enrichment of peptides that are linked to glycans can be used for the analysis of the structure and attachment site identification of glycans that are present on the Glyco-EMPs populations (Nilsson J et al., 2009). Series of glycan- binding agents may be used for detecting and isolating different types of Glyco-EMPs subpopulations in parallel. However, glycan-binding agents that are compatible with technologies that maintain MPs integrity (such as flow cytometry, immunological assays, ELISA, or magnetic beads-based isolation) are preferred. For example, a panel of antibodies and binding proteins that have been characterized in the literature for different PolySia-related epitopes into biological materials can be used for detecting, isolating and comparing the corresponding Glyco-EMPs (sub)populations presenting such PolySia- related epitopes.

This analysis can be accompanied also by the detection of known cell type-specific antigens (using antibodies or other molecules that specifically bind such antigens) that can be more or less frequently associated to total MPs populations or Glyco-EMPs (sub) populations within biological fluids and/or in cell culture conditions. List of antigens that can be used to further distinguish Glyco-EMPs populations according to their cell origin has been published (Burnier L et al., 2009; Orozco A and Lewis D, 2010; van der Heyde H et al., 201 1).

The study of Glyco-EMPs populations that expose glycan epitopes (such as Polysia- related epitopes) can be also performed by using experimental in vitro and in vivo approaches for either inhibiting glycosylation or for integrating unnatural precursors for labelling the glycan epitopes (Bork E et al., 2007). Preferably, the analysis of Glyco-EMPs populations comprises detecting alterations in one or more features of sialylation, including the type of linkage, the degree of polymerization, modifications of sialic acids (including sulfation, branching, presence or absence of a bisecting N-acetylglucosamine), and changes in the number of polysialylated proteins, lipids and/or molecule-specific sites on Glyco-EMPs. These studies, that can be extended to sites that contain mannose or fucose at high or low levels, can be performed in parallel with the determination of other features of Glyco-EMPs populations such as the presence of cell type-specific antigens or of biological activities that are established in vitro using cell line-based assays (as for procoagulant activity).

The methods of the invention provide Glyco-EMPs populations that are separated from the acellular fraction by means of at least one further binding agent recognizing a phospholipid, a protein, a lipid, or a glycan other than the one used for isolating the Glyco- EMPs population. This additional binding agent preferably binds a phospholipid which is phosphatidylserine. In particular, said phosphatidylserine-binding agent is Annexin V or Lactadherin (Shi J and Gilbert G, 2003; Logue et al. 2009) or other proteins, peptides, or chemicals (Thapa N et al., 2008; Lemmon M, 2008; Stace C and Ktistakis N et al., 2006). Phospholipid-binding agents, and in particular phosphatidylserine-binding agents, are preferred, with preference for those not requiring calcium for binding phosphatidylserine and not altering MPs integrity. MPs population have been often isolated and characterized from biological fluids using phosphatidylserine-binding agents that are in a solid or a liquid phase.

Alternatively (or in addition) the methods may involve the use of one or more additional binding agents that bind a protein, lipid, or glycan of the cell surface that is transferred to the surface of Glyco-EMPs populations following their release from the cells, thus identifying their origin. These methods can be performed using a binding agent that bind a protein, lipid, or glycan of the cell surface is defined according to a specific cell type, tissue, organ, drug treatment, age, sex, pathology, genotype, phenotype, predisposition, viral infection, and/or clinical status. Additional binding agents to be used according to the invention can be defined as agents that bind a cell type-specific antigen (e.g. an antibody).

The separation and/or the detection of Glyco-EMPs populations with an additional binding agent can be performed prior to, simultaneously, or following the separation step that involve the glycan-binding agent and may be used as well for separating MPs populations. This step may be performed in liquid or solid phase, and in the latter case the solid phase can be in the forms of beads, and in particular magnetic beads, a support that has been already used for immobilizing Annexin V and sorting apoptotic cells from a biological fluid (Said T et al., 2008). These additional binding agents can be used not only as a mean for positively selecting the Glyco-EMPs population but as a negative selection tool (e.g. for eliminating specific MPs populations presenting a specific cell surface antigen or cell debris and other undesired entities within the sample of biological fluid). Equally, specific Glyco- EMPs subpopulation may be negatively selected in order to define an MPs subpopulation having a potential medical or biological interest. For instance, a MPs subpopulation that is deprived of specific glycan epitopes may be enriched in other antigens that may be of interest for defining and validating a biomarker in a population.

A further aspect of the present invention is an isolated Glyco-EMP or Glyco-EMPs population that is obtained according to the methods defined above. The Glyco-EMPs populations can be provided in a liquid or a solid phase, and in association or not with the binding agent (e.g. a lectin or a phosphatidylserine-binding agent). These novel biological entities can provide novel biomarkers that are associated to Glyco-EMPs in general or to specific Glyco-EMPs subpopulations of interest that are defined in connection to specific cell types and/or disorders. For instance, such biomarker can allow screening subjects at risk of being affected by a disorder, since it can be identified by using common technologies such as flow cytometry, mass spectrometry, gel electrophoresis, an immunoassay (e.g. immunoblot, immunoprecipitation, ELISA), nucleic acid amplification, procoagulant activity, and/or electron microscopy on biological fluids or Glyco-EMPs populations obtained from such subjects in a singleplex or multiplex formats as summarized in Figures 1 and 3. For instance, the examples show that a significant fraction of a Annexin V-positive MPs population can be isolated and characterized as a Glyco-EMPs population by double staining in flow cytometry.

The Glyco-EMPs population provide means for defining novel biomarkers. In particular, the biomarkers can be defined by means of the concentration of Glyco-EMPs population only or the concentration of Glyco-EMPs and of the concentration of at least another population of MPs that present a protein, lipid, phospholipid, and/or glycan of the cell surface. Such biomarkers can correspond to a peptide, a protein, a phospholipid, a lipid, a nucleic acid, a glycan, or any combinations of such Glyco-EMPs components. The biomarker can be specific for a disorder and may be identified by means of one or more technologies such as flow cytometry, mass spectrometry, gel electrophoresis, immunoassay, nucleic acid amplification, or in vitro assays for a biological activity.

The term "biomarker" refers to a molecule, a parameter, a characteristic, or an entity that is objectively measured and evaluated as an indicator of a specific state of an organism, in particular in association to a normal or pathogenic process, or the response to a medical treatment.

In the present case, this factor can be defined by the concentration and/or the composition of Glyco-EMPs that are isolated from biological fluids of humans or animals (rodents or primates, in particular). Accordingly, the biomarkers can be defined by means of:

a) The composition and/or the concentration of a Glyco-EMPs presenting a specific glycan epitope (for example, a PolySia-related epitope);

b) The ratio between the composition and/or the concentration of total MPs population and of a Glyco-EMPs population of (a);

c) The ratio between the composition and/or the concentration of a Glyco-EMPs population of (a) and a different Glyco-EMPs population;

d) The composition and/or the concentration of MPs not presenting the specific glycan epitope of (a); and/or

e) The composition and/or the concentration of MPs populations of (a) and/or (d) that present a protein, lipid, phospholipid, and/or glycan other than the specific glycan epitope of (a),

in an appropriate biological fluid sample obtained from a subject.

The biomarker can be found associated to the whole Glyco-EMPs population and/or to specific Glyco-EMPs subpopulations defined by any molecular parameter of interest (for example, the presence of a cell-type specific antigen). The quantitative evaluation of Glyco-EMPs in specific volume of a biological fluid can be, or not, associated to a quantitative evaluation of total MPs in such volume. The concentration of Glyco-EMPs (sub)populations that present (in particular on the surface) a cell component (e. g. a protein, a protein variant, a phospholipid, a nucleic acid, a glycan, a glycoconjugate) or any other organic or inorganic elements may be used as biomarker. Such component that is found associated to a Glyco-EMPs population can be used as a biomarker that allows establishing a specific status of the cells originating the Glyco MPs (sub) population and/or the possible interactions of Glyco-EMPs (sub)populations with the surface of the specific cell types or of a virus, with a drug, an antibody, and any other compound present in the biological fluid.

An association between a biomarker (such as a Glyco-EMPs population or a Glyco-

EMPs component) and a disorder can be established independently from the cause of the disorder but only from its effects and other associated biological evidences.

The Glyco-EMPs populations can allow the identification of biomarkers for characterizing the state of a subject (such as normal, affected or at a risk of disorder, responding or not to a therapy) by using samples of one or more biological fluid obtained from such subject. Once that biomarkers are found associated to Glyco-EMPs, such biomarkers can be identified in the subjects of interest (e.g. animal models, patients, at risk individuals) for obtaining information of medical interest on a subject, throughout the time (e.g. before, during, and/or after a medical intervention or treatment) and/or in comparison to reference populations (e.g. control, healthy subjects or subjects affected by a disorder). Such biomarkers may be or detectable even without using Glyco-EMPs populations but, given the complexity of biological fluids, Glyco-EMPs population may provide a more precise and reliable analysis of biomarkers otherwise undetectable. Alternatively, the subtraction of specific Glyco-EMPs subpopulation may provide a MPs population that is particularly enriched (or deprived of) specific antigens in defined test conditions, thus representing alternative means to identify biomarkers.

The present invention also relates to kits for isolating and/or using Glyco-EMPs for medical or veterinary applications. The kits for isolating Glyco-EMPs populations comprise at least (i) a glycan-binding agent (for example, specific for a PolySia-related epitope as listed in Table 2), preferably a lectin as defined above and (ii) at least one further binding agent recognizing a phospholipid (for instance, a phosphatidylserine- binding agent), a protein, a lipid, or a glycan other than the one used for isolating the Glyco-EMPs population, preferably recognizing a CD antigen from endothelial origin as defined above. In this latter case, said kit may optionally includes a binding agent recognizing a phospholipid as defined above. The binding agents for the desired molecules can be provided in a liquid or a solid phase, with or without means for detecting and comparing effectively the interaction with Glyco-EMPs (and consequently for quantifying the Glyco-EMPs (sub)population of interest) by using one or more proteomic, immunological, biochemical, chemical, biological, or nucleic acid detection method.

More preferably said kit comprises at least three binding agents, namely: (i) a glycan-binding agent (for example, specific for a PolySia-related epitope as listed in Table 2), preferably a lectin as defined above, (ii) a binding agent recognizing a CD antigen from endothelial origin as defined above and (iii) a binding agent recognizing a phospholipid (for instance, a phosphatidylserine-binding agent and/or a phosphatidylethanolamine- binding agent and/or a phosphatidylinositol-binding agent).

The present invention also provides the use of a Glyco-EMPs population or of a kit as defined above for identifying biomarkers of medical interest in a sample of biological fluid. The Glyco-EMPs populations of the invention can be isolated, compared, and used according to desired medical application. Examples of the process for analyzing and comparing Glyco-EMPs and identifying biomarkers of medical interest are summarized in Figures 1 and 3, but many other possibilities can be envisaged in connection to specific medical goals, features of the biomarker, and/or the type of populations to be evaluated, as shown in the Examples.

In particular, the biological fluids into which Glyco-EMPs features are studied, can be obtained from distinct groups of subjects that are appropriately selected (e. g. on the basis of drug treatment, age, sex, pathologies, genotype, phenotype, exposure to risk factors, viral infection, or clinical status) and then compared at the level of Glyco-EMPs (sub)populations using biomarkers that can be evaluated by means of one or more proteomic, immunological, biochemical, chemical, biological, or nucleic acid detection method. This comparison may also involve the use of appropriate statistical and/or imaging methods (including MRI, CAT, and ultrasound, immunodiagnostic test, detection of protein levels, or biopsy), should allow confirming the identification of a biomarker associated to Glyco-EMPs that can be further used in diagnostic and drug discovery/validation methods for a disorder, as well as of any other disorder that may alter the structure and/or the activity of an organ, a tissue, or a cell type.

The present invention also provides the use of a Glyco-EMPs population or of a kit as defined above for identifying biomarkers of an endothelial dysfunction.

More preferably said use is characterized in that said endothelial dysfunction is selected from the group consisting of septic shock, hypertension, metabolic diseases (such as hypercholesterolaemia, non-alcoholic fatty liver disease, non-alcoholic steatohepatitis and diabetes), atherosclerosis, cardiovascular diseases (such as coronary artery diseases, abdominal or thoracic aortic aneurysm), pulmonary artery diseases, antiphospholipid syndrome (APS) and CNS disorders. In a further aspect, the present invention provides in vitro or ex vivo methods for prognosing, diagnosing or monitoring a disorder, preferably an endothelial dysfunction that comprises the identification of a biomarker that have been characterized using Glyco- EMPs. Such medical methods involve the isolation, the characterization, and the comparison of Glyco-EMPs populations in a sample of biological fluid. Glyco-EMPs quantitative and/or qualitative features might be of considerable value for diagnosing and monitoring of human disorders, as well as for evaluating drug candidates and drug treatments for any disease, and in particular for establishing their effects on biological fluids.

In a preferred embodiment of said aspect of the present invention, the in vitro or ex vivo method for prognosing, diagnosing or monitoring an endothelial dysfunction in a subject, comprises:

a) isolating Glyco-EMPs populations in a test sample of biological fluid of said subject by means of an agent binding a specific glycan epitope and at least one further binding agent recognizing a phospholipid, a protein, a lipid or a glycan other that said selected glycan epitope, preferably said at least further binding agent recognizes a CD antigen from endothelial origin;

b) characterizing said test sample Glyco-EMPs populations, for instance by determining the concentration (qualitatively or quantitatively) and/or the composition (for instance in terms of glycan epitopes and CD antigens) of said Glyco-EMPs populations and

c) comparing said characterized test sample Glyco-EMPs populations with a control sample of Glyco-EMPs populations, a difference in said characterized Glyco-EMPs populations in said test sample and said control sample constituting a biomarker of the presence or the evolution of said endothelial dysfunction.

Preferably, said control sample consists of a physiological sample or a test sample of Glyco-EMPs populations previously determined in the same said subject.

Preferably, said agent binding a specific glycan epitope is a lectin; preferably a lectin selected from the group consisting of GNA (Snowdrop lectin), PNA (Peanut agglutinin), VVL (Hairy vetch lectin), WGA (Wheat Germ agglutinin), SNA (Sambuscus nigra lectin), MAL/MAA (Maackia amurensis leukoagglutinin), MAH (Maackia amurensis hemoagglutinin), LFA (Limax flavus agglutinin), BS-1 (Bandeiraea Simplicifolia-1), Hippeastrum hybrid (HHA), UEA-I (Ulex Europeaus Agglutinin-I), UEA-II (Ulex Europeaus Agglutinin-II) and ConA (Concanavalin A), more preferably selected from the group consisting of GNA, MAA and UEA-I.

Preferably, said CD antigen from endothelial origin is selected from the group consisting of CD31 , CD34, CD51 , CD61 , CD62E, CD105, CD106, CD141 , CD142, CD144 and CD 146, more preferably selected from the group consisting of CD62E, CD31 and CD 144.

Preferably, in step a), a further binding agent recognizing a phospholipid, preferably, phosphatidylserine, phosphatidylethanolamine and phosphatidylinositol, more preferably phosphatidylserine, is also used for isolating said Glyco-EMPs populations.

Thus, preferably, said in vitro or ex vivo method for prognosing, diagnosing or monitoring an endothelial dysfunction in a subject comprises, in step a), for isolating Glyco-EMPs populations, the use of three binding agents: (i) an agent binding a specific glycan epitope, (ii) a binding agent recognizing a CD antigen from endothelial origin and (iii) a binding agent recognizing a phospholipid, preferably phosphatidylserine, phosphatidylethanolamine and phosphatidylinositol.

In a particular embodiment of said method, said characterized Glyco-EMPs populations (biomarkers) are selected in the group consisting of a Glyco-EMPs population positive for (i) GNA (GNA⁺), for (ii) CD 144 (CD144⁺), CD31 (CD31⁺) and/or CD62E (CD62E⁺), and (iii) for phosphatidylserine (PS⁺); a Glyco-EMPs population positive for (i) MA A (MAA⁺), for (ii) CD 144 (CD144⁺), CD31 (CD31⁺) and/or CD62E (CD62E⁺) and for (iii) phosphatidylserine (PS⁺); and a Glyco-EMPs population positive for (i) UEA-I (UEA- Ϋ , for (ii) CD31 (CD31⁺) and for (iii) phosphatidylserine (PS⁺).

For instance, an increase of the level of said Glyco-EMPs populations, namely the Glyco-EMPs population (i) GNA+, (ii) CD144+, CD31+ and/or CD62E+, and (iii) PS+; the Glyco-EMPs population (i) MAA+, (ii) CD144+, CD31+, and/or CD62E+ and (iii) PS⁺; the Glyco-EMPs population (i) UEA-I+, (ii) CD31+ and (iii) PS+; the Glyco-EMPs population (i) UEA-I⁺, (ii), CD62E⁺ and (iii) PS⁺ or the Glyco-EMPs population (i) UEA- I⁺, (ii) CD144⁺ and (iii) PS⁺; constitutes a biomarker of the presence of an endothelial dysfunction in said subject, whereas a modulation of said populations may be a biomarker of the progression (evolution) of said endothelial dysfunction.

According to another referred embodiment of said method, it includes comparing the concentration and/or composition of total MPs.

As used herein a Glyco-EMPs population positive for a lectin {e.g., GNA, MMA or UEA-I) refers to the population of Glyco-EMPs that binds to said lectin.

As used herein a Glyco-EMPs population positive for a CD antigen refers to a population of Glyco-EMPs that expresses said CD antigen at its surface.

As used herein a Glyco-EMPs population positive for phosphatidylserine (PS) refers to the population of Glyco-EMPs that expresses phosphatidylserine at its surface. The presence of phosphatidylserine at the surface of a Glyco-EMP can be determined using annexin V or lactadherin, for instance. As used herein the term "endothelial dysfunction" refers to any pathology involving endothelium; it refers preferably to septic shock, hypertension, metabolic diseases (such as hypercholesterolaemia, non-alcoholic fatty liver disease, non-alcoholic steatohepatitis and diabetes), atherosclerosis, cardiovascular diseases (such as coronary artery diseases, abdominal or thoracic aortic aneurysm), pulmonary artery diseases as well as some antiphospholipid syndrome (APS) and CNS disorders.

The term "cardiovascular disease" encompasses all disorders characterized by insufficient, undesired or abnormal cardiac function, such as ischemic heart disease, hypertensive heart disease and pulmonary hypertensive heart disease, toxic or infectious cardiomyopathies, heart failure, stroke and ischemia.

These methods can involve determining Glyco-EMPs concentration and/or composition in test and control samples by using technologies such as flow cytometry, mass spectrometry, gel electrophoresis, an immunoassay (e.g. immunoblot, immunoprecipitation, ELISA, immunonephelometry assay), nucleic acid amplification, procoagulant activity, and/or electron microscopy on biological fluids or Glyco-EMPs populations (that is, by applying technologies that allow the identification of biomarkers of interest). These methods may involve the isolation and the comparison of Glyco-EMPs populations within selected biological fluids (e.g. blood) where Glyco-EMPs populations can be isolated.

Optionally these methods may involve comparing the concentration and/or composition of total MPs. Still optionally, the methods may involve the detection of the biomarker(s) that can be found associated with Glyco-EMPs populations within a tissue. Such tissues can be the ones from which Glyco-EMPs populations can be originated (e.g. obtained from biopsies of the CNS) but can also be any other cell types or biological material of interest for diagnosing or monitoring a disorder. In particular, Glyco-EMPs populations of interest can be originated by microvasculature at the interface between blood and tissues. These microvascular endothelial cells can have specific functions in patho-physiological processes that are associated to the tissues or organ from which they derived from. For example Glyco-EMPs may be isolated ex vivo from pancreatic tissues that are obtained from an animal model or a human affected by Type-I/-II diabetes in tissue/organ culture conditions.

According to another aspect of the invention, it provides an in vitro or ex vivo method for prognosing, diagnosing or monitoring a disorder, preferably an endothelial dysfunction, in a subject, comprising determining:

a) the concentration and/or the composition of a Glyco-EMPs population presenting a specific glycan epitope; b) the ratio between the concentration and/or the composition of total MPs population and of a Glyco-EMPs population of (a);

c) the ratio between the concentration and/or the composition of a Glyco-EMPs population of (a) and a different Glyco-EMPs population and/or

d) the concentration and/or the composition of MPs not presenting a specific glycan epitope of (a);

in an appropriate biological fluid sample obtained from said subject.

It will be understood that it is not absolutely essential that an actual control sample be run at the same time that assays are being performed on a test sample. Once "normal" (i.e. control) concentration and/or composition of Glyco-EMPs populations (i.e. Glyco- EMPs vs total plasma MPs population ratios, Glyco-EMPs populations presenting or not a specific cell surface antigen) have been established, these levels can provide a basis for comparison without the need to rerun a new control sample with each assay. The comparison between the test and control samples using appropriate statistical methods and criteria should provide a basis for a conclusion on the state of the subject, for instance whether the disorder is progressing or regressing in response to a treatment, if the subject is (or will be) affected by a disorder, or if the subject has been exposed to a drug, to a traumatic insult, or any other event that alters the metabolism within a biological fluid, and in particular the generation of Glyco-EMPs populations.

The term "diagnosing" refers to diagnosis, prognosis, monitoring a disorder in a subject individual that either has not previously had the disorder or that has had the disease but who was treated and is believed to be cured. This application of the methods of the invention can be extended to the selection of participants in (pre) clinical trials, and to the identification of patients most likely to respond to a particular treatment.

The term "monitoring" refers to tests performed on patients known to have a disorder for the purpose of measuring its progress or for measuring the response of a patient to a therapeutic or prophylactic treatment.

The term "treatment" refers to therapy, prevention and prophylaxis of a disorder, in particular by the administration of medicine or the performance of medical procedures, for either prophylaxis (prevention) or to cure the infirmity or malady in the instance where the patient is affected.

The quantitative and/or qualitative in vitro/ex vivo analysis of Glyco-EMPs populations provide relevant information for evaluating a subject, having a predefined clinical status, disorder predisposition, positive/negative response to a treatment, and/or sensibility to a drug or a pathogen (see Figure 1). In fact, this analysis may lead to a definition of a profile in which different elements characterizing Glyco-EMPs populations are used, including concentration (absolute or relative to total MPs or cell type-specific MPs in the preferred biological fluid), presence/absence of one or more antigens, reduced/increased presence of one or more antigens, size, phospholipid composition, or intracellular components. In this latter case, this analysis may include the identification of molecules already characterized as being present within MPs including proteins such as glycan processing enzymes (Varki A et al, 2009), PPARalpha (Benameur T et al., 2010), or caspase-3 (Abid Hussein M et al., 2007), and nucleic acids such as DNA, mRNA, or miRNA (Orozco A and Lewis D, 2010). This method of evaluation can be also applied for characterizing a cell population that is maintained in vitro/ ex vivo (such as cell lines, primary cells, stem cell, tissue material preparations) whereby the presence of Glyco- EMPs population is established in the cell culture supernatant (see Figure 3). In this manner, human Glyco-EMPs can be identified and characterized in patients that have been selected by different criteria (for example at risk, suffering, or under treatment for a disorder).

The Examples demonstrated that, by combining in vitro and in vivo studies of biological fluids, Glyco-EMPs populations can be detected, isolated, and characterized for identifying biomarkers of medical interest, in particular by using proteomic and other technologies that are described in the literature for studying specifically membrane proteins within biological fluids (Lai Z et al, 2009; Cordwell S and Thingholm T, 2010).

All references cited herein are fully incorporated by reference in their entirety. Having now fully described the invention, it will be understood by those of skill in the art that the invention may be practiced within a wide and equivalent range of conditions, parameters and the like, without affecting the spirit or scope of the invention or any embodiment thereof.

EXAMPLES

Example 1 ; Quantification of Glyco-EMPs in the human blood

Materials & Methods

Preparation of platelet-free plasma from human blood in control-case studies

In humans, blood was collected using a classical 19-gauge needle by slow-pull syringe venipuncture and, after discarding the first 2 ml of blood, the subsequent 6 ml of blood transferred into 1/10 volume of 3.2% sodium citrate (Vacutainer Tubes, BD).

In the case-control study for evaluating MPs populations in coronary patients compared to control subjects, two parallel groups (each comprising 6 subjects) were established in open conditions: stable coronary patients (coronarian) versus control subjects.

All participants were Caucasian male aged from 50 to 75 years, in a seemingly good health at the time of selection, and provided written informed consent prior to enrolment. The stable coronary patients were included on the basis of a status defined as "stable coronary disease" that is defined by one or more of the following conditions: stable or unstable angina pectoris with positive ECG stress test or positive myocardial scintigraphy or stenosis of >50% of coronary artery, history of myocardial infarction, history of coronary revascularization, under treatment by Aspirin and/or statin at stable dose for at least 3 months. The control subjects were included as not having a history of coronary disease or family history of coronary disease.

In both cases, within one-two hour of collection, platelet-rich plasma was obtained from the blood samples by centrifugation at l ,500g for 15 minutes at room temperature. The supernatant is then carefully removed and transferred to a new tube. Platelet-free plasma is then obtained by centrifugation at 13,000g for 2 minutes at room temperature. Again, the supernatant is carefully transferred into a new tube and snap-frozen using liquid nitrogen. Samples were stored at -80°C until use.

Production and purification of recombinant human Annexin V protein

The DNA encoding human Annexin V (Genebank NM_001 154) was used for producing Histidine-tagged, recombinant Annexin V in bacteria (E. Coli strain BL21 star PI OS). The recombinant protein results from the fusion of the DNA sequence coding for a synthetic sequence (MGRSHHHHHHGMASMTGGQQMGRDLYDDDKDRWGSE; SEQ ID NO: 1) that includes an hexahistidine tag (HHHHHH; SEQ ID NO: 2) and the Xpress epitope (DLYDDDK; SEQ ID NO: 3; Invitrogen Life Technologies), in 5' to the DNA encoding human Annexin V (amino acids 1-320). The Histidine-tagged, recombinant Annexin V was purified using an HIS-Trap column (GE Healthcare). Purity was assessed by SDS-PAGE gel and sequence was verified by MALDI-TOF mass spectrometry. His- tagged Annexin V was then labelled with FITC (NHS-Fluoroscein; Thermo-Scientific, Pierce Protein Research Products; Cat. No. 46410), following the manufacturer's protocol. Quantification of phospholipid-positive MPs and Glyco-EMPs populations

Phospholipid-positive MPs populations presenting phosphatidylserine or Glyco- EMPs populations that present a PolySia-related epitope as relevant targets on their surface were identified in Platelet-free plasma as cell particles having a diameter comprised between 0.5 and Ι μη and positively stained with labelled Annexin V. The samples of platelet-free plasma (each having a volume of 30μ1) were incubated in the dark for 30 minutes with either His-tagged, FITC-labeled Annexin V (60ng) in incubation buffer (2.5mM CaCl₂, 140mM NaCl, l OmM Hepes pH 7.4), or with the indicated lectins and primary antibodies (10 μΐ of labeled lectin or primary antibody at 50-300μg/ml were mixed, spinned 5 minutes at 450g to remove the excess of lectin and analyzed by flow cytometry). In addition, counting beads (Flowcount Fluorospheres, Beckman-Coulter, France; 30μ1 / sample with a 0.5μηι cut-off to obtain reproducible results) were added to each sample in order to express MPs counts as absolute number per μΐ of biological fluid.

The analysis was performed for each sample by FACS (FC500 flow cytometer; Beckman-Coulter, France), using the two MPs-labelling compounds in parallel, or simultaneously on the same aliquot, then analysing separately the two signals.

Results & Conclusions

The molecular features of potential medical interest that are present in Glyco-EMPs populations can be characterized by applying a process in which samples of biological fluid having different origins (e. g. before or after treatment; in control individuals or patients) are isolated and compared for the quantitative and qualitative features using appropriate technologies (Figure 1).

At least initially, the Glyco-EMPs populations were obtained and defined by using a process that allows comparing the concentration of glycan-specific and phospholipid- specific MPs populations in different subjects, in particular for assessing the contribution of glycan-specific Glyco-EMPs populations to the overall MPs concentration in a sample of biological fluid, as generally defined on the basis of the presence of phosphatidylserine on their surface. This approach was applied for initiating a more general evaluation of quantitative, size, and/or molecular features of Glyco-EMPs populations that are obtained using samples of biological fluid isolated from different subjects in conditions that maintain the MPs integrity.

In a preliminary study, a group of human subjects were selected according to their clinical status associated to Coronary Artery Disease. This disorder is the main death cause in the industrialized world and is mainly due to atherosclerosis, which is a chronic inflammatory disease, initiated and propagated by continuous damage to the vascular endothelium, referred to as endothelial dysfunction. Apoptosis is a process of cell death that has been shown to occur during the progression of human atherosclerotic plaque. The consequences of apoptosis are deleterious because it enhances plaque thrombogenicity after plaque rupture and leads to acute ischemic events and infarction. Such mechanism is associated with a systemic pro-inflammatory state, with changes in multiple cytokines and shedding of MPs population in blood circulation that are originated from diverse apoptotic or activated cell types. The levels of circulating MP (sub)populations are positively correlated with the development of number of chronic inflammatory diseases including cardiovascular diseases (Mayr M et al., 2008; Boulanger C et al. 2006; Shantsila E, 2009).

In a case-control study that was performed in human subjects, the case group defined as stable coronary patients showed a decrease in the concentration of Annexin V-positive MPs population (data not shown). However, the Glyco-EMPs and EMPs subpopulation that was selected by using combinations of Annexin V, lectins, and an antibody for an endothelial-cell associated antigen (CD31) showed that not only different ratio between the distinct MPs subpopulations in the same group but also, for specific Glyco-EMPs subpopulations (for instance those MAA-/CD31 -positive and SNA-/CD31 -positive) differences that are statistically significant between the two groups that are not found by analyzing CD31 -/Annexin V-positive EMPs subpopulation (Figure 2).

The study of specific Glyco-EMPs populations can be also performed in human subjects before or after a medical intervention, such as a drug treatment or a surgical intervention. In another case-control study, the blood of patients that underwent bariatric surgery due to their obesity (and affected or not by type 2 diabetes) was obtained before the surgery and 12 months following the surgery. Total annexin V-positive MPs and Glyco-EMPs populations were isolated and measured as indicated in the previous study, alone or in combination with selected CD antigens. In particular, the presence of Glyco- MPs that are characterized by their UEA-I binding was determined, since the presence of UEA-I binding epitopes is often associated to endothelial cells (Goon P et al., 2006).

It has been observed that, following the intervention, total MPs and Glyco-EMPs were reduced by at least 80% when compared to initial levels, with a constant ratio between AnnexinV -positive MPs vs. UEA-I positive, Thus, Glyco-EMPs can be used in support of the medical management of diseases related to obesity or metabolism in particular when combining a lectin (such as UEA-I) and another binding agent, for instance Annexin V and/or an antibody that is specific for an endothelium cell-related CD antigen (such as CD31 or CD 144). This analysis can be extended to further MPs subpopulations that present one or more CD antigens that also characterize cells circulating in the blood (such as CD31 , CD51 , CD59, CD62e, or CD105) at scope of comparing the trends in the concentration of different MPs and Glyco-EMPs sub-populations in association to medical intervention, drug treatment, disease symptoms, subject weight, and/or metabolic indexes.

A similar analysis can be made not only in blood samples of human origin but also in rodent models for human diseases such as mice models for atherosclerosis research (Whitman S, 2004), as well as rat models for diabetes (Chen D and Wang M, 2005), or db/db mice model for type 2 diabetes and dyslipidemia (Reifel-Miller A et al, 2005).

It can be suggested that, by using binding agents specific for different glycans and/or different biological fluids, different percentages of total MPs populations are identified as Glyco-EMPs populations. However, the specific Glyco-EMPs populations (such as PolySia-positive Glyco-EMPs or Glyco-EMPs subpopulations identified according to lectin binding features as UEA-I positive Glyco-EMPs subpopulations) are of particular interest. Glyco-EMPs subpopulations appear as being generated and metabolized within biological fluids in manners that appears being either very different or very similar to the overall MPs populations (even if such Glyco-EMPs populations have almost certainly different cell origins and even if a different cell surface components are known to be modified with glycans) depending on glycan epitope/binding agent (such as PolySia- related epitopes and corresponding binding agents) that is used.

This finding is of major importance since there are no evidence in the literature on if and how the presence of a specific glycan on the surface of EMPs allow identifying alternative EMPs populations that have relevant features for assessing the overall EMPs and MPs concentration in a sample of biological fluid, as well as and drug- and/or disease induced changes of potential medical interest, as in the rodent models described in this Example.

Rodent and (more importantly) human plasma can provide detectable amounts of Glyco-EMPs populations that can be isolated and characterized in conditions that maintain MPs integrity and that can provide a different understanding of MPs relevance in pathophysiological mechanisms and/or in the definition of biomarkers. Thus, Glyco-EMPs subpopulations can be compared and used (as such or as proxy) for establishing novel biomarkers within biological fluids that reflects a patho-physiological state of a subject and therefore could be used for disease prediction, diagnosis, prognosis and follow-up, with a strong focus on both chronic conditions (associated to atherosclerosis, diabetes, cancer, and in general metabolic and cardiovascular diseases) and acute conditions (including transplantation, stroke, or traumatic injuries).

Example 2: Quantification and characterization of Glyco-EMPs subpopulations that are generated in cell culture conditions.

Materials & Methods

Cell culture protocols

Human Umbilical Vascular Endothelial Cells (HUVEC) (CC-2517; Cambrex,

Walkersville, USA) were cultured in EBM®, Endothelial Basal Medium (EBM media; Cambrex, CC-3121 ; Walkersville, USA) and EBM plus SingleQuots of growth supplements (EGM BulletKit, Cambrex CC-3124; Walkersville, USA). Cells were seeded at 2 500 to 5 000 cells / cm² into 150 cm² culture flask (BD Bioscience; Le Pont-De-Claix, France) and maintained at 37°C in 5% C02 atmosphere. Cells were used for experiments when they reached 80% confluence. HUVEC were then treated for 24 hours at 37°C without (control condition) or with camptothecin at 0.5 μΜ.

Analysis of cell and MPs populations by flow cytometry

At the end of treatment, cells and MPs populations were analyzed separately (see Example 1 for flow cytometry).

In the case of cells, after removing the cell culture supernatant, the cells were trypsined and counted. 500 000 cells were analyzed by flow cytometry using Annexin V (fluorescein isothiocyanate (FITC) or phycoerythrin (PE)) and the glycan-binding agents: Maackia amurensis leukoagglutinin (MAA FITC; EY laboratories, CA, USA), Wheat Germ agglutinin (WGA), Sambuscus nigra lectin SNA), 2-2B antibody (IgM)-PE; (MACs, 130-093-274), or KM93 antibody (anti-sialyl Lewis X, 10 g ml; VWR International, Fontenay sous bois, France). ΙΟΟμΙ of cells and 10 μΐ of each glycan-binding agent at 50μg/ml were mixed, spined 5 minutes at 450g to remove the excess of glycan-binding agent and analyzed by flow cytometry. When needed, the following secondary antibodies were used, goat anti-IgG-PE (Abeam; Cat. No. ab97070), goat anti-mouse IgM-PE (Invitrogen; Cat. No. M31504), rat anti-mouse IgG2A-FITC (BD Pharmingen; Cat. No. 553390).

In the case of MPs populations, cell culture supernatants were collected and spun down for 15 minutes at l ,500g at room temperature. Total MPs were then isolated by ultracentrifugation at 30,000g for 45 minutes at room temperature. Supernatants were discarded and pellets were re-suspended in PBS buffer (final volume of 400 microliter per condition). Samples were stored at -80°C until use. Distinct MPs populations were analysed by flow cytometry using Annexin V PE (Beckman Coulter, 731726) and the same antibodies and glycan-binding agents indicated above.

Results & Conclusions

EMPs populations within biological fluids may be therefore considered as a reporter of the patho-physiological status of large variety of cell types for which specific biomarkers can be identified in MPs populations, or more specifically Glyco-EMPs populations that are associated to specific cell types, biological activities, exposure to chemicals, and/or diseases.

The molecular features that allow differentiating cell type- and/or disease-specific Glyco-EMPs (sub)populations of potential major interest can be initially defined by means of specific primary cells or cell lines that are related to the disease which and that can be used to generated Glyco-EMPs within cell culture, in more tightly controlled conditions than in physiological biological fluids (Figure 3). However, on the basis of the findings of Example 1, the presence and metabolism of cell type-specific, PolySia-positive Glyco- EMPs populations can be also evaluated to understand if and how its ratio with Annexin V-positive MPs populations changes between cell types, and thus if different cell type contributes in different manner to the PolySia-positive Glyco-EMPs populations.

The release of Annexin V-positive MPs populations that are also positive for different PolySia-related epitopes was then tested in HUVEC cultures to verify if these cell lines release distinct Glyco-EMPs populations at different rates.

In a first set of experiments Glyco-EMPs subpopulations present in HUVEC cultures were performed by flow cytometry but also extended to monoclonal antibody recognizing variants of PolySia-related epitopes (KM93 and 2-2B antibody, see Table 2), In fact, the concentration of the selected Glyco-EMPs subpopulations not only differ in a significant manner but also provide a profile of glycan epitopes that differ with the one on HUVEC surface (Figure 4). For instance, SNA- and WGA-positive, Glyco-EMPs populations almost correspond to the whole Annexin V-positive EMPs population, meanwhile MAA or KM93 provide a more limited subset of EMPs.

These data confirm that specific Glyco-EMPs subpopulations can provide substantially different information on how different apoptotic and other pathophysiological events affect MPs shedding and/or metabolism, and on how biomarkers can be found characterized as being associated to MPs subpopulations (being Glyco-EMPs subpopulations or MPs subpopulations not presenting specific glycan epitopes). Moreover, the information on such apoptotic and other patho-physiological events that have been characterized by studying cell surface and composition cannot be equally applied to MPs surface and composition since shedding, metabolic and other biological mechanisms can clearly distinguish them from the originating cells.

The study of qualitative and quantitative effects of a (non)-physiological stimulus on Glyco-EMPs (sub-)populations can be also performed in other human cell lines that are originated from endothelium in specific localization (i.e. of microvascular or macrovascular origin), that are commonly used for studying more specific patho- physiological processes and drug effects. Examples of such cell lines are HMVECs (human mitral valve endothelial cells), HAEC (human aortic endothelial cells), HCAEC (human coronary artery endothelial cells), and HPAEC (human pulmonary artery endothelial cells).

Cell-based assays involving HUVEC or HMVEC can be established by exposing the cells to a culture medium with or without exposure to a stress or apoptotic agent such as homocysteine. In fact, hyperhomocysteinemia is an independent risk factor for the development of atherosclerosis and homocysteine or homocysteine derivatives are used studying their in vitro cytotoxicity and pro-inflammatory properties (Kerkeni M et al., 2006).

Total, annexin V-positive MPs and Glyco-EMPs populations were isolated and measured as indicated in the previous study, alone or in combination with selected CD antigens, using HUVECs or HMVECs. In particular, the presence of Glyco-MPs that are characterized by their UEA-I or MAA binding was determined, showing that the two lectins identify Glyco-EMPs sub-populations that provide a different response to homocysteine and/or folic acid (a compound counteracting effects of homocysteine) in a important fraction of total EMPs (between 10% and 85% of total HUVEC-originated EMPs). If the concentration of Glyco-EMPs populations that are UEA-I positive appear following the trend of the one of Annexin V positive MPs populations (with an increase following homocysteine exposure an decrease following further addition of folic acid), Glyco-EMPs populations that are MAA positive appear increasing only following the treatment with folic acid.

Thus, Glyco-EMPs populations (and in particular Glyco-EMPs populations that are positive for PolySia-related epitopes, UEA-I binding, or MAA binding) are biological products that are different from those used so far for studying MPs and that provide novel opportunities to identify MPs-associated biomarkers, in particular using proteomic or immunological technologies. Glyco-EMPs populations can be used for studying the presence of cell-type specific antigens that are specifically associated to this MPs population, and consequently Glyco-EMPs populations can be of even greater medical interest for evaluating the clinical status of a patient, the progression of a disease or the therapeutic efficacy of a drug treatment. At this scope, another binding agent, for instance Annexin V and/or an antibody that is specific for an endothelium-related CD antigen (such as CD41 or CD 144) can be used at scope of comparing the trends in the concentration of the different MPs and Glyco-EMPs sub-populations in association to the cell exposure to specific cellular stress and/or (anti-)apoptotic agents.

Further analysis can be pursued in the HUVECs- or HMVECs-based model system using proteomics techniques for detecting total proteins and glycoproteins in the cell culture conditions known in the literature for altering EMPs production, function, and/or composition (Abid Hussein M et al., 2008; Abid Hussein M et al., 2003; Abid Hussein M et al., 2007; Szotowski B et al., 2007; Combes V et al., 1999; Simak J et al., 2002; Fu Y et al., 2010; Gelderman M et al., 2008; Diamant M et al. 2008). Specific studies were also performed in specific locations such as pulmonary microvascular endothelium (Bauer N and Smith C, 2010) or bone marrow-derived endothelial progenitor cells (Benameur T et al., 2010), as well as general apoptotic events in HUVEC (Hayot C et al., 2002). The results can be compared with already generated for EMPs proteome (Banfi, C et al., 2005; Mayr M et al., 2008) and for HUVEC proteome in different conditions (Bruneel A et al., 2003; Qiu J et al., 2008; Bruneel A et al., 2005).

A similar proteomic approach can be applied to compare different MPs populations of interest such as Annexin V-positive, antigen-specific, and/or Glyco-EMPs populations isolated from the same biological fluid or in different subjects (e. g. affected or not by a disease, treated or not with a drug). The resulting proteomic maps can be then analyzed to identify candidate biomarkers (presenting or not a glycan) using different technologies that are specific for cell surface glycans, membrane microdomains, and/or membrane proteins (Lai Z et al., 2009; Cordwell S and Thingholm TE, 2009; Apweiler R et al., 2009a; Apweiler R et al., 2009b) in order to define the surfaceome of Glyco-EMPs populations. Together with flow cytometry, this approach can be used to establish if and how Glyco-EMPs and total MPs populations are differentially metabolized within the biological fluid, expose cell surface components in specific manners, and/or interact differentially with cells and other components of the biological fluid. The analysis can be performed considering both known endothelial-specific antigens that have been already characterized in-EMPs (sub) populations within biological fluids and/or in cell culture conditions (Burnier L et al., 2009; Orozco A and Lewis D, 2010; van der Heyde H et al., 201 1), and the extensive lists of glycans and of the corresponding glycan-binding agents (such as such as monoclonal antibodies, plant lectins, and chemicals) that have been published (see Table l and Table 2).

Example 3 : Quantification and characterization of CD and Glyco-EMPs subpopulations that are generated by different endothelial cell lines after physiological stimuli

The goal of this experiment was to differentiate subtypes (or subpopulations) of endothelial cell population based on their MPs surface marker exposition: CD31 , CD 144, CD62e, GNA, MAA, UEA1 and their combination.

Materials & Methods

Lipoproteins preparation and modification (acetylation)

LDL with density 1.030-1.063 g/ml was isolated by sequential ultracentrifugation of EDTA-anticoagulated fasting plasma obtained from healthy normolipidemic volunteers applying the KBr standard method as described by Delalla et al., 1954. The isolated LDL fractions were dialyzed over night against PBS-EDTA 0.1 g/1 pH 7.4 (3 cycles of dialysis).

LDL fraction was acetylated according to the method of Basu et al, 1976 with acetic anhydride at 4°C. The acetylated LDL (acLDL) were dialyzed for 36 hours at 4°C against standard buffer (PBS 0.01 M, pH 7.4).

Cell culture protocols

Human Coronary Artery Endothelial Cells (HCAEC) (Lonza, CC-2585), Human

Aortic Endothelial Cells (HAEC) (Lonza, CC-2535), Human Microvascular Endothelial

Cells (HMVEC) (Lonza, CC-7030), Human Umbilical Vein Endothelila Cells (HUVEC) (Lonza, CC-2517) were cultured in EBM®, Endothelial Basal Medium (EBM media;

Cambrex, CC-3121 ; Walkersville, USA) containing 2% Fetal Calf Serum (FCS) for maintenance.

Cells were seeded at 2 500 to 5 000 cells / cm² into 150 cm² culture flask (BD Bioscience; Le Pont-De-Claix, France) and maintained at 37°C in 5% C0₂ atmosphere. Cells were used for experiments when they reached 80% confluence. HCAEC were then treated for 24 hours at 37°C without (control condition) or with native or Acetylated LDL (LDL- Ac) at 100 μg/ml in EBM medium supplemented with 0.4% SR3. In another experiment, cells were treated with increasing doses of glucose: 5.5-16.7 and 35 mM in the same medium of treatment.

Analysis of cells and MPs populations by flow cytometry

In the case of cells, after removing the cell culture supernatant, the cells were treated with accutase and suspended at around 3.10⁶ cells in a PBS/EDTA (l OmM). Cells were then analyzed by flow cytometry using: Annexin V (FITC or PE), the glycan-binding agents: Maackia amurensis leukoagglutinin (MAA FITC F-7801-2; EY laboratories, CA, USA), Galantus Nivalis Agglutinin GNA (GNA FITC F-7401-2; EY Laboratories, CA, USA), Ulex Europeaus Agglutinin UEA1 (UEA1 FITC F-2201 -2; EY Laboratories, CA, USA) and anti-CD 144 PE (10 test Beckman; PN AZ07481), anti-CD31PE (IO Test Beckman; IM2409) and anti-CD62E (BD Pharmingen; 551 145). 100 μΐ of cells (300 000 cells) and 10 μΐ of each glycan-binding agent at 50 μg/ml were mixed, spined 5 minutes at 1500 g to remove the excess of binding agent and analyzed by flow cytometry. For negative control anti CD antibodies, IgGl (mouse PE) (IO test Beckman; A07796), was used.

In the case of MPs populations, cell culture supematants were collected and spun down for 15 minutes at l ,500g at room temperature. Total MPs were then isolated by ultracentrifugation at 30,000g for 45 minutes at 18°C. Supematants were discarded and pellets were re-suspended in PBS buffer (MP preparation from 20 ml of medium was resuspended in 150 μΕ). Samples were stored at -80°C until use. Distinct MPs populations were analysed by flow cytometry using Annexin V PE (Beckman Coulter, 731726) and the same antibodies and glycan-binding agents as indicated above for cell analysis.

Results & Conclusions

A) Regulation of antigenic exposition of endothelial dysfunction in a control situation.

Table 3 showed that in control conditions, the glycosylation profile is not discriminant between HAEC, HCAEC and HUVEC. Only HMVEC present a specific pattern of glycosylation detected with GNA and MAA lectin binding by flow cytometry.

Conclusion: CD markers did not allow the differentiation of endothelial sub population cells when used alone.

Table 4 showed that when the analysis was performed on Glyco-EMPs generated from different cell subtypes, the complete antigenic profile was drastically different; UEA1 lectin binds ubiquitously MP from different cell origins. In contrast, GNA and MAA lectin binding differentiate HCAEC and HUVEC-derived Glyco-EMPs from HAEC and HMVEC-derived Glyco-EMPs.

These results show that biomarker analysis in the circulation need to be carried out on Glyco-EMPs rather than on cells.

B) Regulation of antigenic exposition with physiological inducers of endothelial dysfunction.

Endothelial cells were treated with acetylated LDL (acLDL) as described in material and methods. Control conditions were cells incubated with native LDL (100 μg/mL).

B. l) Stimulation by acLDL: induction of Glyco-EMPs only

Tables 5 and 6 below illustrate the interest of Glyco-EMPs as biomarkers of an endothelial dysfunction.

B.2) HCAEC profiles compared to HCAEC Glyco-EMPs profiles

Table 7

Tables 7 and 8 correspond to representative data from HCAEC. At the cell level, no significant antigenic profile change was observed between control and acLDL treated cells.

At the MP level, changes are numerous and concerned CD antigen, lectin ligands and their combinations. Physiological stimuli of endothelial dysfunction affect antigenic profile of Glyco-EMPs rather than cell surface pattern. MAA binding is decreased after treatment whereas UEA1 or GNA are increased.

As shown in Tables 7 and 8, Glyco-EMPs profiles are clearly pertinent to identify an endothelial subpopulation, the modulation of the amount of said subpopulation being correlated with an endothelial dysfunction.

HCAEC Glyco-EMPs are thus representative of coronary dysfunction. The most representative Glyco-EMPs profiles of said PS⁺subpopulation are MAA CD62E and UEA1 CD62E which are significantly increased in the model corresponding to coronary dysfunction. B.3) HAEC profiles compared to HAEC Glyco-EMPs profiles

Tables 9 and 10 correspond to representative data from HAEC. At the cell level, no significant antigenic profile change was observed between control and acLDL treated cells.

At the MP level, changes are numerous and concerned CD antigens, lectin ligands and their combinations. Physiological stimuli of endothelial dysfunction affect antigenic profile of Glyco-EMPs.

HAEC Glyco-EMPs are thus representative of artery endothelial dysfunction. The results of Tables 9 and 10 confirm the pertinence of the Glyco-EMPs profiles in detection of endothelial dysfunction.

The most representative Glyco-EMPs profiles of said PS⁺ subpopulations are GNA

CD 144 and MAA CD 144 which are significantly increased in the model corresponding to aortic dysfunction.

B.4) HUVEC profiles compared to HUVEC Glyco-EMPs profiles

Tables 1 1 and 12 correspond to representative data from HUVEC. At the cell level, no significant antigenic profile change was observed between control and acLDL treated cells.

B.5) HMVEC profiles compared to HMVEC Glyco-EMPs profiles

Tables 13 and 14 correspond to representative data from HMVEC cells. At the cell level, no significant antigenic profile change was observed between control and acLDL treated cells.

At MP level, changes are numerous and concerned CD antigens, lectin ligands and their combinations. Physiological stimuli of endothelial dysfunction affect antigenic profile of Glyco-EMPs. HMVEC cells represent a microvascular model of endothelial dysfunction.

The results of Tables 13 and 14 confirm the pertinence of the Glyco-EMPs profiles in detection of endothelial dysfunction.

The most representative Glyco-EMPs profiles of said PS⁺ subpopulation are GNA

CD31 and UEA 1 CD 3 1 which are significantly increased in the model corresponding to microvascular dysfunction.

B.6) Induction of Glyco-EMPs by glucose

In order to determine if other inducers of endothelial dysfunction have the same effect of acLDL on Glyco-EMP surface markers, the Inventors studied also the influence of glucose on endothelial cells and Glyco-EMPs.

Glucose did not affect significantly antigenic profile at cell surface but:

1) Glyco-EMPs production was induced by 1.8-fold for HMVEC derived Glyco- EMPs, 3.6-fold for HAEC derived Glyco-EMPs, 1.9-fold for HCAEC derived Glyco- EMPs and 2.7-fold for HUVEC derived Glyco-EMPs as shown in Tables 15 and 16 below.

2) CD 144 exposition was increased in all cells derived Glyco-EMPs whereas other markers were not affected (CD31 , CD62e) as shown in Tables 17 and 18 below.

B.7) Conclusions

Both glucose and acLDL induce a functional stimulation of endothelial cells. Such a stimulation is detected by an increase of production of Glyco-EMPs only; i.e. this means that no increase may be detected at the cell level. The modifications of MP surface antigenic expression are different when different inducers are used to stimulate an endothelial dysfunction as illustrated in Tables 3 to 18.

REFERENCES

Abid Hussein M et al., 2003. J Thromb Haemost; 1 : 2434-43.

Abid Hussein M et al, 2007. Thromb Haemost; 98: 1096-107.

Abid Hussein M et al., 2008. Thromb Res; 121 : 865-71.

Alley W and Novotny M, 2010. J Proteome Res; 9:3062-72.

Amabile N et al., 2009. J Heart Lung Transplant; 28: 1081-6.

Amabile N et al., 2010. Semin Thromb Hemost ; 36 : 907-916.

Apweiler R et al., 2009a. Clin Chem Lab Med; 47: 724-44.

Apweiler R et al., 2009b. Cytometry Part A; 75 A: 816-832.

Bain J et al., 2009; Diabetes; 58: 2429-43.

Banerjee M and Otonkoski T, 2009. Diabetologia; 52: 621-5.

Banfi C et al., 2005. Proteomics; 5: 4443-55.

Bartik P et al., 2008. Int J Oncol; 32: 1021-31.

Basu SK et al., 1976. Proc. Natl. Acad. Sci. U.S. A; 73: 3178-3182.

Bauer N and Smith C, 2010. Am J Respir Crit Care Med; 181 , A3433.

Benameur T et al., 2010) PLoS One ; 5: el2392.

Bonfanti L, 2006. Prog Neurobiol; 80: 129-64.

Bork E et al, 2007. J Neurochem; 103 Suppl 1 : 65-71.

Brinkman-Van der Linden E et al., 2002. Anal Biochem; 303 : 98-104.

Bruneel A et al., 2003. Proteomics; 3: 714-23.

Bruneel A et al., 2005. Proteomics. 5:3876-84

Burnier L et al., 2009. Thromb Haemost; 101 : 439-51.

Carlsson S et al., 2007. Glycobiology; 17: 663-76.

Chen D and Wang M, 2005. Diabetes Obes Metab; 7: 307-17.

Combes V et al., 1999. J Clin Invest; 104: 93-102.

Cordwell S and Thingholm T, 2010. Proteomics; 10: 61 1 -27.

Cuchacovich M et al. 2001. Clin Exp Rheumatol; 19: 673-80.

DeLalla et al., 1954. Methods of Anal. Biochem; 1 : 459-478.

Diamant M et al. 2008. Thromb Haemost; 100: 489-97. Dignat-GeorgeF and Boulanger C, 201 1. Arterioscl Throm Vase Biol; 31 : 27-33.

Doeuvre L et al, 2009. J Neurochem; 110: 457-68.

Dotan N et al., 2006. Lupus; 15: 442-50.

Fu Y et al., 2010. Thromb Haemost;104: 1235-41.

Galuska S et al., 2007. Anal Chem; 79: 7161-9.

Gelderman M et al., 2008. Int J Nanomedicine; 3: 59-68.

Gemeiner P et al., 2009. Biotechnol Adv; 27: 1-15.

Ghesquiere B et al.,2007. J Proteome Res; 6: 4304-12.

Goon P et al, 2006. Neoplasia; 8: 79-88.

Gupta G et al., 2010. OMICS;14: 419-36.

Hayot C et al., 2002. Int J Oncol; 21 : 417-25

Hayrinen J et al, 1995. J Infect Dis; 171 : 1481-90.

Hayrinen J et al., 2002. Mol Immun; 39: 399-41 1.

Helal O et al., 2010. Nutr Metab Cardiovasc Dis. [Epub ahead of print].

Hu S et al., 2006. Proteomics; 6: 6326-6353.

Iskratsch T et al, 2009. Anal Biochem; 386: 133-46.

Jy W et al., 2006. ASH Annual Meeting Abstracts; 108: abst. 1802.

Kannagi R and Hakomori S, 2001. Adv Exp Med Biol; 491 : 587-630.

Kannagi R et al., 2009. Glycoconjugate J; 26: 923-928.

Kerkeni M et al., 2006. Mol Cell Biochem;291 : 119-26.

Keshava Prasad T et al., 2009. Nucleic Acids Res;37: D767-72.

Kitazume S et al., 2010. J Biol Chem; 285: 6515-21.

Klinkner D et al., 2006. Shock; 25: 575-80.

Kobzdej M et al., 2002. Blood; 100: 4485-94.

Krishnamoorthy L and Mahal L, 2009. ACS Chem. Bio; 4: 715-732.

Krueger K and Srivastava S, 2006. Mol Cell Proteomics; 5:1799-810.

Kunding A et al., 2008. Biophys J; 95: 1 176-88.

Kurogochi M et al.,2010. Mol Cell Proteomics; 9: 2354-68.

Lai Z et al., 2009; Curr Opin Mol Therap; 11 : 623-631.

Lai S et al., 2009. Mol Cell Proteomics; 8: 799-804.

Lawrie A et al. 2009. Vox Sang; 96: 206-12.

Lee T et al, 2009. BMC Res Notes; 2: 1 1 1.

Lemmon M, 2008. Nat Rev Mol Cell Biol. 9: 99-11 1.

Leroyer A et al., 2007. J Am Coll Cardiol 49: 772-777.

Leroyer A et al., 2010. Thromb Haemost; 104: 456-463.

Lester E and Babensee, 2003. J Biomed Mater Res A; 64: 397-410.

Li S et al., 2009. Nucleic Acids Res; 37: D907-912. Logue et al. 2009. Nat Protoc; 4: 1383-95.

Mahal L, 2008. Anti-Cancer Ag Med Chem; 8: 37-51.

Mayr M et al., 2008. Circ Cardiovasc Genet; 2: 379 - 388.

Merrick B, 2008. Brief Funct Genomic Proteomic; 7: 35-49.

Mezentsev A et al., 2005. Am J Physiol Heart Circ Physiol; 289: HI 106-1 1 14

Miguet, L. et al., 2007. Subcell Biochem; 43: 21-34.

Mitoma J et al.,. 2009. Glycoconj J; 26: 51 1-23.

Mirza S and Olivier M, 2008. Physiol Genomics; 33: 3-1 1.

Morel O et al, 2009. Haematologica; 94:313-317.

Nilsson J et al, 2009. Nat Methods; 6: 809- 11.

Nomura S et al., 2009. J Atheroscler Thromb;16: 878-87.

Nozaki T et al., 2009. J Am Coll Cardiol; 54: 601 -8.

Nozaki T et al., 2010. Eur J Heart Fail; 12: 1223-8.

Orozco A and Lewis D, 2010. Cytometry A; 77: 502-14.

Patane J et al., 2009. Carbohydr Res; 344: 820-4.

Peterson D et al., 2008. Proteomics ; 8 : 2430-2446.

Piccin A et al., 2007. Blood Rev; 21 : 157-171.

Priego Capote F and Sanchez J, 2009. Mass Spectrom Rev; 28: 135-46.

Qiu J et al., 2008. J Cell Biochem; 104: 1054-64.

Reifel-Miller A et al., 2005. Mol Endocrinol; 19: 1593-605.

Rek A et al., 2009. Br J Pharmacol; 157: 686-94.

Robert S et al., 2009. J Thromb Haemost; 7: 190-197.

Rukoyatkina N et al., 2009. Br J Haematol; 144: 591-602.

Sato C et al, 1995. J Biol Chem; 270: 18923-8.

Sato C et al., 2000. J Biol Chem; 275: 15422-31.

Shah M et al., 2007. Blood (ASH Annual Meeting Abstracts) 1 10: Abstract 702.

Shi J and Gilbert G, 2003. Blood; 101 : 2628-2636.

Simak J et al., 2002. BMC Cell Biol.;3: l 1.

Stace C and tistakis N. 2006. Bioch Biophys Acta; 1761 : 913-926.

Szotowski B et al, 2007. Cardiovasc Res; 73: 806-12.

Tate E, 2008. J Chem Biol; 1 : 17-26.

Taylor M and Drickamer , 2009. Glycobiology; 19: 1 155-62.

Terrisse A et al., 2010. J Thromb Haemost; 8: 2810-9.

Thapa N et al, 2008; J Cell Mol Med; 12: 1649-60.

Thery C et al., 2009. Nat Rev Immunol; 9: 581-93.

Thornalley P, 2005,. Perit Dial Int; 25: 522-33.

Thouvenot E et al., 2008. J Proteome Res; 7: 4409-4421. Tramontano A et al., 2010. Mediators Inflamm; 2010:250476.

Tushuizen M et al., 201 1. Arterioscl Throm Vase Biol; 31 : 4-9.

Turer A et al;, 2009; Circulation; 1 19: 1736-46.

van der Heyde H et al., 201 1. Methods Mol Biol; 699: 337-54.

van Gaal E et al., 2010. J Control Release; 141 :328-38.

Varki A et al., 2009. Essentials of Glycobiology. Cold Spring Harbor Laboratory Press Varki A, 2008. Trends Mol Med; 14: 351-360.

Wang J et al, 2009. Eur J Pharmacol.; 614: 77-83.

Whitman S, 2004. Clin Biochem Rev; 25: 81-93.

Woywodt A et al.,2003. Lancet; 361 : 206-10.

Wu et al., 2009. Glycoconjugate J; 26: 899-913.

Yabe U et al., 2003. J Biol Chem; 278: 13875-80.

Yuana Y et al. 2010. J Thromb Haemost; 8: 315-23.

Zuber C et al., 1992. J Biol Chem; 267: 9965-71.

Claims

1. An in vitro or ex vivo method for prognosing, diagnosing or monitoring an endothelial dysfunction in a subject, comprising:

a) isolating Glyco-EMPs populations in a test sample of biological fluid of said subject by means of an agent binding a specific glycan epitope and at least one further binding agent recognizing a phospholipid, a protein, a lipid or a glycan other that said selected glycan epitope, preferably said at least further binding agent recognizing a CD antigen from endothelial origin;

b) characterizing said test sample Glyco-EMPs populations, and

2. The method according to claim 1, characterized in that said agent binding a specific glycan epitope is a lectin.

3. The method according to claim 2, characterized in that said lectin is selected from the group consisting of GNA (Snowdrop lectin), PNA (Peanut agglutinin), VVL (Hairy vetch lectin), WGA (Wheat Germ agglutinin), SNA (Sambuscus nigra lectin), MAL/MAA (Maackia amurensis leukoagglutinin), MAH (Maackia amurensis hemoagglutinin), LFA (Limax flavus agglutinin), BS-1 (Bandeiraea Simplicifolia-1), Hippeastrum hybrid (HHA), UEA-I (Ulex Europeaus Agglutinin-I), UEA-II (Ulex Europeaus Agglutinin-II) and ConA (Concanavalin A), more preferably selected from the group consisting of GNA, MAA and UEA-I.

4. The method according to any of claims 1 to 3, characterized in that said CD antigen from endothelial origin is selected from the group consisting of CD31 , CD34, CD51, CD61 ,

CD62E, CD105, CD106, CD141, CD142, CD144 and CD146, more preferably selected from the group consisting of CD62E, CD31 and CD 144.

5. The method according to any of claims 1 to 4, characterized in that in step a), a further binding agent recognizing a phospholipid, preferably, phosphatidylserine, phosphatidylethanolamine and phosphatidylinositol; more preferably phosphatidylserine.

6. The method according to any of claims 1 to 5, characterized in that said step b) of characterizing said Glyco-EMPs populations is performed by determining the concentration and/or the composition of said Glyco-EMPs populations.

7. The method according to any of claims 1 to 6, characterized in that said control sample consists of a physiological sample or a test sample Glyco-EMPs populations previously determined in the same said subject.

8. The method according to any of claims 1 to 7, characterized in that said characterized Glyco-EMPs populations are selected in the group consisting of a Glyco- EMPs population (i) GNA⁺, (ii) CD144⁺, CD31⁺ and/or CD62E⁺ and (iii) PS⁺; a Glyco- EMPs population (i) MAA⁺, (ii) CD144⁺, CD31⁺ and/or CD62E⁺ and (iii) PS⁺; a Glyco- EMPs population (i) UEA-I⁺, (ii) CD31⁺ and PS⁺; a Glyco-EMPs population (i) UEA-I⁺, (ii), CD62E⁺ and (iii) PS⁺ or a Glyco-EMPs population (i) UEA-I⁺, (ii) CD144⁺ and (iii) PS⁺.

9. The method according to any of claims 1 to 8, characterized in that said endothelial dysfunction is selected from the group consisting of septic shock, hypertension, metabolic diseases, atherosclerosis, cardiovascular diseases, pulmonary artery diseases, antiphospholipid syndrome (APS) and CNS disorders.

10. An in vitro or ex vivo method for prognosing, diagnosing or monitoring a disorder, preferably an endothelial dysfunction, in a subject, comprising determining:

a) the concentration and/or the composition of a Glyco-EMPs population presenting a specific glycan epitope;

b) the ratio between the concentration and/or the composition of total MPs population and of a Glyco-EMPs population of (a);

in an appropriate biological fluid sample obtained from said subject.

1 1. An in vitro or ex vivo method for identifying Glycan-containing Endothelial Microparticles (Glyco-EMPs) populations characterized in that it comprises the following steps:

12. The method according to claim 1 1, characterized in that the Glyco-EMPs populations having a size comprised between 100 and 1000 nanometres are separated from said acellular fraction.

13. The method according to any of claims 1 1 and 12, characterized in that said CD antigen from endothelial origin is selected from the group consisting of CD31, CD34, CD51,

CD61, CD62E, CD105, CD106, CD141 , CD142, CD144 and CD146, more preferably selected from the group consisting of CD62E, CD31 and CD 144.

14. The method according to any of claims 1 1 or 12, characterized in that said agent binding a specific glycan epitope is selected from the group consisting of antibodies, lectins, enzymes containing carbohydrate recognition domain, cytokines, chaperone and transport proteins, microbial carbohydrate-binding proteins, glycosaminoglycan-binding proteins, preferably a lectin.

15. The method according to claim 14, characterized in that said lectin is selected from the group consisting of GNA (Snowdrop lectin), PNA (Peanut agglutinin), VVL (Hairy vetch lectin), WGA (Wheat Germ agglutinin), SNA (Sambuscus nigra lectin), MAL/MAA (Maackia amurensis leukoagglutinin), MAH (Maackia amurensis hemoagglutinin), LFA (Limax flavus agglutinin), BS-1 (Bandeiraea Simplicifolia-1), Hippeastrum hybrid (HHA), UEA-I II (Ulex Europeaus Agglutinin-I), UEA-II (Ulex Europeaus Agglutinin-II) and ConA (Concanavalin A), more preferably selected from the group consisting of GNA, MAA and UEA-I.

16. The method according to any of claims 1 1 to 15, characterized in that the selected glycan epitope is a PolySia-related epitope or a Sia related epitope.

17. The method according to any of claims 1 1 to 15, characterized in that the selected glycan epitope is a fucosylated group or a high mannose group.

18. The method according to any of claims 1 to 17, characterized in that the biological fluid is selected from the group consisting of plasma, blood, cerebrospinal fluid, proximal fluid and cell culture supernatants.

19. The method according to any of claims 1 1 to 18, characterized in that the Glyco- EMPs populations are separated from said acellular fraction by means of a further binding agent that recognizes a phospholipid, preferably, phosphatidylserine, phosphatidylethanolamine and phosphatidylinositol.

20. The method according to claim 19, characterized in that the phospholipid is phosphatidylserine.

21. The method of any of the preceding claims wherein one of said agent is immobilized on a solid phase and/or is labelled.

22. An isolated Glyco-EMP or Glyco-EMP population obtained according to the methods of claims 1 1 to 21.

23. A Kit for isolating Glyco-EMPs comprising an agent binding a glycan epitope as defined in any of claims 14 to 17 and at least one further binding agent recognizing a CD antigen from endothelial origin as defined in claim 13, and optionally a binding agent recognizing a phospholipid as defined in claim 19.

24. Use of the Glyco-EMP population of claim 22 or of a kit of claim 23 for identifying biomarkers of medical interest in a sample of biological fluid from a subject.

25. Use of the Glyco-EMP population of claim 22 or of a kit of claim 23 for identifying biomarkers of an endothelial dysfunction.

26. Use according to claim 25, characterized in that said endothelial dysfunction is selected from the group consisting of septic shock, hypertension, metabolic diseases, atherosclerosis, cardiovascular diseases, pulmonary artery diseases, antiphospholipid syndrome (APS) and CNS disorders.