Keywords

11.1 Introduction

New blood vessel formation in adults was considered to result exclusively from sprouting of preexisting endothelial cells, a process referred to angiogenesis. Vasculogenesis, the formation of new blood vessels from endothelial progenitor cells, was thought to occur only during embryonic life. Discovery of adult endothelial progenitor cells (EPCs) in 1997 by Isner’s group had major implications for vasculogenic concepts in human but also opened the door for angiogenic therapy [1]. Endothelial progenitor cells contribute to vascular repair and are now well established as postnatal vasculogenic cells in humans [2]. It is now admitted that endothelial colony-forming cells (ECFCs) are the vasculogenic subtype. ECFCs are progenitor cells committed to endothelial lineage and have strong vasculogenic properties in preclinical models of vascularization [2]. They express specific endothelial lineage markers [3] and originate from bone marrow although some organs, and particularly the lung, can serve as a cellular reservoir for ECFCs [2, 4]. ECFCs could be used as a cell therapy product; however, their expansion is quite difficult when they are obtained from adult blood [5, 6]. ECFCs could also be used to prevascularize tissue-engineering construct. Moreover, whatever systemic or local injection, ECFCs have been proposed as vector for gene therapy in several diseases. However, despite a huge interest in these cells, their tissue and molecular origin is still unclear. We recently proposed that endothelial progenitor could come from very small embryonic-like stem cells (VSELs) [7].

This chapter discusses the reported findings from in vitro data and also preclinical studies that aimed to explore stem cells at the origin of vasculogenesis in human and then explore the potential use of ECFCs to promote newly formed vessels or serve as liquid biopsy to understand vascular pathophysiology.

11.2 Endothelial Progenitor Cells Definition in Culture: Future Cell Therapy Product?

Since Asahara first reported the existence of EPCs in peripheral blood, several studies have shown significant heterogeneity among adult ex vivo expanded EPC populations. Likewise, early and late outgrowing EPCs showed comparable in vivo vasculogenic capacity in improving neovascularization in myocardial infarction [8], in vascular graft survival [9], in tumor angiogenesis [10], or in matrigel plug in vivo [11, 12].

Classical isolation methods include adherence culture of total peripheral blood mononuclear cells and the use of magnetic microbeads coated with anti-CD133, anti-CD34, anti-CD14, or anti-CD146 antibodies. At least two types of EPCs have been described [11, 13]. “Early” EPCs or circulating angiogenic cells appear within 4–7 days of culture, are spindle-shaped, and express both endothelial (von Willebrand factor) and leukocyte/monocytic (CD 45 +/− 14) markers, whereas “late” EPCs or endothelial colony-forming cells (ECFCs) appear after 2–3 weeks of culture and have the characteristic of precursor cells committed to the endothelial lineage; they have a cobblestone pattern in culture, and their long-term proliferative potential or their reactivity for growth factor depends on their origin (umbilical or adult blood) [6, 14,15,16].

Early EPCs, which express CD45 antigen, were the cells identified by Asahara et al. in 1997 and have been the most studied EPC population between 1997 and 2010. There are many uncertainties as to their origin and their progenitor properties. Elsheikh et al. [17] tried to identify the subpopulation within monocytic cells that exerts “EPC properties.” These authors isolated CD14+ monocytic cells and purified those cells that expressed vascular endothelial growth factor receptor 2 (VEGF-R2 or KDR). CD14+VEGF-R2+ cells but not CD14+ VEGF-R2− cells contributed to reendothelialization in mice after denuding injury. These data showed that VEGF-R2 is a fundamental receptor-identifying cell with endothelial capacity. A specific subfraction of circulating CD14+ monocytic cells was recently shown to express the stem-cell markers Nanog and octamer-binding transcription factor 4 (oct-4) [18]. These Nanog+ monocytic cells were positive for VEGF-R2 and showed low CD34 expression. CD14+/CD34low/Nanog+ cells appear to represent the active fraction of CD14+/VEGF-R2+ cells isolated by Elsheikh et al. and confirm the origin of stem cells of a subpopulation of early EPCs [19]. Leukocyte-derived EPC is an area that has been abandoned since no clear endothelial or vessel differentiation has been clearly proved in CD45-positive cells. Multiparametric labeling of cells and new sorting strategies will probably reopen this research field in the next years.

ECFCs are now commonly considered as the true vasculogenic progenitor cell population, and their culture is now consensual with a position paper from vascular biology standardization subcommittee of International Society on Thrombosis and Haemostasis [20]. ECFCs include expression of endothelial cell markers (e.g., CD31, CD144, CD146, EGFL-7, and VEGFR2) and lack of hematopoietic markers (CD45 and CD14) [2, 3]. Several mechanisms have been proposed to explain their vasculogenic effects observed in preclinical studies:

  1. 1.

    The first mechanism is a direct incorporation in host tissue to form vessels. In contrast to early EPC (or circulating/myeloid angiogenic cells), ECFC can graft in tissue or form human vessels in mice models of hind limb ischemia [5, 21].

  2. 2.

    The second mechanism proposed more recently is a paracrine effect. Firstly, ECFCs have been described as non-secreting cells compared to early EPC since they do not secrete VEGF-A and low levels of IL-8 [13]. Since this first description, we know that ECFCs have a secretion potential in inflammatory, stress, or senescence condition for IL-8 [22, 23]. Moreover, they are able to recruit inflammatory cells by secreting other CXCL or CCL molecules [24] or perivascular cells by secreting PDGF-BB so enhancing cell engraftment [25]. Moreover, ECFC has been shown to secrete microvesicle and/or exosomes that could be at the origin of their beneficial effects. This aspect of ECFC secretion of microvesicles and/or exosomes will be treated at the end of this chapter.

  3. 3.

    The third mechanism is a direct supportive effect of mesenchymal stem cells from different origin [26, 27] or perivascular cells [28]. They probably help mesenchymal stem/progenitor cells to differentiate in real perivascular cells by a Notch/jagged-1-induced mechanism [29].

  4. 4.

    The last mechanism proposed is a direct adhesive property to perivascular cells by an endoglin-dependent mechanism. Indeed, ECFC endoglin binding when MSC are co-injected in hind limb ischemia model allows a quicker engraftment of perivascular cells and an accelerated recovery [30, 31].

However, a major barrier to ECFC development as an autologous cell therapy product is their paucity in the peripheral circulation. Attempts have been made to expand ECFCs ex vivo by priming cells with growth factors [32,33,34], peptides [35, 36], platelet lysates [37], hypoxia [38], or acidosis [39]. In vivo mobilization has also been tested [40]. ECFCs have been isolated from the blood of patients with critical limb ischemia patients [41], CAD [42], acute MI [43], and pulmonary hypertension [44, 45], but their therapeutic potential has never been tested in clinical trials because of difficulties to obtain and expand them in a sufficient number of cell allowing human injection with clinically approved procedures [40]. However, the main problem for therapeutic use of ECFCs is probably addressing cells to specific area, and expansion of “adult” cells for autologous cell therapy is probably not the way to improve vasculogenesis in adult.

11.3 ECFCs Stemness and Ontogeny of Endothelial Lineage in Human

Despite the fact that ECFCs and mature endothelial cells share a similar phenotype in vitro, they show different properties in endothelial homeostasis and repair [2, 3]. Whatever their origin, mature endothelial cells do not retain the ability of revascularization when injected to preclinical models, by contrast to progenitor cells [13, 46, 47]. We and other groups have described for a long time functional difference between ECFCs and HUVECs in terms of proliferation potential and survival after growth factor activation [3, 6, 14], and several authors reported a decreased resistance to apoptosis of mature endothelial cells compared to ECFCs [48]. Moreover, ECFCs were described to be reprogrammed into induced pluripotent stem cells (iPSCs) in a better way than adult endothelial cells [49]. There is therefore a need to understand the link between ECFC stemness and their vasculogenic potential. A relative plasticity of ECFCs has been described, probably related to their stem cell/progenitor nature [50]. Membrane expression of CD34 and/or CD133 in ECFCs is a controversial field [51, 52]. One major marker of stemness of hematopoietic cells is CD34. CD34 is heterogeneously expressed on ECFC in culture, and its expression has been correlated to their vasculogenic properties [53, 54]. However, CD34 is also expressed by many mature endothelial cells in culture and has been used in human tissues to identify mature vessels. Therefore, CD34 is not a good candidate to fully explain the relationship between vasculogenic potential and stemness. Endothelial progenitors and/or angiogenic population have been also described coming from CD133+ cells [55]. However, ECFC membrane expression of CD133 is negative despite a mRNA expression that can vary along culture and passages [35, 56].

CD133, a Pentaspan membrane glycoprotein, has been used as a stem cell marker for stem cell isolation from several normal and pathological tissues, but its functional involvement is not clearly defined. In a muscle injury rat model, granulocyte colony-stimulating factor-mobilized peripheral blood CD133+ cells can differentiate into endothelial and myogenic lineages [57]. In addition, bone marrow-derived CD133+ stem cell therapy has been used in clinical trials for patients with refractory angina [58, 59], chronic total occlusion and ischemia [60], or myocardial infarction [61]. Thus, we recently described intracellular expression of the stemness marker CD133 in ECFCs [62]. CD133 gene expression inhibition abolishes ECFC vasculogenic effects in hind limb ischemia model. These findings could resume the discrepancies found in the literature concerning CD133 positivity in endothelial progenitors. Our results could explain why circulating cells expressing surface CD133 have not been correlated to ECFCs in several clinical physiological or pathological situations [40, 44, 63] and also why CD133 isolation in cord and adult blood do not give rise to ECFCs [51, 52]. A better approach would be to quantify and/or isolate ECFC by intracellular expression of CD133. Such approach has been partially evocated with cytosolic aldehyde dehydrogenase (ALDH) [64]. Our hypothesis needs to be confirmed by a prospective correlation between circulating cells expressing intracellular CD133 and ECFCs obtained in culture.

In terms of human postnatal vasculogenic stem cells, the two cell types described so far have been isolated from CD133+ populations. The first one is the stem cell isolated from proliferating phase of infantile hemangioma (Hem-SC), using anti-CD133-coated magnetic beads. These cells are scarce, representing between 0.1% and 1% of the cells from hemangioma. Hem-SCs display two essential properties of stem cells: the ability to self-renew and to undergo multilineage differentiation, including endothelial and mesenchymal lineage [65,66,67]. The second stem cells described to give rise to endothelial cells are very small embryonic-like stem cells (VSELs) [7, 68, 69]. Indeed, VSELs are pluripotent stem cells defined as lineage-negative, CD133-positive, and CD45-negative cells of small size (<6 μm in diameter) [68, 69]. VSELs have a large nucleus-to-cytoplasm ratio, and a high expression level of pluripotency core factors. They could represent a clinically relevant alternative to embryonic stem cells (ESCs) and induced pluripotent stem cells (iPS) for cell therapy since VSELs do not complete blastocyst development and do not form teratomas after transplantation into deficient mice [70, 71]. VSELs are small cells probably at the top of the stem cell hierarchy in adult tissues, able to differentiate into cells from different germ layers. We have previously described that human bone marrow (BM) VSELs isolated from patients with critical limb ischemia were able to differentiate into endothelial cells but also in perivascular cells and to foster post-ischemic revascularization in experimental model of critical limb ischemia [7, 72]. Endothelial cell differentiation has been obtained from VSELs after a step of mesenchymal phenotype as previously described in hemangioma that has been confirmed by another group [73]. VSEL’s ability to differentiate into several lineages could explain the discrepancies observed in the literature in the field of EPCs. Indeed, ECFCs have been firstly described in an elegant study of BM transplantation from Hebbel’s group [74]. However, Yoder’s group in Indianapolis proposed that these ECFCs directly come from vessels [4, 15] and from CD45-negative cells [5, 51, 52]. VSELs have a huge motility and are theoretically able to migrate from bone marrow to vessels. They have all the characteristics of potential vasculogenic CD45-negative stem cells able to give rise to vessels since they have been described to differentiate in endothelial cells by several independent groups with human, mouse, or rat VSELs [7, 75, 76]. However, further work is required to stimulate the proliferative capacity of VSELs to make them a cell therapy product. The critical expansion step should be to reverse VSEL-quiescent state and expand them. VSELs can now be expanded ex vivo in the presence of nicotinamide or valproic acid [69] or in the presence of the small-molecule UM177 [77] without transduction by DNA or RNA or by using supportive feeder cells. These first preclinical steps could lead to clinical trials with expanded cells with or without differentiation step.

Other bone marrow-derived cells have been proposed in human to give rise in endothelial cells. Mesenchymal stem cells (MSCs) have been described in some publication to be able to differentiate into endothelial cells [78] and improve neovascularization in vivo [79]. However, no clear proof of this differentiation ability has been convincing, and probably MSCs are just paracrine cells that support angiogenesis [21, 80] by secreting growth factor, modulate immunologic responses, and differentiate into perivascular cells without endothelial differentiation ability [27]. The same kind of results can be proposed for fibrocytes derived from idiopathic pulmonary fibrosis patient’s blood [81]. Discrepancy of MSC differentiation ability could come from the presence of MSC derived from VSEL. In 2002, Verfaillie’s group [10] also reported that multipotent adult progenitor cells (MAPCs) can be isolated from postnatal human bone marrow. Likewise, these data have been retracted and seem controversial. However, clear differentiation of stem cells in ECs needs to be deeply done in bone marrow cells and postnatal cells, and vasculogenic ability is resumed in Fig. 11.1.

Fig. 11.1
figure 1

Human postnatal vasculogenic stem and progenitor cells

Full size image

Since ECFC number from adult blood is reduced in contrast to cord blood [20] and has low expansion properties [5, 6], differentiation of human pluripotent stem cells into ECFC-like phenotype and characterization of endothelial differentiation pathways may allow to have enough vasculogenic and highly proliferative blood vessel-forming cells to create newly formed vessels in patients with vascular disease. A protocol to convert human-induced pluripotent stem cells (hiPSCs) or embryonic stem cells (hESCs) into ECFC-like cells has been proposed by Mervin Yoder’s group. This neuropilin-1 (NRP-1)-mediated differentiation is a promising tool; however, using hESCs or hiPSCs for regenerative medicine is still a controversial field. VSEL expansion and differentiation protocol could allow us to have an autologous cell therapy product in vascular disease. ECFC ontogeny needs also to be explored according to Peg3/PW1 [82], CD157 [83], and PROCR [84] expression to establish in human a clear hierarchy in endothelial stem and progenitor cells in human.

11.4 ECFC and Secretion of Extracellular Vesicles

More than soluble factors, the release of extracellular vesicles (EV) constitutes a new mechanism for intercellular communication [85]. Indeed, EVs may affect cell function in impaired tissues by horizontal transfer of proteins, mRNA or miRNA. EVs are distinguished on the basis of their size including respectively from small to large vesicles: exosomes, microvesicles, and apoptotic bodies. These effects of MV and exosomes that could recapitulate effects of original cells might justify skipping the use ECFCs and limiting their role as “MVs/exosomes ex vivo producers ” which would then be administered to the patient. All the findings demonstrating that benefits of progenitors or stem cells derived from MVs/exosomes can be recapitulated by their sole injection are an enthusiastic cell-​free approach that would have clinically relevant advantages in terms of cost and probably also immunological properties. One of the perfect examples in cardiovascular disease has been demonstrated in Pr Menashe’s group in a postinfarct heart failure model. Indeed, EVs from progenitor cells originated from ESCs increase cardiac function recovery in the same way than originated cells. These results support the hypothesis that a paracrine mechanism is enough to enhance chronic heart failure recovery in cell-based therapies [86]. Concerning endothelial MVs , they have been largely described as causative agents or markers of endothelial dysfunction in vascular and cardiovascular disease [87]. More than a biomarker, these endothelial MVs have been described to have several biological effects with both protective and deleterious effects. Indeed, endothelial MVs have been described as anticoagulant, profibrinolytic, and proangiogenic agents, while at the same time, they can act as procoagulant/thrombotic/inflammatory agents as well as antiangiogenic properties depending of technical condition for isolation and also depending on patient condition [88]. We will focus here on preclinical data about MVs and exosomes from ECFCs.

11.4.1 Microvesicles from ECFC

ECFCs can be considered as a liquid biopsy of in situ endothelium (see Chap. 4). Thus, MVs obtained from adult peripheral blood-derived ECFCs could also be a valuable tool to understand vascular disease. Indeed, Pr Dignat-George’s group demonstrated that ECFC expanded from cord blood of healthy neonates has angiogenic potential in vitro and promotes blood flow recovery in mouse hind limb ischemia model. However, when obtained from senescent ECFC isolated form low birth weight neonates, the EVs collected in similar culture conditions are able to alter endothelial cell homeostasis in vitro and transfer senescence in target cells [89]. In addition, we recently demonstrated that MVs from ECFC isolated from patients with idiopathic pulmonary fibrosis (IPF) induce profibrotic effects, have higher fibrinolytic properties than controls, and could correlate with IPF severity [90]. Based on these two observations, we can observe that, according to the pathological environment associated to the donor health status, the properties of ECFC in terms of angiogenic or tissue repair potential can be significantly affected. We anticipate that EVs from patient-derived ECFC can have some deleterious effect on vascular system that has to be documented and reversed before their use in an autologous context. More than that, MVs from disease models could help to understand vascular pathophysiology.

As previously described, ECFC may act as cell therapy product in some vascular disease. However, there are several situations with a very low level of ECFC engraftment and injection of ECFC-conditioned media, and in particular, MVs are probably enough to have vascular repair in retinal ischemia [91] or bronchopulmonary dysplasia [92].

11.4.2 Exosomes from ECFCs

Exosomes are produced in the endosomal compartment of most eukaryotic cells and have a size between 40 and 100 nm of diameter. These are very small membrane-bound vesicles that are over produced by most proliferating cell types during normal and pathological states. Their levels have been proposed as diagnostic and therapeutic tools in cardiovascular disease or cancer. Exosomes contain a large variety of RNAs and noncoding RNAs (ncRNAs) including miRNAs and long noncoding RNAs but also proteins and lipids, which are representative to their cellular origin and shuttle from a donor cell to a recipient cell. Since 2011 [93], ECFCs have been described to be able to secrete exosomes. Their functions and secretion specificity are specific of cell and clinical situation. Their roles are different from MVs. Indeed, in senescence associated to premature neonates [89], ECFC has a higher number of MVs, while exosomes are secreted at the same level. In ECFC efficacy, as previously described here, ECFC improves vascular function, and their conditioned media are enough to have beneficial effects. In acute kidney injury (AKI), intravenous injection of ECFC significantly attenuated increases in plasma creatinine, tubular necrosis, macrophage infiltration, oxidative stress, and apoptosis, without cell engraftment in the kidneys [94]. Total conditioned media and exosomes from ECFC reproduce the same effect than original ECFC, while MVs do not have any beneficial effects [94]. ECFC exosomes reduce AKI via transferring miR-486-5p-targeting PTEN and with a CXCR-4/SFD-1 mobilization that helps in exosome uptake [95, 96]. ECFC-derived exosomes have also been described to reduce cardiac fibroblast activation [97] and restore blood-brain barrier continuity in mice after a traumatic brain injury [98].

11.5 ECFC as a Liquid Biopsy to Understand Vascular Disease?

Recently growing interest has been reported for “noninvasive” liquid biopsy as a valuable source for molecular profiling and/or cellular pathophysiology. A biomarker and/or composition of biomarkers capable of detecting vascular abnormalities could help to characterize or deeply understand vascular diseases. ECFCs, after being described in early 2000 from bone marrow [74], have been proposed originating from vascular endothelium [15] and, in particular, lung vessel wall [99]. Since ECFCs have been described to come from vessel wall, they could be considered as a potential liquid biopsy, in particular, in pulmonary vascular disorders. In this chapter, we focus on ECFC involvement in lung diseases and also in haemostasis disorders.

11.5.1 ECFC in Lung Diseases

11.5.1.1 Pulmonary Arterial Hypertension (PAH)

Our team was particularly interested in the endothelial compartment circulating in pulmonary hypertension. It is a rare, rapidly fatal disease in the absence of treatment, characterized by pulmonary vascular obstruction leading to a progressive increase in resistance to blood flow and ultimately to right heart failure. It is defined as mean pulmonary arterial pressure (mPAP) >25 mmHg at rest or >30 mmHg at exercise, measured during right heart catheterization, which is the gold standard for diagnosing pulmonary hypertension [100]. Endothelial dysfunction plays a key role in the development of pulmonary hypertension [101]. This disease is characterized by pulmonary vascular remodeling of small arteries and precapillary arterioles. Plexiform lesions, an uncontrolled proliferation of endothelial cells, are characteristic of idiopathic PAHs, and an antiapoptotic profile of endothelial cells has been described in the irreversible forms of PAH secondary to congenital heart disease [102]. In order to evaluate the involvement of the circulating endothelial compartment during PAH, we quantified ECFCs but also circulating endothelial cells (CECs) in pediatric PAH. We found strong correlation between CECs and remodeling process or vasodilator therapy efficacy [103,104,105], while we never found any modification of ECFC level in PAH compared to controls or reversible PAH. However, we demonstrated that prostacyclin analog treprostinil increases ECFC level [104] by a VEGF-A-dependent mechanism [106]. Thus, we do not think that ECFCs are a good biomarker of vascular remodeling process in PAH, and conflicting results in literature probably paved the way of absence of quantitative relationship between ECFC and vessel pathophysiology [107]. However, functional properties of ECFCs are related to PAH vascular function. Indeed, ECFC proliferation has been correlated to disease severity [108] suggesting a link between their function and clinical settings. Moreover, hereditary PAH can be the consequence of bone morphogenetic protein receptor type 2 gene (BMPRII) mutation. ECFC from PAH patients with BMPRII mutations has been shown to have a defective ability to form vessels in vitro [45]. Studies on ECFC from patients with BMPRII allowed finding new therapeutic targets like translationally controlled tumor protein (TCTP) [109], miR-124 [110], chloride intracellular channel 4 (CLICL4) [111], or interferon type 1 [112] but also proposed new treatment like chloroquine [113] or BMP9 [114].

11.5.1.2 Chronic Obstructive Pulmonary Disease (COPD)

COPD is associated with chronic airway inflammation including chronic bronchitis and emphysema, characterized by alveolar loss and enlargement. Cigarette smoke is in general at the origin of COPD. The contribution of the endothelial progenitor to COPD pathogenesis is not been fully understood because of the diversity of cells phenotype quantified [115,116,117,118]. We recently demonstrated that VSELs (described to be at the origin of endothelial lineage; see earlier in this chapter) were mobilized in COPD patients with PaO2 under 92% [119]. Concerning ECFC, their number seem decreased in most chronic pulmonary diseases [120], but more interesting is their abnormal function in terms of adhesion [121] or senescence [122]. Indeed, Pr Anna Randi’s group evidenced that ECFC from smokers and COPD patients had an accelerated aging due to epigenetic dysfunction [122]. They also found that miR-126-3p is downregulated in ECFCs isolated from COPD patients and promotes increased DNA damage at the origin of endothelial dysfunction [123].

11.5.1.3 Idiopathic Pulmonary Fibrosis (IPF)

Idiopathic pulmonary fibrosis (IPF) is a devastating disease characterized by obliteration of alveolar architecture, resulting in declining lung function and ultimately death. We and others have previously demonstrated that EPCs [63, 124, 125] and, in particular, ECFCs are downregulated in stable IPF but increased significantly in patients with impaired gas transfer [diffusing capacity of the lung for the carbon monoxide (DLco) <40%] [63]. ECFCs from IPF patients could participate to vascular remodeling in fibrotic lung diseases by a direct vasculogenic effect but also by cooperating with fibrocytes, a cell type well known to contribute to organ fibrosis [81]. Cell therapy approaches have been proposed in several chronic lung diseases; thus, we tested ECFC injection in fibrogenesis induced by bleomycin in nude mice. Mice were injected with ECFCs isolated from cord blood or IPF patients. We assessed morbidity, weight variation, collagen deposition, lung imaging by microCT, Fulton score, and microvascular density. No modulation of fibrosis or vascular density during fibrogenesis or when fibrosis was constituted was observed with ECFCs whatever their origin [126]. We then postulated that ECFCs might behave as a liquid biopsy in IPF patients, and we have demonstrated that senescent and apoptotic states were increased in ECFCs from IPF patients as shown by galactosidase staining, p16 expression, and annexin V staining but also increased interleukin-8 secretion [22]. We showed that IL-8 secretion from ECFCs of IPF patients induced migration of neutrophils in vitro and in vivo in a matrigel implant model in nude mice. To check clinical relevance of these results, we showed an infiltration by neutrophils in IPF lung biopsies, and we found, in a prospective clinical study, a higher level of neutrophils in peripheral blood in IPF patients with a poor prognosis [22]. Finally, we also demonstrated that microparticles released from ECFC isolated from IPF patients compared to controls had an increased plasminogen activation and could stimulate fibroblast migration [90], suggesting involvement of ECFC-derived endothelial microparticles to pulmonary fibrogenesis. Altogether, our results are in favor of a real correlation of ECFCs in IPF with the phenotype observed in lungs, while no active process by injecting them in mice is observed [127]. So, ECFCs in IPF are clearly a “liquid biopsy” of vessels.

11.5.2 ECFC in Haemostasis and Thrombosis

11.5.2.1 von Willebrand Disease (VWD)

von Willebrand factor (vWF) is a glycoprotein, highly involved mainly in haemostasis and bleeding disorders, produced uniquely by endothelial cells and megakaryocytes. Its quantification or labeling is routinely used in pathology lab to identify vessels. vWF is increased during angiogenesis since fibroblast growth factor-2 and vascular endothelial growth factor have been shown to increase its expression in a variety of endothelial cells [128]. vWF has been used to affirm endothelial origin of ECFCs and more recently explored mainly by Pr Anna Randi’s group to explore endothelial dysfunction in patients with VWD. Congenital forms of VWD are due to mutations in the vWF gene. There are three subtypes: types 1 and 3 with quantitative defects (respectively partial and total for types 1 and 3) and type 2 with qualitative defects. ECFC culture in patients with VWD evidenced involvement of vWF in angiogenesis processes [129, 130]. Indeed, significant enhancement of in vitro tube formation, proliferation, and migration was observed in ECFCs isolated from VWD patients [131], with a variability in different subtypes of VWD [132, 133]. These studies in VWD are also pointing a role of ECFC as potential liquid biopsy since they allowed studying vascular function of different defect in vWF that could explain angiogenic disorders like angiodysplasia in type 2 patients. ECFC in VWD could also help to understand physiology of vWF storage in endothelial cells and help to decipher between desmopressin (which induces release of vWF from endothelial cells) or replacement therapy in VWD treatment. Finally, gene therapy of VWD type 3 has been proposed by using ECFCs as a cellular vehicle for therapy [134]. These approaches also proposed to cure hemophilia (factor VIII) or anemia (erythropoietin ) are still at a very preliminary step for a potential clinical application [135, 136].

11.5.2.2 Myeloproliferative Neoplasms (MPN)

Initial studies in patients with Philadelphia chromosome-positive or with Janus kinase 2 V617F (JAK2V617F) mutation in MPN were found only in hematopoietic cells and not in ECFC [5, 137, 138]. We recently confirmed this result with ECFCs from a patient double mutated with BCR-ABL1 and JAK2V617F mutations that lack these two mutations [139]. However, Teofili et al. found in a subgroup JAK2V617F mutated patients with thrombotic events that ECFC could express the mutation [140]. Another group also found a regulation of ECFCs in subpopulation of patient with thrombotic disorder in nonactive MPN [141]. Thus, as found in lung disorders and VWD previously, probably in MPN, ECFCs could be marker of a “clinical vascular phenotype” and be the reflection of prothrombotic state or disease evolution from a hematopoietic disorder to a systemic disorder. Another consequence of these studies is about stem cell at the origin of hematopoietic and vascular lineage. Probably mutation present at the origin only affects hematopoietic stem cells, while after a disease evolution perhaps, mutation – i.e., or at least JAK2V617F – can appear on putative hemangioblast or potentially on VSELs at the origin of both lineages. Hypothesis of the prothrombotic phenotype of endothelial cells that could acquire JAK2V617F mutation has been described by two French groups [142, 143]. Indeed, both groups, with different in vitro models, found an increased expression of P-selectin associated with prothrombotic, proinflammatory, and proadhesive phenotype of endothelial cells .

11.5.2.3 Hereditary Hemorrhagic Telangiectasia (HHT)

Mutations in endoglin or activin receptor-like kinase-1 (ACVRL1/ALK1) genes can give rise to hereditary hemorrhagic telangiectasia (HHT). These patients present epistaxis, telangiectases, and arteriovenous malformations in the lung, brain, or liver. ECFC from HHT patients had impaired angiogenesis potentially related to decreased endoglin expression [144, 145]. We explored in our group endoglin involvement in ECFC angiogenic potential and found that endoglin is an adhesive molecule necessary for vessel stabilization and ECFC regenerative potential [30, 31, 146].

11.5.2.4 Venous Thromboembolism (VTE) Disease

Thrombus resolution has been described to be related to endothelial progenitor mobilization [147, 148]. ECFCs have been also involved in fibrin infiltration in vitro, and thrombin could be a chemoattractant for ECFC [149]. Indeed, ECFC has a whole panel of hemostatic receptor and molecule including inducible tissue factor [150], fibrinolytic properties [149], or thrombospondin-1, largely described as a platelet aggregation stabilizer and an active partner for vWF [151, 152]. We also described that ECFC expresses thrombin receptor PAR-1 [6, 23, 24, 35, 36, 149]. Indeed, thrombin involvement is essential in thrombosis. Activation of its main receptor PAR-1 by thrombin- or PAR-1-activating peptide (which mimics the effect thrombin on its PAR-1 receptor without causing cleavage) on cord blood ECFC favored all stages of angiogenesis that are the cell proliferation, migration, and differentiation [6, 23, 24, 35, 36, 149]. Thrombin has also been described to induce endothelial differentiation from bone marrow mononuclear cells [153]. PAR-1 activation on ECFCs improves angiogenesis in vitro by activating the pathways of angiopoietin 2, which promotes cell proliferation, and that of SDF-1/CXCR4, which promotes differentiation. In addition, the modulation of the PAR-1 on ECFCs by an autologous fibrin network that can constitute a matrix enables ECFCs to acquire anticoagulant and antifibrinolytic properties in addition to their angiogenic properties [149]. It is recognized that leukocytes play an important role in this recanalization of the thrombus. Thus, thrombin, which is also a proinflammatory factor, and interleukin-8 (IL-8) have angiogenic properties by interaction with its CXCR1 and CXCR2 receptors. Indeed, IL-8 is strongly expressed in early EPCs, and its level in the conditioned medium is unchanged after activation of PAR-1. On the other hand, the secretion of IL-8 by ECFC, very weak under basal conditions, is strongly increased after activation of the PAR-1 and induces the Boyden chamber migration of the early EPCs. These results suggest that thrombin receptor activation on these ECFCs allows cooperation between different angiogenic cells during paracrine-mediated neovascularization. We further explain how PAR-1 activation in ECFC could interact with inflammatory cells and demonstrated a large panel of chemotactic gene expression increase and an effect of ECFC-conditioned media on leukocyte recruitment at ischemic sites through a COX-2-dependent mechanism [24]. We finally showed that thrombin PAR-1-activating peptide was able to increase inflammatory cell recruitment on other postnatal vasculogenic stem cell. Indeed, activation of infantile hemangioma tumor-derived stem cells (HemSCs) incubated with PAR-1-ap increased leukocyte recruitment in a Matrigel(®) implant model in nude mice [24]. In addition, ECFCs obtained from healthy volunteers are able to express TF following stimuli such as TNF-α [150, 154]. ECFC stimulation by TNF-α also induces the generation of procoagulant microparticles [150], which may represent an angiogenic but also thrombogenic vector. Hubert et al. demonstrate in a well-recognized cremaster arterial laser-induced injury model of thrombosis that neutrophils present at the site of thrombus can recruit ECFCs [155]. Involvement of ECFC in VTE has been demonstrated by the group of Dr. Alvalrado-Moreno [156, 157]. In recurrent and unprovoked VTE, they described dysfunctional ECFC and proposed an association with these defects and the risk of thrombotic events. One of their findings is an increased proinflammatory cytokine secretion of ECFC from VTE patients [156]. Further studies are required to determine whether dysfunctional ECFCs are involved in thrombus formation or recanalization. However, in this VTE situation, we have one more time an ECFC phenotype different from control that could be associated with endothelial dysfunction found in thrombotic patients .

11.6 Conclusion

After several years of inconsistence and ambiguity about endothelial progenitor cell definition, ECFCs have been largely described as the progenitor cell linked to clear vasculogenic potential and vascular diseases, and now, standardization protocols are on the way [20]. Ontogeny of these cells needs a lot of effort yet to determine which stem cell is at the origin of endothelial lineage in human adult [68]. New technologies allowing multiparametric evaluation of heterogeneous cell populations could probably help us to decipher different cell population and explore new marker of vascular stem cells (Peg3/PW1, PROCR, or CD157) in human vasculogenesis. ECFC involvement as a liquid biopsy is now very clear, and we have with ECFC probably the perfect tool to explore vascular disease and find new therapeutical targets or at least demonstrate pathophysiology of vessel abnormalities. Several clinical applications have been proposed as a cell therapy product, a vector for gene therapy, or a tissue-engineering strategy. Recent research about hemocompatibility of biomaterials, in particular, with development of a bioprosthetic total artificial heart [158,159,160], could open a new area of biomaterial bioengineering with ECFCs and/or endothelial stem cells.