D e novo vasculogenesis in the heart *

The formation of the embryonic heart vasculature is a complex process and is the result of vasculogenic, angiogenic and arteriogenic mechanisms, involving cells from distinct origins. In the neonate and the adult, several sources of endothelial precursor cells (EPCs) have been identified that contribute to physiological and pathological vascularization, consistent with the concept of de novo vasculogenesis after birth, including in the heart. The existence of EPCs in the adult has offered the possibility to use these cells for revascularization of ischemic tissues. An important challenge for vascular regeneration in ischemic and post-infarction patients is now to identify the most adequate cell source and cell dose for sufficient revascularization. This review gives an overview of the cellular and molecular cues involved in the formation of the heart vasculature before and after birth and discusses some of the recent insights and outstanding questions on EPCs and other vascular progenitors—both from a biological and therapeutic perspective.  2003 European Society of Cardiology. Published by Elsevier Science B.V. All rights reserved.

. The developing mouse heart and its vasculature. Schematic representation of heart and coronary development in the mouse. The primitive heart tube consists of an endothelium (the endocardium) lining an avascular myocardium that is supplied with oxygen by diffusion. As the myocardium thickens, large invaginations are formed ('trabeculae') in order to increase diffusion capacity. However, shortly after heart looping, the epicardium (originating from the proepicardial organ that attaches to the heart) proliferates and migrates over the heart, closely followed by (hem)angioblasts that form a primitive vasculature (vasculogenesis). During heart chamber formation, maturation and septation, this primitive vasculature expands by sprouting of new vessels from pre-existing ones (angiogenesis) and is transformed into a more organized network with smaller and larger vessels. At this time, the vascular network is still not connected to the aorta. Once the network connects to the aorta (around E13 in the mouse), the vessels become invested with a smooth muscle cell (SMC) coat (arteriogenesis) and a fibroblast-rich adventitia. After birth, as the myocardium further thickens, capillary density increases three-to fourfold and the number of SMC-covered coronaries at least 10-fold during the first 3 weeks. EC: endothelial cell. cardiomyocyte hypertrophy. The increased metabolic de-2 . Origins and heterogeneity of vascular cells in the mands of hypertrophying cardiomyocytes are compensated developing heart by a significant expansion of the myocardial vasculature. In the mouse, capillary density increases three-to fourfold Much of what is known about the origin of vascular and the number of SMC-covered coronaries at least 10-cells in the heart is derived from retroviral labeling and fold during the first 3 weeks [2] (Fig. 1). These vessels quail-chick chimera experiments. An important cell source, undergo significant remodeling in order to assume adult the proepicardial organ, is located between the sinus branching characteristics and acquire specific properties of venosus and the primordial liver. Around E8.5-9 in the coronary arteries or veins. Until now, it has been post-mouse, proepicardial cells migrate and form patches of ulated that postnatal expansion of the myocardial vascular epithelial-like cells that fuse to provide a continuous sheath bed only proceeds through angiogenesis. However, recent enveloping the heart, the epicardium (Fig. 1). A subpopufindings suggest that bone marrow (BM)-derived endo-lation of epicardial cells, the epicardium-derived cells thelial precursor cells (EPCs) contribute to (physiologic) (EPDCs) undergoes an epithelial-mesenchymal transition, neovascularization in the heart and the liver during the delaminates from the growing epicardium and migrates neonatal period, consistent with a vasculogenic paradigm into the myocardium giving rise to at least three different [3].
cell types: SMCs, coronary and intermyocardial fibroblasts [4]. Endocardial cells have embryonic origins distinct from revealed that neural crest-derived SMCs behave differently coronary vessels. These cells arise from within the myocar-from mesoderm-derived cells, as evidenced by their disdial plate, adjacent to the foregut endoderm, where they tinct reaction pattern to transforming growth factor (TGF)assemble into a loose vascular plexus that eventually forms b [6]. a single endocardial tube [5].
Coronary SMCs have different origins, depending on their location in the heart and their function. While SMCs 3 . Molecular control of vessel formation in the of the proximal coronary arteries (as well as the large developing heart arteries in the thorax, head and neck) originate from the neural crest, the rest of the coronary arteries derive their In addition to identifying the cellular origins of vascular smooth muscle coat from the proepicardial organ [6] (Fig. cells, defining the molecules that determine cell fate and 1). In contrast, coronary vein medial cells are from atrial function is of invaluable importance for vascular therapy. cardiomyocyte origin [7]. It has been shown that the first Studies in the chick, the zebrafish and gene-deficient mice layers of SMCs in the dorsal aorta transdifferentiate from have significantly contributed to our current understanding the endothelium, but there is no evidence as yet that such a of molecular signaling during vessel development. The transdifferentiation mechanism contributes to the develop-general molecular cues involved in vessel formation have ment of the coronary vasculature [6]. SMC markers, like been reviewed previously [15] and many of these mole-SMC a-actin, become only expressed in the coronary cules may also drive vessel growth and remodeling in the vasculature upon connection of the coronary plexus with heart. In addition, the formation of a specialized vascular the circulation, indicating that-unlike vasculogenesis-network fine-tuned to a complex organ such as the heart 'muscularization' of coronary vessels is flow-dependent may require specific angiogenic signals. Recently, EG-( Fig. 1). Coronary veins develop before arteries, however, VEGF (endocrine gland-derived vascular endothelial the coronary veins become enveloped with a media only at growth factor) was described, an angiogenic factor only later time points than coronary arteries. Presumably, the affecting endothelial cells in endocrine glands [16]. This pressure increase and blood flow alteration after con-was the first organ-specific angiogenic factor to be denection to the aorta, is a trigger for SMC differentiation in scribed and it is likely that others exist, for instance in the coronary arteries, while differentiation is delayed in the heart. low-pressure environment in coronary veins [7].
The formation of the endocardial tube is one of the first Endothelial cells in vascular beds of different organs steps in cardiac development. Endocardial cells first aggreacquire specialized characteristics to allow optimal func-gate into a loose vascular network that subsequently tion in that specific organ. For example, endothelial cells in coalesces into progressively large tubes, eventually the brain are tightly linked to each other and are sur-generating a single endocardial tube lining the avascular rounded by numerous peri-endothelial cells, which consti-myocardium [5]. VEGF secreted from the adjacent endotute a barrier that protects brain cells from potentially toxic derm appears to play a role in directing endocardial blood-derived molecules. In contrast, vessels in endocrine development [17]. Subsequent trabeculation of the venglands are leaky and their endothelial cells have fenestra-tricular myocardium expands the endocardial endothelial tions, allowing hormone trafficking. In addition to endo-surface area in order to increase diffusion capacity for thelial cell heterogeneity between distinct organs, endo-oxygen and nutrients prior to development of the thelial cells within the same organ can be heterogeneous.
coronaries. This process requires reciprocal interactions In the heart, the heterogeneity of endothelial cells in between the endocardium and the myocardium. distinct locations of the heart vascular tree has been Neuregulins, a family of secreted and cell-membranesuggested by differences in expression patterns for several associated factors generated by alternative splicing of a molecules, like the endothelial constitutive nitric oxide single gene, have been implicated as potential endocardialsynthase (NOS) isoform (ecNOS) [8], brain derived neuro-derived signals in trabeculation. Neuregulins interact with trophic factor (BDNF) [9]or adhesion molecules [10,11]. erbB receptors of the epidermal growth factor family, some In addition, functional differences have been observed of which are expressed in the myocardium (erbB2 and between micro-and macrovascular endothelial cells, iso-erbB4). Targeted deletion of neuregulin, erbB2 or erbB4 lated from human hearts [12]. This heterogeneity within all block trabeculation and deficient embryos die early in cardiac endothelial cells was shown to play a critical role gestation [18][19][20]. Deficiency of RXR-a, a co-receptor for in inducing conduction cell differentiation in the em-retinoid acid receptors, results in defective ventricular bryonic myocardium [13]. In the chick embryo, cells from maturation and trabeculation [21,22]. Formation of the distal Purkinje fibers differentiate in close association trabeculae was also impaired in mice lacking VEGF with coronary arteries, through a neighboring vascular [23,24], angiopoietin-1 (Ang-1) [25] or the basic helixsignal, likely endothelin [14]. Like cardiac endothelial loop-helix transcription factor dHAND (deciduum, heart, cells, heart SMCs seem heterogeneous. Culture studies autonomic nervous system, neural crest derived) [26].
As mentioned above, the generation of coronary vessels tion of a single allele of VEGF-A results in early emis integrally associated with the development of the bryonic lethality due to severe vascular defects including epicardium. Migration of epicardial epithelium is blocked defective trabeculation (see above), precluding study of its in vascular cellular adhesion molecule (VCAM)-1 deficient role at later stages, like during coronary vessel developmice and these mice therefore do not develop coronary ment [23,24]. In addition, different splice variants (isovessels [27]. In addition, adhesion between epicardium and forms) of VEGF-A exist (VEGF , VEGF and 120 164 myocardium is severely disrupted in a integrin-deficient VEGF being the most prominent ones in the mouse), 4 188 mice, resulting in the lack of proper coronary circulation. with possible differential functions due to their distinct A recent knock-in study revealed that a b integrin-me-chemotactic properties, receptor specificity, endothelial 4 1 diated adhesion is necessary for migration of progenitor mitogenicity and tissue-specific expression (reviewed in cells to form the epicardium (Ref. [28] and references Ref. [35]). Therefore, we generated mice expressing the 120 / 120 therein). Coronary vessel development in the embryo not VEGF isoform alone (VEGF mice) using the 120 only depends on the epicardium but is now known to be Cre / loxP system to remove exons 6 and 7, encoding basic coordinated by signals derived from at least four distinct domains only present in VEGF and / or VEGF cells types: endothelial, myocardial, (pro)epicardial and fraction of homozygous VEGF embryos died shortly 120 neural crest cells [29]. An important family of regulators is after birth and the remainder gained less weight and died the GATA family of transcription factors. The members of before day 14 after birth of cardiac failure, exhibiting the GATA-4 /5/6 subfamily are expressed principally in depressed myocardial contractility and cardiac dilation. 120 / 120 the heart and gastrointestinal tract with overlapping expres-Analysis of VEGF hearts revealed severe angiogenic sion patterns. Their transcriptional activities are regulated defects in the myocardium, suggesting that VEGF by 120 by physical association with other transcription factors. itself is insufficient for normal blood vessel growth. In One of these factors is FOG-2 (friend of GATA-2). Mice contrast to wild type mice, capillary and coronary artery deficient in FOG-2 have an intact epicardium, but com-density did not increase during the first weeks of life. As pletely lack coronary vessels [30]. Since mice harboring a mentioned above, the expansion of the myocardial vessel single amino acid mutation in GATA-4 preventing the network is not only due to angiogenesis (involving mature FOG / GATA-4 interaction also feature lack of a coronary endothelial cells), but also to vasculogenesis (involving vasculature, this interaction is likely involved in coronary EPCs) and VEGF was shown to modulate EPC engraft-2 / 2 vessel development [31]. However, unlike FOG-2 ment in the neonatal heart [3]. The fact that the formation mice, the GATA-4 mutant mice show impaired outflow of SMC-surrounded coronary arteries was reduced in 120 / 120 tract septation, resulting in mice with double-outlet right VEGF mice suggests a VEGF-mediated effect on ventricle. This suggests that for the latter event distinct SMCs, either directly [36,37] or indirectly via an effect on mechanisms or interactions other than with FOG-2 are the production of recruitment factors such as plateletinvolved. Fibroblast growth factors (FGFs) have been derived growth factor-B (PDGF-B). Expression levels of associated with embryonic vasculogenesis by determining PDGF-B and its receptor PDGF-receptor b were indeed 120 / 120 angioblast specification from the mesoderm (reviewed in reduced in VEGF hearts [2]. A recent study revealed Ref. [32]). In addition, the spatiotemporal expression that cardiomyocyte-derived VEGF is an important signal pattern of FGF-1 and its receptors during cardiac mor-for normal development of the coronary microvasculature phogenesis correlate with proliferation and differentiation [38]. of endothelial cells and SMCs of the coronary vessels, as The role of VEGF-A homologues has generally been well as of cardiomyocytes [33]. In culture studies, FGFs, less extensively evaluated. PlGF-deficient mice survived particularly FGF-1, -2 and -7, positively regulate the beyond birth with no obvious vascular defects in the heart epicardial-mesenchymal transformation that preceeds for-or other vital organs [39], suggesting that this factor is mation of coronaries. FGF-2 was shown to act both in an redundant for physiological vessel development. Although autocrine (from the epicardium) and a paracrine (from the the most prominent expression of VEGF-B in the developmyocardium) way. Myocardium-derived TGF-b3 inhibits ing myocardium [40] suggested a role in vascularization of this transformation process [29]. A role for FGF-1 later in the embryonic heart, mice lacking VEGF-B were viable life was indicated in mice overexpressing this factor in the and had no gross abnormalities in the heart, except for an heart. The myocardium of these mice showed an increased atrial conduction defect [41] and vascular dysfunction after density of SMC a-actin-positive coronary arteries and an coronary occlusion [42]. For VEGF-C and the closely increased number of coronary branches [34].
related VEGF-D, no specific functions in the development VEGF is a well-studied angiogenic factor involved in of the murine heart vasculature have been described. vasculogenesis as well as angiogenesis (reviewed in Ref. Expression levels of VEGF-C were reduced in the hearts 120 / 120 [15]). The mammalian VEGF family comprises at least five of VEGF mice, however, it is unknown whether this homologues (VEGF-A, -B, -C, -D, and placental growth directly contributed to the coronary vascular defects in factor or PlGF) with different receptor specificities. Dele-these mice. One recent report, using a quail heart explant model, suggests a role for both VEGF-A, -B and -C in like in the ischemic myocardium [50-52]-a process coronary vessel formation [43].
termed postnatal vasculogenesis (Fig. 2). In addition, the Only a few factors specifically important in heart vessel in situ development of large collateral vessels from predevelopment have been described. This may be-in part-existing arteriolar anastomoses (collateral growth or 'adapdue to the fact that many studies in knock-out mice have tive arteriogenesis'; Fig. 2) may not result from ischemia, not included an extensive investigation of the fetal heart but rather from shear-stress-induced upregulation of anphenotype in these animals. Bves (blood vessel / epicardial giogenic and inflammatory factors (reviewed in Ref. [53]). substance) is an adhesion molecule expressed in the Although the myocardial vascular bed can expand, its proepicardial organ, the migrating epicardium, the delami-expansion capacity is limited and soon the heart becomes nated vasculogenic mesenchyme and coronary SMCs [44].
ischemic due to an imbalance between oxygen supply and It was suggested that this factor regulates coronary vessel consumption. Initially, the myocardium develops a protecdevelopment through a dynamic subcellular redistribution tive response-hibernation-in order to preserve highmechanism involved in cell migration and adhesion. energy metabolites at the expense of contractile dysfunc-Antibodies against Bves were able to block epicardial tion. The hibernating myocardium is still viable and able to sheet migration in proepicardial organ explant cultures restore its contractile function upon proper revasculariza- [45]. Secondly, capsulin (also known as epicardin or pod-tion. However, when the ischemic insult becomes too 1) is a basic helix-loop-helix factor highly expressed in severe, the hibernating myocardium may undergo irreversthe proepicarcial organ and expression is apparent immedi-ible structural changes and die to become replaced by 120 / 120 ately before mature coronary vessels develop [46][47][48]. fibrotic scar tissue. VEGF mice featuring impaired Based on this spatio-temporal expression pattern, this myocardial angiogenesis developed ischemia with signs of molecule has been proposed to predict coronary vascular hibernation, that over time, progressed to cardiac failure SMC fate, however direct evidence for such a role from [2]. This mouse model provided genetic evidence that knock-out studies is lacking. Differentiation of coronary insufficient availability of an angiogenic growth factor SMCs from proepicardial cells involves serum response results in ischemic heart disease and constitutes a rationale factor (SRF), a member of the MADS box family of DNA for administering angiogenic factors or their corresponding binding proteins. Rather than affecting epicardial-mesen-genes ('therapeutic angiogenesis'; Fig. 2) to rescue hiberchymal transformation, this factor regulates transformation nating myocardium or to prevent myocardial necrosis and of mesenchyme to SMCs through RhoA-Rho kinase cardiac failure. Therapeutic angiogenesis might also be signaling, thereby mediating cytoskeletal changes [49].
indicated for post-infarct patients to prevent ischemia in BDNF, a member of the neurotrophin family, and its the remodeling, viable myocardium, remote to an infarcted receptor trkB are selectively expressed by neonatal and region. In addition, neovascularization of the infarct area is adult heart vessels. BDNF was suggested as a specific essential for a normal healing process after myocardial survival factor for coronary endothelial cells. Deficiency of infarction [54]. Therapeutic angiogenesis might be used BDNF in the mouse impaired survival of endothelial cells alone or in combination with coronary bypass surgery. in myocardial arteries and capillaries in the early postnatal However, insufficient availability of angiogenic growth period, resulting in intraventricular wall hemorrhage, de-factors might not be the only reason for patients to develop creased cardiac contractility and early postnatal death [9]. myocardial ischemia. Several conditions (aging, diabetes and hypercholesterolemia) in which the adult endothelium becomes less responsive to angiogenic stimuli have been 4 . The adult heart and its vasculature identified in mice and these conditions are also present in a large group of patients that suffer from myocardial is-In contrast to the active vessel growth in the embryo and chemia. The impaired endothelial function likely contrithe newborn, the adult myocardial endothelium remains in butes to the increased severity of cardiovascular disease in a quiescent state. Only when provoked by stress or the geriatric population [55]. In addition, dysfunction of pathologic conditions, does the coronary vascular bed EPCs has been suggested to occur in diabetic mice [56] expand. Vascular expansion in the adult heart may en-and an inverse correlation between the number of carcompass three different mechanisms, possibly driven by diovascular risk factors (including diabetes, hypertension) distinct signals. Until recently, it was generally accepted and the number and migratory activity of EPCs has been that vessels in adult ischemic tissues could only grow by measured in humans [57]. Dysfunction of mature endoangiogenic mechanisms (Fig. 2), i.e., the sprouting of thelium and EPCs is a possible indication for vascular mature endothelial cells from pre-existing vessels, likely in regeneration with EPCs (Fig. 2). EPC transplantation response to angiogenic growth factors. However, recent might complement current strategies of therapeutic angiostudies have revealed that EPCs also circulate postnatally genesis for patients in whom endogenous endothelial in the peripheral blood and may be recruited from the BM (precursor) cells fail to sufficiently respond to growth and incorporated into sites of active neovascularization, factor treatment.  collateral vessels [61]. Orlic et al. used the Lin c-Kit EPCs were initially isolated from peripheral blood and fraction from murine bone marrow for transplantation into identified on the basis of their expression of VEGFR-2 and the peri-infarct region. Transplanted cells gave rise to CD34, antigens shared by the angioblast and the hemato-endothelial cells in the regenerating myocardium [62]. poietic progenitor [58]. These EPCs were subsequently Transplantation of freshly obtained granulocyte-colony 1 shown to express VE-cadherin and AC133, an orphan stimulating factor (G-CSF)-mobilized human CD34 receptor which is specifically expressed on EPCs but blood cells into nude rats generated an angiogenic miwhose expression is lost once they differentiate into mature croenvironment by incorporation into the infarct bed endothelial cells [59]. Their high proliferation rate dis-(vasculogenesis) and by stimulating proliferation of pretinguishes circulating marrow-derived EPCs in the adult existing endothelium in the infarct border (angiogenesis) 2 from mature endothelial cells shed from the vessel wall [63]. Also a subpopulation of the CD34 fraction 2 / low 1 1 [60].

. Cellular sources for vascular regeneration
(CD34 c-Kit Sca-1 ; the so-called 'side population') Cells with the capacity to form endothelial cells in vitro of murine bone marrow was shown to contain cells with or in vivo have been identified in different cell fractions the potential to incorporate into new vessels in the periisolated from the BM or the peripheral blood of adult infarct region [64]. In some cases unfractionated bone animals and humans (Table 1). Some of these cell popula-marrow was used for transplantation. Delivery of freshly tions were shown to be effective in the formation of new aspirated whole bone marrow increased vascularity and vessels in ischemic tissues, including the myocardium.
perfusion in the ischemic myocardium in pigs [65] and rats  isolated cells were expanded ex vivo before transplantation CD34 CD45 cKit GlyA fraction of BM from mice, rats by culture for 7 days in the presence of angiogenic growth and humans and were shown to differentiate into endofactors [67].
thelial cells in the presence of VEGF and contributed to Cord blood constitutes an alternative source of EPCs.
neoangiogenesis in tumors and wounds [73]. Whether Human cord blood was shown to contain a higher con-MAPCs also participate in vessel formation in the ischemic centration of EPCs as compared to peripheral blood and myocardium remains to be tested. cord blood derived EPCs were reported to show faster differentiation rates than peripheral blood derived EPCs 5 .2. Smooth muscle cells 1 [68]. The CD34 fraction of umbilical cord mononuclear cells contributed to neovascularization in the ischemic Only recently, one report in humans has shown SMC hindlimbs of nude rats and increased blood flow [68] outgrowth after culturing blood mononuclear cells in the ( Table 1).
presence of PDGF-BB [75] (Table 1). However, several Totipotent murine embryonic stem cells were shown to animal studies have given indirect evidence for the existdifferentiate into endothelial cells in vitro [69,70]. After ence of circulating SMC progenitors from the BM. After removal of leukemia inhibitory factor, these cells differen-heterotypic cardiac [76,77] or aortic [76] transplantation in tiate into cells of the different germlayers and form mice, most of the neointimal a-actin positive SMCs in the embryoid bodies in which vascular-like structures can be donor coronary arteries or aortas were from host origin, identified consisting of endothelial cells. Endothelial dif-suggesting that these SMCs might at least in part be ferentiation could be enhanced by adding angiogenic derived from BM-derived smooth muscle progenitor cells. growth factors (VEGF and FGF-2) to the culture [71].
In support of this concept, transplantation of b-galacto-Recently, endothelial cells were derived from cultured sidase-expressing BM into irradiated aortic allograft recipihuman embryonic stem cells by the use of antibodies ent mice revealed that part of the neointimal SMC-like against platelet endothelial cell-adhesion molecule-1.
population consisted of marrow-derived cells [78]. Two These endothelial cells participated in new vessel forma-recent studies revealed that bone marrow-derived SMCs tion after implantation into severe combined immuno-(and endothelial cells) importantly (10-50%) contribute to deficiency mice [72]. Mainly due to the ethical considera-neointima formation and re-endothelialization in the contions linked to embryonic-derived cells, scientists have text of transplant atherosclerosis, balloon injury and prisearched for adult pluripotent cell sources. Only recently, mary atherosclerosis [79,80]. However, other reports demadult progenitors (called multipotent adult progenitor cells onstrated a more modest (1-10%) contribution of BMor MAPCs) with the ability to differentiate into cell types derived cells during these events [81][82][83]. Orlic et al.

1
of all three germlayers have been described [73,74]. These revealed that after implantation of Lin c-Kit bone mar-row cells into the ischemic myocardium, 44% of the newly 6 . Vascular and myocardial regeneration formed SMCs were BM-derived [62]. Analysis of human hearts from sex-mismatched transplantations, have, how-Although vascular regeneration can relieve ischemia, ever, revealed conflicting results, reporting substantial (up rescue viable myocardium and accelerate the post-infarcto 60%) contribution of bone marrow derived cells (includ-tion healing process, the necrotic area will be replaced by a ing SMC and cardiomyocyte precursors) in some [84], but collagenous scar instead of contractile cardiomyocytes. rather low (0.04-2.6%) in other studies [85][86][87]. The Functional cardiomyocyte loss causes progressive ventricupersistence of bipotential vascular progenitors giving rise lar dysfunction leading to end-stage cardiac failure. Alto both endothelial cells and SMCs and their contribution though recent findings have indicated that cardiomyocytes to adult neovascularization, as recently described in the can divide [95], the level of proliferation reported in these embryo [88], requires further study. studies is likely insufficient to rescue cardiomyocyte loss resulting from a large infarct. Therefore, myocardial regeneration or repopulation of post-infarction scar with 5 .3. Hematopoietic stem cells contractile cells has been proposed as a potential therapeutic strategy to prevent progression to fatal cardiac failure. Depending on the location, endothelial cells in the Like for vascular regeneration, different cell types have embryo derive either from angioblasts (committed to the been tested for their ability to engraft the infarcted area endothelial lineage) or from hemangioblasts, (common and improve cardiac function: fetal and neonatal carprecursors for endothelial and hematopoietic cells). Endo-diomyocytes, skeletal myoblasts, BM-derived cells and thelial and hematopoietic stem cells (HSCs) not only share embryonic stem cell-derived cells (reviewed in [96]). An a common origin, the latter can also stimulate the assembly important issue is that the success of cardiac cell transof endothelial cells into nascent blood vessels in the plantation depends on the long-term survival and funcembryo. Indeed, HSCs are present at sites of active tional integration of sufficient numbers of grafted cells. vascular expansion. By producing angiopoietin (Ang)-1, Recently, Zhang et al. estimated that cell survival falls to they stimulated endothelial growth in the embryo [89].
about 10% a week after transplantation of rat neonatal HSCs are also present in adult BM and can be mobilized in cardiomyocytes into cryo-injured hearts [97]. Likely, an response to angiogenic factors like VEGF, PlGF and Ang-1 adequate microenvironment that supports nutrient delivery [90,91]. Conversely, blocking of VEGFR-1-but not and waste removal is necessary to sustain survival, growth VEGFR-2-attenuated their recruitment [37,90]. Recently, and differentiation of the transplanted cells. Therefore, a 1 co-recruitment of VEGFR-1 BM-derived hematopoietic concurrent revascularization must keep pace with repopu-1 (precursor) cells and VEGFR-2 EPCs, was shown to lation of the infarct with new myocardial cells. Two recent contribute to tumor growth and vascularization, suggesting studies have combined myocardial regeneration with a mutually vascular supporting role for these cell types stimulation of blood vessel growth by an angiogenic 1 [92]. Possibly, co-mobilization of VEGFR-1 hemato-factor. Yang et al. implanted VEGF-transfected embryonic poietic (stem) cells is essential for incorporation of EPCs stem cell-derived cardiomyocytes into infarcted mouse into newly formed vessels. In agreement with the in-hearts and the density of newly formed capillaries was 1 volvement of VEGFR-1 cells in angiogenesis, we demon-significantly higher than in animals implanted with nonstrated that VEGFR-1 blocking resulted in reduced angio-transfected cells [98]. In a second study, neonatal cargenesis in ischemic retinopathy, tumor growth and diomyocytes transfected with hepatocyte growth factor rheumatoid arthritis [37]. The existence and identity of were implanted in the infarcted area of rat hearts and a hemangioblasts in adult tissues and their contribution to significant improvement in myocardial perfusion was postnatal vessel growth has remained uncertain. Proof of obtained but not when non-transfected cells were used concept requires experiments in which is demonstrated that [99]. Alternatively, the transplanted cells may give rise to a single cell can give rise to both endothelial and hemato-both cardiomyocytes and vessels. Orlic [62]. Transcontributed to both hematopoietic and endothelial cells plantation of the 'side population' resulted in engraftment [93], suggesting that these transplanted cells had heman-of both endothelial cells and cardiomyocytes [64]. gioblastic potential. In addition, Pelosi et al. reported that a 1 1 small subset of CD34 KDR cells from bone marrow or cord blood have long-term proliferative potential and can 7 . Therapeutic perspectives differentiate in endothelial and blood lineage [94]. Although a single MAPC generated cells of all germ layers Hopes are raised that vascular regeneration in the heart including endothelium, in vitro differentiation into hemato-will be of therapeutic value in the future. However, some poietic cells was never observed [73,74]. However, in issues need to be resolved before its therapeutic potential vivo, MAPCs gave rise to cells of all blood lineages [74].
can be rigorously tested in patients. First, reports on the (consisting of 0.65% AC133 and 2.1% CD34 cells) in ly, distinct degrees of engraftment have been reported in 1 patients 7 days after myocardial infarction [102]. In a hindlimb ischemia. Transplantation of human CD34 cells similar set-up, Assmus et al. reported that transplantation into nude mice resulted in incorporation of human cells in of either bone marrow mononuclear cells or blood-derived |13% of the capillaries in the ischemic limbs [58]. After ex vivo expanded progenitors (of which 90% carried ex vivo expansion of human EPCs (by culture for 7-10 endothelial markers), increased myocardial viability, perfudays in the presence of angiogenic growth factors), human sion and function with comparable efficiency [103]. Alcells were found in |56% of the vessels of the ischemic though these results need to be confirmed in large-scale limb [51]. The percentages mentioned above might be an randomized placebo-controlled trials, they are encouraging underestimation, as EPCs and / or other bone marrow cells to continue in this rapidly progressing scientific field. may provide a local source of angiogenic and inflammatory factors that further enhance new vessel formation [63,66].
R eferences Secondly, the transplanted cells should contribute to the formation of functional vessels that restore blood flow to 1 isolated CD34 cells [63] or total bone marrow [65] [2] Carmeliet P, Ng YS, Nuyens D et al. Impaired myocardial angioresulted in increased vessel formation and improved genesis and ischemic cardiomyopathy in mice lacking the vascular endothelial growth factor isoforms VEGF164 and VEGF188. Nat cardiac function, indicating that the newly formed vessels in EPC requirement for limb revascularization [100]. This