Estradiol Stimulates Capillary Formation by Human Endothelial Progenitor Cells
Role of Estrogen Receptor-α/β, Heme Oxygenase 1, and Tyrosine Kinase
Endothelial progenitor cells (EPCs) repair damaged endothelium and promote capillary formation, processes involving receptor tyrosine kinases (RTKs) and heme oxygenase 1 (HO-1). Because estradiol augments vascular repair, we hypothesize that estradiol increases EPC proliferation and capillary formation via RTK activation and induction of HO-1. Physiological concentrations of estradiol (10 nmol/L) increased EPC-induced capillary sprout and lumen formation in matrigel/fibrin/collagen systems. Propyl-pyrazole-triol (PPT; 100 nmol/L; estrogen receptor [ER]-α agonist), but not diarylpropionitrile (ER-β agonist), mimicked the stimulatory effects of estradiol on capillary formation, and methyl-piperidino-pyrazole (ER-α antagonist) abolished the effects of estradiol and PPT. Three different RTK activators (vascular endothelial growth factor, hepatocyte growth factor, and stromal derived growth factor 1) mimicked the capillary-stimulating effects of estradiol and PPT. SU5416 (RTK inhibitor) blocked the stimulatory effects of estradiol and PPT on capillary formation. Estradiol increased HO-1 expression by 2- to 3-fold, an effect blocked by SU5416, and PPT mimicked the effects of estradiol on HO-1. The ability of estradiol to enhance capillary formation, increase expression of HO-1, and augment phosphorylation of extracellular signal–regulated kinase 1/2, Akt, and vascular endothelial growth factor receptor 2 was mimicked by its cell-impermeable analog BSA estradiol. Actinomycin (transcription inhibitor) did not alter the effects of estradiol on RTK activity or vascular endothelial growth factor secretion. We conclude that estradiol via ER-α promotes EPC-mediated capillary formation by a mechanism that involves nongenomic activation of RTKs and HO-1 activation. Estradiol in particular and ER-α agonists in general may promote healing of injured vascular beds by promoting EPC activity leading to more rapid endothelial recovery and capillary formation after injury.
- hormone replacement therapy
- vascular remodeling
- cardiovascular disease
- estrogen receptors
- endothelial progenitor cells
Evidence from epidemiological studies suggests that endogenous human estrogens and estrogen replacement therapy protect women from progression of cardiovascular disease.1,2 Also, multiple animal studies and some small clinical trials support a cardioprotective action of estrogens.1,2 In contrast, 2 large randomized clinical trials (Heart and Estrogen/Progestin Replacement Study and Women’s Health Initiative) fail to demonstrate that exogenous estrogens protect against cardiovascular disease.1–4 Although the reasons for these discordant findings remain unclear, a re-evaluation of data from the Heart and Estrogen/Progestin Replacement Study and Women’s Health Initiative suggest that, in older participants with established cardiovascular pathology, estrogen replacement therapy is ineffective, whereas in younger healthy women estrogen therapy is protective.3,4 Similarly, Hodis et al5 report that estradiol inhibits age-associated increases in intimal thickening in women. These findings have generated a renewed interest in the mechanisms by which estrogens influence the cardiovascular system. To correctly interpret the results of completed and future clinical studies, it is critical to elucidate the mechanisms by which estrogens influence the vessel wall and to identify the independent variables that may influence the vascular actions of estrogens.
Although estrogens induce protective effects on the cardiovascular system via multiple mechanisms,6 the effects of estrogens on endothelial cell growth and function may play important roles. For example, estradiol promotes endothelial cell growth, protects endothelial cells against damage by oxidants and cholesterol, and induces the generation of endothelial-derived vasodilators, such as NO and prostaglandins.6–8 Because estradiol promotes endothelial function and because recent studies suggest that circulating bone marrow–derived endothelial progenitor cells (EPCs) contribute to tissue repair by inducing angiogenesis and neovasculogenesis,9 we hypothesize that estradiol may promote endothelial repair in part by stimulating EPC-induced capillary formation.
Estradiol influences cellular growth and differentiation in a variety of tissues in part via estrogen receptors (ERs) α and β.1,6 Within the cardiovascular system, both ER-α and ER-β mediate the protective actions of estradiol; however, these receptors do not necessarily engage common mechanisms. Indeed, ER-α mediates estradiol-induced NO release, inhibits vascular smooth muscle cell growth and lesion formation, and induces endothelial cell growth.1,6–10 Arterial blood pressure is increased in ER-β knockout mice,11 and in Japanese postmenopausal women a specific polymorphism in the ER-β gene is associated with hypertension,12 suggesting that ER-β plays a critical role in lowering blood pressure. These findings provide evidence that ER-α and ER-β may perform distinct functions and that the endogenous regulation of the expression of these 2 ER subtypes may be important in defining the actions of estradiol on the vessel wall. Because ER-α regulates cell growth, we hypothesize that estradiol may activate EPC-induced capillary formation via ER-α.
Capillary formation (angiogenesis or neovascularization) is a dynamic process regulated by growth factors and signaling pathways. Activators of receptor tyrosine kinases (RTKs), such as vascular endothelial growth factor (VEGF), hepatocyte growth factor (HGF), and stromal derived growth factor-1 (SDF-1) induce EPCs to form capillaries.13 Also, angiogenesis appears to depend on heme oxygenase 1 (HO-1), a 32-kDa stress-inducible enzyme that catalyzes the rate-limiting step in the degradation of heme, resulting in the liberation of iron, CO, and biliverdin.13 Growth factors and cytokines induce HO-1 as an adaptive and beneficial response to tissue injury.14 Proangiogenic factors, such as VEGF, increase HO-1 expression in endothelial cells in vitro, and inhibition of HO-1 blocks angiogenesis in vivo.15 Because estradiol is known to induce VEGF synthesis and endothelial cell growth,2,4,6 we hypothesize that estradiol may induce EPC-induced capillary formation via sequential activation of RTKs and HO-1.
The main purpose of the present study was to investigate whether estradiol promotes EPC-induced capillary formation and whether these effects are mediated via ER-α, ER-β, or both. Because EPC-induced capillary formation likely is regulated by sequential activation of RTKs and HO-1, a second aim of our study was to investigate their roles in mediating the stimulatory effects of estradiol on capillary formation.
Isolation and Culture of EPCs
EPCs were isolated, cultured, and characterized as described previously.9 For detailed Materials and Methods, please see the expanded data in the online Data Supplement and Table S1, available online at http://hyper.ahajournals.org.
Capillary Formation Studies
Well-established gel (matrigel, collagen, and Cytodex-3 beads in fibrin)-based assays were used to study capillary, lumen, and sprout formation by EPC-derived endothelial cells. Capillary, lumen, and sprout formations were assessed microscopically by measuring the length of capillaries or counting lumens and sprout formations. For details see the expanded methods in the online Data Supplement at http://hyper.ahajournals.org.
Protein Expression Studies
Western blotting, ELISA assays, and flow cytometry were used to assess the role of various proteins in mediating the angiogenic effects of estradiol on EPCs. For details see the expanded methods in the online Data Supplement at http://hyper.ahajournals.org.
Data were analyzed by ANOVA and statistical significance (P<0.05) calculated using the Fisher least significant difference test.
A homogeneous population of CD34+ cells was isolated after antibody-based magnetic separation (Figure 1A). The purity of EPCs was also confirmed by staining with AC133, a marker for stem cell glycoprotein selectively expressed in CD34+ progenitor cells (Figure 1A). To assess whether progenitor cell–derived endothelial cells were phenotypically similar to endothelial cells, the cells were immunostained for endothelial cell-specific markers. As shown in the photomicrographs (Figure 1B through 1D), EPCs stained positive for von Willebrand factor, platelet endothelial cell adhesion molecule 1, and melanoma cell adhesion molecule 1. Human aortic smooth muscle cells were used as negative controls and did not show any positive staining. To assess whether EPCs express ERs, EPC lysates were analyzed by Western blotting. Both ER-α and ER-β were highly expressed in EPCs (Figure 1E). To assess whether EPCs were able to form capillaries on matrigel, EPCs were plated on matrigel-coated slides and treated with 10% steroid-free serum. Treatment with serum resulted in robust capillary formation by EPCs (Figure 1F).
Treatment of EPCs with estradiol increased capillary formation from 168±16 μm in vehicle-treated control cells to 487±24 μm in cells treated with 10 nmol/L of estradiol (P<0.05). The stimulatory effects of estradiol on capillary formation were mimicked by the ER-α agonist propyl-pyrazole-triol (PPT; 100 nmol/L), which stimulated capillary formation by 330±23% (P<0.05 versus untreated control; Figure 2A). In contrast to estradiol and PPT, treatment of EPCs with 100 nmol/L of the ER-β agonist diarylpropionitrile (DPN) failed to induce capillary formation (Figure 2A). The stimulatory effects of estradiol on capillary formation were abrogated in EPCs cotreated with the ER-α antagonist methyl-piperidino-pyrazole (MPP; 1 μmol/L). MPP also blocked the stimulatory effects of the ER-α agonist PPT on capillary formation (P<0.05 versus EPCs treated with estradiol or PPT alone; Figure 2A). Similar to increasing capillary length, estradiol induced capillary junction formation via ER-α (please see Figure S1 available in the online Data Supplement at http://hyper.ahajournals.org). Moreover, estradiol promoted lumen (Figure 2B) and sprout (Figure 2C) formation in EPCs cultured for 24 hours in collagen gels or Cytodex beads for 7 days in fibrin gels, respectively.
Similar to estradiol and PPT, treatment of EPCs with 100 ng/mL of VEGF, HGF, or SDF-1 stimulated EPC-induced capillary formation (Figure 3A). The stimulatory effects of estradiol on lumen and sprout formation were also mimicked by VEGF and phorbol myristate acetate (Figure 2B and 2C). The efficacies of the RTK activators on capillary formation were comparable to those of estradiol and PPT. Also, the stimulatory effects of VEGF, HGF, and SDF-1 on capillary formation were abolished when EPCs were cotreated with 5 μmol/L of the RTK inhibitor SU5416 (Figure 3A). Treatment of EPCs with SU5416 also abrogated the stimulatory effects of estradiol and PPT on EPC-induced capillary formation by 93% (P<0.05 versus EPCs treated with estradiol or PPT alone; Figure 3B). Similar to the effects on capillary length, estradiol and PPT induced junction formation, and these effects were mimicked by RTK activators and blocked by SU5416 (please see Figure S2).
As shown in Figure 4A, treatment of EPCs with 10 nmol/L of estradiol induced HO-1 expression by 2-fold (from 100.0±0.3% to 228.0±7.8%; P<0.05 versus untreated control). Treatment of EPCs with PPT, but not DPN, mimicked the stimulatory effects of estradiol on HO-1 expression (Figure 4A). Treatment with PPT induced HO-1 expression by ≈280% (P<0.05 versus control). The ER-α antagonist MPP abolished the stimulatory effects of estradiol and PPT on HO-1 expression (Figure 4B). Moreover, the stimulatory effects of estradiol and PPT on HO-1 expression were also abrogated by the RKT inhibitor SU5416 (P<0.05 versus EPCs treated with estradiol or PPT; Figure 5A).
Extracellular Signal–Regulated Kinase 1/2 and Akt Phosphorylation
Treatment of EPCs with estradiol (10 nmol/L) significantly upregulated the expression of phosphorylated extracellular signal–regulated kinase (ERK) 1/2 and Akt (Figure 5B). Moreover, these stimulatory effects were abolished by the RTK inhibitor SU5416 (Figure 5B). Treatment with the ERK1/2 pathway inhibitor PD98059 blocked the stimulatory effects of estradiol and PPT on capillary formation (Figure 5C, top). Similarly, the Akt pathway inhibitor LY 294002 blocked estradiol- and PPT-induced capillary formation (Figure 5C, bottom). Both ERK1/2 and Akt pathway inhibitors also blocked the stimulatory effects of estradiol on capillary junction formation (please see Figure S3).
Nongenomic Actions on Capillary Formation
In EPCs treated with fluorescein isothiocyanate–labeled BSA-tagged estradiol, nuclear staining was observed in permeabilized but not intact cells (Figure 6A). Similar to estradiol, treatment of EPCs with impermeable BSA-estradiol (10 nmol/L) significantly induced capillary formation (Figure 6B), and these effects were blocked by ICI182780 (ER-α/β antagonist), MPP (ER-α antagonist), SU5416 (RTK inhibitor), PD98059 (ERK1/2 inhibitor), and LY294002 (Akt pathway inhibitor). Similar stimulatory effects of BSA-estradiol were also observed on capillary junction formation (please see Figure S4). Moreover, like estradiol, treatment with impermeable BSA-estradiol upregulated the expression of HO-1, as well as phosphorylated ERK1/2 and Akt (Figure 6C).
Tyrosine Kinase Activation
Treatment of EPCs with estradiol (10 nmol/L) for 10 minutes significantly induced tyrosine kinase activity, and these effects were not blocked by 100 nmol/L of actinomycin (Figure 6D). Treatment of EPCs with estradiol (10 nmol/L) did not induce VEGF receptor (VEGFR) 2 expression (Figure 6E); however, the levels of phospho–VEGFR-2 were significantly induced in lysates of EPCs treated for 10 minutes with estradiol or its impermeable analog BSA-estradiol (Figure 6F), and these effects of estradiol were not blocked by actinomycin (Figure 6F). Treatment of EPCs with estradiol for 8 hours significantly induced VEGF production, and this effect also was not blocked by actinomycin (Figure 6G). Estradiol did not enhance the stimulatory effects of VEGF on tube formation (Figure 6H).
This study demonstrates several important findings in human circulating EPCs: (1) estradiol stimulates capillary formation by EPCs; (2) the capillary-stimulating effects of estradiol are mimicked by a specific agonist for ER-α (PPT) but not for ER-β (DPN); (3) MPP (ER-α–specific antagonist) blocks the stimulatory effects of estradiol and PPT on capillary formation; (4) the stimulatory effects of estradiol and PPT on capillary formation are mimicked by RTK activators (VEGF, HGF, and SDF-1) and blocked by the RTK inhibitor SU5416; (5) expression of the well-known angiogenesis-stimulating enzyme HO-1 is induced by estradiol and by the ER-α agonist PPT but not by the ER-β agonist DPN; (6) the stimulatory effects of estradiol and PPT on HO-1 expression are blocked by the ER-α antagonist MPP, as well as by the RTK inhibitor SU5416; (7) treatment with estradiol induces Akt and ERK1/2 phosphorylation, and inhibition of the ERK1/2 and Akt pathways by PD98059 and LY294002, respectively, abrogates the capillary stimulatory effects of both estradiol and PPT; and (8) estradiol stimulates tyrosine kinase activity and phosphorylation of the RTK VEGFR-2. Together, our findings provide strong evidence that estradiol promotes EPC-induced capillary formation, that these effects are ER-α mediated, and that the stimulatory effects of estradiol and PPT are mediated via activation of RTKs leading to induction of HO-1 with involvement of the Akt and ERK1/2 signal transduction pathways. Importantly, our findings that the stimulatory effects of estradiol on tube formation, HO-1 expression, and phosphorylation of Akt, ERK1/2, and VEGFR-2 are mimicked by the membrane-impermeable BSA-estradiol and that the stimulatory effects of estradiol on tyrosine kinase activity and VEGF production are not blocked by actinomycin indicate that estradiol induces tube formation via a nongenomic mechanism involving membrane ERs.
Some of the actions of estradiol would be expected to improve vascular health. For example, estradiol induces growth of endothelial cells and speeds the recovery of denuded endothelium,4,7 thus enabling the vessel wall to recover more rapidly after injury. Alternative sources of endothelial cells, for example, bone marrow–derived circulating EPCs,9 may also participate in vascular repair, and studies show that estradiol promotes EPC growth and recruits EPCs to sites of injury. The present study shows for the first time that estradiol actually activates EPC-induced capillary formation. Our finding that estradiol promotes EPC-induced capillary formation suggests that estradiol may facilitate tissue repair in part via this mechanism.
The biological effects of estradiol are largely mediated via ER-α and ER-β.1,6 Our finding that the stimulatory effect of estradiol on capillary formation is mimicked by the ER-α agonist PPT, but not by the ER-β agonist DPN, provides evidence that the effects are indeed ER-α mediated. It is important to note that, because selective ER modulators are being developed for safer and more effective hormone replacement therapy in postmenopausal women, the use of PPT as a selective ER modulator to promote endothelial and tissue recovery may be of therapeutic importance.16 This concept is further supported by the facts that, via ER-α, estradiol inhibits injury-induced neointima formation,10 induces NO release, and prevents bone loss and hot flushes.16
The mechanisms involved in EPC-induced capillary formation or angiogenesis remain unclear. However, it is well established that activation of the RTK pathway by ligands such as VEGF and HGF enhances EPC-induced capillary formation.17 Our findings that RTK activators mimic the effects of estradiol on capillary formation and that the RTK inhibitor SU5416 blocks the effects of estradiol and PPT on capillary formation suggest that the stimulatory effects of estradiol on capillary formation are mediated by activation of RTKs. Also, our observations that estradiol induces ERK1/2 and Akt phosphorylation and that the Akt pathway inhibitor LY294002 and the ERK1/2 pathway inhibitor PD98059 attenuate the stimulatory effects of estradiol and PPT on capillary formation suggest that, via ER-α, estradiol can stimulate capillary formation via activation of these signal transduction mechanisms. It is likely that the Akt and ERK1/2 pathways are downstream of RTK activation because it is well known that RTKs activate these classic signal transduction pathways.17
Previous studies implicate a role for HO-1 in angiogenesis.13 Pharmacological or genetic manipulations that increase HO-1 expression enhance proliferation and tube formation in human microvascular endothelial cells in vitro,18 whereas inhibition of HO-1 decreases tube formation, a phenomenon independent of HO-2.18 RTKs and the Akt and ERK1/2 signal transduction pathways activate HO-1,19 and HO-1 mediates their effects on angiogenesis.17 Our finding that estradiol and the ER-α agonist PPT, but not the ER-β agonist DPN, induce HO-1 expression links the proangiogenic effects of estradiol to HO-1. This notion is further supported by the fact that the stimulatory effects of estradiol and PPT on both capillary formation and HO-1 expression are blocked by the ER-α antagonist MPP and by the RTK inhibitor SU5416. Although our findings implicate HO-1 as a likely intermediate in estradiol-induced capillary formation, the mediators downstream from HO-1 remain unknown. In this regard, detailed studies are required to determine whether HO-1–derived CO or another molecule, such as vasodilator-stimulated phosphoprotein, which is known to stimulate capillary formation by EPCs,13,17 is involved in transducing the angiogenic effects of estradiol.
Estradiol induces multiple vascular actions via nongenomic mechanisms20 and increases VEGF synthesis and mobilizes the RTK VEGFR-2.21 Our finding that the effects of estradiol are mimicked by its impermeable analog BSA-estradiol suggests the participation of a nongenomic mechanism in estradiol-induced tube formation. The observations that estradiol induces RTK, that these effects are not blocked by actinomycin, and that the membrane-impermeable BSA-estradiol induces VEGFR-2 phosphorylation suggest that estradiol can activate this key pathway responsible for capillary induction. Because the effect of estradiol to increase VEGF secretion is not affected by actinomycin, it is likely that estradiol stimulates the release of VEGF from intracellular stores via a nongenomic mechanism.22 Based on the above findings it is feasible that estradiol induces capillary formation by stimulating the release of VEGF, which, in turn, activates RTKs, ERK1/2, Akt, and HO-1. Although our findings suggest that estradiol can stimulate capillary formation via a nongenomic mechanism, the participation of nuclear receptors or genomic activation cannot be ruled out and requires further investigated. Because the TRK family constitutes >58 receptors, it is feasible that receptors other than VEGFR-2 also participate in mediating estradiol’s tube-forming effects.
Our observation that estradiol stimulates capillary formation via a nongenomic mechanism suggests that estradiol may repair endothelial damage within vascular beds. Consistent with this concept, in vivo studies provide evidence that estradiol promotes accumulation of progenitor endothelial cells at sites of injury.23 Our finding that estradiol promotes capillary formation within hours would imply that estradiol could facilitate tissue repair by re-establishing capillary formation to improve perfusion within the damaged vascular sites.
In summary, the present study provides evidence that estradiol induces capillary formation by human EPCs, and these effects are ER-α mediated. The stimulatory effects of estradiol are likely attributed to RTK-mediated induction of HO-1. The Akt and ERK1/2 signal transduction pathways are also involved, perhaps as intermediates between RTK activation and HO-1 induction.
Here we provide strong evidence that estradiol promotes EPC-mediated capillary formation via ER-α and that the stimulatory actions of estradiol are mediated via RTK and HO-1 activation. Because HO-1 induction occurs as an adaptive and beneficial response to tissue injury, estradiol in particular and ER-α agonists in general may protect the vascular system by promoting EPC activity, leading to more rapid endothelial recovery and capillary formation after injury. Apart from repairing the vascular endothelium, estradiol-induced capillary formation would also facilitate tissue repair in the heart after myocardial infarction by re-establishing capillaries and blood supply.
Sources of Funding
This work was supported by Swiss National Science Foundation grants 32-64040.00 and 320000-117998 (to R.K.D.); Oncosuisse OCS-01551-08-2004 (to R.K.D.); EMDO Forschung (to R.K.D.); and California Cancer Research Coordinating Committee (to C.C.W.H.) and National Institutes of Health grants RO1-HL60067 and RO1-HL086959 (to C.C.W.H.), and HL069846, DK068575, and DK079307 (to E.K.J.).
- Received March 12, 2010.
- Revision received March 30, 2010.
- Accepted June 23, 2010.
Dubey RK, Jackson EK. Estrogen-induced cardiorenal protection: potential cellular, biochemical, and molecular mechanisms. Am J Physiol Renal Physiol. 2001; 280: F365–F388.
Dubey RK, Imthurn B, Barton M, Jackson EK. Vascular consequences of menopause and hormone therapy: importance of timing of treatment and type of estrogen. Cardiovasc Res. 2005; 66: 295–306.
Clarkson TB, Appt SE. Coronary artery disease and postmenopausal hormone therapy: is there a time for prevention? Gynaecol Forum. 2004; 9: 11–14.
Miller VM, Duckles SP. Vascular actions of estrogens: functional implications. Pharmacol Rev. 2008; 60: 210–241.
Losordo DW, Inser JM. Estrogen and angiogenesis: a review. Arterioscler Thromb Vasc Biol. 2001; 21: 6–12.
Rosselli M, Imthurn B, Keller PJ, Jackson EK, Dubey RK. Circulating nitric oxide (nitrite/nitrate) levels in postmenopausal women substituted with 17β-estradiol and norethisterone acetate: a two-year follow-up study. Hypertension. 1995; 25: 848–853.
Dzau VJ, Gnecchi M, Pachori AS, Morello F, Melo LG. Therapeutic potential of endothelial cells in cardiovascular diseases. Hypertension. 2005; 46: 7–18.
Pare G, Krust A, Karas RH, Dupont S, Aronovitz M, Chambon P, Mendelsohn ME. Estrogen receptor-α mediates the protective effects of estrogen against vascular injury. Circ Res. 2002; 90: 1087–1092.
Zhu Y, Bian Z, Lu P, Karas RH, Bao L, Cox D, Hodgin J, Shaul PW, Thorén P, Smithies O, Gustafsson JA, Mendelsohn ME. Abnormal vascular function and hypertension in mice deficient in estrogen receptor-β. Science. 2002; 295: 505–508.
Dulak J, Deshane J, Jozkowicz A, Agarwal A. Heme oxygenase-1 and carbon monoxide in vascular pathobiology: focus on angiogenesis. Circulation. 2008; 117: 231–241.
Fang J, Sawa T, Akaike T, Akuta T, Sahoo SK, Khaled G, Hamada A, Maeda H. In vivo antitumor activity of pregylated zinc protoporphyrin: targeted inhibition of heme oxygenase in solid tumor. Cancer Res. 2003; 63: 3567–3574.
Deshane J, Chen S, Caballero S, Grochot-Przeczek A, Was H, Calzi SL, Lach R, Hock TD, Chen B, Hill-Kapturczak N, Siegal GP, Dulak J, Jozkowicz A, Grant MB, Agarwal A. Stromal cell-derived factor-1 promotes angiogenesis via a heme oxygenase 1-sependent mechanism. J Exp Med. 2007; 204: 605–618.
Koehne P, Willam C, Strauss E, Schindler R, Eckardt R-U, Bührer C. Lack of hypoxic stimulation of VEGF secretion from neutrophils and platelets. Am J Physiol Circ Physiol. 2000; 279: H817–H824.
Hamada H, Kim MK, Iwakura A, Ii M, Thorne T, Qin G, Asai J, Tsutsumi Y, Sekiguchi H, Silver M, Wecker A, Bord E, Zhu Y, Kisore R, Losordo DW. Estrogen recptors α and β mediate contrinution of bone marrow-derived endothelial progenitor cells to functional recovery after myocardial infarction. Circulation. 2006; 114: 2261–2270.