Estrogen Increases Bone Marrow–Derived Endothelial Progenitor Cell Production and Diminishes Neointima Formation
Background— Estrogens improve endothelial function and accelerate reendothelialization after vascular injury via largely unknown mechanisms. Bone marrow–derived endothelial progenitor cells (EPCs) are thought to positively influence endothelialization, vascular repair, and angiogenesis.
Methods and Results— In mice subjected to sham operation, ovariectomy, or ovariectomy and estrogen replacement treatment, estrogen deficiency significantly decreased EPCs circulating in the peripheral blood and residing in the bone marrow, as well as EPCs that were in vitro expanded from spleen-derived mononuclear cells. These effects were completely prevented by estrogen replacement. Human women with increased estrogen plasma concentrations also displayed profoundly increased levels of circulating EPCs. Estrogens increase EPC numbers through a decreased apoptosis rate, which is mediated via a caspase-8–dependent pathway. Estrogen deficiency increased neointima formation after carotid artery injury in mice, but this effect was diminished by estrogen replacement therapy. In mice transplanted with green fluorescent protein–positive bone marrow, reendothelialization of injured vessel segments by bone marrow–derived cells was decreased during estrogen deficiency and increased in response to estrogen treatment.
Conclusions— Estrogens increase numbers of EPCs by antiapoptotic effects leading to accelerated vascular repair and decreased neointima formation.
Received April 16, 2003; accepted April 29, 2003.
Vascular function not only depends on cells residing within the vessel wall but is significantly modulated by circulating bone marrow–derived cells.1 A subset of these stem cells, endothelial progenitor cells (EPCs), enhance angiogenesis and vascular repair, diminish atherosclerosis, and increase ventricular function after myocardial infarction.2–6 Myocardial infarction and coronary bypass surgery increase and the accumulation of risk factors reduces the circulating numbers of these bone marrow–derived cells, which suggests that vascular health and also repair processes after injury require increased numbers of these potentially beneficial cells.7–9
Besides their effects on classical risk factors such as lipoproteins, estrogens exert their potentially vasoprotective properties through diverse, direct effects on vascular cells.10–13 Estrogens increase the production of nitric oxide (NO) and reduce the production of reactive oxygen species, which in turn are putatively important for growth-stimulating and antiapoptotic effects on the endothelium.10,14,15 Consequently, experiments in several animal models revealed that estrogen accelerates reendothelialization processes after injury.16 Supply of a functional endothelial monolayer is crucial for protection against atherosclerosis and restenosis after injury.17
Because EPCs have been implicated in various events requiring endothelialization, we reasoned that estrogens could influence the action of EPCs and that this could be important for vasculoprotection exerted by the hormone. In order to corroborate this assumption, we quantified EPCs in estrogen-deficient and estrogen-treated animals and humans and characterized the underlying molecular mechanisms. In addition, we investigated the influence of EPCs after estrogen therapy on vascular repair mechanisms.
Mononuclear cells were isolated from the blood of healthy young volunteers by density gradient centrifugation with Ficoll separating solution (Cedarlane). Cells (1×107) were plated in endothelial basal medium (EBM) (Cellsystem) with supplements (1 μg/mL hydrocortisone, 3 μg/mL bovine brain extract, 30 μg/mL gentamicin, 50 μg/mL amphotericin B, 10 μg/mL hEGF, 20% fetal calve serum) and stimulated on extraction day or after 3 days of culture with 100 nmol 17β-estradiol (E2) for the respective time points. Adherent cells were washed with medium, incubated with 2.4 μg/mL 1,1′-dioctadecyl-3,3,3′,3′-tetramethylindocarbocyanine–labeled acetylated LDL (DiLDL; Harbor Bio-Products) for 1 hour, fixed with 2% paraformaldehyde, and counterstained with fluorescein isothiocyanate–labeled lectin (Sigma). The numbers of DiLDL- and lectin-positive cells were evaluated by two independent investigators.
All animal procedures were approved and were in accordance with the institutional guidelines. Female mice (C57BL6; Charles River, Schweinfurt, Germany) were put on a standard chow and were ovariectomized or sham-operated 6 weeks after birth. E2 pellets (containing 0.18 mg E2 each, 60-day release; Innovative Research) were implanted subcutaneously. The mice were euthanized 4 weeks after surgery, and blood and tissue samples were collected immediately. Spleens were mechanically minced, and mononuclear cells were isolated as described above. Splenocytes (8×106) were seeded into fibronectin-coated 24-well plates. After 5 days, EPCs were stained for DiLDL/lectin and counted by two independent investigators. E2 levels were measured by RIA assay.
Carotid Artery Injury Model
Carotid artery injury was induced as described previously.9 Four weeks after ovariectomy, mice were anesthetized with 150 mg/kg body weight ketamine hydrochloride (Ketanest; Pharmacia) and 0.1 mg/kg body weight xylazine hydrochloride (Rompun 2%; Bayer). Using a dissecting microscope (MZ6; Leica), the bifurcation of the left carotid artery was exposed via a midline incision of the ventral side of the neck. Two ligatures were placed proximally and distally around the external carotid artery. The distal ligature was then tied off. After temporary occlusion of the internal and common carotid artery with ligatures, a transverse arteriotomy was performed between the ligatures of the external carotid artery to introduce a 0.13-mm-diameter curved flexible wire. The wire was passed 3 times along the common carotid artery in a rotating manner. After removal of the wire, the proximal ligature of the external carotid artery was tied off. Restoration of normal blood flow was confirmed, and the skin was closed with single sutures of 6/0 silk. Animals were euthanized 2 weeks after induction of carotid artery injury.
Perfusion-fixed carotid arteries were embedded in Tissue Tek O.C.T. embedding medium (Miles), snap-frozen, and stored at −80°C. Samples were sectioned on a Leica cryostat (7 μm) and placed on poly-l-lysine–coated (Sigma) slides. Cryosections were postfixed in 4% formaldehyde for 2 minutes. Complete reendothelialization was confirmed by staining representative sections with von Willebrand factor (vWF), polyclonal antibody (clone A 0082, Dako), and 4′,6-diamidino-2-phenylindole (Dapi; Linaris). Isotype-specific antibodies (Santa Cruz) were used for negative controls. For morphometric analyses, hematoxylin-eosin staining was performed according to standard protocols. All sections were examined under a Nikon E600 microscope. Lucia Measurement Version 4.6 software was used to measure external elastic lamina, internal elastic lamina, and lumen circumference, as well as medial and neointimal area of 25 sections per animal.
Retroviral Gene Transfer and Bone Marrow Transplantation
Reconstitution of hematopoiesis with green fluorescent protein (GFP)–marked bone marrow cells was performed as described previously.9,18,19 Briefly, bone marrow was harvested from 8- to 10-week-old female C57/BL6 mice 2 days after treatment with 5-fluorouracil (Sigma). Bone marrow cells were stimulated in DMEM/15% fetal calve serum supplemented with 20 ng/mL recombinant murine interleukin-3, 50 ng/mL human interleukin-6 (PromoCell), and 50 ng/mL rat stem cell factor (generously provided by Amgen, Thousand Oaks, Calif) for 48 hours. Subsequently, bone marrow cells were cocultured with irradiated (13-Gy) viral producer cells. The ecotropic packaging cell line GP+E86 was used for the generation of MGirL22Y retroviral vector particles. The MGirL22Y vector was generated by cloning the enhanced GFP (Clontech) gene upstream from the internal ribosomal entry site from the encephalomyocarditis virus linked to a mutant dihydrofolate reductase gene (L22Y) in a murine stem cell virus vector. The presence of recombinant retrovirus was excluded in assays that used a Mus dunni test cell line. After 48 hours of coculture, nonadherent bone marrow cells were rinsed off the producer cell monolayer. Transduced bone marrow cells (5×106) were transplanted by tail vein injection into lethally irradiated 8- to 10-week-old female C57/BL6 mice (11-Gy cumulative dose). Flow cytometric analysis of GFP expression was performed in peripheral blood leukocytes of mice 1 month after transplantation of GFP-expressing bone marrow cells and in leukocytes of control mice that did not receive GFP-labeled bone marrow. Approximately 70% of white blood cells were stably expressing the fluorophore in GFP–bone marrow chimeras. All myeloid and lymphocytic populations were labeled with varying levels of GFP expression within lineages.
Blood samples of 15 patients from the gynecology outpatient clinic undergoing planned in vitro fertilization were investigated. Controlled ovarian hyperstimulation as described by long protocol was initiated in all patients with a gonadotropin-releasing hormone analogue (triptorelin acetate 0.1 mg subcutaneously daily) starting in the midluteal phase of the previous cycle until pituitary desensitization was achieved. To induce follicular growth, gonadotropin therapy was started using recombinant follicle-stimulating hormone 150-200 (Gonal-F; Serono). Gonadotropin-releasing hormone analogue injection was continued up to and including the day of ovulation induction (days 10 to 12). EDTA plasma (30 mL) was taken before and after 6 to 10 days of follicle-stimulating hormone treatment, and expression of the surface markers CD 34 and VEGF-R2 was evaluated by fluorescence-activated cell sorter (FACS) analysis. The local ethics committee of the University of the Saarland approved the studies.
The rate of apoptosis in E2-treated EPCs was assessed using the Cell Death Detection ELISA PLUS System (Roche Molecular Biochemicals). The test principle is based on determination of the amount of nucleosomes generated during the apoptotic fragmentation of cellular DNA. Mononuclear cells were cultured in EBM for 5 days. Adherent cells were washed with EBM twice and incubated in 100-μL lysis buffer for 30 minutes. After centrifugation for 10 minutes at 15 000 rpm and 4°C in a microcentrifuge, 50 μL of each sample was incubated in anti-histone–coated microtiter plate wells for 90 minutes, the wells were washed 3 times with incubation buffer, and 100 μL of anti-DNA peroxidase-linked antibody was added, followed by further incubation for 90 minutes. After 3 washing steps with incubation buffer, 100 μL of ABTS substrate solution for the peroxidase was added; after 10 to 20 minutes, the rate of apoptosis was determined by photometric measurement at 492 nm.
The results were confirmed by FACS analyses with the Annexin V Fluos labeling kit (Roche Molecular Biochemicals).
Caspase-8-activity was measured by using the Caspase-8 colorimetric assay kit according to the instructions of the manufacturer (Calbiochem). Cell lysates were incubated with 5 mmol/L dithiothreitol and the caspase-8 colorimetric substrate, which was conjugated to IETD-p-nitroaniline in reaction buffer for 2 hours at 37°C. Cleavage of the substrate was determined by absorbance at 405 nm.
Results are presented as mean±SEM. Significance of the difference between two measurements was determined by unpaired Student’s t test, and multiple comparisons were evaluated by the Newman-Keuls multiple comparison test. Values of P<0.05 were considered significant.
Estrogen Increases EPC Levels in Mice
Mice were sham-operated or ovariectomized with or without concomitant estrogen replacement treatment. Samples of peripheral blood and bone marrow were obtained after 4 weeks and were submitted to FACS analysis to quantify the expression of the mouse stem cell marker Sca-1 and the endothelial lineage marker VEGF-R2. Quantitative analyses are depicted in Figure 1. Numbers of circulating EPCs decreased during estrogen deficiency to 42.6±27.5% of controls (sham-operated), but this effect was prevented by estrogen treatment (109.3±34.6% of control). EPCs residing within the bone marrow showed a similar response. Ovariectomy diminished cells to 41.1±7.1% of control, whereas estrogen replacement significantly inhibited this effect (75.0±11.5% of control). In mice, hematopoiesis is also accomplished within the spleen. Therefore, mononuclear cells derived from spleen homogenates were isolated 4 weeks after ovariectomy and were cultured for 5 days. EPC numbers were assessed after staining with DiLDL and lectin using fluorescence microscopy. Figure 2 shows the quantitative assessment demonstrating that estrogen deficiency decreased spleen-derived EPCs to 37.6±14.7% of control levels, whereas estrogen treatment prevented this decline (141.7±33.6% of control).
Estrogen Increases EPC Levels in Humans
In order to test the relevance of these findings for humans, the number of circulating EPCs was assessed in women with increased estrogen plasma levels due to ovarian hyperstimulation in the context of planned in vitro fertilization. Figure 3 displays estrogen levels and corresponding EPC numbers. Circulating EPCs were increased to 463±102% of control during estrogen stimulation.
Estrogen Reduces Apoptosis of EPCs
Cultured human EPCs were identified through double-staining with DiLDL and lectin. Cell counts were assessed in cultures incubated for 24 hours with vehicle or 100 nmol/L E2. Figure 4 shows that EPC numbers increased from 125±31 to 216±30 after 24 hours in response to estrogen.
To detect a potentially decreased apoptosis as the underlying mechanism, cell death detection and annexin assays were employed in cultured human EPCs after stimulation with vehicle, 100 nmol/L E2, 25 nmol/L tumor necrosis factor (TNF)-α, or both. Figure 5A displays that estrogens significantly reduced basal and TNFα-induced apoptosis of EPCs. Caspase-8 activity was significantly decreased under estrogen treatment, which suggests that caspase-8 could serve as intracellular target in the apoptosis cascade (Figure 5B).
Estrogen Prevents Neointima Formation
Neointima formation after carotid artery injury is dependent on the rate of reendothelialization,17 which in turn is influenced by the number of circulating EPCs.9 To explore a vasculoprotective effect of estrogen, sham-operated and ovariectomized mice with or without estrogen replacement therapy were subjected to a carotid artery injury procedure. Fourteen days after lesion induction, vessels were harvested and histologically analyzed. Figure 6 shows representative examples of neointima formation and the quantitative histomorphometric measurements. Estrogen deficiency increases neointima area to 1 165 729±67 170 μm2, whereas estrogen treatment reduces the formation of neointima to 860 341±98 665 μm2. Media thickness was increased by estrogen deficiency (436 397±32 115 μm2) as opposed to estrogen replacement (307 227±43 628 μm2). Consequently, lumen circumference was significantly reduced in ovariectomized mice in comparison to sham-operated animals.
Estrogens Increase Vascular Repair by Bone Marrow–Derived Cells
After reconstitution with GFP-expressing bone marrow cells, mice were subjected to carotid artery injury procedures.9,17 Fourteen days after injury, harvested vessels were analyzed by direct fluorescent microscopy to visualize GFP-positive cells. The animal model has been extensively characterized, as published previously.9,18,19 No GFP-positive cells were detected in uninjured vessels. At the lesion side, GFP-positive cells were frequently detected at the endothelial monolayer (Figure 7, A and C). Immunohistological analyses with monoclonal antibodies against vWF (Figure 7, B and D) were employed, demonstrating that luminal GFP-positive cells represent endothelial cells. Control stainings with Dapi (data not shown) show the spindle-shaped nucleus of endothelial cells. These events were markedly decreased in ovariectomized mice (Figure 7, A and B) as opposed to ovariectomized, estrogen-treated animals (Figure 7, C and D) and sham-operated animals (data not shown). Quantification of bone marrow–derived cells at the lesion site committed to the endothelial cell lineage is depicted in Figure 7E.
There is accumulating evidence that cardiovascular diseases such as atherosclerosis, restenosis after injury, and myocardial regeneration after infarction not only depend on cells formerly residing in the location of the vascular or myocardial insult but are decisively influenced by bone marrow–derived cells.1–9 EPCs resemble such premature, circulating, bone marrow–derived cells with putative transdifferentiation potential.20 On the one hand, accumulation of risk factors leads to a decreased number of circulating EPCs.21 On the other hand, these cells have been implicated in revascularization, vascular repair, and myocardial regeneration.1–9 In terms of cardiovascular diseases, an increase of EPCs could be considered a potential benefit.
Estrogen deficiency after menopause correlates with inclined rates of cardiovascular events. Estrogens exert multiple immediate effects on cardiovascular cells, including enhancement of NO bioavailability and reduction of oxidative stress.10 Because estrogens have been shown to fortify reendothelialization after injury,16 and in the light of the effects of EPCs on endothelial regeneration, the herein-demonstrated enhancing effect of estrogens on EPC numbers and adhesion partly closes the gap in our pathophysiological understanding. Estrogens increase the bone marrow–located production and survival of EPCs, which amplifies the circulating levels of these cells. This potentially facilitates beneficial effects on vascular lesion sites. Especially after injury and atherosclerosis, the absence of an intact endothelial monolayer represents one of the driving forces that accelerates the progression of the disease, ultimately leading to enhanced neointima formation.17,22,23 The GFP-chimera animal model indicates not only that estrogens increase the numbers of EPCs in the circulation but that bone marrow–derived cells arrive more frequently at the vascular lesion, which in turn reduces neointima formation. In any event, estrogens exert multiple cellular and molecular effects that could be also related to the reduced neointima formation, including growth-inhibiting properties on vascular smooth muscle cells.10
The cellular and molecular events involved in regulation of EPC numbers are less clear. Statins increase EPC numbers via a PI-3-kinase–dependent pathway, which is putatively related to the release of NO, causing reduced apoptosis rates of EPCs.2,4 Estrogens are known to induce NO and PI-3-kinase in the endothelium.10 Whether this occurs also in EPCs is so far undetermined. However, our data imply involvement of the caspase-8 pathway. It remains unknown how agents such as estrogens or statins exactly influence the production, mobilization, egress, and adhesion of EPCs at the target sites. The detailed understanding of these cascades is essential for successful drug-based induction of subsets of bone marrow–derived cells in order to treat cardiovascular diseases.
Finally, estrogens exert a variety of biological effects that are not exclusively of beneficial nature. Large-scale studies have failed to provide evidence for cardiovascular benefits in the primary as well as secondary prevention setting.24,25 It is thought that this may be due to concomitant adverse events such as prothrombotic properties of estrogens. If we were able to omit these harmful effects of estrogens, the well-documented impact on the vasculature would probably result in a better clinical outcome for patients on hormone replacement treatment. In this respect, selective estrogen receptor modulators may be of special interest.
The induction of EPCs may not only be of a beneficial nature. Estrogens increase the incidence of various malignancies.24 Notably, statins also induce malignant diseases in the elderly.26 Both tumor growth and metastasis require neovascularization, which in turn is readily enhanced by EPCs. It is tempting to speculate that neoplasia induction during estrogen therapy is dependent on the herein-shown effects on EPCs.
Estrogens increase the number of EPCs that may participate in accelerated reendothelialization and reduced neointima formation after vascular injury. The presented findings not only add another intriguing aspect to the multiple biological effects of estrogens but also portend a starting point for novel treatment regimens for vascular diseases such as atherosclerosis and restenosis after injury.
This study was supported by the Deutsche Forschungsgemeinschaft. We thank Anja Geitlinger, Simone Jäger, Sybille Karl, and Isabel Paez-Maletz for excellent technical assistance.
↵*Drs Strehlow and Werner contributed equally.
This article originally appeared Online on June 16, 2003 (Circulation. 2003;107:r119–r125).
In mice, estrogen deficiency significantly decreased EPCs circulating in the peripheral blood, in the bone marrow, and in vitro–expanded spleen-derived mononuclear cells. These effects were completely prevented by estrogen replacement. Human women with increased estrogen plasma concentrations also displayed increased levels of circulating EPCs. Estrogen deficiency increased neointima formation after carotid artery injury in mice, but this effect was diminished by estrogen replacement therapy. In mice transplanted with green fluorescent protein–positive bone marrow, reendothelialization of injured vessel segments by bone marrow–derived cells was decreased during estrogen deficiency and increased in response to estrogen treatment. Estrogens increase numbers of EPCs based on antiapoptotic effects leading to accelerated vascular repair and decreased neointima formation.
Asahara T, Murohara T, Sullivan A, et al. Isolation of putative progenitor endothelial cells for angiogenesis. Science. 1997; 275: 964–967.
Kalka C, Masuda H, Takahashi T, et al. Transplantation of ex vivo expanded endothelial progenitor cells for therapeutic neovascularization. Proc Natl Acad Sci U S A. 2000; 97: 3422–3427.
Assmus B, Schachinger V, Teupe C, et al. Transplantation of Progenitor Cells and Regeneration Enhancement in Acute Myocardial Infarction (TOPCARE-AMI). Circulation. 2002; 106: 3009–3017.
Vasa M, Fichtlscherer S, Adler K, et al. Increase in circulating endothelial progenitor cells by statin therapy in patients with stable coronary artery disease. Circulation. 2001; 103: 2885–2890.
Walter DH, Rittig K, Bahlmann FH, et al. Statin therapy accelerates reendothelialization: a novel effect involving mobilization and incorporation of bone marrow–derived endothelial progenitor cells. Circulation. 2002; 105: 3017–3024.
Werner N, Priller J, Laufs U, et al. Bone marrow–derived progenitor cells modulate vascular reendothelialization and neointimal formation: effect of 3-hydroxy-3-methylglutaryl coenzyme a reductase inhibition. Arterioscler Thromb Vasc Biol. 2002; 22: 1567–1572.
Nickenig G, Strehlow K, Wassmann S, et al. Differential effects of estrogen and progesterone on AT1 receptor gene expression in vascular smooth muscle cells. Circulation. 2000; 102: 1828–1833.
Wassmann S, Bäumer AT, Strehlow K, et al. Endothelial dysfunction and oxidative stress during estrogen deficiency in spontaneously hypertensive rats. Circulation. 2001; 103: 435–441.
Brouchet L, Krust A, Dupont S, et al. Estradiol accelerates reendothelialization in mouse carotid artery through estrogen receptor-alpha but not estrogen receptor-beta. Circulation. 2001; 103: 423–428.
Lindner V, Fingerle J, Reidy MA. Mouse model of arterial injury. Circ Res. 1993; 73: 792–796.
Priller J, Persons DA, Klett FF, et al. Neogenesis of cerebellar Purkinje neurons from gene-marked bone marrow cells in vivo. J Cell Biol. 2001; 155: 733–738.