VEGF Family Members Regulate Myocardial Tubulogenesis and Coronary Artery Formation in the Embryo
http://www.100md.com
Robert J. Tomanek, Yasuo Ishii, Jennifer
参见附件。
the Department of Anatomy & Cell Biology and the Cardiovascular Center (R.J.T., J.S.H., C.L.S., H.K.H.), Carver College of Medicine, The University of Iowa, Iowa City
Department of Cell Biology (Y.I., T.M.), Weill Medical College, Cornell University, New York, NY.
Abstract
This study tested the hypothesis that coronary tubulogenesis and coronary artery formation require VEGF family members. Quail embryos were injected with soluble vascular endothelial growth factor (VEGF) receptors R1 (Flt-1), R2 (Flk-1), R3 (Flt-4), VEGF-Trap (a chimera of R1 and R2), or neutralizing antibodies to VEGF-A, VEGF-B, or fibroblast growth factor (FGF)-2. Our data document that tubulogenesis is temporally dependent on multiple VEGF family members, because the early stage of tubulogenesis was markedly inhibited by VEGF-Trap and to a lesser extent by soluble VEGFR-1. Some inhibition of tubulogenesis was documented when anti-FGF-2, but not anti-VEGF-A, antibodies were injected at embryonic day 6 (E6). Most importantly, we found that VEGF-Trap injected at either E6 or E7 prevented the formation of coronary arteries. Soluble VEGFR-1 and soluble VEGFR-2 modified the formation of coronary arteries, whereas soluble VEGFR-3 was without effect. Antibodies to VEGF-B, but not VEGF-A, had a strong inhibitory effect on coronary artery development. The absence of coronary artery stems, and thus a functional coronary circulation, in the embryos injected with VEGF-Trap caused an accumulation of erythrocytes in the subepicardium and muscular interventricular septum. Using retroviral cell tagging, we showed that some of the erythrocytes in blood islands and small vascular tubes were progeny of the proepicardium. Thus, another salient finding of this study is the first definitive documentation of proepicardially derived hemangioblasts, which can differentiate into erythrocytes.
Key Words: FGF-2 VEGF-B vasculogenesis angiogenesis hemangioblast Flt-1 Flk-1
Introduction
Formation of the coronary vasculature requires a series of closely temporally and spatially regulated events. (For review, see Tomanek1 and Majesky.2) All of the cells that contribute to the coronary vasculature (endothelial, smooth muscle, pericytes, and fibroblasts) migrate to the heart from the proepicardium, a transitional structure located posterior to the septum transversum.3,4 The cells of the proepicardial organ form the epicardium and subepicardium and undergo epithelial-mesenchymal transformation (reviewed by Olivey et al5). Recent evidence shows that hematopoietic precursors, ie, CD45+ cells, are also present on the surface of the quail heart in blood islands before vascularization of the myocardium.6 Migration of angioblasts into the myocardium and their assembly into vascular tubes constitutes the process of vasculogenesis and is followed by tubular expansion via branching (angiogenesis). Subsequent to these events, a capillary-like network (peritruncal ring) surrounding the base of the outflow tract fuses and penetrates the aorta at 2 specific sites, recruits smooth muscle cells, and consequently forms the 2 main coronary arteries.7–9 The remainder of the endothelial strands in contact with the aorta disappear.10 It is only at this point in time, E8 to E9 in the quail,10 that the coronary vasculature is perfused by blood from the aorta. The development of a venous system, like that of the capillary network, also develops before the coronary perfusion from the aorta.1 Thus, these events are not flow dependent.
We have documented the roles of several key growth factors in the regulation of coronary tubulogenesis, defined as vasculo- and angiogenesis.11–13 Having established an association between tube formation and 2 key growth factors, ie, vascular endothelial growth factor (VEGF) and fibroblast growth factor (FGF)-2, in vivo,11–13 we focused on in vitro studies using explanted quail embryonic hearts, a system that enables the establishment of a tubular system on collagen gels.14–18 These studies revealed that tubulogenesis is dependent on multiple growth factors,17 including members of the VEGF family and their receptors.16 Our data suggested that the tyrosine kinase receptors VEGFR-1 (flt-1) as well as VEGFR-2 (flk-1) may play a role in tube formation in the embryonic heart. The former is a receptor for VEGF-A, VEGF-B, and placental growth factor (PlGF), whereas the latter binds VEGFs A, C, D, and E (reviewed by Hoeben et al19). Like VEGFs, FGFs play a role in the endothelial cell proliferation, migration, and survival. The importance of these and other growth factors, eg, angiopoietins, ephrins, and platelet-derived growth factor, in the regulation of vessel development has been reviewed.20
To determine the temporal effect of VEGF growth factors on tube formation and the subsequent formation of the coronary arteries, we conducted in ovo experiments to test 3 hypotheses: (1) the timing of growth factor expression is critical for coronary tubulogenesis; (2) penetration of the tubular network to form the 2 main coronary arteries is dependent on VEGF family members; and (3) proepicardially derived epicardial cells are precursors of blood cells in that they contribute to blood islands. Because we previously documented, in vitro, that FGF-2 plays a major role in tubulogenesis,16 we also included experiments on its role in ovo.
Materials and Methods
Antibodies and Soluble Receptors
Monoclonal neutralizing antibodies to VEGF-A and FGF-2 were provided by Dr T. Brock and Encysive Pharmaceutical Corp (Houston, Tex), and anti-VEGF-B was purchased from R&D (Minneapolis, Minn). Soluble VEGF receptors R1 (Flt-1), R2 (Flk-1), and R3 (Flt-4) were purchased from R&D. These reagents have been described previously in detail.15 VEGF-Trap, a receptor 1 and 2 fusion protein, comprises portions of the human VEGFR-1 and human VEGFR-2 extracellular domains fused to the constant region of human IgG1 and was provided by Regeneron (Tarrytown, NY). The purified protein has been shown to be >90% pure. VEGF-Trap binds ligands for either VEGFR-1 or VEGFR-2. We previously used this chimera for in vitro tubulogenesis.15
Experimental Procedures
Fertilized quail eggs (Coturnix japonica) were incubated for 6 or 7 days at 37.8°C and 80% humidity. The embryo was exposed via a window in a shell and an incision of the inner-shell membrane, as previously described for chicken.12 With the aid of a micromanipulator, neutralizing antibodies or soluble receptors (12 μg in 12 μL of saline) were injected into the vitelline vein, the shell was sealed with paraffin film, and the egg was returned to the incubator for an additional 2 days. Sham controls were injected with 12 μL of saline. In preliminary experiments, we found that doses of <6 μg were without effect. Our previous in vitro experiments revealed an optimal response with 12 μg added to the media. The hearts of the 8- or 9-day embryos were perfused with 4% paraformaldehyde and embedded in paraffin.
Establishing Presence of Coronary Arteries
Serial cross-sections (6 μm thick) from the basal 1/2 of the heart were stained with hemotoxylin and eosin and systematically scanned for the presence of major coronary arteries at their origin from the coronary ostia. Select sections from some hearts were stained with smooth muscle -actin and QH1 antibodies.
Immunofluorescence and Image Analysis
Deparaffinized sections were blocked by immersing in PBS containing 1% BSA for 1 hour at room temperature. Incubation in QH1 was at room temperature; after blocking with 1% BSA in normal goat serum, the sections were incubated in fluorescein isothiocyanate (FITC)-labeled goat anti-mouse IgG Fab. The slides were viewed with a Nikon fluorescence microscope, images were digitized, and microvascular volume density and vessel diameter were determined using Image Pro Plus software (Media Cybernectics). The QH1 antibody reaction demarcated microvessels. From these images, we determined vessel volume density, ie, cross-sectional areas of vessels/test area, in the compact region of the left ventricle.
Apoptosis Detection
To detect in situ apoptosis, we used ApopTag Detection Kit (Chemicon Int, Temecula, Calif) on paraformaldehyde-fixed and paraffin-embedded sections. This method of detection is based on TUNEL technology, ie, modification of genomic DNA using deoxynucleotidyl transferase for detection of positive cells via specific staining. Positive nuclei were visualized by FITC and all nuclei revealed by nuclear counterstain with 4',6-diamidino-2-phenylindole iodide (DAPI).
In Ovo Retroviral Tagging
A cell line producing CXL, a replication-defective spleen necrosis virus encoding lacZ,21 was generated as described elsewhere.22,23 The packaging cell line, D17.2G,24 was cotransfected with a retroviral plasmid, pCXL, and a neomycin-resistance plasmid, pSV40-neo. (For details, see the online data supplement available at http://circres.ahajournals.org.) Retroviral tagging was performed in Hanburger Hamilton (HH) stage 16 to 17 (E2.5) quail embryos by injecting the viral suspension through a small window in the egg shell, membrane, and chorion and amnion with a glass capillary attached to Picospritzer (General Value Corp, Fairfield, NJ). The eggs were resealed with parafilm and incubated at 37.5°C. After the embryos were allowed to develop until E6, a small window was made at the pointed end of the egg shell. Twelve micrograms of VEGF Trap or hFC was pressure injected into the vitelline vessel through a glass capillary. The embryos were allowed to develop until E8 and the heart was fixed in 2% paraformaldehyde and processed for X-Gal histochemistry as described previously.21 Serial cross-sections were cut on a microtome and stained with eosin or incubated in QH1.
Statistics
Group comparisons for thickness of the ventricular compact region and vascular volume density were based on the Dunnett multiple comparison test. The Wilcoxon 2-sample exact test was used to test for intergroup differences for the presence of 0, 1, or 2 coronary ostia. A value of P0.05 was considered statistically significance.
Results
Growth Factor Deprivation Affects Cardiac Phenotype
Two phenotypic changes related to growth factor deprivation were documented. First, the compact region of the myocardium was significantly thinner (by 33% to 50%) in all of the treatment groups compared with shams at E9, but not at E8 (see online data supplement). The thinner walls were attributable to fewer numbers of cardiac myocytes, as readily discerned in histological cross sections. The second phenotypic alteration, an accumulation of erythrocytes subepicardially, and sometimes in the muscular interventricular septum, was seen in all Trap-injected hearts. These changes are discussed later in this section.
Tubulogenesis Is Dependent on Temporally Expressed Multiple VEGF Family Members
To test the hypothesis that various VEGF family members play a role in tubulogenesis, initially, we injected VEGF-Trap, or soluble VEGFR-1, VEGFR-2, or VEGFR-3 into the vitelline vein at either E6 or E7 and studied the hearts at E8 or E9. VEGFs A and B and PIGF are ligands for R1, whereas A, C, and D are ligands for R2. R3 binds VEGF-C and VEGF-D. When VEGF-Trap was injected at E6 tubulogenesis (as indicated by vascular volume 2 days later) was inhibited by 65% (P<0.0001) (Figures 1 and 2A and 2B). However, when injection was delayed until E7 and tubulogenesis evaluated on E9, the degree of inhibition (25%) was less (P=0.07). This finding underscores the importance of the stage of tube formation as a factor in responsiveness to growth factors. Although soluble VEGFR-2 did not affect tubulogenesis when injected at either time point, soluble VEGFR-1 had the most pronounced effect (32% inhibition) when administered at E7. Soluble VEGFR-3 (Flt-4) had no effect on tubulogenesis (data not shown). We then injected anti-VEGF-A and anti-FGF-2 neutralizing antibodies because VEGF-A and FGF-2 are interdependent in their actions.17 Anti-FGF-2 tended to limit vascular growth when injected at E6, but the probability value (0.15) was beyond statistical significance. Values for anti-VEGF-A-treated embryos were similar to the shams. Measurement of vessel diameters of profiles cut in cross-section revealed that mean tubule diameter (μm) in sham embryos increased from 10.1±0.4 at E8 to 14.0±1.3 at E9. The lower volume density in the E8 Trap group was not attributable to vessel size, because mean diameter (10.7±1.4) was similar to the sham group. In contrast, the lower volume density in the E9 embryo hearts was attributable, at least in part, to smaller vessel diameters (10.2±0.6).
Formation of the Coronary Arteries Requires VEGF Family Members
To test the hypothesis that VEGF signaling is necessary for the capillary plexus at the root of the aorta to penetrate that vessel at 2 distinct points, thereby establishing the roots of the 2 main coronary arteries, we examined heart serial cross-sections of E8 or E9 embryos treated 2 days earlier (Figure 2). Sham-injected hearts showed that both coronary arteries were present in all E9 and at least 1 was present in E8 hearts (Figure 3). Consistent with the data on tubulogenesis, anti-VEGF-A had no effect on the formation of coronary artery stems when administered at either time point. In contrast, 8 of 9 E8 hearts injected with VEGF-Trap had no coronary arteries. Of 4 hearts injected at E7 and studied at E9, 3 lacked coronary arteries, whereas 1 had only 1. Penetration of the aorta by the vascular tubes encircling the root of the vessel was precluded by VEGF-Trap. As seen in Figure 2D, a blood island forms a slight indentation into the aorta but does not extend further into the wall, as verified by serial sections. VEGF-Trap treatment did not prevent formation of the capillary-like network at the base of the aorta. Thus, failure of this tubular network to penetrate the aorta was caused by a lack of signaling rather than absence of this microvascular component. In addition to precluding the formation of coronary arteries, VEGF-Trap also caused massive accumulations of erythrocytes in the ventricular walls (Figure 2). These accumulations occurred primarily in the subepicardium but, in some hearts, were seen in the central portion of the interventricular septum.
Data shown in Figure 3 also illustrate the effects of soluble VEGFR-1 and soluble VEGFR-2 and anti-VEGF and –FGF-2 on coronary artery stem formation. Soluble VEGFR-1 treatment tended to reduce the number of coronary arteries formed at E8. Three of 4 E8 hearts lacked coronary artery stems. Forty percent of the E9 hearts treated with either soluble receptor (R1 or R2) had only 1 coronary artery. To test the possibility that the inhibitory effects of soluble VEGFR-1 were attributable to a higher affinity than VEGFR-2, we injected 15 μg of the receptor into 4 embryos. Both E9 embryos had 2 coronaries, whereas in the 2 E8 embryos, 1 lacked coronary arteries and the other had developed both coronary artery stems. Tubulogenesis, as evidenced by vascular volume density, was similar to the sham-injected group. Thus, the higher dose did not alter the results. Because anti-VEGF-A had no effect on coronary artery formation, we increased its amount to 15 or 20 μg (data not shown). However, the results were not different. Accordingly, we conclude from these experiments that because VEGFR-1 binds VEGF-B and PlGF, these growth factors may contribute to coronary artery formation. In 7 embryos treated with soluble VEGFR-3, all had well-formed coronary arteries. A role for FGF-2 in coronary artery formation is suggested by the finding that inhibition of this growth factor between E7 and E9 resulted in the formation of only 1 coronary artery in 4 of 9 hearts. Anomalous coronary stems were also seen in aortae from embryos treated with either Flt-1 or Flk-1. Multiple coronary channels could be seen with a connection to a single coronary artery (Figure 4). This is in contrast to the single channels seen in sham E9 embryos. Apoptosis was not widespread in the ventricular walls of sham or treated hearts. Although apoptotic cells were occasionally found in the ventricular walls, they were more likely to occur in the subepicardium (Figure 4). Previous work has documented apoptosis of the periaortic vascular plexus during the initial formation of the coronary artery stems.25
Having found that anti-VEGF-A neutralizing antibodies did not prevent coronary artery formation, we tested the hypothesis that VEGF-B, a ligand for VEGFR-1, may play a crucial role. Although anti-VEGF-B (12 μg/mL) had no significant effect on tube formation, coronary artery stems were not formed in most of the embryos studied (Figure 5). None of the 4 E8 hearts developed coronary arteries, whereas 5 of 6 E9 hearts had only 1 coronary artery. Anti-VEGF-B also compromised growth of the compact region of the ventricle in E9 hearts (112±4 μm versus 260±9 in sham).
Proepicardial Cells as Hemangioblast Precursors
The finding that large accumulations of erythrocytes are present in the quail hearts deprived of VEGF suggested to us that these cells differentiate from precursors derived from the proepicardium. Origin from the circulation was ruled out because the coronary circulation was absent in these hearts. To test this hypothesis, we injected a retroviral tag encoding lacZ into the proepicardium of E2.5 (HH16 to HH17) quail (Figure 6). Data are based on 3 VEGF-Trap treated and 3 hFC-treated hearts, all of which were staged as HH31 or HH32 (approximately E7.5 to 8.0).
As seen in Figure 6, cells from the proepicardium could be found in association with accumulations of erythrocytes under the epicardium (as observed in whole mounts and sectioned hearts). In all of the hearts, we documented -galactosidase cells not only as mesothelial, smooth muscle, and endothelial cells but also erythrocytes. The latter occurred in blood islands and in small vascular tubes. In both VEGF-trap and hFC injected embryos, blood islands could be found with erythroblasts -galactose positive. Thus, our findings indicate that VEGF deprivation (which precludes a coronary circulation via inhibition of coronary artery formation) is associated with enhanced differentiation of erythrocytes from proepicardial cells.
Discussion
This study documents VEGF family members as essential regulators of both coronary tubulogenesis and arteriogenesis during in vivo embryonic development in quail. First, we show that (1) tubulogenesis and (2) the formation of the 2 main coronary arteries are dependent on multiple VEGF family members rather than on VEGF-A. This conclusion is based on the following findings: (1) that administration of the flt-1/flk-1 receptor chimera (VEGF-Trap) nearly completely inhibits tube formation and precludes the formation of the coronary arteries; and (2) that anti-VEGF-B, a ligand for VEGFR-1 (flt-1), significantly impairs coronary artery formation. Although we were unable to document a significant effect of anti-VEGF-B on in ovo tubulogenesis, we previously showed, in vitro, that anti-VEGF-B limits coronary vascular tube formation by approximately 65%.16 Second, using a retrovirus, we provide proof that at least some proepicardial cells are hemangioblasts because they differentiate into erythrocytes in myocardial blood islands. Additionally, our findings indicate that FGF-2 contributes to tubulogenesis, and plays a minor role in the initial stage of coronary artery formation. Moreover, our study indicates that VEGFs as well as FGF-2 are important for growth of the myocardium, because their inhibition between E7 and E9 limited growth of the ventricular compact region. We previously documented, in vitro, that these growth factors are interdependent for their optimal effects.17 Whereas FGF-2 has direct effects on the myocardium, the effects of VEGF may be via limiting those of FGF-2. As indicated in recent reviews,1,2 previous studies focused mainly on the proepicardium, epicardial-mesenchymal transformation, and vascular progenitor cells. The current study provides the first in vivo data on the growth factor regulation of (1) tubulogenesis and (2) establishment of the coronary artery stems.
Coronary Tubulogenesis
Progenitor cells from the proepicardium migrate to the epicardium and subepicardium where they differentiate into fibroblasts and endothelial and smooth muscle cells.3,26 Cells that become immunoreactive for VEGFR-2 (flk-1) increase in number27 and subsequently form vascular tubes in an epicardial to endocardial direction, as shown here and as previously documented in the rat.11 This vascular gradient correlates with the density of VEGF protein and VEGF mRNA transcripts.13 Moreover, as shown by our in vitro studies, tube formation by endothelial cells from embryonic hearts is enhanced by hypoxia, a response that is prevented by anti-VEGF neutralizing antibodies.14 Our previous work showed that tube formation is inhibited by 62% when neutralizing antibodies to FGF-2 are added to the embryonic heart explants and by approximately 30% with anti-VEGF-A or soluble Tie-2 receptor.17 In addition, we documented an interdependence of VEGF and FGF-2 and a dependence of VEGF on angiopoietins.17 These findings support the concept that multiple growth factors regulate tubulogenesis in the embryonic heart. Moreover, our current experiments show that the effectiveness of growth factors on tubulogenesis is dependent on their precise temporal expression. Thus, VEGF-Trap was most effective in curtailing tubulogenesis when administered at E6, a stage of early tubulogenesis. Administration of VEGF-Trap 1 day later, when tubulogenesis is more advanced, had a more mild effect.
A role for FGF-2 in coronary tubulogenesis has been previously demonstrated in our laboratory both in vivo11,12 and in vitro.17 Of interest in this regard is the recent report that indicates that FGF-2–mediated capillary morphogenesis requires signaling via Flt-1.28 This finding supports our in vitro data implicating VEGF-B and/or PIGF as an important contributor to coronary tubulogenesis.16 In summary, the current in vivo study supports the concept of multiple growth factor regulation and brings into focus the importance of the VEGF family because we have documented a 2/3 reduction in myocardial vascularization when VEGF-Trap is injected at an early stage of microvessel formation.
Coronary Artery Formation
Previous studies have established the importance of neural crest cells and parasympathetic neurons29,30 and transcription factors in the epicardium, eg, Fog-231 and Ets-1 and Ets-2,32 in the formation of the coronary arteries. The current study has documented the necessity of VEGF family members for (1) tubulogenesis and (2) capillary plexus penetration of the aortic root and formation of coronary arteries. The hypothesis that this family of growth factors provides the key signaling mechanism for coronary ostia and artery formation was based on our observation of high expression of VEGF in epicardial and subepicardial cells at the aortic root and the high density of VEGF-R2 and -R3 transcripts at the aortic sites of coronary artery roots.16 The premise that VEGF family members in addition to VEGF-A play a role in vascularization is supported by the finding that VEGF-C acts synergistically with VEGF-A on bovine aortic endothelial cells33 and is found in epicardial vessels.34 Furthermore, both 167 and 186 splice variants of VEGF-B have been shown to improve the ischemic score in the hindlimb,35 a finding that suggests growth of arterioles and/or arteries.
Our data clearly indicate that VEGF family members are required for the formation of the main coronary arteries via endothelial cell penetration of the aorta, because VEGF-Trap prevented this event. The finding that soluble VEGFR-1 prevented coronary artery formation suggested that VEGF-B and/or PIFG may play a role, because these ligands do not directly activate VEGFR-2. A role for VEGF-B in the developing heart was suggested by its high expression in this organ36,37 and by our finding that anti-VEGF-B inhibits tubulogenesis in the explanted quail heart.15 VEGF-B, as well as VEGF-A, gene delivery also improves the hyperemic response in this model in rabbits.38 Both VEGF-B and PIGF have been shown to increase receptor crosstalk. The former overlaps with NRP1, and the latter has been shown to cause transphosphorylation of VEGFR-2 by activating VEGFR-1.39 The current study documents a key role for VEGF-B in the formation of the coronary artery ostia and stems. This novel finding reveals that the effects of this growth factor extend beyond its role in tubule formation.
Proepicardial Cells Include a Hemangioblast Population
Our observation that VEGF-Trap causes a massive increase in erythrocytes in the subepicardium suggested that these cells arise from epicardial-derived precursors. The presence of hematopoietic (CD45+) cells as early as HH23 (E3.5 to 4.0) in the quail subepicardium were previously reported.6 These cells preceded the occurrence of blood islands. Our study reveals that -galactosidase-positive cells from the proepicardium include erythrocytes, as well as endothelial, smooth muscle, and fibroblast cells. The positive cells were found in sham hearts injected with hFC as well as in those injected with VEGF-Trap. Accordingly, our data provide the first definitive evidence that erythrocytes, which are components of blood islands, arise from proepicardial cells. Moreover, our study indicates that VEGF deprivation is associated with an increase in differentiation of precursor cells into erythrocytes. A limitation of the retrovirus injection of the proepicardium is that only a small portion of the cells are tagged. It is, therefore, possible that erythrocytes could have been derived from another source. As previously noted, these cells could not have emerged from the circulation because they were observed in the absence of a coronary circulation. The only other avenue for their migration would be retrograde via the venous system. This possibility is not likely considering the poor development of a tubular network in VEGF-Trap E8 hearts.
Limitations of Study
Our experiments focused primarily on VEGF family members during 1 period of embryonic development. Therefore, no conclusions can be drawn with regard to other growth factors that may interact with VEGFs or play a role during other periods of development. For example, Donovan et al40 have shown that brain-derived neurotrophic factor is important for endothelial cell survival of intramyocardial arteries and capillaries during early postnatal life. Moreover, we have previously documented the interplay and differential modulation of early postnatal coronary angiogenesis by FGF-2 and VEGF.41
Acknowledgments
This work was supported by NIH grants HL075446 (R.J.T.) and HL078921 (T.M.).
Footnotes
Original received October 5, 2005; revision received February 21, 2006; accepted February 28, 2006.
References
Tomanek RJ. Formation of the coronary vasculature during development. Angiogenesis. 2005; 8: 273–284. [Order article via Infotrieve]
Majesky MW. Development of coronary vessels. Curr Top Dev Biol. 2004; 62: 225–259. [Order article via Infotrieve]
Mikawa T, Fischman DA. Retroviral analysis of cardiac morphogenesis: discontinuous formation of coronary vessels. Proc Natl Acad Sci U S A. 1992; 89: 9504–9508.
Mikawa T, Gourdie RG. Pericardial mesoderm generates a population of coronary smooth muscle cells migrating into the heart along with ingrowth of the epicardial organ. Dev Biol. 1996; 174: 221–232. [Order article via Infotrieve]
Olivey HE, Compton LA, Barnett JV. Coronary vessel development: the epicardium delivers. Trends Cardiovasc Med. 2004; 14: 247–251. [Order article via Infotrieve]
Kattan J, Dettman RW, Bristow J. Formation and remodeling of the coronary vascular bed in the embryonic avian heart. Dev Dyn. 2004; 230: 34–43. [Order article via Infotrieve]
Bogers AJ, Gittenberger-de Groot AC, Poelmann RE, Peault BM, Huysmans HA. Development of the origin of the coronary arteries, a matter of ingrowth or outgrowth Anat Embryol (Berl). 1989; 180: 437–441. [Order article via Infotrieve]
Waldo KL, Willner W, Kirby ML. Origin of the proximal coronary artery stems and a review of ventricular vascularization in the chick embryo. Am J Anat. 1990; 188: 109–120. [Order article via Infotrieve]
Ratajska A, Fiejka E. Prenatal development of coronary arteries in the rat: morphologic patterns. Anat Embryol (Berl). 1999; 200: 533–540. [Order article via Infotrieve]
Ando K, Nakajima Y, Yamagishi T, Yamamoto S, Nakamura H. Development of proximal coronary arteries in quail embryonic heart: multiple capillaries penetrating the aortic sinus fuse to form main coronary trunk. Circ Res. 2004; 94: 346–352.
Tomanek RJ, Haung L, Suvarna PR, O’Brien LC, Ratajska A, Sandra A. Coronary vascularization during development in the rat and its relationship to basic fibroblast growth factor. Cardiovasc Res. 1996; 31 Spec No: E116–E126. [Order article via Infotrieve]
Tomanek RJ, Lotun K, Clark EB, Suvarna PR, Hu N. VEGF and bFGF stimulate myocardial vascularization in embryonic chick. Am J Physiol. 1998; 274: H1620–H1626.
Tomanek RJ, Ratajska A, Kitten GT, Yue X, Sandra A. Vascular endothelial growth factor expression coincides with coronary vasculogenesis and angiogenesis. Dev Dyn. 1999; 215: 54–61. [Order article via Infotrieve]
Yue X, Tomanek RJ. Stimulation of coronary vasculogenesis/angiogenesis by hypoxia in cultured embryonic hearts. Dev Dyn. 1999; 216: 28–36. [Order article via Infotrieve]
Yue X, Tomanek RJ. Effects of VEGF(165) and VEGF(121) on vasculogenesis and angiogenesis in cultured embryonic quail hearts. Am J Physiol Heart Circ Physiol. 2001; 280: H2240–H2247.
Tomanek RJ, Holifield JS, Reiter RS, Sandra A, Lin JJ. Role of VEGF family members and receptors in coronary vessel formation. Dev Dyn. 2002; 225: 233–240. [Order article via Infotrieve]
Tomanek RJ, Zheng W, Peters KG, Lin P, Holifield JS, Suvarna PR. Multiple growth factors regulate coronary embryonic vasculogenesis. Dev Dyn. 2001; 221: 265–273. [Order article via Infotrieve]
Holifield JS, Arlen AM, Runyan RB, Tomanek RJ. TGF-beta(1), -beta(2) and -beta(3) cooperate to facilitate tubulogenesis in the explanted quail heart. J Vasc Res. 2004; 41: 491–498. [Order article via Infotrieve]
Hoeben A, Landuyt B, Highley MS, Wildiers H, Van Oosterom AT, De Bruijn EA. Vascular endothelial growth factor and angiogenesis. Pharmacol Rev. 2004; 56: 549–580.
Carmeliet P. The development of blood vessels: cellular and molecular mechanisms. In: Tomanek RJ, ed. Assembly of the Vasculature and Its Regulation. Boston: Birkhauser; 2002: 1–20.
Mikawa T, Fischman DA, Dougherty JP, Brown AM. In vivo analysis of a new lacZ retrovirus vector suitable for cell lineage marking in avian and other species. Exp Cell Res. 1991; 195: 516–523. [Order article via Infotrieve]
Hyer J, Mikawa T. Retroviral techniques for studying organogenesis with a focus on heart development. Mol Cell Biochem. 1997; 172: 23–35. [Order article via Infotrieve]
Ishii Y, Reese DE, Mikawa T. Somatic transgenesis using retroviral vectors in the chicken embryo. Dev Dyn. 2004; 229: 630–642. [Order article via Infotrieve]
Watanabe S, Temin HM. Construction of a helper cell line for avian reticuloendotheliosis virus cloning vectors. Mol Cell Biol. 1983; 3: 2241–2249.
Velkey JM, Bernanke DH. Apoptosis during coronary artery orifice development in the chick embryo. Anat Rec. 2001; 262: 310–317. [Order article via Infotrieve]
Poelmann RE, Gittenberger-de Groot AC, Mentink MM, Bokenkamp R, Hogers B. Development of the cardiac coronary vascular endothelium, studied with antiendothelial antibodies, in chicken-quail chimeras. Circ Res. 1993; 73: 559–568.
Perez-Pomares JM, Macias D, Garcia-Garrido L, Munoz-Chapuli R. Immunolocalization of the vascular endothelial growth factor receptor-2 in the subepicardial mesenchyme of hamster embryos: identification of the coronary vessel precursors. Histochem J. 1998; 30: 627–634. [Order article via Infotrieve]
Kanda S, Miyata Y, Kanetake H. Fibroblast growth factor-2-mediated capillary morphogenesis of endothelial cells requires signals via Flt-1/vascular endothelial growth factor receptor-1: possible involvement of c-Akt. J Biol Chem. 2004; 279: 4007–4016.
Hood LC, Rosenquist TH. Coronary artery development in the chick: origin and deployment of smooth muscle cells, and the effects of neural crest ablation. Anat Rec. 1992; 234: 291–300. [Order article via Infotrieve]
Waldo KL, Kumiski DH, Kirby ML. Association of the cardiac neural crest with development of the coronary arteries in the chick embryo. Anat Rec. 1994; 239: 315–331. [Order article via Infotrieve]
Tevosian SG, Deconinck AE, Tanaka M, Schinke M, Litovsky SH, Izumo S, Fujiwara Y, Orkin SH. FOG-2, a cofactor for GATA transcription factors, is essential for heart morphogenesis and development of coronary vessels from epicardium. Cell. 2000; 101: 729–739. [Order article via Infotrieve]
Lie-Venema H, Gittenberger-de Groot AC, van Empel LJ, Boot MJ, Kerkdijk H, de Kant E, DeRuiter MC. Ets-1 and Ets-2 transcription factors are essential for normal coronary and myocardial development in chicken embryos. Circ Res. 2003; 92: 749–756.
Pepper M, Mandriota S, Jeltsch M, Kumar V, Alitalo K. Vascular endothelial growth factor (VEGF)-C synergizes with basic fibroblast growth factor and VEGF in the induction of angiogenesis in vitro and alters endothelial cell extracellular proteolytic activity. J Cell Physiol. 1998; 177: 439–452. [Order article via Infotrieve]
Lymboussaki A, Olofsson B, Eriksson U, Alitalo K. Vascular endothelial growth factor (VEGF) and VEGF-C show overlapping binding sites in embryonic endothelia and distinct sites in differentiated adult endothelia. Circ Res. 1999; 85: 992–999.
Silvestre JS, Tamarat R, Ebrahimian TG, Le-Roux A, Clergue M, Emmanuel F, Duriez M, Schwartz B, Branellec D, Levy BI. Vascular endothelial growth factor-B promotes in vivo angiogenesis. Circ Res. 2003; 93: 114–123.
Lagercrantz J, Farnebo F, Larsson C, Tvrdik T, Weber G, Piehl F. A comparative study of the expression patterns for vegf, vegf-b/vrf and vegf-c in the developing and adult mouse. Biochim Biophys Acta. 1998; 1398: 157–163. [Order article via Infotrieve]
Aase K, von Euler G, Li X, Ponten A, Thoren P, Cao R, Cao Y, Olofsson B, Gebre-Medhin S, Pekny M, Alitalo K, Betsholtz C, Eriksson U. Vascular endothelial growth factor-B-deficient mice display an atrial conduction defect. Circulation. 2001; 104: 358–364.
Wright CE. Effects of vascular endothelial growth factor (VEGF)A and VEGFB gene transfer on vascular reserve in a conscious rabbit hindlimb ischaemia model. Clin Exp Pharmacol Physiol. 2002; 29: 1035–1039. [Order article via Infotrieve]
Autiero M, Waltenberger J, Communi D, Kranz A, Moons L, Lambrechts D, Kroll J, Plaisance S, De Mol M, Bono F, Kliche S, Fellbrich G, Ballmer-Hofer K, Maglione D, Mayr-Beyrle U, Dewerchin M, Dombrowski S, Stanimirovic D, Van Hummelen P, Dehio C, Hicklin DJ, Persico G, Herbert JM, Shibuya M, Collen D, Conway EM, Carmeliet P. Role of PlGF in the intra- and intermolecular cross talk between the VEGF receptors Flt1 and Flk1. Nat Med. 2003; 9: 936–943. [Order article via Infotrieve]
Donovan MJ, Lin MI, Wiegn P, Ringstedt T, Kraemer R, Hahn R, Wang S, Ibanez CF, Rafii S, Hempstead BL. Brain derived neurotrophic factor is an endothelial cell survival factor required for intramyocardial vessel stabilization. Development. 2000; 127: 4531–4540.
Tomanek RJ, Sandra A, Zheng W, Brock T, Bjercke RJ, Holifield JS. Vascular endothelial growth factor and basic fibroblast growth factor differentially modulate early postnatal coronary angiogenesis. Circ Res. 2001; 88: 1135–1141. [Order article via Infotrieve]
the Department of Anatomy & Cell Biology and the Cardiovascular Center (R.J.T., J.S.H., C.L.S., H.K.H.), Carver College of Medicine, The University of Iowa, Iowa City
Department of Cell Biology (Y.I., T.M.), Weill Medical College, Cornell University, New York, NY.
Abstract
This study tested the hypothesis that coronary tubulogenesis and coronary artery formation require VEGF family members. Quail embryos were injected with soluble vascular endothelial growth factor (VEGF) receptors R1 (Flt-1), R2 (Flk-1), R3 (Flt-4), VEGF-Trap (a chimera of R1 and R2), or neutralizing antibodies to VEGF-A, VEGF-B, or fibroblast growth factor (FGF)-2. Our data document that tubulogenesis is temporally dependent on multiple VEGF family members, because the early stage of tubulogenesis was markedly inhibited by VEGF-Trap and to a lesser extent by soluble VEGFR-1. Some inhibition of tubulogenesis was documented when anti-FGF-2, but not anti-VEGF-A, antibodies were injected at embryonic day 6 (E6). Most importantly, we found that VEGF-Trap injected at either E6 or E7 prevented the formation of coronary arteries. Soluble VEGFR-1 and soluble VEGFR-2 modified the formation of coronary arteries, whereas soluble VEGFR-3 was without effect. Antibodies to VEGF-B, but not VEGF-A, had a strong inhibitory effect on coronary artery development. The absence of coronary artery stems, and thus a functional coronary circulation, in the embryos injected with VEGF-Trap caused an accumulation of erythrocytes in the subepicardium and muscular interventricular septum. Using retroviral cell tagging, we showed that some of the erythrocytes in blood islands and small vascular tubes were progeny of the proepicardium. Thus, another salient finding of this study is the first definitive documentation of proepicardially derived hemangioblasts, which can differentiate into erythrocytes.
Key Words: FGF-2 VEGF-B vasculogenesis angiogenesis hemangioblast Flt-1 Flk-1
Introduction
Formation of the coronary vasculature requires a series of closely temporally and spatially regulated events. (For review, see Tomanek1 and Majesky.2) All of the cells that contribute to the coronary vasculature (endothelial, smooth muscle, pericytes, and fibroblasts) migrate to the heart from the proepicardium, a transitional structure located posterior to the septum transversum.3,4 The cells of the proepicardial organ form the epicardium and subepicardium and undergo epithelial-mesenchymal transformation (reviewed by Olivey et al5). Recent evidence shows that hematopoietic precursors, ie, CD45+ cells, are also present on the surface of the quail heart in blood islands before vascularization of the myocardium.6 Migration of angioblasts into the myocardium and their assembly into vascular tubes constitutes the process of vasculogenesis and is followed by tubular expansion via branching (angiogenesis). Subsequent to these events, a capillary-like network (peritruncal ring) surrounding the base of the outflow tract fuses and penetrates the aorta at 2 specific sites, recruits smooth muscle cells, and consequently forms the 2 main coronary arteries.7–9 The remainder of the endothelial strands in contact with the aorta disappear.10 It is only at this point in time, E8 to E9 in the quail,10 that the coronary vasculature is perfused by blood from the aorta. The development of a venous system, like that of the capillary network, also develops before the coronary perfusion from the aorta.1 Thus, these events are not flow dependent.
We have documented the roles of several key growth factors in the regulation of coronary tubulogenesis, defined as vasculo- and angiogenesis.11–13 Having established an association between tube formation and 2 key growth factors, ie, vascular endothelial growth factor (VEGF) and fibroblast growth factor (FGF)-2, in vivo,11–13 we focused on in vitro studies using explanted quail embryonic hearts, a system that enables the establishment of a tubular system on collagen gels.14–18 These studies revealed that tubulogenesis is dependent on multiple growth factors,17 including members of the VEGF family and their receptors.16 Our data suggested that the tyrosine kinase receptors VEGFR-1 (flt-1) as well as VEGFR-2 (flk-1) may play a role in tube formation in the embryonic heart. The former is a receptor for VEGF-A, VEGF-B, and placental growth factor (PlGF), whereas the latter binds VEGFs A, C, D, and E (reviewed by Hoeben et al19). Like VEGFs, FGFs play a role in the endothelial cell proliferation, migration, and survival. The importance of these and other growth factors, eg, angiopoietins, ephrins, and platelet-derived growth factor, in the regulation of vessel development has been reviewed.20
To determine the temporal effect of VEGF growth factors on tube formation and the subsequent formation of the coronary arteries, we conducted in ovo experiments to test 3 hypotheses: (1) the timing of growth factor expression is critical for coronary tubulogenesis; (2) penetration of the tubular network to form the 2 main coronary arteries is dependent on VEGF family members; and (3) proepicardially derived epicardial cells are precursors of blood cells in that they contribute to blood islands. Because we previously documented, in vitro, that FGF-2 plays a major role in tubulogenesis,16 we also included experiments on its role in ovo.
Materials and Methods
Antibodies and Soluble Receptors
Monoclonal neutralizing antibodies to VEGF-A and FGF-2 were provided by Dr T. Brock and Encysive Pharmaceutical Corp (Houston, Tex), and anti-VEGF-B was purchased from R&D (Minneapolis, Minn). Soluble VEGF receptors R1 (Flt-1), R2 (Flk-1), and R3 (Flt-4) were purchased from R&D. These reagents have been described previously in detail.15 VEGF-Trap, a receptor 1 and 2 fusion protein, comprises portions of the human VEGFR-1 and human VEGFR-2 extracellular domains fused to the constant region of human IgG1 and was provided by Regeneron (Tarrytown, NY). The purified protein has been shown to be >90% pure. VEGF-Trap binds ligands for either VEGFR-1 or VEGFR-2. We previously used this chimera for in vitro tubulogenesis.15
Experimental Procedures
Fertilized quail eggs (Coturnix japonica) were incubated for 6 or 7 days at 37.8°C and 80% humidity. The embryo was exposed via a window in a shell and an incision of the inner-shell membrane, as previously described for chicken.12 With the aid of a micromanipulator, neutralizing antibodies or soluble receptors (12 μg in 12 μL of saline) were injected into the vitelline vein, the shell was sealed with paraffin film, and the egg was returned to the incubator for an additional 2 days. Sham controls were injected with 12 μL of saline. In preliminary experiments, we found that doses of <6 μg were without effect. Our previous in vitro experiments revealed an optimal response with 12 μg added to the media. The hearts of the 8- or 9-day embryos were perfused with 4% paraformaldehyde and embedded in paraffin.
Establishing Presence of Coronary Arteries
Serial cross-sections (6 μm thick) from the basal 1/2 of the heart were stained with hemotoxylin and eosin and systematically scanned for the presence of major coronary arteries at their origin from the coronary ostia. Select sections from some hearts were stained with smooth muscle -actin and QH1 antibodies.
Immunofluorescence and Image Analysis
Deparaffinized sections were blocked by immersing in PBS containing 1% BSA for 1 hour at room temperature. Incubation in QH1 was at room temperature; after blocking with 1% BSA in normal goat serum, the sections were incubated in fluorescein isothiocyanate (FITC)-labeled goat anti-mouse IgG Fab. The slides were viewed with a Nikon fluorescence microscope, images were digitized, and microvascular volume density and vessel diameter were determined using Image Pro Plus software (Media Cybernectics). The QH1 antibody reaction demarcated microvessels. From these images, we determined vessel volume density, ie, cross-sectional areas of vessels/test area, in the compact region of the left ventricle.
Apoptosis Detection
To detect in situ apoptosis, we used ApopTag Detection Kit (Chemicon Int, Temecula, Calif) on paraformaldehyde-fixed and paraffin-embedded sections. This method of detection is based on TUNEL technology, ie, modification of genomic DNA using deoxynucleotidyl transferase for detection of positive cells via specific staining. Positive nuclei were visualized by FITC and all nuclei revealed by nuclear counterstain with 4',6-diamidino-2-phenylindole iodide (DAPI).
In Ovo Retroviral Tagging
A cell line producing CXL, a replication-defective spleen necrosis virus encoding lacZ,21 was generated as described elsewhere.22,23 The packaging cell line, D17.2G,24 was cotransfected with a retroviral plasmid, pCXL, and a neomycin-resistance plasmid, pSV40-neo. (For details, see the online data supplement available at http://circres.ahajournals.org.) Retroviral tagging was performed in Hanburger Hamilton (HH) stage 16 to 17 (E2.5) quail embryos by injecting the viral suspension through a small window in the egg shell, membrane, and chorion and amnion with a glass capillary attached to Picospritzer (General Value Corp, Fairfield, NJ). The eggs were resealed with parafilm and incubated at 37.5°C. After the embryos were allowed to develop until E6, a small window was made at the pointed end of the egg shell. Twelve micrograms of VEGF Trap or hFC was pressure injected into the vitelline vessel through a glass capillary. The embryos were allowed to develop until E8 and the heart was fixed in 2% paraformaldehyde and processed for X-Gal histochemistry as described previously.21 Serial cross-sections were cut on a microtome and stained with eosin or incubated in QH1.
Statistics
Group comparisons for thickness of the ventricular compact region and vascular volume density were based on the Dunnett multiple comparison test. The Wilcoxon 2-sample exact test was used to test for intergroup differences for the presence of 0, 1, or 2 coronary ostia. A value of P0.05 was considered statistically significance.
Results
Growth Factor Deprivation Affects Cardiac Phenotype
Two phenotypic changes related to growth factor deprivation were documented. First, the compact region of the myocardium was significantly thinner (by 33% to 50%) in all of the treatment groups compared with shams at E9, but not at E8 (see online data supplement). The thinner walls were attributable to fewer numbers of cardiac myocytes, as readily discerned in histological cross sections. The second phenotypic alteration, an accumulation of erythrocytes subepicardially, and sometimes in the muscular interventricular septum, was seen in all Trap-injected hearts. These changes are discussed later in this section.
Tubulogenesis Is Dependent on Temporally Expressed Multiple VEGF Family Members
To test the hypothesis that various VEGF family members play a role in tubulogenesis, initially, we injected VEGF-Trap, or soluble VEGFR-1, VEGFR-2, or VEGFR-3 into the vitelline vein at either E6 or E7 and studied the hearts at E8 or E9. VEGFs A and B and PIGF are ligands for R1, whereas A, C, and D are ligands for R2. R3 binds VEGF-C and VEGF-D. When VEGF-Trap was injected at E6 tubulogenesis (as indicated by vascular volume 2 days later) was inhibited by 65% (P<0.0001) (Figures 1 and 2A and 2B). However, when injection was delayed until E7 and tubulogenesis evaluated on E9, the degree of inhibition (25%) was less (P=0.07). This finding underscores the importance of the stage of tube formation as a factor in responsiveness to growth factors. Although soluble VEGFR-2 did not affect tubulogenesis when injected at either time point, soluble VEGFR-1 had the most pronounced effect (32% inhibition) when administered at E7. Soluble VEGFR-3 (Flt-4) had no effect on tubulogenesis (data not shown). We then injected anti-VEGF-A and anti-FGF-2 neutralizing antibodies because VEGF-A and FGF-2 are interdependent in their actions.17 Anti-FGF-2 tended to limit vascular growth when injected at E6, but the probability value (0.15) was beyond statistical significance. Values for anti-VEGF-A-treated embryos were similar to the shams. Measurement of vessel diameters of profiles cut in cross-section revealed that mean tubule diameter (μm) in sham embryos increased from 10.1±0.4 at E8 to 14.0±1.3 at E9. The lower volume density in the E8 Trap group was not attributable to vessel size, because mean diameter (10.7±1.4) was similar to the sham group. In contrast, the lower volume density in the E9 embryo hearts was attributable, at least in part, to smaller vessel diameters (10.2±0.6).
Formation of the Coronary Arteries Requires VEGF Family Members
To test the hypothesis that VEGF signaling is necessary for the capillary plexus at the root of the aorta to penetrate that vessel at 2 distinct points, thereby establishing the roots of the 2 main coronary arteries, we examined heart serial cross-sections of E8 or E9 embryos treated 2 days earlier (Figure 2). Sham-injected hearts showed that both coronary arteries were present in all E9 and at least 1 was present in E8 hearts (Figure 3). Consistent with the data on tubulogenesis, anti-VEGF-A had no effect on the formation of coronary artery stems when administered at either time point. In contrast, 8 of 9 E8 hearts injected with VEGF-Trap had no coronary arteries. Of 4 hearts injected at E7 and studied at E9, 3 lacked coronary arteries, whereas 1 had only 1. Penetration of the aorta by the vascular tubes encircling the root of the vessel was precluded by VEGF-Trap. As seen in Figure 2D, a blood island forms a slight indentation into the aorta but does not extend further into the wall, as verified by serial sections. VEGF-Trap treatment did not prevent formation of the capillary-like network at the base of the aorta. Thus, failure of this tubular network to penetrate the aorta was caused by a lack of signaling rather than absence of this microvascular component. In addition to precluding the formation of coronary arteries, VEGF-Trap also caused massive accumulations of erythrocytes in the ventricular walls (Figure 2). These accumulations occurred primarily in the subepicardium but, in some hearts, were seen in the central portion of the interventricular septum.
Data shown in Figure 3 also illustrate the effects of soluble VEGFR-1 and soluble VEGFR-2 and anti-VEGF and –FGF-2 on coronary artery stem formation. Soluble VEGFR-1 treatment tended to reduce the number of coronary arteries formed at E8. Three of 4 E8 hearts lacked coronary artery stems. Forty percent of the E9 hearts treated with either soluble receptor (R1 or R2) had only 1 coronary artery. To test the possibility that the inhibitory effects of soluble VEGFR-1 were attributable to a higher affinity than VEGFR-2, we injected 15 μg of the receptor into 4 embryos. Both E9 embryos had 2 coronaries, whereas in the 2 E8 embryos, 1 lacked coronary arteries and the other had developed both coronary artery stems. Tubulogenesis, as evidenced by vascular volume density, was similar to the sham-injected group. Thus, the higher dose did not alter the results. Because anti-VEGF-A had no effect on coronary artery formation, we increased its amount to 15 or 20 μg (data not shown). However, the results were not different. Accordingly, we conclude from these experiments that because VEGFR-1 binds VEGF-B and PlGF, these growth factors may contribute to coronary artery formation. In 7 embryos treated with soluble VEGFR-3, all had well-formed coronary arteries. A role for FGF-2 in coronary artery formation is suggested by the finding that inhibition of this growth factor between E7 and E9 resulted in the formation of only 1 coronary artery in 4 of 9 hearts. Anomalous coronary stems were also seen in aortae from embryos treated with either Flt-1 or Flk-1. Multiple coronary channels could be seen with a connection to a single coronary artery (Figure 4). This is in contrast to the single channels seen in sham E9 embryos. Apoptosis was not widespread in the ventricular walls of sham or treated hearts. Although apoptotic cells were occasionally found in the ventricular walls, they were more likely to occur in the subepicardium (Figure 4). Previous work has documented apoptosis of the periaortic vascular plexus during the initial formation of the coronary artery stems.25
Having found that anti-VEGF-A neutralizing antibodies did not prevent coronary artery formation, we tested the hypothesis that VEGF-B, a ligand for VEGFR-1, may play a crucial role. Although anti-VEGF-B (12 μg/mL) had no significant effect on tube formation, coronary artery stems were not formed in most of the embryos studied (Figure 5). None of the 4 E8 hearts developed coronary arteries, whereas 5 of 6 E9 hearts had only 1 coronary artery. Anti-VEGF-B also compromised growth of the compact region of the ventricle in E9 hearts (112±4 μm versus 260±9 in sham).
Proepicardial Cells as Hemangioblast Precursors
The finding that large accumulations of erythrocytes are present in the quail hearts deprived of VEGF suggested to us that these cells differentiate from precursors derived from the proepicardium. Origin from the circulation was ruled out because the coronary circulation was absent in these hearts. To test this hypothesis, we injected a retroviral tag encoding lacZ into the proepicardium of E2.5 (HH16 to HH17) quail (Figure 6). Data are based on 3 VEGF-Trap treated and 3 hFC-treated hearts, all of which were staged as HH31 or HH32 (approximately E7.5 to 8.0).
As seen in Figure 6, cells from the proepicardium could be found in association with accumulations of erythrocytes under the epicardium (as observed in whole mounts and sectioned hearts). In all of the hearts, we documented -galactosidase cells not only as mesothelial, smooth muscle, and endothelial cells but also erythrocytes. The latter occurred in blood islands and in small vascular tubes. In both VEGF-trap and hFC injected embryos, blood islands could be found with erythroblasts -galactose positive. Thus, our findings indicate that VEGF deprivation (which precludes a coronary circulation via inhibition of coronary artery formation) is associated with enhanced differentiation of erythrocytes from proepicardial cells.
Discussion
This study documents VEGF family members as essential regulators of both coronary tubulogenesis and arteriogenesis during in vivo embryonic development in quail. First, we show that (1) tubulogenesis and (2) the formation of the 2 main coronary arteries are dependent on multiple VEGF family members rather than on VEGF-A. This conclusion is based on the following findings: (1) that administration of the flt-1/flk-1 receptor chimera (VEGF-Trap) nearly completely inhibits tube formation and precludes the formation of the coronary arteries; and (2) that anti-VEGF-B, a ligand for VEGFR-1 (flt-1), significantly impairs coronary artery formation. Although we were unable to document a significant effect of anti-VEGF-B on in ovo tubulogenesis, we previously showed, in vitro, that anti-VEGF-B limits coronary vascular tube formation by approximately 65%.16 Second, using a retrovirus, we provide proof that at least some proepicardial cells are hemangioblasts because they differentiate into erythrocytes in myocardial blood islands. Additionally, our findings indicate that FGF-2 contributes to tubulogenesis, and plays a minor role in the initial stage of coronary artery formation. Moreover, our study indicates that VEGFs as well as FGF-2 are important for growth of the myocardium, because their inhibition between E7 and E9 limited growth of the ventricular compact region. We previously documented, in vitro, that these growth factors are interdependent for their optimal effects.17 Whereas FGF-2 has direct effects on the myocardium, the effects of VEGF may be via limiting those of FGF-2. As indicated in recent reviews,1,2 previous studies focused mainly on the proepicardium, epicardial-mesenchymal transformation, and vascular progenitor cells. The current study provides the first in vivo data on the growth factor regulation of (1) tubulogenesis and (2) establishment of the coronary artery stems.
Coronary Tubulogenesis
Progenitor cells from the proepicardium migrate to the epicardium and subepicardium where they differentiate into fibroblasts and endothelial and smooth muscle cells.3,26 Cells that become immunoreactive for VEGFR-2 (flk-1) increase in number27 and subsequently form vascular tubes in an epicardial to endocardial direction, as shown here and as previously documented in the rat.11 This vascular gradient correlates with the density of VEGF protein and VEGF mRNA transcripts.13 Moreover, as shown by our in vitro studies, tube formation by endothelial cells from embryonic hearts is enhanced by hypoxia, a response that is prevented by anti-VEGF neutralizing antibodies.14 Our previous work showed that tube formation is inhibited by 62% when neutralizing antibodies to FGF-2 are added to the embryonic heart explants and by approximately 30% with anti-VEGF-A or soluble Tie-2 receptor.17 In addition, we documented an interdependence of VEGF and FGF-2 and a dependence of VEGF on angiopoietins.17 These findings support the concept that multiple growth factors regulate tubulogenesis in the embryonic heart. Moreover, our current experiments show that the effectiveness of growth factors on tubulogenesis is dependent on their precise temporal expression. Thus, VEGF-Trap was most effective in curtailing tubulogenesis when administered at E6, a stage of early tubulogenesis. Administration of VEGF-Trap 1 day later, when tubulogenesis is more advanced, had a more mild effect.
A role for FGF-2 in coronary tubulogenesis has been previously demonstrated in our laboratory both in vivo11,12 and in vitro.17 Of interest in this regard is the recent report that indicates that FGF-2–mediated capillary morphogenesis requires signaling via Flt-1.28 This finding supports our in vitro data implicating VEGF-B and/or PIGF as an important contributor to coronary tubulogenesis.16 In summary, the current in vivo study supports the concept of multiple growth factor regulation and brings into focus the importance of the VEGF family because we have documented a 2/3 reduction in myocardial vascularization when VEGF-Trap is injected at an early stage of microvessel formation.
Coronary Artery Formation
Previous studies have established the importance of neural crest cells and parasympathetic neurons29,30 and transcription factors in the epicardium, eg, Fog-231 and Ets-1 and Ets-2,32 in the formation of the coronary arteries. The current study has documented the necessity of VEGF family members for (1) tubulogenesis and (2) capillary plexus penetration of the aortic root and formation of coronary arteries. The hypothesis that this family of growth factors provides the key signaling mechanism for coronary ostia and artery formation was based on our observation of high expression of VEGF in epicardial and subepicardial cells at the aortic root and the high density of VEGF-R2 and -R3 transcripts at the aortic sites of coronary artery roots.16 The premise that VEGF family members in addition to VEGF-A play a role in vascularization is supported by the finding that VEGF-C acts synergistically with VEGF-A on bovine aortic endothelial cells33 and is found in epicardial vessels.34 Furthermore, both 167 and 186 splice variants of VEGF-B have been shown to improve the ischemic score in the hindlimb,35 a finding that suggests growth of arterioles and/or arteries.
Our data clearly indicate that VEGF family members are required for the formation of the main coronary arteries via endothelial cell penetration of the aorta, because VEGF-Trap prevented this event. The finding that soluble VEGFR-1 prevented coronary artery formation suggested that VEGF-B and/or PIFG may play a role, because these ligands do not directly activate VEGFR-2. A role for VEGF-B in the developing heart was suggested by its high expression in this organ36,37 and by our finding that anti-VEGF-B inhibits tubulogenesis in the explanted quail heart.15 VEGF-B, as well as VEGF-A, gene delivery also improves the hyperemic response in this model in rabbits.38 Both VEGF-B and PIGF have been shown to increase receptor crosstalk. The former overlaps with NRP1, and the latter has been shown to cause transphosphorylation of VEGFR-2 by activating VEGFR-1.39 The current study documents a key role for VEGF-B in the formation of the coronary artery ostia and stems. This novel finding reveals that the effects of this growth factor extend beyond its role in tubule formation.
Proepicardial Cells Include a Hemangioblast Population
Our observation that VEGF-Trap causes a massive increase in erythrocytes in the subepicardium suggested that these cells arise from epicardial-derived precursors. The presence of hematopoietic (CD45+) cells as early as HH23 (E3.5 to 4.0) in the quail subepicardium were previously reported.6 These cells preceded the occurrence of blood islands. Our study reveals that -galactosidase-positive cells from the proepicardium include erythrocytes, as well as endothelial, smooth muscle, and fibroblast cells. The positive cells were found in sham hearts injected with hFC as well as in those injected with VEGF-Trap. Accordingly, our data provide the first definitive evidence that erythrocytes, which are components of blood islands, arise from proepicardial cells. Moreover, our study indicates that VEGF deprivation is associated with an increase in differentiation of precursor cells into erythrocytes. A limitation of the retrovirus injection of the proepicardium is that only a small portion of the cells are tagged. It is, therefore, possible that erythrocytes could have been derived from another source. As previously noted, these cells could not have emerged from the circulation because they were observed in the absence of a coronary circulation. The only other avenue for their migration would be retrograde via the venous system. This possibility is not likely considering the poor development of a tubular network in VEGF-Trap E8 hearts.
Limitations of Study
Our experiments focused primarily on VEGF family members during 1 period of embryonic development. Therefore, no conclusions can be drawn with regard to other growth factors that may interact with VEGFs or play a role during other periods of development. For example, Donovan et al40 have shown that brain-derived neurotrophic factor is important for endothelial cell survival of intramyocardial arteries and capillaries during early postnatal life. Moreover, we have previously documented the interplay and differential modulation of early postnatal coronary angiogenesis by FGF-2 and VEGF.41
Acknowledgments
This work was supported by NIH grants HL075446 (R.J.T.) and HL078921 (T.M.).
Footnotes
Original received October 5, 2005; revision received February 21, 2006; accepted February 28, 2006.
References
Tomanek RJ. Formation of the coronary vasculature during development. Angiogenesis. 2005; 8: 273–284. [Order article via Infotrieve]
Majesky MW. Development of coronary vessels. Curr Top Dev Biol. 2004; 62: 225–259. [Order article via Infotrieve]
Mikawa T, Fischman DA. Retroviral analysis of cardiac morphogenesis: discontinuous formation of coronary vessels. Proc Natl Acad Sci U S A. 1992; 89: 9504–9508.
Mikawa T, Gourdie RG. Pericardial mesoderm generates a population of coronary smooth muscle cells migrating into the heart along with ingrowth of the epicardial organ. Dev Biol. 1996; 174: 221–232. [Order article via Infotrieve]
Olivey HE, Compton LA, Barnett JV. Coronary vessel development: the epicardium delivers. Trends Cardiovasc Med. 2004; 14: 247–251. [Order article via Infotrieve]
Kattan J, Dettman RW, Bristow J. Formation and remodeling of the coronary vascular bed in the embryonic avian heart. Dev Dyn. 2004; 230: 34–43. [Order article via Infotrieve]
Bogers AJ, Gittenberger-de Groot AC, Poelmann RE, Peault BM, Huysmans HA. Development of the origin of the coronary arteries, a matter of ingrowth or outgrowth Anat Embryol (Berl). 1989; 180: 437–441. [Order article via Infotrieve]
Waldo KL, Willner W, Kirby ML. Origin of the proximal coronary artery stems and a review of ventricular vascularization in the chick embryo. Am J Anat. 1990; 188: 109–120. [Order article via Infotrieve]
Ratajska A, Fiejka E. Prenatal development of coronary arteries in the rat: morphologic patterns. Anat Embryol (Berl). 1999; 200: 533–540. [Order article via Infotrieve]
Ando K, Nakajima Y, Yamagishi T, Yamamoto S, Nakamura H. Development of proximal coronary arteries in quail embryonic heart: multiple capillaries penetrating the aortic sinus fuse to form main coronary trunk. Circ Res. 2004; 94: 346–352.
Tomanek RJ, Haung L, Suvarna PR, O’Brien LC, Ratajska A, Sandra A. Coronary vascularization during development in the rat and its relationship to basic fibroblast growth factor. Cardiovasc Res. 1996; 31 Spec No: E116–E126. [Order article via Infotrieve]
Tomanek RJ, Lotun K, Clark EB, Suvarna PR, Hu N. VEGF and bFGF stimulate myocardial vascularization in embryonic chick. Am J Physiol. 1998; 274: H1620–H1626.
Tomanek RJ, Ratajska A, Kitten GT, Yue X, Sandra A. Vascular endothelial growth factor expression coincides with coronary vasculogenesis and angiogenesis. Dev Dyn. 1999; 215: 54–61. [Order article via Infotrieve]
Yue X, Tomanek RJ. Stimulation of coronary vasculogenesis/angiogenesis by hypoxia in cultured embryonic hearts. Dev Dyn. 1999; 216: 28–36. [Order article via Infotrieve]
Yue X, Tomanek RJ. Effects of VEGF(165) and VEGF(121) on vasculogenesis and angiogenesis in cultured embryonic quail hearts. Am J Physiol Heart Circ Physiol. 2001; 280: H2240–H2247.
Tomanek RJ, Holifield JS, Reiter RS, Sandra A, Lin JJ. Role of VEGF family members and receptors in coronary vessel formation. Dev Dyn. 2002; 225: 233–240. [Order article via Infotrieve]
Tomanek RJ, Zheng W, Peters KG, Lin P, Holifield JS, Suvarna PR. Multiple growth factors regulate coronary embryonic vasculogenesis. Dev Dyn. 2001; 221: 265–273. [Order article via Infotrieve]
Holifield JS, Arlen AM, Runyan RB, Tomanek RJ. TGF-beta(1), -beta(2) and -beta(3) cooperate to facilitate tubulogenesis in the explanted quail heart. J Vasc Res. 2004; 41: 491–498. [Order article via Infotrieve]
Hoeben A, Landuyt B, Highley MS, Wildiers H, Van Oosterom AT, De Bruijn EA. Vascular endothelial growth factor and angiogenesis. Pharmacol Rev. 2004; 56: 549–580.
Carmeliet P. The development of blood vessels: cellular and molecular mechanisms. In: Tomanek RJ, ed. Assembly of the Vasculature and Its Regulation. Boston: Birkhauser; 2002: 1–20.
Mikawa T, Fischman DA, Dougherty JP, Brown AM. In vivo analysis of a new lacZ retrovirus vector suitable for cell lineage marking in avian and other species. Exp Cell Res. 1991; 195: 516–523. [Order article via Infotrieve]
Hyer J, Mikawa T. Retroviral techniques for studying organogenesis with a focus on heart development. Mol Cell Biochem. 1997; 172: 23–35. [Order article via Infotrieve]
Ishii Y, Reese DE, Mikawa T. Somatic transgenesis using retroviral vectors in the chicken embryo. Dev Dyn. 2004; 229: 630–642. [Order article via Infotrieve]
Watanabe S, Temin HM. Construction of a helper cell line for avian reticuloendotheliosis virus cloning vectors. Mol Cell Biol. 1983; 3: 2241–2249.
Velkey JM, Bernanke DH. Apoptosis during coronary artery orifice development in the chick embryo. Anat Rec. 2001; 262: 310–317. [Order article via Infotrieve]
Poelmann RE, Gittenberger-de Groot AC, Mentink MM, Bokenkamp R, Hogers B. Development of the cardiac coronary vascular endothelium, studied with antiendothelial antibodies, in chicken-quail chimeras. Circ Res. 1993; 73: 559–568.
Perez-Pomares JM, Macias D, Garcia-Garrido L, Munoz-Chapuli R. Immunolocalization of the vascular endothelial growth factor receptor-2 in the subepicardial mesenchyme of hamster embryos: identification of the coronary vessel precursors. Histochem J. 1998; 30: 627–634. [Order article via Infotrieve]
Kanda S, Miyata Y, Kanetake H. Fibroblast growth factor-2-mediated capillary morphogenesis of endothelial cells requires signals via Flt-1/vascular endothelial growth factor receptor-1: possible involvement of c-Akt. J Biol Chem. 2004; 279: 4007–4016.
Hood LC, Rosenquist TH. Coronary artery development in the chick: origin and deployment of smooth muscle cells, and the effects of neural crest ablation. Anat Rec. 1992; 234: 291–300. [Order article via Infotrieve]
Waldo KL, Kumiski DH, Kirby ML. Association of the cardiac neural crest with development of the coronary arteries in the chick embryo. Anat Rec. 1994; 239: 315–331. [Order article via Infotrieve]
Tevosian SG, Deconinck AE, Tanaka M, Schinke M, Litovsky SH, Izumo S, Fujiwara Y, Orkin SH. FOG-2, a cofactor for GATA transcription factors, is essential for heart morphogenesis and development of coronary vessels from epicardium. Cell. 2000; 101: 729–739. [Order article via Infotrieve]
Lie-Venema H, Gittenberger-de Groot AC, van Empel LJ, Boot MJ, Kerkdijk H, de Kant E, DeRuiter MC. Ets-1 and Ets-2 transcription factors are essential for normal coronary and myocardial development in chicken embryos. Circ Res. 2003; 92: 749–756.
Pepper M, Mandriota S, Jeltsch M, Kumar V, Alitalo K. Vascular endothelial growth factor (VEGF)-C synergizes with basic fibroblast growth factor and VEGF in the induction of angiogenesis in vitro and alters endothelial cell extracellular proteolytic activity. J Cell Physiol. 1998; 177: 439–452. [Order article via Infotrieve]
Lymboussaki A, Olofsson B, Eriksson U, Alitalo K. Vascular endothelial growth factor (VEGF) and VEGF-C show overlapping binding sites in embryonic endothelia and distinct sites in differentiated adult endothelia. Circ Res. 1999; 85: 992–999.
Silvestre JS, Tamarat R, Ebrahimian TG, Le-Roux A, Clergue M, Emmanuel F, Duriez M, Schwartz B, Branellec D, Levy BI. Vascular endothelial growth factor-B promotes in vivo angiogenesis. Circ Res. 2003; 93: 114–123.
Lagercrantz J, Farnebo F, Larsson C, Tvrdik T, Weber G, Piehl F. A comparative study of the expression patterns for vegf, vegf-b/vrf and vegf-c in the developing and adult mouse. Biochim Biophys Acta. 1998; 1398: 157–163. [Order article via Infotrieve]
Aase K, von Euler G, Li X, Ponten A, Thoren P, Cao R, Cao Y, Olofsson B, Gebre-Medhin S, Pekny M, Alitalo K, Betsholtz C, Eriksson U. Vascular endothelial growth factor-B-deficient mice display an atrial conduction defect. Circulation. 2001; 104: 358–364.
Wright CE. Effects of vascular endothelial growth factor (VEGF)A and VEGFB gene transfer on vascular reserve in a conscious rabbit hindlimb ischaemia model. Clin Exp Pharmacol Physiol. 2002; 29: 1035–1039. [Order article via Infotrieve]
Autiero M, Waltenberger J, Communi D, Kranz A, Moons L, Lambrechts D, Kroll J, Plaisance S, De Mol M, Bono F, Kliche S, Fellbrich G, Ballmer-Hofer K, Maglione D, Mayr-Beyrle U, Dewerchin M, Dombrowski S, Stanimirovic D, Van Hummelen P, Dehio C, Hicklin DJ, Persico G, Herbert JM, Shibuya M, Collen D, Conway EM, Carmeliet P. Role of PlGF in the intra- and intermolecular cross talk between the VEGF receptors Flt1 and Flk1. Nat Med. 2003; 9: 936–943. [Order article via Infotrieve]
Donovan MJ, Lin MI, Wiegn P, Ringstedt T, Kraemer R, Hahn R, Wang S, Ibanez CF, Rafii S, Hempstead BL. Brain derived neurotrophic factor is an endothelial cell survival factor required for intramyocardial vessel stabilization. Development. 2000; 127: 4531–4540.
Tomanek RJ, Sandra A, Zheng W, Brock T, Bjercke RJ, Holifield JS. Vascular endothelial growth factor and basic fibroblast growth factor differentially modulate early postnatal coronary angiogenesis. Circ Res. 2001; 88: 1135–1141. [Order article via Infotrieve]
您现在查看是摘要介绍页，详见ORG附件。