Fibroblast Growth Factor Receptor-1 Expression Is Required for Hematopoietic but not Endothelial Cell Development
http://www.100md.com
动脉硬化血栓血管生物学 2005年第5期
From the Department of Genetics and Pathology (P.U.M., P.C., L.J., A.D., L.C.-W.), Rudbeck Laboratory, Uppsala University, Uppsala, Sweden; the Unit of General Pathology and Immunology (R.R., P.D.E.), Department of Biomedical Sciences and Biotechnology, University of Brescia, Brescia, Italy; and the Institute of Biotechnology (J.P.), University of Helsinki, Helsinki, Finland
Correspondence to Lena Claesson-Welsh, Department of Genetics and Pathology, Uppsala University, The Rudbeck Laboratory, Dag Hammarskj?ldsv. 20, 751 85 Uppsala, Sweden. E-mail Lena.Welsh@genpat.uu.se
Abstract
Objective— The purpose of this study was to clarify the role of fibroblast growth factors (FGFs) and FGF receptors (FGFRs) in hematopoietic/endothelial development.
Methods and Results— Using several different FGFR-1–specific antibodies and FGFR-1 promoter-driven LacZ activity, we show that FGFR-1 is expressed and active as a tyrosine kinase in a subpopulation of endothelial cells (20% of the endothelial pool) during development in embryoid bodies. In agreement, in stem cell-derived teratomas, expression of FGFR-1 was detected in some but not all vessels. The FGFR-1 expressing endothelial cells were mitogenically active in the absence and presence of vascular endothelial growth factor (VEGF). Expression of FGFR-1 in endothelial cell precursors was not required for vascular development, and vascularization was enhanced in FGFR-1–deficient embryoid bodies compared with wild-type stem cells. In contrast, hematopoietic development was severely disturbed, with reduced expression of markers for primitive and definitive hematopoiesis.
Conclusions— Our data show that FGFR-1 is expressed in early hematopoietic/endothelial precursor cells, as well as in a subpool of endothelial cells in tumor vessels, and that it is critical for hematopoietic but not for vascular development.
Immunostaining using several different antibodies and scoring for FGFR-1 promoter-driven Lac Z activity show that some but not all endothelial cells express FGFR-1. FGFR-1 gene inactivation leads to a block in hematopoietic development; in contrast, endothelial cell development is enhanced.
Key Words: angiogenesis ? endothelial cells ? FGF receptor-1 ? hematopoiesis ? vasculogenesis
Introduction
Fibroblast growth factors (FGF) constitute a family of at least 23 related heparin-binding growth factors with vital and broad functions in health and disease. There are 4 distinct FGF receptor (FGFR) tyrosine kinases, denoted FGFR-1, -2, -3, and -4.1 Binding of FGF to its receptor leads to receptor dimerization and activation of the kinase, followed by tyrosine phosphorylation of the receptor itself, as well as of downstream signal transduction molecules.2
See page 883 and
FGF-2 was the first angiogenic growth factor to be identified,3 and endothelial cells in culture respond mitogenically to FGF-2 mainly through FGFR-1.4 The question of whether endothelial cells in vivo express FGFR-1 and respond to various FGFs has remained a matter of debate. There is no clear vascular phenotype in FGF-2 or FGFR-1 knockout animals. Gene inactivation of FGFR-1 leads to embryonic death in conjunction with gastrulation between embryonic days 7.5 to 9.5.5,6 Embryos homozygous-null for FGFR-1 expression display severe reduction in paraxial mesoderm formation, whereas development of node, axial, and extraembryonic mesoderm is unaffected. Gene inactivation of FGF-2, a ligand for FGFR-1, is compatible with normal development, although the gene-targeted animals display decreased vascular tone and low blood pressure.7 The possibility of compensation through any of the other FGFs or FGFRs cannot be excluded.
Vessels are established through vasculogenesis in embryonic development at approximately day 7.5 in the mouse. The vital role for vascular endothelial growth factor (VEGF)-A and its receptors VEGFR-1 and -2 for blood vascular development has been clearly demonstrated by gene inactivation.8 Cross-talk between the VEGF and FGF families of growth factors have been described; expression of VEGF-A and VEGFR-2 may be induced by FGF/FGFRs in vitro9 and in vivo.10–12 The purpose of this study was to determine whether FGFR-1 is expressed in endothelial cells using several different strategies and high-quality models.
Materials and Methods
Embryonic Stem Cell and Embryoid Body Culture
The following murine embryonic stem (ES) cell lines were used: R1,13 either wild-type or clone FGFR-1+/lacZfgfr-1; J1,6 homozygous or heterozygous for FGFR-1 gene inactivation; or clone Lentivirus-human FGFR-1 (Lv–hFGFR-1), in which FGFR-1 was reintroduced. For creation of FGFR-1+/lacZfgfr-1, the LacZ expression cassette was excised from internal ribosomal entry site (IRES)-LacZ and inserted into exon 13 in the pPNT–FGFR-1 vector encompassing FGFR-1 exons 11 to 17, corresponding to part of the tyrosine kinase domain. The FGFR-1/LacZ cDNA was integrated into R1 Sv129 by homologous recombination. For generation of Lv–hFGFR-1 stem cells, see http://atvb.ahajournals.org. The ES cell lines were cultured as described.14
Magnetic Cell Sorting to Enrich for Endothelial Cells
Embryoid bodies were dissociated by incubation in 0.25% collagenase and the single cell suspension was incubated with rat anti-mouse CD31 antibody (Becton Dickinson Biosciences), followed by mixing with magnetic cell sorting goat anti-rat IgG micro beads (Miltenyi Biotech). The CD31+ eluate was then incubated with mouse monoclonal FGFR-1 antibody (QED Bioscience) and goat anti-mouse IgG microbeads to isolate a CD31+FGFR-1+ fraction. Total RNA was extracted using the Qiagen RNeasy mini kit.
Fluorescence Activated Cell Sorting and Cell Cycle Assays
Embryoid bodies were dissociated in 0.25% collagenase. The single-cell suspension was incubated with mouse FGFR-1 monoclonal antibody (QED Bioscience) and secondary Alexa 488 goat anti-mouse highly cross-absorbed IgG (Molecular Probes). The cells were then stained with rat anti-mouse CD31-phycoerythrin antibody (Becton Dickinson). To identify the cell cycle profile, 5 μmol/L Hoechst 33342 was added to the single cell suspension in DMEM with 15% fetal bovine serum during 1 hour followed by antibody staining as described. Viable cells were identified by use of propidium iodide. The results were analyzed on a FACSVantage SE DiVa machine (Becton Dickinson) equipped with an Enterprise laser (Coherent Inc) for ultraviolet and 488 nm excitation using ModFit software (Verity Software House, Inc).
Embryoid Bodies in 3-Dimensional Collagen I Gels
Collagen gels were prepared by mixing 10x Ham’s F12 medium (Invitrogen) with 0.12% NaHCO3, 50 mmol/L HEPES, 5 mmol/L NaOH, and 1.5 mg/mL collagen I (Cohesion). Embryoid bodies (day 4 after leukemia inhibitory factor [LIF] withdrawal) were distributed on the polymerized collagen, followed by instant covering with a second layer of collagen. Medium containing 30 ng/mL of VEGF-A was added and culture continued for 8 days. To visualize the FGFR-1 promoter-driven ?-galactosidase activity, FGFR-1+/lacZfgfr-1 embryoid bodies in 3-dimensional collagen gel were fixed in 4% p-formaldehyde for 30 minutes, followed by overnight incubation at 37°C in 5 mmol/L K-ferricyanide, 5 mmol/L K-ferrocyanide, 4 mmol/L MgCl2, and 1.25 mg/mL X-gal. Incorporation of bromodeoxyuridine (BrdU) was visualized using anti-BrdU+nuclease kit (Amersham Pharmacia Biotech) and secondary antibody Alexa 488 goat anti-mouse highly cross-absorbed IgG (Molecular Probes). Nuclei were visualized by incorporation of Hoechst 33342 (1 μg/mL).
Preparing, Sectioning, and Staining of Teratomas
1x106 R1 (FGFR-1+/+) or J1 (FGFR-1+/–) ES cells were injected subcutaneously on the back flank of female Naval Medical Research Institute (NMRI)-nu mice (n=13; M&B Animal Models) and teratomas were grown for 30 days. Animal handling was performed with ethical permission approved by the Uppsala University ethics committee and according to the United Kingdom Coordinating Committee Cancer Research (UKCCCR) guidelines for the welfare of animals in experimental neoplasia.15 Acetone fixed frozen 6-μm sections immunostained with rabbit anti-mouse FGFR-1 (sc-121 and sc-4975; Santa Cruz Biotechnology) and mouse monoclonal FGFR-1 (30104; QED Bioscience Inc) were analyzed by confocal laser-scanning microscope (Zeiss LSM 510 META) or by use of a Nikon Eclipse E1000 microscope. Please see http://atvb.ahajournals.org.
Results
FGFR-1 Expression in Endothelial Cells During Development
We analyzed expression of FGFR-1 in endothelial cells during development in differentiating ES cell cultures, which display several hallmarks of vascular development in mouse embryos.14 Embryoid bodies (strain R1) were created by aggregation of stem cells and seeded individually in 8-well chambers in the presence of 20 ng/mL FGF-2 for 8 days, to induce vessel formation. Coimmunostaining showed that expression of FGFR-1 overlapped with that of CD31 on endothelial cells (Figure 1A). In embryoid bodies created from ES cells homozygous-null for FGFR-1, the anti–FGFR-1 antibody gave rise to an even, low background without vessel-specific staining (Figure 1B); in contrast, immunostaining for CD31 showed vessel-like structures. This result demonstrates the specificity of the anti–FGFR-1 antibody. To show that the FGFR-1 was functional in endothelial cells, an antibody was used that is specific for the activated receptor, phosphorylated at Y653 in the kinase domain. Staining for Y653-phosphorylated FGFR-1 colocalized with CD31 expression in endothelial cells (Figure 1C). There was no specific immunostaining by the Yp653 antibody in FGFR-1–/– embryoid bodies (Figure 1D).
Figure 1. Expression of FGFR-1 in embryoid bodies and in purified endothelial cells. Whole-mount immunostaining of day 8 wild-type R1 embryoid bodies (A, C) for CD31 or FGFR-1 (antibodies directed against the N-terminal part or against FGFR-1 phosphorylated at Y653). Staining of FGFR-1–/– embryoid bodies (B, D) was used as a negative control. Bars, 100 μm (A, C), 200 μm (B, D). E. Semi-quantitative PCR for CD31, FGFR-1, VEGFR-2, and VE-cadherin transcript expression in CD31+FGFR-1+ purified endothelial cells. Analysis of -fetoprotein and ?-actin transcript expression in flow-through (lane 1) and CD31+FGFR-1+ cell pool (lane 2). F, Fluorescence-activated cell sorting analysis of CD31 and FGFR-1 expression (blue) in embryoid bodies. The insert (red) shows background fluorescence of IgG-phycoerythrin and IgG-488. An average of 4 individual experiments is presented. G, Cell cycle profile by staining for DNA content using Hoechst 33342 of the dispersed CD31/FGFR-1 antibody-labeled cells. Figures indicate the fraction of cells in the different cell cycle phases, in percent.
To further evaluate expression of FGFR-1 in endothelial cells, day 8 wild-type embryoid bodies were dissociated and magnetic beads were used to purify CD31+FGFR-1+ coexpressing cells using a protocol ensuring that the proteins were expressed on the same cell. Semi-quantitative polymerase chain reaction (PCR) was performed to show that the CD31+FGFR-1+ also expressed the endothelial cell markers VEGFR-2 and VE-cadherin. The endodermal marker -fetoprotein served as a negative control and was present in the flow through (CD31–FGFR-1–) from the magnetic bead separation, but absent in the CD31+FGFR-1+ pool (Figure 1E).
To allow analysis of cell cycle parameters of the CD31+FGFR-1+ cells, fluorescence-activated cell sorting analyses were performed on the collagenase-digested embryoid bodies. A relatively small pool of viable CD31+FGFR-1+ cells was identified; 3% of all cells (Figure 1F), possibly representing more than one population of cells. In total, 12% of all viable cells expressed CD31 and 23% expressed FGFR-1. Staining with Hoechst 33342 allowed visualization of the cell cycle profile in the different fractions. As shown in Figure 1G, 70% of the CD31+FGFR-1+ cells were in the S/G2 phase, indicating that a larger proportion of these cells are progressing through the cell cycle, compared with cells individually expressing either CD31 or FGFR-1. Staining for expression of Annexin V indicated that there was no significant difference in the extent of apoptosis in the different cell pools (data not shown).
FGFR-1 Promoter-Driven LacZ Expression Demonstrates FGFR-1 Expression in Endothelial Cells
To analyze expression of FGFR-1 in endothelial cells independently of FGFR-1 antibodies, we created an ES cell line expressing ?-galactosidase controlled by the FGFR-1 promoter (FGFR-1+/LacZfgfr-1). This strain of ES cells behaved similar to wild-type R1 stem cells with regard to proliferation rate and FGF-2-driven DNA synthesis (Figure I, available online at http://atvb.ahajournals.org). To facilitate analysis of individual endothelial cells, we used a sprouting angiogenesis assay in which embryoid bodies are placed in 3-dimensional collagen gels. Inclusion of VEGF-A is required to induce sprouting of vessel-like structure into the gel16 (Figure 2A). FGFR-1+/LacZfgfr-1 embryoid bodies were cultured between days 4 and 12 in 3-dimensional collagen and ?-galactosidase activity was identified by X-gal staining. Costaining with anti-CD31 and anti–-smooth muscle cell actin antibodies was performed to identify pericyte-coated vessel sprouts invading the collagen gels. Some but not all endothelial cells contained ?-galactosidase activity as a result of FGFR-1 promoter activity (Figure 2B). By analyzing the distal ends of sprouts where cells clearly could be classified as single endothelial cells, the number of CD31/?-galactosidase–positive cells (ie, CD31+FGFR-1+ cells) was estimated to 15% of all endothelial cells in the sprouts. This indicates a lower fraction of coexpressing cells in the angiogenic sprouts than in the total CD31-positive pool (25%; see Figure 1F), however, the fluorescence-activated cell sorting-derived value may be influenced by the fact that CD31 is expressed also on hematopoietic cells. Moreover, in the sprouting assay, only endothelial cells in the tip of the sprouts could be clearly identified. Fluorescence-activated cell sorting analyses combining the only commercial FGFR-1 antibody of several tested that allowed specific visualization of the FGFR-1–expressing cells with antibodies against other endothelial cell markers were not technically feasible because of specificity of the reagents.
Figure 2. Detection of FGFR-1 expression in endothelial cells using FGFR-1+/lacZfgfr-1 ES cells. A, 3-dimensional culture of embryoid body in collagen gel in the presence of VEGF-A induced angiogenic sprouting visualized by CD31 staining. Bar, 1 mm. B, VEGF-A–induced sprouting by FGFR-1+/lacZfgfr-1 embryoid body in 3-dimensional collagen gel shows FGFR-1 expressing (arrow) or nonexpressing (arrowhead) endothelial cells. Left panel, Fluorescence staining, red; CD31, green; -smooth muscle cell actin, purple; Hoechst staining. Right panel, Light microscopy in which X-gal stain (blue) marks FGFR-1 promoter-positive cells. Bars, 200 μm. C, Left panel shows fluorescence staining for BrdU (green) and CD31 (red); right panel shows X-gal staining (blue). BrdU-positive CD31+FGFR-1+ (ring marked) and CD31+FGFR-1– (arrowhead) endothelial cells in FGFR-1+/lacZfgfr-1-derived sprouts. Bars, 50 μm.
Data in Figure 1G showed that the CD31+FGFR-1+ cells were actively going through the cell cycle, whereas cells individually expressing CD31 or FGFR-1 were resting to a higher extent. We wished to test the ability of the cell pools to respond mitogenically to VEGF-A. Embryoid bodies in 3-dimensional collagen gels were treated with VEGF-A for different periods of time; BrdU was added during the last 24 hours of culture. The embryoid bodies were then stained to visualize CD31/?-galactosidase–positive cells (Figure 2C shows 10-day-old embryoid bodies). Quantification by counting the number of X-gal–stained and BrdU-stained endothelial cells in the angiogenic sprouts showed the fraction of mitogenically active CD31+ endothelial cells was higher initially at day 6 (50%) and decreased with time in conjunction with stabilization of the endothelial sprouts (30%). The mitogenic activity of CD31/?-galactosidase–positive cells showed the same pattern with an initial activity in 50% of the CD31+FGFR-1+ pool followed by a decrease to 25% to 30% (data not shown). These data show that the total endothelial cell pool responds to exogenous VEGF-A with induction of a mitogenic response also engaging the CD31+FGFR-1+ cells. However, in the absence of exogenous VEGF-A, the mitogenic activity is markedly higher in the CD31+FGFR-1+ pool than in the total CD31+ pool (Figure 1G).
FGFR-1 Is Expressed in Teratoma Vessels
We next asked whether FGFR-1 is expressed on vessels in vivo. For this purpose, we analyzed vessels in teratomas, created by inoculation of immune-deficient mice with J1 or R1 ES cells expressing FGFR-1. As shown in Figure 3A (arrowheads), CD31/FGFR-1 coexpressing cells were readily detected in the J1 teratomas by use of an antibody reactive with the intracellular domain of FGFR-1. The result was further verified by use of a second commercial antibody against the FGFR-1 extracellular domain to stain sections of R1 teratomas. The FGFR-1 immunostaining colocalized with that for CD31 in vessel structures of the teratoma (Figure 3B). A third commercial FGFR-1 antibody specific for the unique NH2-terminal IgG loop of the receptor also costained CD31-positive vessel-like structures in R1 teratomas (Figure 3C). We estimated that 25% of the endothelial cells in teratomas expressed FGFR-1, although the frequency varied between different regions of the tumors.
Figure 3. Immunostaining for FGFR-1 expression in mouse teratomas. A, FGFR-1 expression in J1 teratoma (green; antibody against FGFR-1 intracellular domain; Santa Cruz, sc-121). Arrowhead indicates coexpression with CD31. B, R1 teratoma (green; antibody against the extracellular domain of FGFR-1; Santa Cruz sc-4975). C, R1 teratoma (green; antibody against the unique NH2-terminal Ig loop of FGFR-1; QED Bioscience, mAb 30104). All panels show CD31 staining in red. Bars, 100 μm.
Function of FGFR-1 in Endothelial Cells During Development
Embryoid bodies derived from stem cells lacking one (+/–) or both (–/–) alleles of FGFR-1 still form blood vessels.14 In fact, even though VEGFR-2 expression is decreased in the FGFR-1–/– stem cells, blood vessels form to a considerably increased extent even in the absence of exogenous growth factors (Figure 4A). Thus, endothelial cell development still proceeds in the absence of FGFR-1; in contrast, hematopoietic development is attenuated.17 Development of the endothelial and hematopoietic lineages diverges downstream of a common mesodermal precursor cell, the hemangioblast.18 To pinpoint the level at which loss of FGFR-1 expression guides the common endothelial/hematopoietic development toward a preference for the endothelial lineage, expression of markers of primitive and definitive hematopoiesis, as well as endothelial cell markers, was investigated by real-time PCR in differentiating embryoid bodies (Figure 4B). Expression of brachyury, one of the earliest markers for mesoderm development,19 was slightly increased by day 4 in the absence of FGFR-1. Runx1, which is expressed at the earliest stage of blood island development,20 was not affected by elimination of FGFR-1 expression. Hemangioblasts are known to express brachyury, VEGFR-2, and Tal-1/SCL.17,21,22 In the absence of FGFR-1, there was a 50% reduction in VEGFR-2 transcripts, whereas the level of Tal-1 expression was <10% of that in control FGFR-1+/– stem cells. These data indicate that loss of FGFR-1 expression affects the endothelial/hematopoietic differentiation at the level of the hemangioblast. Further real-time PCR analyses showed that expression of the primitive hematopoietic marker CD4123 and the embryonic form of ?-globin, ?-H1, a marker for primitive erythropoiesis,20 was essentially lost. In accordance, there was little or no expression of GATA-1, a transcription factor expressed in hematopoietic but absent in endothelial cells,24 or of the pan hematopoietic marker CD4525 in FGFR-1–/– stem cells at day 8 and day 12, respectively. Despite changes on the hemangioblast level as indicated by decreased expression of VEGFR-2 and Tal-1/SCL, development of the endothelial cell lineage proceeded and expression of CD31 and VE-cadherin transcripts was higher or similar to the control in the FGFR-1–/– stem cells at day 8. Transcript levels of the general smooth muscle cell marker -smooth muscle actin was reduced (Figure 4B) in agreement with the impaired mesodermal development in the absence of FGFR-1 expression. Time course analyses at days 4, 8, and 12 (Figure II, available online at http://atvb.ahajournals.org) showed there was no rescue in expression of hematopoietic markers in the FGFR-1–/– embryoid bodies excluding that the FGFR-1–/– phenotype was caused by a delay in development.
Figure 4. Endothelial and hematopoietic development in FGFR-1 lacking embryoid bodies. A, Increased vascularization in FGFR-1–/– embryoid bodies compared with FGFR-1+/– control embryoid bodies as visualized by staining for VEGFR-2. B, Real-time PCR analysis of expression in FGFR-1+/– or FGFR-1–/– embryoid bodies of hematopoietic or endothelial cell markers at day 8, or for brachyury at day 4, and CD45 at day 12. C, Localization of staining for CD41 (red) and VE-cadherin (green) in blood island-like structures in FGFR-1+/– embryoid bodies. D, Localization of CD41-positive (red) cells inside VE-cadherin–positive (green) vessels. E, Immunofluorescent staining for CD41 (red; arrow) and VE-cadherin (green) expression in FGFR-1–/– embryoid bodies. Bars, 100 μm.
Immunostaining for CD41 and VE-cadherin in FGFR-1+/– embryoid bodies showed the presence of specifically stained cells in blood island-like structures, which form in the absence of exogenous angiogenic growth factors in the conditions used here (Figure 4C). CD41-expressing cells were localized inside VE-cadherin–positive vessels (Figure 4D). In the FGFR-1–/– embryoid bodies, a mature vascular plexus was visualized by immunostaining for VE-cadherin, but very few CD41 expressing cells were detected (Figure 4E, arrow), in agreement with the real-time PCR data in Figure 4B.
To verify that the vascular and hematopoietic phenotype of FGFR-1–/– embryoid bodies was caused by loss of the receptor, FGFR-1 was reintroduced by lentivirus-mediated gene transfer26 into the gene-targeted ES cells to create Lv–hFGFR-1. Introduction of FGFR-1 restored CD41 expression in the Lv–hFGFR-1 embryoid bodies. Spontaneous vessel formation decreased and vessels assumed a more primitive morphology. Furthermore, CD45 expression in Lv–hFGFR-1 was similar to that in wild-type embryoid bodies at day 12 (Figure III, available online at http://atvb.ahajournals.org).
Discussion
FGFR-1 is expressed on endothelial cells in tissue culture and FGF-2 is a potent mitogen for such endothelial cells.27 It has not been generally accepted that this reflects the in vivo situation, because it is well known that explanted cells display an inflammatory response leading to induction of genes not normally expressed in the in vivo context.28 We report expression of FGFR-1 in a subpopulation of endothelial cells in embryoid bodies and in teratomas. Our data are in accordance with a report on expression of FGFR-1 in a subpopulation (4.5±2.1) of CD34+ endothelial progenitor cells.29
What distinguishes an FGFR-1–expressing endothelial cell from one that lacks FGFR-1 expression? It has been suggested that replicating but not quiescent endothelial cells express FGFR-1.30 Our data agree with this report in that fluorescence-activated cell sorting analyses of the CD31+FGFR-1+ cells displayed an increased fraction of actively cycling cells compared with cells individually expressing CD31 or FGFR-1 (Figure 1). Exposure to VEGF-A induced a similar extent of mitogenic response in the CD31+ pool and the double-positive CD31+FGFR-1+ pool (Figure 2C). Expression of FGFR-1 in endothelial cells in tumor vessels may promote vascularization of the tumor and facilitate tumor growth. This is indicated by the fact that expression of dominant-negative FGFR is accompanied by decreased vascularization and growth of different model tumors in mice.11,31 Giavazzi et al32 showed that overexpression of FGF-2 leads to increased vascularization and growth of endometrial carcinoma in mice. Furthermore, suppression of tumor angiogenesis was also achieved by vaccination of mice with a xenogeneic FGFR-1.33 However, many endothelial cells in teratomas (Figure 3) as well as in human kidney cancer (data not shown) lack FGFR-1 expression; therefore, the different manipulations of FGF/FGFR function in tumor endothelial cells reported in these articles may be exerted via other types of FGFRs than FGFR-1. Alternatively, the treatment may target the tumor cells as well as the endothelial cells.
Our data show clearly that FGFR-1 expression is not required for endothelial cell development. This conclusion is further supported by the fact that endothelial cell-specific FGFR-1 gene inactivation is compatible with normal development and growth of the mouse (Partanen J and Rossant J, unpublished, 2002). In contrast, Lee et al34 showed that expression of dominant-negative FGFR delivered to the vasculature of E9 mouse embryos led to disrupted embryonic and extra-embryonic vasculature. Possibly, the dominant-negative FGFR used in this study acted as a general FGF-binding decoy, blocking the function of all FGFRs. We have previously reported that although differentiating embryoid bodies express reduced levels of VEGFR-2 transcript and protein, the VEGFR-2 appears hyperactive and promotes elevated vascularization of embryoid bodies.14 In agreement with Faloon et al,17 we show that lack of FGFR-1 expression is accompanied by decreased expression of markers of primitive and definitive hematopoiesis and an apparent arrest in hematopoietic development. Furthermore, we pinpoint the FGFR-1–dependent switch toward endothelial cell development to the level of the hemangioblast. Expression of brachyury was slightly increased, whereas VEGFR-2 and Tal-1/Stem Cell Leukemia (SCL) expression was reduced by 50% and 90%, respectively, in the FGFR-1–deficient embryoid bodies. This indicates that differentiation of the hemangioblast was disturbed in conjunction with segregation toward hematopoietic and endothelial cell development,22 and loss of FGFR-1 supports further development only of the endothelial cell linage. FGFR-1 function in this context appears to be linked to that of Tal-1/SCL as inferred from the study by Robertson et al,35 who showed that elimination of Tal-1/SCL expression attenuates hematopoiesis but allows progression of endothelial development.
Acknowledgments
This study was supported by grants from the Swedish Cancer Society (project no. 3820-B04-09XAC), the Swedish Science council (project no. K2005-32X-12552-08A), and the Novo foundation to Lena Claesson-Welsh. Expert assistance by Charlotte Wikner is deeply appreciated. The expert guidance of Dr Jan Grawés in the fluorescence-activated cell sorting analyses was crucial for our work. We are also grateful to Dr Andras Nagy, Samuel Lunenfeld Research Institute, Toronto, Canada, and Dr Chuxia Deng, Mammalian Genetics Section GDDB, National Institutes of Health, Bethesda, Md, for R1 and FGFR-1 gene-inactivated ES cell lines, respectively. Vesicular stomatitis virus-pseudotyped lentiviral vectors (Lvs) were kind gifts of Drs Luigi Naldini and Michele De Palma, San Raffaele Telethon Institute for Gene Therapy, Milan, Italy. The kind gift of the PY653 FGFR-1 antibody from Dr Jiang Wu, Cell Signaling, was much appreciated.
References
Klint P, Claesson-Welsh L. Signal transduction by fibroblast growth factor receptors. Front Biosci. 1999; 4: D165–D177.
Cross MJ, Lu L, Magnusson P, Nyqvist D, Holmqvist K, Welsh M, Claesson-Welsh L. The Shb adaptor protein binds to tyrosine 766 in the FGFR-1 and regulates the Ras/MEK/MAPK pathway via FRS2 phosphorylation in endothelial cells. Mol Biol Cell. 2002; 13: 2881–2893.
Folkman J, Klagsbrun M, Sasse J, Wadzinski M, Ingber D, Vlodavsky I. A heparin-binding angiogenic protein–basic fibroblast growth factor–is stored within basement membrane. Am J Pathol. 1988; 130: 393–400.
Kanda S, Landgren E, Ljungstrom M, Claesson-Welsh L. Fibroblast growth factor receptor 1-induced differentiation of endothelial cell line established from tsA58 large T transgenic mice. Cell Growth Differ. 1996; 7: 383–395.
Yamaguchi TP, Harpal K, Henkemeyer M, Rossant J. fgfr-1 is required for embryonic growth and mesodermal patterning during mouse gastrulation. Genes Dev. 1994; 8: 3032–3044.
Deng CX, Wynshaw-Boris A, Shen MM, Daugherty C, Ornitz DM, Leder P. Murine FGFR-1 is required for early postimplantation growth and axial organization. Genes Dev. 1994; 8: 3045–3057.
Zhou M, Sutliff RL, Paul RJ, Lorenz JN, Hoying JB, Haudenschild CC, Yin M, Coffin JD, Kong L, Kranias EG, Luo W, Boivin GP, Duffy JJ, Pawlowski SA, Doetschman T. Fibroblast growth factor 2 control of vascular tone. Nat Med. 1998; 4: 201–207.
Cross MJ, Dixelius J, Matsumoto T, Claesson-Welsh L. VEGF-receptor signal transduction. Trends Biochem Sci. 2003; 28: 488–494.
Pepper MS, Mandriota SJ. Regulation of vascular endothelial growth factor receptor-2 (Flk-1) expression in vascular endothelial cells. Exp Cell Res. 1998; 241: 414–425.
Seghezzi G, Patel S, Ren CJ, Gualandris A, Pintucci G, Robbins ES, Shapiro RL, Galloway AC, Rifkin DB, Mignatti P. Fibroblast growth factor-2 (FGF-2) induces vascular endothelial growth factor (VEGF) expression in the endothelial cells of forming capillaries: an autocrine mechanism contributing to angiogenesis. J Cell Biol. 1998; 141: 1659–1673.
Auguste P, Gursel DB, Lemiere S, Reimers D, Cuevas P, Carceller F, Di Santo JP, Bikfalvi A. Inhibition of fibroblast growth factor/fibroblast growth factor receptor activity in glioma cells impedes tumor growth by both angiogenesis-dependent and -independent mechanisms. Cancer Res. 2001; 61: 1717–1726.
Tille JC, Wood J, Mandriota SJ, Schnell C, Ferrari S, Mestan J, Zhu Z, Witte L, Pepper MS. Vascular endothelial growth factor (VEGF) receptor-2 antagonists inhibit VEGF- and basic fibroblast growth factor-induced angiogenesis in vivo and in vitro. J Pharmacol Exp Ther. 2001; 299: 1073–1085.
Nagy A, Rossant J, Nagy R, Abramow-Newerly W, Roder JC. Derivation of completely cell culture-derived mice from early-passage embryonic stem cells. Proc Natl Acad Sci U S A. 1993; 90: 8424–8428.
Magnusson P, Rolny C, Jakobsson L, Wikner C, Wu Y, Hicklin DJ, Claesson-Welsh L. Deregulation of Flk-1/vascular endothelial growth factor receptor-2 in fibroblast growth factor receptor-1-deficient vascular stem cell development. J Cell Sci. 2004; 117: 1513–1523.
Workman P, Balmain A, Hickman JA, McNally NJ, Rohas AM, Mitchison NA, Pierrepoint CG, Raymond R, Rowlatt C, Stephens TC, et al. UKCCCR guidelines for the welfare of animals in experimental neoplasia. Lab Anim. 1988; 22: 195–201.
Feraud O, Cao Y, Vittet D. Embryonic stem cell-derived embryoid bodies development in collagen gels recapitulates sprouting angiogenesis. Lab Invest. 2001; 81: 1669–1681.
Faloon P, Arentson E, Kazarov A, Deng CX, Porcher C, Orkin S, Choi K. Basic fibroblast growth factor positively regulates hematopoietic development. Development. 2000; 127: 1931–1941.
Choi K, Kennedy M, Kazarov A, Papadimitriou JC, Keller G. A common precursor for hematopoietic and endothelial cells. Development. 1998; 125: 725–732.
Wilkinson DG, Bhatt S, Herrmann BG. Expression pattern of the mouse T gene and its role in mesoderm formation. Nature. 1990; 343: 657–659.
Lacaud G, Gore L, Kennedy M, Kouskoff V, Kingsley P, Hogan C, Carlsson L, Speck N, Palis J, Keller G. Runx1 is essential for hematopoietic commitment at the hemangioblast stage of development in vitro. Blood. 2002; 100: 458–466.
Drake CJ, Fleming PA. Vasculogenesis in the day 6.5 to 9.5 mouse embryo. Blood. 2000; 95: 1671–1679.
Huber TL, Kouskoff V, Fehling HJ, Palis J, Keller G. Haemangioblast commitment is initiated in the primitive streak of the mouse embryo. Nature. 2004; 432: 625–630.
Ferkowicz MJ, Starr M, Xie X, Li W, Johnson SA, Shelley WC, Morrison PR, Yoder MC. CD41 expression defines the onset of primitive and definitive hematopoiesis in the murine embryo. Development. 2003; 130: 4393–4403.
Orkin SH. GAT. A-binding transcription factors in hematopoietic cells. Blood. 1992; 80: 575–581.
Mikkola HK, Fujiwara Y, Schlaeger TM, Traver D, Orkin SH. Expression of CD41 marks the initiation of definitive hematopoiesis in the mouse embryo. Blood. 2003; 101: 508–516.
De Palma M, Naldini L. Transduction of a gene expression cassette using advanced generation lentiviral vectors. Methods Enzymol. 2002; 346: 514–529.
Javerzat S, Auguste P, Bikfalvi A. The role of fibroblast growth factors in vascular development. Trends Mol Med. 2002; 8: 483–489.
Liaw L, Schwartz SM. Microtubule disruption stimulates DNA synthesis in bovine endothelial cells and potentiates cellular response to basic fibroblast growth factor. Am J Pathol. 1993; 143: 937–948.
Burger PE, Coetzee S, McKeehan WL, Kan M, Cook P, Fan Y, Suda T, Hebbel RP, Novitzky N, Muller WA, Wilson EL. Fibroblast growth factor receptor-1 is expressed by endothelial progenitor cells. Blood. 2002; 100: 3527–3535.
Lindner V, Reidy MA. Expression of basic fibroblast growth factor and its receptor by smooth muscle cells and endothelium in injured rat arteries. An en face study. Circ Res. 1993; 73: 589–595.
Compagni A, Wilgenbus P, Impagnatiello MA, Cotten M, Christofori G. Fibroblast growth factors are required for efficient tumor angiogenesis. Cancer Res. 2000; 60: 7163–7169.
Giavazzi R, Giuliani R, Coltrini D, Bani MR, Ferri C, Sennino B, Tosatti MP, Stoppacciaro A, Presta M. Modulation of tumor angiogenesis by conditional expression of fibroblast growth factor-2 affects early but not established tumors. Cancer Res. 2001; 61: 309–317.
He QM, Wei YQ, Tian L, Zhao X, Su JM, Yang L, Lu Y, Kan B, Lou YY, Huang MJ, Xiao F, Liu JY, Hu B, Luo F, Jiang Y, Wen YJ, Deng HX, Li J, Niu T, Yang JL. Inhibition of tumor growth with a vaccine based on xenogeneic homologous fibroblast growth factor receptor-1 in mice. J Biol Chem. 2003; 278: 21831–21836.
Lee SH, Schloss DJ, Swain JL. Maintenance of vascular integrity in the embryo requires signaling through the fibroblast growth factor receptor. J Biol Chem. 2000; 275: 33679–33687.
Robertson SM, Kennedy M, Shannon JM, Keller G. A transitional stage in the commitment of mesoderm to hematopoiesis requiring the transcription factor SCL/tal-1. Development. 2000; 127: 2447–2459.(Peetra Ulrica Magnusson; )
Correspondence to Lena Claesson-Welsh, Department of Genetics and Pathology, Uppsala University, The Rudbeck Laboratory, Dag Hammarskj?ldsv. 20, 751 85 Uppsala, Sweden. E-mail Lena.Welsh@genpat.uu.se
Abstract
Objective— The purpose of this study was to clarify the role of fibroblast growth factors (FGFs) and FGF receptors (FGFRs) in hematopoietic/endothelial development.
Methods and Results— Using several different FGFR-1–specific antibodies and FGFR-1 promoter-driven LacZ activity, we show that FGFR-1 is expressed and active as a tyrosine kinase in a subpopulation of endothelial cells (20% of the endothelial pool) during development in embryoid bodies. In agreement, in stem cell-derived teratomas, expression of FGFR-1 was detected in some but not all vessels. The FGFR-1 expressing endothelial cells were mitogenically active in the absence and presence of vascular endothelial growth factor (VEGF). Expression of FGFR-1 in endothelial cell precursors was not required for vascular development, and vascularization was enhanced in FGFR-1–deficient embryoid bodies compared with wild-type stem cells. In contrast, hematopoietic development was severely disturbed, with reduced expression of markers for primitive and definitive hematopoiesis.
Conclusions— Our data show that FGFR-1 is expressed in early hematopoietic/endothelial precursor cells, as well as in a subpool of endothelial cells in tumor vessels, and that it is critical for hematopoietic but not for vascular development.
Immunostaining using several different antibodies and scoring for FGFR-1 promoter-driven Lac Z activity show that some but not all endothelial cells express FGFR-1. FGFR-1 gene inactivation leads to a block in hematopoietic development; in contrast, endothelial cell development is enhanced.
Key Words: angiogenesis ? endothelial cells ? FGF receptor-1 ? hematopoiesis ? vasculogenesis
Introduction
Fibroblast growth factors (FGF) constitute a family of at least 23 related heparin-binding growth factors with vital and broad functions in health and disease. There are 4 distinct FGF receptor (FGFR) tyrosine kinases, denoted FGFR-1, -2, -3, and -4.1 Binding of FGF to its receptor leads to receptor dimerization and activation of the kinase, followed by tyrosine phosphorylation of the receptor itself, as well as of downstream signal transduction molecules.2
See page 883 and
FGF-2 was the first angiogenic growth factor to be identified,3 and endothelial cells in culture respond mitogenically to FGF-2 mainly through FGFR-1.4 The question of whether endothelial cells in vivo express FGFR-1 and respond to various FGFs has remained a matter of debate. There is no clear vascular phenotype in FGF-2 or FGFR-1 knockout animals. Gene inactivation of FGFR-1 leads to embryonic death in conjunction with gastrulation between embryonic days 7.5 to 9.5.5,6 Embryos homozygous-null for FGFR-1 expression display severe reduction in paraxial mesoderm formation, whereas development of node, axial, and extraembryonic mesoderm is unaffected. Gene inactivation of FGF-2, a ligand for FGFR-1, is compatible with normal development, although the gene-targeted animals display decreased vascular tone and low blood pressure.7 The possibility of compensation through any of the other FGFs or FGFRs cannot be excluded.
Vessels are established through vasculogenesis in embryonic development at approximately day 7.5 in the mouse. The vital role for vascular endothelial growth factor (VEGF)-A and its receptors VEGFR-1 and -2 for blood vascular development has been clearly demonstrated by gene inactivation.8 Cross-talk between the VEGF and FGF families of growth factors have been described; expression of VEGF-A and VEGFR-2 may be induced by FGF/FGFRs in vitro9 and in vivo.10–12 The purpose of this study was to determine whether FGFR-1 is expressed in endothelial cells using several different strategies and high-quality models.
Materials and Methods
Embryonic Stem Cell and Embryoid Body Culture
The following murine embryonic stem (ES) cell lines were used: R1,13 either wild-type or clone FGFR-1+/lacZfgfr-1; J1,6 homozygous or heterozygous for FGFR-1 gene inactivation; or clone Lentivirus-human FGFR-1 (Lv–hFGFR-1), in which FGFR-1 was reintroduced. For creation of FGFR-1+/lacZfgfr-1, the LacZ expression cassette was excised from internal ribosomal entry site (IRES)-LacZ and inserted into exon 13 in the pPNT–FGFR-1 vector encompassing FGFR-1 exons 11 to 17, corresponding to part of the tyrosine kinase domain. The FGFR-1/LacZ cDNA was integrated into R1 Sv129 by homologous recombination. For generation of Lv–hFGFR-1 stem cells, see http://atvb.ahajournals.org. The ES cell lines were cultured as described.14
Magnetic Cell Sorting to Enrich for Endothelial Cells
Embryoid bodies were dissociated by incubation in 0.25% collagenase and the single cell suspension was incubated with rat anti-mouse CD31 antibody (Becton Dickinson Biosciences), followed by mixing with magnetic cell sorting goat anti-rat IgG micro beads (Miltenyi Biotech). The CD31+ eluate was then incubated with mouse monoclonal FGFR-1 antibody (QED Bioscience) and goat anti-mouse IgG microbeads to isolate a CD31+FGFR-1+ fraction. Total RNA was extracted using the Qiagen RNeasy mini kit.
Fluorescence Activated Cell Sorting and Cell Cycle Assays
Embryoid bodies were dissociated in 0.25% collagenase. The single-cell suspension was incubated with mouse FGFR-1 monoclonal antibody (QED Bioscience) and secondary Alexa 488 goat anti-mouse highly cross-absorbed IgG (Molecular Probes). The cells were then stained with rat anti-mouse CD31-phycoerythrin antibody (Becton Dickinson). To identify the cell cycle profile, 5 μmol/L Hoechst 33342 was added to the single cell suspension in DMEM with 15% fetal bovine serum during 1 hour followed by antibody staining as described. Viable cells were identified by use of propidium iodide. The results were analyzed on a FACSVantage SE DiVa machine (Becton Dickinson) equipped with an Enterprise laser (Coherent Inc) for ultraviolet and 488 nm excitation using ModFit software (Verity Software House, Inc).
Embryoid Bodies in 3-Dimensional Collagen I Gels
Collagen gels were prepared by mixing 10x Ham’s F12 medium (Invitrogen) with 0.12% NaHCO3, 50 mmol/L HEPES, 5 mmol/L NaOH, and 1.5 mg/mL collagen I (Cohesion). Embryoid bodies (day 4 after leukemia inhibitory factor [LIF] withdrawal) were distributed on the polymerized collagen, followed by instant covering with a second layer of collagen. Medium containing 30 ng/mL of VEGF-A was added and culture continued for 8 days. To visualize the FGFR-1 promoter-driven ?-galactosidase activity, FGFR-1+/lacZfgfr-1 embryoid bodies in 3-dimensional collagen gel were fixed in 4% p-formaldehyde for 30 minutes, followed by overnight incubation at 37°C in 5 mmol/L K-ferricyanide, 5 mmol/L K-ferrocyanide, 4 mmol/L MgCl2, and 1.25 mg/mL X-gal. Incorporation of bromodeoxyuridine (BrdU) was visualized using anti-BrdU+nuclease kit (Amersham Pharmacia Biotech) and secondary antibody Alexa 488 goat anti-mouse highly cross-absorbed IgG (Molecular Probes). Nuclei were visualized by incorporation of Hoechst 33342 (1 μg/mL).
Preparing, Sectioning, and Staining of Teratomas
1x106 R1 (FGFR-1+/+) or J1 (FGFR-1+/–) ES cells were injected subcutaneously on the back flank of female Naval Medical Research Institute (NMRI)-nu mice (n=13; M&B Animal Models) and teratomas were grown for 30 days. Animal handling was performed with ethical permission approved by the Uppsala University ethics committee and according to the United Kingdom Coordinating Committee Cancer Research (UKCCCR) guidelines for the welfare of animals in experimental neoplasia.15 Acetone fixed frozen 6-μm sections immunostained with rabbit anti-mouse FGFR-1 (sc-121 and sc-4975; Santa Cruz Biotechnology) and mouse monoclonal FGFR-1 (30104; QED Bioscience Inc) were analyzed by confocal laser-scanning microscope (Zeiss LSM 510 META) or by use of a Nikon Eclipse E1000 microscope. Please see http://atvb.ahajournals.org.
Results
FGFR-1 Expression in Endothelial Cells During Development
We analyzed expression of FGFR-1 in endothelial cells during development in differentiating ES cell cultures, which display several hallmarks of vascular development in mouse embryos.14 Embryoid bodies (strain R1) were created by aggregation of stem cells and seeded individually in 8-well chambers in the presence of 20 ng/mL FGF-2 for 8 days, to induce vessel formation. Coimmunostaining showed that expression of FGFR-1 overlapped with that of CD31 on endothelial cells (Figure 1A). In embryoid bodies created from ES cells homozygous-null for FGFR-1, the anti–FGFR-1 antibody gave rise to an even, low background without vessel-specific staining (Figure 1B); in contrast, immunostaining for CD31 showed vessel-like structures. This result demonstrates the specificity of the anti–FGFR-1 antibody. To show that the FGFR-1 was functional in endothelial cells, an antibody was used that is specific for the activated receptor, phosphorylated at Y653 in the kinase domain. Staining for Y653-phosphorylated FGFR-1 colocalized with CD31 expression in endothelial cells (Figure 1C). There was no specific immunostaining by the Yp653 antibody in FGFR-1–/– embryoid bodies (Figure 1D).
Figure 1. Expression of FGFR-1 in embryoid bodies and in purified endothelial cells. Whole-mount immunostaining of day 8 wild-type R1 embryoid bodies (A, C) for CD31 or FGFR-1 (antibodies directed against the N-terminal part or against FGFR-1 phosphorylated at Y653). Staining of FGFR-1–/– embryoid bodies (B, D) was used as a negative control. Bars, 100 μm (A, C), 200 μm (B, D). E. Semi-quantitative PCR for CD31, FGFR-1, VEGFR-2, and VE-cadherin transcript expression in CD31+FGFR-1+ purified endothelial cells. Analysis of -fetoprotein and ?-actin transcript expression in flow-through (lane 1) and CD31+FGFR-1+ cell pool (lane 2). F, Fluorescence-activated cell sorting analysis of CD31 and FGFR-1 expression (blue) in embryoid bodies. The insert (red) shows background fluorescence of IgG-phycoerythrin and IgG-488. An average of 4 individual experiments is presented. G, Cell cycle profile by staining for DNA content using Hoechst 33342 of the dispersed CD31/FGFR-1 antibody-labeled cells. Figures indicate the fraction of cells in the different cell cycle phases, in percent.
To further evaluate expression of FGFR-1 in endothelial cells, day 8 wild-type embryoid bodies were dissociated and magnetic beads were used to purify CD31+FGFR-1+ coexpressing cells using a protocol ensuring that the proteins were expressed on the same cell. Semi-quantitative polymerase chain reaction (PCR) was performed to show that the CD31+FGFR-1+ also expressed the endothelial cell markers VEGFR-2 and VE-cadherin. The endodermal marker -fetoprotein served as a negative control and was present in the flow through (CD31–FGFR-1–) from the magnetic bead separation, but absent in the CD31+FGFR-1+ pool (Figure 1E).
To allow analysis of cell cycle parameters of the CD31+FGFR-1+ cells, fluorescence-activated cell sorting analyses were performed on the collagenase-digested embryoid bodies. A relatively small pool of viable CD31+FGFR-1+ cells was identified; 3% of all cells (Figure 1F), possibly representing more than one population of cells. In total, 12% of all viable cells expressed CD31 and 23% expressed FGFR-1. Staining with Hoechst 33342 allowed visualization of the cell cycle profile in the different fractions. As shown in Figure 1G, 70% of the CD31+FGFR-1+ cells were in the S/G2 phase, indicating that a larger proportion of these cells are progressing through the cell cycle, compared with cells individually expressing either CD31 or FGFR-1. Staining for expression of Annexin V indicated that there was no significant difference in the extent of apoptosis in the different cell pools (data not shown).
FGFR-1 Promoter-Driven LacZ Expression Demonstrates FGFR-1 Expression in Endothelial Cells
To analyze expression of FGFR-1 in endothelial cells independently of FGFR-1 antibodies, we created an ES cell line expressing ?-galactosidase controlled by the FGFR-1 promoter (FGFR-1+/LacZfgfr-1). This strain of ES cells behaved similar to wild-type R1 stem cells with regard to proliferation rate and FGF-2-driven DNA synthesis (Figure I, available online at http://atvb.ahajournals.org). To facilitate analysis of individual endothelial cells, we used a sprouting angiogenesis assay in which embryoid bodies are placed in 3-dimensional collagen gels. Inclusion of VEGF-A is required to induce sprouting of vessel-like structure into the gel16 (Figure 2A). FGFR-1+/LacZfgfr-1 embryoid bodies were cultured between days 4 and 12 in 3-dimensional collagen and ?-galactosidase activity was identified by X-gal staining. Costaining with anti-CD31 and anti–-smooth muscle cell actin antibodies was performed to identify pericyte-coated vessel sprouts invading the collagen gels. Some but not all endothelial cells contained ?-galactosidase activity as a result of FGFR-1 promoter activity (Figure 2B). By analyzing the distal ends of sprouts where cells clearly could be classified as single endothelial cells, the number of CD31/?-galactosidase–positive cells (ie, CD31+FGFR-1+ cells) was estimated to 15% of all endothelial cells in the sprouts. This indicates a lower fraction of coexpressing cells in the angiogenic sprouts than in the total CD31-positive pool (25%; see Figure 1F), however, the fluorescence-activated cell sorting-derived value may be influenced by the fact that CD31 is expressed also on hematopoietic cells. Moreover, in the sprouting assay, only endothelial cells in the tip of the sprouts could be clearly identified. Fluorescence-activated cell sorting analyses combining the only commercial FGFR-1 antibody of several tested that allowed specific visualization of the FGFR-1–expressing cells with antibodies against other endothelial cell markers were not technically feasible because of specificity of the reagents.
Figure 2. Detection of FGFR-1 expression in endothelial cells using FGFR-1+/lacZfgfr-1 ES cells. A, 3-dimensional culture of embryoid body in collagen gel in the presence of VEGF-A induced angiogenic sprouting visualized by CD31 staining. Bar, 1 mm. B, VEGF-A–induced sprouting by FGFR-1+/lacZfgfr-1 embryoid body in 3-dimensional collagen gel shows FGFR-1 expressing (arrow) or nonexpressing (arrowhead) endothelial cells. Left panel, Fluorescence staining, red; CD31, green; -smooth muscle cell actin, purple; Hoechst staining. Right panel, Light microscopy in which X-gal stain (blue) marks FGFR-1 promoter-positive cells. Bars, 200 μm. C, Left panel shows fluorescence staining for BrdU (green) and CD31 (red); right panel shows X-gal staining (blue). BrdU-positive CD31+FGFR-1+ (ring marked) and CD31+FGFR-1– (arrowhead) endothelial cells in FGFR-1+/lacZfgfr-1-derived sprouts. Bars, 50 μm.
Data in Figure 1G showed that the CD31+FGFR-1+ cells were actively going through the cell cycle, whereas cells individually expressing CD31 or FGFR-1 were resting to a higher extent. We wished to test the ability of the cell pools to respond mitogenically to VEGF-A. Embryoid bodies in 3-dimensional collagen gels were treated with VEGF-A for different periods of time; BrdU was added during the last 24 hours of culture. The embryoid bodies were then stained to visualize CD31/?-galactosidase–positive cells (Figure 2C shows 10-day-old embryoid bodies). Quantification by counting the number of X-gal–stained and BrdU-stained endothelial cells in the angiogenic sprouts showed the fraction of mitogenically active CD31+ endothelial cells was higher initially at day 6 (50%) and decreased with time in conjunction with stabilization of the endothelial sprouts (30%). The mitogenic activity of CD31/?-galactosidase–positive cells showed the same pattern with an initial activity in 50% of the CD31+FGFR-1+ pool followed by a decrease to 25% to 30% (data not shown). These data show that the total endothelial cell pool responds to exogenous VEGF-A with induction of a mitogenic response also engaging the CD31+FGFR-1+ cells. However, in the absence of exogenous VEGF-A, the mitogenic activity is markedly higher in the CD31+FGFR-1+ pool than in the total CD31+ pool (Figure 1G).
FGFR-1 Is Expressed in Teratoma Vessels
We next asked whether FGFR-1 is expressed on vessels in vivo. For this purpose, we analyzed vessels in teratomas, created by inoculation of immune-deficient mice with J1 or R1 ES cells expressing FGFR-1. As shown in Figure 3A (arrowheads), CD31/FGFR-1 coexpressing cells were readily detected in the J1 teratomas by use of an antibody reactive with the intracellular domain of FGFR-1. The result was further verified by use of a second commercial antibody against the FGFR-1 extracellular domain to stain sections of R1 teratomas. The FGFR-1 immunostaining colocalized with that for CD31 in vessel structures of the teratoma (Figure 3B). A third commercial FGFR-1 antibody specific for the unique NH2-terminal IgG loop of the receptor also costained CD31-positive vessel-like structures in R1 teratomas (Figure 3C). We estimated that 25% of the endothelial cells in teratomas expressed FGFR-1, although the frequency varied between different regions of the tumors.
Figure 3. Immunostaining for FGFR-1 expression in mouse teratomas. A, FGFR-1 expression in J1 teratoma (green; antibody against FGFR-1 intracellular domain; Santa Cruz, sc-121). Arrowhead indicates coexpression with CD31. B, R1 teratoma (green; antibody against the extracellular domain of FGFR-1; Santa Cruz sc-4975). C, R1 teratoma (green; antibody against the unique NH2-terminal Ig loop of FGFR-1; QED Bioscience, mAb 30104). All panels show CD31 staining in red. Bars, 100 μm.
Function of FGFR-1 in Endothelial Cells During Development
Embryoid bodies derived from stem cells lacking one (+/–) or both (–/–) alleles of FGFR-1 still form blood vessels.14 In fact, even though VEGFR-2 expression is decreased in the FGFR-1–/– stem cells, blood vessels form to a considerably increased extent even in the absence of exogenous growth factors (Figure 4A). Thus, endothelial cell development still proceeds in the absence of FGFR-1; in contrast, hematopoietic development is attenuated.17 Development of the endothelial and hematopoietic lineages diverges downstream of a common mesodermal precursor cell, the hemangioblast.18 To pinpoint the level at which loss of FGFR-1 expression guides the common endothelial/hematopoietic development toward a preference for the endothelial lineage, expression of markers of primitive and definitive hematopoiesis, as well as endothelial cell markers, was investigated by real-time PCR in differentiating embryoid bodies (Figure 4B). Expression of brachyury, one of the earliest markers for mesoderm development,19 was slightly increased by day 4 in the absence of FGFR-1. Runx1, which is expressed at the earliest stage of blood island development,20 was not affected by elimination of FGFR-1 expression. Hemangioblasts are known to express brachyury, VEGFR-2, and Tal-1/SCL.17,21,22 In the absence of FGFR-1, there was a 50% reduction in VEGFR-2 transcripts, whereas the level of Tal-1 expression was <10% of that in control FGFR-1+/– stem cells. These data indicate that loss of FGFR-1 expression affects the endothelial/hematopoietic differentiation at the level of the hemangioblast. Further real-time PCR analyses showed that expression of the primitive hematopoietic marker CD4123 and the embryonic form of ?-globin, ?-H1, a marker for primitive erythropoiesis,20 was essentially lost. In accordance, there was little or no expression of GATA-1, a transcription factor expressed in hematopoietic but absent in endothelial cells,24 or of the pan hematopoietic marker CD4525 in FGFR-1–/– stem cells at day 8 and day 12, respectively. Despite changes on the hemangioblast level as indicated by decreased expression of VEGFR-2 and Tal-1/SCL, development of the endothelial cell lineage proceeded and expression of CD31 and VE-cadherin transcripts was higher or similar to the control in the FGFR-1–/– stem cells at day 8. Transcript levels of the general smooth muscle cell marker -smooth muscle actin was reduced (Figure 4B) in agreement with the impaired mesodermal development in the absence of FGFR-1 expression. Time course analyses at days 4, 8, and 12 (Figure II, available online at http://atvb.ahajournals.org) showed there was no rescue in expression of hematopoietic markers in the FGFR-1–/– embryoid bodies excluding that the FGFR-1–/– phenotype was caused by a delay in development.
Figure 4. Endothelial and hematopoietic development in FGFR-1 lacking embryoid bodies. A, Increased vascularization in FGFR-1–/– embryoid bodies compared with FGFR-1+/– control embryoid bodies as visualized by staining for VEGFR-2. B, Real-time PCR analysis of expression in FGFR-1+/– or FGFR-1–/– embryoid bodies of hematopoietic or endothelial cell markers at day 8, or for brachyury at day 4, and CD45 at day 12. C, Localization of staining for CD41 (red) and VE-cadherin (green) in blood island-like structures in FGFR-1+/– embryoid bodies. D, Localization of CD41-positive (red) cells inside VE-cadherin–positive (green) vessels. E, Immunofluorescent staining for CD41 (red; arrow) and VE-cadherin (green) expression in FGFR-1–/– embryoid bodies. Bars, 100 μm.
Immunostaining for CD41 and VE-cadherin in FGFR-1+/– embryoid bodies showed the presence of specifically stained cells in blood island-like structures, which form in the absence of exogenous angiogenic growth factors in the conditions used here (Figure 4C). CD41-expressing cells were localized inside VE-cadherin–positive vessels (Figure 4D). In the FGFR-1–/– embryoid bodies, a mature vascular plexus was visualized by immunostaining for VE-cadherin, but very few CD41 expressing cells were detected (Figure 4E, arrow), in agreement with the real-time PCR data in Figure 4B.
To verify that the vascular and hematopoietic phenotype of FGFR-1–/– embryoid bodies was caused by loss of the receptor, FGFR-1 was reintroduced by lentivirus-mediated gene transfer26 into the gene-targeted ES cells to create Lv–hFGFR-1. Introduction of FGFR-1 restored CD41 expression in the Lv–hFGFR-1 embryoid bodies. Spontaneous vessel formation decreased and vessels assumed a more primitive morphology. Furthermore, CD45 expression in Lv–hFGFR-1 was similar to that in wild-type embryoid bodies at day 12 (Figure III, available online at http://atvb.ahajournals.org).
Discussion
FGFR-1 is expressed on endothelial cells in tissue culture and FGF-2 is a potent mitogen for such endothelial cells.27 It has not been generally accepted that this reflects the in vivo situation, because it is well known that explanted cells display an inflammatory response leading to induction of genes not normally expressed in the in vivo context.28 We report expression of FGFR-1 in a subpopulation of endothelial cells in embryoid bodies and in teratomas. Our data are in accordance with a report on expression of FGFR-1 in a subpopulation (4.5±2.1) of CD34+ endothelial progenitor cells.29
What distinguishes an FGFR-1–expressing endothelial cell from one that lacks FGFR-1 expression? It has been suggested that replicating but not quiescent endothelial cells express FGFR-1.30 Our data agree with this report in that fluorescence-activated cell sorting analyses of the CD31+FGFR-1+ cells displayed an increased fraction of actively cycling cells compared with cells individually expressing CD31 or FGFR-1 (Figure 1). Exposure to VEGF-A induced a similar extent of mitogenic response in the CD31+ pool and the double-positive CD31+FGFR-1+ pool (Figure 2C). Expression of FGFR-1 in endothelial cells in tumor vessels may promote vascularization of the tumor and facilitate tumor growth. This is indicated by the fact that expression of dominant-negative FGFR is accompanied by decreased vascularization and growth of different model tumors in mice.11,31 Giavazzi et al32 showed that overexpression of FGF-2 leads to increased vascularization and growth of endometrial carcinoma in mice. Furthermore, suppression of tumor angiogenesis was also achieved by vaccination of mice with a xenogeneic FGFR-1.33 However, many endothelial cells in teratomas (Figure 3) as well as in human kidney cancer (data not shown) lack FGFR-1 expression; therefore, the different manipulations of FGF/FGFR function in tumor endothelial cells reported in these articles may be exerted via other types of FGFRs than FGFR-1. Alternatively, the treatment may target the tumor cells as well as the endothelial cells.
Our data show clearly that FGFR-1 expression is not required for endothelial cell development. This conclusion is further supported by the fact that endothelial cell-specific FGFR-1 gene inactivation is compatible with normal development and growth of the mouse (Partanen J and Rossant J, unpublished, 2002). In contrast, Lee et al34 showed that expression of dominant-negative FGFR delivered to the vasculature of E9 mouse embryos led to disrupted embryonic and extra-embryonic vasculature. Possibly, the dominant-negative FGFR used in this study acted as a general FGF-binding decoy, blocking the function of all FGFRs. We have previously reported that although differentiating embryoid bodies express reduced levels of VEGFR-2 transcript and protein, the VEGFR-2 appears hyperactive and promotes elevated vascularization of embryoid bodies.14 In agreement with Faloon et al,17 we show that lack of FGFR-1 expression is accompanied by decreased expression of markers of primitive and definitive hematopoiesis and an apparent arrest in hematopoietic development. Furthermore, we pinpoint the FGFR-1–dependent switch toward endothelial cell development to the level of the hemangioblast. Expression of brachyury was slightly increased, whereas VEGFR-2 and Tal-1/Stem Cell Leukemia (SCL) expression was reduced by 50% and 90%, respectively, in the FGFR-1–deficient embryoid bodies. This indicates that differentiation of the hemangioblast was disturbed in conjunction with segregation toward hematopoietic and endothelial cell development,22 and loss of FGFR-1 supports further development only of the endothelial cell linage. FGFR-1 function in this context appears to be linked to that of Tal-1/SCL as inferred from the study by Robertson et al,35 who showed that elimination of Tal-1/SCL expression attenuates hematopoiesis but allows progression of endothelial development.
Acknowledgments
This study was supported by grants from the Swedish Cancer Society (project no. 3820-B04-09XAC), the Swedish Science council (project no. K2005-32X-12552-08A), and the Novo foundation to Lena Claesson-Welsh. Expert assistance by Charlotte Wikner is deeply appreciated. The expert guidance of Dr Jan Grawés in the fluorescence-activated cell sorting analyses was crucial for our work. We are also grateful to Dr Andras Nagy, Samuel Lunenfeld Research Institute, Toronto, Canada, and Dr Chuxia Deng, Mammalian Genetics Section GDDB, National Institutes of Health, Bethesda, Md, for R1 and FGFR-1 gene-inactivated ES cell lines, respectively. Vesicular stomatitis virus-pseudotyped lentiviral vectors (Lvs) were kind gifts of Drs Luigi Naldini and Michele De Palma, San Raffaele Telethon Institute for Gene Therapy, Milan, Italy. The kind gift of the PY653 FGFR-1 antibody from Dr Jiang Wu, Cell Signaling, was much appreciated.
References
Klint P, Claesson-Welsh L. Signal transduction by fibroblast growth factor receptors. Front Biosci. 1999; 4: D165–D177.
Cross MJ, Lu L, Magnusson P, Nyqvist D, Holmqvist K, Welsh M, Claesson-Welsh L. The Shb adaptor protein binds to tyrosine 766 in the FGFR-1 and regulates the Ras/MEK/MAPK pathway via FRS2 phosphorylation in endothelial cells. Mol Biol Cell. 2002; 13: 2881–2893.
Folkman J, Klagsbrun M, Sasse J, Wadzinski M, Ingber D, Vlodavsky I. A heparin-binding angiogenic protein–basic fibroblast growth factor–is stored within basement membrane. Am J Pathol. 1988; 130: 393–400.
Kanda S, Landgren E, Ljungstrom M, Claesson-Welsh L. Fibroblast growth factor receptor 1-induced differentiation of endothelial cell line established from tsA58 large T transgenic mice. Cell Growth Differ. 1996; 7: 383–395.
Yamaguchi TP, Harpal K, Henkemeyer M, Rossant J. fgfr-1 is required for embryonic growth and mesodermal patterning during mouse gastrulation. Genes Dev. 1994; 8: 3032–3044.
Deng CX, Wynshaw-Boris A, Shen MM, Daugherty C, Ornitz DM, Leder P. Murine FGFR-1 is required for early postimplantation growth and axial organization. Genes Dev. 1994; 8: 3045–3057.
Zhou M, Sutliff RL, Paul RJ, Lorenz JN, Hoying JB, Haudenschild CC, Yin M, Coffin JD, Kong L, Kranias EG, Luo W, Boivin GP, Duffy JJ, Pawlowski SA, Doetschman T. Fibroblast growth factor 2 control of vascular tone. Nat Med. 1998; 4: 201–207.
Cross MJ, Dixelius J, Matsumoto T, Claesson-Welsh L. VEGF-receptor signal transduction. Trends Biochem Sci. 2003; 28: 488–494.
Pepper MS, Mandriota SJ. Regulation of vascular endothelial growth factor receptor-2 (Flk-1) expression in vascular endothelial cells. Exp Cell Res. 1998; 241: 414–425.
Seghezzi G, Patel S, Ren CJ, Gualandris A, Pintucci G, Robbins ES, Shapiro RL, Galloway AC, Rifkin DB, Mignatti P. Fibroblast growth factor-2 (FGF-2) induces vascular endothelial growth factor (VEGF) expression in the endothelial cells of forming capillaries: an autocrine mechanism contributing to angiogenesis. J Cell Biol. 1998; 141: 1659–1673.
Auguste P, Gursel DB, Lemiere S, Reimers D, Cuevas P, Carceller F, Di Santo JP, Bikfalvi A. Inhibition of fibroblast growth factor/fibroblast growth factor receptor activity in glioma cells impedes tumor growth by both angiogenesis-dependent and -independent mechanisms. Cancer Res. 2001; 61: 1717–1726.
Tille JC, Wood J, Mandriota SJ, Schnell C, Ferrari S, Mestan J, Zhu Z, Witte L, Pepper MS. Vascular endothelial growth factor (VEGF) receptor-2 antagonists inhibit VEGF- and basic fibroblast growth factor-induced angiogenesis in vivo and in vitro. J Pharmacol Exp Ther. 2001; 299: 1073–1085.
Nagy A, Rossant J, Nagy R, Abramow-Newerly W, Roder JC. Derivation of completely cell culture-derived mice from early-passage embryonic stem cells. Proc Natl Acad Sci U S A. 1993; 90: 8424–8428.
Magnusson P, Rolny C, Jakobsson L, Wikner C, Wu Y, Hicklin DJ, Claesson-Welsh L. Deregulation of Flk-1/vascular endothelial growth factor receptor-2 in fibroblast growth factor receptor-1-deficient vascular stem cell development. J Cell Sci. 2004; 117: 1513–1523.
Workman P, Balmain A, Hickman JA, McNally NJ, Rohas AM, Mitchison NA, Pierrepoint CG, Raymond R, Rowlatt C, Stephens TC, et al. UKCCCR guidelines for the welfare of animals in experimental neoplasia. Lab Anim. 1988; 22: 195–201.
Feraud O, Cao Y, Vittet D. Embryonic stem cell-derived embryoid bodies development in collagen gels recapitulates sprouting angiogenesis. Lab Invest. 2001; 81: 1669–1681.
Faloon P, Arentson E, Kazarov A, Deng CX, Porcher C, Orkin S, Choi K. Basic fibroblast growth factor positively regulates hematopoietic development. Development. 2000; 127: 1931–1941.
Choi K, Kennedy M, Kazarov A, Papadimitriou JC, Keller G. A common precursor for hematopoietic and endothelial cells. Development. 1998; 125: 725–732.
Wilkinson DG, Bhatt S, Herrmann BG. Expression pattern of the mouse T gene and its role in mesoderm formation. Nature. 1990; 343: 657–659.
Lacaud G, Gore L, Kennedy M, Kouskoff V, Kingsley P, Hogan C, Carlsson L, Speck N, Palis J, Keller G. Runx1 is essential for hematopoietic commitment at the hemangioblast stage of development in vitro. Blood. 2002; 100: 458–466.
Drake CJ, Fleming PA. Vasculogenesis in the day 6.5 to 9.5 mouse embryo. Blood. 2000; 95: 1671–1679.
Huber TL, Kouskoff V, Fehling HJ, Palis J, Keller G. Haemangioblast commitment is initiated in the primitive streak of the mouse embryo. Nature. 2004; 432: 625–630.
Ferkowicz MJ, Starr M, Xie X, Li W, Johnson SA, Shelley WC, Morrison PR, Yoder MC. CD41 expression defines the onset of primitive and definitive hematopoiesis in the murine embryo. Development. 2003; 130: 4393–4403.
Orkin SH. GAT. A-binding transcription factors in hematopoietic cells. Blood. 1992; 80: 575–581.
Mikkola HK, Fujiwara Y, Schlaeger TM, Traver D, Orkin SH. Expression of CD41 marks the initiation of definitive hematopoiesis in the mouse embryo. Blood. 2003; 101: 508–516.
De Palma M, Naldini L. Transduction of a gene expression cassette using advanced generation lentiviral vectors. Methods Enzymol. 2002; 346: 514–529.
Javerzat S, Auguste P, Bikfalvi A. The role of fibroblast growth factors in vascular development. Trends Mol Med. 2002; 8: 483–489.
Liaw L, Schwartz SM. Microtubule disruption stimulates DNA synthesis in bovine endothelial cells and potentiates cellular response to basic fibroblast growth factor. Am J Pathol. 1993; 143: 937–948.
Burger PE, Coetzee S, McKeehan WL, Kan M, Cook P, Fan Y, Suda T, Hebbel RP, Novitzky N, Muller WA, Wilson EL. Fibroblast growth factor receptor-1 is expressed by endothelial progenitor cells. Blood. 2002; 100: 3527–3535.
Lindner V, Reidy MA. Expression of basic fibroblast growth factor and its receptor by smooth muscle cells and endothelium in injured rat arteries. An en face study. Circ Res. 1993; 73: 589–595.
Compagni A, Wilgenbus P, Impagnatiello MA, Cotten M, Christofori G. Fibroblast growth factors are required for efficient tumor angiogenesis. Cancer Res. 2000; 60: 7163–7169.
Giavazzi R, Giuliani R, Coltrini D, Bani MR, Ferri C, Sennino B, Tosatti MP, Stoppacciaro A, Presta M. Modulation of tumor angiogenesis by conditional expression of fibroblast growth factor-2 affects early but not established tumors. Cancer Res. 2001; 61: 309–317.
He QM, Wei YQ, Tian L, Zhao X, Su JM, Yang L, Lu Y, Kan B, Lou YY, Huang MJ, Xiao F, Liu JY, Hu B, Luo F, Jiang Y, Wen YJ, Deng HX, Li J, Niu T, Yang JL. Inhibition of tumor growth with a vaccine based on xenogeneic homologous fibroblast growth factor receptor-1 in mice. J Biol Chem. 2003; 278: 21831–21836.
Lee SH, Schloss DJ, Swain JL. Maintenance of vascular integrity in the embryo requires signaling through the fibroblast growth factor receptor. J Biol Chem. 2000; 275: 33679–33687.
Robertson SM, Kennedy M, Shannon JM, Keller G. A transitional stage in the commitment of mesoderm to hematopoiesis requiring the transcription factor SCL/tal-1. Development. 2000; 127: 2447–2459.(Peetra Ulrica Magnusson; )