Novel Vascular Endothelial Growth Factor Binding Domains
of Fibronectin Enhance Vascular Endothelial Growth Factor
Biological Activity
Errol S. Wijelath,* Jacqueline Murray,* Salman Rahman,* Yatin Patel,* Atsushi Ishida, Kurt Strand,
Salim Aziz, Carlos Cardona, William P. Hammond, Geoffrey F. Savidge, Shahin Rafii, Michael Sobel
Downloaded from http://circres.ahajournals.org/ by guest on June 14, 2017
Abstract—Interactions between integrins and growth factor receptors play a critical role in the development and healing
of the vasculature. This study mapped two binding domains on fibronectin (FN) that modulate the activity of the
angiogenic factor, vascular endothelial growth factor (VEGF). Using solid-phase assays and surface plasmon resonance
analysis, we identified two novel VEGF binding domains within the N- and C-terminus of the FN molecule. Native FN
bound to VEGF enhanced endothelial cell migration and mitogen-activated protein (MAP) kinase activity, but FN that
is devoid of the VEGF binding domains failed to do so. Coprecipitation studies confirmed a direct physical association
between VEGF receptor-2 (Flk-1) and the FN integrin, ␣5␤1, which required intact FN because FN fragments lacking
the VEGF binding domains failed to support receptor association. Thrombin-activated platelets released intact
VEGF/FN complexes, which stimulated endothelial cell migration and could be inhibited by soluble high affinity VEGF
receptor 1 and antibodies to ␣5␤1 integrin. This study demonstrates that FN is potentially a physiological cofactor for
VEGF and provides insights into mechanisms by which growth factor receptors and integrins cooperate to influence
cellular behavior. (Circ Res. 2002;91:25-31.)
Key Words: vascular endothelial growth factor 䡲 fibronectin 䡲 binding domains 䡲 integrins 䡲 endothelial cells
T
he growth, repair, and regeneration of blood vessels are
complex processes that involve coordinated regulation of
endothelial cell proliferation, migration, and differentiation.1
One of the most important vascular morphogens is vascular
endothelial growth factor (VEGF). VEGF has been shown to
play a major role in vasculogenesis and angiogenesis by gene
deletion studies.2,3 Targeted disruption of the VEGF receptor
Flk-1 (VEGFR-2) in mice resulted in failure of blood-island
formation and endothelial differentiation.4 Flk-1 is also the
first endothelial receptor tyrosine kinase to be expressed in
the hemangioblast.5 We and others recently demonstrated that
the hematopoietic progenitor cell CD34⫹ can differentiate
into endothelial cells, and that VEGF was one of the critical
factors promoting this differentiation.6,7 Interactions between
cells and their extracellular matrix (ECM) play an integral
role in blood vessel development. The earliest ECM protein
expressed in the embryo during vasculogenesis is fibronectin
(FN).8 Gene deletion studies have demonstrated that both FN
and its major integrin receptor, ␣5␤1, are critical for vasculogenesis and angiogenesis in the developing embryo.9 –11
Collectively, these observations suggest important roles for
FN and its integrin receptor, ␣5␤1, in vasculogenesis and
angiogenesis.
In this study, we show that novel VEGF binding domains
of FN are required for promoting the specific association of
the FN receptor integrin ␣5␤1 with the VEGF receptor, Flk-1.
This association between VEGF and FN is required for the
full effects of VEGF-induced endothelial cell migration and
proliferation. This study demonstrates that FN can profoundly
affect VEGF biological activity and consequently the behavior of endothelial cells through their coordinated effects on
Flk-1 and ␣5␤1.
Materials and Methods
Solid-Phase VEGF Binding Assay
ECM proteins and FN peptides were purchased from Sigma and
Gibco and were purified further by gel filtration and ion exchange
chromatography. Microtiter plates were coated with the appropriate
ECM proteins (50 ␮L; 10 ␮g/mL) in 100 mmol/L bicarbonate buffer
(pH 9) overnight at 4°C. 125I-VEGF165 (NEN) in binding buffer (PBS
containing 2% BSA) was added to the microtiter plates and
Original received June 5, 2001; resubmission received April 1, 2002; revised resubmission received June 4, 2002; accepted June 5, 2002.
From the Department of Molecular Biology (E.S.W., J.M., A.I., K.S., C.C., W.P.H.), The Hope Heart Institute, Seattle, Wash; the Department of
Surgery (E.S.W., J.M., K.S., M.S.), Division of Vascular Surgery, University of Washington School of Medicine and VA Puget Sound Health Care
System, Seattle, Wash; Coagulation Research Laboratory (S. Rahman, Y.P., G.F.S.), GKT Medical School, St Thomas Hospital, London, UK; Division
of Cardiothoracic Surgery (S.A.), University of Colorado, Denver, Colo; and the Division of Hematology-Oncology (S. Rafii), Cornell Medical College,
New York, NY.
*These authors contributed equally to this work.
Correspondence to E.S. Wijelath, PhD, Research Service-151, VA Puget Sound Health Care System, 1660 S Columbian Way, Seattle, WA 98108.
E-mail [email protected]
© 2002 American Heart Association, Inc.
Circulation Research is available at http://www.circresaha.org
DOI: 10.1161/01.RES.0000026420.22406.79
25
26
Circulation Research
July 12, 2002
incubated for 30 minutes at room temperature. After washing,
radioactivity was eluted with 100 mmol/L NaOH and determined
using a gamma counter. To determine nonspecific binding, 100-fold
excess of cold VEGF was added to the binding buffer and counts
subtracted from the total binding.
Slot Blot Assay
ECM protein or FN peptides (2.5 ␮g of each ) were immobilized on
nitrocellulose membranes and incubated with 50 ng/mL VEGF165
(R&D Systems) for 1 hour at 37°C in 20 mmol/L Tris, pH
7.5/0.15 mmol/L NaCl/0.1% BSA. The membranes were incubated
with mouse anti-human VEGF antibody (R&D Systems), followed
by goat anti-mouse HRP-conjugated in binding buffer for 30 minutes
each. All blots were visualized by chemiluminescence (Pierce).
Surface Plasmon Resonance Analysis (SPR)
Downloaded from http://circres.ahajournals.org/ by guest on June 14, 2017
SPR analysis was performed on the BIAcore X (Biacore). FN
fragments were coupled to CM5 dextran chips by amine coupling
chemistry according to the manufacturer’s protocols. The reference
cell had immobilized mouse IgG. In competition experiments with
the FN 40-kDa fragment, 1.3 ␮mol/L VEGF was incubated with
increasing amounts of FN 40-kDa (0.5 to 8.0 ␮mol/L) for 30 minutes
at 37°C in a reaction volume of 30 ␮L. Samples were then injected
across the FN 70-kDa biosensor chip and sensograms recorded. The
value for IC50 was determined using the ASSAY program (Biosoft).
Assuming a simple competition between the FN fragments for
binding to VEGF, an estimate of the Kd for VEGF binding to the FN
40-kDa fragment was determined using the equation; Kdcomp⫽IC50/
(1⫹Lt/KdLig), where Kdcomp is the Kd value for the competitor (FN-40
kDa fragment), Lt is the ligand concentration, and KdLig is the
dissociation constant for VEGF binding to the 70-kDa fragment.
Immunoprecipitation of VEGF/FN Complex From
Platelet Supernatants
Washed platelets were prepared as previously described.12 Platelets
were resuspended in the presence of 1.5 mmol/L calcium and at a
count of 30⫻108/mL. One milliliter of platelets was stimulated with
either saline (resting) or thrombin (1 U/mL) for 10 minutes.
Supernatants were immunoprecipitated with an antibody to FN
(Chemicon). After SDS-PAGE and electrotransfer to PVDF membranes, VEGF was detected with a polyclonal antibody (Santa Cruz)
by immunoblotting and chemiluminesescence detection.
Immunoprecipitation
Human microvessel endothelial cells (HMVECs) in serum-free
MCDB-131 medium (BioWhittaker) supplemented with 0.1% BSA
were plated on polylysine (1 mg/mL), FN (10 ␮g), vitronectin (VN;
10 ␮g/mL), or FN peptide (50 ␮g/mL) coated plates containing
VEGF (50 ng/mL) for 1 hour. Cells were lysed with lysis buffer
(20 mmol/L HEPES, pH 7.5, 0.5% Brij 35, 0.5% NP-40,
100 mmol/L NaCl, 5% glycerol, 0.1% BSA and protease inhibitors)
and immunoprecipitated with antibodies to ␣5␤1, ␣v␤1, ␣v␤3, or ␣v␤5
integrin (Chemicon). After SDS-PAGE and protein transfer, membranes were immunoblotted with antibodies to Flk-1 (Santa Cruz).
Bands were detected by chemiluminescence.
Migration Assay
Migration studies were carried out using 6.5-mm Transwells
(Costar). VEGF (50 ng/mL) and ECM (10 ␮g/mL) or FN fragment
(50 ␮g/mL) mixtures in MCDB 131 medium containing 0.5% bovine
serum albumin were added to the bottom chambers of the Transwell
and incubated at 37°C for 30 minutes. HMVECs (5⫻104) were then
placed in the upper chamber in the same medium and the Transwells
were incubated for a further 6 hours at 37°C. Transwells were
processed as described previously.13 For studies on the effect of
antibodies to integrins, HMVECs were preincubated for 30 minutes
with the indicated antibodies before addition to the upper chamber of
Transwells. The lower chamber contained VEGF (50 ng/mL), FN,
and VN (10 ␮g/mL).
Figure 1. FN potentiates VEGF-induced endothelial cell migration. HMVECs (5⫻104) were incubated in the upper chamber of
Transwell plates; lower chamber contained VEGF (50 ng/mL) ⫾
the indicated ECM protein (5 ␮g/mL). Control lower wells did
not contain VEGF or ECM. Number of migrating cells were
determined after 6 hour. Data are represented as mean⫾SD.
MAPK In Vitro Kinase Assay
HMVECs were lysed, and MAPK was immunoprecipitated using a
pan MAPK antibody (Pharmigen), washed in 50 mmol/L Tris-HCL,
pH 7.4, containing 120 mmol/L NaCl, 0.1% Triton-X-100, and 10%
glycerol. After washing in kinase buffer (50 mmol/L Tris-HCL, pH
7.4, 0.5 mmol/L DTT, 10 mmol/L MgCl2, 10 mmol/L MnCl2,
120 mmol/L NaCl, and 10% glycerol), the immunoprecipitates were
incubated for 15 minutes at 30°C in 30 ␮L of kinase buffer
containing 2.5 ␮g myelin basic protein, 20 ␮mol/L ATP, and 10
␮Ci/nmol [␥-32P]ATP (3000 Ci/mmol). Reactions were stopped with
4⫻ SDS-PAGE sample buffer, resolved by 10% SDS-PAGE. Radioactivity incorporated into the myelin basic protein bands were
determined by Cerenkov counting.
Results
Intact FN Promotes VEGF-Induced Endothelial
Cell Migration
Endothelial cell migration was slightly enhanced by either
VEGF or FN alone. But the combination of VEGF/FN increased
migration by 2.5-fold over VEGF alone or VEGF/vitronectin
(VN) and VEGF/collagen mixtures (Figure 1). Because FN has
proteolytic fragments that are chemotatic for endothelial cells,14
we tested the effect of FN fragments combined with VEGF on
VEGF-induced endothelial cell migration (see map of fragments, Figure 5B). When endothelial cells were exposed to
VEGF and only the FN 120-kDa cell-binding domain peptide,
no enhancement of endothelial cell migration was observed
(Figure 1). Adding both the 70-kDa N-terminal and the 40-kDa
C-terminal FN peptides to the VEGF/120-kDa FN peptide
mixture failed to restore endothelial migration to levels observed
with the VEGF/intact FN combination (Figure 1). To determine
whether intact FN was enhancing VEGF-induced migration by
protecting VEGF from proteolysis, we compared the rates of
VEGF degradation by plasmin in the absence or presence of FN
and VN. VEGF alone was completely degraded within 15
Wijelath et al
VEGF Binding Domains of FN
27
Downloaded from http://circres.ahajournals.org/ by guest on June 14, 2017
Figure 2. ␣5␤1 integrin mediates VEGF/FN-induced migration.
Effect of integrin blocking antibodies on HMVEC migration
induced by VEGF in the presence mixed FN and VN matrix.
Cells were preincubated for 30 minutes with the indicated antibodies before addition of the transwells to the lower chamber.
VEGF (50 ng/mL) was added to the lower chambers containing
both FN and VN. After 6 hours, migrated cells were quantified.
Data are represented as mean⫾SD.
minutes. Both FN and VN did significantly protect VEGF from
degradation, but VN was as effective as FN (⬇50% degraded
after 4 hours in the presence of FN or VN; data not shown).
␣5␤1 Integrin Mediates VEGF/FN-Induced Migration
Although ␣5␤1 is the key receptor for FN, other integrin
receptors such as ␣v␤3, the major receptor for VN, can also bind
FN. To determine the integrin responsible for the enhanced
migration, endothelial cells were exposed to a mixture of
VEGF/FN/VN, and migration measured across a combined
Figure 3. VEGF/FN combination promotes Flk-1 association
with ␣5␤1 integrin. A, Endothelial cells were incubated for 1 hour
on either VEGF/FN- or VEGF/VN-coated plates. After cell lysis,
immunoprecipitation was carried out with the indicated integrin
antibodies followed by immunoblotting with a Flk-1 monoclonal
antibody. Results of a representative experiment are shown. B,
Endothelial cells were lysed after incubation with FN alone,
polylysine/VEGF, VEGF/FN, or VEGF/120-, 70-, and 40-kDa FN
peptides for 1 hour. Lysates were immunoprecipitated with an
antibody to ␣5␤1 followed by immunoblotting with a Flk-1 monoclonal antibody.
Figure 4. FN promotes VEGF-induced MAPK activation. A,
HMVECs were incubated on VEGF/FN (filled circles), VEGF/120kDa FN peptide (triangles), and VEGF/VN (open circles) coated
plates for the indicated times. HMVEC lysates were assayed for
MAPK activity as described in Materials and Methods section
(B). HMVECs were preincubated with U0126 (50 ␮mol/L) and
wortmannin (10 nmol/L) for 1 hour. HMVECs were then placed
in the upper chamber of Transwells and stimulated with
VEGF/FN placed in the lower chamber. Migrated cells were
determined after 6 hours. Results are expressed as mean⫾SD.
FN/VN substrate. By using specific integrin-blocking antibodies, the integrin responsible for migration could be identified. In
the presence of antibodies to ␣5␤1, VEGF-induced cell migration
across the mixed FN/VN substrate was suppressed, whereas
antibodies to ␣v␤1, ␣v␤3, or ␣v␤5 had no effect (Figure 2). This
inhibition of migration by antibodies to ␣5␤1 was not due to
suppression of cell adhesion because endothelial cells were still
able to attach to the FN/VN through ␣v␤3 (data not shown). In
addition, soluble flt-1, a high-affinity receptor for VEGF,
blocked migration by over 70%.
VEGF Receptor Flk-1 Associates With Integrin ␣5␤1
We next studied how the VEGF/FN mixtures might influence
the association of their respective receptors. Incubating endothelial cells on VEGF/FN-coated plates promoted the
association of Flk-1 receptor (VEGFR-2) with ␣5␤1 as demonstrated by immunoprecipitation and Western blotting (Figure 3A). Flt-1 receptor (VEGFR-1) did not coprecipitate with
28
Circulation Research
July 12, 2002
Downloaded from http://circres.ahajournals.org/ by guest on June 14, 2017
Figure 6. Real-time interaction of VEGF with FN 70- and 40-kDa
fragments. A, Saturation analysis of VEGF binding to immobilized FN 70-kDa by SPR. Figure shows the sensograms from a
single experiment in which increasing concentrations of VEGF
were injected across an FN 70-kDa biosensor chip. Arrows indicate the injection start and end. B, Figure shows a combined
set of data from two separate experiments in which the equilibrium response is plotted as a function of VEGF concentration
and shows that saturation was achieved at 4 ␮mol/L VEGF. C,
Competition analysis of VEGF binding to immobilized FN
70-kDa by FN 40-kDa. D, Figure shows the combined set of
data from 2 independent experiments in which the equilibrium
response is plotted as a function of VEGF concentration.
Figure 5. VEGF binding sites on FN and fibrinogen. A, ECM
proteins (10 ␮g/mL) were immobilized on microtiter plates and
incubated with 125I-VEGF for 30 minutes. After washing, bound
VEGF was eluted with 100 mmol/L NaOH. Data represented as
mean⫾SD. B, Schematic diagram of the domain structure of
human FN showing type I (rectangles), type II (circles), and type
III (squares) repeats and the known binding sites for various
ligands. FN fragments (70-kDa, 120-kDa, and 40-kDa) tested are
depicted as solid lines. VEGF was incubated with full-length FN
or FN fragments immobilized on nitrocellulose membranes.
Bound VEGF was detected with a monoclonal antibody.
␣5␤1 (data not shown). When endothelial cells were incubated
on VEGF/VN-coated plates, we observed only modest coprecipitation of Flk-1 with the ␣v␤3 integrin (Figure 3A),
consistent with previous reports.15,16 Association of Flk-1
with ␣5␤1 integrin was not observed when endothelial cells
were incubated on FN alone without VEGF. Similarly,
coprecipitation of Flk-1 with ␣5␤1 integrin was not observed
when endothelial cells were either incubated on plates coated
with VEGF and the FN 120-kDa, 70-kDa, and 40-kDa
fragments or polylysine (Figure 3B). Coprecipitation of Flk-1
with ␣5␤1 was only observed when endothelial cells were
incubated with VEGF and intact FN.
Association of Flk-1 With the ␣5␤1 Integrin
Promotes Prolonged MAP Kinase Activation
Endothelial cells incubated on VEGF/FN-coated plates demonstrated sustained MAP kinase activation compared with
cells incubated on VEGF/120-kDa FN– coated plates or
VEGF/VN. In addition, an intact FN molecule is required to
mediate the VEGF-induced activation of MAP kinase be-
cause the 120-kDa FN peptide failed to promote VEGF
induced MAP kinase activation (Figure 4A). Figure 4B shows
that U0126, a specific MAP kinase inhibitor, blocked endothelial cell migration by 90%, whereas wortmannin, a PI3kinase inhibitor, suppressed migration by 20%.
FN Contains Two VEGF Binding Sites
Using the solid-phase assay, the binding of VEGF to a variety
of ECM proteins was tested. 125I-VEGF165 bound mainly to
FN (Figure 5A). Binding of VEGF was also observed with
fibrin and fibrinogen (recently reported17). VEGF did not
bind to vitronectin or collagen I, III, or IV. To locate the
VEGF binding site on the FN molecule, slot blot assays were
performed using purified proteolytically cleaved FN fragments immobilized onto nitrocellulose membranes. VEGF
bound strongly to the 70-kDa N-terminal fragment (70-kDa
FN peptide) and the 40-kDa C-terminal fragment (40 kDa FN
peptide; Figure 5B). Binding was not observed with the
120-kDa internal cell binding domain fragment (120-kDa
peptide; data not shown). Equivalent binding to FN was also
observed with VEGF121 (data not shown).
To confirm the observations of the slot blot assays, the
equilibrium binding of VEGF to the 70-kDa FN peptide was
quantified over a range of concentrations using surface
plasma resonance analysis (SPR). As shown in Figure 6,
VEGF bound to the 70-kDa FN peptide immobilized on the
sensor chip in a specific and saturable manner. The estimated
Kd was 2 ␮mol/L. VEGF binding to the 40-kDa FN peptide
could not be measured directly by SPR because immobilization of the 40-kDa FN peptide appeared to mask the VEGF
binding site. Accordingly, the Kd for VEGF binding to the
Wijelath et al
VEGF Binding Domains of FN
29
Figure 8. Schematic diagram showing VEGF/FN complexes
released by activated platelets at sites of vascular injury. VEGF
binding domains on FN promote integration of signals generated
by Flk-1 and ␣5␤1.
Downloaded from http://circres.ahajournals.org/ by guest on June 14, 2017
the VEGF/FN complex was biologically active, supernatants
from thrombin-activated platelets were filtered through Amicon filters (105-kDa cut-off) to obtain VEGF/FN complexes.
The presence of these complexes was confirmed by immunoprecipitation. Filtered supernatants from thrombinstimulated platelets promoted endothelial cell migration (Figure 7B). Addition of soluble Flt-1, a high-affinity receptor for
VEGF, inhibited endothelial cell migration by 25%. Blocking
antibody to ␣5␤1 inhibited migration by 45%. The combination of both soluble Flt-1 and anti-␣5␤1 inhibited migration by
more than 60%.
Discussion
Figure 7. VEGF/FN complex secreted by activated platelets is
biologically active. A, One milliliter of platelets (30⫻108) was
stimulated with either saline (resting) or thrombin (1 U/mL) for 10
minutes. Resting supernatant or thrombin-activated supernatant
was immunoprecipitated with an antibody to FN or control IgG1
followed by immunoblotting with a VEGF polyclonal antibody. B,
Resting and thrombin supernatants were spun through Amicon
100. Sample concentration was kept constant by adding resuspension buffer. For migration assay, the supernatants (resting
and thrombin-stimulated) were used at 0.3 mg/mL and carried
out as described in the Materials and Methods section. Soluble
Flt-1 receptor was added at a concentration of 1 ␮g/mL. Antibody to ␣5␤1 was added at 10 ␮g/mL.
40-kDa FN peptide was determined by competition experiments in which increasing amounts of the 40-kDa FN peptide
were preincubated with VEGF before injection across the
SPR sensor chip coated with the 70-kDa FN peptide surface
(Figure 6C and 6D). These experiments showed that the
40-kDa FN peptide blocked binding of VEGF to the immobilized 70-kDa FN peptide with an IC50 of 1 ␮mol/L and a
calculated Kd of 200 nmol/L for the 40-kDa fragment.
VEGF/FN Complex Formation In Vivo
To determine whether VEGF/FN complexes were spontaneously formed after platelet activation, we immunoprecipitated
the supernatants from resting and thrombin-activated platelets
with an antibody to FN. Precipitation of FN from supernatant
of thrombin-activated platelets caused significant coprecipitation of VEGF compared with resting platelets or to a
negative control antibody (Figure 7A). To determine whether
In this study, we demonstrate that VEGF binding to FN
serves to amplify the biological effects of VEGF. Specific
VEGF binding domains within FN were required to promote
sustained MAP kinase activation and endothelial cell migration. The observation that both VEGF-165 and VEGF-121
can bind FN suggests that their FN binding ability is
contained within exons 1 to 5 or 8. The VEGF binding
domains on FN were mapped to the 70-kDa N-terminal and
40-kDa C-terminal ends of the FN molecule. Although the
40- and 70-kDa FN fragments bound VEGF, these fragments
on their own were not sufficient to exert the full biological
effect of enhanced migration seen with the intact FN molecule. One possible explanation is that the 70- and 40-kDa
fragments do not fully support ␣5␤1 adhesion and that an
intact FN molecule containing both the cell binding and
VEGF binding domains is required to facilitate the association of ␣5␤1 with Flk-1 to promote enhanced cell migration.
These findings identify the VEGF binding domains on intact
FN as important cofactors for initiating a signaling pathway
mediated by the ␣5␤1/Flk-1 complex (Figure 8).
The integration of signals from integrins and receptor
tyrosine kinases is essential in mediating cellular events such
as cell proliferation, migration, and differentiation.18 –22 Several recent studies have demonstrated that integrins and
growth factor receptors can interact to form functional complexes although the mechanism(s) by which they associate to
integrate their signals is unclear. For example, recent reports
have demonstrated association of the ␤1 integrin with
VEGFR-3 (Flt4) that is required for cell migration23 and that
30
Circulation Research
July 12, 2002
Downloaded from http://circres.ahajournals.org/ by guest on June 14, 2017
platelet-derived growth factor (PDGF) in the presence of
vitronectin induces the association of the PDGF-␤ receptor
with the ␣v␤3 integrin to enhance PDGF-BB induced proliferation and migration of fibroblasts.16,24 –26 The ␣6␤4 and ␣6␤1
integrins have been shown to associate with ErB-2 in human
carcinoma cell lines after stimulation with epidermal growth
factor or insulin.27 In addition, binding of tenascin-C to ␣v␤3
was shown to promote epidermal growth factor (EGF) receptor recruitment to focal adhesions, which resulted in increased
smooth muscle cell proliferation, and that sustained activation
of MAP kinase by EGF required ␣v␤3 aggregation.28,29
Recently, it was demonstrated that PDGF-␤ and Flk-1 associated with the ␤3 integrin through its extracellular domain.
However, it is still unclear how growth factors complexed to
ECM proteins mediate the association of growth factor
receptors and integrins. Indeed, several recent reports have
demonstrated binding of soluble growth factors to ECM
proteins. Fibrinogen was shown to bind both bFGF and
VEGF and promote the proliferative effects of bFGF,17,30
whereas tenascin-X was shown to bind VEGF-B.31 Vitronectin was also shown to bind VEGF.32 The binding of VEGF to
heparan sulfate proteoglycan in ECM protects VEGF from
proteolytic degradation. Bound VEGF can be released in a
soluble and bioactive form by heparin and plasmin.33
Although a strong association between Flk-1 and ␣5␤1 was
observed when endothelial cells were plated on FN/VEGFcoated plates, we also, consistent with two other reports,
observed a weak physical association between Flk-1 and ␣v␤3
when VN was the substrate.15,16 When compared with the
Flk-1/␣5␤1 association, Flk-1/␣v␤3 association did not translate into prolonged MAP kinase activity or endothelial cell
migration. One plausible explanation for VEGF/FN promotion of MAP kinase activation and endothelial cell migration
versus VEGF/VN is that the VEGF binding domains on FN
may help bridge the Flk-1 and ␣5␤1 receptors for signal
amplification. This is supported by the observation that the
synergistic effects of VEGF/FN on endothelial cell migration,
VEGF/Flk-1 association, and MAP kinase activation requires
intact FN molecules. In support of our observations, it was
shown recently that breast cancer cells had a higher rate of
proliferation and migration in response to VEGF when
cultured on a FN substrate.34 The enhanced biological effects
of the VEGF/FN complexes observed in this study were not
due to the protective effects of FN on VEGF because both FN
and VN equally protected VEGF degradation by plasmin.
Further work will be needed to elucidate the structurefunction relationship important for the coordinate actions of
VEGF/FN.
Using specific inhibitors to MAP kinase and PI3-kinase,
we demonstrate that MAP kinase activation is important for
VEGF/FN-induced endothelial cell migration. MAP kinase
inhibition resulted in almost total suppression of endothelial
cell migration, whereas PI3-kinase inhibition only resulted in
⬇20% inhibition. This observation, in contrast to a recent
study,35 may be due to the presentation to the cell of VEGF
as a complex with FN. Consistent with our studies, fibroblasts
obtained from MEK-1– deficient mice failed to migrate on
FN. Re-expression of functional MEK-1 in the mutant fibroblasts restored their ability to migrate on FN,36 suggesting
that MAP kinase may play an important role in ␣5␤1-mediated
migration. Indeed, several other studies have provided evidence that sustained activation of MAP kinase may play a
role in cell migration37 by phosphorylation and activation of
myosin light chain kinase as well as regulating focal adhesion
assembly.38,39
It is known that platelets are a major source of VEGF, and
activated platelets release VEGF.40 This present study shows
that platelets release VEGF complexed to FN. This is a
significant finding with regard to the process of neovascularization in wound healing and tumor angiogenesis. This
suggests that growth factors in general are released from
activated platelets complexed to ECM. The formation of a
complex may serve to protect the growth factor from degradation and also to integrate the signals generated by integrins
and receptor tyrosine kinases.
In summary, we have identified two novel VEGF binding
domains on the 70-kDa N-terminal and 40-kDa C-terminal
FN molecule, which we propose are necessary for promoting
the physical association of ␣5␤1 and Flk-1. This association of
integrin and receptor tyrosine kinase enhances the amplification of signals required for sustained activation of MAP
kinase and subsequent endothelial cell migration. These
present data provide further insights into mechanisms by
which growth factors and ECM cooperate to influence cellular behavior.
Acknowledgments
This work was supported by grants from the Murdock Charitable
Foundation and Patterson Foundation (E.S.W., J.M., K.S.), Grupo
Grifols and Centeon (S. Rahman, Y.P.), and National Institutes of
Health Grant HL39903 (M.S.).
References
1. Folkman J, Shing Y. Angiogenesis. J Biol Chem. 1992;267:10931–10934.
2. Ferrara N, Carver-Moore K, Chen H, Dowd M, Lu L, O’Shea KS,
Powell-Braxton L, Hillan KJ, Moore MW. Heterozygous embryonic
lethality induced by targeted inactivation of the VEGF gene. Nature.
1996;380:439 – 442.
3. Carmeliet P, Ferreira V, Breier G, Pollefeyt S, Kieckens L, Gertsenstein
M, Fahrig M, Vandenhoeck A, Harpal K, Eberhardt C, Declercq C,
Pawling J, Moons L, Collen D, Risau W, Nagy A. Abnormal blood vessel
development and lethality in embryos lacking a single VEGF allele.
Nature. 1996;380:435– 439.
4. Shalaby F, Rossant J, Yamaguchi TP, Gertsenstein M, Wu XF, Breitman
ML, Schuh AC. Failure of blood-island formation and vasculogenesis in
Flk-1– deficient mice. Nature. 1995;376:62– 66.
5. Yamaguchi TP, Dumont DJ, Conlon RA, Breitman ML, Rossant J. flk-1,
an flt-related receptor tyrosine kinase is an early marker for endothelial
cell precursors. Development. 1993;118:489 – 498.
6. Shi Q, Rafii S, Wu MH, Wijelath ES, Yu C, Ishida A, Fujita Y, Kothari
S, Mohle R, Sauvage LR, Moore MA, Storb RF, Hammond WP.
Evidence for circulating bone marrow-derived endothelial cells. Blood.
1998;92:362–367.
7. Asahara T, Murohara T, Sullivan A, Silver M, van der Zee R, Li T,
Witzenbichler B, Schatteman G, Isner JM. Isolation of putative progenitor
endothelial cells for angiogenesis. Science. 1997;275:964 –967.
8. Risau W, Lemmon V. Changes in the vascular extracellular matrix during
embryonic vasculogenesis and angiogenesis. Dev Biol. 1988;125:
441– 450.
9. Taverna D, Hynes RO. Reduced blood vessel formation and tumor
growth in ␣5-integrin–negative teratocarcinomas and embryoid bodies.
Cancer Res. 2001;61:5255–5261.
10. Yang JT, Rayburn H, Hynes RO. Embryonic mesodermal defects in ␣5
integrin– deficient mice. Development. 1993;119:1093–1105.
Wijelath et al
Downloaded from http://circres.ahajournals.org/ by guest on June 14, 2017
11. George EL, Baldwin HS, Hynes RO. Fibronectins are essential for heart
and blood vessel morphogenesis but are dispensable for initial specification of precursor cells. Blood. 1997;90:3073–3081.
12. Patel Y, Rahman S, Siddiqua A, Wilkinson JM, Kakkar VV, Authi KS.
Functional characterization of PM6/13, a ␤3-specific (GPIIIa/CD61)
monoclonal antibody that shows preferential inhibition of fibrinogen
binding over fibronectin binding to activated human platelets. Thromb
Haemost. 1998;79:177–185.
13. Leavesley DI, Schwartz MA, Rosenfeld M, Cheresh DA. Integrin ␤1- and
␤3-mediated endothelial cell migration is triggered through distinct signaling mechanisms. J Cell Biol. 1993;121:163–170.
14. Grant MB, Caballero S, Bush DM, Spoerri PE. Fibronectin fragments
modulate human retinal capillary cell proliferation and migration.
Diabetes. 1998;47:1335–1340.
15. Soldi R, Mitola S, Strasly M, Defilippi P, Tarone G, Bussolino F. Role of
␣v␤3 integrin in the activation of vascular endothelial growth factor
receptor-2. EMBO J. 1999;18:882– 892.
16. Borges E, Jan Y, Ruoslahti E. Platelet-derived growth factor receptor beta
and vascular endothelial growth factor receptor 2 bind to the ␤3 integrin
through its extracellular domain. J Biol Chem. 2000;275:39867–39873.
17. Sahni A, Sporn LA, Francis CW. Potentiation of endothelial cell proliferation by fibrin(ogen)-bound fibroblast growth factor-2. J Biol Chem.
1999;274:14936 –14941.
18. Eliceiri BP. Integrin and growth factor receptor crosstalk. Circ Res.
2001;89:1104 –1110.
19. Hynes RO. Integrins. versatility, modulation, and signaling in cell
adhesion. Cell. 1992;69:11–25.
20. Giancotti FG. Integrin signaling: specificity and control of cell survival
and cell cycle progression. Curr Opin Cell Biol. 1997;9:691–700.
21. Yamada KM, Even-Ram S. Integrin regulation of growth factor receptors.
Nat Cell Biol. 2002;4:E75–E76.
22. Schwartz MA, Ginsberg MH. Networks and crosstalk: integrin signalling
spreads. Nat Cell Biol. 2002;4:E65–E68.
23. Wang JF, Zhang XF, Groopman JE. Stimulation of ␤1 integrin induces
tyrosine phosphorylation of vascular endothelial growth factor receptor-3
and modulates cell migration. J Biol Chem. 2001;276:41950 – 41957.
24. Bartfeld NS, Pasquale EB, Geltosky JE, Languino LR. The ␣v␤3 integrin
associates with a 190-kDa protein that is phosphorylated on tyrosine in
response to platelet-derived growth factor. J Biol Chem. 1993;268:
17270 –17276.
25. Schneller M, Vuori K, Ruoslahti E. Alphavbeta3 integrin associates with
activated insulin and PDGF␤ receptors and potentiates the biological
activity of PDGF. EMBO J. 1997;16:5600 –5607.
26. Woodard A, García-Cardeña G, Leong M, Madri J, Sessa W, Languino L.
The synergistic activity of ␣v␤3 integrin and PDGF receptor increases cell
migration. J Cell Sci. 1998;111:469 – 478.
VEGF Binding Domains of FN
31
27. Falcioni R, Antonini A, Nistico P, Di Stefano S, Crescenzi M, Natali PG,
Sacchi A. ␣6␤4 and ␣6␤1 integrins associate with ErbB-2 in human
carcinoma cell lines. Exp Cell Res. 1997;236:76 – 85.
28. Jones PL, Crack J, Rabinovitch M. Regulation of tenascin-C, a vascular
smooth muscle cell survival factor that interacts with the ␣v␤3 integrin to
promote epidermal growth factor receptor phosphorylation and growth.
J Cell Biol. 1997;139:279 –293.
29. Miyamoto S, Teramoto H, Gutkind JS, Yamada KM. Integrins can collaborate with growth factors for phosphorylation of receptor tyrosine
kinases and MAP kinase activation: roles of integrin aggregation and
occupancy of receptors. J Cell Biol. 1996;135:1633–1642.
30. Sahni A, Francis CW. Vascular endothelial growth factor binds to
fibrinogen and fibrin and stimulates endothelial cell proliferation. Blood.
2000;96:3772–3778.
31. Ikuta T, Ariga H, Matsumoto K. Extracellular matrix tenascin-X in
combination with vascular endothelial growth factor B enhances endothelial cell proliferation. Genes Cells. 2000;5:913–927.
32. Schoppet M, Chavakis T, Al-Fakhri N, Kanse SM, Preissner KT.
Molecular interactions and functional interference between vitronectin
and transforming growth factor-␤. Lab Invest. 2002;82:37– 46.
33. Park JE, Keller GA, Ferrara N. The vascular endothelial growth factor
(VEGF) isoforms: differential deposition into the subepithelial extracellular matrix and bioactivity of extracellular matrix-bound VEGF. Mol
Biol Cell. 1993;4:1317–1326.
34. Miralem T, Steinberg R, Price D, Avraham H. VEGF165 requires extracellular matrix components to induce mitogenic effects and migratory
response in breast cancer cells. Oncogene. 2001;20:5511–5524.
35. Byzova VT, Goldman KC, Pampori N, Thomas AK, Bett A, Shattil JS,
Plow FE. A mechanism for modulation of cellular responses to VEGF:
activation of the integrins. Mol Cell. 2000;6:851– 860.
36. Giroux S, Tremblay M, Bernard D, Cardin-Girard JF, Aubry S, Larouche
L, Rousseau S, Huot J, Landry J, Jeannotte L, Charron J. Embryonic
death of Mek1-deficient mice reveals a role for this kinase in angiogenesis in the labyrinthine region of the placenta. Curr Biol. 1999;9:
369 –372.
37. Eliceiri BP, Klemke R, Stromblad S, Cheresh DA. Integrin ␣v␤3
requirement for sustained mitogen-activated protein kinase activity
during angiogenesis. J Cell Biol. 1998;140:1255–1263.
38. Klemke RL, Cai S, Giannini AL, Gallagher PJ, de Lanerolle P, Cheresh
DA. Regulation of cell motility by mitogen-activated protein kinase.
J Cell Biol. 1997;137:481– 492.
39. Fincham VJ, James M, Frame MC, Winder SJ. Active ERK/MAP kinase
is targeted to newly forming cell-matrix adhesions by integrin
engagement and v-Src. EMBO J. 2000;19:2911–2923.
40. Mohle R, Green D, Moore MA, Nachman RL, Rafii S. Constitutive
production and thrombin-induced release of vascular endothelial growth
factor by human megakaryocytes and platelets. Proc Natl Acad Sci U S A.
1997;94:663– 668.
Downloaded from http://circres.ahajournals.org/ by guest on June 14, 2017
Novel Vascular Endothelial Growth Factor Binding Domains of Fibronectin Enhance
Vascular Endothelial Growth Factor Biological Activity
Errol S. Wijelath, Jacqueline Murray, Salman Rahman, Yatin Patel, Atsushi Ishida, Kurt Strand,
Salim Aziz, Carlos Cardona, William P. Hammond, Geoffrey F. Savidge, Shahin Rafii and
Michael Sobel
Circ Res. 2002;91:25-31; originally published online June 20, 2002;
doi: 10.1161/01.RES.0000026420.22406.79
Circulation Research is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231
Copyright © 2002 American Heart Association, Inc. All rights reserved.
Print ISSN: 0009-7330. Online ISSN: 1524-4571
The online version of this article, along with updated information and services, is located on the
World Wide Web at:
http://circres.ahajournals.org/content/91/1/25
Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally published
in Circulation Research can be obtained via RightsLink, a service of the Copyright Clearance Center, not the
Editorial Office. Once the online version of the published article for which permission is being requested is
located, click Request Permissions in the middle column of the Web page under Services. Further information
about this process is available in the Permissions and Rights Question and Answer document.
Reprints: Information about reprints can be found online at:
http://www.lww.com/reprints
Subscriptions: Information about subscribing to Circulation Research is online at:
http://circres.ahajournals.org//subscriptions/