| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Regular Articles |
From the Laboratories of Molecular
Immunoregulation*
and Experimental
Immunology,
Division of Basic Sciences,
Frederick Cancer Research and Development Center, Frederick, the
Intramural Research Support Program,§
Pathology
Histotechnology Laboratory, SAIC Frederick, Frederick, and the Cell
Biology Section,¶
National Institute of Dental
Research,||
Bethesda, Maryland; and the
Department of Medicine,
St. George Clinical
School, Kogarah, New South Wales, Australia
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
(SDF-1). Competitive binding studies
using 125I-labeled SDF-1 with Scatchard analysis
indicated that VEGF or bFGF induced an average number of approximately
16,600 CXCR4 molecules per endothelial cell, with a
Kd = 1.23 x 10-9 mol/L.
These receptors were functional as HUVECs and human aorta
endothelial cells (HAECs) migrated toward SDF-1. Although
SDF-1-induced chemotaxis was inhibited by the addition of a
neutralizing monoclonal CXCR4 antibody, endothelial chemotaxis
toward VEGF was not altered; therefore, the angiogenic effect
of VEGF is independent of SDF-1. Furthermore, subcutaneous
SDF-1 injections into mice induced formation of local small blood
vessels that was accompanied by leukocytic infiltrates. To test whether
these effects were dependent on circulating leukocytes, we
successfully obtained SDF-1-induced neovascularization from cross
sections of leukocyte-free rat aorta. Taken together, our data
indicate that SDF-1 acts as a potent chemoattractant for endothelial
cells of different origins bearing CXCR4 and is a participant in
angiogenesis that is regulated at the receptor level by VEGF and
bFGF.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
A number of mediators induce angiogenesis, including members of the
fibroblast growth factor (FGF)1
family, vascular
endothelial growth factor (VEGF), epidermal growth factor (EGF), tumor
necrosis factor (TNF)-, and certain members of the CXC chemokine
family.1-5
Chemokines are characterized by their ability
to induce cell migration and are divided into four groups based on
structural properties and primary amino acid sequence: CXC, CC, C, or
CX3C.6-10
The CXC chemokine subfamily includes
interleukin (IL)-8, NAP-2, ENA-78, GRO, interferon-inducible protein
(IP-10, MIG, PF4, and stromal-cell-derived factor (SDF)-1 among
others. All of the genes encoding known CXC chemokines are clustered on
human chromosome 4, with the exception of the SDF-1 gene, which is
located on chromosome 10.11
The presence of an
amino-terminal proximal Glu-Leu-Arg (ELR) motif is a reported correlate
of angiogenic chemokines, whereas CXC chemokines lacking this motif,
such as PF4 and IP-10, are reported to be angiostatic.12
SDF-1 is highly conserved between human and mouse. It was originally
isolated from murine bone marrow stromal cells, and it is highly
expressed on stromal cells of several tissues, including pancreas,
spleen, ovary, and small intestine, but not in peripheral blood
leukocytes.11
SDF-1 acts as pre-B-cell growth factor in
the presence of IL-713
and as a chemoattractant for
leukocytes14
and hematopoietic progenitor
cells.15
SDF-1-deficient mice are grossly normal but die
shortly after birth, lack B cell lymphopoiesis during embryonic
development, have defective bone marrow myelopoiesis, and have a
ventricular septal defect.16
CXCR4, also known as LESTR,
HUMSTER, or Fusin, specifically binds SDF-1. CXCR4 is expressed on
lymphocytes, monocytes, neutrophils, and epithelial
cells17,18
and was recently found on human
microglia19
and various endothelial
cells.20-22
Although ELR-positive members of the CXC chemokine family are reported
to induce corneal angiogenesis, expression of their receptors (CXCR1
and CXCR2) on endothelial cells, a prerequisite of endothelial cell
chemotaxis and direct induction of angiogenesis, remains a
controversial issue.23,24
It has been recently reported
that human umbilical vein endothelial cells (HUVECs) express
CXCR421,22
and that this expression can be modulated by
the inflammatory mediators interferon (IFN)-
, TNF-, IL-1ß or
lipopolysaccharde (LPS).21
However, in this study, the
investigators cultured their endothelial cells in media containing
either VEGF or added basic (b)FGF.21
In this paper we investigated the ability of known angiogenic and/or
inflammatory factors to induce expression of CXC chemokine receptors on
endothelial cells. Under the conditions used, we found that either VEGF
or bFGF induced elevated cell surface expression of CXCR4 only and
substantially increased HUVEC migration toward SDF-1. Furthermore,
we demonstrate that SDF-1 induces local neovascularization in
vivo. Thus, even though SDF-1 lacks an ELR motif, it nevertheless
can act as a novel angiogenic factor in conjunction with other
angiogenic growth factors.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
HUVECs were isolated by treatment of umbilical cords with
trypsin/EDTA (0.25%/0.02%) in PBS for 10 minutes at 37°C. After
elution of the HUVECs with RPMI 1640 (Life Technologies, Gaithersburg,
MD) containing 20% FCS (Hyclone, Logan, UT), HUVECs were cultured on
gelatin-coated (Sigma Chemical Co., St. Louis, MO) flasks in 199 medium
(Life Technologies) or EBM medium (Clonetics, Walkersville, MD)
containing 10% fetal calf serum, glutamine (2 mmol/L), penicillin (100
U/ml), and streptomycin (100 µg/ml). The cells were characterized by
morphological criteria and positive staining with CD31 or inducibility
of vascular cell adhesion molecule (VCAM)-1 after TNF- stimulation.
Human aorta endothelial cells (HAECs) were prepared as
described.25
Experiments were performed on subcultures
between the third and sixth in vitro passage.
Chemokines and Antibodies
Recombinant human SDF-1, recombinant human VEGF, recombinant
human bFGF, recombinant human epidermal growth factor (EGF), and
recombinant human TNF- were purchased from Pepro Tech (Rocky Hill,
NJ). Recombinant human IFN- was purchased from Biogen (Cambridge,
MA), and phorbol ester and LPS was purchased from Sigma. Monoclonal
anti-human CXCR4 (12G5) was purchased from R&D Systems (Minneapolis,
MN). Anti-human VCAM-1 and anti-human ELAM-1 were purchased from
Immunotech (Westbrook, ME), and mouse IgG and Rabbit IgG (Coulter,
Miami, FL) were used as negative controls.
RNAse Protection Assay
After a 4-hour incubation of the cells with VEGF (50 ng/ml), bFGF
(50 ng/ml), EGF (100 ng/ml), SDF-1 (100 ng/ml), TNF- (200 U/ml),
LPS (10 µg/ml), IFN- (100 U/ml), and phorbol ester (160 nmol/L)
alone or in different combinations, RNA was isolated by the TRIZOL
method as directed (Life Technologies) and thereafter used for analysis
of mRNA expression using the Riboquant RNAse protection assay system
(human CR6 probe set, PharMingen, San Diego, CA) according to the
manufacturer's instructions. Briefly 33P-labeled antisense
RNA probes were synthesized from the human chemokine receptor 6
template by T7 RNA polymerase. The probe (1.5 x 106
cpm) was hybridized in solution overnight in excess to target RNA (10
µg total RNA/treatment) in a total reaction volume of 10 µl. The
free probe and other single-stranded RNA were digested with RNAses A
and T1 per instructions provided by the manufacturer. The remaining
RNAse-protected probes were precipitated, dissolved in 5 µl of sample
buffer (PharMingen), and resolved on denaturing polyacrylamide gels
followed by autoradiography for 1 to 7 days at -70°C.
Immunofluorescence Flow Cytometry
HUVECs and HAECs were stimulated with escalating concentrations of
VEGF, bFGF, TNF-, and LPS. Optimal doses for each factor were then
used in a kinetic assay at 12, 24, 48, and 72 hours for cell surface
expression of CXCR4 by immunofluorescence. For TNF--stimulated
cells, cells were exposed for approximately 40 minutes to 200 U/ml
TNF-, which was removed by washing. Treated cells were thereafter
cultured as described. Indirect immunofluorescence was performed by
exposing cells to saturating amounts of mouse antibodies to human
CXCR4, VCAM-1, or ELAM-1. As the second antibody,
fluorescein-conjugated F(ab)2 fragments of goat anti-mouse
(Sigma) were used at a 1:20 dilution. After staining, cells were
analyzed in a FACScan flow cytometer (Becton Dickinson, Mountain View,
CA). Mouse IgG was used as the negative control.
Endothelial Cell Migration Assay
Endothelial cell chemotaxis was performed using micro Boyden's chambers. Briefly, polycarbonate filters of 8-µm pore size (Nucleopore, NeuroProbe, Cabin John, MD) were coated with 0.5% collagen type I (Collaborative-Biomedical Products, Bedford, MA) overnight at 4°C. Binding buffer containing 1.0% bovine serum albumin in RPMI 1640 with or without the stimulus was placed in the lower compartment of the chamber, and 106 HUVECs or HAECs/ml resuspended in binding medium were then seeded into the upper compartment. The chambers were incubated for 2 hours at 37°C. After the filters were removed, the upper surface was scraped, fixed with methanol, and stained with Leukostat (Fisher Scientific). Membranes were analyzed using the BIOQUANT program (R & M Biometrics, Nashville, TN), and the results were expressed as the mean number of migrated cells/10 fields at x10 magnification.
Receptor Binding Assay
Binding of chemokines to their receptors was assessed using 1
ng/ml of 125I-labeled SDF-1 (New England Nuclear, Boston
MA) in the presence of various concentrations of unlabeled SDF-1
(Pepro Tech), as previously described.26
Stimulated HUVECs
at 107/ml in RPMI 1640 containing 1% bovine serum albumin
(w/v) and 25 mmol/L HEPES were incubated in the presence of SDF-1
for 45 minutes at room temperature and pelleted through 10% sucrose in
PBS, and cell-pellet-associated radioactivity was determined in a gamma
counter. Binding data were analyzed using the computer program LIGAND.
Assessment of Receptor Redistribution Using Confocal Laser Microscopy
HUVECs were grown on gelatin-treated tissue culture chamber slides
(Nunc, Naperville, IL) in the presence or absence of VEGF (100 ng/ml)
at 37°C in a humidified atmosphere containing 5% CO2 for
24 hours. They were then washed three times with PBS and cultured in
199 medium containing SDF-1 (1 µg/ml) for 30 minutes, washed once
with ice-cold PBS, and fixed in 2% paraformaldehyde/PBS. Cells were
permeabilized in 0.15% saponin before incubation for 60 minutes with
biotin-labeled 12G5 monoclonal antibody (MAb; R&D Systems). After
washing, cells were incubated with streptavidin-conjugated rhodamine
(Molecular Probes, Eugene, OR). Finally, cells were washed and stained
with DAPI (Sigma) for 10 minutes, and slides were then examined
using a Zeis 310 confocal laser scanning microscope. Nomarski,
rhodamine (543 nmol/L, red), and DAPI (ultraviolet 364 nm, blue) images
were prepared for each specimen and were subsequently superimposed
using the Nomarski image as a base.
In Vivo Angiogenesis Assay
Female BALB/c mice were used at 6 to 8 weeks of age. Animal
housing and management were in accordance with the procedures outlined
in the Guide for the Care and Use of Laboratory Animals (National
Academy of Sciences, Institute of Laboratory Animal Resources, National
Research Council, 1996), and the protocol used was approved by the
National Cancer Institute-Frederick Cancer Research and Development
Center Animal Care and Use Committee. Angiogenesis was assessed using
the technique of Weidner et al27
as an increase in the
number of microvessels in the subcutaneous tissue injected with
chemokines. Mice, three per group per time point, were injected
subcutaneously on days 0, 1, 2, and 3 with SDF-1 (1 µg) or VEGF (1
µg) or both in 0.1 ml of PBS. Injection sites were removed at
necropsy on day 7 or day 14, fixed in Bouin's, and embedded in
paraffin, and 5-µm-thick sections were prepared.
Immunohistochemical staining using rat anti-mouse CD31 (PharMingen), also known as PECAM-1, was performed at a dilution of 1:50 after pretreatment of the sections for 30 minutes with Sigma tissue-culture-grade trypsin diluted 1:10 in PBS. Biotinylated rabbit anti-rat IgG, mouse absorbed, was the secondary antibody used with the Vectastain ABC kit (Vector Laboratories, Burlingame, CA). Diaminobenzidine was used as a chromogen. Slides were also stained with hematoxylin and eosin (H&E) to evaluate inflammatory lesions.
ELISA
Growth factor release was quantitated by ELISA as directed by the
manufacturer (R&D Systems). HUVECs were cultured (0.5 x
106
per/well) in six-well plates pretreated with gelatin as
described above in medium containing VEGF (100 ng/ml) and bFGF (10
ng/ml). After 24 hours, the medium was removed, and the cells were
washed eight times with PBS. Fresh 199 medium containing 2% fetal calf
serum was added (2 ml/well), and treated wells received SDF-1 (1
µg/ml final concentration). Supernatant samples were collected at 24,
48, and 72 hours.
Rat Aortic Ring Assay
Rat aortic rings were prepared as previously
described28
with modifications. The thoracic and abdominal
aorta was obtained from 100- to 150-g male Sprague-Dawley rats (Taconic
Farms, Germantown, NY). Excess perivascular tissue was removed,
transverse sections (1 to 2 mm) were made, and the resulting aortic
rings were then extensively washed in medium 199 (Mediatech).
The rings were then embedded in 2 µg/ml rat tail collagen in Nunc
eight-well chamber slides (Nalgene Nunc International) so that
the lumen was parallel to the base of the slide. After the collagen I
gelled (by adjustment of pH to neutral with NaOH), serum-free medium
(endothelial basal medium supplemented with antibiotics and
e-aminocaproic acid, 0.3 µg/ml) was added to each well, and the
slides were incubated at 37°C, 5% CO2, for 3 days. Once
sprouts began to appear, SDF-1 was added at concentrations of 1 to
100 ng/ml (n = 6 per dose). ECGS was used
as the positive control at concentrations of 200 µg/ml. The rings
were incubated for 3 additional days and then fixed, stained, and
photographed. The ring assay was repeated two times.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In an effort to identify chemokines that might be direct inducers
of angiogenesis, we first established the capacity of unstimulated
HUVECs and HUVECs preincubated with a variety of stimuli to express
known CXC subfamily receptor genes. We used the RNAse protection assay,
which allows for the detection of multiple chemokine receptor messages
in a single RNA preparation. Freshly isolated HUVECs (passage 4) were
grown in basal (199) medium and subsequently stimulated for 4 hours, as
described in Materials and Methods, with the following combinations:
bFGF, VEGF, and EGF together; TNF-, LPS, and IFN- together; or
phorbol ester alone. As shown in Figure 1A
, HUVECs constitutively expressed low
levels of CXCR4 mRNA, and this expression was increased approximately
10-fold after treatment either with phorbol ester or the combination of
VEGF, bFGF, and EGF.
|
, LPS, TNF-, or
IFN- under the same conditions as described above. As shown in
Figure 1B
, bFGF treatment increased CXCR4 mRNA twofold and VEGF
fourfold; however, neither EGF nor SDF-1 had any stimulatory effect.
In addition, TNF-, IFN-, and LPS did not have any effect on CXCR4
mRNA levels within 4 hours when tested alone (data not shown). Despite
reports of endothelial cell CXC chemokine receptor expression and that
the ligands for CXCR1 and CXCR2 are angiogenic and for CXCR3 are
angiostatic, we did not detect CXCR1, CXCR2, or CXCR3 mRNAs in HUVECs
under basal conditions or after stimulation with any of the stimuli
used above by using the RNAse protection assay (Figure 1)
. Cell Surface Expression of CXCR4 on Endothelial Cells
Consistent with the low constitutive levels of CXCR4 mRNA in
cultured endothelial cells, we detected low expression of CXCR4 protein
on the cell surface. After combined stimulation with VEGF (50 ng/ml)
and bFGF (10 ng/ml), CXCR4 expression on HUVECs and HAECs was
significantly enhanced, as detected by FACS analysis (Figure 2A)
.
No detectable increase in cell surface expression of this receptor was
observed after either EGF, SDF-1, or LPS treatment, whereas IFN-
reduced CXCR4 levels (Figure 2B)
. The highest expression of CXCR4 was
detected using HUVECs cultured for 24 hours in the presence of 10 to 50
ng/ml VEGF. This high level of expression was maintained beyond 48
hours of stimulation but returned to basal levels by 72 hours (data not
shown). We also observed that TNF- induced increased CXCR4
expression on HUVECs; however, this effect required 48 hours or more to
become evident, whereas the effect of VEGF or bFGF peaked at 24 to 48
hours (data not shown).
|
Binds Specifically and with High Affinity to Stimulated
HUVECs and HAECs
To investigate the binding capacity of CXCR4 on bFGF- and
VEGF-stimulated HUVECs, we performed competitive binding studies with
radio-iodinated SDF-1. The 125I-labeled SDF-1 (10
nmol/L) bound rapidly (after 30 minutes) to HUVECs, and unlabeled
SDF-1 competitively reduced the binding of labeled SDF-1 (Figure 3)
. Scatchard analysis for SDF-1
binding indicated that stimulated HUVECs expressed an average number of
16,608 receptors/cell, which bound SDF-1 with high affinity
(Kd = 1.23 x 10-9
mol/L; Figure 3
; Table 1
).
|
|
Induced Redistribution of CXCR4 on Stimulated HUVECs
Interaction of chemokines with their receptors typically results
in receptor redistribution and/or internalization. To analyze the
effect of SDF-1 binding to CXCR4 on activated HUVECs, immunostaining
for CXCR4 followed by confocal microscopy was performed. As shown in
Figure 4
, CXCR4 on VEGF-stimulated HUVECs
was primarily localized on the cell surface and in intracellular pools
(Figure 4A)
. After SDF-1 treatment, reduction of cell surface CXCR4
with an increase in the intracellular pools of this receptor was
observed (Figure 4B)
. Indeed, cell surface expression of CXCR4 as
assessed by flow cytometric analysis, was decreased after SDF-1
treatment. Furthermore, permeabilized SDF-1-treated cells were
clearly CXCR4 positive (data not shown). These findings are in
agreement with the effects of phorbol myristate acetate and SDF-1, both
of which also cause the internalization of CXCR4 by T
cells.29
|
Induced Endothelial Cell Chemotaxis via
CXCR4
We next assessed the ability of SDF-1 to induce endothelial
cell migration in vitro. SDF-1 was capable of
inducing chemotaxis of VEGF- and bFGF-activated
HUVECs and HAECs in a dose-dependent fashion. The
dose-response curve had the characteristic bell shape typical of other
chemokines with an optimal chemotactic dose between 10 and 100 ng/ml
(Figure 5A)
. Interestingly, the
chemotactic index for VEGF was approximately threefold lower than
SDF-1 (data not shown).
|
on HUVECs, we performed assays by placing different concentrations of
SDF-1 in the upper and/or lower wells of the chemotaxis chamber.
Checkerboard analysis revealed that the response of bFGF- and
VEGF-stimulated HUVECs to SDF-1 was chemotactic rather than
chemokinetic (data not shown).
As a test of specificity, we determined the effects of a MAb directed
against human CXCR4 on SDF-1-induced migration of activated HUVECs
and HAECs. The results demonstrated that 12G5, which recognizes an
epitope located in the second extracellular loop of CXCR4, inhibited
HUVEC and HAEC migration by more than 80% (Figure 5B)
. The CXCR4
antibody, however, did not inhibit the chemotactic response of
endothelial cells toward VEGF, indicating that the effect of this
growth factor on HUVECs and HAECs is direct and not dependent on
SDF-1 (data not shown).
SDF-1 Induced Neovascularization in Vivo
In our hands, SDF-1 had a low but significant proliferative
effect on VEGF-pretreated HUVECs or HAECs (data not shown).
Nonetheless, our in vitro chemotaxis findings suggested that
SDF-1 might act as an angiogenic factor. To test this possibility,
subcutaneous injections (day 0) in mouse skin of 1 µg of either
SDF-1, VEGF (positive control), or PBS daily for 4 consecutive days
were performed, as described in Materials and Methods. At day 7,
significant induction of microvessel formation was observed using
SDF-1 alone, which was equivalent to the VEGF-induced response
(Figures 6A and 7)
. At 14 days after the initial
injection the average number of microvessels was reduced, but the
number of microvessels in both the SDF-1- and VEGF-stimulated skin
sections was still significantly elevated relative to the PBS control
(Figure 6B)
. We did not observe a significant increase in microvessel
formation when both agents were used in combination relative to VEGF
alone (P 
0.375). Furthermore, an inflammatory
reaction consisting of neutrophils and mononuclear cells, including
CD3-positive lymphocytes was observed at day 7 in both VEGF- and
SDF-1-stimulated skin sections (data not shown).
|
|
-Induced Rat Aortic Endothelial Cell Sprouting
As SDF-1 is chemotactic for lymphocytes and VEGF is chemotactic
for mononuclear cells, which were found at sites of cutaneous
injections of either SDF-1 or VEGF, we sought to investigate the
possibility that the angiogenesis we observed was leukocyte dependent.
To elucidate the role of SDF-1 in angiogenesis in the absence of
inflammatory cell infiltrates, SDF-1 was tested using the
ex-vivo rat aortic ring sprouting assay. Transverse sections
of rat aorta embedded in collagen were stimulated with SDF-1, as
described in Materials and Methods, and thereafter examined for the
degree of sprouting vessels. Cell culture medium and endothelial cell
growth supplement medium were used as negative and positive controls,
respectively. As shown in Figure 8
,
SDF-1 stimulated numerous long capillary sprouts at a concentration
as low as 1 ng/ml (0.125 nmol/L). Thus, the data indicated that
SDF1- can induce angiogenic sprouting at subnanomolar concentrations
from rat aortic rings in the absence of inflammatory cell infiltrates.
|
-Induced VEGF Release from Cultured HUVECs
To learn whether SDF-1-CXCR4 interaction on HUVECs induced the
production of angiogeneic factors, ELISA of culture supernatant samples
was performed as described in Materials and Methods. Relative to
resting cultures, VEGF- and bFGF-conditioned HUVECs, in response to
SDF-1, produced from 7- to 13-fold more VEGF, depending on the donor
of origin (Figure 9)
. VEGF
production was diminished by 48 hours and returned to basal levels by
72 hours (data not shown). In contrast, no significant increase in bFGF
or IL-8 was observed after addition of SDF-1, under these
conditions, at any of the time points examined (data not shown).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
, IL-1, or LPS induces the release of IL-8, GRO- and -ß,
and ENA-78 and expression of chemokine receptors such as CXCR1 and
CXCR2.30-32
However, under our culture conditions for
fresh HUVECs, mRNA for CXCR1, CXCR2, and CXCR3 receptors were not
detected even after stimulation with either inflammatory mediators (ie,
LPS, TNF-, or IFN-) or angiogenic factors (ie, VEGF, bFGF, or
EGF) or with phorbol ester. Our RNAse protection assay results obtained
using HUVECs in the absence of leukocytes after stimulation with either
bFGF, VEGF, or phorbol ester showed inducible expression only of CXCR4
mRNA by 4 hours and by FACS analysis CXCR4 protein expression by 24
hours. In agreement with published results,21
IFN-,
TNF-, and LPS, acting alone, failed to induce CXCR4 mRNA synthesis
by 4 hours after stimulation. We observed late CXCR4 expression by 48
to 72 hours on TNF--stimulated HUVECs, and at a lower level than on
VEGF- or bFGF-stimulated cells. TNF- is also reported to be an
angiogenic factor33
and was capable of inducing VEGF and
bFGF production from endothelial cells.32,34
Indeed,
antibodies to VEGF blocked TNF--induced neovascularization in the
rabbit cornea.32
We therefore tested the possibility that
the capacity of TNF- to induce delayed expression of CXCR4 was
mediated by these angiogenic factors. Antibodies to VEGF and bFGF, when
used in combination, inhibited up to 43% of the CXCR4 expression on
TNF--stimulated HUVECs (data not shown). This suggests
that the induction of CXCR4 by TNF- is only in part mediated by VEGF
and bFGF, and additional mediators are involved.
The induction of CXCR4 expression presumably precedes responsiveness to
its ligand, and therefore the angiogenic effect of the ligand for
CXCR4, SDF-1, was investigated. In vitro chemotaxis,
receptor binding, and receptor redistribution data indicated that VEGF-
and bFGF-induced CXCR4 on HUVECs was indeed responsive to SDF-1,
which led us to experimentally demonstrate that injections of this
chemokine can induce angiogenesis in vivo.
We observed that endothelial cell responsiveness to SDF-1 is not
restricted to freshly isolated HUVECs; HAECs and spontaneously
immortalized endothelial cell lines derived from human umbilical cords
(HUVEC B019809 and HUVEC B019810) also migrated toward SDF-1 after
VEGF and bFGF stimulation (Figure 5
; data not shown). Furthermore, as
SDF-1 induced endothelial cell proliferation and microvessel
formation at peripheral injection sites, mouse capillary endothelial
cells are also SDF-1 responsive. Moreover, rat aorta endothelial
cells are also responsive to SDF-1. Interestingly, CXCR4 knockout
mice lack the ability to properly vascularize the intestine during
fetal development.35
These data indicate that
SDF-1-CXCR4 interaction is necessary for some, but not all, types of
neovascularization. The capacity of SDF-1 knockout fetuses to
vascularize tissues and heal wounds has not been evaluated.
Despite the reports that only ELR-containing CXC chemokines have a role
in angiogenesis,12,36
we found that SDF-1, a member of
the CXC chemokine family and ligand for CXCR4, even though it does not
have an ELR motif, acted as a direct chemoattractant for endothelial
cells in vitro and as an angiogenic factor in
vivo. The presence of leukocytic infiltrates prompted the
assessment of angiogenic activity using cross sections of rat aorta.
The ability of SDF-1 to induce capillary sprouting is therefore
leukocyte independent.
It is well established that IL-8, which binds to CXCR1 and CXCR2, can
act as an angiogenic factor.24,37-39
In addition, Yoshida
et al32
reported inducibility of CXCR2 by TNF- on human
microvascular endothelial cells. However, our observation on the lack
of receptors for IL-8 on endothelial cells is in accordance with the
data of others21,23
showing that HUVECs and human dermal
microvascular endothelial cells lack these receptors. The lack of CXCR1
and CXCR2 expression is consistent with an indirect model as proposed
by Hu et al39
in which pro-inflammatory cytokines derived
from leukocytes contribute to angiogenesis triggered by IL-8 in
vivo. Consequently, cytokines released by leukocytes at
inflammatory sites, such as TNF-, by inducing bFGF and VEGF, might
be important indirect stimulants of the angiogenic response.
Under the conditions used, we were unable to induce detectable levels
of CXCR3 mRNA after multiple stimuli (shown above). The ligand for
CXCR3, IP-10, is a known potent angiostatic factor.38
The
lack of IP-10 receptor expression on HUVECs is puzzling and suggests
that the effect of IP-10 may be indirect, or an as yet unidentified
receptor for IP-10 might exist on endothelial cells. The inhibitory
ability of mediators such as IFN- on angiogenesis might exert their
effects via an IP-10-independent mechanism and might be
based on the down-modulation of CXCR4.21
The fact that all three known VEGF, acid and basic FGF, and SDF-1 genes
are widely expressed in normal organs of adult mice and humans and that
their receptors (flk, flt-1, flt-4, FGFR1–4, and CXCR4) are expressed
on vascular endothelial cells1,11,20
suggests that these
interactions contribute to maintenance of the endothelium. Angiogenesis
is therefore indirectly triggered by up-regulation of receptor levels
by inflammatory mediators such as TNF-32
or by enhanced
levels of angiogenic factors such as VEGF and bFGF. Indeed,
SDF-1-CXCR4 interaction further amplifies angiogenesis by inducing
more VEGF release from CXCR4-bearing HUVECs. In addition, SDF-1 might
contribute to angiogenesis during pathological neovascularization
induced by angiogenic factors, such as bFGF or VEGF, which by
increasing the expression of CXCR4 render endothelial cells more
responsive to SDF-1.
Consideration of the potential for interactions between SDF-1 and
VEGF led to the possibility that these factors can act in an additive
or synergistic manner. However, additivity and synergy are best
determined when a single outcome is assessed, in this case, mitogenesis
versus chemotaxis. The mitogenic influence of SDF-1, in
contrast to the proliferative effects of VEGF, appears limited at best,
whereas VEGF is a less potent chemoattractant than SDF-1. The two
factors appear to act in a complementary fashion by performing
different functions. Taken together, our results are consistent with a
model for angiogenesis in which mediators such as TNF- act
indirectly by inducing the release of factors such as VEGF and
bFGF.32
Elevated levels of these growth factors in turn
and in the absence of IFN-22,40
promote enhanced
expression of CXCR4 on endothelial cells, which can then respond to
stromal-cell-derived SDF-1. SDF-1, in turn, and in addition to
its role as a chemoattractant, acts as an amplifying factor by
enhancing VEGF release (Figure 10)
.
|
|
Acknowledgements |
|---|
, Dr. Gerald A. Evans for his help with the
confocal microscopy (Figure 4)
, and Kimberly Roderick, R.N., of
Frederick Memorial Hospital for kindly providing human umbilical cords.
R. Salcedo was supported by Abgenix, Inc. By acceptance of this article, the publisher or recipient acknowledges the right of the U.S. Government to retain a non-exclusive, royalty-free license in and to any copyright covering the article.
|
Footnotes |
|---|
Accepted for publication January 13, 1999.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
and ß chemokines. FASEB J 1995, 9:57-62[Abstract]
-dependent angiogenesis. Mol Cell Biol 1997, 17:4015-4023[Abstract]
. Nature 1987, 329:630-632[Medline]
This article has been cited by other articles:
 |
|
 |
|
  Z. Sameermahmood, M. Balasubramanyam, T. Saravanan, and M. Rema Curcumin Modulates SDF-1{alpha}/CXCR4-Induced Migration of Human Retinal Endothelial Cells (HRECs) Invest. Ophthalmol. Vis. Sci., August 1, 2008; 49(8): 3305 - 3311. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. M. Kerfoot, G. Andonegui, C. S. Bonder, and L. Liu Exogenous stromal cell-derived factor-1 induces modest leukocyte recruitment in vivo Am J Physiol Heart Circ Physiol, June 1, 2008; 294(6): H2524 - H2534. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Yang, L. K. Pham, C.-P. Liao, B. Frenkel, A. H. Reddi, and P. Roy-Burman A Novel Bone Morphogenetic Protein Signaling in Heterotypic Cell Interactions in Prostate Cancer Cancer Res., January 1, 2008; 68(1): 198 - 205. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Z. Pasha, Y. Wang, R. Sheikh, D. Zhang, T. Zhao, and M. Ashraf Preconditioning enhances cell survival and differentiation of stem cells during transplantation in infarcted myocardium Cardiovasc Res, January 1, 2008; 77(1): 134 - 142. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Lima e Silva, J. Shen, S. F. Hackett, S. Kachi, H. Akiyama, K. Kiuchi, K. Yokoi, M. C. Hatara, T. Lauer, S. Aslam, et al. The SDF-1/CXCR4 ligand/receptor pair is an important contributor to several types of ocular neovascularization FASEB J, October 1, 2007; 21(12): 3219 - 3230. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 I. Kryczek, S. Wei, E. Keller, R. Liu, and W. Zou Stroma-derived factor (SDF-1/CXCL12) and human tumor pathogenesis Am J Physiol Cell Physiol, March 1, 2007; 292(3): C987 - C995. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Chen, H. Bai, Y. Shao, M. Arzigian, V. Janzen, E. Attar, Y. Xie, D. T. Scadden, and Z. Z. Wang Stromal Cell-Derived Factor-1/CXCR4 Signaling Modifies the Capillary-Like Organization of Human Embryonic Stem Cell-Derived Endothelium In Vitro Stem Cells, February 1, 2007; 25(2): 392 - 401. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. K. Martin, A. L. Dewar, A. N. Farrugia, N. Horvath, S. Gronthos, L. B. To, and A. C.W. Zannettino Tumor Angiogenesis Is Associated with Plasma Levels of Stromal-Derived Factor-1{alpha} in Patients with Multiple Myeloma Clin. Cancer Res., December 1, 2006; 12(23): 6973 - 6977. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 O. Salvucci, M. de la Luz Sierra, J. A. Martina, P. J. McCormick, and G. Tosato EphB2 and EphB4 receptors forward signaling promotes SDF-1-induced endothelial cell chemotaxis and branching remodeling Blood, November 1, 2006; 108(9): 2914 - 2922. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. M. Burns, B. C. Summers, Y. Wang, A. Melikian, R. Berahovich, Z. Miao, M. E. T. Penfold, M. J. Sunshine, D. R. Littman, C. J. Kuo, et al. A novel chemokine receptor for SDF-1 and I-TAC involved in cell survival, cell adhesion, and tumor development J. Exp. Med., September 4, 2006; 203(9): 2201 - 2213. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Mieno, M. Boodhwani, B. Ramlawi, J. Li, J. Feng, C. Bianchi, R. J. Laham, J. Li, and F. W. Sellke Human Coronary Microvascular Effects of Cardioplegia-Induced Stromal-Derived Factor-1{alpha} Ann. Thorac. Surg., August 1, 2006; 82(2): 657 - 663. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 I A Bhutto, D S McLeod, C Merges, T Hasegawa, and G A Lutty Localisation of SDF-1 and its receptor CXCR4 in retina and choroid of aged human eyes and in eyes with age related macular degeneration Br. J. Ophthalmol., July 1, 2006; 90(7): 906 - 910. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Kakinuma and S. T. Hwang Chemokines, chemokine receptors, and cancer metast |
), VEGF- and
bFGF-stimulated HUVECs
(
), unstimulated HAECs
(
), VEGF- and
bFGF-stimulated HAECs
(
). B:
Inhibition of the chemotactic response of HUVECs and HAECs toward
SDF-1
