Vol. 283, Issue 4, C1196-C1205, October 2002
Role of 
v
3-integrin in
TNF--induced endothelial cell migration
Baochong
Gao1,2,3,
Thomas M.
Saba1,2, and
Min-Fu
Tsan3
1 Department of Physiology and
2 Center for Cell Biology and Cancer Research,
Albany Medical College, Albany, New York 12208, and
3 Laboratory of Cell Physiology, Veterans Affairs
Medical Center, Washington, District of Columbia 20422
 |
ABSTRACT |
Tumor necrosis factor-
(TNF-), one
of the major inflammatory cytokines, is known to influence endothelial
cell migration. In this study, we demonstrate that exposure of calf
pulmonary artery endothelial cells to TNF- caused an increase in the
formation of membrane protrusions and cell migration. Fluorescence
microscopy revealed an increase in v
3
focal contacts but a decrease in 51 focal
contacts in TNF--treated cells. In addition, both cell-surface and
total cellular expression of v3-integrins
increased significantly, whereas the expression of
51-integrins was unaltered. Only focal
contacts containing v3- but not
51-integrins were present in membrane
protrusions of cells at the migration front. In contrast, robust focal
contacts containing 51-integrins were present in cells behind the migration front. A blocking antibody to
v3, but not a blocking antibody to
5-integrins, significantly inhibited TNF--induced
cell migration. These results indicate that in response to TNF-,
endothelial cells may increase the activation and ligation of
v3 while decreasing the activation and
ligation of 51-integrins to facilitate
cell migration, a process essential for vascular wound healing and angiogenesis.
integrins; focal contacts; tumor necrosis factor-
| |
INTRODUCTION |
ADHESION AND
MIGRATION are distinct functions of endothelial cells essential
for maintaining the integrity of the endothelium and repairing or
forming blood vessels during wound healing or angiogenesis. The balance
between adhesion and migration is precisely regulated in response to
changing environments in the blood stream. Strong adhesion to the
extracellular matrix is required for resting endothelial cells to
maintain the integrity of the endothelium (8, 21, 31),
whereas modulated adhesion to the matrix is necessary to facilitate
cell migration (17, 18, 44). One way cells can modulate
the strength of adhesion and facilitate migration is to change the
expression and distribution of integrins on the cell surface.
Functional cell-surface integrins are complexes of an - and a
-subunit. More than 20 integrin complexes have been identified representing different combinations of at least 16 - and 8 -subunits (19, 31). The difference in the subunit
composition determines the specificity of the integrin complex for its
substrate in the matrix. For example, the
51-integrin complex essentially interacts only with fibronectin in the matrix, whereas the
61 complex interacts preferentially with
laminin (19, 28, 32). Some integrin complexes have
multiple preferred substrates in the matrix. One example is
v3-integrin, which interacts with
vitronectin and fibronectin, as well as laminin. Integrins
51 and v3
are predominant integrin complexes expressed in endothelial cells (8, 37). Both integrin complexes have been implicated in endothelial cell adhesion and migration (17, 18, 33, 42).
The regulation of cell adhesion and migration involves coordinated
events including cell signaling, cytoskeleton rearrangement, and
surface integrin redistribution. These cellular events are known to be
influenced by inflammatory cytokines such as tumor necrosis factor-
(TNF-). TNF- is a 17-kDa polypeptide that forms homotrimers on
the cell surface. It is synthesized and secreted by many cell types
upon stimulation with a variety of toxins and cytokines including
TNF- itself. Activated macrophages and monocytes are major sources
of TNF-, and a primary target of this specific cytokine is the
endothelial cell (23, 25, 36, 38).
Over the past decade, considerable effort has been focused on
TNF--induced apoptosis, whereas the mechanism of
TNF--induced endothelial cell migration is relatively unknown.
Studies show that TNF- can display either proangiogenic or
antiangiogenic effect depending on experimental conditions (12,
22, 26). One of these conditions appears to be the dosage or
concentration of TNF- used in vivo or in vitro. It promotes the
formation of tubular structure at relatively low dosages but becomes
inhibitory to angiogenesis and induces apoptosis at relatively
high dosages (12, 22, 29). In vitro, TNF-
concentrations between 100 and 250 units/ml induced the highest levels
of tubule formation, whereas tubule formation was significantly reduced
at TNF- concentrations higher than 500 units/ml (43).
TNF- concentrations around 250 units/ml were also observed in the
blood of patients with serious inflammation and sepsis or in healthy
human subjects challenged with endotoxin (36, 39).
Accordingly, we used TNF- at a concentration known to induce cell
migration to identify the role of cell-surface integrins in
TNF--induced endothelial cell migration.
| |
MATERIALS AND METHODS |
Materials.
Bovine pulmonary artery endothelial (CPAE) cells were obtained from
American Type Culture Collection (Manassas, VA). Recombinant human
TNF- (20 units/ng) was obtained from Cellular Products (Buffalo,
NY). Monoclonal antibodies to v3 (clone
LM609), 51 (clone HA5), and actin
(MAB1501) and polyclonal antibodies to 5- (AB1928) and
3- (AB1932) integrins were obtained from Chemicon International, (Temecula, CA). The blocking antibody to
5 (clone BIIG2) was developed by C. H. Damsky and
obtained from the Developmental Studies Hybridoma Bank established
under the auspices of the National Institute of Child Health and Human
Development (NICHD) and maintained by the Department of Biological
Sciences, The University of Iowa, Iowa City, IA. All integrin
antibodies used in this study recognize both activated and nonactivated
form of integrins. Protease inhibitors phenylmethylsulfonyl fluoride
(PMSF) and N-p-tosyl-L-lysine
chloromethyl ketone (TLCK) were purchased from Sigma (St. Louis, MO).
Endothelial cell culture.
CPAE cells at passage 16 were cultured as described
previously (14). The cells were cultured in minimum
essential medium (MEM; GIBCO Invitrogen, Carlsbad, CA) containing 20%
fetal bovine serum (FBS; GIBCO Invitrogen). TNF- exposure was
carried out in MEM containing 5% FBS. All cells used in this study
were cultured to confluence and treated with or without TNF- at 200 units/ml for 18 h before analysis (migration assay, adhesion
assay, immunofluorescence microscopy, or immunoprecipitation).
Determination of membrane protrusion formation and cell migration
with an in vitro wound-healing assay.
Confluent endothelial cells on glass coverslips were treated with or
without TNF-, and wounds were created on cell monolayers by using
the "scratch wound" protocol (10, 15, 34) with a razor
blade. The debris was removed by washing the cells with serum-free MEM,
and the cells were incubated in a 37°C incubator for 5 h in
serum-free MEM. The cells were photographed, and the number of
migrating cells and the percentage of cells with membrane protrusions
were determined under an inverted microscope. A total of nine areas
were selected randomly on each coverslip under a 40× objective. Cells
on three to six coverslips of either control or TNF--treated sample
were quantified in each experiment. To detect integrins in focal
contacts, the cells were fixed, permeabilized, and incubated with
antibodies to v3 (LM609) or
51 (HA5) and fluorescence-labeled
secondary antibodies (Molecular Probes, Eugene, OR).
To determine the effects of blocking antibodies on cell migration,
confluent endothelial cells on glass coverslips were treated with or
without TNF- and scraped with a razor blade. The debris was removed
by washing the cells with serum-free MEM. The cells were then incubated
in a 37°C incubator for 5 h in the presence or absence of
blocking antibodies to either 5 (BIIG2)- or
v3 (LM609)-integrin complexes. The number
of cells migrated into the wound area was determined as described above.
Determination of membrane protrusion formation with cell adhesion
assay.
Human fibronectin was purified from cryoprecipitate (American Red
Cross) by using geletin-sepharose affinity chromatography according to
the procedure of Engvall and Ruoslahti (11). Human cryoprecipitate (15 ml) was diluted 1:1 with the column equilibration buffer and loaded onto a 10-ml gelatin-sepharose column (Pharmacia Biotech, Piscataway, NJ) at a flow rate of 0.5 ml/min. The column was
washed with 1 M NaCl in phosphate-buffered saline (PBS) and eluted with
4 M urea in the washing buffer. The eluted fraction was dialyzed
overnight in 0.2 M phosphate buffer, pH 7.4, and the fibronectin
concentration was determined by using the extinction coefficient
= 12.8.
Glass coverslips in 12-well plates were incubated overnight with
purified fibronectin at 2 µg/ml in coating buffer (50 mM NaHCO3, pH 9.6) at 4°C. Endothelial cells treated with or
without TNF- were lifted into suspension with trypsin-EDTA buffer
(GIBCO Invitrogen) and seeded onto fibronectin-coated or noncoated
coverslips at 105 cells/well. The cells were incubated in
either serum-free medium on fibronectin-coated surfaces or MEM
containing 20% FBS on noncoated surfaces at 37°C for 30 min.
Nonadhered cells were removed by washing with PBS, and adhered cells
were examined and photographed under an inverted microscope. Cells with
membrane protrusions were quantified as described above.
Determination of the effect of blocking antibodies on endothelial
cell adhesion on fibronectin-coated surfaces.
Endothelial cells in suspension were preincubated with blocking
antibodies to either v3 (LM609)- or
5 (BIIG2)-integrins on ice for 30 min before being
seeded onto glass coverslips coated with 2 µg/ml fibronectin.
Coverslips coated with 10 µg/ml bovine serum albumin (BSA) were used
as controls for nonspecific adhesion. Cells were incubated in
serum-free medium at 37°C for 30 min. Nonadhered cells were removed
by washing with PBS. The number of adhered cells was determined by
counting under an inverted microscope as described above.
Determination of cell-surface integrin expression by surface
biotinylation, immunoprecipitation, and Western blotting.
Confluent CPAE cell monolayers treated with or without TNF- were
labeled with Biotin (Pierce, Rockford, IL) at 0.5 mg/ml in PBS for 60 min at 4°C. Cells were then lysed in the lysis buffer (150 mM NaCl, 5 mM EDTA, 1% sodium deoxycholate, 1% Triton X-100, and 20 mM Tris at
pH 7.4) containing protease inhibitors (0.3 mM PMSF and 0.1 mM TLCK).
The cell lysate was clarified by centrifugation in a Microfuge and
precleared by incubation with protein G agarose (GIBCO Invitrogen).
Integrins v3 or
51 were immunoprecipitated with
antibodies LM609 and HA5, respectively, followed by incubation with
protein G agarose. The agarose-bound integrins were solubilized in
boiled SDS-gel sample buffer under nonreducing conditions and clarified
by spinning in a Microfuge. Precipitated integrins were separated on
two identical 7.5% SDS gels and transferred onto two nitrocellulose
membranes. One membrane was used to determine cell-surface integrins
with streptavidin conjugated to horseradish peroxidase and enhanced
chemiluminescence (ECL) Western blotting detection solutions (both from
Amersham, Piscataway, NJ). The other membrane was used to
determine total cellular integrins in biotinylated cells with
antibodies to either 5 (AB1928) or 3
(AB1932) and ECL Western blotting detection solutions. The bands on
films were quantified by densitometric scanning using a BioRad imaging
densitometer (Bio-Rad, Hercules, CA).
Determination of total cellular integrin expression by
immunoprecipitation and Western blotting.
Confluent CPAE cells treated with or without TNF- were lysed in the
lysis buffer, and v3- or
51-integrins were immunoprecipitated from
the cell lysate by using monoclonal antibodies to the integrins as
described above. Precipitated integrins were separated on SDS gels and
transferred onto nitrocellulose membranes. The nitrocellulose membranes
were probed for either 5- or 3-integrins
by using polyclonal antibodies (AB1928 and AB1932). The integrins were quantified by densitometric scanning of Western blot films. The amount
of protein in cell lysate used in immunoprecipitation was determined on
a separate gel and Western blot probed for actin by using antiactin
antibody MAB1501.
Immunofluorescence microscopy.
CPAE cells cultured on coverslips were fixed with 3% formaldehyde,
permeabilized in 0.5% Triton, and stained with either an antibody
against human 51-integrin (clone HA5) or
an antibody to human v3-integrin (clone
LM609) at 2 µg/ml. This was followed by incubations with secondary
antibodies conjugated to Alexa-488 (Molecular Probes, Eugene, OR). The
coverslips were mounted with ProLong Anti-Fade (Molecular Probes) and
examined under a BX60 fluorescence microscope (Olympus, Melville, NY)
and photographed using a SPOT digital camera (Diagnostic Instruments,
Sterling Heights, MI).
Statistical analysis.
All measurements were performed at least three times with duplicate
samples. Results are presented as means ± SD. Levels of significance are determined by a two-tailed Student's
t-test (13), and a confidence level of >95%
(P < 0.05) was used to established statistical significance.
| |
RESULTS |
Effect of TNF- on endothelial cell migration and the formation
of membrane protrusions.
We examined the migration of endothelial cells treated with TNF- at
200 units/ml for 18 h, because previous studies indicate that
functional changes in endothelial monolayers occur between 12 and
24 h of TNF- exposure at this dosage. These functional changes
include dissociation of 51-integrins from
focal contacts (14, 30), increased recycling of integrins
(14), reduced cell adhesion to fibronectin
(30), cell-cell gap formation (7, 14, 30),
and increase in protein permeability (6, 7, 40).
The migration of endothelial cells was evaluated by using a
well-established in vitro wound-healing assay (10, 15,
34). In these experiments, endothelial cell monolayers treated
with or without TNF- were wounded with a razor blade. After a 5-h incubation, the number of cells migrated into the wound area was determine under an inverted microscope. Results showed that
TNF--treated cells displayed a significant increase in cell
migration (Fig. 1,
A-C). In
addition, increased formation of membrane protrusions was observed in
TNF--treated cells at the migration front (Fig. 1, A,
B, and D), suggesting a possible role of the
membrane protrusions in the increased cell migration.
View larger version (90K):
[in this window]
[in a new window]
|
Fig. 1.
Tumor necrosis factor- (TNF-) induced increases in cell
migration and in the formation of membrane protrusions. Confluent
endothelial cell monolayers were treated with (B) or without
(A) TNF- at 200 units/ml for 18 h, and cell
monolayers were wounded with a razor blade. The cells were then
incubated for 5 h at 37°C in a serum-free medium. The number of
cells migrated into the open area (C) and the percentage of
cells with membrane protrusion (D) were determined under an
inverted microscope. Results represent means ± SD of 5 experiments. *P < 0.05. Bar, 100 µm.
|
|
We next asked the question whether the increased formation of membrane
protrusion was a characteristic of all TNF--treated cells, not only
cells at the migration front. One way to answer this question is to
determine the formation of membrane protrusions in a cell adhesion
assay under subconfluent conditions. To determine the effect of matrix
proteins, we determined the formation of membrane protrusions on
surfaces coated with fibronectin, a common substrate for both
v3- and
51-integrins. In this experiment, cells
in monolayers treated with or without TNF- were lifted into
suspension and seeded onto glass coverslips coated with or without
fibronectin. Cells were then incubated briefly at 37°C either in a
serum-free medium on fibronectin-coated surfaces or in the presence of
20% serum on noncoated surfaces. Nonadhered cells were removed by a
washing with PBS. Adhered cells were photographed under an inverted
microscope. Results in Fig. 2 show
clearly that TNF--treated cells display more membrane protrusions
than control cells in this assay on both fibronectin-coated and
noncoated surfaces. These observations under subconfluent conditions
suggest that increased formation of membrane protrusions may be a
characteristic of all TNF--treated cells, not only cells at the
migration front in the wound-healing assay.
View larger version (68K):
[in this window]
[in a new window]
|
Fig. 2.
Increased formation of membrane protrusions was a
characteristic of all TNF--treated endothelial cells. Confluent
endothelial cell monolayers treated with (C and
D) or without (A and B) TNF- (200 units/ml, 18 h) were lifted into suspension and seeded onto glass
coverslips. Cells were incubated at 37°C for 30 min either in a
serum-free medium on fibronectin-coated surfaces (Fn-coated,
A and C) or in the presence of 20% serum on
noncoated surfaces (noncoated, B and D). The
percentage of cells with membrane protrusions was quantified
(E) under an inverted microscope. Results represent
means ± SD of 5 experiments. *P < 0.05. Bar, 100 µm.
|
|
Effect of TNF- on the localization of
v3- and
51-integrins in focal contacts.
v3 and 51
are predominant integrin complexes expressed in endothelial cells.
These integrin complexes have been shown to play important roles in
cell migration (31, 32, 37). It is possible that the
increased formation of membrane protrusions and cell migration after
TNF- exposure were mediated by an increase in the ligation of these
integrins. If this were the case, one would expect to see
integrin-containing focal contacts in membrane protrusions, especially
in cells at the migration front.
To test this possibility, endothelial cells treated with or without
TNF- were assayed for cell migration as in Fig. 1, and cells were
fixed and stained with antibodies recognizing either v3- or
51-integrin complexes. As shown in Fig.
3, only
v3-containing focal contacts were
detected in membrane protrusions of cells at the migration front (Fig.
3, B and D). In contrast,
51-integrins in cells at the migration
front were observed only in structures resembling endocytic vesicles
(Fig. 3, A and C), not in focal contacts.
However, focal contacts containing
51-integrins were readily identified in
cells immediately behind the migration front (Fig. 3, A and
C). In comparison, control cells not treated with TNF-
had much fewer v3 focal contacts and
significantly lower levels of membrane protrusion formation (Fig. 3,
compare B and D). These data support the notion
that TNF--induced formation of membrane protrusion and cell
migration may rely on the increase in the ligation of
v3-integrins.
View larger version (115K):
[in this window]
[in a new window]
|
Fig. 3.
Focal contacts containing
v3- but not
51-integrins were observed in cells at
the migration front. Confluent endothelial cells on coverslips treated
with (C and D) or without (A and
B) TNF- were wounded as described in Fig. 1. The cells
were incubated for 5 h at 37°C in a serum-free medium before
being fixed and incubated with antibodies to either
51 (A and C)- or
v3 (B and
D)-integrins. Arrows point to some of the focal contacts
containing either v3 (B and
D)- or 51 (A and
C)-integrins. Arrowheads in A and C
indicate antibody-labeled 51-integrins in
structures resembling endocytic vesicles. Bar, 50 µm.
|
|
To obtain a closer look at the formation of
v3 focal contacts on migrating cell, we
carried out a time course of cell migration into the wound area and
compared the rate of migration of control and TNF--treated
endothelial cells. Results (Fig. 4)
showed that cell migration could be detected in TNF--treated cell
monolayers 1 h after wounding (Fig. 4E), whereas
similar levels of cell migration were not observed until 4 h after
wounding in control monolayers (Fig. 4C). In addition, focal
contacts containing v3-integrins formed
in all migrating cells, especially on membrane protrusions (Fig. 4,
C-H).
View larger version (78K):
[in this window]
[in a new window]
|
Fig. 4.
Focal contacts containing
v3 were expressed on all migrating cells.
Confluent endothelial cell monolayers on coverslips treated with
(E-H) or without
(A-D) TNF- were wounded as in Fig. 1. The
cells were incubated at 37°C for 1 (A and E), 2 (B and F), 4 (C and G), or
8 h (D and H) in a serum-free medium before
being fixed and labeled with an antibody to
v3-integrins. Arrows in
C-H point to some of the
v3-containing focal contacts. Bar, 100 µm.
|
|
An important question was whether the increased
v3 and decreased
51 focal contacts occurred not only in
cells at the migration front but also in cells in confluent monolayers.
To answer this question, endothelial monolayers treated with or without
TNF- were fixed and stained with antibodies recognizing both
activated and nonactivated form of either
v3- or
51-integrins. Results show (Fig.
5) that the expression of
v3-integrins in control cells in
monolayers is only detectable at cell-cell junctions (Fig.
5B), whereas v3-containing
focal contacts can be readily identified around the cell periphery in
TNF--treated cells (Fig. 5D). In addition,
TNF--treated cells also display increased gap formation, suggesting
a loss of cell-cell interactions after cell monolayers were exposed to
TNF-. This is consistent with earlier observations under similar
conditions (6, 14, 30). Increased v3 focal contacts around the periphery of
the TNF--treated cells may be a cellular response to increase cell
adhesion in compensating the loss of cell-cell interactions. In
contrast to v3 focal contacts, robust
51-containing focal contacts were detected in control cells (Fig. 5A), and an apparent
decrease in 51 focal contacts was
observed in TNF--treated cells (Fig. 5C). These results
indicate that TNF- caused an increase in the activation/ligation of
v3 and a decrease in the activation/ligation of
51-integrins in all TNF--treated
endothelial cells, not only in cells at the migration front.
View larger version (156K):
[in this window]
[in a new window]
|
Fig. 5.
TNF- induced a decrease in
51- but an increase in
v3- containing focal contacts on cells in
monolayers. Confluent endothelial cells cultured on coverslips treated
with (C and D) or without (A and
B) TNF- were fixed and stained with antibodies to either
51 (A and C) or
v3 (B and D).
Arrows point to some of the focal contacts containing either
51 (A and C)- or
v3 (D)-integrins. Arrows in
B indicate some of the
v3-integrins expressed at cell-cell
junctions of control cells. Bar, 50 µm.
|
|
Effect of TNF- on the expression of
v3- and
51-integrins.
Changes in focal contacts observed in Figs. 3 and 4 could have been
caused by changes in cell-surface expression and/or total cellular
expression of the integrins. However, individual integrins cannot be
detected by microscopy unless they have been recruited into focal
contacts. We therefore investigated TNF--induced changes in the
expression of v3- and
51-integrins in endothelial cells using
biochemical approaches.
Integrins can display different activation states, and the state of
integrin activation is influenced by their interactions with ligands,
antibodies, and cations such as Mn2+ (1, 24, 28,
41). Binding of a ligand or Mn2+ can switch an
integrin complex from a "low-affinity state" (nonactivated form) to
a "high-affinity state" (activated form). The transition of the
affinity states involves conformational changes of the integrins, which
can be detected by specific antibodies recognizing motifs exposed only
when integrins are activated. To quantify the expression of all forms
of integrins, we used antibodies to recognize both activated and
nonactivated forms of integrins for immunoprecipitation.
To determine the surface expression of the integrins, the cell surface
was first biotinylated and then v3- or
51-integrins were immunoprecipitated from
the cell lysate. The immunoprecipitated integrins were then quantified
by Western blotting using streptavidin conjugated to horseradish
peroxidase (Fig. 6). To determine the total cellular expression of the integrins,
v3 or 51
was immunoprecipitated from the whole cell lysate and the integrins
were quantified by Western blotting using antibodies to either
5- or 3-integrin subunit (Fig.
7). These antibodies were used to
quantify 51- and
v3-integrin complexes, because
5-subunit has been found only in
51 complexes and 3-
subunit forms complexes only with v in endothelial cells
(9, 28, 31).
View larger version (34K):
[in this window]
[in a new window]
|
Fig. 6.
TNF- induced an increase in cell-surface expression of
v3- but not
51-integrins. The surface of confluent
endothelial cells treated with or without TNF- was labeled with
biotin, and 51- or
v3-integrin complexes were
immunoprecipitated from the cell lysate. Cell-surface integrins were
quantified by Western blotting using horseradish peroxidase conjugated
streptavidin (A). Total cellular integrins were determined
on separate Western blots using antibodies to either 5-
or 3-integrins (B). Shown in A and
B are representative Western blots. Results from
densitometric scanning expressed as optical density (OD) ratios of
cell-surface integrins to total integrins are plotted in C.
Values represent means ± SD of 3 experiments. *P < 0.05.
|
|
View larger version (31K):
[in this window]
[in a new window]
|
Fig. 7.
TNF- induced an increase in total cellular expression
of v3- but not
51-integrins. Endothelial cells treated
with or without TNF- were lysed, and
51- or
v3-integrin complexes were
immunoprecipitated from the cell lysate. Precipitated integrins were
quantified by Western blotting using antibodies to 5- or
3-integrins (A). The cell lysate used for
immunoprecipitation was analyzed on a separate gel to probe for actin
on Western blots (B). Shown in A and B
are representative Western blots. Results from densitometric scanning
expressed as OD ratios of integrins to total actin are plotted in
C. Values represent means ± SD of 3 experiments.
*P < 0.05.
|
|
Results (Figs. 6 and 7) indicated that TNF- caused a significant
increase in both the cell-surface and total cellular expression of
v3-integrins. In contrast, the expression
of 51-integrins did not change
significantly despite the clear decrease in
51-containing focal contacts observed
(Fig. 5). Thus the increase in
v3-containing focal contacts observed in
TNF--treated endothelial cells (Figs. 3-5) was at least
partially due to the increased surface expression of the integrins. On
the other hand, the data were consistent with the concept that an
inactivation, rather than a decrease in surface expression of
51-integrins, was the basis for the reduction of focal contacts containing
51-integrins observed in TNF--treated
endothelial cells (Fig. 5).
Effect of blocking antibodies to v3-
and 51-integrins on TNF--induced cell
migration.
The above observations suggest that increased
v3-containing focal contacts may have
served as anchors for membrane protrusions, without which membrane
protrusions may retract and cell migration may be abolished. If this
were true, one would expect to see an attenuation of cell migration
when the v3-ligand interactions are blocked.
To test this hypothesis, the cell migration assay was performed in the
presence of blocking antibodies to either
v3- or 51-integrins. A blocking antibody to
5-subunit was used to block the function of
51-integrin complexes, because
5-subunit has only been found in
51-integrin complexes (28,
31). We first examined the effect of the antibodies on cell
adhesion to determine the concentration at which the antibodies can act
effectively. Because fibronectin is a substrate for both
v3- and
51-integrins in the matrix, we determined
whether the antibodies could block cell adhesion on fibronectin-coated
surfaces. We observed that both blocking antibodies inhibited the
adhesion of endothelial cells with significant blocking effects
observed at 5 µg/ml for anti-v3 and a
fivefold dilution for anti-5 antibodies (Fig. 8).
View larger version (35K):
[in this window]
[in a new window]
|
Fig. 8.
Blocking antibodies to v3-
or 5-integrins inhibited the adhesion of endothelial
cells in a concentration-dependent manner. Confluent endothelial cells
were lifted into suspension and preincubated with or without blocking
antibodies to either v3 or to
5 at indicated concentrations for 30 min on ice. Cells
were diluted into serum-free MEM and seeded at the same cell density on
glass coverslips precoated with 2 µg/ml fibronectin. Coverslips were
coated with BSA as background controls. Nonadhering cells were removed
by a washing with PBS after a 30-min incubation at 37°C. Adhered
cells were counted under an inverted microscope. Results represent
means ± SD of 3 experiments. *P < 0.05.
|
|
We next determined the effect of the blocking antibodies on cell
migration. Results showed that cell migration was inhibited in
TNF--treated cells by the blocking antibody to
v3-integrins in a concentration-dependent
manner (Fig. 9). In contrast, the blocking antibody to 5-integrins had little effect on
TNF--induced cell migration (Fig.
10), even at concentrations that
significantly inhibited cell adhesion (Fig. 8). These observations
suggest that v3-integrins play an
important role in TNF--induced cell migration.
View larger version (68K):
[in this window]
[in a new window]
|
Fig. 9.
The blocking antibody (Ab) to
v3-integrins significantly inhibited
TNF--induced cell migration in a concentration-dependent manner.
Confluent endothelial cells on coverslips were treated with
(B, D, and F) or without
(A, C, and E) TNF-, and cell
monolayers were wounded as in Fig. 1. Cells were then incubated at
37°C for 5 h in the presence of LM609, a blocking Ab to
v3-integrins, at indicated
concentrations. The number of cells migrated into the wound area was
determined under an inverted microscope (G). Results
represent means ± SD of 3 experiments. *P < 0.05. Bar, 100 µm.
|
|
View larger version (72K):
[in this window]
[in a new window]
|
Fig. 10.
The blocking Ab to 5-integrins had no
significant effect on TNF--induced cell migration. Confluent
endothelial cells on coverslips were treated with (B,
D, and F) or without (A, C,
and E) TNF-, and cell monolayers were wounded as in Fig.
1. Cells were then incubated at 37°C for 5 h in the presence of
BIIG2, a blocking Ab to 5-integrins, at indicated
dilutions. The number of cells migrated into the wound area was
determined under an inverted microscope (G). Results
represent means ± SD of 3 experiments. Bar, 100 µm.
|
|
| |
DISCUSSION |
The results presented in the current study demonstrated that
TNF- at 200 units/ml, a concentration commonly found in severely septic patients, could cause endothelial cells to increase the formation of membrane protrusions and cell migration. These changes were accompanied by an increase in both cell-surface and total cellular
expression of v3-integrins. In contrast,
the expression of 51-integrins remained
unchanged. The increased formation of membrane protrusions and cell
migration in TNF--treated cells was facilitated by the increased
expression of v3 on the cell surface and
increased recruitment of v3-integrin into
focal contacts. Several lines of evidence presented in this study
support these conclusions. First, a significant increase in
v3-integrin expression was detected on
the surface of TNF--treated endothelial cells. Second, a marked
increase in v3-containing focal contacts was observed after cells were exposed to TNF-. Third, only
v3-containing focal contacts, but not
51-containing focal contacts, were
detected in membrane protrusions of cells at the migration front.
Fourth, a blocking antibody to
v3-integrins, but not a blocking antibody to 5-integrin subunit, significantly inhibited
TNF--induced cell migration.
The development of inflammation is mediated by cytokines released upon
bacterial infection. Proinflammatory cytokines such as TNF- mediate
vascular inflammation by inducing cell-cell and cell-matrix
dissociation of endothelial cells (7, 14, 23, 30). In
vitro, the dissociation of either cell-cell or cell-matrix interactions
can cause increased protein permeability across the endothelial
monolayer (4, 7, 30, 40). This may be the basis for the
increased endothelial protein permeability across the endothelium
observed in vivo with inflammation and sepsis. A similar process occurs
in the formation of new blood vessels. Angiogenic factors such as VEGF
cause cell-cell and cell-matrix dissociation followed by migration and
proliferation of endothelial cells (5). On the other hand,
many angiogenic factors have also been shown to cause increased
permeability across the endothelial monolayer and inflammatory response
(5, 9, 43). It is therefore likely that both processes
share a part of the same cell-signaling pathway.
TNF- has been shown to induce the release of metalloproteinases
(35), vascular endothelial growth factor A (VEGF-A), and interleukin-8 (43), all of which are potent angiogenic
factors. TNF- has also been shown to modulate the expression of VEGF
receptors (16, 26). The current study has demonstrated a
possible involvement of integrin signaling in TNF--induced cell
migration via a coordinated regulation between
v3- and
51-integrins. On the other hand, it is
well known that the angiogenic effect of TNF- varies with cell lines
and experimental conditions (12, 22, 26). Therefore it
remains to be determined whether the TNF--induced coordinated regulation of v3- and
51-integrins observed in CPAE cells also
occurs in other endothelial cell lines or under in vivo conditions.
The integrin complex v3 interacts with a
wide range of matrix proteins. It is, however, not expressed at high
levels compared with 51 on resting
endothelial cells (37). A likely reason for its increased
expression on TNF--treated cells is to allow cells to survive on a
changing matrix. Resting endothelial cells produce a fibronectin-rich
matrix both in vivo and in vitro, and their interactions with the
matrix are mediated predominately by
51-integrins (8, 9). TNF-
has been shown to cause the release of proteinases that can modify the
matrix of endothelial cells (35). This matrix modification
may be one reason for the observed decreased localization of
51-integrins and the increased localization of v3-integrins in focal contacts.
The current study demonstrated changes in
v3 surface expression and focal contacts
in endothelial cells after TNF- exposure. It also suggested a
possible coordinated regulation on the expression and ligation of two
different integrins. This is evident not only in protein expression but
also in the localization of these integrins in focal contacts. Integrin
v3 was detected only at cell-cell junctions in untreated cells, whereas focal contacts containing v3-integrins were readily identified in
cells after TNF- exposure. In contrast,
51-integrins were present in robust focal
contacts in untreated cells, and the number of
51 contacts was dramatically reduced
after TNF- exposure. No focal contacts containing
51-integrins were observed in membrane
protrusions of cells at the migration front. These coordinated changes
in v3- and
51-integrins induced by TNF- may
mediate the observed membrane protrusion formation and cell migration.
Considerable evidence suggests that signaling among integrins is
modulated by "cross talk" mediators. Integrin
v5-mediated vitronectin internalization
appeared to require the ligation of 51-integrins (27). Ligation
of v3-integrins was found to suppress
51-mediated activation of
calcium/calmodulin-dependent protein kinase II (CamKII)
(2), which appeared to be required for integrin-mediated
phagocytosis and cell migration. CamKII at high levels, however, may
inhibit the interaction of 51-integrin with fibronectin (3). Kim et al. (20)
demonstrated that the ligation of
51-integrins could potentiate
v3-mediated endothelial cell migration on
vitronectin by suppressing the activity of protein kinase A. It is
possible that the differential regulation on the expression of
51- and
v3-integrins induced by TNF- is
mediated by a cross-talk mediator. Future studies to identify such a
mediator may provide a better understanding of the mechanism by which
TNF- induces the increase in
v3-integrin expression and endothelial cell migration, processes that may be essential for vascular wound healing and angiogenesis.
| |
ACKNOWLEDGEMENTS |
We thank Kara L. Powell and Alice Damrau-Abney for technical
assistance and Debbie Moran for administrative assistance.
| |
FOOTNOTES |
This study was supported by research grant RG-133N (B. Gao) from the
American Lung Association of New York, National Institute of General
Medical Sciences Grant GM-21447 (T. M. Saba), and a Veterans
Affairs Merit Review Award (M.-F. Tsan).
Address for reprint requests and other correspondence: B. Gao, VA Medical Center (10R), 50 Irving St., N.W., Washington, DC 20422 (E-mail: baochong.gao{at}med.va.gov).
The costs of publication of this
article were defrayed in part by the
payment of page charges. The article
must therefore be hereby marked
"advertisement"
in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
June 26, 2002;10.1152/ajpcell.00064.2002
Received 12 February 2002; accepted in final form 19 June 2002.
| |
REFERENCES |
1.
Bazzoni, G,
and
Hemler M.
Are changes in integrin affinity and conformation overemphasized?
Trends Biochem Sci
23:
30-34,
1998[ISI][Medline].
2.
Blystone, SD,
Slater SE,
Williams MP,
Crow MT,
and
Brown EJ.
A molecular mechanism of integrin crosstalk: v3 suppression of calcium/calmodulin-dependent protein kinase II regulates 51 function.
J Cell Biol
145:
889-897,
1999[Abstract/Free Full Text].
3.
Bouvard, D,
Molla A,
and
Block MR.
Calcium/calmodulin-dependent protein kinase II controls 51-integrin-mediated inside-out signaling.
J Cell Sci
111:
657-665,
1998[Abstract].
4.
Brett, J,
Gerlach H,
Nawroth P,
Steinberg S,
Godman G,
and
Stern D.
Tumor necrosis factor/cachectin increases permeability of endothelial cell monolayers by a mechanism involving regulatory G proteins.
J Exp Med
169:
1977-1991,
1989[Abstract/Free Full Text].
5.
Bussolino, F,
Matovani A,
and
Persico G.
Molecular mechanisms of blood vessel formation.
Trends Biochem Sci
22:
251-256,
1997[ISI][Medline].
6.
Chen, R,
Gao B,
Huang C,
Olsen B,
Rotundo R,
Blumenstock F,
and
Saba TM.
Transglutaminase-mediated fibronectin multimerization in lung endothelial matrix in response to TNF-.
Am J Physiol Lung Cell Mol Physiol
279:
L126-L138,
2000.
7.
Curtis, TM,
McKeown-Longo PJ,
Vincent PA,
Homan SM,
Wheatley EM,
and
Saba TM.
Fibronectin attenuates increased endothelial monolayer permeability after RGD peptide, anti-51, or TNF- exposure.
Am J Physiol Lung Cell Mol Physiol
269:
L248-L260,
1995[Abstract/Free Full Text].
8.
Defilippi, P,
Silengo L,
and
Tarone G.
Regulation of adhesion receptor expression in endothelial cells.
Curr Top Microbiol Immunol
184:
87-98,
1993[Medline].
9.
Dejana, E,
Raiteri M,
Resnati M,
and
Lampugnati MG.
Endothelial integrins and their role in maintaining the integrity of the vessel wall.
Kidney Int
43:
61-65,
1993[ISI][Medline].
10.
Dimmeler, S,
Dernbach E,
and
Zeiher AM.
Phosphorylation of the endothelial nitric oxide synthase at ser-1177 is required for VEGF-induced endothelial cell migration.
FEBS Lett
477:
258-262,
2000[ISI][Medline].
11.
Engvall, E,
and
Ruoslahti E.
Binding of soluble form of fibroblast surface protein, fibronectin, to collagen.
Int J Cancer
20:
1-5,
1977[ISI][Medline].
12.
Fajardo, LF,
Kwan HH,
Kowaiski J,
Prionas SD,
and
Allison AC.
Dual role of tumor necrosis factor in angiogenesis.
Am J Pathol
140:
539-544,
1992[Abstract].
13.
Fisher, RA,
and
Yates F.
Statistical Tables for Biological, Agricultural and Medical Research (6th ed.). Edinburgh: Oliver and Boyd, 1963.
14.
Gao, B,
Curtis T,
Blumenstock F,
Minnear F,
and
Saba T.
Increased recycling of 51-integrins by lung endothelial cells in response to tumor necrosis factor.
J Cell Sci
113:
247-257,
2000[Abstract].
15.
Galvez, BG,
Matias-Roman S,
Albar JP,
Sanchez-Madrid F,
and
Arroyo AG.
Membrane type 1-matrix metalloproteinase is activated during migration of human endothelial cells and modulates endothelial motility and matrix remodeling.
J Biol Chem
276:
37491-37500,
2001[Abstract/Free Full Text].
16.
Giraudo, E,
Primo L,
Audero E,
Gerber HP,
Koolwijk P,
Soker S,
Klagsbrun M,
Ferrara N,
and
Bussolino F.
Tumor necrosis factor- regulates expression of vascular endothelial growth factor receptor-2 and of its co-receptor neuropilin-1 in human vascular endothelial cells.
J Biol Chem
273:
22128-22135,
1998[Abstract/Free Full Text].
17.
Horwitz, AR,
and
Parsons JT.
Cell migration
movin' on.
Science
286:
1102-1103,
1999[Free Full Text].
18.
Huttenlocher, A,
Ginsberg MH,
and
Horwitz AF.
Modulation of cell migration by integrin-mediated cytoskeletal linkages and ligand-binding affinity.
J Cell Biol
134:
1551-1562,
1996[Abstract/Free Full Text].
19.
Hynes, RO.
Integrins: versatility, modulation, and signaling in cell adhesion.
Cell
69:
11-25,
1992[ISI][Medline].
20.
Kim, S,
Harris M,
and
Varner JA.
Regulation of integrin v3-mediated endothelial cell migration and angiogenesis by integrin 51 and protein kinase.
J Biol Chem
275:
33920-33928,
2000[Abstract/Free Full Text].
21.
Lampugnani, MG,
Resnani M,
Dejana E,
and
Marchisio PC.
The role of integrins in the maintenance of endothelial monolayer integrity.
J Cell Biol
112:
479-490,
1991[Abstract/Free Full Text].
22.
Leibovich, SJ,
Polverini PJ,
Shepard HM,
Wiseman DM,
Shively V,
and
Nuseir N.
Macrophage-induced angiogenesis is mediated by tumor necrosis factor .
Nature
329:
630-632,
1987[Medline].
23.
Mantovani, A,
Bussolino F,
and
Dejana E.
Cytokine regulation of endothelial cell function.
FASEB J
6:
2591-2599,
1992
TOP
ABSTRACT
INTRODUCTION
