| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Laboratoire de Migration Cellulaire et Infiltration Tumorale, Institut Albert Bonniot, Grenoble, France
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
2 integrin-mediated
leukocyte adhesion, and migration across EC junctions. This last step
or transmigration consists in leukocyte migration across endothelial
junctions. Little is known about the adhesion molecules and the
intracellular mechanisms implicated in transendothelial migration. This
process involves proteins expressed predominantly at intercellular
junctions, such as platelet EC adhesion molecule-1 (PECAM-1)
(2), a cell adhesion molecule of the Ig superfamily, which
has been shown to play a major role in leukocyte transmigration
(3). Proteins expressed at the apical surface of the cell
may also be involved, either by transmitting a signal or by
facilitating leukocyte migration toward the cell junctions
(4).
ICAM-1 (5) is a key molecule in leukocyte-endothelium
adhesion through its recognition of 2 integrin
counterreceptors CD11a/CD18 (LFA-1) (6) and CD11b/CD18
(Mac-1) (7). In addition to its interaction with
cell-associated counterreceptors, ICAM-1 recognizes fibrinogen (fg)
(8), an abundant plasma glycoprotein of 340 kDa, which is
also a ligand for the neutrophil 2 integrin
Mac-1 (9). We have previously shown that fg interacts
through its 
-chain (117–133)
(10) with the first Ig-like domain of ICAM-1
(11), and that fg enhances the attachment of leukocytes to
HUVEC by acting as a molecular bridge between the two cell types
(8). In addition, fg-ICAM-1 interactions have been shown
to enhance neutrophil transmigration, stimulating this process by 20-
to 30-fold (12), but it remained to be determined whether
this enhancement was due to a direct effect from endothelial or
leukocyte ICAM-1 or to a more complex pathway involving other adhesion
molecules expressed by EC.
Although Abs against ICAM-1 have been shown to inhibit transmigration in in vitro systems (13, 14, 15), these experiments failed to demonstrate the direct implication of ICAM-1 in leukocyte transmigration, as these mAbs may, in fact, block the prerequisite step of tight adhesion to endothelium. Thus, the implication of ICAM-1 in leukocyte transmigration is not clearly demonstrated. To investigate the specific contribution of ICAM-1-fg interaction in neutrophil emigration, we have transfected the cDNA encoding ICAM-1, PECAM-1, or the two in combination into Chinese hamster ovary (CHO) cells, a cell type not normally expressing these molecules, to obtain stable CHO cell lines. Using cells expressing ICAM-1 alone, we have been able to specifically study the role of ICAM-1 in neutrophil transmigration without the interference of PECAM-1, which is implicated in this step. The role of fg-ICAM-1 interactions in PMN transmigration was investigated in adhesion and transmigration assays in the presence of a well-known neutrophil chemoattractant, fMLP (16). In this paper, we show that ICAM-1 is able to support transmigration in the absence of other adhesion molecules expressed in EC. This enhancement of PMN migration across CHO ICAM-1 monolayers was fg dependent, as evidenced by the inhibition of transmigration obtained in the presence of fg-depleted plasma. This phenomenon was also blocked using function-blocking anti-ICAM-1 mAbs. These results demonstrate that ICAM-1, through its interaction with fg, is directly implicated in the process of neutrophil transmigration. Furthermore, a mutant of ICAM-1 lacking its cytoplasmic portion is not able to support PMN emigration, while PMN adhesion is not modified on these monolayers. This demonstrates that the enhanced transmigration induced by ICAM-1 is not a direct consequence of a facilitated adhesion. Finally, we show in this paper that ICAM-1-mediated transmigration involves the small GTP-binding protein Rho.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The following mAbs were used: mouse mAbs 3D6 and 2D5 to human
ICAM-1 (12); mouse Ab 5F4-10 (Hemeris, Grenoble, France)
to human PECAM-1. Mouse mAb OKM1 directed against
2 integrin Mac-1 was a kind gift from Dario
Altier (Boyer Center for Molecular Medicine, Yale University School of
Medicine). Anti-LFA-1 mAb 25.3.1 was obtained from Immunotech
(Marseille, France). Tetramethylrhodamine isothiocyanate and
FITC-conjugated goat anti-mouse IgG F(ab')2
secondary Abs were purchased from Jackson ImmunoResearch
(Asniéres, France). C3 transferase was kindly provided by M.
Popoff (Institut Pasteur, Paris, France).
Cell culture
HUVEC were isolated from human umbilical cords by collagenase
treatment as described by Jaffe et al. (17). Cells were
maintained in medium 199 (M199; BioWhittaker, Gagny, France) containing
20% FBS (Life Technologies, Cergy Pontoise, France), 100 U/ml
penicillin, 100 µg/ml streptomycin, 0.25 µg/ml amphotericin B (Life
Technologies), 50 µg/ml EC growth supplement (prepared from bovine
brain), and 100 µg/ml heparin (Sigma, St. Louis, MO). In some
experiments, HUVEC were stimulated with 200 U/ml TNF-
(Boehringer
Mannheim, Meylan, France) for 4 h at 37°C. PMN were isolated by
differential centrifugation on Ficoll-Hypaque gradients (Sigma),
followed by dextran sedimentation and hypotonic lysis of
erythrocytes.
Transfection experiments
A full-length cDNA clone encoding human PECAM-1 (18) in the mammalian expression vector pcDNA3 (Invitrogen, Groningen, The Netherlands) was transfected in wild-type (WT) CHO cells by electroporation. Recombinant CHO cells were selected in DMEM (BioWhittaker) containing 10% FBS, 2 mM L-glutamine, 100 U/ml penicillin, 100 µg/ml streptomycin, and nonessential amino acids (BioWhittaker) plus 1 mg/ml G418 (Geneticin; Life Technologies) for selection or 500 µg/ml for routine culture.
CHO cells were transfected with a full-length cDNA clone encoding human ICAM-1 (19) in the mammalian expression vector pRC/CMV by electroporation (Invitrogen, San Diego, CA), with selection of stable transfectants in DMEM complemented with 10% FBS, 100 U/ml penicillin, 100 µg/ml streptomycin, 1 mM L-glutamine, and nonessential amino acids plus G418 as described previously (10).
CHO cells expressing ICAM-1 cytoplasmic deletion mutant (CHO ICAM-1
cyt) were generated. A PCR procedure was performed to delete the
cytoplasmic domain of ICAM-1 using the pcDNA3 plasmid containing the
human ICAM-1 coding sequence. The following primers were used: a sense
primer
(5'-931TGTGCTGGAATTCCTCAGCC950-3')
containing the EcoRI restriction site and an antisense
primer
(5'-GCTCTAGATCA2467GTTATAGAGG2458-3')
containing a stop codon (underlined) and the XbaI
restriction site to facilitate subsequent cloning. The PCR product was
digested by EcoRI and XbaI and ligated into the
pcDNA3 vector containing native human ICAM-1 that was also cut by the
same enzymes. CHO cells were then transfected with the resulting
construction and recombinant cells were selected in DMEM (BioWhittaker)
containing 10% FBS, 2 mM L-glutamine, 100 U/ml
penicillin, 100 µg/ml streptomycin, and nonessential amino acids
(BioWhittaker) plus 1 mg/ml G418 (Geneticin; Life Technologies) for
selection or 500 µg/ml for routine culture.
CHO ICAM-1/PECAM-1 were obtained by transfection of CHO ICAM-1 cells with a full-length cDNA clone encoding human PECAM-1 in the mammalian expression pUT-SV1 (Eurogentec, Seraing, Belgium). Selection of stable transfectants was performed in DMEM complemented with 10% FBS, 1 mM L-glutamine, 100 U/ml penicillin, 100 µg/ml streptomycin, 500 µg/ml G418, and nonessential amino acids plus 50 µg/ml phleomycin (Eurogentec) for selection or 25 µg/ml for routine culture.
Phenotypic characterization of WT, ICAM-1, ICAM-1 cyt,
PECAM-1, or ICAM-1/PECAM-1 transfectants was conducted by flow
cytometry and immunofluorescence staining on adherent cells with
anti-ICAM-1 (3D6) or anti-PECAM-1 (5F4-10) mAbs.
Flow cytometry analysis
Single-cell suspensions obtained after trypsinization were centrifuged (400 x g, 5 min) and resuspended in medium containing 5% FBS. After incubation with 20 µg/ml primary mAbs for 30 min at 4°C, cell suspensions were washed and mixed with a FITC-conjugated secondary Ab (20 µg/ml) (Jackson ImmunoResearch) for an additional 30 min at 4°C. After washing, cells were analyzed on a Becton Dickinson FACScan (Mountain View, CA).
Immunofluorescence staining and confocal imaging
CHO cells were seeded on glass coverslips and grown to confluency before immunofluorescence staining. Glass coverslips were coated with 25 µg/ml human plasma fibronectin (prepared from human plasma, as described by Engvall and Ruoslahti (20)). Cells were fixed with 2% paraformaldehyde, permeabilized with 0.5% Triton X-100, and processed for indirect immunofluorescence microscopy. Incubation with the primary Ab was followed by incubation with tetramethylrhodamine isothiocyanate-conjugated secondary Ab (Jackson ImmunoResearch), with three washes in 1x PBS, Ca2+, Mg2+, 1% BSA, and 0.05% Tween 20 between the various steps. Coverslips were then mounted in Aquamount (BDH Laboratory Supplies, Dorset, U.K.) before microscopic observation. Cells were imaged using a Zeiss confocal microscope (Zeiss, Oberkochen, Germany). Cross-sectional reconstruction was accomplished with a Metamorph program (Princeton Instruments, Evry, France) to combine the entire Z series into a stack projection and to overlay and process the final image.
Permeability assay
Cells (104 at seeding in 100 µl) were cultured for 4 days on 6.5-mm diameter polycarbonate membranes of 0.4-µm pore size (Costar; Dominique Dutscher, Brumath, France), previously coated with human fibronectin (25 µg/ml) for 2 h at 37°C and rinsed with 1x PBS. Six hundred microliters of culture medium filled the lower compartment. Before the experiment, the culture medium of both the upper and lower compartments was replaced, without washing, with DMEM supplemented with 5% FBS (0.1 ml, upper chamber and 0.6 ml, lower chamber). Special culture medium for HUVEC (described above) was used for assessment of HUVEC permeability. HRP (35 nM; VI-A type, 44,000 m.w.; Sigma) was added in the upper compartment. After 1 h at 37°C, 10 µl of the upper compartment was collected and kept on ice until enzymatic activity of HRP was assayed using a Sigma Fast kit (Sigma). The reaction was allowed to proceed for 20 min at room temperature (RT) before measuring the absorbance at 490 nm. To control monolayer permeability in the same condition as in the transmigration assay, CHO monolayers were assessed for cellular integrity 3 days after seeding on 0.4-µm polycarbonate membranes. Medium from the lower compartments was replaced by DMEM/HEPES (20 mM, pH 7.4). The upper compartments were filled with the same medium plus plasma (1:5 dilution) and 100 µM D-phenylalanyl-L-prolyl-L-arginine chloromethyl ketone (Boehringer Mannheim). After 3 h at 37°C in 5% CO2, 7 µl of HRP at 0.5 µM was added in the upper compartment, followed by gentle agitation and a 30-min incubation at 37°C in 5% CO2. Quantification of HRP presence in the lower compartment was determined as described above.
Adhesion assay
WT or recombinant CHO cells were grown to confluency onto 48-well plates (Falcon; Becton Dickinson). In some cases, CHO cell monolayers were pretreated with a specific inhibitor of Rho proteins, C3 transferase (50 µg/ml in complete medium). After a 4 h-incubation at 37°C, cells were extensively washed before addition of labeled PMN. PMN (at 5 x 106/ml) were fluorescently labeled with 7.5 µM calcein acetoxymethyl (Molecular Probes, Leiden, The Netherlands) in RPMI 1640/HEPES (20 mM, pH 7.4) for 30 min at 37°C, washed in the same medium, and resuspended in RPMI 1640/HEPES. PMN were incubated with 10 µM fMLP (Calbiochem, La Jolla, CA), with or without purified fg at 300 µg/ml (Diagnostica Stago, Asnieres, France). After a 20-min incubation at RT, 2 x 105 calcein acetoxymethyl-labeled PMN were added to CHO cell monolayers for increasing time intervals between 10 and 45 min at RT. At the end of each incubation, cells were washed and fluorescent adherent cells were evaluated using a fluorescence plate reader (Fluoroskan Ascent; Labsystems, Helsinki, Finland) set at excitation 490 nm and emission 535 nm.
Transmigration assay
CHO cell monolayers were grown to confluency on fibronectin-coated porous membranes (6.5-mm diameter and 5-µm pore size; Costar) for 3 days at 37°C. CHO cells were then incubated in duplicate with PMN (2 x 105/well); to maintain the viability of human PMN, autologous normal human plasma (1:5 dilution in DMEM/HEPES (20 mM, pH 7.4) was added with the PMN in the upper compartment in the presence of 100 µM D-phenylalanyl-L-prolyl-arginine chloromethyl ketone (Boehringer Mannheim), which is a specific inhibitor of thrombin. A total of 2 x 10-8 M fMLP or 10-8 M IL-8 (Calbiochem) in DMEM/HEPES (20 mM, pH 7.4) was added in the lower compartment to create a chemotactic gradient for PMN. In some experiments, plasma was fg depleted before the experiment by incubation with 1 U/ml thrombin (from human plasma; Sigma) for 5 min at 37°C, followed by centrifugation (18 000 x g, 5 min). PMN were incubated with fg-depleted plasma alone or in the presence of purified fg at 300 µg/ml (Diagnostica Stago). In mAb inhibition experiments, PMN suspensions and cell monolayers were separately preincubated with anti-ICAM-1 mAbs 2D5 or 3D6 (20 µg/ml) for 5 min at RT for PMN and 10 min at 37°C for cell monolayers before the addition of autologous plasma. Anti-Mac-1 or anti-LFA-1 mAbs (20 µg/ml) were incubated for 10 min at RT with PMN before the addition of plasma and incubation on CHO monolayers. In assays performed with C3 transferase, monolayers were treated with C3 transferase (50 µg/ml, 37°C) before addition of PMN, following extensive washing and replacement into new medium; PMN were added after a 4-h incubation. Migrated neutrophils were recovered from the bottom of the well after 3 h at 37°C in 5% CO2, centrifuged, stained with trypan blue, and counted microscopically.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
We have developed stable CHO cell lines expressing ICAM-1 or PECAM-1 (CHO ICAM-1 and CHO PECAM-1, respectively) or the two adhesion molecules in combination (CHO ICAM-1/PECAM-1). Expression levels of ICAM-1 and PECAM-1 in the three recombinant cell lines were quantified by FACScan analysis.
As expected, WT CHO did not express ICAM-1 or PECAM-1 (Fig. 1
A), whereas CHO ICAM-1
expressed ICAM-1 at a high expression level comparable to
TNF--stimulated HUVEC (Fig. 1B). In CHO PECAM-1, PECAM-1
is expressed at nearly the same level as in HUVEC, where its expression
is constitutive. Similar expression levels of both proteins were found
for the cell line coexpressing ICAM-1 and PECAM-1 (Fig. 1A).
|
C) while PECAM-1 was
concentrated at the intercellular junctions (Fig. 1D),
indicating a correct localization of these recombinant proteins in CHO
cells. Similar results were found for the cell line coexpressing ICAM-1
and PECAM-1 (data not shown).
Since we intended to use these CHO cell lines for transmigration assays
instead of HUVEC monolayers, we assessed their capacity to form
confluent monolayers in comparison to resting or TNF--stimulated
HUVEC. The integrity of the monolayers and their function as barriers
were studied by a permeability assay using HRP (21).
Three days after seeding, resting and TNF--stimulated HUVEC
presented a low permeability level with 19.4 ± 0.1% and
21.6 ± 0.1% of controls, respectively (Fig. 2). CHO ICAM-1, CHO PECAM-1, and CHO
ICAM-1/PECAM-1 formed monolayers that were nearly impermeable to HRP,
with 12.9 ± 3.1%, 17.2 ± 4.1%, and 15.7 ± 4.4% of
controls, respectively; these values are similar to those obtained with
WT CHO monolayers (11.9 ± 1.1%) (Fig. 2). Permeability across WT
CHO or CHO ICAM-1 was also investigated in the conditions established
for transmigration assays. Monolayers were incubated for 3 h at
37°C with a 1:5 dilution of plasma (in the medium used in
transmigration) before the addition of HRP in the upper compartment to
verify if plasma could have any effect on monolayer integrity. In these
conditions, similar results were obtained with no significant
difference between the two cell lines (data not shown). These results
show that recombinant CHO cells are able to form correct, continuous
monolayers with a high cohesion of cell–cell contacts and a
permeability level comparable to that observed with EC. Finally, the
cellular location of recombinant proteins was identical to that
observed for ICAM-1 and PECAM-1 in EC.
|
To study the role of ICAM-1 by itself in neutrophil transmigration, we used CHO ICAM-1 monolayers in an in vitro transmigration assay. The implication of ICAM-1-fg interactions in our model was first studied in a protein-free medium with or without fg. In these conditions, few PMN were able to migrate across CHO monolayers after a 3-h incubation time period and most of migrated PMN were either dead or adherent to the well surface. To maintain the viability of human neutrophils, all subsequent experiments were done in the presence of autologous human plasma.
In preliminary experiments, we tested PMN transmigration across WT CHO
and CHO cells expressing either ICAM-1 or PECAM-1 or the two in
combination toward fMLP or IL-8 gradients, which are well-known
neutrophil chemoattractants (16). It has previously been
shown that PECAM-1 plays a crucial role in leukocyte transendothelial
migration during inflammation (3). Fig. 3 clearly shows that PECAM-1 expressed on
CHO cells can support transmigration, with a 4-fold increase in the
number of transmigrated cells in comparison to WT CHO (29.8 ±
4.5% vs 6.7 ± 3.9%), demonstrating that the function of PECAM-1
is maintained in these transfected cells.
|
). Similar results
were obtained when IL-8 was used instead of fMLP. All subsequent
experiments were conducted in the presence of a fMLP gradient. It can
be noted that, in our model, coexpression of the two proteins did not
provide any additional effect in comparison to CHO cells expressing
ICAM-1 or PECAM-1 alone (Fig. 3). These data clearly show that ICAM-1
is directly implicated in PMN transmigration.
We next investigated the effect of the absence of fg on PMN
transmigration. fg was depleted from human plasma by action of
thrombin, which is known to specifically degrade fg. Interestingly,
results in Fig. 4 show that the addition
of fg-depleted plasma reduced PMN transmigration across CHO ICAM-1 to
only 22 ± 3%, while no effect was observed on PECAM-1-dependent
transmigration or across WT CHO monolayers. This demonstrates that
ICAM-1-fg interactions are implicated in this process.
|
, addition of purified fg in fg-depleted plasma completely
restored PMN migration across CHO ICAM-1 monolayers, demonstrating that
fg depletion of plasma has no side effects and that the observed
inhibition was due to the lack of fg. Furthermore, depletion of fg had the same effect when the transmigration assay was performed toward a gradient of IL-8 (data not shown). This demonstrates that fg via its interaction with ICAM-1 is directly involved in the passage of PMN across the cell barrier, independently of the chemoattractant used.
Fg-dependent transmigration is blocked by anti-ICAM-1 mAbs
To better understand this mechanism, we used anti-ICAM-1 mAbs in transmigration assays. A panel of mAbs recognizing different ICAM-1 domains is available in our laboratory (12). Among these, 2D5 mAb binds to domain 1 of ICAM-1 and is known to completely suppress ICAM-1-fg interactions, while 3D6, which interacts with the second domain, has no effect on these interactions (11). The effect of 2D5 and 3D6 mAbs was tested on transmigration of PMN across CHO ICAM-1 and CHO PECAM-1 monolayers in the presence of human plasma.
As shown in Fig. 5, neither 2D5 nor 3D6
had inhibitory effects on PMN transmigration across WT CHO or CHO
PECAM-1 monolayers. However, 2D5 mAb completely inhibited
ICAM-1-dependent PMN transmigration across CHO ICAM-1 monolayers,
resulting in only 19.2 ± 6.1% of transmigrated PMN.
Interestingly, 3D6 did not have any effect on ICAM-1-mediated
transmigration. These inhibition experiments with 2D5 mAb provide
evidence for a direct role of fg-ICAM-1 interactions in neutrophil
transmigration.
|
PMN transmigration was almost totally abolished when 2D5 was added on
both PMN and CHO cells with only 4.7 ± 1.1% of transmigrated PMN
(Fig. 6). The inhibition was similar when
2D5 was added on CHO ICAM-1 alone (3.6 ± 2% of transmigrated
PMN), whereas 3D6 used in the same conditions, as a control, had no
effect. In contrast, preincubation of 2D5 with PMN alone did not
inhibit their transmigration (24.3 ± 3.2% of transmigrated PMN)
(Fig. 6). These results demonstrate that the enhancement of PMN
emigration is due to interaction of fg with ICAM-1 expressed on CHO
cells.
|
Because we wanted to understand the mechanism that regulates ICAM-1-mediated transmigration, we evaluated the implication of ICAM-1 in a signaling cascade.
Fg-dependent transmigration requires the presence of ICAM-1 cytoplasmic sequence
The ability of ICAM-1 to function as a signaling molecule has been
demonstrated in previous studies. In brain EC, ICAM-1 cross-linking
leads to tyrosine phosphorylation of cytoskeleton proteins and enhanced
lymphocyte migration across brain EC (22, 23). To assess
whether ICAM-1 might be implicated in a signaling cascade in our
system, we have produced a recombinant cell line expressing ICAM-1
without its cytoplasmic portion (CHO ICAM-1 cyt). Expression level
of this mutated protein was comparable to native ICAM-1 expressed in
CHO ICAM-1 (Fig. 7A).
Furthermore, this mutation did not have any effect on ICAM-1
localization on the cell (Fig. 7B) since we found the
protein expressed uniformly on the whole cell surface.
|
clearly show that ICAM-1 expressed
on CHO cells are able to support PMN adhesion in the presence of
purified fg, with a 2-fold increase of adherent cells in comparison to
WT CHO (63.3 ± 4.8% and 30 ± 3.3%, respectively).
Furthermore, the absence of fg led to a decrease in adherent PMN,
whereas no effect was observed on WT CHO monolayers. Interestingly, CHO
ICAM-1 cyt are able to support PMN adhesion in the presence of fg
(62 ± 2.7%) to the same level as that of native protein (Fig. 8A), while ICAM-1-dependent adhesion was inhibited in the
absence of fg (40.2 ± 5.3%). Surprisingly, PMN transmigration
across CHO ICAM-1 cyt monolayers was totally abolished in the
presence or absence of fg (2.3 ± 3.1% vs 0.2 ± 2.3%,
respectively) (Fig. 8B). This defect in PMN transmigration
is not due to an impaired interaction between fg and ICAM-1 cyt
protein. First, adhesion of PMN is enhanced in the presence of purified
fg, as it does on CHO ICAM-1 monolayers. Second, using anti-fg
polyclonal Ab followed by an immunofluorescence experiment, we have
demonstrated that fg was able to interact with WT ICAM-1 and mutated
ICAM-1 (data not shown).
|
-actinin, an actin-binding cytoskeletal
protein (24). The small GTP-binding protein Rho is a key
mediator of actin cytoskeletal remodeling induced by intracellular
signals (25).The observation that ICAM-1 can associate
with components of the actin cytoskeleton suggested that Rho protein
could be involved in regulating adhesion and transmigration mediated by
this receptor. Furthermore, endothelial ICAM-1 has been implicated
recently in lymphocyte migration across brain EC via a Rho-dependent
pathway (23). Since deletion of the ICAM-1 cytoplasmic
sequence dramatically inhibited PMN transmigration, the potential role
of Rho was assessed in our PMN transmigration assay. CHO cell
monolayers were treated with the bacterial toxin C3 transferase, which
is capable of entering the cells and specifically inactivating Rho
proteins by ADP ribosylation (26). As shown in Fig. 9A, treatment of the cells
with 50 µg/ml C3 transferase did not influence PMN adhesion on CHO
ICAM-1, whereas it significantly reduced PMN transmigration to the
level reached across WT CHO cells (16.15 vs 15.44%, respectively). The
same treatment had no effect on transmigration achieved across WT CHO
(Fig. 9B). These results clearly demonstrate that ICAM-1 is
implicated in a cascade of signaling events involving the small
GTP-binding protein Rho; this signaling pathway is probably involved in
the subsequent transmigration of leukocytes.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
To test the specific role of ICAM-1 in the process of transmigration,
we have engineered stable CHO cell lines expressing human ICAM-1 or
PECAM-1 or coexpressing both proteins. To be used as a model of
leukocyte transendothelial migration, these transfectants must share
common characteristics with EC expressing the native molecules in terms
of expression level, localization at the cell surface, and the ability
to form cohesive and confluent monolayers. In these recombinant cell
lines, we obtained expression levels similar to those found in EC, with
an ICAM-1 expression level comparable to TNF--stimulated HUVEC. As
revealed by immunofluorescence experiments, ICAM-1 is localized
uniformly at the cell surface, whereas PECAM-1 is concentrated at the
cell-cell contacts of these recombinant cells; these findings are in
agreement with ICAM-1 and PECAM-1 localization on EC and on recombinant
cells described in other studies (31, 32, 33, 34).
The results obtained with our model of CHO cells expressing ICAM-1
alone indicate that in addition to its well-known function in leukocyte
adhesion, ICAM-1 plays a direct role in the migration of these cells
across the junction. This enhanced transmigration mediated by ICAM-1
could be the end result of the increase in adhesion, with other
molecules expressed at the cell junctions being responsible for the
transmigration. Our results indicate that this is not the case. First,
if the above hypothesis was correct, CHO ICAM-1 would have an increase
in leukocyte adhesion, with no modification of the transmigration
level, while CHO ICAM-1/ PECAM-1 would be expected to show an even
higher increase in transmigration. However, transmigration across CHO
ICAM-1/PECAM-1 or ICAM-1 monolayers was very similar, suggesting that
in our system, ICAM-1 participates extensively in PMN transmigration,
independently of PECAM-1. This is supported by the fact that
CD31-deficient mice exhibit leukocyte transendothelial migration to the
same degree as WT mice (35), thus suggesting the existence
of a CD31-independent pathway to sustain leukocyte migration. Second,
results obtained with cells expressing ICAM-1 cyt clearly
demonstrate that adhesion can be dissociated from transmigration. Cells
transfected with this mutant protein are able to support adhesion, to a
level similar to the one obtained with WT ICAM-1, but can no longer
support PMN transmigration.
One surprising result was that there was no additive effect of ICAM-1 and PECAM-1 in our model of recombinant CHO cells. One possibility is that ICAM-1 is already stimulating transmigration to its maximum and that the addition of PECAM-1 produces no further effect. This is supported by the fact that in our model, ICAM-1-mediated transmigration reached the same level as PMN transmigration across CHO PECAM-1 monolayers.
All of the above experiments were done in the presence of autologous plasma to overcome the problem of neutrophil viability in the absence of protein. As we have previously shown that fg-ICAM-1 interactions enhance leukocyte transmigration across resting HUVEC monolayers (12), we have performed experiments to address the question of the importance of fg-ICAM-1 interactions using our recombinant model. Replacement of normal plasma with fg-depleted plasma led to a decrease of the observed transmigration to control level, demonstrating the involvement of fg-ICAM-1 interaction in neutrophil transmigration. Addition of purified fg to fg-depleted plasma counteracted the inhibition due to the depletion of fg. These data clearly demonstrate the fundamental implication of fg in ICAM-1-mediated neutrophil transmigration. In contrast, fg depletion has no effect on PMN transmigration across WT CHO or CHO PECAM-1 monolayers.
To further confirm the implication of fg-ICAM-1 interaction in PMN
transmigration, we used anti-ICAM-1 mAbs. 2D5, which interacts with
domain 1 and inhibits fg-ICAM-1 recognition (12), had no
effect on PECAM-1-induced transmigration but completely suppressed PMN
migration across CHO ICAM-1. Interestingly, this 2D5 inhibitory effect
was observed when CHO monolayers were preincubated with the mAb, but
not when PMN alone were preincubated. These results indicate that the
enhanced PMN transmigration is directly mediated by ICAM-1 expressed on
CHO cells. Using anti-Mac-1 Abs, we conclude that on PMN, fg
interacts with the 2 integrin Mac-1, as
previously demonstrated by Altieri et al. (9).
The previous data are consistent with the hypothesis of an
intracellular signal in the cell monolayer. The mechanism by which
fg-ICAM-1 interaction leads to the increase of transmigrated
neutrophils remains to be determined. Several mechanisms can explain
this marked increase. Interaction between endothelial ICAM-1 and a
ligand expressed on leukocytes or a soluble ligand, like fg, could
induce an intracellular signaling cascade leading to the opening of
intercellular junctions associated with leukocyte transmigration.
Indeed, Hicks et al. (36) have shown that fg, via its
binding to ICAM-1, initiates signaling pathways leading to the
synthesis of vasoactive mediators. ICAM-1 has also been implicated in
signal transduction mechanism involving tyrosine phosphorylation of
cytoskeleton-associated proteins (22, 37), resulting in
enhanced lymphocyte migration across brain EC monolayers
(23). To test the hypothesis of an intracellular signaling
pathway, we investigated the implication of ICAM-1 cytoplasmic domain.
Interestingly, in the presence or absence of fg, ICAM-1 cyt protein
supports PMN adhesion, whereas ICAM-1-mediated PMN transmigration is
totally abolished.
This result is consistent with a role for ICAM-1 in intracellular signaling events which may lead to facilitated PMN infiltration into tissues. Since the Rho signaling pathway was previously implicated in lymphocyte migration across brain EC (23), we investigated its implication in our model of PMN transmigration by pretreatment of CHO monolayers with C3 transferase, a specific inhibitor of Rho proteins. PMN adhesion is not affected by C3, whereas ICAM-1-mediated transmigration is totally abolished. These results demonstrate that ICAM-1 is implicated in a signaling cascade involving Rho proteins expressed in CHO cells. As previously reported (38), Rho itself may be activated through cell surface signals propagated through the actin cytoskeleton. A possible mechanism is that ICAM-1-fg interactions induced strong adhesion of PMN on cell monolayers which led to intracellular signals involving ICAM-1 cytoplasmic domain and Rho proteins. One hypothesis to explain the observed effect of fg-ICAM-1 interaction on PMN transmigration is that the subsequent activation of Rho might induce some modifications of junctional proteins which may facilitate PMN infiltration.
Another likely possibility is that neutrophils may migrate more easily on the surface of EC using fg-ICAM-1 interactions, leading to an increase in the number of leukocytes reaching the intercellular junctions. This hypothesis is supported by the demonstration that VCAM-1 was able to increase monocyte transendothelial chemotaxis (4) by facilitating lateral migration of monocytes on EC. We are currently investigating this possibility.
Taken together, these results demonstrate the implication of ICAM-1 in neutrophil transmigration via its interaction with fg. It is the first time that the role of ICAM-1 by itself is directly demonstrated in the process of neutrophil transmigration, without the interfering presence of other adhesion molecules expressed by EC. In addition, we demonstrate that this effect is mediated through the ICAM-1 cytoplasmic domain and the activation of Rho.
Further studies are now in progress to elucidate the mechanisms underlying this phenomenon. In addition, this recombinant CHO model will be useful to precisely determine which amino acids in the ICAM-1 sequence are important in fg-dependent transmigration.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Address correspondence and reprint requests to Dr. Emmanuelle Sans or Dr. Alain Duperray, Laboratoire de Migration Cellulaire et Infiltration Tumorale, Institut Albert Bonniot, Domaine de la Merci, 38706 La Tronche Cedex, France.
3 Abbreviations used in this paper: EC, endothelial cell; PECAM-1, platelet/EC adhesion molecule-1; fg, fibrinogen; CHO, Chinese hamster ovary; WT, wild type; RT, room temperature.
Received for publication July 18, 2000. Accepted for publication October 3, 2000.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
chain sequence (117–133) by intercellular adhesion molecule-1 mediates leukocyte-endothelium interaction. J. Biol. Chem. 270:696.-actinin. J. Cell Biol. 118:1223.This article has been cited by other articles:
|
|
 |
|
  E. Kanters, J. van Rijssel, P. J. Hensbergen, D. Hondius, F. P. J. Mul, A. M. Deelder, A. Sonnenberg, J. D. van Buul, and P. L. Hordijk Filamin B Mediates ICAM-1-driven Leukocyte Transendothelial Migration J. Biol. Chem., November 14, 2008; 283(46): 31830 - 31839. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. D. van Buul, M. J. Allingham, T. Samson, J. Meller, E. Boulter, R. Garcia-Mata, and K. Burridge RhoG regulates endothelial apical cup assembly downstream from ICAM1 engagement and is involved in leukocyte trans-endothelial migration J. Cell Biol., September 24, 2007; 178(7): 1279 - 1293. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. D. van Buul, E. Kanters, and P. L. Hordijk Endothelial Signaling by Ig-Like Cell Adhesion Molecules Arterioscler. Thromb. Vasc. Biol., September 1, 2007; 27(9): 1870 - 1876. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. D. Sithu, W. R. English, P. Olson, D. Krubasik, A. H. Baker, G. Murphy, and S. E. D'Souza Membrane-type 1-Matrix Metalloproteinase Regulates Intracellular Adhesion Molecule-1 (ICAM-1)-mediated Monocyte Transmigration J. Biol. Chem., August 24, 2007; 282(34): 25010 - 25019. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. M. Rao, L. Yang, G. Garcia-Cardena, and F. W. Luscinskas Endothelial-Dependent Mechanisms of Leukocyte Recruitment to the Vascular Wall Circ. Res., August 3, 2007; 101(3): 234 - 247. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H.-M. Oh, S. Lee, B.-R. Na, H. Wee, S.-H. Kim, S.-C. Choi, K.-M. Lee, and C.-D. Jun RKIKK Motif in the Intracellular Domain Is Critical for Spatial and Dynamic Organization of ICAM-1: Functional Implication for the Leukocyte Adhesion and Transmigration Mol. Biol. Cell, June 1, 2007; 18(6): 2322 - 2335. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. Celli, J.-J. Ryckewaert, E. Delachanal, and A. Duperray Evidence of a Functional Role for Interaction between ICAM-1 and Nonmuscle {alpha}-Actinins in Leukocyte Diapedesis J. Immunol., September 15, 2006; 177(6): 4113 - 4121. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. J. Chancey, K. V. Khanna, J. F. M. L. Seegers, G. W. Zhang, J. Hildreth, A. Langan, and R. B. Markham Lactobacilli-Expressed Single-Chain Variable Fragment (scFv) Specific for Intercellular Adhesion Molecule 1 (ICAM-1) Blocks Cell-Associated HIV-1 Transmission across a Cervical Epithelial Monolayer J. Immunol., May 1, 2006; 176(9): 5627 - 5636. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Q. Wang, M. Yerukhimovich, W. A. Gaarde, I. J. Popoff, and C. M. Doerschuk MKK3 and -6-dependent activation of p38{alpha} MAP kinase is required for cytoskeletal changes in pulmonary microvascular endothelial cells induced by ICAM-1 ligation Am J Physiol Lung Cell Mol Physiol, February 1, 2005; 288(2): L359 - L369. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
V. L. Tsikitis, N. A. Morin, E. O. Harrington, J. E. Albina, and J. S. Reichner The Lectin-Like Domain of Complement Receptor 3 Protects Endothelial Barrier Function from Activated Neutrophils J. Immunol., July 15, 2004; 173(2): 1284 - 1291. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. A. Adams, M. Passino, B. D. Sachs, T. Nuriel, and K. Akassoglou Fibrin Mechanisms and Functions |
