1. No category

From www.bloodjournal.org by guest on November 19, 2014. For personal use only.
1994 84: 1116-1123
Clustering of vitronectin and RGD peptides on microspheres leads to
engagement of integrins on the luminal aspect of endothelial cell
membrane
A Zanetti, G Conforti, S Hess, I Martin-Padura, E Ghibaudi, KT Preissner and E Dejana
Updated information and services can be found at:
http://www.bloodjournal.org/content/84/4/1116.full.html
Articles on similar topics can be found in the following Blood collections
Information about reproducing this article in parts or in its entirety may be found online at:
http://www.bloodjournal.org/site/misc/rights.xhtml#repub_requests
Information about ordering reprints may be found online at:
http://www.bloodjournal.org/site/misc/rights.xhtml#reprints
Information about subscriptions and ASH membership may be found online at:
http://www.bloodjournal.org/site/subscriptions/index.xhtml
Blood (print ISSN 0006-4971, online ISSN 1528-0020), is published weekly by the American
Society of Hematology, 2021 L St, NW, Suite 900, Washington DC 20036.
Copyright 2011 by The American Society of Hematology; all rights reserved.
From www.bloodjournal.org by guest on November 19, 2014. For personal use only.
Clustering of Vitronectin and RGD Peptides on Microspheres
Leads to Engagement of Integrins on the Luminal Aspect of
Endothelial Cell Membrane
By Adriana Zanetti, Grazia Conforti, Sibylle Hess, In& Martin-Padura, Elena Ghibaudi, Klaus T. Preissner,
and Elisabetta Dejana
In previous work (Conforti et al. Blood 80:437,1992). we
have shown that integrins in endothelial cells (EC) are not
polarized to the basal cell membrane, but are also exposed
on the apical cell surface, in contact with blood. Therefore,
endothelial integrins might be available for binding circulating plasmaproteins.Howeversolubleplasmavitronectin
(vn) bound very poorlyto EC apical surfaceand this interaction was unaffected by Arg-Gly-Asp (RGD) peptides oran
anti-avp3 serum. In contrast, beads(diameter, 4.5 pm) coupled with plasmavnassociated to EC apicalsurfacein a
time- and concentration-dependent way. Addition of antibodies directed to vn, avp3, and RGD-containing peptides
blocked the interaction ofvnbeads with EC. In contrast,
heparinandantibodiesdirected to avp5 and p 1 integrin
chain had no effect. Beads coupled with Gly-Arg-Gly-AspSer-Pro boundto the EC surface, but not those coupledwith
Gly-Arg-Gly-Glu-Ser-Pro. Thisinteraction was blockedby
avp3 antibodies and RGD peptides, but not by avp5 antibody. Overall, these results indicate that luminal avp3 retains its bindingcapacityforsurface-linked vn and RGDcontaining ligands, but binding is observed only when the
ligand is offered
in a clustered,multivalent form. We propose
boundto circuthat when vn or RGD-containing proteins are
promoting
lating cells, they can act asbridging molecules by
adhesion ofthe cells to the endothelium via apical integrins.
0 1994 by The American Society of Hematology.
T
integrin receptors not only on their abluminal, but also on
their luminal surface. The distribution of integrins on the
endothelialhlood interface might be an important mechanism for binding plasma proteins and circulating cells on
their surface. However, direct evidence that apically located
endothelial integrins are functionally active is still lacking.
To address this issue we investigated vitronectin (vn)
binding to EC. Vn can bind to cells through at least two
domains; the Arg-Gly-Asp (RGD sequence) located close to
the aminoterminal domain of the molecule serves as binding
site for integrins during cellular adhesion to immobilized vn,
whereas the heparin-binding domain within the carboxytermina1 portion mediates binding of soluble vn complexes to
c e k x Vn occurs in at least two conformational configurations: The plasma formthat is intramolecularly stabilized
and does not expose the heparin-binding domain, and the
multimeric form that, after treatment with urea or low pH,
undergoes a conformational transition and exposes the cryptic heparin-binding site. The RGD sequence is apparently
exposed on both forms and is notaffected by conformational
changes of vn.'""
In the present report, we present evidence that integrins
on the luminal aspect of EC can bind soluble vn neither in
its plasma norinthe multimeric form. However, integrin
binding could be detected when vn was present in a highly
aggregated state or immobilized onbeads. Thus, because
specific binding of the circulating plasma form of vn to EC is
very weak, if existent at all, EC integrins might discriminate
between soluble and clustered forms of RGD ligands.
HE ENDOTHELIUM is the innermost lining of the entire cardiovascular system and forms a crucial permselective barrier between bloodand the underlying tissues.
Like other types of epithelia, endothelial cells (EC) are functionally polarized. The luminal front of EC facing the blood
is continuously exposed to circulating cells and plasma components, whereas the opposite abluminal side is in contact
withthe extracellular matrixand the surrounding tissues.
The EC luminal surface can modulate important processes
such as binding and inactivatiodactivation of many elements
of the coagulation and fibrinolytic systems.'.' In addition,
EC can apically expose adhesive receptors for circulating
EC adhesion to the subendothelium is mediated by specific
adhesive receptors located on the basal aspect of the membrane. Mostof these belong to the integrin family of adhesive
molecules' and specifically recognize and bind subendothelial components.
In previous investigations we have shown, by multiple
experimental approaches in vitro and in situ,' that ECexpose
From lstituto di Ricerche Farmacologiche Mario Negri, Milano,
Italy; CEA, Laboratoire d'Hematologie, Department de Biologie
Moleculaire et Structurale, INSERM U217, CEN-G,Grenoble,
France; and Haemostasis Research Unit, Kerckhoff-Klinik ManPlanck-Institut, Bad Nauheirn, Germany.
Submitted October 15, 1993; accepted April 11, 1994.
Supported by The Italian National Research Council (P. F. Applicazioni Cliniche della Ricerca Oncologica), Telethon Project No.
A.03, Associazione Italiana per la Ricerca sul Cancro, and by Commission of the European Communities (BRIDGE: BIOT-CT90-0195).
I.M.P. is a recipient of a Science Plan of European Communities
fellowship.
Address reprint requests to Adriana Zunetti, MD, Instituto di Ricerche Farmacologiche Mario Negri, Via Eritrea 62, 20157 Milano,
Italy.
The publication costs of this article were defrayed in part by page
charge payment. This article must therefore be hereby marked
"advertisement" in accordance with 18 U.S.C. section 1734 solely to
indicate this fact.
0 1994 by The American Society of Hematology.
0006-4971/94/8404-0022$3.00/0
1116
MATERIALS AND METHODS
Materials. Chemical reagents were purchased from the following sources: bovine serum albumin (BSA), unfractionated heparin
sodium salt from porcine intestinal mucosa, insulin-transfenin-sodium selenite media supplement for cell culture, antimouse polyclonal IgG-peroxidase conjugate, crystal violet, o-phenylenediamine
dihydrochloride, TRIS (hydroxymethyl) aminomethane, 2-deoxy-Dglucose, Tween-20 (polyoxyethylenesorbitan monolaurate) (Sigma
Chemical Co., St Louis, MO); heparan sulfate was a kind gift from
Dr L. A. Fransson, University of Lund, Lund, Sweden; culture reagents (GIBCO, Paisley, UK); tissue culture plates and flasks (Falcon, Becton Dickinson Labware, NJ, and Costar, Cambridge, MA);
Blood, VOI 84, NO 4 (August 15). 1994: pp 1116-1 123
From www.bloodjournal.org by guest on November 19, 2014. For personal use only.
1117
BINDING OF PLASMAVN-COATEDBEADS
carrier-free NaIz5I(New England Nuclear, Boston, MA); paraformaldehyde, sucrose and urea (Merck, Darmstadt, Germany) Diff-Quick
fixing solution (Men + Dade AG, Dudingen, Switzerland); Dynal
M-450, tosyl activated magnetic beads and Dynal magnetic device
(Dynal AS, Oslo, Norway); IODO-GEN (Pierce Europe BV, OudBeijerland, The Netherlands); column PD-l0 (Pharmacia LKB Biotechnology, Uppsala, Sweden).
Antibodies. Some of the antibodies usedin this study were
kindly donated by the following investigators: anti-avg3 monoclonal
antibody (MoAb) LM609 and anti-av MoAb LM142 under the form
of ascitic fluid" by Dr D. A. Cheresh (The Scripps Research Institute,
La Jolla, CA); anti-P3 MoAb 7E3 purified immunoglobulins (IgGs)
by Dr B. S. Coller (State University of New York at Stony Brook);
anti-Dl MoAb Lia ~upematant'~
by Dr F. Sanchez-Madrid (Hospital de la F'rincesa,Madrid, Spain); anti-vn MoAb 16A7 purifiedIgGs
by Dr P. Declerck, (University of Leuven, Leuven, Belgium). Rabbit
sera raised by injecting placenta-purified avg3 and a5p1 receptors
and anti-avp5 MoAb (clone P1F6)I4 under the form of ascitic fluid
were from Telios, (Pharmaceuticals Inc, San Diego, CA). Irrelevant
MoAbs (ascitic fluid or purified IgG) and preimmune rabbit serum
were used as control.
In blocking experiments we used antibody concentrations able to
induce maximal effect as reported in the pertinent references.
Anti-vn MoAb BVI and MoAb BV2, supernatants and purified
IgGs, were prepared in our laboratory by standard hybridoma techniques using purified human denatured vnas immunogen. The
MoAbs were characterized by their ability to bind purified vn in the
enzyme-linked immunosorbent assay (ELISA) technique" (see Fig
1, A and B) and by immunoprecipitation" (not shown). The ELISA
technique was performed essentially as des~ribed.'~
Briefly, microtiter wells were coated with 30 pL of 7 pg/mL solution of either
plasma vn and multimeric vn in carbonate bicarbonate buffer pH
9.6 and incubated overnight at 4°C in humidified chamber. Plates
were washed three times with 0.05% Tween 20 in phosphate-buffered saline (PBS) containing 0.5% BSA and then incubated with
the same buffer containing 1% BSA (100 &/well) for 1 hourat
37°C. After wash, the plates were incubated with different concentrations of anti-vn: MoAbs BV1, BV2, and nonimmune mouse IgGs
(0.001 to 50 pg/mL) in PBS 0.05% Tween 20 containing 0.5% BSA
for 1 hour at room temperature.
The wells were washed three times with PBS, incubated with
peroxidase-conjugate goat antimouse IgGs (0.15 pg/mL, 1 0 0 pL/
well) for 1 hour at room temperature. Binding of IgGs was quantified
by addition of chromogenic substrate as described below for soluble
vn binding to EC. As reported in Fig 1, A and B, MoAbs BVI
and BV2 recognized vn bothin its plasma and denatured form.
Recognition of plasma vn was slightly better than that of multimeric
vn. MoAbs BV1 and BV2 showed a comparable activity.
Both MoAbs BV1 and BV2 inhibited EC adhesion and spreading
on vn-coated (7 pg/mL) substrates (Fig 1C). These assays have been
previously described in detail." Briefly, wells were precoated with
7 pg/rnL of multimeric vn (2 hours at 37°C) residual protein binding
sites on wells were saturated by further incubation with 1% BSA in
PBS (30 minutes at 37°C). EC (18 X 103/well)in suspension were
added to the wells in the presence of increasing concentration of
MoAbs BVI, BV2, and nonimmune mouse IgGs. After 2 hours at
37°C unbound cells were removed by washing twice in PBS-Ca"
Mg2' and adherent cells were fixed and stained with crystal violet
as described." The number of adherent and spread cells was evaluated by phase-contrast microscopy (at least IO fields per well were
counted). As reported in Fig IC, MoAb BV2 was more effective
than MoAb BVl in inhibiting cell adhesion and spreading.
By ELISA, MoAb BV 1 selectively recognized human vn, but not
bovine vn (measured by coating the wells with undiluted bovine
serum, not shown).
O.D.
1.4
0.8
I."
O.D.
A
1.2
l 7
1.4
1.2
1 .o
0.8
0.6
0.4
0.2
O/O
NI IgG
60
BV2
0
100
200
ANTIBODIES pg/ml
Fig 1. Characterizationof MoAbs BV1 and BV2. Bindingof MoAbs
B V l and BV2 to multimeric vn (A) and plasma vn(B).Increasing
concentrations of MoAbs BV1 (A) and BV2 10)or nonimmune mouse
lgGs (0)
were addedto vn- (7 pg/mL) coatedplates for 1 hour
at room
temperature. The amount of antibody bound was then evaluated
by ELISA technique as described in Materials and Methods. Values
expressed in optical density (OD) are means If: SD of three replicates.
(C) Effect of MoAbs BV1 and BV2 on EC adhesion and spreading to
vn-coated substrate.EC adhasion assaywas performed as described
previourly." EC in suspension were added to the wells in the presence of increasingconcentrationof BV1 (A, A), BV2 (0. and nonimmune mouse lgGs (0. 0).The number of adherent (A, U, 0)
and
spread U , m, 0 )cells was evaluated by phase-contrast microscopy.
Values (means of two experiments performed in triplicate) are expressed as percentage of the number of cells adherent and spread
in absence of antibodies.
From www.bloodjournal.org by guest on November 19, 2014. For personal use only.
1118
Proteins and peptides. Humanplasma vn was purified under
nondenaturing conditions as previously described.Ix Multimeric
vn
was prepared by incubating plasma vn for 1 hour at 37°C in 6 moll
L urea followed by extensive dialysis against PBS. Plasma vn was
radiolabeled with "'l by the iodogen procedure" reaching specific
radioactivity of 1.4 pCi/pg protein. Synthetic hexapeptides Gly-ArgGly-Asp-Ser-Pro (GRGDSP) and Gly-Arg-Gly-Glu-Ser-Pro (GRGESP)wereeithersynthesized
in our laboratory as described'" or
purchased
from
Telios.
The
synthetic
(cyclical)
peptide
GlyPen-Gly-Arg-Gly-Asp-Ser-Pro-Cys-Ala
(GPenGRGDSPCA) was
obtainedfromTelios.ThecyclicpeptidesGly-Pen-Arg-Ala-ArgGly-Asp-Asn-Pro-Cys-Ala
(GPenRARGDNPCA)
and
Arg-PmcGly-His-Lys-Gly-Flu-Leu-Arg-Cys-Arg
R(Pmc)GHKGELRCR were
a kind gift of Dr J. F. Tschopp (Telios Pharmaceuticals Inc, San
Diego, CA). GRGDSP and GRGESP coupled to albumin were prepared as described."
EC culture. EC were isolated from normal-term
umbilical cord
veins by collagenase perfusion as previously described in detail."
Cells were grown in tissue culture dishes coated with gelatin (0.5%)
in 20% newborn calf serum-medium 199 supplemented with endothelial cell growth supplement (prepared from bovine brain) (50 p,g/
mL), heparin (100 pg/mL), and penicillin (100 U/mL)/streptomycin
(100 pg/mL) and used between the second and
the third passage.
For binding assays, cells were seeded in gelatin-coated 96-microtiter
wells and used at confluency (2 X IO4 cells/well).
Gelatin was used because preliminary experiments showed
that
under these conditions, RGD peptides
or anti-avp3 serum (within
the time frame and concentrationsused in this study) did not induce
a detectable detachment of the cells from the wells.
Soluble vn binding to EC. Microtiter wells of confluent EC were
incubated overnight with 1 0 0 pllwell of medium 199, 4% BSA, and
0.08% insulin-transferrin-sodium selenite supplement. Medium was
then removed and cells were incubated with increasing amounts of
either plasma vn or multimeric vn in 60 p1 medium 199,0.5% BSA,
0.01% sodium azide at 37°C. Sodium azide was included to prevent
vn internalization. In a few experiments, plasma vn and multimeric
vn were added to the cells in PBS-Ca" Mg", 0.5% BSA, 0.01%
azide, and 2-deoxy-D-glucose (S0 mmol/L) with comparable results.
Afterdifferentincubationtimes,medium
was removed,thecells
were fixed with 3% paraformaldehyde and 2% sucrose in PBS (15
minuteroomtemperature).Fixation
was required to preventcell
detaching from the culture wells during the following washes. After
or the same concentrathree washes in PBS, MoAb BVI (25 pg/mL)
tion of irrelevant MoAb IgGs in PBS (60 pL/well) were added for
2 hoursatroomtemperature.Afterrinsing
in PBS,peroxidaseconjugated goat-antimouse IgGs (0.15 pg/mL) were applied for 90
minutesatroomtemperature
in 60pL/well of PBS.Afterthree
washes,thechromogenicsubstrateo-phenylene-diaminedihydrochloride was diluted as indicated by the manufacturer and 100 pL
of this solution was added for 30 minutes at room temperature. The
rate of color development at 30 minutes was measured at 450 nm
in an automated micro plate lambda reader (Perkin-Elmer. Brachburg, NJ).
In each experiment, the binding of MoAb BVI to the cells in the
absence of exogenous vn was evaluated. This value, considered as
background was subtracted from each measurement.
Binding ofaggreguted vn to EC. ConfluentECmonolayers in
48-well culture plates were incubated overnight
in serum-free medium containing 0.5% BSA and insulin-transferrin-sodium selenite
supplement as described above. The cells were rinsed twice with
BSA-containing serum-free medium and were further incubated at
37°C with 200 pL of serum-free medium/0.2% BSA containing 330
ng/mL'ZSI-labeledaggregated vn, which wasobtained by boiling
3 minutes before the binding experiments.
radiolabeled plasma vn for
As competitorsGPenGRGDSPCA ( I to 100 pmol/L),GPenRA-
ZANETTI ET AL
RGDNPCA (200 pmol/L) and RPrncGHKGELRCR (200 pmol/L),
heparin (500 pg/mL) and heparan sulfate (8 pmoUL) were included
in the binding solution. After 4 hours incubation time, binding mewashed 3 times with
dium was removedandthecelllayerwas
serum-free medium/0,2% BSA. Cells and matrix were solubilized
by incubating the cell layer with 1 mol/L NaOH for l hour at room
temperature and counted.
Coating beads with vn and ulbumin-coupledpeptides. Approximately 10' magnetic beads (diameter, 4.5 pm) were taken from the
stock undiluted suspension" and added to1 mL of 0.05 molL borate
buffer, pH 9.5 containing ( I ) different concentrations of plasma vn
(0.024 to 320 pg/mL), (2) BSA-linked GRGDSP or GRGESP peptide (corresponding to 9 to 14 ng peptide1246 pg BSNmL), and (3)
BSA alone (246 pg/mL). The mixture was then incubated 16 to 18
hours by slow end-over-end rotation atroomtemperature. At the
end of incubation, unreacted tosyl groups were blocked by reaction
with 1 mL of 0.05 mol/L TRIS-HCI buffer pH 9.6. The buffer was
then removedusingtheDynalmagneticdeviceandbeadswere
washed four times with 1 mL TRIS-HCI buffer and once with sterile
Hank's Balanced Salt Solution (HBSS). The amount ofvn bound
to beads was quantified by including trace amounts of "'l-vn, (1.5
X IO" cpm in 100 p L of the protein solution) in the protein-binding
buffer. After the binding (see above) beads were washed and counted
in a gamma counter.
Binding o
f profein-coated beads to EC. Different concentrations
of vn. GRGDSP, or GRGESPlinkedtoBSAandBSA-coupled
beads in 60 pL/well of medium 199,0.S% BSA, and 0.01% sodium
azide were added to confluent EC. After various timesof incubation
at 3 7 T , the bead solution was aspirated and the wells were washed
twice with PBS containing Ca2" and Mg". Cells were then fixed
with Diff-Quick (100 pL/well) for 3 minutes, stained with crystal
violet (0.5% in 20% methanol solution) for
5 minutes and washed
twice with distilled water. These procedures were all done at room
temperature. The beads adhering to the cell surface were countedby
phase contrast microscopy. At least 10 fields per well were counted
(corresponding to about 300 cells at confluency).
RESULTS
We first measured the binding of soluble plasmavn to EC
monolayers. To more selectively evaluate vn binding to the
apical cell surface an ELISA set-up was used. Binding of vn
to receptors located on thebasal aspect of the cell membrane
would be poorly detected by this assay because preliminary
data showed that, because of their high molecular weight
IgG cannot effectively traverse EC cell-to-cell junctions. As
reported in Fig 2, purified soluble vn bound very little to
confluent EC monolayers. This association was not dependent on vn concentration in the medium (Fig 2) or time of
incubation (in the range
of 5 to 120 minutes, not shown).
Heparin (up to 570 pg/mL), GRGDSP peptides (up t o 800
pmol/L), and anti-avp3 serum (up to 0.02 dilution) added
during thebindingassay
did not affect vn binding (not
shown). For comparison, wetested the binding of multimeric
vn,theheparin
binding form of the adhesive protein. As
reported in Fig 2 , multimeric vn associated to EC in a concentration-dependent way. Time-course experiments showed
that at 30 pg/mL, the binding of multimeric vn was already
apparent at 5 minutes and reached the equilibrium at 1 hour
(not shown). As reported in Fig 2, heparin blocked binding
of multimeric vn to the cells, but similar to plasma vn, no
effect was found by GRGDSP peptides (up to 800 pmo1L)
and anti-avb3 serum (up to 0.02 dilution) (not shown).
From www.bloodjournal.org by guest on November 19, 2014. For personal use only.
1119
BINDING OF PLASMA VN-COATEDBEADS
These results were in agreement with previously published
data24.25
and did not support the hypothesis that integrins on
EC luminal surface can bind soluble vn.
However, modification of vn by heating resulted in aggregated forms of the protein (reaching mw greater than 10')
that could associate to EC monolayers in an RGD-dependent
way. As reported in Fig 3, unlike multimeric vn, heparin
or heparan sulfate were ineffective in competing for the
aggregated vn binding. In contrast, the RGD containing cyclic peptides were able to substantially reduce binding of the
aggregated form of vn to about 20% of control suggesting
major involvement of integrins in this interaction. These
data were also supported by the concentration dependence
of RGD peptides with half maximal inhibition at about 7
pmollL. Under these conditions, cells did not detach from
their substratum. Compared with 37"C, less than 20% total
binding of aggregated vn was observed at 4"C, suggesting
that active metabolism was required for integrin-mediated
recognition (data not shown). Overall, these data suggested
that integrins could be involved in vn binding whenthe
density of the ligand was sufficiently high.
To further investigate this point, plasma vn was linked
to beads andthebead suspension added to confluent EC
monolayers. The beads (diameter, 4.5 pm) were large
enough to be counted directly using phase-contrast microscopy when attached on EC surface (Fig 4).
As reported in Fig 5 , vn-bead binding to EC was dependent on ( l ) time of incubation, reaching a plateau at 60
minutes (A); ( 2 ) the number of beads added (B); and (3) the
concentration of vn present on the bead surface (C). The
minimal vn concentration able to give a detectable binding
was in the range of 0.3 to 0.5 X
ghead.
Irrelevant binding of albumin-coated beads was observed
(less than 6 beads per 10' cells adding 5 X IO5 beads/mL
for up to 120 minutes incubation, Fig 4C).
The mechanism of vn-bead binding to EC was further
established by competition assays.
0.0
0
10
20
30
40
C19/ml
Fig 2. Binding ofsolubleplasmavn
and multimeric vn to EC
monolayers. Increasingconcentrationsof
multimeric vn (U) or
were added to cultured EC for 60 minutes at 37°C.
plasma vn 11.
Heparin (57 pg/mL) ( A ) was added to multimeric vn during the binding assay. Data are means ? SEM of three separate experiments.
a
z
400001
3
32000
A
3
z?
i:
k5
$2
(3
CY
<
24000
16000
8000
0
n
z-J
zo
>U
0
i
10
100
PEPTIDE (PM)
Fig 3. Binding of aggregated vn to EC. Radiolabeled aggregated
vn (330 ng/mL; 200.000 cpm amount of radioactivity per well) was
incubated for 4 hours at 37°C with confluent cell monolayers. (A)
Binding of '%aggregated vn was performed in the absence or presence of 100 pmol/L cyclic peptides as indicated as well as heparin
(500 pg/mL) and heparan sulfate(8 pmollL). (B) Binding of '%aggregated vn was followed in the presence ofdifferent concentrations of
the cyclic RGD peptide (GPenGRGDSPCA)1.1 and control peptide
RlPmc) GHKGELRCR IO). Data (means k SD, n = 3) are representative
of four separate experiments.
As reported in Fig 6, anti-avp3 serum blocked vn-bead
binding toEC, whereas anti-a5/31 serum was ineffective.
Table 1 reports the effect of other agents on vn-bead binding.
Two anti-vn MoAbs able to inhibit EC adhesion to vn (16A7
and BV2) blocked vn-bead binding. Two inhibitory MoAbs,
LM609 directed to avP3 and 7E3 directed to P3,"." inhibited vn-bead binding, whereas the nonblocking MoAbs
LM142, directed to av," P1F6 (blocking MoAb, directed to
av/35),I4 and Lia L/2 (blocking MoAb, directed to pl)" were
inactive. Finally, the cyclic GPenGRGDSPCA (but not
GRGESP) peptide inhibited vn-bead binding to EC (see also
Fig 4), whereas heparin, up to 570 pg/mL was without effect.
To further investigate the presence of recognition sites for
RGD ligands on the apical surface of EC, binding of albumin-linked GRGDSP and GRGESP peptides coated to beads
was studied. GRGDSP, but not GRGESP-coated beads significantly bound to EC in a time-dependent manner reaching
From www.bloodjournal.org by guest on November 19, 2014. For personal use only.
ZANElTI ET AL
1120
3
J
W
V
(U
0
r-
3j
n
4m
20
0
40
60
80
100 120 140
TIME (min)
500-
3
J
W
V
(U
0
r-
300 200 -
4
m
100 -
'3
.i;
J
W
V
(U
0
Y
Fig 4. Light microscopy analysis of plasma vn-coated bead binding t o €C. Phase-contrast microscopy of EC incubated with plasma
vn-coated beads (lo6beadslmLfor 60 minutes at37°C) in the absence
(a) or presence of GPenGRGDSPCA (400pmol/L) Ib) and
EC incubated
with BSA-coated beads (c). Cells were fixed with Diff-Quick and
stained with crystal violet.
400 -
G
n
,. .
B
3j
n
a
W
m
100
0.001
0.01
0.1
1
10
10 - l 3 g/BEAD
a plateau at 2 hours (Fig 7A). The binding was blocked by
anti-avp3 serum, LM609 and 7E3 MoAbs (directed to a v o 3
and 03, respectively), and GRGDSP and GPenGRGDSPCA
peptides, whereas MoAbs PIF6 (directed to cuv/3S) and Lia
'/? (directed to 01) were inactive (Fig 7B).
DISCUSSION
The major finding of this paper is that apically located
EC integrins (and inparticular a v 0 3 ) retaintheir biologic
function andbind surface-linked vnand RGD peptides. In
Fig 5. Binding of plasma vn-coated beads t o EC. (A) Time course
of vn-bead binding t o €C. Beads (3.3 x 105/mL)coated with vnwere
added t o confluent EC monolayers for different times at 37°C. (B)
Binding of vn
beads t o EC in function ofbead concentration. Different
numbers ofvn-coated beads were added t o confluent EC for 60 minutes at 37T.
(C) Binding of vn beads t o EC in function of vn concentration on beads. The binding of vn-coated beads (3.3 x 105/mL, for 60
minutes at37°C) was dependent on theamount of vn linkedt o the
bead surface. Different concentrations of vn (0.024 t o 320 pglmLl
were used in the bead coupling buffert o obtain different concentrations of the protein on the bead surface. The beads bound were
counted directly on the EC surface by phase-contrast microscopy.
Data are means 2 SEM of three separate experiments.
From www.bloodjournal.org by guest on November 19, 2014. For personal use only.
1121
BINDING OF PLASMAVN-COATEDBEADS
-
2 100 l0
80 -
a
60 -
40 20 1
O
.
'O
0
0
0
0.02 0.04 0.06 0.08 0.10
ANTISERUM DILUTIONS
Fig 6. Effect of anti-mp3 and anti-a5pl sera on plasma vn-coated
bead bindingto EC. Vn-beads (3.3x 106/mL)were added to confluent
EC monolayen for 80 minutes at 37°C in the presence of increasing
concentrations of a n t i e p 3 (.l and antia5pl sera (0).Data are
means of three separate experiments. Rerub ara expressed as the
percentage of bead binding in the presence of nonimmune serum.
different types of cell-binding mechanisms depending on
their conformation. For instance, only when it is coated on
a surface and treated by reducing agents, thromb~spondin~~
canbind avP3, presumably by exposure of its otherwise
hindered RGD domain. Thrombin can be bound by avp3
through its RGD domain only when the molecule is modified
by different treatment^.^" In addition, integrins bind more
effectively to their specific ligands when they are allowed
to cluster or bind to a ~ u r f a c e , ~and
" ~ ~this appears to be
correlated with ligand d e n ~ i t y . ~ ~ . ~ ~
Interestingly, whereas heparin was most effective in inhibiting soluble multimeric/denaturated vn binding, it did not
change the interaction of vn-coated beads with EC. This
suggests that if a change in conformation occurs, the heparinbinding domain remains hindered or not available for promoting adhesion. Thus, vn linked to beads presents different
features than multimeric soluble vn.
The biologic relevance of the binding of vn beads to EC
outlined in this work remains to be elucidated. Circulating
cells, including tumor cells and bacteria37can bindvnin
different ways and mediate adhesion of the latter to EC?*
Vn bound to beads can be taken as a model system for vn
association with cell surfaces. Thereby, vn might be recognized by the luminal EC integrins and thus, act as a bridge
between circulating cells and the endothelium. Tumor cell
adhesion to EC was found in fact to be dependent on surfaceexposed avp3.39,40
The observation reported here that RGD-coated beads can
also bind to the luminal surface of EC suggests that besides
vn, other RGD containing proteins (such as fibrinogen, von
Willebrand factor or fibronectin) can associate to the endothelium whenbound to circulating cells. Fibrinogen is of
particular interest because it contains two times three puta-
agreement with previous work?25 we were unable to show
the involvement of EC integrins in binding soluble vn. We
found that multimeric vn could bind to EC, but this association appeared to be entirely dependent on the heparin binding
and not the RGD domain of the molecule. In addition, polyclonal antibodies to avo3 did not modify this type of interaction, indicating that vn binding integrins (see below) were
not implicated in this association.
Surprisingly, we found that high-Mr aggregated forms of
vn bound to EC in an RGD-dependent manner. These data
suggest that clustered forms (greater than lo6) are able to
engage luminally expressed integrins, whereas multimeric
vn (Mr, 0.5 to 1 X lo6) is not. Moreover, in aggregated vn,
Table 1. Effect of Competitors on vn-Bead Bindingto €C
no heparidheparan sulfate Competition in binding to EC
Competitor
Concentration
BeadsllO'
Cells
96 Inhibition
was found, although this clustered ligand possesses heparinNA
103 2 10
NA
None
binding capacity. It remains to be established whether addi8 t 1
92
BV2
0.1 (dil)
tional nonintegrin cell surface binding sites may account for
70
16A7
31 t 3
50 (pg/mL)
scavenging aggregated vn, and whether these "scavengingLM 609
61 -t 7
41
0.05 (dil)
type" binding sites may act in cooperation with integrins to
LM 142
172 2 52
0
0.05 (dil)
recognize (and possibly internalize) clustered vn.
7E3
55 2 7
47
20 (pg/mL)
To further investigate whether clustered vn could be recLia 1/2
103 t 27
0
0.1 (dill
ognized by EC integrins, vn was linked to microbeads. Vn
P1F6
106 t 15
0
0.005 (dil)
GPenGRGDSPCA
73 2 9
30
50 (pmol/L)
immobilized on beads significantly bound to the EC apical
200
45 t 6
57
surface inan RGD, butnot heparin-dependent way. The
55 2 6
45
400
integrin receptor involved in this association is most likely
0
114 t 12
800
GRGESP
(pmol/L)
avP3 because vn-bead binding is blocked by anti-avp3 se7
96 t 9
Heparin
570 (pg/mL)
rum. The other integrin receptors that can bind vn, 0 v P 1 , ~ ~
and avP514.2?,28 are not expressed in detectable amounts in
Vn-coated beads (3.3 x 105/mL)were added to confluent EC monohuman umbilical vein EC.7 In addition, blocking antibodies
directed to P I and avP5 (Fig 6 and Table 1) did not modify
vn-bead binding to EC, thus further indicating that these
integrins do not play a significant role in this interaction.
The reason why avo3 recognizes vn only when it is immobilized on surfaces is intriguing. This phenomenon could be
caused by a change in the ligand conformation and/or to an
increase in its local density. Other integrin ligands present
layers for 60 minutes at 37°C in the presence or absence of inhibitors.
BV2. anti-vn MoAb; 16A7, anti-vn MoAb; LM609 anti-avo3 MoAb;
LM142 anti-uv MoAb: 7E3, anti-83 MoAb; Lia 1/2 anti-81 MoAb; P1F6
anti-avp5 MoAb. Blocking antibodies were used at concentrations
able to produce maximal inhibitory activity in appropriate assays.
Percent inhibition was calculated by comparing binding of beads in
the absence and presence of competitors. Data are means ? SEM of
three experiments.
Abbreviations: NA, not applicable; dil, dilution.
From www.bloodjournal.org by guest on November 19, 2014. For personal use only.
1122
ZANElTI ET AL
250
v)
j 200
W
cu
0
150
100
0
2m
50
n-
0
40
80 120 160
200
240
TIME (min)
280
300
B
T
GRGDSP anti-avp3 LM609
7E3
P1F6
Abs
tive available integrin recognition domains4’ and has been
shown to act as bridging molecule in the adherence of bacteria to EC4’ In addition, fibrinogen mediates platelet binding
to avfl3-transfected Chinese Hamster Ovary cells by bridging aIIb-@3and a ~ f l 3 . Therefore,
4~
fibrinogen bound to activated platelets or to other cell types could be recognized and
bound by endothelial av@3and help to localize circulating
cells to the intact vascular surface.
In addition to EC integrins, ICAM-1 has been found to
act as a fibrinogen receptor that thereby acts as bridging
molecule between leukocytes and EC.44 Interestingly,
ICAM-1 antibodies only partially inhibited fibrinogen-mediated adhesion of leukocytes to resting EC suggesting that
additional adhesive mechanisms contribute to this effect.
REFERENCES
1. Gimbrone MA Jr: Vascular endothelium: Nature’s blood con-
tainer, in Gimbrone MA Jr (ed): Vascular Endothelium in Hemostasis and Thrombosis. Edinburgh, UK, Churchill Livingstone, 1986,
P1
2. Miles LA, Levin EG, Plescia J, Collen D, Plow EF: Plasminogen receptors, urokinase receptors, and their modulation on human
endothelial cells. Blood 72:628, 1988
3. Bevilacqua MP, Nelson RM: Selectins. J Clin Invest 91:379,
I993
Liall2 GPenGRG GRGDSP
DSPCA
PEPTIDES
Fig
7.
Binding
of
GRGDSP
and
GRGESP
linked to BSAcoated beads to €C. (A) Time
course of beads coated with albumin-linked GRGDSP l.) or
GRGESP (0).Beads (3.3 x lo5/
mL) were added to EC at 37°C.
(B) GRGDSP bead (3.3 x 10’lrnL)
binding was also evaluated in
the presence of competitors
added during the binding assay
(120 minutes at 37°C): anti-rYvp3
serum (0.005 dilution), LM609
a n t i d o 3 Mob (0.0016 dilution), 7E3 anti-fl3 MoAb (25 pa/
mL),
P1F6
a n t i - m p MoAb
(0.005 dilution), Lia ‘h anti-81
MoAb (0.1 dilution), GPenGRGDSPCA
cyclic
peptide (400
pmollL), and
GRGDSP
(400
pmollL). Data are means 2 SEM
of three separate experiments.
Nonimmune rabbit serum and
nonimmune ascitic fluid or nonimmune purified IgG were ineffective (not shown).
4. Osborn L: Leukocyte adhesion to endothelium in inflammation.
Cell 62:3, 1990
5. Mantovani A, Bussolino F, Dejana E: Cytokine regulation of
endothelial cell function. FASEB J 6:2591, 1992
6. Hynes RO: Integrins: Versatility, modulation, and signaling in
cell adhesion. Cell 69:11, 1992
7. Conforti G, Dominguez-Jimenez C, Zanetti A, Gimbrone MA,
Cremona 0, Marchisio PC, Dejana E: Human endothelial cells express integrin receptors on the luminal aspect of their membrane.
Blood 80:437, 1992
8. F’reissner KT: Structure and biological role of vitronectin. Ann
Rev Cell Biol 7:275, 1991
9. Tomasini BR, Mosher DF: Conformational states of vitronectin: Preferential expression of an antigenic epitope when vitronectin
is covalently and non-covalently complexed with thrombin-antithrombin 111 or treated with urea. Blood 72:903, 1988
10. Izumi M, Yamada KM, Hayashi M: Vitronectin exists in
two structurally and functionally distinct forms in human plasma.
Biochim Biophys Acta 990:101, 1989
11. Cheresh DA, Smith JW, Cooper HM, Quaranta V: A novel
vitronectin receptor integrin (crvpx) is responsible for distinct adhesive properties of carcinoma cells. Cell 5759, 1989
12. Charo IF, Bekeart S, Phillips D R Platelet glycoprotein IIbIIIa like proteins mediate endothelial cell attachment to adhesive
proteins and the extracellular matrix. J Biol Chem 262:9935, 1987
13. Campanero MR, Arroyo AG, Pulido R, Ursa A, de Matias
MS, Shnchez-Mateos P, Kassner PD, Chan BMC, Hemler ME, Corbi
From www.bloodjournal.org by guest on November 19, 2014. For personal use only.
BINDING OF PLASMA VN-COATED BEADS
AL, de LandLuri MO, Sanchez-Madrid F: Functional role of a2Pl
and a4b1 integrins in leukocyte intercellular adhesion induced
through the common 8 1 subunit. Eur J Immunol22:3111, 1992
14. Wayner EA, Orlando RA, Cheresh DA: Integrins avp3 and
avP5 contribute to cell attachment to vitronectin but differentially
distribute on the cell surface. J Cell Biol 113:919, 1991
15. Engvall E Enzyme immunoassay ELISA and EMIT. Methods
Enzymol 70:419, 1981
16. Conforti G, Zanetti A, Pasquali-Ronchetti I, Quaglino D,
Neyroz P, Dejana E: Modulation of vitronectin receptor binding by
membrane lipid composition. J Biol Chem 2654011, 1990
17. Dejana E, Colella S, Conforti G, Abbadini M, Gaboli M,
Marchisio PC: Fibronectin and vitronectin regulate the organization
of their respective Arg-Gly-Asp adhesion receptors in cultured human endothelial cells. J Cell Biol 107:1215, 1988
18. Preissner KT, Wassmuth R, Muller-Berghaus G: Physicochemical characterization of human S-protein and its function in the
blood coagulation system. Biochem J 231:349, 1985
19. Fraker PJ, Speck JC Jr: Protein and cell membrane iodinations
with a sparingly soluble chloroamide, 1,3,4,6-tetrachloro-3a, 6a-diphenylglycoluril. Biochem Biophys Res Commun 80:849, 1978
20. Conforti G , Zanetti A, Colella S, Abbadini M, Marchisio PC,
Pytela R, Giancotti FG, Tarone G, Languino LR, Dejana E: Interaction of fibronectin with cultured human endothelial cells: Characterization of the specific receptor. Blood 73:1576, 1989
21. Harlow E, Lane D: Coupling peptides to carrier proteins using
glutaraldehyde. Harlow E, Lane D (eds): Antibodies. Cold Spring
Harbor Laboratory, Cold Spring Harbor, NY, 1988, p 78
22. Barbieri B, Balconi G, Dejana E, Donati MB: Evidence that
vascular endothelial cells can induce the retraction of fibrin clots.
Proc SOCExp Biol Med 168:204, 1981
23. Curtis ASG, McGrath M, Gasmi L: Localised application of
an activating signal to a cell: Experimental use of fibronectin bound
to beads and the implications for mechanisms of adhesion. J Cell
Sci 101:427, 1992
24. Volker W, Hess S, Vischer P, Preissner KT: Binding and
processing of multimeric vitronectin by vascular endothelial cells. J
Histochem Cytochem 41:1823, 1993
25. Hess S, Stockmann A, Volker W, Preissner KT: Multimeric
vitronectin: Structure and function, in Preissner KT, Rosenblatt S,
Kost A, Wegerhoff J, Mosher DF (eds): Biology of Vitronectins and
Their Receptors. Amsterdam, The Netherlands, Elsevier, 1993, p 21
26. Vogel BE, Tarone G, Giancotti FG, Gailit J, Ruoslahti E: A
novel fibronectin receptor with an unexpected subunit composition
(avbl). J Biol Chem 2655934, 1990
27. Vogel BE, Lee SJ, Hildebrand A, Craig W, Pierschbacher
MD, Wong-Staal F, Ruoslahti E: A novel integrin specificity exemplified by binding of the avb5 integrin to the basic domain of the
HIV Tat protein and vitronectin. J Cell Biol 121:461, 1993
28. Scalise Panetti T, McKeown-Longo PJ: The avP5 integrin
receptor regulates receptor-mediated endocytosis of vitronectin. J
Biol Chem 268: 11492, 1993
29. Sun X, Skorstengaard K, Mosher D F Disulfides modulate
RGD-inhibitable cell adhesive activity of thrombospondin. J Cell
Biol 118:693, 1992
1123
30. Bar-Shavit R, Sabbah V, Lampugnani MG, Marchisio PC,
Fenton JW 11, Vlodavsky I, Dejana E: An Arg-Gly-Asp sequence
within thrombin promotes endothelial cell adhesion. J Cell Biol
112:335, 1991
31. Grinnell F Fibroblast receptor for cell substratum adhesion:
Studies on the interaction of baby hamster kidney cells with latex
beads coated by cold insoluble globulin (plasma fibronectin). J Cell
Biol 86:104, 1980
32. Savage B, Ruggeri ZM: Selective recognition of adhesive
sites in surface-bound fibrinogen by glycoprotein IIb-IIIa on nonactivated platelets. J Biol Chem 266:11227, 1991
33. Coller BS: Interaction of normal, thromboasthenic, and Bernard-Soulier platelets with immobilized fibrinogen: Defective platelet-fibrinogen interaction in thromboasthenia. Blood 55:169, 1980
34. Massia SP, Hubbel JA: An RGD spacing of 440 nm is sufficient for integrin avb3-mediated fibroblast spreading and 140 nm
for focal contact and stress fiber formation. J Cell Biol 114: 1089,
1991
35. Grinnell F, Geiger B: Interaction of fibronectin-coated beads
with attached and spread fibroblasts. Exp Cell Res 162:449, 1986
36. Singer 11, Kawka DW, Scott S, Mumford RA, Lark MW: The
fibronectin cell attachment sequence Arg-Gly-Asp promotes focal
contact formation during early fibroblast attachment and spreading.
J Cell Biol 104573, 1987
37. Chhatwal GS, Preissner KT, Muller-Berghaus G, Blobel H:
Specific binding of human S protein (vitronectin) to streptococci,
Staphylococcus aureus and Escherichia coli. Infect Immun 55:1878,
1987
38. Valentin Weigand P, Grulich-Henn J, Chhatwal CS, MullerBerghaus G, Blobel H, Preissner K T Complement S protein
(vitronectin) mediates the adherence of streptococci to human endothelial cells. Infect Immun 56:2851, 1988
39. Lafrenie RM, Podor TJ, Buchanan MR, Orr F W : Up-regulated
biosynthesis and expression of endothelial cell vitronectin receptor
enhances cancer cell adhesion. Cancer Res 52:2202, 1992
40. Tang DG, Grossi IM, Chen YQ, Diglio CA, Honn KV: 12(S)HETE promotes tumor-cell adhesion by increasing surface expression of ( ~ v b integrins
3
on endothelial cells. Int J Cancer 54:102,
1993
41. Doolittle RF, Watt KWK, Cottrell BA, Strong DD, Riley M:
The amino acid sequence of the a-chain of human fibrinogen. Nature
280:464, 1979
42. Cheung AL, Krishnan M, Jaffe EA, Fischetti VA: Fibrinogen
acts as a bridging molecule in the adherence of Staphylococcus
aureus to cultured human endothelial cells. J Clin Invest 87:2236,
1991
43. Gawaz MP, Loftus JC, Bajt ML, Frojmovic MM, Plow EF,
Ginsberg MH: Ligand bridging mediates integrin aIIbb3 (platelet
GPIIB-IIIA) dependent homotypic and heterotypic cell-cell interactions. J Clin Invest 88:1128, 1991
44. Languino LR, Plescia J, Duperray A, Brian AA,Plow EF,
Geltosky JE, Altieri DC: Fibrinogen mediates leukocyte adhesion to
vascular endothelium through an ICAM-I -dependent pathway. Cell
73:l. 1993