Protease Nexin-1 Interacts With Thrombomodulin and
Modulates Its Anticoagulant Effect
Marie-Christine Bouton, Laurence Venisse, Benjamin Richard, Cécile Pouzet,
Véronique Arocas, Martine Jandrot-Perrus
Downloaded from http://circres.ahajournals.org/ by guest on July 12, 2017
Abstract—The endothelial cell membrane glycoprotein thrombomodulin (TM) plays a critical role in the regulation of
coagulation. TM is an essential cofactor in protein C activation by thrombin, and a direct inhibitor of thrombin-induced
platelet activation and fibrinogen clotting. Protease nexin-1 (PN-1) is a serpin synthesized and secreted by a variety of
cells including endothelial cells. PN-1 bound to the cell surface through interactions with glycosaminoglycans, is an
efficient inhibitor of thrombin and controls thrombin-induced cell responses. An investigation of the interaction of PN-1
with TM using purified proteins and cultured human aortic endothelial cells was performed. Purified PN-1 was observed
to bind to purified TM in a concentration-dependent manner. Double immunofluorescence studies indicated that PN-1
and TM were colocalized at the endothelial cell surface from which they were coprecipitated. Pretreatment of the cells
with chondroitinase ABC greatly decreased the amount of the PN-1 associated to TM at the cell surface demonstrating
the involvement of the TM chondroitin-sulfate chain in the formation of complexes. The inhibitory activity of the
PN-1/TM complexes on the catalytic activity of thrombin, and on thrombin-induced fibrinogen clotting, was markedly
enhanced when compared with the inhibitory activity of each partner. PN-1– overexpressing human aortic endothelial
cells and PN-1– underexpressing human aortic endothelial cells exhibited respectively a significantly reduced ability and
enhanced capacity to activate protein C. Furthermore, PN-1 decreased the cofactor activity of TM on thrombin activable
fibrinolysis inhibitor activation by thrombin. These data show for the first time that PN-1 forms complexes with TM and
modulates its anticoagulant activity. (Circ Res. 2007;100:1174-1181.)
Key Words: protease nexin-1 䡲 thrombin 䡲 thrombomodulin 䡲 serpins 䡲 endothelial cells
T
hrombin is the only protease in the coagulation cascade
that possesses both coagulant and anticoagulant activities. The anticoagulant action of thrombin is dependent on
thrombomodulin (TM), a transmembrane glycoprotein predominantly synthesized by vascular endothelial cells. The
extracellular domain of TM consists of an N-terminal lectinlike domain followed by 6 EGF-like domains and a serine/
threonine-rich region which contains potential sites for
O-linked glycosylation supporting the attachment of a
chondroitin-sulfate (CS) moiety. Thrombin/TM complex formation prevents thrombin-induced fibrinogen clotting, activation of Factor V and of platelets.1 In contrast, TM-bound
thrombin activates protein C (PC) to generate the anticoagulant active protein C (APC).2 The TM/thrombin complex can
also activate the latent inhibitor of fibrinolysis, Thrombin
Activable Fibrinolysis Inhibitor (TAFI).3 Furthermore, TM
exhibiting an attached CS moiety binds thrombin ⬇10 times
tighter and accelerates inactivation of bound thrombin by
antithrombin4,5 and PC inhibitor.6 The wide distribution of
TM within the vascular system,7 and the marked changes in
the activity pattern of thrombin on complex formation with
TM, explain the major role of this protein, in the physiological anticoagulant mechanism of control in the hemostatic
system.
The catalytic activity of thrombin can be inhibited by a
variety of serine protease inhibitors (serpins), including
antithrombin (AT), heparin cofactor II (HCII), the plasminogen activator inhibitor-1 (PAI-1) and protease-nexin-1 (PN1). PN-1, a 43 to 50 kDa glycoprotein, is a potent inhibitor of
thrombin and also inhibits other serine proteases such as,
u-PA (urokinase-plasminogen activator), t-PA (tissue-typeplasminogen activator) and plasmin. However, in the presence of glycosaminoglycans (GAGs), such as heparin, thrombin becomes the preferential target of PN-1.8 In the presence
or absence of heparin, PN-1 is a more potent thrombin
inhibitor than AT.9 In contrast to AT and HC II, PN-1 is
barely detectable in plasma.10 PN-1 is synthesized and secreted by a variety of cells including vascular smooth muscle
cells,11 endothelial cells,12 human foreskin fibroblasts,13 human skeletal muscle myotubes14 and glial cells or neurons.15
Original received September 14, 2006; resubmission received February 21, 2007; revised resubmission received March 14, 2007; accepted March 14,
2007.
From the INSERM, U698, CHU Xavier Bichat (M.-C.B., L.V., B.R., V.A., M.J.-P.), Paris, France; IFR 02, Faculté Médecine Xavier Bichat (C.P.),
Paris, France.
Correspondence to Dr Marie-Christine Bouton, Unité INSERM U698, CHU Xavier Bichat, 46 rue Henri Huchard 75877 Paris Cedex 18, France. E-mail
[email protected]
© 2007 American Heart Association, Inc.
Circulation Research is available at http://circres.ahajournals.org
DOI: 10.1161/01.RES.0000265066.92923.ee
1174
Bouton et al
Thrombomodulin and PN-1 Are Complexed on HAECs
1175
Downloaded from http://circres.ahajournals.org/ by guest on July 12, 2017
At the cell surface, PN-1 forms SDS-stable equimolecular
complexes with target proteases.13 Once formed, these complexes are rapidly internalized and degraded as has been
reported in human foreskin fibroblasts.16,17 PN-1, therefore,
has been suggested to be an important specific regulator of
protease activities in the pericellular environment.
The aim of this study was to determine whether PN-1 could
interact with TM, thereby improving the efficiency of thrombin inactivation by endothelial cells. For this purpose, an
investigation of the capacity of TM and PN-1 to interact
under purified conditions as well as on endothelial cells was
performed. Furthermore the effects of TM/PN-1 complexes
on the regulation of thrombin activity were analyzed by using
wild-type endothelial cells or endothelial cells that both
overexpress or underexpress PN-1.
the presence or the absence of thrombomodulin, under pseudo-first
order conditions. The dependence of the kon value for the inhibition
of thrombin on the concentration of thrombomodulin was determined by incubating PN-1 (5 nmol/L) with increasing concentrations
of thrombomodulin (0 to 5 nmol/L in 20 mmol/L phosphate,
100 mmol/L NaCl, 0.1 mmol/L EDTA, 0.1% PEG 8000, pH 7.5) for
5 minutes at 37°C before the addition of the chromogenic substrate
S-2238 (0.3 mmol/L). The reactions were started by the addition of
thrombin (0.5 nmol/L). Thrombin activity was determined by measuring the rate of substrate hydrolysis at 405 nm using a microtiter
plate reader and the Biolyse 2 application (Labsystem, Courtaboeuf,
France). Values for kon were calculated as described previously.17,18
The uncatalyzed second order rate constant, ie, in the absence of
thrombomodulin, was determined using the above method but with
PN-1 (150 nmol/L) present in the medium.
Materials and Methods
Thrombin (0.5 nmol/L in TBS-20 mmol/L Tris, 150 mmol/L NaCl
pH 7.5- containing 2.5 mmol/L CaCl2, and 0.1% PEG 8000) was
incubated with confluent cells which have been or not pre-incubated
with PN-1 (5 or 10 nmol/L) and rinsed before use. In other
experiments, thrombin was incubated with chondroitinase-treated
confluent HAEC monolayers. After 10 minutes incubation of thrombin on confluent cells, aliquots were removed and transferred into a
microtiter plates containing S-2238 (0.3 mmol/L). Thrombin activity
was measured as above and residual thrombin activity was
calculated.
Cell culture, immunofluorescence, cell transfections, enzymatic
treatment of cells, reverse transcription and quantitative real-time
polymerase chain reaction, TAFI activation, statistical analysis,
antibodies and reagents are described in the online data supplement,
available at http://circres.ahajournals.org.
Binding of PN-1 to Thrombomodulin
Binding of PN-1 to rabbit TM was analyzed in 96-multiwell plates
(Immulon II, Dynatech, Chantilly, Va): Rabbit TM (0.5 ␮g/well in
50 mmol/L bicarbonate buffer, pH 9.6) was allowed to adsorb for 18
hour at 4°C. After saturation with 1% BSA in phosphate-buffered
saline (PBS), pH 7.5, recombinant PN-1 (1.25 to 100 ␮g/mL in PBS,
0.1% Tween-20, 0.1% BSA) was incubated with immobilized TM
for 90 minutes at room temperature, in the absence or presence of
various competitors: heparin, fucoidans, polybrene, heparin- or
chondroitan-sulfates. After washing, bound PN-1 was detected using
a polyclonal rabbit anti-PN-1 antibody (8 ␮g /mL) followed by a
peroxidase-coupled secondary anti-rabbit IgG and OPD. The absorbance at 492 nm was monitored in a microtiter plate reader (ieMS,
Labsystem, Courtaboeuf, France). The nonspecific binding of PN-1
to BSA-coated wells was subtracted from the data. Association
curves were constructed, and the apparent Kd was estimated using
GraphPad Prism for one site binding regression analysis.
Immunoprecipitation and Immunoblot Analysis
Human aortic endothelial cells (HAECs) were grown to confluence.
Control cells and chondroitinase-treated cells were washed twice in
ice-cold PBS and solubilized in 500 ␮L of ice-cold lysis buffer
(10 mmol/L Tris/HCl, pH 8, 150 mmol/L NaCl, 2 mmol/L EDTA,
2 mmol/L sodium orthovanadate, 0.5% Nonidet P-40), containing a
complete cocktail of protease inhibitors. Protein concentration was
determined and immunoprecipitations were realized using the same
amount of protein for each sample. After preclearing the samples
with protein A-Sepharose for 30 minutes at 4°C and centrifugation,
lysates were incubated with the anti-human PN-1 (25 ␮g/mL) or
anti-human TM monoclonal antibodies (25 ␮g/mL) overnight at 4°C
followed by the addition of protein A/G-coated magnetic beads and
incubation was continued for 2 hour at 4°C. Magnetic separation
process was allowed and immunoprecipitated proteins were solubilized with 2% SDS, separated by SDS-PAGE and transferred to
nitrocellulose membranes. Membranes were blocked with 5% nonfat
dry milk, probed with the anti- PN-1 and anti-human TM monoclonal antibodies followed by horseradish peroxydase-conjugated secondary antibodies. Immunoreactivity was visualized by
chemiluminescence.
Thrombin Inhibition by PN-1 in the Presence of
Thrombomodulin in a Fluid Phase Assay
Progress curve kinetics were used to estimate the value of second
order rate constant kon for the interaction of thrombin with PN-1, in
Thrombin Inhibition at the Surface of
Endothelial Cells
Fibrinogen Clotting
Fibrinogen (2.5 mg/mL in 10 mmol/L Imidazole, 150 mmol/L NaCl,
10 mmol/L CaCl2, 0.1% PEG 8000, pH 7.5) was mixed with
different concentrations of thrombomodulin and PN-1. After 5
minutes at 37°C, clotting was initiated by the addition of 1 nmol/L
thrombin. The time to clot formation was measured using a KC 10
automatic coagulometer. A standard curve (␣-thrombin 0.125
nmol/L to 2 nmol/L) was used to calculate the percentage of residual
thrombin activity.
Protein C Activation in a Fluid Phase Assay
Thrombin (0.5 nmol/L) was incubated for 10 minutes at 37°C with
rabbit thrombomodulin (0.5 nmol/L) in TBS containing 10 mmol/L
CaCl2 and 0.1% PEG 8000, in absence or presence of PN-1 (0.25 to
2 nmol/L). Bovine protein C (80 nmol/L) was added and the
incubation continued at 37°C. At specified timed, aliquots were
removed; thrombin was inactivated by 100 U/mL hirudin, and
activated protein C (APC) was measured using 0.2 mmol/L S-2366
in TBS, pH 7.5.
Protein C Activation at the Endothelial
Cell Surface
Thrombin (0.5 nmol/L in TBS containing 2.5 mmol/L CaCl2 and
0.1% human serum albumin) was incubated for 10 minutes with
confluent cells which have been or not preincubated with PN-1 (5 or
10 nmol/L) and rinsed before further use. In other experiments,
thrombin was incubated with PN-1- over or underexpressing cells or
with control cells. Protein C (80 nmol/L) was, then added, and the
incubation at 37°C was continued for 90 minutes. Activated protein
C was quantified as described above.
Results
PN-1 Binds to TM
A concentration-dependent and saturable binding of PN-1 to
immobilized rabbit TM was observed with an apparent Kd of
105.1⫾4.9 nmol/L (mean⫾SD) (Figure 1A). Heparin (20 ␮g
/mL) inhibited PN-1 binding to TM by 83.4%⫾7.3% (Figure
1B). High-molecular weight fucoidans (20 ␮g/mL) are sul-
1176
Circulation Research
April 27, 2007
A
A
OD 492nm
0.4
TM
PN-1
TM
PN-1
Merge
0.3
0.2
0.1
B
0.0
0
25
50
75
Merge
100
PN-1 (µg/ml)
Downloaded from http://circres.ahajournals.org/ by guest on July 12, 2017
PN-1 binding (% of control)
B
100
80
60
*
40
*
*
*
*
20
0
control
heparin
fucoidan polybrene heparan chondroitin
sulfate
sulfate
Figure 1. PN-1 binds to TM. A, Increasing concentrations of
PN-1 were added to TM-coated wells (0.5 ␮g/well) for 90 minutes. Bound PN-1 was detected using a polyclonal anti–PN-1
antibody followed by a peroxydase coupled anti-rabbit IgG and
subsequent colorimetric reaction. Data are presented as
means⫾SD of 2 independent experiments each performed in
triplicate. B, PN-1 (10 ␮g/mL) was mixed with heparin (20
␮g/mL) or fucoidan (20 ␮g/mL), or polybrene (10 ␮g/mL),
heparin- or chondroitin-sulfate (10 ␮g/mL). Bound PN-1 was
detected as described above. Data are presented as
means⫾SD of 3 independent experiments each performed in
triplicate. *P⬍0.05 is significantly different from control. When
using 25 ␮g/mL PN-1, heparin- or chondroitin-sulfate (10
␮g/mL) still blocked PN-1–TM interaction by ⬇ 60%.
fated polysaccharides with polyanionic characteristics comparable to those of heparin and also inhibited PN-1 binding to
TM by 71.6%⫾17.7% (Figure 1B). In addition, polybrene
(10 ␮g/mL), a polycation that neutralizes the glycosaminoglycan-dependent interactions, blocked PN-1-TM interaction
by 71.8%⫾3.2% (Figure 1B). Finally, chondroitin-sulfate
(CS) (10 ␮g/mL) or heparan-sulfate (10 ␮g/mL) also blocked
the binding of PN-1 to TM (Figure 1B). PAI-1 is another
serpin synthesized by endothelial cells that shares with PN-1
the capacity to bind to vitronectin19,20 and to inhibit thrombin
but with a lower efficiency. PAI-1 (100 ␮g/mL) had no effect
on PN-1 binding to TM, the percentage of PN-1 binding to
TM being 111.7%⫾6.8%.
PN-1 and Thrombomodulin Are Complexed at the
Surface of Endothelial Cells
The localization of PN-1 and TM at the surface of endothelial
cells was analyzed by immunofluorescence using anti-PN-1
and anti-TM antibodies and confocal microscopy analysis.
Double immunofluorescence studies revealed that TM colocalized with PN-1 on the surface of HAECs (Figure 2A).
Figure 2. TM and PN-1 are colocalized on HAECs. A, Control
HAECs or (B) HAECs treated with chondroitinase ABC (0.2
U/mL) for 45 minutes at 37°C were processed for double indirect immunofluorescence with the anti-PN-1 monoclonal antibody (red) and the anti-TM polyclonal antibody (green). Fluorescence labeling images were obtained by confocal laserscanning microscopy. No staining was detected when the
labeling was performed without primary antibody or with an
irrelevant primary antibody (data not shown). The colocalization
of the 2 antibodies was analyzed using the colocalization Zeiss
LSM 510 3.2 Image Browser and shown in the Merge (white).
Images are representative of 3 experiments. (Bar 10 ␮m).
When cells were pre-treated with chondroitinase ABC, we
observed a robust decrease in the labeling with the anti-PN-1
antibody whereas the labeling with the anti-TM antibody
remained unchanged (Figure 2B) leading to a disruption of
the protein colocalization. This indicates that the CS chain on
TM is important for its colocalization with PN-1.
To determine whether PN-1 and TM colocalization was
caused by protein association, immuno-precipitation experiments were performed using either the anti-PN-1 or anti-TM
monoclonal antibodies. A band at ⬇50 kDa, corresponding to
PN-1 was detected in the samples precipitated by the anti-TM
antibody (Figure 3A) and reciprocally a spread signal at ⬇70
kDa corresponding to TM was present in the sample precipitated by the anti-PN-1 antibody (Figure 3C) confirming that
the two proteins were complexed. Pretreatment of HAEC
with chondroitinase ABC resulted in a decreased intensity of
the PN-1 band in the cell lysates (Figure 3B) and in the TM
immuno-precipitate (Figure 3A). After the treatment with
chondroitinase, TM was still present in cells but detected as a
narrow band, because of the loss of the highest molecular
weight species of TM after deglycosylation (Figure 3D).
Inhibition of Thrombin Activity by the
TM-PN-1 Complexes
Because CS appeared to be involved in PN-1 binding to TM
and are known to accelerate thrombin inhibition by PN-1,21
the effect of TM on the inactivation of thrombin by PN-1 was
investigated in a fluid phase assay. The uncatalyzed rate
constant for thrombin inhibition by PN-1 was
6.6⫻105⫾0.2⫻105 M-1s-1, and is in agreement with previous
IP : TM
B
Cell lysate
IB:
PN-1
CTL
C
ChABC
IP : PN-1
CTL
D
ChABC
Cell lysate
IB:
TM
CTL
ChABC
CTL
ChABC
Figure 3. TM and PN-1 are complexed in HAECs. Control
HAECs (CTL) or chondroitinase ABC- treated HAEC (ChABC)
lysates were submitted to immunoprecipitation using an anti-TM
(A) or an anti-PN-1 (C) antibody. HAEC lysates were analyzed
by immunoblotting using the anti-PN-1 monoclonal antibody (A
and B) or the anti-TM antibody (C and D) followed by a revelation with a peroxydase-coupled anti-mouse antibody and
chemiluminescence. Results are from 1 representative experiment of 3.
Downloaded from http://circres.ahajournals.org/ by guest on July 12, 2017
studies.18,21 When TM and PN-1 were added together, the rate
of S2238 hydrolysis by thrombin decreased as a function of
TM concentration (Figure 4A). At an equimolecular ratio
with PN-1, TM (5 nmol/L) accelerated approximately 20-fold
thrombin inhibition by PN-1. Polybrene (50 ␮g/mL) completely abolished the effect of TM on thrombin inhibition by
PN-1 (data not shown) indicating that the TM-dependent
acceleration of thrombin inhibition by PN-1 involves primarily glycosaminoglycan-protein interactions.
Inhibition of Fibrinoformation by TMPN-1 Complexes
Experimental conditions were chosen in such a way as rabbit
TM alone just slightly inhibited thrombin-induced fibrinogen
clotting [clotting time increasing from 44 sec to 120 s
respectively in the absence or presence of TM (6 nmol/L)]
(Figure 4B), and as indicated by the intersection points of the
different curves with the Y axis on the Figure 4B, PN-1 alone
(0 to 4 nmol/L) did not prolonged the clotting time. In
contrast, a progressive increase in the thrombin clotting time
followed by a sharp upward tendency was observed in the
presence of increasing amounts of the TM/PN-1 mixture. At
a fixed concentration of TM (6 nmol/L), the fibrinogen
clotting time increased from 120 sec in the absence of PN-1
to 329 sec in the presence of 1 nmol/L PN-1. In the presence
of 2 nmol/L PN-1, the clotting time was further increased to
908 sec. This corresponded to a decrease in the residual
activity of thrombin from 72.5%⫾0.7% in the absence of
PN-1 to 24.0%⫾0.8% and 16.5%⫾2.9% in the presence of
PN-1, 1 nmol/L and 2 nmol/L, respectively.
Inhibition of Thrombin Catalytic Activity at the
Surface of Endothelial Cells
Thrombin was incubated for 10 minutes with cells which
have been pretreated with PN-1. The part of unbound active
thrombin was then measured (Figure 5A). Incubation of
thrombin with control HAEC monolayers resulted in a
decrease in thrombin catalytic activity, the residual activity
being of 86.5%⫾1.5% (Figure 5A). When thrombin was
incubated with HAECs pretreated with 5 nmol/L or 10
A
1177
1.00
S-22338 hydrolysis
(OD 405nm)
A
Thrombomodulin and PN-1 Are Complexed on HAECs
0.75
0.50
0.25
0.00
0
25
50
75
100
time (min)
B
Thrombin clotting time (sec)
Bouton et al
1000
750
500
250
0
0
1
2
3
4
5
6
7
TM (nM)
Figure 4. TM accelerates the inhibition of thrombin activity by
PN-1 in a fluid phase assay. A, S-2238 hydrolysis was measured as described in the “Methods” section. In a fluid phase
assay, thrombin (0.5 nmol/L) was incubated with PN-1 (5 nmol/
L), in the absence (Œ) or in the presence of TM [() 1 nmol/L,
(⽧) 2 nmol/L, (●) 5 nmol/L]. Data are presented as means⫾SD
of 3 independent experiments each performed in duplicate.
Standard deviations were small where deviation bars are not
seen. B, Thrombin-catalyzed fibrin formation was measured as
described in the Materials and Methods section. Fibrinogen (2.5
mg/mL) was mixed with increasing concentrations of TM, in the
absence (f), or presence of PN-1 [(Œ) 1 nmol/L, () 2 nmol/L,
(⽧) 4 nmol/L] for 5 minutes at 37°C. Thrombin (1 nmol/L) was
then added and the clotting time was determined. Results are
from 1 representative experiment of 3.
nmol/L PN-1, the thrombin residual activity decreased respectively to 78.8%⫾2.8% and 73.6%⫾0.4% (Figure 5A),
indicating that the additional PN-1 bound to HAEC surface
increased the inhibitory capacity of the cells. Interestingly,
when thrombin was incubated with chondroitinase ABCpretreated HAECs, the catalytic activity of thrombin was not
blocked anymore, the residual activity being of 97.8%⫾4.7%
(Figure 5B), indicating that the chondroitinase treatment
abolished the inhibitory effect of cells on thrombin.
Inhibition of Protein C Activation by TMPN-1 Complexes
As TM is a critical cofactor for thrombin-mediated activation
of PC, an investigation of the effect of PN-1 on APC
production was performed in a fluid phase assay (Figure 6A).
Circulation Research
A
% of r es id u a l a ct iv it y
100
75
90
*
80
*
70
60
50
0
50
25
0
0
0
5
B
25
50
75
100
125
activation time (min)
10
PN-1 (nM)
B
Downloaded from http://circres.ahajournals.org/ by guest on July 12, 2017
*
100
% of r es id u a l a c tiv it y
generated APC
( p m o l e /m i n )
A
April 27, 2007
90
80
70
100
Rate of PC activation
(% of control)
1178
90
80
*
70
60
50
0
0
60
50
0
5
10
PN-1 (nM)
HAEC CTL
HAEC ChABC
Figure 5. Inhibition of thrombin catalytic activity at the surface
of HAEC. A, Thrombin (0.5 nmol/L) was incubated with confluent HAEC which have been preincubated or not with PN-1 (5 or
10 nmol/L). B, Thrombin (0.5 nmol/L) was incubated with control
HAEC (HAEC CTL) or with chondroitinase-treated HAEC (HAEC
ChABC). Residual thrombin catalytic activity was measured as
described in the Materials and Methods section by measuring
the rate of S2238-hydrolysis. Results are presented as residual
thrombin activity [(thrombin activity after 10 minutes incubation
on HAEC /thrombin activity at time 0) x 100] and are the
means⫾SD of 3 determinations. *P⬍0.05 is significantly different
from data obtained with control HAEC.
PC was efficiently activated by an equimolar (0.5 nmol/L)
mixture of thrombin and TM, whereas in the absence of TM,
thrombin failed to generate APC. PN-1 decreased the rate of
protein C activation by TM/thrombin. The rate of PC activation was reduced by 2-fold in the presence of PN-1 (0.5
nmol/L), APC generation being of 6.1⫾0.2 pmole/min in the
presence of PN-1 versus 13.1⫾1.2 pmole/min in the absence
of PN-1.
Inhibition of Protein C Activation at the Surface
of Endothelial Cells
PC activation was also performed on HAEC. In such conditions, HAECs were the source of thrombomodulin. The rate
of protein C activation was reduced by 34.7%⫾4.7% on
HAECs which have been preincubated with 10 nmol/L PN-1
(Figure 6B). To better address the functional contribution of
Figure 6. PN-1 inhibits protein C activation by thrombinthrombomodulin. A, Thrombin/TM complexes were preformed
by mixing the two proteins in equimolecular concentration (0.5
nmol/L). Protein C was activated by the thrombin/TM complex
for 15, 30, 60, 90 and 120 minutes, in the absence (f), or presence of PN -1 [(Œ) 0.25 nmol/L, () 0.5 nmol/L (⽧) 1 nmol/L, (●)
2 nmol/L]. Activated protein C was measured using 0.2 mmol/L
S-2366 in TBS, pH 7.5. Results are the means⫾SD of 3 independent experiments. Standard deviations were small where
deviation bars are not seen. B, Thrombin (0.5 nmol/L) was incubated with confluent cells which have been preincubated or not
with PN-1 (5 or 10 nmol/L). Protein C was added and its activation was then measured as described in the Materials and
Methods section. *P⬍0.05 significantly different from the
respective control cells in the absence of PN-1.
PN-1/TM interaction on the cell surface, PN-1 siRNA was
used to transiently knock-down PN-1 expression. In another
set of experiments, a vector containing the complete PN-1
coding sequence was used to overexpress PN-1 (Figure 7A).
We observed that 48 hours after transfection, PN-1underexpressing HAECs exhibited a significantly enhanced
capacity to activate protein C (30%⫾18% increase) and
reciprocally, PN-1-overexpressing HAECs exhibited a significantly reduced ability to activate protein C (24%⫾4%
decrease) (Figure 7B).
Inhibition of Thrombin Activatable Fibrinolysis
Inhibitor Activation
In addition to protein C activation, the thrombin/TM complex
mediates TAFI activation, a procarboxypeptidase U that
Bouton et al
Thrombomodulin and PN-1 Are Complexed on HAECs
A
0
PN-1 mRNA expression
( % o f co n t r o l)
200
10
60
1179
60 min
kDa
150
60
50
TAFI
40
100
30
PN-1
TAFIa
25
20
50
15
10
0
Irrelevant
siRNA
B
PN-1
siRNA
mock
PN-1
overexpression
Downloaded from http://circres.ahajournals.org/ by guest on July 12, 2017
R at e o f P C a c t i v a t i o n
(% of control)
*
140
*
120
100
- PN-1
+ PN-1
Figure 8. PN-1 inhibits TAFI activation by thrombinthrombomodulin. TAFI (450 nmol/L) was activated by an
equimolecular (10 nmol/L) complex of thrombin/TM for 0, 10 or
60 minutes at 37°C in the absence (⫺PN-1) or presence of 10
nmol/L PN-1 (⫹PN-1). The reaction was stopped by the addition
of 2% SDS and the samples analyzed by SDS-PAGE followed
by silver staining. The migration of TAFI (56 kDa), activated
TAFIa (36 kDa) and products of degradation of TAFIa (25 kDa
and 11 kDa) is indicated.
80
60
50
0
Irrelevant
siRNA
PN-1
siRNA
mock
PN-1
overexpression
Figure 7. Reducing or increasing the expression of PN-1 on
HAECs modulates protein C activation by thrombinthrombomodulin. As described in the data supplement, HAECs
were transiently transfected with an irrelevant siRNA, siRNA targeting PN-1 to knockdown PN-1 (PN-1 siRNA), or a vector containing the complete PN-1 coding sequence to over-express
PN-1 (PN-1 surexpression), or an empty vector (mock). Control
cells were treated with the transfection reagents alone. A, PN-1
and GAPDH were quantified by RT-PCR as described in the
data supplement section. Data are expressed relatively to control values obtained from control cells (100%). B, Thrombin (0.5
nmol/L) was incubated with cells transfected with an irrelevant
siRNA or the PN-1 siRNA or with an empty vector (mock), or a
vector containing the PN-1 cDNA (PN-1-overexpressing cells).
48 hour after transfection, protein C was added and its activation was measured as described in the Materials and Methods
section. Results are the means⫾SD of 3 to 5 determinations.
*P⬍0.05 significantly different from the respective controls.
protects the fibrin clot against lysis.22 Thrombin alone is a
relatively poor activator of TAFI.3 In the present study, TAFI
activation was performed by an equimolar (10 nmol/L)
mixture of thrombin and TM, in the absence or the presence
of 10 nmol/L PN-1. On 10 minutes activation of TAFI, a
36-kDa band corresponding to activated TAFI was generated
(Figure 8). Subsequently, after 60 minutes, the typical degradation products of 25- and 11 kDa were formed (Figure 8).
TAFI activation was completely blocked in the presence of
PN-1 even after 60 minutes activation with thrombin and TM.
Discussion
The results of the present study demonstrate that PN-1 binds
to TM and that the CS moiety of TM is critical for the
interaction. These proposals are supported by the following
evidences: First, PN-1 bound to immobilized TM. Second,
PN-1 binding to thrombomodulin was inhibited by various
polysaccharides such as CS and by the polycation, polybrene.
Third, thrombin inhibition by PN-1 was accelerated in the
presence of TM and polybrene reversed the effect of TM.
Two other serpins, AT and the protein C inhibitor, have
already been shown to bind to different sites on TM. TM has
been shown to enhance the rate of thrombin inactivation by
AT5,23,24 and CS is important for AT binding to TM, as
deglycosylated TM lacks the AT-dependent anticoagulant
activity but retains its PC activation cofactor activity.25 Thus,
TM, thrombin and PN-1 or AT appear to associate as
trimolecular complexes in which the catalytic site of thrombin remains accessible to the inhibition by serpins.23,26 Nevertheless, using our solid phase assay we neither detected
direct binding of AT to TM nor observed any inhibition of
PN-1 binding to TM in the presence of an excess of AT
(10-fold) over PN-1 (data not shown). Furthermore, TM
accelerated thrombin inhibition by PN-1 by about ⬇20-fold
whereas it has been reported to accelerate thrombin inhibition
by AT by only 4 to 8-fold.23 Together these data suggest that
PN-1 binds to TM with a higher efficacy than AT does. PN-1
has been actually reported to have a higher affinity for CS
than AT.27
In contrast, PAI-1, another serpin with poor antithrombin
activity does not bind to the thrombin/TM complex.28 Similarly, PAI-1 had no effect on PN-1 interaction with TM.
The interaction of PN-1 with TM appears to be physiologically relevant since PN-1/TM complexes were detected on
HAECs. PN-1 and TM were efficiently coprecipitated from
lysates of HAECs and their colocalization was observed by
confocal laser-scanning microscopy. Secreted PN-1 is known
to bind at the cell surface through interactions with polysaccharides and is released on cell treatment with heparin.29 The
present report identified TM as a major PN-1-binding proteoglycan. The role of the CS moiety of TM in PN-1 binding
on endothelial cells is demonstrated by the treatment with
chondroitinase ABC, which resulted in the disruption of
protein colocalization as well as in the reduction of their
1180
Circulation Research
April 27, 2007
Downloaded from http://circres.ahajournals.org/ by guest on July 12, 2017
coprecipitation. In contrast, no difference in the colocalization or coprecipitation was observed after heparinase treatment (data not shown). Therefore, the efficacy of PN-1
binding to TM may be related to the TM content in CS which
is variable depending on the vascular origin.30 Interestingly,
in the present report, HAECs have been shown to limit the
catalytic activity of thrombin by both a CS- and PN-1dependent mechanism. Thus, TM and PN-1 variations are
likely to determine the anti-thrombin activity of the endothelium. Therefore, arterial cells which express a high proportion
of CS–associated TM are assumed to carry out a better
protection against thrombin than venous cells which express
a low proportion of CS-associated TM. The fact that we did
not succeed in coprecipitating TM and PN-1 in lysates of
HUVEC (data not shown) is in favor of this hypothesis.
Therefore the amount of TM/PN-1 complexes at the endothelial cell surface may vary according to the vascular
territory. Whether or not PN-1 expression is variable in
different vascular beds remains to be established.
When TM and PN-1 at concentrations too low to prolong
the thrombin clotting time were mixed, a synergistic effect on
clot formation was observed as indicated by a sharp prolongation of the clotting time. This striking anticoagulant activity
of the TM/PN-1 complex is explained by the observation that
TM prevents fibrinogen binding to the thrombin exosite 1,31
and enhances the inhibition by PN-1 of fibrinogen proteolysis
by thrombin. The TM/PN-1 complex thus inhibits clotting in
a fashion similar to hirudin, with the simultaneous blockade
of the thrombin exosite 1 and catalytic site.
On another hand, TM increases the rate of PC activation by
thrombin by a factor ⬇1 000.1 A recent study demonstrated
that the affinity of PC for the thrombin/TM complex is
determined in a primary way by active site dependent
interactions.32 In the present report, PN-1 was shown to
reduce the rate of PC activation by thrombin in the presence
of thrombomodulin, not only in a fluid phase assay but also at
the surface of endothelial cells. Indeed, transient knock-down
or overexpression of PN-1 respectively increased or reduced
PC activation. Membrane bound–PN-1 has thus a direct
influence on PC activation and therefore a significant impact
on the antithrombotic properties of endothelial cells.
TM is also a cofactor for the thrombin-catalyzed activation
of TAFI. The active form of TAFI prevents fibrinolysis by
removing lysine residues from fibrin. The present study
indicates that PN-1 blocks TAFI activation by thrombin/TM
and thus may favor fibrinolysis. Consequently, PN-1 can
modulate the regulation of both the coagulation and fibrinolytic cascades by its interaction with TM. Nevertheless, the
net effect of the PN-1/TM interaction on the different
activities of thrombin remains to be determined.
In summary, PN-1 binds to endothelial cell TM. This
interaction appears to account for an important improvement
in the reactivity of the serpin with thrombin. The coordinated
action of TM with PN-1 has direct consequences on thrombin
activity among which a dramatic enhancement in the inhibition of fibrinoformation. The net effect of thrombin on
endothelial cells might thus be regulated by the concentration
of the TM/PN-1 complexes. Because TM and PN-1 are both
expressed in the vasculature, local variations of expression of
PN-1 and/or TM, are likely determinants for the regulation of
thrombin activity in vivo.
Acknowledgments
We acknowledge Dr Ann Gils for providing the recombinant TAFI,
Dr Daniel Hantaı̈ for the polyclonal antibody anti–PN-1, and Pr
Denis Monard for the recombinant rat PN-1. We thank Dr Charles
Brink for critical reading of the manuscript.
Sources of Funding
This work was supported by grants from INSERM, Université Paris
7 and Fondation de France n° 2006005672. Benjamin Richard was
supported by training grants from the Groupe d’Etude sur
l’Hémostase et la Thrombose (GEHT) and the Société française
d’Hématologie (SFH).
Disclosures
None
References
1. Esmon CT, Esmon NL, Harris KW. Complex formation between
thrombin and thrombomodulin inhibits both thrombin-catalyzed fibrin
formation and factor V activation. J Biol Chem. 1982;257:7944 –7947.
2. Esmon CT, Owen WG. Identification of an endothelial cell cofactor for
thrombin-catalyzed activation of protein C. Proc Natl Acad Sci U S A.
1981;78:2249 –2252.
3. Bajzar L, Morser J, Nesheim M. TAFI, or plasma procarboxypeptidase B,
couples the coagulation and fibrinolytic cascades through the thrombinthrombomodulin complex. J Biol Chem. 1996;271:16603–16608.
4. Bourin MC, Boffa MC, Bjork I, Lindahl U. Functional domains of rabbit
thrombomodulin. Proc Natl Acad Sci U S A. 1986;83:5924 –5928.
5. Preissner KT, Delvos U, Muller-Berghaus G. Binding of thrombin to
thrombomodulin accelerates inhibition of the enzyme by antithrombin III.
Evidence for a heparin-independent mechanism. Biochemistry. 1987;26:
2521–2528.
6. Rezaie AR, Cooper ST, Church FC, Esmon CT. Protein C inhibitor is a
potent inhibitor of the thrombin-thrombomodulin complex. J Biol Chem.
1995;270:25336 –25339.
7. Maruyama I, Bell CE, Majerus PW. Thrombomodulin is found on endothelium of arteries, veins, capillaries, and lymphatics, and on syncytiotrophoblast of human placenta. J Cell Biol. 1985;101:363–371.
8. Evans DL, McGrogan M, Scott RW, Carrell RW. Protease specificity and
heparin binding and activation of recombinant protease nexin I. J Biol
Chem. 1991;266:22307–22312.
9. Pratt CW, Church FC. General features of the heparin-binding serpins
antithrombin, heparin cofactor II and protein C inhibitor. Blood Coagul
Fibrinolysis. 1993;4:479 – 490.
10. Baker JB, Gronke RS. Protease nexins and cellular regulation. Semin
Thromb Hemost. 1986;12:216 –220.
11. Bouton MC, Richard B, Rossignol P, Philippe M, Guillin MC, Michel JB,
Jandrot-Perrus M. The serpin protease-nexin 1 is present in rat aortic
smooth muscle cells and is upregulated in L-NAME hypertensive rats.
Arterioscler Thromb Vasc Biol. 2003;23:142–147.
12. Leroy-Viard K, Jandrot-Perrus M, Tobelem G, Guillin MC. Covalent
binding of human thrombin to a human endothelial cell-associated
protein. Exp Cell Res. 1989;181:1–10.
13. Baker JB, Low DA, Simmer RL, Cunningham DD. Protease-nexin: a
cellular component that links thrombin and plasminogen activator and
mediates their binding to cells. Cell. 1980;21:37– 45.
14. Mbebi C, Hantai D, Jandrot-Perrus M, Doyennette MA, VerdiereSahuque M. Protease nexin I expression is up-regulated in human skeletal
muscle by injury-related factors. J Cell Physiol. 1999;179:305–314.
15. Reinhard E, Meier R, Halfter W, Rovelli G, Monard D. Detection of
glia-derived nexin in the olfactory system of the rat. Neuron. 1988;1:
387–394.
16. Low DA, Baker JB, Koonce WC, Cunningham DD. Released
protease-nexin regulates cellular binding, internalization, and degradation
of serine proteases. Proc Natl Acad Sci U S A. 1981;78:2340 –2344.
17. Rovelli G, Stone SR, Guidolin A, Sommer J, Monard D. Characterization
of the heparin-binding site of glia-derived nexin/protease nexin-1. Biochemistry. 1992;31:3542–3549.
Bouton et al
Thrombomodulin and PN-1 Are Complexed on HAECs
Downloaded from http://circres.ahajournals.org/ by guest on July 12, 2017
18. Myles T, Church FC, Whinna HC, Monard D, Stone SR. Role of
thrombin anion-binding exosite-I in the formation of thrombin-serpin
complexes. J Biol Chem. 1998;273:31203–31208.
19. Declerck PJ, De Mol M, Alessi MC, Baudner S, Paques EP, Preissner KT,
Muller-Berghaus G, Collen D. Purification and characterization of a
plasminogen activator inhibitor 1 binding protein from human plasma.
Identification as a multimeric form of S protein (vitronectin). J Biol
Chem. 1988;263:15454 –15461.
20. Rovelli G, Stone SR, Preissner KT, Monard D. Specific interaction of
vitronectin with the cell-secreted protease inhibitor glia-derived nexin and
its thrombin complex. Eur J Biochem. 1990;192:797– 803.
21. Farrell DH, Cunningham DD. Glycosaminoglycans on fibroblasts
accelerate thrombin inhibition by protease nexin-1. Biochem J. 1987;245:
543–550.
22. Bajzar L. Thrombin activatable fibrinolysis inhibitor and an antifibrinolytic pathway. Arterioscler Thromb Vasc Biol. 2000;20:2511–2518.
23. Hofsteenge J, Taguchi H, Stone SR. Effect of thrombomodulin on the
kinetics of the interaction of thrombin with substrates and inhibitors.
Biochem J. 1986;237:243–251.
24. Koyama T, Parkinson JF, Sie P, Bang NU, Muller-Berghaus G, Preissner
KT. Different glycoforms of human thrombomodulin. Their glycosaminoglycan-dependent modulatory effects on thrombin inactivation by
heparin cofactor II and antithrombin III. Eur J Biochem. 1991;198:
563–570.
25. Aritomi M, Watanabe N, Ohishi R, Gomi K, Kiyota T, Yamamoto S,
Ishida T, Maruyama I. Recombinant human soluble thrombomodulin
delivers bounded thrombin to antithrombin III: thrombomodulin asso-
26.
27.
28.
29.
30.
31.
32.
1181
ciates with free thrombin and is recycled to activate protein c. Thromb
Haemost. 1993;70:418 – 422.
Bourin MC, Ohlin AK, Lane DA, Stenflo J, Lindahl U. Relationship
between anticoagulant activities and polyanionic properties of rabbit
thrombomodulin. J Biol Chem. 1988;263:8044 – 8052.
Herndon ME, Stipp CS, Lander AD. Interactions of neural glycosaminoglycans and proteoglycans with protein ligands: assessment of selectivity,
heterogeneity and the participation of core proteins in binding. Glycobiology. 1999;9:143–155.
Dekker RJ, Pannekoek H, Horrevoets AJ. A steady-state competition
model describes the modulating effects of thrombomodulin on thrombin
inhibition by plasminogen activator inhibitor-1 in the absence and
presence of vitronectin. Eur J Biochem. 2003;270:1942–1951.
Richard B, Arocas V, Guillin MC, Michel JB, Jandrot-Perrus M, Bouton
MC. Protease nexin-1: a cellular serpin down-regulated by thrombin in rat
aortic smooth muscle cells. J Cell Physiol. 2004;201:138 –145.
Lin JH, McLean K, Morser J, Young TA, Wydro RM, Andrews WH,
Light DR. Modulation of glycosaminoglycan addition in naturally
expressed and recombinant human thrombomodulin. J Biol Chem. 1994;
269:25021–25030.
Fuentes-Prior P, Iwanaga Y, Huber R, Pagila R, Rumennik G, Seto M,
Morser J, Light DR, Bode W. Structural basis for the anticoagulant
activity of the thrombin-thrombomodulin complex. Nature. 2000;404:
518 –525.
Lu G, Chhum S, Krishnaswamy S. The affinity of protein C for the
thrombin.thrombomodulin complex is determined in a primary way by
active site-dependent interactions. J Biol Chem. 2005;280:15471–15478.
Protease Nexin-1 Interacts With Thrombomodulin and Modulates Its Anticoagulant Effect
Marie-Christine Bouton, Laurence Venisse, Benjamin Richard, Cécile Pouzet, Véronique
Arocas and Martine Jandrot-Perrus
Downloaded from http://circres.ahajournals.org/ by guest on July 12, 2017
Circ Res. 2007;100:1174-1181; originally published online March 22, 2007;
doi: 10.1161/01.RES.0000265066.92923.ee
Circulation Research is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231
Copyright © 2007 American Heart Association, Inc. All rights reserved.
Print ISSN: 0009-7330. Online ISSN: 1524-4571
The online version of this article, along with updated information and services, is located on the
World Wide Web at:
http://circres.ahajournals.org/content/100/8/1174
Data Supplement (unedited) at:
http://circres.ahajournals.org/content/suppl/2007/03/22/01.RES.0000265066.92923.ee.DC1
Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally published
in Circulation Research can be obtained via RightsLink, a service of the Copyright Clearance Center, not the
Editorial Office. Once the online version of the published article for which permission is being requested is
located, click Request Permissions in the middle column of the Web page under Services. Further information
about this process is available in the Permissions and Rights Question and Answer document.
Reprints: Information about reprints can be found online at:
http://www.lww.com/reprints
Subscriptions: Information about subscribing to Circulation Research is online at:
http://circres.ahajournals.org//subscriptions/
Bouton MC et al
Thrombomodulin and PN-1 are complexed on HAEC
ONLINE DATA SUPPLEMENT
Materials
Human α-thrombin was purified as previously described 1. Recombinant thrombin activable
fibrinolysis inhibitor, TAFI-AT (Ala147-Thr325)2 was a generous gift from Dr A. Gils
(Laboratory for Pharmaceutical Biology and Phytopharmacology, Leuven, Belgium).
Recombinant rat PN-1 (a generous gift from Dr D. Monard, Friedrich Miescher Institute,
Basel, Switzerland) was produced in yeast as previously described 3. The specific siRNA
duplexes targeted against human PN-1 were purchased from Ambion (Austin,TX, USA). The
monoclonal antibody (1F6) directed against human PN-1 sequence was obtained by
immunizing mice with the cDNA encoding human PN-1 in collaboration with Agrobio (La
Ferté St Aubin, France). Anti-human PN-1 IgGs were selected for their ability to bind PN-1 in
ELISA assays. IgGs were purified from ascites by chromatography on a HiTrap-protein A
Sepharose (Amersham Biosciences, Uppsala, Sweden). The polyclonal rabbit anti-PN-1
antibody was a gift from Dr D. Hantaï (Paris, France). Bovine protein C was from Enzyme
Research Laboratories (South Bend, IN, USA). Rabbit TM and the goat anti-rabbit TM IgG
were from American Diagnostica (Greenwich, CT, USA). The monoclonal anti-human TM
IgG was from Abcam (Cambridge, UK). Alexa 568-conjugated goat anti-mouse IgG, Alexa
488-conjugated rabbit anti-goat IgG and Oligofectamine were from Invitrogen (Cergy
Pontoise, France). Heparin was from Sanofi-Aventis (Paris, France). Bovine serum albumin
(BSA), polybrene (Hexadimethrine bromide), a synthetic quaternary polyamine used to
neutralized glycosaminoglycans, high molecular weight fucoidans, human plasma fibrinogen,
heparan- and chondroitin-sulfate, protease inhibitor cocktail for mammalian tissues,
chondroitinase ABC, O-Phenylenediamine Dihydrochloride (OPD) were from Sigma-Aldrich
(Saint Quentin-Fallavier, France). Hirudin was from Serbio (France). The chromogenic
Bouton MC et al
Thrombomodulin and PN-1 are complexed on HAEC
substrates S-2238 (H-D-Phe-pipecolyl-Arg-p-nitroanilide) and S-2366 (pyro-Glu-Pro-Arg-pnitroanilide) were purchased from Biogenic (Mauguio, France). Protein A/G-coated magnetic
beads were from Ademtech (Pessac, France), horseradish peroxydase coupled secondary
antibodies from Jackson ImmunoResearch (West Grove, PA, USA) and ECL from Amersham
Bioscience (Uppsala, Sweden).
Methods
Cell culture.
HAECs (human aortic endothelial cells, pooled donors) were purchased from
Cambrex (Rockland, ME, USA) and cultured according to manufacturer’s procedures. All
experiments were carried out by the 2nd to 5th cell passage.
Enzymatic treatment of cells.
Chondroitinase ABC was used to specifically cleave chondroitin-sulfate from the
endothelial cell surface. The enzyme was used at a concentration of 0.2 U/mL for 45 min in
the cell incubator 4. The efficacy of TM deglycosylation was assessed by western blot using a
lectin (Bandeiraea simplicifolia) that binds specifically to the N-acetyl-D-galactosamine
residue present on chondroitin-sulfate.
Immunocytochemical analysis.
Human aortic endothelial cells (HAECs) were seeded on gelatine-coated glass
coverslips and enzymatically treated or not as described above. Cells fixation was performed
in 2 % paraformaldehyde at room temperature for 10 minutes followed by PBS containing 5
% BSA for 1 h. After washings in PBS, cells were incubated overnight at 4°C with the antihuman PN-1 monoclonal antibody and the anti-rabbit TM polyclonal antibody (20 µg/mL and
50 µg/mL respectively in PBS containing 0.5 % BSA). After washings in PBS, cells were
incubated with Alexa 568-conjugated goat anti-mouse IgG and Alexa 488-conjugated rabbit
Bouton MC et al
Thrombomodulin and PN-1 are complexed on HAEC
anti-goat IgG for 2 h at room temperature, mounted and visualized with a confocal laserscanning microscope (LSM-510-META, Zeiss, Mannheim, Germany) equipped with a x63
oil-immersion objective. Simultaneous two-channel recording was performed with a pinhole
size of 1.00 Airy Units by using excitation wavelengths of 488 and 588 nm. The specificity of
the labelling was proved by the absence of signal when the primary antibody was omitted or
when using an irrelevant antibody. The colocalization of the two antibodies was analyzed
using the co-localization Zeiss LSM 510 3.2 Image Browser software.
Expression plasmids of PN-1 and transfection.
The cDNA coding sequence for human PN-1 preceded by a Kozak consensus
translation initiation site was inserted as a KpnI/EcoRV fragment into a pcDNA3 expression
vector (Invitrogen, Cergy Pontoise, France). The pcDNA3 vector containing the PN-1 coding
sequence or an empty pcDNA3 vector (mock) were transfected into HAECs by using
FuGENE 6 (Roche Applied Science, Meylan, France) according to the manufacturer’s
instructions. Cells treated with the transfection reagents alone are referred to as “control
cells”. Cells were assayed after 48 hours of transfection.
siRNA and cell transfection.
The following pre-designed annealed siRNA was chosen for PN-1 silencing: sense
sequence 5’-GGUUUUCAAUCAGAUUGUGtt-3’ and antisense sequence 5’CACAAUCUGAUUGAAAACCtg-3’. The pre-designed annealed irrelevant siRNA from
Eurogentec (Searing, Belgium) was used as negative control. The duplexes (150 pmol per
well of 12-well cell culture plates) were introduced into subconfluent (~ 80-85 %) cultured
HAECs using Oligofectamine reagents according to the manufacturer’s instructions. Cells
Bouton MC et al
Thrombomodulin and PN-1 are complexed on HAEC
treated with the transfection reagents alone are referred to as “control cells”. Cells were
assayed after 48 hours of transfection. To verify PN-1 extinction, RNA were collected 48 h
post-transfection and analyzed by quantitative RT-PCR.
Reverse transcription and quantitative real-time polymerase chain reaction.
Total RNA was extracted with Trizol (Invitrogen, Cergy Pontoise, France) and was
reverse-transcribed using the Superscript II Reverse transcriptase (Invitrogen, Cergy Pontoise,
France) as previously described 5. The resulting cDNA was used as a template for quantitative
PCR analysis of PN-1 and GAPDH mRNA expression in a LightCycler system with SYBR
Green detection (Roche Applied Science, Mannhein, Germany). PN-1 primers were: forward
5’-CCGCTGAAAGTTCTTGG-CA-3’ and reverse 5’-CAGCACCTGTAGGATTATGTCG3’. The following run protocol for PN-1 was used: denaturation: 95°C, 10 min; amplification
and quantification (40 cycles): 60°C, 10 sec; 72°C, 20 sec. The lightCycler run protocol for
GAPDH was as follows: denaturation: 95°C, 10 min; amplification and quantification (40
cycles): 65°C, 10 sec; 72°C, 20 sec. GAPDH primers were: forward 5’GGGCACCCTGGGCTAAACTGA-3’ and reverse 5’-TGCTCTTGCTGGGGCTGGT-3’.The
level of mRNA encoding PN-1 was normalized relative to GAPDH mRNA level.
Activation of TAFI.
Recombinant TAFI (450 nmol/L) was incubated with 10 nmol/L thrombin and 10
nmol/L thrombomodulin in 30 µl TBS containing 5 mmol/L CaCl2 and 0.1 % Tween 80 at
37°C for 0, 10, or 60 min, in the presence or absence of 10 nmol/L PN-1. The reactions were
stopped at the indicated time points by the addition of 2% SDS and used for SDS-PAGE
followed by silver staining.
Statistical analysis.
Bouton MC et al
Thrombomodulin and PN-1 are complexed on HAEC
Results are shown as means ± SD. Statistical evaluation was performed using a student’s ttest. P values < 0.05 were considered statistically significant.
References
1.
Bezeaud A, Denninger MH, Guillin MC. Interaction of human alpha-thrombin and
gamma-thrombin with antithrombin III, protein C and thrombomodulin. Eur J
Biochem. 1985;153:491-6.
2.
Gils A, Alessi MC, Brouwers E, Peeters M, Marx P, Leurs J, Bouma B, Hendriks D,
Juhan-Vague I, Declerck PJ. Development of a genotype 325-specific proCPU/TAFI
ELISA. Arterioscler Thromb Vasc Biol. 2003;23:1122-7.
3.
Sommer J, Meyhack B, Rovelli G, Buergi R, Monard D. Synthesis of glia-derived
nexin in yeast. Gene. 1989;85:453-459.
4.
Ainslie KM, Garanich JS, Dull RO, Tarbell JM. Vascular smooth muscle cell
glycocalyx influences shear stress-mediated contractile response. J Appl Physiol.
2005;98:242-249.
5.
Bouton MC, Richard B, Rossignol P, Philippe M, Guillin MC, Michel JB, JandrotPerrus M. The serpin protease-nexin 1 is present in rat aortic smooth muscle cells and
is upregulated in L-NAME hypertensive rats. Arterioscler Thromb Vasc Biol.
2003;23:142-7.