Clinical Science (2002) 102, 167–175 (Printed in Great Britain)
Adenosine inhibits thrombin-induced
expression of tissue factor on endothelial cells
by a nitric oxide-mediated mechanism
Esteban C. GABAZZA*†, Tatsuya HAYASHI*, Masaru IDO*, Yukihiko ADACHI†
and Koji SUZUKI*
*Department of Molecular Pathobiology, Mie University School of Medicine, Edobashi 2-174, Tsu, Mie 514-8507, Japan,
and †Third Department of Internal Medicine, Mie University School of Medicine, Edobashi 2-174, Tsu, Mie 514-8507, Japan
A
B
S
T
R
A
C
T
Enhanced expression of tissue factor (TF) is associated with the occurrence of coronary disease,
strokes and arterial thrombosis. We demonstrated previously that adenosine inhibits TF
expression in human umbilical vein endothelial cells (HUVECs) stimulated with inflammatory
mediators. In the present study, we evaluated the mechanism of adenosine-induced inhibition of
TF expression in HUVECs. The adenosine inhibitory activity on thrombin-induced TF expression
in HUVECs was potentiated by the NO precursor, L-arginine, but it was significantly suppressed
by the NO scavenger, 2(4-carboxyphenyl)-4,4,5,5-tetramethyl-imidazoline-1-oxyl-3-oxide,
and by endothelial NO synthase inhibitors, NG-monomethyl-L-arginine and NG-nitro-L-arginine
methyl ester, in a dose-dependent manner. The concentrations of nitrites, cGMP and cAMP in
the culture medium of HUVECs treated with a mixture of thrombin and adenosine were
significantly higher compared with the culture medium of HUVECs treated with thrombin
alone. Northern blotting showed that thrombin decreases and adenosine increases the eNOS
mRNA expression in HUVECs. A cAMP-dependent protein kinase inhibitor suppressed NOmediated TF inhibition in a dose-dependent manner. Overall, these results suggest that
adenosine inhibits thrombin-induced TF expression in endothelial cells by a NO-mediated
mechanism, and that increased intracellular formation of cAMP is implicated in this inhibitory
activity of NO.
INTRODUCTION
Tissue factor (TF) is a transmembrane receptor, which
after binding to activated Factor VII (Factor VIIa)
initiates the activation of the extrinsic pathway of blood
coagulation [1]. Under normal physiological conditions,
TF is not expressed in vascular endothelial cells and
monocytes, but it can be induced by inflammatory
stimulants such as lipopolysaccharide, thrombin, oxi-
dized lipoproteins and cytokines [tumour necrosis factor-α (TNF-α) and interleukin-1 (IL-1)] [1,2]. Enhanced
expression of TF on endothelial cells, monocytes and
macrophages is associated with the occurrence of coronary disease, atherosclerosis, strokes and arterial or
venous thrombosis [3,4]. Down-regulation of TF on
endothelial cells occurs in response to antioxidants,
retinoic acid and various anti-inflammatory cytokines
(IL-4, IL-10 and IL-13) [5,6]. We reported previously
Key words: adenosine receptor, coagulation, phosphodiesterase.
Abbreviations: carboxy-PTIO, 2(4-carboxyphenyl)-4,4,5,5-tetramethyl-imidazoline-1-oxyl-3-oxide ; EHNA, erythro-9-(2-hydroxy-3-nonyl)adenine:HCL ; NOS, NO synthase ; eNOS, endothelial NOS ; FBS, fetal bovine serum ; HUVEC, human umbilical
vein endothelial cell ; IL, interleukin ; iNOS, inducible NOS ; L-NAME, NG-nitro-L-arginine methyl ester ; L-NMMA, NGmonomethyl-L-arginine ; NF-κB, nuclear factor κB ; Rp-cAMPS, Rp-adenosine-3h,5h-cyclic monophosphorothioate ; S2222,
N-benzoyl-L-isoleucyl-L-glutamyl-glycyl-L-arginine-p-nitroanilide hydrochloride ; 8-SPT, 8-( p-sulphophenyl)-theophylline ;
TF, tissue factor ; TNF, tumour necrosis factor.
Correspondence : Dr Esteban Cesar Gabazza (e-mail gabazza!clin.medic.mie-u.ac.jp)
# 2002 The Biochemical Society and the Medical Research Society
167
168
E. C. Gabazza and others
[7,8] that the adenosine potentiator, dilazep, and adenosine itself can also down-regulate the expression of TF
induced by TNF-α or thrombin on endothelial cells ; this
mechanism of adenosine-induced down-regulation of TF
remains to be clarified.
Adenosine is a naturally occurring purine nucleoside
that may be intracellularly produced by the hydrolysis of
AMP or by the catabolism of S-adenosylhomocysteine
[9]. Under conditions of hypoxia or cellular activation
the intracellular pool of adenosine increases, and it may
be released from the cell or tissue affected [9]. In the
cardiovascular system, adenosine is known to protect
vascular endothelial cells against ischaemia-reperfusion
injury, oxidative stress and to play an important role in
the maintenance of vascular wall tone [9]. In addition to
adenosine, NO has been also shown to exert cytoprotective activity against vascular endothelial injury [10]. NO
is a free gas generated via the oxidation of a terminal
guanidinium nitrogen on the amino acid L-arginine
through a reaction catalysed by NO synthase (NOS),
yielding the co-product L-citrulline in addition to NO
[11]. Inhibition of platelet activation and TF expression
on monocytes are some of the cytoprotective activities of
NO [12–14]. In the present study, we evaluated whether
the inhibitory activity of adenosine on thrombin-induced
TF activity in endothelial cells is mediated by NO.
sulphate, were purchased from Dojindo (Kumamoto,
Japan). Epithelial cell and fibroblast growth factors were
purchased from Boehringer Mannheim (Mannheim,
Germany) and [α-$#P]dCTP from Daiichi Kagahu
(Tokyo, Japan). The NO donor FK409 was a kind gift
from Fujisawa (Osaka, Japan). The inhibitor of the
cAMP-dependent protein kinase inhibitor Rp-adenosine-3h,5h-cyclic monophosphorothioate (Rp-cAMPS),
and the inhibitor of the phosphodiesterase II, erythro9-(2-hydroxy-3-nonyl)adenine:HCL (EHNA), were
purchased from Biomol (Plymouth Meeting, PA,
U.S.A.).
Cell culture
METHODS
Human umbilical vein endothelial cells (HUVECs) were
purchased from Sanko Junyaku (Tokyo, Japan). The cells
were cultured in MCDB 131 medium (Chlorella Industry, Tokyo, Japan) supplemented with 5 % heatinactivated fetal bovine serum (FBS) (Gibco BRL, Grand
Island, NY, U.S.A.), 50 µg\ml gentamycin (Nacalai
Tesque), 50 ng\ml amphotericin B (Bristol–Myers
Squibb, New Brunswick, NJ, U.S.A.), 50 µg\ml ascorbic
acid, 0.5 µg\ml hydrocortisone, 10 µg\ml vascular endothelial cell growth factor (Beckon Dickinson, Bedford,
MA, U.S.A.), 8 ng\ml fibroblast growth factor and
20 ng\ml epithelial cell growth factor at 37 mC under an
atmosphere of 95 % air and 5 % CO . The viability of the
#
cells in different experiments was evaluated using a cell
counting kit.
Materials
Determination of cell surface TF activity
Human Factor VIIa and Factor X were kindly provided
by Dr Tomohiro Nakagaki (Chemo-Sera Therapeutic
Institute, Kumammoto, Japan). Human thrombin was
purchased from Yoshitomi Pharmaceuticals (Osaka,
Japan). The synthetic peptide substrate for Factor Xa,
N-benzoyl-L-isoleucyl-L-glutamyl-glycyl-L-arginine-pnitroanilide hydrochloride (S2222), was purchased from
Chromogenic (Mo$ lndal, Sweden). Hydrocortisone and
the inhibitors of NOS, NG-nitro-L-arginine methyl
ester (L-NAME) and NG-monomethyl-L-arginine
(L-NMMA), were purchased from Wako Chemicals
(Osaka, Japan). The NO scavenger 2(4-carboxyphenyl)-4,4,5,5-tetramethyl-imidazoline-1-oxyl-3-oxide
(carboxy-PTIO) was purchased from Funakoshi (Tokyo,
Japan) and L-arginine from Nacalai Tesque (Kyoto,
Japan). Dulbecco’s phosphate-buffered saline, BSA and
adenosine were from Sigma (St. Louis, MO, U.S.A.).
Glyceraldehyde-3-phosphate dehydrogenase and endothelial NOS (eNOS) DNA probes were from Cayman
(Ann Arbor, MI, U.S.A.), and 8-( p-sulphophenyl)theophylline (8-SPT) from Research Biochemicals
International (Natick, MA, U.S.A.). The reagents used
to check the viability of the cells, 2-(4-iodophenyl)-3-(4nitrophenyl)-5-(2,4-disulphophenyl)-2H-tetrazoliumsodium and 1-methoxy-5-methyl-phenazinium methyl-
TF activity was determined by the degree of Factor X
activation by the Factor VIIa–TF complex on HUVECs
after stimulation with thrombin. The cells (48-well plate,
5i10& cells\well) were incubated for 5 h in the presence
of various concentrations of thrombin diluted in MCDB
131 medium without FBS and supplements, and then the
cells were washed twice with Hepes-buffered saline
(20 mM Hepes and 150 mM NaCl, pH 7.5) containing
5 mM CaCl . Thereafter, the cells were incubated in the
#
presence of a mixture of 50 µl of 4 nM Factor VIIa, 50 µl
of 1 µM Factor X and 150 µl of Hepes-buffered saline,
containing 5 mM CaCl , for 60 min at 25 mC. The reaction
#
was stopped by adding 10 µl of 100 mM EDTA, and
generated Factor Xa was determined using 100 µM S2222.
The reaction was stopped by the addition of 10 µl of 20 %
acetic acid, and colour development was determined by
measuring the absorbance at 405 nm with a EAR 340
microplate reader (SLT-Lab, Salzburg, Austria).
# 2002 The Biochemical Society and the Medical Research Society
Assessment of production of NO, cGMP and
cAMP induced by adenosine in HUVECs
NO is an unstable free radical that rapidly converts to
nitrite and nitrate in the presence of oxygen [14]. To
evaluate whether adenosine induces the increased formation of NO in HUVECs, these cells were incubated
Adenosine inhibits tissue factor expression via nitric oxide
overnight in FBS- and supplement-free MCDB 131
medium, at 37 mC under an atmosphere of 95 % air and
5 % CO , in the presence of various concentrations of the
#
nucleoside. In similar assays, HUVECs were also stimulated with 10 µM adenosine in the presence of 1 mM 8SPT or 2 mM NOS inhibitors, L-NMMA or L-NAME.
The total concentration of nitrite and\or nitrate in the
medium was measured by a modified Griess reaction by
using a commercial colourimetric assay kit (Cayman)
according to the manufacturer’s instructions [15]. The
sensitivity of the assay was 2 µM nitrite\nitrate, and the
inter-assay and intra-assay coefficients of variation were
less than 10 %. For all experiments, background levels of
nitrite\nitrate in the culture medium were determined
by a parallel incubation of medium containing the
appropriate reagents in the absence of HUVECs. The
production of NO by HUVECs was determined by subtracting the background levels of nitrite\nitrate obtained
in the absence of HUVECs from values obtained in the
presence of the cells. Interaction of NO with the haem
component of soluble guanylate cyclase stimulates enzymic conversion of GTP into cGMP ; thus, as a marker
of NO generation, the intracellular levels of cGMP were
also determined in adenosine-treated HUVECs. The
total content of cGMP or cAMP in adenosine-treated
HUVECs was measured using a commercial cGMP or
cAMP enzyme immunoassay kit from Amersham Pharmacia Biotech (Little Chalfont, Buckinghamshire, U.K.).
The intra-assay and inter-assay coefficients of variation
of the cGMP and cAMP assays were 7 % and 10 %
respectively.
Assessment of the effect of NO scavenger,
NO inhibitors and L-arginine on adenosineinduced inhibition of cell surface TF activity
HUVECs were cultured in 48-well microplates to confluency, and then stimulated with 500 nM thrombin and
10 µM adenosine in the presence of various concentrations of carboxy-PTIO, L-NMMA, L-NAME or
L-arginine. After 5 h, the cells were washed, and the effect
of TF activity on HUVECs was determined as described
above.
Evaluation of adenosine effect on eNOS or
inducible NOS (iNOS) gene expression in
HUVECs by Northern-blot analysis
For total RNA preparation, confluent HUVEC monolayers in 100-mm dishes were incubated overnight in
FBS- and supplement-free MCBD 131 medium containing no stimulant, 500 nM thrombin alone, 10 µM
adenosine alone, a mixture of 500 nM thrombin
and 10 µM adenosine, or a mixture of 500 nM thrombin,
10 µM adenosine and 1 mM 8-SPT. Total RNA was
extracted from cells by the guanidine–isothiocyanate
procedure using Trizol Reagent. Total RNA (15 µg) were
run in a 2 % agarose gel containing 660 mM de-ionized
formaldehyde, transferred for 24 h by capillary action on
to a nylon membrane and then dried for 2 h under
vacuum at 80 mC. The membrane was pre-hybridized in a
solution containing 20 mM Tris\HCl, pH 7.5, 750 mM
NaCl, 2.5 mM EDTA, 1 % SDS, 0.5 % Denhardt’s
solution and 50 % de-ionized formamide, and then
hybridized with eNOS, iNOS or GADPH cDNA probes
labelled with [α-$#P]dCTP. The membrane was then
washed twice in 0.2iSSC (0.15 M NaCl\0.015 M sodium citrate) and 0.1 % SDS at 60 mC for 30 min, exposed
to imaging plates and analysed using BAS-2000 Image
Analyzer. The amount of eNOS or iNOS mRNA was
normalized against the glyceraldehyde-3-phosphate dehydrogenase mRNA.
Evaluation of the effect of cAMP-dependent
protein kinase inhibitor on NO-mediated
inhibition of TF activity on HUVECs
To evaluate the role of cAMP-dependent protein kinase
in the inhibitory activity of NO on thrombin-induced
TF activity, HUVECs cultured in 48-well plates were
incubated overnight in the presence of 250 nM thrombin,
25 µM FK409, a NO donor, and increasing concentrations of Rp-cAMPS, a cAMP-dependent protein kinase
inhibitor. The cells were then washed, and the effect of
TF activity on HUVECs was determined as described
above.
Evaluation of the effect of
phosphodiesterase II inhibitor on TF activity
on HUVECs
Intracellular cGMP is known to inhibit the activity of
phosphodiesterase II leading to increased intracellular
formation of cAMP. To evaluate the role of phosphodiesterase II on TF activity on HUVECs induced by
thrombin, HUVECs cultured in 48-well plates were
incubated overnight in the presence of 250 nM thrombin,
a mixture of 250 nM thrombin and 10 µM adenosine,
or a mixture of 250 nM thrombin and 10 µM adenosine plus increasing concentrations of EHNA, an inhibitor of phosphodiesterase II. The cells were then
washed, and the effect of TF activity on HUVECs was
determined as described above.
Measurement of adenosine secretion
induced by NO donor from HUVECs
To evaluate whether NO itself stimulates the secretion of
adenosine in HUVECs, these cells were cultured in FBSfree medium in the presence or absence of the NO donor,
FK409, at 37 mC for 24 h under an atmosphere of 95 % air
and 5 % CO . Adenosine concentration in the con#
ditioned medium was then measured using a commercial
radio-immunoassay kit from Yamasa (Chiba, Japan).
# 2002 The Biochemical Society and the Medical Research Society
169
170
E. C. Gabazza and others
Statistical analysis
Results are expressed as the meanpS.E.M., unless otherwise specified. The difference between the means of two
variables was calculated by the Mann–Whitney U-test,
and between three or more variables by analysis of
variance. A P 0.05 was considered as statistically
significant. Statistical analyses were carried out using the
StatView 4.5 package for the Macintosh (Abacus Concepts, Berkeley, CA, U.S.A.).
RESULTS
Effect of adenosine, 8-SPT and NOS
inhibitors on NO, cGMP and cAMP
production
To determine the effect of adenosine on NO production
by HUVECs, we measured the levels of nitrite\nitrate
and cGMP in conditioned media and homogenates from
HUVECs treated with adenosine. High levels of nitrite\
Figure 2 Effect of NO scavenger and NOS inhibitors on
adenosine-induced inhibition of TF activity on HUVECs stimulated with thrombin
(A) HUVECs were stimulated with 500 nM thrombin and 10 µM adenosine in the
presence of various concentrations of carboxy-PTIO. NO scavenger blocked
the inhibitory activity of adenosine on TF expression. TF activity is expressed as
a percentage of the control activity (thrombin alone). *P 0.05 compared with
thrombin alone, and **P 0.05 compared with 500 nM thrombinj10 µM
adenosine without NO scavenger. (B) HUVECs were stimulated with 500 nM
thrombin and 10 µM adenosine in the presence of various concentrations of the
NOS inhibitor L-NMMA (closed columns) or L-NAME (open columns). NOS inhibitors
blocked the inhibitory activity of adenosine on TF expression. TF activity is
expressed as a percentage of the control activity (thrombin alone). Results are
expressed as the meanpS.E.M. for four independent experiments, each done in
quadruplicate. *P 0.05 compared with thrombin alone, and **P 0.05
compared with 500 nM thrombinj10 µM adenosine without NO scavenger.
Figure 1 Assessment of adenosine-induced production of
NO, cGMP and cAMP
HUVECs were incubated in the presence of various concentrations of adenosine, 8SPT, L-NMMA or L-NAME. The concentrations of nitrites (A) were measured in the
conditioned medium, and cGMP and cAMP levels (B) in the cell lysates. Adenosine
induced increased production of nitrites, cGMP (open columns) and cAMP (closed
columns) ; cAMP production was inhibited by NOS inhibitors. Results are expressed
as the meanpS.E.M. for four independent experiments, each carried out in
quadruplicate. *P 0.05 compared with no adenosine, †P 0.05 compared
with 10 µM adenosine, §P 0.05 compared with no adenosine, and ¶P
0.05 compared with adenosine alone.
# 2002 The Biochemical Society and the Medical Research Society
nitrate (Figure 1A) were observed in the conditioned
medium and homogenates from HUVECs cultured in
the presence of adenosine ; this stimulatory activity of
adenosine was inhibited by 8-SPT, L-NMMA and LNAME, suggesting the important role of adenosine in
NO formation. The levels of cGMP, which increased
in the presence of adenosine, were also decreased by
8-SPT, L-NMMA and L-NAME, but not to a significant
level ; cAMP was also increased in the homogenates from
HUVECs treated with adenosine (Figure 1B). The
increase in cAMP induced by adenosine was inhibited by
the presence of the NOS inhibitors L-NMMA and LNAME, suggesting that this enhancement in cAMP levels
Adenosine inhibits tissue factor expression via nitric oxide
Figure 3 Effect of the NO precursor, L-arginine, on adenosine-induced inhibition of TF activity on HUVECs stimulated
with thrombin
Subconfluent HUVECs were incubated in the presence of 500 nM thrombin, 10 µM
adenosine and various concentrations of L-arginine. After washing the cells with
Hepes-buffered saline, TF activity was determined. TF activity is expressed as a
percentage of the control activity (thrombin alone). L-Arginine potentiated the
inhibitory activity of adenosine on TF activity. Further, L-arginine itself was also
able to inhibit thrombin-induced TF activity on HUVECs. Values are meanpS.E.M.
for four independent experiments, each done in quadruplicate. *P 0.05.
depended, at least in part, on the intracellular production
of cGMP.
Effect of adenosine-induced production of
NO on TF activity in HUVECs
To investigate the role of adenosine-induced NO in the
inhibitory effect of adenosine on TF activity, HUVECs
pre-treated with thrombin and adenosine were incubated
in the presence of various concentrations of the NO
scavenger, carboxy-PTIO, or NOS inhibitors, L-NMMA
and L-NAME. The results of the experiments showed
that carboxy-PTIO (Figure 2A) and NOS inhibitors
(Figure 2B) block the inhibition of the cell surface TF
activity by adenosine in a dose-dependent manner. The
blockade of adenosine inhibitory activity was significant
at concentrations over 6 µM and 1 mM for carboxyPTIO and the NOS inhibitors respectively. The effect of
the precursor of NO, L-arginine, on the inhibition of cell
surface TF activity by adenosine was also evaluated. LArginine potentiated the adenosine-induced inhibition of
TF activity on HUVECs (Figure 3). Furthermore, Larginine itself was also able to inhibit thrombin-induced
TF activity on HUVECs. These results implicate NO
production in the mechanism of adenosine-induced
inhibition of cell surface TF activity on endothelial cells.
The concentrations used of carboxy-PTIO or those of
the NOS inhibitors did not affect the viability of
HUVECs.
Figure 4 Northern-blot analysis of adenosine-induced eNOS
mRNA in HUVECs
(A) Northern-blot analysis of the following : untreated HUVECs (lane 1) ; cells
treated with 500 nM thrombin alone (lane 2) ; 500 nM thrombin and 10 µM
adenosine (lane 3) ; 500 nM thrombin, 10 µM adenosine and 1 mM 8-SPT (lane 4) ;
or with 10 µM adenosine alone (lane 5). (B) Densitometric analysis and
eNOS/GADPH ratio. Adenosine significantly blocked the decreased transcription
of eNOS gene induced by thrombin in HUVECs. The results shown are representative
of three independent experiments. *P 0.05, compared with untreated HUVECs
(lane 1), and †P 0.05, compared with cells treated with thrombin and
adenosine (lane 3).
Effect of thrombin and adenosine on eNOS
or iNOS mRNA expression in HUVECs
To evaluate the effect of thrombin and adenosine on the
expression of eNOS or iNOS mRNA, HUVECs were
incubated in the presence of thrombin, adenosine or 8SPT, and the level of eNOS or iNOS gene transcription
was evaluated by Northern blotting. Thrombin alone
decreased and adenosine alone increased the steady-state
transcript levels of eNOS gene (Figure 4). Adenosine
inhibited the down-regulation of eNOS mRNA levels
induced by thrombin, and this effect was blocked by
8-SPT. Densitometric analysis showed that thrombin
decreases and adenosine increases the steady state of
eNOS mRNA levels significantly (Figure 4B). The eNOS
mRNA levels in HUVECs treated with thrombin and
adenosine were significantly higher compared with cells
treated with thrombin alone. 8-SPT suppressed the inhibitory activity of adenosine on thrombin-induced
inhibition of eNOS mRNA expression. On the other
hand, the expression of iNOS mRNA was not significantly changed by the addition of adenosine to the
HUVECs (results not shown).
# 2002 The Biochemical Society and the Medical Research Society
171
172
E. C. Gabazza and others
Effect of cAMP-dependent protein kinase
inhibitor on NO-mediated inhibition of TF
The cAMP-dependent protein kinase has been reported
to decrease the transcription of TF through components
of the nuclear factor κB (NF-κB)\Rel family of transcription factors. To evaluate whether the cAMPdependent protein kinase is involved in the inhibitory
activity of NO on thrombin-induced TF activity,
HUVECs were cultured in the presence of thrombin, the
NO donor FK409 and increasing concentrations of RpcAMPS, and then TF activity was measured at the surface
of these cells. The results showed that FK409 significantly
inhibited TF activity on HUVECs stimulated by thrombin (Figure 5B). This suppressive effect of NO on TF
activity was significantly blocked by Rp-cAMPS at
concentrations over 0.6 µM, suggesting the important
role of cAMP-dependent protein kinase in the inhibition
of TF expression by NO.
Effect of NO donor on adenosine secretion
by HUVECs
Figure 5 Effect of phosphodiesterase II inhibitor and cAMPdependent protein kinase inhibitor on TF activity
(A) TF activity was measured on HUVECs after incubating in the presence of
thrombin, a mixture of thrombin and adenosine, or a mixture of thrombin and
adenosine plus increasing concentrations of EHNA. Adenosine significantly inhibited
TF activity induced by thrombin on HUVECs ; the addition of EHNA in the inhibition
assay provoked further inhibition of TF in a dose-dependent manner. (B) TF
activity was measured on HUVECs after incubation in the presence of thrombin,
FK409 and increasing concentrations of Rp-cAMPS. The NO donor, FK409,
significantly inhibited TF activity induced by thrombin on HUVECs. This suppressive
effect of NO on TF activity was significantly blocked by Rp-cAMPS in a dosedependent manner. FP 0.01, compared with thrombin alone, and *P 0.05
compared with thrombin plus FK409.
It is known that both NO and adenosine increase and
exert a cytoprotective activity at foci of tissue injury or
hypoxia. As shown above, adenosine may stimulate the
production of NO from endothelial cells in these
pathological conditions. However, it is also conceivable
that NO itself may stimulate the local production of
adenosine. In the present study, this potential effect
of NO on adenosine secretion was also evaluated in
HUVECs. Significant increased levels of adenosine were
observed in the supernatants of HUVECs treated with
50 µM NO donor (FK409) as compared with the controls
(results not shown), suggesting that adenosine may also
be secreted from endothelial cells by NO via an autocrine
mechanism.
Effect of phosphodiesterase II inhibitor on
TF activity
DISCUSSION
Activation of phosphodiesterase II is known to decrease
intracellular generation of cAMP. Increased intracellular concentration of cGMP may up-regulate intracellular
formation of cAMP by blocking the activity of phosphodiesterase II. To evaluate the effect of phosphodiesterase
II inhibition on TF activity induced by thrombin,
HUVECs were incubated in the presence of thrombin, a
mixture of thrombin and adenosine, or a mixture of
thrombin and adenosine plus increasing concentrations
of EHNA, an inhibitor of phosphodiesterase II, and then
TF activity was measured on the cells. It can be seen in
Figure 5(A) that adenosine significantly inhibited TF
activity induced by thrombin on HUVECs ; the addition
of EHNA provoked further inhibition of TF activity in a
dose-dependent manner.
The vascular endothelium plays a critical role in the
regulation of coagulation through the constitutive expression and release of anticoagulants, and the inducible
expression of procoagulant substances [16]. TF is the
clotting factor which mediates the procoagulant activity
of the vascular endothelium. It is expressed on endothelial
cells upon stimulation with inflammatory stimuli ; once
expressed, plasma Factor VIIa binds to TF, and this
TF–Factor VIIa complex activates the extrinsic pathway
of blood coagulation [17]. Activation of this coagulation
pathway, with subsequent thrombin generation, has been
implicated in the mechanism of several pathological
conditions, such as unstable angina, atherosclerosis,
chronic inflammation and cancer [17]. The cellular effect
of thrombin may be mediated by the protease-activated
# 2002 The Biochemical Society and the Medical Research Society
Adenosine inhibits tissue factor expression via nitric oxide
receptors-1, -3 or -4 present on the surface of several
cells, including endothelial cells [18]. By acting on these
receptors, thrombin itself may perpetuate coagulation
activation by inducing the expression of TF on endothelial cells [7]. We have shown previously [7,8] that
adenosine may suppress this vicious cycle of thrombin
formation by inhibiting thrombin-induced expression of
TF on endothelial cells. Most of the physiological effects
of adenosine take place by the action of adenosine at
specific cell-surface receptors [19]. In a previous study
[8], we have shown that adenosine receptor antagonists,
including 8-SPT, counteract the inhibitory activity of
adenosine on TF and that adenosine receptor agonists
block the TF expression in HUVECs ; these observations
suggest that the effect of adenosine is mediated by its
receptors.
Adenosine is normally produced during the basal
metabolic activity of the cells. The extracellular concentration of adenosine increases dramatically (up to
micromolar levels) during ischaemia and hypoxia, and it
is believed to confer cytoprotection to the ischaemic
tissue [8]. Adenosine also provides protective activity
against ischaemia-reperfusion injury, and the beneficial
effect of cardiac pre-conditioning has been attributed to
the high concentration of adenosine in affected tissues
[16–21]. The mechanism of this protective effect of
adenosine is not completely clear. Adenosine protects
against injury by enhancing the activity of antioxidant
enzymes (e.g., superoxide dismutase), which are known
to play a key role in the elimination of reactive oxygen
species generated during metabolic activity or ischaemiareperfusion injury [8,22]. TF is also markedly expressed
during hypoxia, ischaemic damage and ischaemiareperfusion injury [23]. Abnormal expression of TF has
been involved in the pathogenesis of stable and unstable
coronary syndromes, acute myocardial infarction, and
acute thrombosis with vascular restenosis [24–28]. It has
also been reported previously [29] that TF expressed on
endothelial cells activates the coagulation system and
induces tissue injury during the post-ischaemic cardiac
reperfusion. Thus the observation that adenosine inhibits
thrombin-induced TF activity may also constitute an
important mechanism by which adenosine protects normal tissue against injury.
NO has been reported to mediate some of the
biological actions of adenosine, such as the regulation of
smooth muscle tone and protection against ischaemiareperfusion injury [30–32]. In the present study, we
investigated whether adenosine inhibits TF expression on
HUVECs by increasing the production of NO from
these cells. Levels of the NO scavenger and NOS
inhibitors decreased, and L-arginine potentiated the
adenosine-mediated inhibition of TF activity on
HUVECs in a dose-dependent manner, suggesting that
the NO pathway is involved in the inhibitory activity
of adenosine on TF expression. This inhibitory activity of
adenosine depended on formation de novo of NO from
endothelial cells, as shown by the increase in the
concentrations of nitrite\nitrate and cGMP in HUVECs
after treatment with adenosine, and by the suppressive
effect on NO formation of the non-selective adenosine
receptor antagonist, 8-SPT. In addition, we found that
NO itself may stimulate the secretion of adenosine from
HUVECs, suggesting the existence of an autocrine
mechanism in the induction of NO by adenosine.
Overall, these observations suggest that NO mediates, at
least in part, the anticoagulant effect of adenosine on
endothelial cells.
The major rate-limiting step in NO formation is the
activity of NOS. There are two forms of NOS in
endothelial cells, a constitutive eNOS and an iNOS [10].
The expression of iNOS may be induced by cytokines
(TNF-α and IL-1) or lipopolysaccharide in endothelial
cells ; however, in our present study, adenosine did not
affect the expression of iNOS in HUVECs. With regards
to eNOS, it was initially thought that eNOS transcript
abundance was not affected by cytokines or other stimuli.
However, previous studies [33–38] have shown that the
levels of eNOS mRNA may also be decreased by
TNF-α, hypoxia, lipopolysaccharide or low-density
lipoproteins, and increased by sex steroids or shear stress.
The present study showed that thrombin may also downregulate the baseline expression of eNOS in HUVECs ;
this inhibitory activity of thrombin on eNOS expression
was suppressed by adenosine. The adenosine receptor
antagonist 8-SPT blocked this protective activity of
adenosine on eNOS gene transcription, suggesting that
its effect is mediated by adenosine receptors. TNF-α and
hypoxia were reported to decrease eNOS gene transcription by decreasing eNOS mRNA stability [38] ;
thus, it is also conceivable that thrombin exerts its effect
by decreasing, and adenosine by increasing, the stability
of eNOS mRNA levels in endothelial cells.
The intracellular events involved in the inhibitory
activity of adenosine on TF expression are not completely
clear. Protein kinase C and protein tyrosine kinase are
involved in the expression of TF by monocytes or
endothelial cells in response to lipopolysaccharide, PMA
and cytokines, such as TNF-α and IL-1 [39–41]. During
the process of cellular activation, cytoplasmic NFκB–Rel
complexes dissociate from the inhibitor protein IκBα,
and the transcription factors translocate to the nucleus to
initiate the expression of TF. Protease inhibitors that
block the proteolysis of the inhibitor protein, IκBα, and
agents that increase intracellular cAMP levels suppress
the cellular expression of TF [42–44]. Activation of the
NO–cGMP pathway is associated with increased intracellular cAMP formation [45]. The increased intracellular levels of cAMP observed in the present study
depended on activation of the NO–cGMP pathway,
because the NO inhibitors, L-NMMA and L-NAME,
decreased cAMP levels in HUVECs. cGMP may increase
# 2002 The Biochemical Society and the Medical Research Society
173
174
E. C. Gabazza and others
ACKNOWLEDGMENTS
We thank Ms Taemi Hibi for her excellent technical
assistance.
REFERENCES
1
2
3
4
5
Figure 6 Potential mechanism of TF inhibition by adenosine
in HUVECs
the intracellular level of cAMP by inhibiting phosphodiesterase II, and by this mechanism it may decrease TF
activity [48]. In agreement with this, in the present study
we found that the addition of EHNA, a phosphodiesterase II inhibitor, potentiates inhibition of TF activity
on HUVECs in a dose-dependent manner. The mechanism of TF expression inhibition by intracellular cAMP
is not clear, but some studies suggest that activation
of cAMP-dependent protein kinases inhibits NFκBmediated transcription of TF in human monocytes and
endothelial cells [46]. To evaluate whether cAMPdependent protein kinase is also involved in the inhibitory activity of NO on thrombin-induced TF activity
in endothelial cells, we stimulated HUVECs with thrombin in the presence of the NO donor, FK409, and the
protein kinase inhibitor, Rp-cAMPS. The dose of FK409
used in this experiment was within the range at which it
inhibits platelet aggregation (approx. 0.32 µM) and vasoconstriction (approx. 0.32 nM), as demonstrated in previous studies [47,48]. The results showed that Rp-cAMPS
blocks the suppressive effect of FK409 on TF activity in
HUVECs in a dose-dependent manner, implicating the
cAMP-dependent protein kinase in the NO-mediated
inhibition of TF in HUVECs (Figure 6).
Overall, the results of this study suggest that adenosine
inhibits thrombin-induced TF expression in endothelial
cells by a NO-mediated mechanism, and that the increased intracellular formation of cAMP is implicated in
this inhibitory activity of NO.
# 2002 The Biochemical Society and the Medical Research Society
6
7
8
9
10
11
12
13
14
15
16
17
18
Scarpati, E. M. and Sadler, J. E. (1989) Regulation of
endothelial cell coagulant properties. Modulation of tissue
factor, plasminogen activator inhibitors, and
thrombomodulin by phorbol 12-myristate 13-acetate and
tumor necrosis factor. J. Biol. Chem. 264, 20705–20713
Bartha, K., Brisson, C., Archipoff, G. et al. (1993)
Thrombin regulates tissue factor and thrombomodulin
mRNA levels and activities in human saphenous vein
endothelial cells by distinct mechanisms. J. Biol. Chem.
268, 421–429
Wada, H., Wakita, Y. and Shiku, H. (1995) Tissue factor
expression in endothelial cells in health and disease.
Blood Coag. Fibrinolysis 6, S26–S31
Osterud, B. (1998) Tissue factor expression by
monocytes : regulation and pathophysiological roles.
Blood Coag. Fibrinolysis 9, 9–14
Herbert, J. M., Savi, P., Laplace, M. C. et al. (1993) IL-4
and IL-13 exhibit comparable abilities to reduce pyrogeninduced expression of procoagulant activity in endothelial
cells and monocytes. FEBS Lett. 328, 268–270
Ishii, H., Horie, S., Kizaki, K. and Kazama, M. (1992)
Retinoic acid counteracts both the down-regulation of
thrombomodulin and the induction of tissue factor in
cultured human endothelial cells exposed to tumor
necrosis factor. Blood 80, 2556–2562
Deguchi, H., Takeya, H., Wada, H. et al. (1997) Dilazep,
an antiplatelet agent, inhibits tissue factor expression in
endothelial cells and monocytes. Blood 90, 2345–2356
Deguchi, H., Takeya, H., Urano, H., Gabazza, E. C.,
Zhou, H. and Suzuki, K. (1998) Adenosine regulates
tissue factor expression on endothelial cells. Thromb. Res.
91, 57–64
Ramkumar, V., Nie, Z., Rybak, L. P. and Maggirwar,
S. B. (1995) Adenosine, antioxidant enzymes and
cytoprotection. Trends Pharmacol. Sci. 16, 283–285
Yang, B. C., Chen, L. Y., Saldeen, T. G. and Mehta, J. L.
(1997) Reperfusion injury in the endotoxin-treated rat
heart : reevaluation of the role of nitric oxide. Br. J.
Pharmacol. 120, 305–311
Moncada, S. and Higgs, A. N. (1993) The L-arginine–
nitric oxide pathway. N. Engl. J. Med. 329, 2002–2012
Langford, E. J., Brown, A. S., Wainwright, R. J.
et al. (1994) Inhibition of platelet activity by
S-nitrosoglutathione during coronary angioplasty.
Lancet 344, 1458–1460
Kaul, S., Makkar, R. R., Nakamura, M. et al. (1996)
Inhibition of acute stent thrombosis under high-shear
flow conditions by nitric oxide donor, DMHD\NO. An
ex vivo porcine arteriovenous shunt study. Circulation
94, 2228–2234
Gerlach, M., Keh, D., Bezold, G. et al. (1998) Nitric
oxide inhibits tissue factor synthesis, expression and
activity in human monocytes by prior formation of
peroxynitrite. Intensive Care Med. 24, 1199–1208
Green, L. C., Wagner, D. A., Glogowski, J., Skipper,
P. L., Wishnok, J. S. and Tannenbaum, S. R. (1982)
Analysis of nitrate, nitrite, and ["&N ]nitrate in biological
fluids. Anal. Biochem. 126, 131–138
Nawroth, P. P. and Stern, D. M. (1987) Endothelial cell
procoagulant properties and the host response. Semin.
Thromb. Hemost. 13, 391–397
Bauer, K. A. (1997) Activation of the Factor VII–tissue
factor pathway. Thromb. Haemostasis 78, 108–111
Brass, L. F. and Molino, M. (1997) Protease-activated
G-protein-coupled receptors on human platelets and
endothelial cells. Thromb. Haemostasis 78, 234–241
Adenosine inhibits tissue factor expression via nitric oxide
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
Janier, M. F., Vanoverschelde, J. L. and Bergmann, S. R.
(1993) Adenosine protects ischemic and reperfused
myocardium by receptor-mediated mechanisms. Am. J.
Physiol. 264, H163–H170
Peralta, C., Hotter, G., Closa, D., Gelpi, E., Bulbena, O.
and Rosello-Catafau, J. (1997) Protective effect of
preconditioning on the injury associated to hepatic
ischemia-reperfusion in the rat : role of nitric oxide and
adenosine. Hepatology 25, 934–937
McCully, J. D., Uematsu, M., Parker, R. A. and Levitsky,
S. (1998) Adenosine-enhanced ischemic preconditioning
provides enhanced cardioprotection in the aged heart.
Ann. Thorac. Surg. 66, 2037–2043
Maggirwar, S. B., Dhanraj, D. N., Somani, S. M. and
Ramkumar, V. (1994) Adenosine acts as an endogenous
activator of the cellular antioxidant defense system.
Biochem. Biophys. Res. Commun. 201, 508–515
Compeau, C. G., Ma, J., De Campos, K. N., Waddell,
T. K., Brisseau, G. F., Slutsky, A. S. and Rotstein, O. D.
(1994) In situ ischemia and hypoxia enhance alveolar
macrophage tissue factor expression. Am. J. Respir. Cell
Mol. Biol. 11, 446–455
Ardissino, D., Merlini, P. A., Ariens, R., Coppola, R.,
Bramucci, E. and Mannucci, P. M. (1997) Tissue factor
antigen and activity in human coronary atherosclerotic
plaques. Lancet 349, 769–771
Annex, B. H., Denning, S. M., Channon, K. M. et al.
(1995) Differential expression of tissue factor protein in
directional atherectomy specimens from patients with
stable and unstable coronary sydromes. Circulation 91,
619–622
Kaikita, K., Ogawa, H., Yasue, H. et al. (1997) Tissue
factor expression on macrophages in coronary plaques in
patients with unstable angina. Arterioscler. Thromb.
Vasc. Biol. 17, 2232–2237
Taubman, M. B., Fallon, J. T., Schecter, A. D. et al. (1997)
Tissue factor in the pathogenesis of atherosclerosis.
Thromb. Haemostasis 7, 200–204
Kobayashi, Y., Yoshimura, N., Nakamura, K., Yamagishi,
H. and Oka, T. (1998) Expression of tissue factor in
hepatic ischemic-reperfusion injury of the rat.
Transplantation 66, 708–716
Reverter, J. C., Beguin, S., Kessels, H., Kumar, R.,
Hemker, H. C. and Coller, B. S. (1996) Inhibition of
platelet-mediated, tissue thrombin generation by the
mouse\human chimeric 7E3 antibody. Potential
implication for the effect of c7E3 fab treatment on acute
thrombin and clinical retenosis. J. Clin. Invest. 98,
863–874
Smits, P., Williams, S. B., Lipson, D. E., Banitt, P.,
Rogen, G. A. and Creager, M. A. (1995) Endothelial
release of nitric oxide contributes to the vasodilator effect
of adenosine in humans. Circulation 92, 2135–2141
Peralta, C., Hotter, G., Closa, D. et al. (1999) The
protective role of adenosine in inducing nitric oxide
synthesis in rat liver ischaemia preconditioning is
mediated by activation of adenosine A2 receptors.
Hepatology 29, 126–132
Miller, K. J. and Hoffman, B. J. (1994) Adenosine A3
receptors regulate serotonin transport via nitric oxide and
cGMP. J. Biol. Chem. 269, 27351–27356
Yoshizumi, M., Perrella, M. A., Burnett, J. C. and Lee,
M. E. (1993) Tumor necrosis factor down-regulates an
endothelial nitric oxide synthase mRNA by shortening its
half-life. Circ. Res. 73, 205–209
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
Mcquillan, L. P., Leung, G. K., Marsden, P. A., Kostyk,
S. K. and Kourenbanas, S. (1994) Hypoxia inhibits
expression of eNOS via transcriptional and posttranscriptional mechanisms. Am. J. Physiol. 267,
H1921–H1927
Myers, P. R., Wright, T. F., Tannerm, M. A. and Adams,
H. R. (1992) EDRF and nitric oxide production in
cultured endothelial cells : direct inhibition by E. coli
endotoxin. Am. J. Physiol. 262, H710–H718
Liao, J. K., Shin, W. S., Lee, W. Y. and Clark, S. L. (1995)
Oxidized low-density lipoprotein decreases the expression
of endothelial nitric oxide synthase. J. Biol. Chem. 270,
319–342
Sessa, W. C., Pritchard, K., Seyedi, N., Wang, J. and
Hintze, T. H. (1994) Chronic exercise in dogs increases
coronary vascular nitric oxide production and endothelial
cell nitric oxide synthase gene expression. Circ. Res. 74,
349–353
Hishikawa, K., Nakaki, T., Marumo, T., Suzuki, H.,
Kato, R. and Saruta, T. (1995) Up-regulation of nitric
oxide synthase by estradiol in human aortic endothelial
cells. FEBS Lett. 369, 291–293
Pettersen, K. S., Wiiger, M. T., Narahara, N., Andoh, K.,
Gaudernack, G. and Prydz, H. (1992) Induction of tissue
factor synthesis in human umbilical vein endothelial cells
involves protein kinase C. Thromb. Haemostasis 67,
473–447
Ternisien, C., Ollivier, V., Khechai, F., Ramani, M.,
Hakim, J. and de Prost, D. (1995) Protein tyrosine kinase
activation is required for LPS and PMA induction of
tissue factor mRNA in human blood monocytes.
Thromb. Haemostasis 73, 413–420
Terry, C. M. and Callahan, K. S. (1996) Protein kinase C
regulates cytokine-induced tissue factor transcription and
procoagulant activity in human endothelial cells. J. Lab.
Clin. Med. 127, 81–93
Mackman, N. (1994) Protease inhibitors block
lipopolysaccharide induction of tissue factor gene
expression in human monocytic cells by preventing
activation of c-Rel\p65 heterodimers. J. Biol. Chem.
269, 26363–26367
Orthner, C. L., Rodgers, G. M. and Fitzgerald, L. A.
(1995) Pyrrolidine dithiocarbamate abrogates tissue factor
(TF) expression by endothelial cells : evidence implicating
nuclear factor-κB in TF induction by diverse agonists.
Blood 86, 436–443
Oeth, P. and Mackman, N. (1995) Salicylates inhibit
lipopolysaccharide-induced transcriptional activation of
the tissue factor gene in human monocytic cells. Blood 86,
4144–4152
Hobbs, A. J. and Ignarro, L. J. (1997) The nitric oxidecyclic GMP signal transduction system. In Nitric Oxide
and the Lung (Zapol, W. M. and Bloch, K. D. eds),
pp. 1–57, Marcel Dekker, New York
Ollivier, V., Parry, G. C. N., Cobb, R. R., de Prost, D.
and Mackman, N. (1996) Elevated cyclic AMP inhibits
NF-κB-mediated transcription in human monocytic cells
and endothelial cells. J. Biol. Chem. 271, 20828–20835
Kita, Y., Ohkubo, K., Hirasawa, Y. et al. (1995)
FR144420, a novel, slow, nitric oxide-releasing agent.
Eur. J. Pharmacol. 275, 125–130
Kita, Y., Suginoto, T., Hirasawa, Y., Yoshida, K. and
Maeda, K. (1994) Close correlation of the
cardioprotective effect of FK409, a spontaneous NO
releaser, with an increase in plasma cyclic GMP level.
Br. J. Pharmacol. 113, 5–6
Received 11 May 2001/2 August 2001; accepted 21 September 2001
# 2002 The Biochemical Society and the Medical Research Society
175