Báo cáo khoa học: Induction of PAI-1 expression by tumor necrosis factor a in endothelial cells is mediated by its responsive element located in the 4G/5G site ppt

Induction of PAI-1 expression by tumor necrosis factor a
in endothelial cells is mediated by its responsive element
located in the 4G/5G site
Maria Swiatkowska
1
, Janusz Szemraj
2
and Czeslaw S. Cierniewski
1,3
1 Department of Molecular and Medical Biophysics, Medical University in Lodz, Poland
2 Department of Biochemistry, Medical University in Lodz, Poland
3 Center of Medical Biology, Polish Academy of Sciences, Lodz, Poland
Plasminogen activator inhibitor type-1 (PAI-1) is
mainly identified as the primary physiological inhibitor
of both urokinase-type (uPA) and tissue-type (tPA)
plasminogen activators, and plays an important role in
regulation of the fibrinolytic system. Under normal
conditions, PAI-1 is present in plasma at low
concentrations, although high levels are seen in a vari-
ety of clinical settings [1]. PAI-1 is an early response
gene product known to be activated by numerous fac-
tors, including transforming growth factor b (TGFb)
and interleukin (IL)-1b [2], platelet-derived growth
factor and b fibroblast growth factor [3], thrombin [4],
Keywords
antioxidants; endothelium; gene regulation;
inflammation; reactive oxygen species
Correspondence
C. S. Cierniewski, Department of Medical
and Molecular Biophysics, Medical
University in Lodz, 6 ⁄ 8 Mazowiecka Street,

92-213 Lodz, Poland
Tel: +48 42 678 3393
E-mail:
(Received 25 April 2005, revised 2
September 2005, accepted 16 September
2005)
doi:10.1111/j.1742-4658.2005.04979.x
Plasminogen activator inhibitor type 1 (PAI-1) is induced by many proin-
flammatory and pro-oxidant factors. Among them, tumor necrosis factor a
(TNFa), a pivotal early mediator that regulates and amplifies the develop-
ment of inflammation, is one of the strongest PAI-1 synthesis activators.
Location of the TNFa response element in the PAI-1 promoter is still
ambiguous. In this study, we attempted to evaluate the significance of the
element located in the 4G ⁄ 5G site of the PAI-1 promoter in the TNFa sti-
mulation of PAI-1 expression in endothelial cells. PAI-1 expression was
monitored at: (a) the level of mRNA using real-time PCR, (b) PAI-1 gene
transcription by transfection reporter assays, and (c) protein synthesis using
the enzyme immunoassay. NF-jB activity was monitored using the elec-
trophoretic mobility shift assay. Its activity was modified by either anti-
sense oligonucleotides or transfection of endothelial cells with the wild-type
or mutated IjBa. We have shown that TNFa-induced expression and gene
transcription of PAI-1 involves a regulatory region present in segment
)664 ⁄ )680 of the PAI-1 promoter. This reaction involves the TNFa-
induced generation of superoxide leading to activation of NF-jB, and can
be abolished by antioxidants and by overexpression of a super-suppressor
phosphorylation-resistant IjBa. Stimulation of PAI-1 under these condi-
tions involves the motif of the PAI-1 promoter adjacent to the 4G ⁄ 5G site,
which can directly interact with NF-jB. We show that activation of PAI-1
gene by TNFa and reactive oxygen species is mediated by interaction of
NF-jB with the cis-acting element located in the )675 4G ⁄ 5G inser-

tion ⁄ deletion in the PAI-1 promoter.
Abbreviations
DCF, dichlorofluorescin; DCFH-DA, 2¢7¢-dichlorofluorescein diacetate; EMSA, electrophoretic mobility shift assay; H
2
O
2
, hydrogen peroxide;
HUVEC, human umbilical vein endothelial cells; IKK, IjB kinase; IL, interleukin; NAC, N-acetylcysteine; O
2
–
, superoxide anion; PAI-1,
plasminogen activator inhibitor-1; PEG, poly(ethylene glycol); ROS, reactive oxygen species; TGFb, transforming growth factor b; TNFa,
tumor necrosis factor a.
FEBS Journal 272 (2005) 5821–5831 ª 2005 The Authors Journal Compilation ª 2005 FEBS 5821
tumor necrosis factor a (TNF a) [5], insulin [6], angio-
tensin II [7] and oxidation products [8].
PAI-1 is inherently unstable and readily converts
from an active to a latent form [9,10]. Thus, self-inacti-
vation of PAI-1 is a crucial regulatory mechanism, by
which this protein functions in circulation. Another
mechanism of PAI-1 regulation results from the tran-
scriptional control of its expression. Several regulatory
elements have been localized in the human PAI-1
promoter and include two Sp1 elements ()73 and
)42 bp) mediating glucose responsiveness [11], a
hypoxia-responsive element ()194 bp) [12], a very low-
density lipoprotein-responsive site ()672 ⁄ )657 bp) [13],
SMAD 3 and -4 protein-binding sites that mediate
TGFb responsiveness ()730, )580, and )280 bp) [14],
oxidative stress and thymosin b4 responsive AP-1 site

at )60 ⁄ 52 [15,16], and a 5¢ distal TNFa responsive
enhancer of the PAI-1 gene located 15 kb upstream
of the transcription start site containing a conserved
NF-jB-binding site [17].
In this report, we provide evidence that activation of
PAI-1 gene by TNFa and reactive oxygen species
(ROS) is mediated by interaction of NF-jB with the
cis-acting element located in the )675 4G ⁄ 5G inser-
tion ⁄ deletion in the PAI-1 promoter. Because TNFa is
a pivotal early mediator that regulates and amplifies
the development of inflammation, this mechanism can
be used primarily under diverse inflammatory condi-
tions, including ischemia, traumatic injury, allograft
rejection, cytokine stimulation and activation by bac-
terial components [18,19]. TNFa is known to induce
the generation of ROS [20], which constitute primary
signals transduced into the cytoplasm and ultimately
alter the expression of specific genes [21–23]. Because
ligand-stimulated NF-jB activation can be blocked by
antioxidants, it appears that the generation of ROS
may be involved in the induction of PAI-1 expression
by TNFa via the activation of NF-jB.
Results
Upregulation of PAI-1 expression in endothelial
cells by ROS
In preliminary experiments, we evaluated the role of
ROS during the stimulation of PAI-1 expression in
endothelial cells by TNFa. For this purpose, endothel-
ial cells were treated with 10 mm N-acetylcysteine
(NAC) for 30 min, followed by incubation with TNFa

(50 ngÆmL
)1
) and the released PAI-1 antigen was deter-
mined by ELISA. Figure 1A shows that the treatment
of endothelial cells with TNFa increased PAI-1 antigen
by almost twofold (180 ± 10%, P < 0.05). Cotreat-
ment with NAC inhibited TNFa-induced PAI-1 antigen
accumulation by 48 ± 7% (P<0.05). Similarly,
incubation of endothelial cells with 100 or 200 lm
H
2
O
2
for 30 min increased PAI-1 by 160 ± 4 and
180 ± 8%, respectively (P<0.05 for both). Also, in
this case, cotreatment with NAC inhibited H
2
O
2
-
induced PAI-1 antigen release by 38 ± 3 and 26 ± 3%
observed at 100 and 200 lm H
2
O
2
, respectively
(P<0.05 for both).
A
B
Fig. 1. Effect of TNFa and H

2
O
2
on PAI-1 expression in vascular
endothelial cells. Expression of PAI-1 was analyzed at the level of
protein synthesis (A) in the presence or absence of NAC (10 m
M).
The PAI-1 antigen was determined using the ELISA test. (B) PAI-1
mRNA expression analyzed by real-time PCR in endothelial cells
induced by TNFa in the absence or presence of different antioxi-
dants. For this purpose, endothelial cells were preincubated for
30 min with NAC (10 m
M), catalase (500 UÆmL
)1
), pyrrolidine dithio-
carbamate (PDTC) (100 l
M), or vitamin C (100 lM) and stimulated
with TNFa (50 ngÆmL
)1
) for 4 h. Data are shown as mean ± SD
obtained during three separate experiments.
The 4G ⁄ 5G site and PAI-1 expression M. Swiatkowska et al.
5822 FEBS Journal 272 (2005) 5821–5831 ª 2005 The Authors Journal Compilation ª 2005 FEBS
To determine whether the changes in PAI-1 antigen
release were due to modulation of PAI-1 mRNA
expression, we performed real-time PCR using endo-
thelial cells treated with TNFa in the presence or
absence of NAC. In addition to NAC, other antioxi-
dants, such as PDTC (100 lm), catalase (500 UÆmL
)1

)
or vitamin C (100 lm) attenuated TNFa-induced
PAI-1 mRNA expression (Fig. 1B).
NF-jB regulates PAI-1 gene transcription
Under basal culture conditions, endothelial cells exhibit
little or no dichlorofluorescin (DCF) fluorescence.
Treatment for 10 min with TNFa used at 50 ngÆmL
)1
increased DCF fluorescence, which was attenuated by
cotreatment with NAC (10 mm) (Fig. 2A,B). Similarly,
incubation of endothelial cells with 100 lm H
2
O
2
for
10 min increased DCF fluorescence, which was then
suppressed by cotreatment with NAC. These findings
indicate that both TNFa and H
2
O
2
increased intracellu-
lar oxidative stress, which was then partially abolished
by the antioxidant NAC. Figure 2C shows the effect
of poly(ethylene glycol) (PEG) and PEG-catalase on
endothelial cells stimulated with TNFa. Results were
obtained after 2 h of TNFa stimulation in the presence
of PEG alone or PEG-catalase (100–1000 UÆmL
)1
).

Experiments were performed three times with similar
results.
To determine whether TNFa-induced PAI-1 expres-
sion involves the activation of NF-jB, we performed
electrophoretic mobility shift assay (EMSA) studies
using the consensus oligonucleotide for NF-jB
(Fig. 3A), and PCR product containing wild-type
sequence from )664 to ) 680 of PAI-1 promoter
(Fig. 3B), 17-bp fragment ()664 to )680) of the PAI-1
promoter containing the putative jB-binding site
(Fig. 3C) [24]. The treatment of endothelial cells with
TNFa or H
2
O
2
resulted in NF-jB activation which
was inhibited in competition experiment using either
the jB consensus oligonucleotide or fragment )664 to
)680 of PAI-1 promoter containing the jB-binding site
but not by mutated fragment )664 to )680 of PAI-1
promoter. NAC abolished NF-jB activation, indica-
ting that TNFa- and H
2
O
2
-induced activation of
NF-jB involves the generation of ROS. The identity
of NF-jB present in the nuclear complex formed with
the NF-jB consensus oligonucleotide was proved by
supershift produced with anti-p65 sera (Fig. 3C). The

effect of oxidative stress-induced NF-jB activation in
TNFa-activated cells was confirmed using an antioxid-
ant NAC. The cells were transfected with pNF-jB-
SEAP plasmid and treated with TNFa in the presence
or absence of NAC. NF-jB promoter activation by
TNFa was inhibited by antioxidant, NAC (Fig. 3D).
To confirm the role of fragment ()664 to )680) of
PAI-1 promoter in this reaction, endothelial cells were
analyzed after transfection with either wild-type PAI-1
promoter or its mutated version. Mutation of the
NF-jB putative binding site within the PAI-1 promo-
ter abolished its sensitivity to induction by TNFa
(Fig. 3E).
To further evidence whether NF-jB activation is
required for TNFa- and H
2
O
2
-induced PAI-1 expres-
sion, in subsequent experiments NF-jB expression was
downregulated with antisense oligodeoxynucleotide
(5¢-GGGGAACAGTTCGTCCATGGC-3¢) that is spe-
cific to the NF-jB subunit, RelA p65. In parallel, endo-
thelial cells incubated with sense oligonucleotide were
used as a control. In contrast to the sense RelA oligonu-
cleotide, the addition of the antisense RelA oligonucleo-
tide to endothelial cells resulted in strong inhibition of
both TNFa- and H
2
O

2
-induced PAI-1 mRNA expres-
sion (Fig. 4A) and promoter activity (Fig. 4B), proving
that NF-jB is required for both stimulated pathways.
The efficacy of antisense oligonucleotides to reduce the
expression of NF-jB is shown in Fig. 4C.
Effects of NF-jB inhibition on PAI-1 expression
The NF-jB complex is retained in the cytoplasm by
IjB proteins. Activation of NF-jB involves its phos-
phorylation by IjB kinase (IKK) and subsequent
degradation of IjB by 26S proteasome. To inhibit
NF-jB, we treated endothelial cells with a relatively
specific IjBa 26S proteasome inhibitor, MG132 used
in the concentration range 0–1000 nm to prevent
degradation of IjBa. In the presence of MG132,
TNFa-induced expression of PAI-1 detected at the
level of mRNA (Fig. 5C) or protein synthesis (Fig. 5A)
was blocked in a concentration-dependent manner.
Similarly, the stimulating effect of TNFa was abol-
ished by salicylate, which is an IKK inhibitor
(Fig. 5B,D). The data are consistent with observations
presenting the effect of TNFa on PAI-1 promoter
activity (Fig. 5E) and mRNA expression (Fig. 5F) tes-
ted in endothelial cells transfected either with empty
vector (pCMV4), or containing wild-type IjBa (WT)
or mutated IjBa (MT). Transfection with pCMV4 or
WT-IjBa had little or no effect on TNFa-induced
PAI-1 mRNA expression or promoter activity. How-
ever, overexpression of MT-IjBa, which cannot be
phosphorylated at Ser32 and Ser36, resulted in a sub-

stantial decrease in TNFa-induced PAI-1 expression
and promoter activity. These findings indicate that
signaling pathways proximal to IjBa phosphorylation
(i.e. IKK) are potential targets for activation by ROS.
M. Swiatkowska et al. The 4G ⁄ 5G site and PAI-1 expression
FEBS Journal 272 (2005) 5821–5831 ª 2005 The Authors Journal Compilation ª 2005 FEBS 5823
Discussion
In this study, we have shown that TNFa-induced
expression and gene transcription of PAI-1 involve a
regulatory region present in segment )664 ⁄ )680 of the
PAI-1 promoter, corresponding to the previously
reported IL-1a-inducible site located between )675 and
)669 [24]. This is supported by the following observa-
tions: (a) the oligonucleotide containing the putative
jB-binding site of PAI-1 promoter ()664 ⁄ )680) can
bind transcriptional factor subunits p50 and p65. This
binding is specific and can be abolished by triple muta-
tion of the oligonucleotide, as seen both in the direct
binding and during competitive inhibition experiments.
(b) Mutation of the regulatory region abolished its
responsiveness in the PAI-1 promoter to both TNFa
and ROS, as demonstrated after the transfection of
cells with p800Luc or its mutated version, respectively.
Furthermore, responsiveness of this element to activa-
tion of endothelial cells by TNFa or ROS was also
A
C
Fig. 2. (A, B) The effect of TNFa
(50 ngÆmL
)1

) on intracellular oxidation ana-
lyzed by DCF fluorescence in vascular endo-
thelial cells with and without of NAC
(10 m
M). (C) The effect of PEG and PEG-
catalase on endothelial cells stimulated with
TNFa. Results were obtained after 2 h of
TNFa stimulation in the presence of PEG
alone or PEG-catalase (100–1000 UÆmL
)1
).
Experiments were performed three times
with similar results.
The 4G ⁄ 5G site and PAI-1 expression M. Swiatkowska et al.
5824 FEBS Journal 272 (2005) 5821–5831 ª 2005 The Authors Journal Compilation ª 2005 FEBS
demonstrated by EMSA experiments. Interestingly,
this regulatory element is adjacent to the polymor-
phic locus, )675 4G ⁄ 5G insertion ⁄ deletion, which has
stimulated increased interest in PAI-1 as a risk factor
for venous thrombosis [25]. In particular, the 4G allele
in this polymorphism has been associated with higher
TNFα
H
2
O
2
[µΜ]
NAC
TNFα
H

2
O
2

TNFα
TNFα
NAC
0
15
30
45
0
60
40
20
p-800Luc
p-800Luc-(664-680)
Induction (%)
80
140
A
B
C
D
E
120
100
NAC
4G/5G
Consensus

Mutant
A-p65
NF-κB-SEAP activity (% of control)
100 200 100 200 100
200
100 200
Fig. 3. Increased expression of NF-jB in endothelial cells with upregulated PAI-1 upon treatment with TNFa. EMSAs using oligonucleotides
derived from consensus sequence of NF-jB (A) and containing the putative of jB binding site of PAI-1 promoter (B). Endothelial cells were
stimulated with TNFa (50 ngÆmL
)1
)orH
2
O
2
(100 and 200 lM) in the presence of NAC (10 mM). (C) EMSAs using oligonucleotides
(ACGTGGGGGAGTCAGCC) containing putative of jB binding site of PAI-1 promoter. Endothelial cells were stimulated with TNFa
(50 ngÆmL
)1
)orH
2
O
2
(200 lM) in the presence or absence of NAC (10 mM). In the competition experiment, an excess amount unlabeled
oligonucleotide containing the putative NF-jB binding site of PAI-1 promoter or unlabeled mutanted oligonucleotide or unlabeled consensus
of NF-jB oligonucleotide was added to binding system 10 min prior to adding labeled oligonucleotide. The nuclear complex produced with
the NF-jB probe is supershifted by the p65 antibody. Endothelial cells were transfected with pNF-jB-SEAP in the presence of the pSEAP as
a control vector. The cells were preincubated with antioxidant NAC (10 m
M) for 30 min and then treated with TNFa (50 ngÆmL
)1
). Secreted

alkaline phosphatase (SEAP) activities in the cells were determined by Chemiluminescence Detection Kit (D). (E) A mutation of the NF-jB
putative binding site ()664 to )680) within PAI-1 promoter abolishes its sensitivity to TNFa-induced transcriptional activity. In this experiment
endothelial cells were transfected either with p800Luc or its mutated version, in which the )664 to )680 fragment (ACGTGGGGGAGT
CAGCC) was substituted by the mutated sequence (ACATGGGCCAGTCAGCC). Mutation of the wild-type sequence was done on a PAI-1 pro-
moter cloned into pLuc vector. Transfected cells were treated with TNFa (50 ngÆmL
)1
) for 24 h and the cells harvested 12 h later. Luciferase
activity was determined by the Dual-Luciferase assay kit. Data are shown as the mean of three separate experiments ± SD.
M. Swiatkowska et al. The 4G ⁄ 5G site and PAI-1 expression
FEBS Journal 272 (2005) 5821–5831 ª 2005 The Authors Journal Compilation ª 2005 FEBS 5825
levels of plasma PAI-1 antigen and PAI-1 activity [26].
Furthermore, in the presence of the 4G allele not only
was the PAI-1 response more pronounced, but so too
was the response of other acute-phase reactants, which
implies that the increases in these reactants are secon-
dary to the increase in PAI-1 [27].
Our data show that activation of endothelial cells
with TNF a to produce PAI-1 is mediated by a ROS-
stimulated increase in NF-jB activity. Treatment with
H
2
O
2
increased PAI-1 expression and the effect was
reduced by cotreatment with the antioxidant, NAC.
Indeed, direct inhibition of NF-jB, either by antisense
oligonucleotide or by overexpression of IjBa, attenu-
ated both TNFa- and H
2
O

2
-induced PAI-1 expression.
TNFa was reported to induce both superoxide anion
and H
2
O
2
production [28,29]. Because TNFa and ROS
have similar effects on PAI-1 expression, these findings
suggest a common mechanism by which these signaling
pathways lead to the induction of PAI-1 in endothelial
cells. Interestingly, PAI-1 upregulation by TGF-b has
also been shown to involve ROS production in mesan-
gial cells [30]. Furthermore, high glucose [31] and
cyclic strain [32] can upregulate PAI-1 expression in
endothelial cells through ROS generation. Indeed,
PAI-1 is regulated by a redox-sensitive mechanism
after the exposure to ionizing radiation in renal tubu-
lar epithelial cells [33]. In addition, IL-1-mediated
upregulation of PAI-1 expression in cardiac micro-
vascular endothelial cells also appears to be ROS
dependent [34].
In this study, we also characterized the inhibitory
mechanism produced by antioxidants on TNFa-
induced PAI-1 expression in endothelial cells. Our
findings indicate that this inhibition occurs via the
suppression of NF-jB. Although TNFa could activate
NF-jB by a ROS-independent mechanism, the ability
of ROS to stimulate NF-jB activity may provide a
synergistic effect in this activation [35]. TNFa is

known to mediate the PAI-1 acute phase response
in vivo and a jB-like sequence has been found in a
number of promoters of acute phase-regulated genes
[36]. Involvement of the NF-jB signaling pathway in
PAI-1 upregulation has been reported in human endo-
thelial cells stimulated by lipopolysaccharide [37]
and in proximal tubular cells exposed to uremic tox-
ins [38]. NF-jB also contributes to Chlamydia pneu-
moniae-induced overexpression of PAI-1 in vascular
smooth muscle cells and endothelial cells [39].
Furthermore, NF-jB-like binding site in the PAI-1
promoter is responsible for IL-1-induced PAI-1
expression and this element is involved in the control
of plasma levels of PAI-1 [24]. Thus, multiple cyto-
kines and infectious agents can upregulate PAI-1
A
B
C
Fig. 4. Effect of oligonucleotides antisense to RelA p65 on PAI-1
mRNA expression and promoter activity in endothelial cells. Endo-
thelial cells were preincubated with the sense or antisense oligo-
nucleotides and then treated with TNFa (50 ngÆmL
)1
)orH
2
O
2
(100 lM). (A) Expression of PAI-1 mRNA determined by real-time
PCR and (B) activity of PAI-1 promoter. (C) Treatment of endothelial
cells with the oligonucleotide antisense to RelA p65 abolished for-

mation of the nuclear complex with the labeled probe containing
NF-jB consensus sequence, as determined by EMSA.
The 4G ⁄ 5G site and PAI-1 expression M. Swiatkowska et al.
5826 FEBS Journal 272 (2005) 5821–5831 ª 2005 The Authors Journal Compilation ª 2005 FEBS
expression, leading to altered vascular hemostasis and
proliferation.
Experimental procedures
Reagents
All standard tissue culture reagents including M199 and
fetal bovine serum were obtained from Gibco-BRL (Beth-
esda, MD, USA). TNFa were obtained from Promega
(Madison, WI, USA). All other reagents were obtained
from Sigma (St. Louis, MO, USA), unless indicated other-
wise. The inhibitor Z-Leu-Leu-Leu-CHO (MG132) was
from BioMol (Plymouth Meeting, PA, USA). The chemilu-
minescence detection reagent for western blotting was from
Pierce (Rockford, IL, USA). The protein determination
assay reagent, acrylamide, TEMED, and ammonium per-
sulfate were from Bio-Rad (Hercules, CA, USA). Lipo-
fectAMINE Plus reagent was obtained from Gibco-BRL.
Plasmid p800LUC with PAI-1 promoter was obtained as a
gift from D. J. Loskutoff (Scripps Institute, La Jolla, CA,
USA). Plasmids with wild-type IjBa (WT) and mutated
IjBa (MT) were obtained as a gift from J. K. Liao (Har-
vard Medical School, Boston, MA, USA). [
32
P]dATP[cP]
(6000 CiÆmmol
)1
) and T7 Sequenase (v.2.0) were purchased

from Amersham (Piscataway, NJ). Protein assay reagents
and polyacrylamide gel chemicals were from Bio-Rad.
P-NF-jB-SEAP vector and SEAP Chemiluminescence
Detection Kit was from BD Bioscience Clontech (Mountain
View, CA, USA).
Cell culture
Human umbilical vein endothelial cells (HUVEC) were cul-
tured in growth medium containing M199 and 20% (v ⁄ v)
fetal bovine serum in a 90–95% humidified atmosphere of
5% (v ⁄ v) CO
2
at 37 °C. To evaluate the effect of different
compounds on PAI-1 expression at the level of protein syn-
thesis, cells were grown on 48-well microplates. To deter-
mine PAI-1 mRNA levels, endothelial cells were cultured in
A
B
C
D
F
E
Fig. 5. Dependence of PAI-1 expression
upon NF-jB activation. (A, B) Inhibition of
PAI-1 antigen expression produced by incu-
bation of TNFa-stimulated endothelial cells
with increasing concentrations of 26S pro-
teasome inhibitor, MG132 and sodium sali-
cylate, respectively. (C, D) Similar reduction
in PAI-1 mRNA expression in the same cells
as analyzed by real-time PCR. All experi-

ments were performed at least 3 times with
reproducible results and data are shown as
mean ± SD. (E, F) Changes in PAI-1 expres-
sion as measured at the level of PAI-1 pro-
moter activity and PAI-1 mRNA evaluated
by real-time PCR, respectively. In these
experiments, endothelial cells were trans-
fected with the empty vector (pCMV4), and
the vector expressing either a wild-type
IjBa (WT) or its nonphosphorylable mutant
IjBa (MT). Results were standardized to
cotransfection RSV-b-Gal expression plasmid
(internal control). All data represent the
mean of three separate experiments ± SD.
M. Swiatkowska et al. The 4G ⁄ 5G site and PAI-1 expression
FEBS Journal 272 (2005) 5821–5831 ª 2005 The Authors Journal Compilation ª 2005 FEBS 5827
10 cm dishes. In all experiments, cellular viability was
assessed by Trypan Blue dye-excluding cells. Relatively pure
(> 95%) vascular endothelial cells were confirmed by their
morphological features using phase-contact microscopy and
immunofluorescent staining with PECAM-1 antibody (data
not shown). At the concentrations used, there were no
observable adverse effects of TNFa, superoxide dismutase,
catalase or other antioxidants on cellular viability for all
treatment conditions.
Treatment conditions
Prior to treatment with TNFa, endothelial cells were
starved for  12 h in M199 supplemented with 0.1% (v ⁄ v)
fetal bovine serum. Endothelial cells were then stimulated
with TNFa (50 ngÆmL

)1
)orH
2
O
2
(100 and 200 lm)in
the presence and absence of NAC (10 mm), pyrroli-
dinedithiocarbamate (100 lm), vitamin C (100 lm) and
catalase (500 UÆmL
)1
). In some experiments, the 26S pro-
teasome inhibitor, MG132 (10 nm to lm), was added
30 min prior to TNFa stimulation. Sodium salicylate
(2 and 4 mm) was added 1 h before the TNFa stimulation.
Real-time quantitative RT-PCR
Total RNA (1 lg) was extracted from endothelial cells
using Trizol reagent (Life Technologies Inc, Rockville, MD,
USA) and processed directly to cDNA synthesis using the
TaqMan Reverse Transcription Reagents kit (Applied Bio-
systems, Foster City, CA, USA), according to the manufac-
turer’s protocol. The PAI-1 and b-actin expression was
quantified by real-time RT-PCR using ABI Prism 7000
Sequence Detection System (Applied Biosystems), according
to the manufacturer’s protocol. Briefly, 2.5, 2.0; 1.5, 1.0;
0.5 and 0.25 lL of synthesized cDNA were amplified in
triplicate for both b-actin and each of the target genes to
create a standard curve. Likewise, 2 lL of cDNA was
amplified in triplicate in all isolated samples for each pri-
mer ⁄ probe combination and b-actin. Each sample was sup-
plemented with both respective 0.3 lm forward and reverse

primers, fluorescent probe, and made up to 50 lL using
qPCR
TM
Mastermix for SYBR Green I (Eurogentec, Sera-
ing, Belgium). All the following PCR primers were designed
using software primerexpress (Applied Biosystems)
forward 5¢-TGCTGGTGAATGCCCTCTACT-3¢, reverse
5¢-CGGTCATTCCCAGGTTCTCTA-3¢ forward 5¢-CGTA
CCACTGGCATCGTGAT-3¢, reverse 5¢-GTGTTGGCGT
ACAGGTCTTTG-3¢ specific for mRNAs of PAI-1 and
b-actin, respectively. b-Actin was used as an active and
endogenous reference to correct for differences in the
amount of total RNA added to the reaction and to com-
pensate for different levels of inhibition during reverse tran-
scription of RNA and during PCR. Each target probe was
amplified in a separate 96-well plate. All samples were incu-
bated at 50 °C for 2 min and at 95 °C for 10 min, and then
cycled at 95 °C for 30 s, 56 °C for 1 min and 72 °C for
1 min for 40 cycles. SYBR Green I fluorescence emission
data were captured and mRNA levels were quantified using
the critical threshold (C
t
) value. Analyses were performed
with abi prism 7000 (SDS Software). Controls without
reverse transcription and with no template cDNA were per-
formed with each assay. To compensate for variations in
input RNA amounts, and efficiency of reverse transcription,
b-actin mRNA was quantified and results were normalized
to these values. Relative gene expression levels were
obtained using the DC

t
method. Amplification specific tran-
scripts were further confirmed by obtaining melting curve
profiles.
Assay of intracellular oxidative stress
HUVEC of fewer than three passages were cultured in
35 mm dishes (Corning, NY, USA) coated with 0.1%
(w ⁄ v) gelatin. Phenol-free M199 medium + 15 mm Hepes
(pH 7.4) were used. Before seeding the endothelial cells, a
sterile cover-slip was placed on the bottom of each dish.
To subtract background fluorescent activity from intracel-
lular fluorescence, these cover-slips were eliminated before
measurement in order to have a cell-free control field. The
preincubation period with antioxidants was 30 min. Intra-
cellular generation of ROS was quantified using 2¢7¢ -
dichlorofluorescein diacetate (DCFH-DA) (Molecular
Probes, Eugene, OR, USA). This esterified form is cell
membrane permeable and undergoes deacetylation by
intracellular esterases. Upon oxidation, DCFH is conver-
ted to dichlorofluorescin (DCF), a fluorescent compound.
Confluent endothelial cell monolayers were incubated
30 min with 30 lm DCFH-DA before stimulation with
TNFa. Briefly, PEG and PEG-catalase were dissolved in
phenol-free culture medium and applied to endothelial
cells 1 h before stimulation with TNFa. Results were
obtained after 2 h of TNFa stimulation in the presence of
PEG alone or PEG-catalase (100–1000 UÆmL
)1
). Fluores-
cence was monitored using an inverted microscope (Zeiss

Axiovert 405M, Oberkochen, Germany) with a specimen
stage for 35 mm dishes. This custom-design stage was
equipped with a temperature and gas control, thus allow-
ing incubator conditions [i.e. 5% (v ⁄ v) CO
2
,37°C] under
the microscope. Culture medium pH was 7.4 under micro-
scopy conditions. A mercury lamp with a 490 nm filter
was used as a light source for excitation. Excitation time
(3 s) was constant for all conditions. The emission wave-
length was set to 525 nm. Images were acquired using a
CCD camera (Photometrics CH 250, Tucson, CA, USA)
with a 512 · 512 pixel format. Digitalized images were
transferred to a Sun SPARC workstation IPX (Sun
Microsystems, Mountain View, CA, USA). Analysis was
performed with isee software v. 3.6 (Inovision, Durham,
NC, USA). For each condition, data were acquired from
six different and representative fields (three separate
The 4G ⁄ 5G site and PAI-1 expression M. Swiatkowska et al.
5828 FEBS Journal 272 (2005) 5821–5831 ª 2005 The Authors Journal Compilation ª 2005 FEBS
regions of interest within two separate microscopic fields).
Fluorescence intensity (excitation wavelength 490 nm;
emission wave length 520 nm; excitation time 3 s) was
recorded on a gray scale from 0 to 16 384. Background
fluorescence (cell-free area) was subtracted from total
fluorescent intensity.
EMSA
Confluent HUVECs were stimulated with TNFa
(50 ngÆmL
)1

) for 2 h. The double-stranded oligonucleo-
tides: (a) transcription factor consensus oligonucleotide for
NF-jB (AGTTGAGGGCACTTTCCCAGG) obtained
from Integrated DNA Technologies, Inc (Coralville, IA,
USA), (b) PCR products containing wild-type sequence
from )664 to )680 of PAI-1 promoter (ACGTGGGGG
AGTCAGCC), and (c) an oligonucleotide from the
sequence )664 to )680 of PAI-1 promoter (ACGTGGGG
GAGTCAGCC) were used as labeled probe. The introduc-
tion the mutations to wild-type sequence from )664 to
)680 of PAI-1 promoter cloned into pLUC vector was
carried out using a Quick Change site-directed mutagenesis
kit (Stratagene, La Jolla, CA, USA) with mutagenic prim-
ers: 5¢-ACATGGGGGAGTCAGCC-3, 5¢-GGCTGACTCC
CCCATGT-3¢;5¢-ACATGGGGCAGTCAGCC-3, 5¢-GG
CTGACTGCCCCATGT-3¢;5¢-ACATGGGCCAGTCAG
CC-3, 5¢-GGCTGACTGGCCCATGT-3¢. The PCR prod-
uct and the oligonucleotides were labeled to high specific
activity with T4 polynucleotide kinase (Promega) using
[
32
P]dATP[cP] (Amersham Biotech) 37 °C for 10 min and
subsequently purified by electrophoresis in 7% polyacryla-
mide gels using 0.5· Tris ⁄ borate ⁄ EDTA, pH 7.8. We used
6–7 · 10
4
c.p.m. for EMSA. Binding reactions were per-
formed in the binding buffer 20 mm Hepes ⁄ KOH, pH 7.5,
32 mm KCL, 5 mm MgCl
2

,1mm dithiothreitol, 10% (v ⁄ v)
glycerol) in the presence of nonspecific competitor
poly(dI:dC) (Amersham Biotech). Crude nuclear extracts
(15 lg) from control cells and stimulated with TNFa
(50 ngÆmL
)1
) cells were incubated with 0.01 pmol of
c
32
P-labeled oligonucleotides and PCR products for
20 min at room temperature in a total volume of 20 lL.
For competition experiments, 200-fold molar excess of un-
labeled double-stranded competitor oligonucleotides
(consensus of NF-jB, the sequence from )664 to )680 of
PAI-1 promoter, mutated fragment )664 to )680 of PAI-1
promoter) were added to the reaction mixtures. To identify
a transcription factor, 4 lg of polyclonal antibodies
against NF-j B subunits (p50, p65) were introduced to the
binding reactions for 30 min prior to the addition of the
radioactive probe. DNA–protein complexes were resolved
from unbound oligonucleotides and PCR products by
elecrophoresis on 5% polyacrylamide gel using 0.5· Tris ⁄
borate ⁄ EDTA buffer (150 V for 2 h). Gels were vacuum-
dried and autoradiographed with intensifying screens for
2–24 h at )20 °C.
Antisense oligonucleotide transfection
To transfect antisense oligonucleotides to RelA, cells were
cultured in six-well plates and grown in M199 containing
20% (v ⁄ v) fetal bovine serum. After 24 h, wells were
washed with serum- and antibiotic-free medium. Phos-

phorothioate oligonucleotides (200 nm) and the oligofecta-
mine reagent (Invitrogen, Life Technologies) were added to
the wells. Cells were incubated for 20 h with sense (5¢-GC
CATGGACGAACTGTTCCCC-3¢) or antisense (5¢-GGG
GAACAGTTCGTCCATGGC-3¢) oligonucleotides and
then treated for additional 4 h with TNFa or H
2
O
2
and
total cellular RNA was isolated by the TRIzol Reagent.
Plasmids construcions
Introduction of mutations to the wild-type sequence from
)664 to )680 of PAI-1 promoter cloned into pLUC vector
was carried out using a Quick Change site-directed muta-
genesis kit (Stratagene) and mutagenic primers: 5¢-AC
ATGGGGGAGTCAGCC-3¢,5¢-GGCTGACTCCCCCAT
GT-3¢;5¢-ACATGGGGCAGTCAGCC-3¢,5¢-GGCTGA
CTGCCCCATGT-3¢;5¢-ACATGGGCCAGTCAGCC-3¢,
5¢-GGCTGACTGGCCCATGT-3¢. The nickled vector
DNA incorporating the desired mutations was than trans-
formed into XL1-Blue cells.
Transfection reporter assays
Semiconfluent cell cultures in six-well tissue culture plates
were transfected with DNA constructs (plasmid p800LUC
with the PAI-1 promoter, p800LUC
mut
containing mutated
sequence from )664 to )680 of the PAI-1 promoter and
pCMV.IjBa expression vector) or with pNF-jB-SEAP vec-

tor. The cells were grown in six wells with a density of
 1 · 10
5
cellsÆmL
)1
and transfected using 1 lg of the plas-
mid DNA and lipoficatmine (Gibco BRL) according to
manufacturer’s instructions. As an internal control for
transfection efficiency, pRSV.bGAL plasmid (0.5 lg) was
cotransfected in all experiments. In parallel experiments,
endothelial cells were transfected with luciferase reporter
vector pGL3 (Promega) and used as control cells to test
whether the effect of inhibitors was specific for the PAI-1
promoter.
Overexpression of IjBa
The following expression plasmids were used: (a) WT IjBa,
encoding the full-length protein, was N-terminus FLAG-
tagged in pCMV4. (b) MT IjBa, lacking the serine
phosphorylation sites (32 and 36) and thus resistant to
degradation by the 26S proteasome, was also N-terminus
FLAG-tagged in pCMV4. Endothelial cells were transfected
with IjBa expression plasmids by lipofectamine method.
M. Swiatkowska et al. The 4G ⁄ 5G site and PAI-1 expression
FEBS Journal 272 (2005) 5821–5831 ª 2005 The Authors Journal Compilation ª 2005 FEBS 5829
After 48 h cells were treated with TNFa (50 ngÆmL
)1
) and
total cellular RNA isolated 4 h later by using the TRIzol
Reagent.
Statistical analysis

All values are expressed as mean ± SE compared with con-
trols and among separate experiments. Paired and unpaired
Student’s t-tests were employed to determine the signifi-
cance of changes. A P-value < 0.05 was considered statisti-
cally significant.
Acknowledgements
This work was supported by the Polish Committee for
Scientific Research (KBN) Grant no. 3 PO4A 01425.
References
1 Horrevoet AJ (2004) Plasminogen activator inhibitor 1
(PAI-1): in vitro activities and clinical relevance. Br J
Haematol 125, 12–23.
2 Sato Y, Tsuboi R, Lyons R, Moses H & Rifkin DB
(1990) Characterization of the activation of latent TGF-
beta by co-cultures of endothelial cells and pericytes or
smooth muscle cells: a self-regulating system. J Cell Biol
111, 757–763.
3 Lau HK (1999) Regulation of proteolytic enzymes and
inhibitors in two smooth muscle cell phenotypes. Cardi-
ovasc Res 43, 1049–1059.
4 Cockell KA, Ren S, Sun J, Angel A & Shen GX (1995)
Effect of thrombin on the release of plasminogen activa-
tor inhibitor-1 from cultured primate arterial smooth
muscle cells. Thromb Res 77, 119–131.
5 Samad F, Yamamoto K & Loskutoff DJ (1996) Distri-
bution and regulation of plasminogen activator inhibi-
tor-1 in murine adipose tissue in vivo. Induction by
tumor necrosis factor-alpha and lipopolysaccharide.
J Clin Invest 97, 37–46.
6 Nordt T, Schneider D & Sobel B (1994) Augmentation

of the synthesis of plasminogen activator inhibitor type-1
by precursors of insulin. A potential risk factor for vas-
cular disease. Circulation 89, 321–330.
7 Brown NJ, Kim KS, Chen YQ, Blevins LS, Nadeau JH,
Meranze SG & Vaughan DE (2000) Synergistic effect of
adrenal steroids and angiotensin II on plasminogen acti-
vator inhibitor-1 production. J Clin Endocrinol Metab
85, 336–344.
8 Dichtl W, Stiko A, Eriksson P, Goncalves I, Calara F,
Banfi C, Ares MP, Hamsten A & Nilsson J (1999) Oxi-
dized LDL and lysophosphatidylcholine stimulate plas-
minogen activator inhibitor-1 expression in vascular
smooth muscle cells. Arterioscler Thromb Vasc Biol 19 ,
3025–3032.
9 Mottonen J, Strand A, Symersky J, Sweet RM, Danley
DE, Geogheggan KF, Gerard RD & Goldsmith EJ,
(1992) Structural basis of latency in plasminogen activa-
tor inhibitor-1. Nature 355, 270–273.
10 Lawrence DA, Olson ST, Palaniappani S & Ginsburg D
(1994) Serpin reactive center loop mobility is required
for inhibitor function but not for enzyme recognition.
Biochemistry 33, 3643–3648.
11 Chen YQ, Su M, Walia RR, Hao Q, Covington JW &
Vaughan DE (1998) Sp1 sites mediate activation of the
plasminogen activator inhibitor-1 promoter by glucose
in vascular smooth muscle cells. J Biol Chem 273, 8225–
8231.
12 Fink T, Kazlauskas A, Poellinger L, Ebbesen P &
Zachar V (2002) Identification of a tightly regulated
hypoxia-response element in the promoter of human

plasminogen activator inhibitor-1. Blood 99, 2077–2083.
13 Eriksson P, Nilsson L, Karpe F & Hamsten A (1998)
Very-low-density lipoprotein response element in the
promoter region of the human plasminogen activator
inhibitor-1 gene implicated in the impaired fibrinolysis
of hypertriglyceridemia. Arterioscler Thromb Vasc Biol
18, 20–26.
14 Dennler S, Itoh S, Vivien D, ten Dijke P, Huet S &
Gauthier JM (1998) Direct binding of Smad3 and
Smad4 to critical TGF beta-inducible elements in the
promoter of human plasminogen activator inhibitor-
type 1 gene. EMBO J 17, 3091–3310.
15 Vulin AI & Stanley FM (2004) Oxidative stress activates
the plasminogen activator inhibitor type 1 (PAI-1) pro-
moter through an AP-1 response element and coop-
erates with insulin for additive effects on PAI-1
transcription. J Biol Chem 279, 25172–25178.
16 Al-Nedawi KN, Czyz M, Bednarek R, Szemraj J, Swiat-
kowska M, Cierniewska-Cieslak A, Wyczolkowska J &
Cierniewski CS (2004) Thymosin beta 4 induces the
synthesis of plasminogen activator inhibitor 1 in cul-
tured endothelial cells and increases its extracellular
expression. Blood 103, 1319–1324.
17 Hou B, Eren M, Painter CA, Covington JW, Dixon JD,
Schoenhard JA & Vaughan DE (2004) Tumor necrosis
factor alpha activates the human plasminogen activator
inhibitor-1 gene through a distal nuclear factor kappaB
site. J Biol Chem 279, 18127–18136.
18 Ranta V, Orpana A, Carpen O, Turpeinen U, Ylikorkala
O & Viinikka L (1999) Human vascular endothelial cells

produce tumor necrosis factor-alpha in response to pro-
inflammatory cytokine stimulation. Crit Care Med 27,
2184–2187.
19 Imaizumi T, Itaya H, Fujita K et al. (2000) Expression
of tumor necrosis factor-alpha in cultured human endo-
thelial cells stimulated with lipopolysaccharide or
interleukin-1alpha. Arterioscler Thromb Vasc Biol 20,
410–415.
The 4G ⁄ 5G site and PAI-1 expression M. Swiatkowska et al.
5830 FEBS Journal 272 (2005) 5821–5831 ª 2005 The Authors Journal Compilation ª 2005 FEBS
20 Keulenaer GW, Alexander WR, Ushio-Fukai M, Ishiz-
aka N & Griendling KK (1998) Tumour necrosis factor
a activates a p22
phox
-based NADH oxidase in vascular
smooth muscle. Biochem J 329, 653–657.
21 Kunsch Ch & Medford RM (1999) Oxidative stress as a
regulator of gene expression in the vasculature. Circ Res
85, 753–766.
22 Sundaresan MYuZX, Ferrans VJ, Irani K & Finkel T
(1995) Requirement for generation of H
2
O
2
for platelet-
derived growth factor signal transduction. Science 270,
298–301.
23 Sen CK & Packer L (1996) Antioxidant and redox regu-
lation of gene transcription. FASEB J 10, 709–720.
24 Dawson SJ, Wiman B, Hamsten A, Green F, Humph-

ries S & Henney A (1993) The two allele sequences of a
common polymorphism in the promoter of the plasmi-
nogen activator inhibitor-1 (PAI-1) gene respond differ-
ently to interleukin-1 in HepG2 cells. J Biol Chem 268,
10739–10745.
25 Lane DA & Grant PJ (2000) Role of hemostatic gene
polymorphisms in venous and arterial thrombotic dis-
ease. Blood 95, 1517–1532.
26 Grancha S, Estelles A, Tormo G, Falco C, Gilabert J,
Espana F, Cano A, Segui R & Aznar J (1999) Plasmi-
nogen activator inhibitor-1 (PAI-1) promoter 4G ⁄ 5G
genotype and increased PAI-1 circulating levels in post-
menopausal women with coronary artery disease.
Thromb Haemostat 81, 516–521.
27 Hoekstra T, Geleijnse JM, Schouten EG & Kluft C
(2004) Plasminogen activator inhibitor-type 1: its plasma
determinants and relation with cardiovascular risk.
Thromb Haemostat 91, 861–872.
28 Murphy HS, Shayman JA, Till GO, Mahrougui M,
Owens CB, Ryan US & Ward PA (1992) Superoxide
responses of endothelial cells to C5a and TNF-a: diver-
gent signal transduction pathways. Am J Physiol Lung
Cell Mol Physiol 263, L51–L59.
29 Matsubara T & Ziff M (1986) Increased superoxide
anion release from human endothelial cells in response
to cytokines. J Immunol 137, 3295–3329.
30 Jiang Z, Seo JY, Ha H, Lee EA, Kim YS, Han SCh,
Uh ST, Park Ch S & Lee HB (2003) Reactive oxygen
species mediate TGF-b1-induced plasminogen activator
inhibitor-1 upregulation in mesantial cells. Biochem

Bioph Res Commun 309, 961–966.
31 Du XL, Edelstein D, Rossetti L, Fantus IG, Goldberg H,
Ziyadeh FN, Wu J & Brownlee M (2000) Hyperglycemia-
induced mitochondrial superoxide overproduction acti-
vates the hexoamine pathway and induces plasminogen
activator inhibitor-1 expression by increasing Sp1 glyco-
sylation. Proc Natl Acad Sci USA 97, 12222–12226.
32 Cheng JJ, Chao YJ, Wung BS & Wang DL (1996)
Cyclic-strain-induced plasminogen activator inhibitor-1
(PAI-1) release from endothelial cells involves reactive
oxygen species. Biochem Biophys Res Commun 225,
100–105.
33 Zhao W, Spitz DR, Oberley LW & Robbins MEC
(2001) Redox modulation of the pro-fibrogenic mediator
plasminogen activator inhibitor-1 following ionizing
radiation. Cancer Res 61, 5537–5543.
34 Okada H, Woodcock-Mitchell J, Mitchell J, Sakamoto
T, Marutsuka K, Sobel BE & Fujii S (1998) Induction
of plasminogen activator inhibitor type 1 and collagen
expression in rat cardiac microvascular endothelial cells
by interleukin-1 and its dependence on oxygen-centered
free radicals. Circulation 97, 2175–2182.
35 Janssen-Heininger YM, Macara I & Mossman BT
(1999) Cooperativity between oxidants and tumor
necrosis factor in the activation of nuclear factor
NF-jB: requirement of Ras ⁄ mitogen-activated protein
kinases in the activation of NF-jB by oxidants. Am J
Respir Cell Mol Biol 20, 942–952.
36 Baeuerle PA & Henkel T (1996) Function and activa-
tion of NF-jB in the immune system. Annu Rev Immu-

nol 12, 141–117.
37 Ruan QR, Zhang WJ & Hunfnagl P (2001) Anisoda-
mine counteracts lipopolysaccharide-induced tissue fac-
tor and plasminogen activator inhibitor-1 expression in
human endothelial cells: contribution of the NF-jB
pathway. J Vasc Res 38, 13–19.
38 Motojima M, Hosokawa A, Yamato H, Muraki T &
Yoshioka T (2003) Uremic toxin of organic anions up-
regulate PAI-1 expression by induction of NF-jB and
free radical in proximal tubular cells. Kidney Int 63,
1671–1680.
39 Dechend R, Maass M, Gieffers J, Dietz R, Scheidereit
C, Leutz A & Gulba DC (1999) Chlamydia pneumoniae
infection of vascular smooth muscle and endothelial
cells activates NF-jB and induces tissue factor and
PAI-1 expression. Circulation 100, 1369–1373.
M. Swiatkowska et al. The 4G ⁄ 5G site and PAI-1 expression
FEBS Journal 272 (2005) 5821–5831 ª 2005 The Authors Journal Compilation ª 2005 FEBS 5831