Selective induction of cyclo-oxygenase-2 activity in the permanent human endothelial cell line HUV-EC-C: Biochemical and pharmacological characterization

Article abstract of DOI:10.1038/sj.bjp.0701112

1. Cyclo-oxygenase (COX), the enzyme responsible for the conversion of arachidonic acid (AA) to prostaglandin H₂ (PGH₂), exists in two forms, termed COX-1 and COX-2 which are encoded by different genes. COX-1 is expressed constitutively and is known to be the site of action of aspirin and other nonsteroidal anti-inflammatory drugs. COX-2 may be induced by a series of pro-inflammatory stimuli and its role in the development of inflammation has been claimed. 2. Endothelial cells are an important physiological source of prostanoids and the selective induction of COX-2 activity has been described for finite cultures of endothelial cells, but not for permanent endothelial cell lines. 3. The HUV-EC-C line is a permanent endothelial cell line of human origin. We have determined the COX activity of these cells under basal conditions and after its exposure to two different stimuli, phorbol 12-myristate 13-acetate (PMA) and interleukin-1β (IL-1β). 4. Both PMA and IL-1β produced dose- and time-dependent increases of the synthesis of the COX-derived eicosanoids. These increases were maximal after the treatment with 10 nM PMA for 6 to 9 h. Under these conditions, the main eicosanoid produced by the cells was PGE₂. 5. The increase of COX activity by PMA or IL-1β correlated with an increase of the enzyme's apparent V(max), whilst the affinity for the substrate, measured as apparent K(m), remained unaffected. 6. Treatment of the cells with PMA induced a time-dependent increase in the expression of both COX-1 and COX-2 mRNAs. Nevertheless, this increase was reflected only as an increase of the COX-2 isoenzyme at the protein level. 7. The enzymatic activity of the PMA-induced COX was measured in the presence of a panel of enzyme inhibitors, and the IC₅₀ values obtained were compared with those obtained for the inhibition of human platelet COX activity, a COX-1 selective assay. Classical non-steroidal anti-inflammatory drugs (NSAIDs) inhibited both enzymes with varying potencies but only the three compounds previously shown to be selective COX-2 inhibitors (SC-58125, NS-398 and nimesulide) showed higher potency towards the COX of PMA-treated HUV-EC-C. 8. Overall, it appears that the stimulation of the HUV-EC-C line with PMA selectively induces the COX-2 isoenzyme. This appears to be a suitable model for the study of the physiology and pharmacology of this important isoenzyme, with a permanent endothelial cell line of human origin.

Full text of DOI:10.1038/sj.bjp.0701112

British Journal of Phar
m
acology (1997) 121, 171 ± 180
 1997 Stockton Press All rights reserved 0007 ± 1188/97 $12.00
Selective induction of cyclo-oxygenase-2 activity in the permanent
human endothelial cell line HUV-EC-C: biochemical and
pharmacological characterization
*
*
Montserrat Miralpeix, Mercedes Camacho, Jesu s Lopez-Belmonte, *Francesca Canalõas,
 Â
a
1
Jordi Beleta, *Jose M Palacios & Luis Vila
Â
Laboratory of In¯ammation Mediators, Institute of Research of the Santa Creu i Sant Pau Hospital and *Department of Molecular
Pharmacology, Laboratorios Almirall S.A., Barcelona, Spain
1 Cyclo-oxygenase (COX), the enzyme responsible for the conversion of arachidonic acid (AA) to
prostaglandin H (PGH ), exists in two forms, termed COX-1 and COX-2 which are encoded by di€erent
2
2
genes. COX-1 is expressed constitutively and is known to be the site of action of aspirin and other non-
steroidal anti-in¯ammatory drugs. COX-2 may be induced by a series of pro-in¯ammatory stimuli and
its role in the development of in¯ammation has been claimed.
2
Endothelial cells are an important physiological source of prostanoids and the selective induction of
COX-2 activity has been described for ®nite cultures of endothelial cells, but not for permanent
endothelial cell lines.
3 The HUV-EC-C line is a permanent endothelial cell line of human origin. We have determined the
COX activity of these cells under basal conditions and after its exposure to two di€erent stimuli, phorbol
1
2-myristate 13-acetate (PMA) and interleukin-1b (IL-1b).
4 Both PMA and IL-1b produced dose- and time-dependent increases of the synthesis of the COX-
derived eicosanoids. These increases were maximal after the treatment with 10 nM PMA for 6 to 9 h.
Under these conditions, the main eicosanoid produced by the cells was PGE
2
.
5
V
The increase of COX activity by PMA or IL-1b correlated with an increase of the enzyme's apparent
max, whilst the anity for the substrate, measured as apparent K , remained una€ected.
Treatment of the cells with PMA induced a time-dependent increase in the expression of both COX-1
m
6
and COX-2 mRNAs. Nevertheless, this increase was re¯ected only as an increase of the COX-2
isoenzyme at the protein level.
7
The enzymatic activity of the PMA-induced COX was measured in the presence of a panel of enzyme
inhibitors, and the IC50 values obtained were compared with those obtained for the inhibition of human
platelet COX activity, a COX-1 selective assay. Classical non-steroidal anti-in¯ammatory drugs
(
NSAIDs) inhibited both enzymes with varying potencies but only the three compounds previously
shown to be selective COX-2 inhibitors (SC-58125, NS-398 and nimesulide) showed higher potency
towards the COX of PMA-treated HUV-EC-C.
8
Overall, it appears that the stimulation of the HUV-EC-C line with PMA selectively induces the
COX-2 isoenzyme. This appears to be a suitable model for the study of the physiology and
pharmacology of this important isoenzyme, with a permanent endothelial cell line of human origin.
Keywords: Inducible cyclo-oxygenase; COX-2; endothelial cell line; HUV-EC-C; phorbol 12-myristate 13-acetate; interleukin-1;
human platelets; COX-1; NSAID; nimesulide; SC-58125; NS-398
Introduction
Cyclo-oxygenase (COX), also termed prostaglandin-H syn-
thase, is the key enzyme involved in the biosynthesis of pros-
tanoids from arachidonic acid (AA), Wlodawer & Samuelson,
2
and TXA -synthase, respectively (Haurand & Ullrich, 1985;
Hecker & Ullrich, 1989; Tanabe & Ullrich, 1995). COX is
expressed in most cell types and tissues (De Witt, 1991; Smith,
1992; O'Neill & Ford-Hutchinson, 1993; Goppelt-Struebe,
1995). Nevertheless, speci®c prostanoids are formed in di€er-
ent tissues due to the di€erentially expressed secondary en-
zymes (Tanabe & Ullrich, 1995; Urade et al., 1995).
COX is the site of action of aspirin and a wide range of
other non-steroidal anti-in¯ammatory drugs (NSAIDs). Inhi-
bition of prostanoid biosynthesis can account for both the
therapeutic and side e€ects associated with the administration
of NSAIDs (Vane, 1971). Studies in recent years have indica-
ted that two isoforms of COX exist, encoded by separate genes,
referred to as COX-1 and COX-2 (Yokoyama & Tanabe, 1989;
Fletcher et al., 1992). Human COX-2 shares about 60%
identity with COX-1 (Jones et al., 1993). COX-1 is constitu-
tively expressed in most tissues (De Witt, 1991; O'Neill &
Ford-Hutchinson, 1993; Goppelt-Struebe, 1995) and in blood
monocytes and platelets (Funk et al., 1991; Hla & Neilson,
1992). COX-1 has been assigned as responsible for `house-
1
973; Smith et al., 1991) and octadecanoids from linoleic acid
Hamberg & Samuelsson, 1967; Kaduce et al., 1989; Camacho
et al., 1995). COX catalyses the oxidative transformation of
AA into prostaglandin H (PGH ), the common precursor for
physiologically important prostaglandins and thromboxane A
TXA ) (Wlodawer & Samuelsson, 1973; Smith et al., 1991;
(
2
2
2
(
2
Smith, 1992). Prostanoids are in¯ammation, thrombosis and
vascular tone modulators and they are classic targets for
therapeutic intervention (Flower, 1974). The isomerization of
PGH
enzymatically catalysed (Nugteren & Christ-Hazelhof, 1980;
Smith, 1992) whereas the formation of PGI and TXA from
is catalysed by speci®c enzymes termed PGI -synthase
2 2 2
to PGE , PGD , and PGF2a may be spontaneous or
2
2
PGH
2
2
1
Author for correspondence at: H.S. Creu i S. Pau (Casa de
Convalecencia), S. Antonio M Claret 167, 08025-Barcelona, Spain.
a

172
M. Miralpeix et al
COX-2 activity in HUV-EC-C cell line
keeping' prostanoid biosynthesis when agonist-activated mo-
bilization of esteri®ed AA from membrane phospholipids oc-
curs. In contrast, COX-2 expression is induced by serum,
growth factors, phorbol esters, cytokines, hormones and bac-
terial toxins, and is over-expressed in the in¯ammatory process
IL-1b or phorbol-12-myristate-13-acetate (PMA). Experiments
were performed with HUV-EC-C passage 19-27.
Preparation of platelet suspensions
(
Maier et al., 1990; Kujubu et al., 1991; 1993; Lee et al., 1992;
Sirois & Richards, 1992; Sirois et al., 1992; O'Sullivan et al.,
992a, b; O'Bannion et al., 1992; Jones et al., 1993). It has been
Platelets were isolated from human peripheral blood obtained
from healthy donors who denied taking any non-steroidal anti-
in¯ammatory drugs (NSAIDs) during at least the previous
1
7
1
postulated that COX-2-derived prostanoids may play a role in
in¯ammation and other pathological situations. Expression of
COX-2 is inhibited by glucocorticoids, probably by a post-
transcriptional mechanism (Lee et al., 1992; O'Bannion et al.,
week. The blood was anticoagulated with 2 mg ml sodium
EDTA and centrifuged at 180 g for 10 min at room tempera-
ture to obtain a platelet-rich plasma. Then, the platelet-rich
plasma was centrifuged at 2,000 g for 20 min at 48C to obtain
a platelet pellet. Cells were washed twice with phosphate-buf-
1
992; Kujubu & Herschmann, 1992; Masferrer et al., 1992),
2+
2+
and this partially contributes to the anti-in¯ammatory activity
exhibited by these compounds.
fered saline (PBS) without Ca and Mg (GIBCO-BRL) and
7
71
resuspended to 5610 cells ml with Hank's balanced salt
solution (HBSS, GIBCO-BRL).
While di€erential regulation and/or tissue speci®c location
of COX-1 and COX-2 has been extensively studied, the cata-
lytic-structural di€erences between these proteins are still not
well-known (Kulmacz et al., 1994; Bakovic & Dunford, 1994;
Lacomte et al., 1994; Laneuville et al., 1995; Capdevila et al.,
Determination of the metabolic pro®le of AA and COX-
activity in HUV-EC-C
1
McKeevwe et al., 1996). Despite this fact, a growing amount of
995; Kulmacz & Wang, 1995; Kurumbail et al., 1996;
After the desired period of time of cell exposure to the indi-
cated concentration of IL-1b and PMA, cells were incubated at
378C in the presence of 0.5 ml Hank's F12K containing 10 mM
HEPES and the desired concentration of [ C]-AA in 5 ml of
ethanol. At the indicated periods of time, the reactions were
stopped by adding 1 N HCl to yield pH 3 followed by one
volume of cold methanol. Samples were kept at 7808C until
analysis. High performance liquid chromatographic (h.p.l.c.)
analysis of eicosanoids was performed as previously described
data indicate that selective inhibition of COX-2 is possible
(
14
Meade et al., 1993; Barnett et al., 1994; Mitchell et al., 1994;
Copeland et al., 1994; Quellet & Percival, 1995; Camacho et
al., 1995). COX-2 selective inhibitors could have therapeutic
advantages compared with the classic non selective com-
pounds, especially regarding gastric side-e€ects (Arai et al.,
1
1
993; Spangler, 1993; Masferrer et al., 1994; Seibert et al.,
994; Klein et al., 1994; Chan et al., 1995). In addition, ex-
(Sola et al., 1992). Brie¯y, samples were injected into a reverse
Â
perimental use of COX-2 selective NSAIDs will also contribute
to our understanding of the physiological role of the two forms
of COX.
Di€erent cell types have been used to study the selectivity of
NSAIDs on COX-1 and COX-2 in vitro, including ®broblasts,
endothelial cells and cos-1 cells transfected with murine COX-1
or COX-2 (Meade et al., 1993). Human endothelial cells are
particularly relevant because the vascular endothelium is an
important physiological source of prostaglandins at sites of
in¯ammation. In vitro, stimulated endothelial cells produce
phase column (Ultrasphere-ODS 46250 mm, Beckman) and
the composition of the mobile phase was 100% of A (aceto-
nitrile : water : acetic acid 33 : 67 : 0.01) for the ®rst 16 min after
the injection. Afterwards, the percentage of B (acetonitrile : -
water : acetic acid 90 : 10 : 0.01) was increased linearly until it
reached 65% 26 min after injection, and then maintained until
71
min 45. Mobile phase ¯ow rate was 1 ml min . When HETEs
had to be analysed the samples were injected into the column
and eluted isocratically with methanol/water/tri¯uoroacetic
71
acid/triethylamine (75 : 25 : 0.1 : 0.05), ¯ow rate 1 ml min
.
PGI
2
, PGF2a and PGE
experimental conditions. Minor amounts of PGD
2
as the main eicosanoids under di€erent
, 12-hydro-
xy-heptadecatrienoic acid (HHT), 15-hydroxy-eicosatetraenoic
2
COX-1 and -2 speci®c mRNA analysis
acid (15-HETE) and 11-HETE are also produced (Marcus et
Cells were incubated in the presence or absence of 10 nM PMA
for the indicated periods of time. Total RNA was isolated by
phenol-chloroform extraction and isopropanol precipitation as
described by Chomczynski & Sacchi (1987). The speci®c
mRNA levels were estimated by means of a quantitative RT ±
PCR protocol. One mg of total RNA was reverse-transcribed
into cDNA by incubating with 50 units of murine leukemia
virus reverse transcriptase in a reaction bu€er containing
al., 1978; Alhenc-Gelas et al., 1982; Hopkins et al., 1984; Ku
et al., 1985; Zavoico et al., 1989), all of which are derived from
COX activity (Lopez et al., 1993). Finite cultures of endothelial
È
hn
Â
cells from human umbilical vein (HUVEC) are the most widely
used tool to study endothelial cell physiology. Nevertheless,
the use of HUVEC as a model for quantitative studies invol-
ving COX, such as screening tests, has serious disadvantages.
A large individual variability in COX activity is found, making
it a time consuming procedure because of the great number of
primary cultures that are necessary. The purpose of the present
work was to characterize the metabolic pro®le of AA via the
COX pathway and the selective induction and expression of
COX-2 in a permanent human endothelial cell line (HUV-EC-
C).
10 mM Tris-HCl pH 8.3, 50 mM KCl, 5 mM MgCl , 2.5 mM
2
random hexamers, 20 u RNAsin, 1 mM dNTPs in a ®nal
volume of 20 ml. The reaction mixture was incubated for
30 min at 428C and then was stopped by heating for 5 min at
998C and cooled for 5 min at 58C. The primers used for COX-
1 and COX-2 were 5'-TGCCCAGCTCCTGGCCCGCCG-
CTT-3' (sense) and 5'-GTGCATCAACACAGGCGCCTC-
TTC-3' (antisense) and 5'-TTCAAATGAGATTGTGGGAA-
AATTGCT-3' (sense) and 5'-AGATCATCTCTGCCTGAG-
TATCTT-3' (antisense), respectively (Hla & Neilson, 1992).
The RNA for glyceraldehyde-3-phosphate-dehydrogenase
(GAPDH) was ampli®ed and used as the internal control. The
sense and antisense primers used for GAPDH were
5'-CCACCCATGGCAAATTCCATGGCA-3' and 5'-TCTA-
GACGGCAGGTCAGGTCCACC-3', respectively (Maier et
al., 1990). The PCR was carried out in a DNA Thermal Cycler
480 (Perkin Elmer) with a reaction mixture (100 ml) containing
Methods
HUV-EC-C culture and treatment with IL-1b and PMA
The HUV-EC-C is a permanent endothelial cell line derived
from the vein of a normal human umbilical cord (ATCC CRL
1730). Cells were grown in F12K medium containing 10%
foetal bovine serum (FBS) and supplemented with 2 mM L-
7
1
glutamine, 1 mM sodium pyruvate, 100 u ml
00 u ml
penicillin,
heparin and
2
10 mM Tris-HCl pH 8.3, 50 mM KCl, 2.5 mM MgCl , 0.2 mM
7
1
71
1
streptomycin, 100 mg ml
dNTP, 1 mM sense/antisense primers, 2.5 u Taq polymerase
71
3
50 mg ml endothelial cell growth supplement (ECGS). Cells
in con¯uent state were maintained without heparin and ECGS
and 4 mCi [ H]-dCTP. Serial half dilutions of the cDNA were
done in order to prove linearity. Twenty seven cycles were
for 48 h before the addition (or not) of the desired amount of
performed as follows, 1.5 min 948C, 1.5 min 588C, 1.5 min

M. Miralpeix et al
COX-2 activity in HUV-EC-C cell line
173
7
28C, for all the samples. The products of the ampli®cation
was then changed and cells were incubated in the presence or
absence of drugs for 30 min at 378C. AA 50 mM was then
added, and the cells were incubated for a further 30 min. The
were separated by electrophoresis in a 1.5% w/v Low Melting
Point Agarose gel, containing ethidium bromide. To eliminate
the excess of [ H]-dCTP and to avoid excessive background,
3
2
production of PGF in response to AA was measured in the
the running bu€er was replaced by a fresh one when the
loading dye had migrated half way through the gel. The bands
were visualized and cut under u.v. light with a circular tem-
plate and melted by heating to 708C. Complete digestion of the
gel pieces was performed by incubating the pieces in a water
bath at 508C for 1 h with 1 ml of the tissue solubilizer NCS-II.
Afterwards, 10 ml of scintillation cocktail were added and the
radioactivity was monitored in a b-counter (LS-3800, Beck-
supernatants by an enzyme-immunoassay system (BIOTRAK,
Amersham). The inhibitory e€ects on the COX-2 activity were
evaluated by incubating each drug with ®ve or six di€erent
concentrations in triplicate. IC50 values were obtained as
mentioned above.
Drugs
man, San Ramo
Â
n, CA). The radioactivity was normalized with
Non-commercial available drugs (nimesulide, NS-398: (N-
[2-cyclohexyloxy - 4 - nitrophenyl] methansulphonamide), SC-
respect to GAPDH.
5
8125: (1-[(4-methylsulphonyl)phenyl]-3-tri¯uoromethyl-5-(4-
Western-blot analysis of COX-1 and COX-2
¯uorophenyl)pyrazole, ketorolac) were synthesized by the
Department of Chemical Synthesis of Laboratorios Almirall.
73
Cells were incubated in the presence or absence of 10 nM PMA
for the indicated periods of time. To solubilize COX proteins,
cells were scraped, sonicated and incubated for 1 h at 48C with
extraction bu€er (50 mM Tris-HCl pH 8.0 containing 10 mM
EDTA, 1 mM PMSF and 1% v/v Tween-20) with gentle
shaking. Cell extracts were centrifuged at 10,000 g for 5 min
and were then boiled for 3 min in Laemmli sample bu€er
Stock solutions (10 M) of the drugs were dissolved in 50%
dimethylsulphoxide or water, and further dilutions were done
with F12 K medium (for COX-2 assay) or HBSS (for COX-1
assay). Drug vehicle, at the concentrations employed, did not
a€ect enzyme activities.
(
Laemmli, 1970) containing 100 mM dithiothreitol. Protein
a
concentration of the supernatants was determined using the
DC Protein Assay (Bio-Rad), a colorimetric assay for protein
measurement of samples solubilized in detergent. The samples
were adjusted by dilution to have an equal amount of protein
for each condition, and 20 mg of protein lysate were separated
by polyacrylamide gel electrophoresis in the presence of so-
dium dodecyl sulphate (SDS) (SDS ± PAGE). An 8% w/v re-
solving gel and a 4% w/v stacking gel were used. The resolved
proteins were electrotransferred onto nitrocellulose mem-
branes as described by Towbin et al. (1979). Nonspeci®c an-
tibody binding to the nitrocellulose was prevented by
incubating the ®lter overnight at 48C with block solution (3%
w/v bovine serum albumin, 20 mM Tris, pH 7.5, 150 mM NaCl
and 0.01% v/v Tween 20). Blocked membranes were incubated
for 3 h at room temperature with either rabbit polyclonal
antiserum against ovine COX-1 (PG20, 1 : 1000 dilution) or
rabbit polyclonal antiserum against a peptide corresponding to
a distinct carboxyl-terminal region of human COX-2 (PG27,
2
1
1
000
500
000
5
00
0
b
IL-1β
2000
1
500
1
: 1000 dilution). The membranes were extensively washed
with 20 mM Tris, pH 7.5, 150 mM NaCl and 0.01% v/v Tween
0. Immunoreaction and detection were performed by using an
2
1000
ECL Chemiluminescence Western blotting Kit (rabbit) fol-
lowing the manufacturer's instructions, and bands were vi-
sualized after exposure to Hyper®lm ECL.
5
00
0
Inhibition of COX-1 activity in human platelets
7
71
A 0.2 ml aliquot of platelets suspension (5610 ml ) were
preincubated in the presence or absence of drugs during
c
2
1
1
000 PMA
500
1
5 min at 378C. AA 50 mM was then added and incubations
were continued for a further 15 min. Then, tubes were placed
on ice and centrifuged at 2,000 g for 10 min at 48C. The pro-
duction of TXB was measured in the supernatants by an en-
2
zyme-immunoassay system (BIOTRAK, Amersham). The
inhibitory e€ects on the COX-1 activity were evaluated by
incubating each drug at ®ve to six di€erent concentrations with
triplicate determinations. IC50 values were obtained by non-
linear regression by use of InPlot, GraphPad Software on an
IBM computer.
000
5
00
0
0
10
20
30
40
50
Inhibition of COX-2 activity in the HUV-EC-C line
Tim e (m in)
HUV-EC-C expresses selectively COX-2 isoenzyme after
treatment with PMA (this paper). HUV-EC-C (2610 per
well) were seeded onto 96-well plates and made quiescent by
removing the growth factor for 48 h before the initiation of the
experiments. Quiescent cells were treated with 10 nM PMA for
4
Figure 1 Representative chromatograms from samples of HUV-EC-
C (a) untreated (control) and treated with (b) 10 u ml interleukin-
71
1
b (IL-1b) or (c) 20 nM phorbol 12-myristate 13-acetate (PMA) for
14
24 h. Cells were incubated at 378C with 25 mM [ C]-arachidonic acid
(AA) for 20 min. Peak identi®cation of each eicosanoid was
performed by coelution with reference standards.
6
h at 378C to induce the COX-2 isoenzyme. Cultured medium

174
M. Miralpeix et al
COX-2 activity in HUV-EC-C cell line
Materials
COX-1 (ram seminal vesicles) and COX-2 (sheep placenta)
electrophoresis standards were from Cayman Chemical
14
14
Foetal bovine serum, phosphate-bu€ered saline (PBS) and
Hank's balanced salt solution (HBSS) were purchased from
GIBCO-BRL (France). F12K medium, hepatin, endothelial
cell growth supplement (ECGS), arachidonic acid, phorbol
Company (U.S.A.). [1- C]-arachidonic acid ([ C]-AA), 55 ±
1 3
58 mCi mmol , [ H]-dCTP 48 ± 71 Ci mmol , tissue solu-
7
71
bilizer NCS-II, enzyme-immunoassay kits for PGE
TXB
luminescence kit were purchased from Amersham Ibe
2
and
2
determinations, Hyper®lm ECL and the ECL chemi-
12-myristate 13-acetate (PMA) and all other compounds
Â
rica
were obtained from Sigma-Aldrich Quõmica S.A. (Spain).
Â
1
(Madrid, Spain). The scintillation cocktail (Ecoscint H) was
from National diagnostics (Atlanta, GA, U.S.A.).
7
Human recombinant IL-1b 50000 u mg , purity 498% was
from Boehringer Mannheim (GmbH, Germany). Solvents
were h.p.l.c. grade from Sharlau S.A. (Barcelona, Spain).
Random hexamers and Rnasin (GeneAmp RNA kit), ethi-
dium bromide and Taq polymerase were from Perkin Elmer
Results
(
Brachburg, NJ, U.S.A.). dNTPs were from Epicentre
H.p.l.c. analysis of eicosanoids produced by the HUV-EC-C
line when cells were incubated with exogenous [ C]-AA
showed that both control and stimulated cells produced
1
4
Technologies (Madison, WI, U.S.A.). Low Melting Point
Agarose was supplied by Gibco BRL (Paisley, Scotland).
The rest of the electrophoresis reagents were from Bio-Rad
Laboratories (Spain). Nitrocellulose membranes were ob-
PGF2a, PGE
as the main eicosanoids, whereas PGD
2
and 12-hydroxy-heptadecatrienoic acid (HHT)
, 6-keto-PGF1a, 15-
2
tained from Schleicher & Schu
È
ell (Germany). Rabbit poly-
clonal antiserums against COX-1 (PG 20) and COX-2 (PG
hydroxy-eicosatetraenoic acid (15-HETE) and 11-HETE were
produced in minor amounts (Figure 1). Treatment of HUV-
EC-C cells with PMA resulted in a higher production of the
2
7) were from Oxford Biomedical Research, Inc. (U.S.A.).
a
d
5
4
3
2
1
0
0
0
0
0
0
6
4
3
1
00
50
00
50
0
b
e
2
2
1
1
50
00
50
00
8
6
4
0
0
0
5
0
0
20
0
c
f
3
2
1
00
00
00
0
1
00
7
5
2
5
0
5
0
0
5
10
15
20
IL-1β (u m l–1
0
5
10
15
20
)
Figure 2 Biosynthesis of eicosanoids ((a) 6-keto-PGF1a, (b) PGE
PGD and (f) hydroxy-eicosatetra-eroic acids (HETEs)) by HUV-EC-C as a function of IL-1b concentration. Cells were incubated
for 24 h at 378C in the presence of 0, 1, 2, 5, 10 and 20 u ml
0 min, and then eicosanoids were analysed by h.p.l.c. Data are shown as the mean, n=4; vertical lines show s.e.mean.
2
, (c) 12-hydroxy-heptadecatrienoic acid (HHT), (d) PGF2a, (e)
2
71
14
IL-1b. Afterwards, cells were incubated with 25 mM [ C]-AA for
2

M. Miralpeix et al
COX-2 activity in HUV-EC-C cell line
175
above mentioned eicosanoids compared to cells treated with
IL-1b or untreated cells. PGE was the most synthesized
2
prostanoid after PMA-stimulation of HUV-EC-C cells. The
HETEs were quantitated together, since they resolved poorly
with the chromatographic technique used.
COX activity was measured as a function of time of expo-
sure to either IL-1b or PMA. Figure 5 shows that COX activity
achieved a maximum 6 to 9 h after addition of both stimuli
and remained elevated for at least 24 h.
The e€ect of varying the AA concentration on enzyme ac-
tivity was measured in control, IL-1b and PMA-treated cells
(Figure 6). In previous time-course assays, a linear relationship
between COX-derived eicosanoids formation and incubation
time was found for the ®rst 10 min of incubation with the
substrate and this incubation time was chosen for kinetic ex-
Biosynthesis of all eicosanoids was increased several fold in
a concentration-dependent fashion after exposure of HUV-
EC-C to IL-1b (Figure 2) and PMA (Figure 3). PMA enhanced
the production of all eicosanoids much more than IL-1b, ex-
cept for PGF2a. PMA induced a maximum eicosanoid bio-
synthetic activity at 10 nM and higher concentrations produced
lower amounts of COX products. A lower, although signi®cant
increase in the production of AA metabolites was also ob-
served when IL-1b was used as the stimulus. The maximum
periments. Apparent K
m
values for AA of 12.2+0.65,
10.9+1.7, and 7.0+1.4 mM (mean+s.e.mean, n=4) were ob-
tained for the control, IL-1b- and PMA-treated cells, respec-
tively. The apparent Vmax values obtained in the same assays
71
6
biosynthetic activity with IL-1b was observed at 10 u ml
.
were 481.9+20.25, 919.7+30.6 and 2577.4+127.3 pmol/10
Indomethacin inhibited the formation of all eicosanoids in a
concentration-dependent manner (Figure 4), which indicated a
COX origin for all of them. Therefore, COX activity, expres-
sed as [ C]-AA transformed through COX pathway, was
evaluated as the sum of all the aforementioned compounds.
cells/10 min (mean+s.e.mean, n=4), respectively, for the same
above-mentioned conditions.
Overall, the e€ect of PMA on COX activity was quantita-
tively more signi®cant than that observed with IL-1b. Therefore,
PMA was chosen as the preferred stimulus and was used
1
4
a
d
3
00
600
2
2
50
00
4
3
1
50
00
50
0
1
1
50
00
5
0
0
b
e
7
6
4
3
1
50
00
50
00
50
0
250
200
1
1
50
00
5
0
0
c
f
8
6
4
2
00
00
00
00
0
600
450
300
150
0
0
10
20
30
40
50
PMA (nM)
0
10
20
30
40
50
2 2
Figure 3 Biosynthesis of eicosanoids ((a) 6-keto-PGF1a, (b) PGE , (c) HHT, (d) PGF2a, (e) PGD and (f) HETEs)) by HUV-EC-C
as a function of phorbol 12-myristate 13-acetate (PMA) concentration. Cells were incubated for 24 h at 378C in the presence of 0, 1,
1
4
5
, 10, 20 and 50 nM PMA. Afterwards, cells were incubated with 25 mM [ C]-AA for 20 min and then eicosanoids were analysed by
h.p.l.c. Data are shown as the mean, n=4; vertical lines show s.e.mean.

176
M. Miralpeix et al
COX-2 activity in HUV-EC-C cell line
a
b
1
00
0
0
0
0
.3
.2
.1
.0
8
6
4
2
0
0
0
0
0
25
50
75
100
3
2
000
500
S
0
0
2000
.001
0.1
1
10
0.001
0.1
1
10
1
1
500
000
Indom ethacin (µM)
Figure 4 E€ect of indomethacin concentration on the synthesis of
(
a) prostaglandins (*), HHT (&); (b) 11-HETE (~) and 15-HETE
^). PMA treated-HUV-EC-C were preincubated for 5 min with
5
00
0
(
1
4
several concentrations of indomethacin before 25 mM
[ C]-AA
addition. Afterwards, cells were incubated for another 20 min. Data
are shown as the mean, n=4; vertical lines show s.e.mean. For
abbreviations used see legends of Figures 2 and 3.
0
25
50
[14C]-AA (µM)
75
100
Figure 6 E€ect of substrate concentration on COX activity in
control (~), PMA-treated (&) and IL-1b-treated (*) HUV-EC-C.
Cells were incubated with several concentrations of [ C]-AA (5, 10,
14
3
3
2
2
1
1
500
000
500
000
500
000
25, 50 and 100 mM) for 10 min. Insert: Woolf plot, were S indicates
6
AA concentration (mM) and
v the enzyme activity (pmol/10
cells 10 min ). Data are shown as the mean, n=4; vertical lines
71
show s.e.mean.
time-course of COX-1 (A) and COX-2 (B) protein expression
by incubating HUV-EC-C with or without 10 nM PMA for
various time intervals. In the absence of PMA, the antibody
against COX-1 (PG 20) recognized a weak band of approxi-
mately 70 kDa corresponding to the migration of puri®ed
COX-1 from ram seminal vesicules (A, lanes a and g, respec-
tively), whereas the antibody against COX-2 peptide (PG 27)
did not immunoreact with any protein band of the unstimu-
lated cells (B, lane a). In contrast, after PMA treatment there
was a signi®cant increase of COX-2 protein expression in a
time-dependent manner (B, lanes b to f), whereas the weak
protein band corresponding to COX-1 was not modi®ed (A,
lanes b to f). COX-2 expression reached a maximum level after
6 to 8 h of PMA exposure (B, lanes d and e), followed by a
decrease at 24 h (B, lane f). The antibody against COX-2
peptide (PG 27) recognized a protein doublet of approximately
70 kDa corresponding to the migration of puri®ed COX-2
from sheep placenta and did not detect the puri®ed COX-1
from ram seminal vesicles (B, lanes h and g, respectively). We
determined the speci®city of the antiserums used in the above
experiments by using solubilized COX-1 from membranes of
human platelets, a cell type that is unable to synthesize in-
dubible proteins. The rabbit polyclonal antiserum against hu-
man COX-2 (PG27) did not immunoreact with the human
platelet COX-1 whereas the antiserum against ovine COX-1
(PG20) did crossreact with the human platelet COX-1 (data
not shown).
5
00
0
0
5
10
15
20
25
Tim e (h)
Figure 5 Time-course response of the COX activity in HUV-EC-C
treated with IL-1b or PMA. Cells were incubated for 0, 1, 3, 6, 9 and
24 h in the presence of 5 u ml
COX activity is expressed as pmol [ C]-AA transformed through
COX pathway by 10 cells after incubation with 25 mM substrate for
20 min. Data are shown as the mean, n=4; vertical lines show
s.e.mean. For abbreviations used see legends of Figures 2 and 3.
71
IL-1b (*) or 10 nM PMA (&).
14
6
to study whether the increase of eicosanoids observed was re-
lated to the de novo synthesis of COX-1 and/or COX-2 isoen-
zymes.
The e€ect of PMA on the transcription of both COX-1 and
COX-2 genes was determined by RT-PCR as described in
Methods. Results in Figure 7 show that after PMA treatment
there was a signi®cant increase (7 fold) in speci®c mRNA levels
of COX-2 in a time-dependent manner, whereas COX-1 spe-
ci®c mRNA levels were increased moderately (2 fold). The
speci®c mRNA levels for both isoenzymes reached a maximum
between 6 to 9 h after exposure to the stimuli. Afterwards, a
slight decrease in the levels of both COX-1 and COX-2 mRNA
was observed.
We tested the e€ect of known NSAIDS and new selective
COX-2 inhibitors on PMA-induced COX-2 activity in the
HUV-EC-C line (Table 1). IC50 values of COX-2 inhibition
were compared with those obtained for the inhibition of hu-
2
man platelet-dependent TXB production, a COX-1 selective
assay. In these models, ketorolac was the most potent inhibitor
of platelet COX and showed the greatest selectivity for COX-1.
Indomethacin was a potent inhibitor of COX-1 but was only 3
times more potent against COX-1 than COX-2. Aspirin was 13
times more active against COX-1, but was less potent than
indomethacin on either assay. Ibuprofen was also an inhibitor
of COX-1 and was 10 fold weaker towards COX-2 activity.
The Western blot analysis of the COX isoforms present in
the HUV-EC-C line is shown in Figure 8. We measured the

M. Miralpeix et al
COX-2 activity in HUV-EC-C cell line
177
The sulphonamide, NS-398, and the pyrazole, SC-58125, were
the most potent inhibitors of COX-2 and also presented the
highest selectivity for COX-2. Another sulphonamide, nime-
sulide, also showed COX-2 selectivity but was 10 times less
potent than NS-398 at inhibiting COX-2.
Discussion
We have found that the pattern of AA metabolism by the
permanent endothelial cell line, HUV-EC-C and of the HU-
VEC are quite similar although PGI (evaluated as 6-keto-
2
PGF1a) was not the main prostanoid produced by the HUV-
EC-C line. Under di€erent experimental conditions, including
endogenous AA and addition of exogenous AA, HUVEC
2 2
produce PGI , PGF2a and PGE as the main eicosanoids and
minor amounts of PGD
2
, HHT, 15-HETE and 11-HETE
a
0
0
0
0
0
0
0
0
0
.8
.7
.6
.5
.4
.3
.2
.1
.0
a
kDa
—
—
106
80
COX-1 ➝
—
—
49
32
a
b
c
d
e
f
g
h
0
5
10
15
20
25
b
Tim e (h)
—
—
106
80
COX-2 ➝
b
Tim e (h)
0
1
3
6
9
24 st
b.p.
—
49
1
1
358
078
8
72
— 32
6
63
COX-1
a
b
c
d
e
f
g
h
3
10
2
2
81
71
Figure 8 Selective expression of COX-2 isoenzyme in PMA-treated
HUV-EC-C. Quiescent con¯uent HUV-EC-C were incubated with
1
0 nM PMA for 0 (lane a), 1 (lane b), 3 (lane c), 6 (lane d), 8 (lane e)
and 24 h (lane f). For each condition, equal amounts of proteins
20 mg) were resolved by SDS ± PAGE on 8% acrylamide gels,
(
transferred to nitrocellulose membranes, and immunoblotted either
with anti-COX-1 (PG-20) (A) or anti-COX-2 (PG-27) (B) antiserums,
followed by enhanced chemiluminescence detection. Electrophoresis
standards of COX-1 (from ram seminal vesicles, 0.3 mg, lane g) and
COX-2 (from sheep placenta, 0.3 mg, lane h) were used as a reference.
Similar results were obtained with cell extracts from two di€erent
batches of cells.
COX-2
Table 1 Inhibition of HUV-EC-C PMA-induced COX-2
activity and human platelet COX-1 activity by COX
inhibitors
IC50 for
COX-1
(mM)
IC50 for
COX-2
(mM)
GAPDH
Ratio
COX-2/COX-1
Compound
Aspirin
1.2+0.3
0.05+0.02
0.0014+0.0003 0.14+0.09
0.34+0.3
12.5+3.1
28.9+16.1
13.3+1.9
15.8+3.9
0.15+0.06
13
3
100
11
0.03
0.001
0.006
Indomethacin
Ketorolac
Ibuprofen
Nimesulide
NS-398
Figure 7 Speci®c COX-1 and COX-2 mRNA expression normal-
ized to GAPDH in HUV-EC-C as a function of time of exposure
to PMA. Cells were incubated for 0, 1, 3, 6, 9 and 24 h in the
presence of 10 nM PMA. (a) Speci®c COX-1 (*) and COX-2 (&)
mRNA levels normalized to GAPDH levels are represented. Data
are the mean value of two independent experiments, vertical lines
indicate the limits of the individual values obtained. (b)
Representative gel photographs from PCR samples corresponding
to the indicated periods of time of treatment with PMA are
shown.
3.8+1.8
0.4+0.09
0.04+0.02
0.07+0.05
SC-58125
Results are expressed as mean+s.e.mean of the IC50 values
obtained from three independent experiments. Assay condi-
tions are described in Methods. The ratio of the IC50 values
for the compounds on COX-2 relative to COX-1 is given in
the far right column.

178
M. Miralpeix et al
COX-2 activity in HUV-EC-C cell line
(
1
Marcus et al., 1978; Alhenc-Gelas et al., 1982; Ku
985; Zavoico et al., 1989; Hopkins et al., 1984), all of them
derived from COX activity (Lopez et al., 1993). The HUV-EC-
C line, produced PGE and PGF2a as the main prostanoids
È
hn et al.,
To test the biochemical selectivity of various NSAIDs to-
wards COX-1 and COX-2, we compared their ability to inhibit
TXB
production by platelets and PGE production by PMA-
2 2
Â
2
stimulated HUV-EC-C respectively. In these models, aspirin,
indomethacin, ibuprofen and ketorolac inhibited preferentially
COX-1 as opposed to COX-2. Ketorolac showed the highest
potency for inhibition of COX-1 followed by indomethacin,
ibuprofen and aspirin, which order of potency correlates with
their tendency to cause gastric damage (Lanza, 1989; Bateman,
1994; Langman et al., 1994). Mitchell et al. (1994), using non-
human cultured cells as a source of COX-1 (bovine aortic en-
dothelial cells) and COX-2 (mouse macrophages), described
higher COX-2/COX-1 ratios for aspirin (166), indomethacin
(60) and ibuprofen (15) than those obtained in this study.
However, Klein et al. (1994) using human platelets as a source
of COX-1 and IL-1 stimulated rat mesangial cells as a source of
COX-2 obtained a COX-2/COX-1 ratio for indomethacin of
5.9 which was very close to the value obtained here. Surpris-
ingly, when the whole human blood system described by Pa-
trignani et al. (1994) was used to measure COX-1 and COX-2
activities, indomethacin was twice as potent as an inhibitor of
COX-2 than COX-1. Thus, it appears that the COX selectivity
of the drugs is very dependent on the assay system used.
NS-398 showed the lowest COX-2/COX-1 ratio, followed by
SC-58125 and nimesulide. NS-398 has been shown to inhibit
selectively the activity of puri®ed sheep COX-2 (Futaki et al.,
1994), recombinant murine COX-2 (Masferrer et al., 1994) and
recombinant human COX-2 (Gierse et al., 1995), with COX-2/
COX-1 ratios40.038, 40.0003 and 40.001, respectively. Re-
cently, Panara et al. (1995) have found that NS-398 is a COX-2
selective inhibitor with a ratio of 0.006 in platelets (COX-1) and
LPS-stimulated monocytes (COX-2) from human whole blood.
In the present study, we have shown that in human cell models
NS-398 has a COX-2/COX-1 ratio of 0.001, identical to that
obtained by Gierse et al. (1995). Similar to NS-398, SC-58125
was approximately a 200 fold more potent as an inhibitor of
COX-2 activity in PMA-stimulated HUV-EC-C than platelet
COX-1. Seibert et al. (1994) obtained similar results with
murine recombinant COX-1 (IC50410 mM) and COX-2
(IC50=0.05 mM). Finally, nimesulide showed a higher COX-2/
COX-1 ratio than NS-398 and SC-58125 in our cell models.
These results are consistent with those described by Barnett et
al. (1994) with human recombinant COX-1 and COX-2.
At present, controversial ®ndings regarding the selectivity of
standard NSAIDs have been described in the literature (Bat-
tistini et al., 1994; Griswold & Adams, 1996), which has ham-
pered progress in this ®eld. It has been suggested that cellular
models are more likely to re¯ect the selectivity exhibited by the
compounds in whole animal models and in clinical practice
(Mitchell et al., 1994). An ideal COX-2 model should couple the
use of cells from human origin with the ease and reproducibility
that permanent cell lines provide. Such a model will be of
bene®t in the study of COX-2 physiology as well as in the
screening of selective COX-2 inhibitors.
thereby exhibiting an AA metabolic pro®le similar to that
found in microvessel-derived endothelial cells (Bull et al., 1991;
Stanimirovic et al., 1993; Szczpanski et al., 1994). As shown in
HUVEC (Lopez et al., 1993), the formation of HETEs by the
Â
HUV-EC-C line was inhibited by indomethacin in a concen-
tration-dependent manner, which suggests that COX is the
enzyme responsible (Hecker et al., 1987).
IL-1b and PMA treatment of the HUV-EC-C line increased
the biosynthesis of prostanoids from exogenous AA. More-
over, COX activity was higher after treatment with PMA than
with IL-1b in HUV-EC-C. In HUVEC it has been clearly
shown that both PMA and IL-1 increased the synthesis of
prostanoids (Maier et al., 1990; Habib et al., 1993; Jones et al.,
1
993), although the induction of COX activity was greater with
IL-1a than in the presence of PMA (Jones et al., 1993). This
discrepancy could be explained by the fact that Jones and
colleagues measured COX activity in HUVEC only by the
production of 6-keto-PGF1a, and it may be that prostanoids,
2 2
such as PGE , increased more than PGI by the action of PMA.
Low levels of COX-1 transcripts are expressed in quiescent
HUV-EC-C line while COX-2 transcripts were undetectable.
PMA stimulation of HUV-EC-C increased signi®cantly the
mRNA levels of COX-2 whereas only a modest increase in
COX-1 mRNA was observed. Interestingly, only COX-2
protein was enhanced after the exposure to the phorbol ester
and the expression of COX-2 mRNA and protein remained
evident after 24 h. These data suggest that in the HUV-EC-C
line, (i) only the COX-1 gene is expressed by quiescent cells; (ii)
PMA induces the COX-2 gene to a much higher extent that the
COX-1 gene; (iii) PMA induces only the expression of the
COX-2 protein; and (iv) in PMA-treated cells there is a cor-
relation between COX activity, COX-2 mRNA levels and
COX-2 protein mass. Taken together, these data indicate that
PMA selectively induces a functionally active COX-2 protein
in the HUV-EC-C line.
The expression of COX-1 mRNA without modi®cation of
COX-1 protein has also been observed in mouse NIH3T3 cells
stimulated by serum (De Witt & Meade, 1993), in EGV-6 cells
(
a tracheal epithelial cell line) stimulated with PMA (Kitzler et
al., 1995) and in HUVEC stimulated with PMA (Habib et al.,
993) although Xu et al. (1996) showed that COX-1 protein is
1
also enhanced by PMA in HUVEC. These results suggest a
post-transcriptional regulation of COX-1 expression in the
HUV-EC-C line and/or a speci®c induction of nonfunctional,
aberrantly spliced COX-1 mRNA, as has been observed in
EGV-6 cells (Kitzler et al., 1995).
Unstimulated HUV-EC-C express undetectable levels of the
COX-2 mRNA and protein. In contrast, there are several
studies that suggest that HUVEC express signi®cant levels of
COX-2 mRNA and protein under basal conditions (Hla &
Neilson, 1992; Jones et al., 1993; Camacho et al., 1995). Fur-
thermore, in HUV-EC-C the e€ect of PMA on COX-2 ex-
pression, both at the mRNA and protein levels, persisted for a
longer time than in HUVEC (Hla & Neilson, 1992; Habib et
al., 1993).
Our results show that the HUV-EC-C line has a very low
basal COX activity and that the dramatic increase in activity
seen when stimulated with PMA is likely to re¯ect selective
COX-2 expression. The main eicosanoid formed by HUV-EC-
C, PGE
treated with 10 nM PMA, when compared with controls.
Therefore, evaluation of the formation of PGE by ELISA in
2
, increased at least 10 fold in concentration in cells
It is currently believed that the unwanted side-e€ects of
NSAIDs are due to inhibition of COX-1 and the therapeutic
e€ects are due to their action on the COX-2 isoenzyme (Vane
2
PMA stimulated HUV-EC-C is a suitable experimental strat-
egy to determine the action of NSAIDs on COX-2 activity.
&
Botting, 1995). Drugs which have the highest potency for
inhibition of COX-2 and a more favourable COX-2/COX-1
activity ratio may have potent anti-in¯ammatory activity with
fewer side e€ects in the stomach and kidney. Conversely to the
classical NSAIDs, COX-2 selective inhibitors such as nime-
sulide, NS-398, SC-58125 among others are anti-in¯ammatory
and analgesic agents that do not produce gastrointestinal le-
sions following oral administration (Arai et al., 1993; Davis &
Brogden, 1994; Seibert et al., 1994). This is likely to be due to
their potency and relative selectivity towards inhibition of
COX-2 isoenzyme.
This work was supported by grants from Institut de Recerca of the
HSCSP, and from Laboratorios Almirall S.A. We thank Drs C. Puig
and F. Pujol at the Department of Chemical Synthesis of Almirall
Laboratories for synthesizing the non-commercial available drugs
used in this study. We would also like to thank Ms M.C. Cabello,
Ms J.E. Dõaz, Ms C. Gerboles and Ms E. Gerboles for excellent
  Â
technical assistance.

M. Miralpeix et al
COX-2 activity in HUV-EC-C cell line
179
References
ALHENC-GELAS, F., TSAI, S.J., CALLAHAN, K.S., CAMPBELL, W.B.
JOHNSON, A.R. (1982). Stimulation of prostaglandin formation
GIERSE, J.K., HAUSER, S.D., CREELY, D.P., KOBOLDT, C., RANG-
WALA, S.H., ISASKSON, P.C. & SEIBERT, K. (1995). Expression
and selective inhibition of the constitutive and inducible forms of
human cyclooxygenase. Biochem. J., 305, 479 ± 484.
&
by vasoactive mediators in cultured human endothelial cells.
Prostaglandins, 24, 723 ± 742.
ARAI, I., HAMASAKA, Y., FUTAKI, N., TAKAHASHI, S., YOSHIKA-
WA, K., HIGUCHI, S. & OTOMO, S. (1993). E€ect of NS-398, a new
nonsteroidal anti-in¯ammatory agent, on gastric ulceration and
acid secretion in rats. Res. Commun. Chem. Pathol. Pharmacol.,
GOPPELT-STRUEBE, M. (1995). Regulation of prostaglandin en-
doperoxide synthase (cyclooxygenase) isozyme expression.
Prostaglandins Leukotrienes Essential Fatty Acids, 5, 213 ± 222.
GRISWOLD, D.E. & ADAMS, J.L. (1996). Constitutive cyclooxygenase
(COX-1) and inducible cyclooxygenase (COX-2): rationale for
selective inhibition and progress to date. Med. Res. Rev., 16,
181 ± 206.
8
1, 259 ± 269.
BAKOVIC, M. & DUNFORD, H.B. (1994). Intimate relation between
cyclooxygenase and peroxidase activities of prostaglandin H
synthase. Peroxidase reaction of feluric acid and its in¯uence on
the reaction of arachidonic acid. Biochemistry, 33, 6475 ± 6482.
BARNETT, J., CHOW, J., IVES, D., CHIOU, M., MACKENZIE, R., OSEN,
E., NGUYEN, B., TSING, S., BACH, C., FREIRE, J., CHAN, H.,
SIGAL, E. & RAMESHA, C. (1994). Puri®cation, characterization
and selective inhibition of human prostaglandin G/H synthase 1
and 2 expressed in the baculovirus system. Biochim. Biophys.
Acta, 1209, 130 ± 139.
HABIB, A., CREMINON, C., FROBERT, Y., GRASSI, J., PRADELLES,
& MACLOUF, J. (1993). Demonstration of an inducible
P.
cyclooxygenase in human endothelial cells using antibodies
raised against the carboxyl-terminal region of the cyclooxygen-
ase-2. J. Biol. Chem., 268, 23448 ± 23454.
HAMBERG, M.
& SAMUELSSON, B. (1967). Oxygenation of
unsaturated fatty acids by the vesticular gland of sheep. J. Biol.
Chem., 242, 5344 ± 5354.
BATEMAN, N. (1994). NSAIDs: time to re-evaluate gut toxicity.
Lancet, 343, 1051 ± 1052.
BULL, H., COHEN, J. & DOWD, P.M. (1991). Responses of human
dermal microvascular endothelial cells to histamine and their
HAURAND, M. & ULLRICH, V. (1985). Isolation and characterization
of thromboxane synthase from human platelets as a cytochrome
P-450 enzyme. J. Biol. Chem., 260, 15059 ± 15067.
HECKER, M.
& ULLRICH, V. (1989). On the mechanism of
modulation by interleukin
Dermatol., 97, 787 ± 792.
1
and substance P. J. Invest.
2
prostacyclin and thromboxane A biosynthesis. J. Biol. Chem.,
264, 141 ± 150.
BATTISTINI, B., BOTTING, R. & BAKHLE, Y.S. (1994). COX-1 and
COX-2: toward the development of more selective NSAIDs.
Drug News & Perspectives, 7, 501 ± 512.
HECKER, M., ULLRICH, V., FISCHER, C. & MEESE, C.O. (1987).
Identi®cation of novel arachidonic acid metabolites formed by
prostaglandin H synthase. Eur. J. Biochem., 169, 113 ± 123.
HLA, T. & NEILSON, K. (1992). Human cyclooxygenase-2 cDNA.
Proc. Natl. Acad. Sci. U.S.A., 89, 7384 ± 7388.
HOPKINS, N.K., OGLESBY, T.D., BUNDY, G.L. & GORMAN, R.R.
(1984). Biosynthesis and metabolism of 15-hydroperoxy-
5,8,11,13- eicosatetraenoic acid by human umbilical vein
endothelial cells. J. Biol. Chem., 259, 14048 ± 14053.
JONES, D.A., CARLTON, D., MCINTYRE, M.T., ZIMMERMAN, G.A. &
PRESCOTT, S.M. (1993). Molecular cloning of human prostaglan-
dinendoperoxidesynthasetypeIIanddemonstrationofexpression
in response to cytokines. J. Biol. Chem., 268, 9049 ± 9054.
KADUCE, T.L., FIGARD, P.H., LEIFUR, R. & SPECTOR, H.H. (1989).
Formation of 9-hydroxyoctadecadienoic acid from linoleic acid
in endothelial cells. J. Biol. Chem., 264, 6823 ± 6830.
Â
Â
CAMACHO, M., GODESSART, N., ANTON, R., GARCIA, M. & VILA, L.
1995). Interleukin-1 enhances the ability of cultured umbilical
vein endothelial cells to oxidize linoleic acid. J. Biol. Chem., 270,
7279 ± 17286.
(
1
CAPDEVILA, J.H., MORROW, J.D., BELOSLUDTSEV, Y.Y., BEU-
CHAMP, D.R., DUBOIS, R.N. & FALK, J.R. (1995). The Catalytic
outcomes of constitutive and the mitogen inducible isoforms of
2
prostaglandin H synthase are markedly a€ected by glutathione
and glutathione peroxidase(s). Biochemistry, 34, 3325 ± 3337.
CHAN, C.-C., BOYCE, S., BRIDEAU, C., FORD-HUTCHINSON, A.W.,
GORDON, R., GUAY, D., HILL, R.G., LI, C.-S., MANCINI, J.,
PENNETON, M., PRASIT, P., RASORI, R., RIENDEAU, D., ROY, P.,
TAGARI, P., VICKERS, P., WONG, E. & RODGER, I.W. (1995).
Pharmacology of a selective cyclooxygenase-2 inhibitor, L-
KITZLER, J., HILL, E., HARDMAN, R., REDDY, N., PHILPOT, R. &
ELING, T.E. (1995). Analysis and quantitation of splicing variants
of the TPA-inducible PGHS-1 mRNA in rat tracheal epithelial
cells. Arch. Biochem. Biophys. 316, 856 ± 863.
7
45,337: A novel nonsteroidal anti-in¯ammatory agent with an
ulcerogenic sparing e€ect in rat and nonhuman primate stomach.
J. Pharmacol. Exp. Ther., 274, 1531 ± 1537.
CHOMCZYNSKI, P. & SACCHI, N. (1987). Single-step method of
RNA isolation by acid guanidiniumthiocyanate-phenol-chloro-
form extraction. Anal. Biochem., 162, 156 ± 159.
COPELAND, R.A., WILLIAMS, J.M., GIANNARAS, J., NURNBERG, S.,
COVINGTON, M., PINTO, D., PINK, S. & TRZASKOS, J.M. (1994).
Mechanism of selective inhibition of the inducible isoform of
prostaglandin G/H synthase. Proc. Natl. Acad. Sci. U.S.A., 91,
KLEIN, T., NUSING, R.M., SCHIFTER, J.P. & ULLRICH, V. (1994).
È
Selective inhibition of cyclooxygenase 2. Biochem. Pharmacol.,
48, 1605 ± 1610.
KU
È
HN, H., PO
FORSTER, W. (1985). Metabolism of [1- C] - arachidonic acid by
È
NICKE, K., HALLE, W., WIESNER, R., SCHEWE, T. &
1
4
È
cultured calf aortic endothelial cells: evidence for the presence of
a lipoxygenase pathway. Prostagland. Leuk. Med., 17, 291 ± 303.
KUJUBU, D.A., FLETCHER, B.S., VARNUM, B.C., LIM, R.W. &
1
1202 ± 11206.
DAVIS, R. & BROGDEN, R.N. (1994). Nimesulide. An update of its
pharmacodynamic and pharmacokinetic properties, and ther-
apeutic ecacy. Drugs, 48, 431 ± 454.
DE WITT, D.L. (1991). Prostaglandin endoperoxide synthase:
regulation of enzyme expression. Biochim. Biophys. Acta, 1083,
HERSCHMAN, H.R. (1991). TIS10, a phorbol ester tumour
promoter inducible mRNA from Swiss 3T3 cells, encodes a
novel prostaglandin synthase/cyclooxygenase homologue. J.
Biol. Chem., 266, 12866 ± 12872.
KUJUBU, D.A. & HERSCHMAN, H.R. (1992). Dexamethasone inhibits
mitogen induction of the TIS10 prostaglandin synthase/cycloox-
ygenase gene. J. Biol. Chem., 267, 7991 ± 7994.
KUJUBU, D.A., REDDY, S.T., FLETCHER, B.S. & HERSCHMAN, H.R.
(1993). Expression of the protein product of the prostaglandin
synthase-2/TIS10 gene in mitogen-stimulated Swiss 3T3 cells. J.
Biol. Chem., 268, 5425 ± 5430.
1
21 ± 134.
DE WITT, D.L. & MEADE, E.A. (1993). Serum and glucocorticoid
regulation of gene transcription and expression of the prosta-
glandin H synthase-1 and prostaglandin H synthase-2 isozymes.
Arch. Biochem. Biophys., 94, 94 ± 102.
FLETCHER, B.S., KUJUBU, D.A., PERRIN, D.M. & HERSCHMAN,
H.R. (1992). Structure of the mitogen-inducible TIS10-encoded
protein is a functional prostaglandin G/H synthase. J. Biol.
Chem., 267, 438 ± 444.
FLOWER, R.J. (1974). Drugs which inhibit prostaglandin biosynth-
esis. Pharmacol. Rev., 26, 33 ± 67.
FUNK, C.D., FUNK, L.B., KENNEDY, M.E., PONG, A.S. & FITZGER-
ALD, G.A. (1991). Human platelet/erythroleukemia cell prosta-
glandin G/H synthase: cDNA cloning, expression, and gene
chromosomal assignment. FASEB J., 5, 2304 ± 2312.
KULMACZ, R.J., PENDLETON, R.B.
& LANDS, W.E.M. (1994).
Interaction between peroxidase and cyclooxygenase activities in
prostaglandin-endoperoxide synthase. Interpretation of reaction
kinetics. J. Biol. Chem., 269, 5527 ± 5536.
KULMACZ, R.J. & WANG, L.-H. (1995). Comparison of hydroper-
oxide initiator requirements for the cyclooxygenase activities of
prostaglandin H synthase-1 and -2. J. Biol. Chem., 270, 24019 ±
24023.
KURUMBAIL, R.G., STALLINGS, W.C., MCDONALD, STEVENS,
A.M., STEGEMAN, R.A., GIERSE, J.K., MIYASHIRO, J.M., PEN-
FUTAKI, N., TAKAHASHI, M., YOKOYAMA, M., ARAI, I., HIGUCHI,
S. & OTOMO, S. (1994). NS-398, a new anti-in¯ammatory agent,
selectively inhibits prostaglandin G/H synthase/cyclooxygenase
NING, T.D., SEIBERT, K.
& ISAKSON, P.C. (1996). Crystal
structure of murine cyclooxygenase-2. Prostaglandins, Leuko-
trienes and Essential Fatty Acids, 55 (suppl. 1), 20.
(
COX-2) activity in vitro. Prostaglandins, 47, 55 ± 59.

180
M. Miralpeix et al
COX-2 activity in HUV-EC-C cell line
LACOMTE, M., LANEUVILLE, O., DE WITT, D.L.
&
SMITH, W.
1994). Acetylation of human prostaglandin endoperoxide
synthase-2 (cyclooxygenase-2) by aspirin. J. Biol. Chem., 269,
3207 ± 13215.
PANARA, M.R., GRECO, A., SANTINI, G., SCIULLI, M.G., ROTONDO,
M.T., PADOVANO, R., DI GIAMBERARDINO, M., CIPOLLONE, F.,
CUCCURULLO, F., PATRONO, C. & PATRIGNANI, P. (1995).
E€ects of the novel anti-in¯ammatory compounds, N-[2-cyclo-
hexyloxy)-4-nitrophenyl]metansulphonamide (NS-398) and 5-
metansulphonamide-6-(2,4-di¯uorothiophenyl)-1-indanone (L-
745,337), on the cyclooxygenase activity of human blood
prostaglandin endoperoxide synthases. Br. J. Pharmacol., 116,
2429 ± 2434.
(
1
LAEMMLI, E.K. (1970). Cleavage of structural proteins during the
assembly of the head of bacteriophage T4. Nature, 227, 680 ±
6
85.
LANEUVILLE, O., BREUER, D.K., XU, N., HUANG, Z.H., GAGE, D.A.,
WATSON, J.T., LAGARDE, M., DE WITT, D.L. & SMITH, W. (1995).
Fatty acid substrate speci®cities of human prostaglandin-
endoperoxide H synthase-1 and -2. Formation of 12-hydroxy-
PATRIGNANI, P., PANARA, M.R., GRECO, A., FUSCO, O., NATOLI,
C., IACOBELLI, S., CIPOLLONE, F., GANCI, A., CREMINON, C.,
(
Biol. Chem., 270, 19330 ± 19336.
9Z,13E/Z,15Z)-octadecatrienoic acids from a-linoleic acid. J.
MACLOUF, J.
& PATRONO, C. (1994). Biochemical and
pharmacological characterization of the cyclooxygenase activity
of human blood prostaglandin endoperoxide synthases. J.
Pharmacol. Exp. Ther., 271, 1705 ± 1712.
LANGMAN, M.J.S., WEIL, J., WAINWRIGHT, P., LAWSON, D.H.,
RAWLINS, M.D., LOGAN, R.F.A., MURPHY, M., VESSEY, M.P. &
COLIN-JONES, D.G. (1994). Risks of bleeding peptic ulcer
associated with individual non-steroidal anti-in¯ammatory
drugs. Lancet, 343, 1075 ± 1078.
QUELLET, M. & PERCIVAL, D. (1995). E€ect of inhibitor time-
dependency on selectivity towards cyclooxygenase isoforms.
Biochem. J., 306, 247 ± 251.
LANZA, F.L. (1989). A review of gastric ulcer and gastroduodenal
injury in normal volunteers receiving aspirin and other non-
steroidal anti-in¯ammatory drugs. Scand. J. Gastroenterol, 24
SEIBERT, K., ZHANG, Y., LEAHY, K., HAUSER, S., MASFERRER, J.,
PERKINS, W., LEE, L. & ISAKSON, P. (1994). Pharmacological and
biochemical demonstration of the role of cyclooxygenase 2 in
in¯ammation and pain. Proc. Natl. Acad. Sci. U.S.A., 91, 12013 ±
12017.
SIROIS, J. & RICHARDS, J.S. (1992). Puri®cation and characterization
of a novel, distinct isoform of prostaglandin endoperoxide
synthase induced by human chorionic gonadotropin in granulosa
cells of rat preovulatory follicles. J. Biol. Chem., 267, 6382 ± 6388.
SIROIS, J., SIMMONS, D.L. & RICHARDS, J.S. (1992). Hormonal
regulation of messenger ribonucleic acid encoding a novel
(
suppl. 163), 24 ± 31.
LEE, S.H., SOYOOLA, E., CHANMUGAM, P., HART, S., SUN, W.,
ZHONG, H., LIOU, S., SIMMONS, D. HWANG, D. (1992).
&
Selective expression of mitogen-inducible cyclooxygenase in
macrophages stimulated with lipopolysaccharide. J. Biol. Chem.,
2
LOPEZ, S., VILA, L., BREVIARIO, F. & DE CASTELLARNAU, C.
67, 25934 ± 25938.
Â
(
1993). Interleukin-1 increases 15-hydroxyeicosatetraenoic acid
formation in cultured human endothelial cells. Biochim. Biophys.
Acta, 1170, 17 ± 24.
MAIER, J.A.M., HLA, T. & MACIAG, T. (1990). Cyclooxygenase is an
immediate-early gene induced by interleukin-1 in human
endothelial cells. J. Biol. Chem., 265, 10805 ± 10808.
MARCUS, A.J., WEKSLER, B.B. & JAFFE, E.A. (1978). Enzymatic
isoform of prostaglandin endoperoxide H synthase in rat
preovulatory follicles. J. Biol. Chem., 267, 11586 ± 11592.
SMITH, W.L. (1992). Prostanoid biosynthesis and mechanisms of
action. Am. J. Physiol., 263, F181 ± F191.
SMITH, W.L., MARNETT, L.J. & DE WITT, D.L. (1991). Prostaglandin
and thromboxane biosynthesis. Pharmacol. Ther., 49, 153 ± 179.
Â
SOLA, J., GODESSART, N., VILA, L., PUIG, L. & DE MORAGAS, J.M.
(1992). Epidermal cell-polymorphonuclear leukocyte coopera-
conversion of prostaglandin endoperoxide H
2
and arachidonic
acid to prostacyclin by cultured human endothelial cells. J. Biol.
Chem., 253, 7138 ± 7141.
MASFERRER, J.L., SEIBERT, K., ZWEIFEL, B. & NEEDLEMAN, P.
4
tion in the formation of leukotriene B by transcellular
biosynthesis. J. Invest. Dermatol., 98, 333 ± 339.
(
1992). Endogenous glucocorticoids regulate an inducible
cyclooxygenase enzyme. Proc. Natl. Acad. Sci. U.S.A., 89,
917 ± 3921.
SPANGLER, R.S. (1993). Gastrointestinal damage demonstrated with
nabumetone or etodolac in preclinical studies. Am. J. Med., 95
(suppl 2A), 35S ± 39S.
3
MASFERRER, J.L., ZWEIFEL, B.S., MANNING, P.T., HAUSER, S.D.,
LEAHY, K.M., SMITH, W.G., ISAKSON, P.C. & SEIBERT, K. (1994).
Selective inhibition of inducible cyclooxygenase 2 in vivo is
antiin¯ammatory and nonulcerogenic. Proc. Natl. Acad. Sci.
U.S.A., 91, 3228 ± 3232.
MCKEEVER, B.M., PANDYA, S., PERCIVAL, M.D., OULLET, M.,
BAYLY, C., O'NEILL, G.P., BASTIEN, L., KENNEDY, B., ADAM,
M., CROMLISH, W., ROY, P., BLACK, C., GUAY, D. & LEBLANC,
Y. (1996). The X-ray crystal structure of recombinant human
prostaglandin G/H synthase-2 at 3.0 A resolution. Prostaglan-
dins, Leukotrienes and Essential Fatty Acids, 55 (suppl. 1), 20.
MEADE, E.A., SMITH, W.L. & DE WITT, D.L. (1993). Di€erential
inhibition of prostaglandin endoperoxide synthase (cyclooxy-
genase) isoenzymes by aspirin and other non-steroidal antiin-
STANIMIROVIC, D.B., BACIC, F., UEMATSU, S. & SPATZ, M. (1993).
Pro®le of prostaglandins induced by endothelin-1 in human brain
capillary endothelium. Neurochem. Int., 23, 385 ± 393.
SZCZEPANSKI, A., MOATTER, T., CARLEY, W.W. & GERRITSEN,
M.E. (1994). Induction of cyclooxygenase II in human synovial
microvessel endothelial cells by interleukin-1. Inhibition by
glucocorticoids. Arthritis Rheum., 37, 495 ± 503.
TANABE, T. & ULLRICH, V. (1995). Prostacyclin and thromboxane
synthases. J. Lipid Mediators Cell Signalling, 12, 243 ± 255.
TOWBIN, H., STAEHELIN, T. & GORDON, J. (1979). Electrophoretic
transfer of proteins from polyacrylamide gel to nitrocellulose
sheet: procedure and some applications. Proc. Natl. Acad. Sci.
U.S.A., 76, 4350 ± 4354.
URADE, Y., WATANABE, K. & HAYAISHI, O. (1995). Prostaglandin D,
E,andFsynthases.J.LipidMediatorsCellSignalling,12,257 ± 273.
VANE, J.R. (1971). Inhibition of prostaglandins as a mechanism of
action for aspirin-like drugs. Nature, 231, 232 ± 235.
VANE, J.R. & BOTTING, R.M. (1995). New insights into the mode of
action of anti-in¯ammatory drugs. In¯amm. Res., 44, 1 ± 10.
WLODAWER, P. & SAMUELSSON, B. (1973). On the organization and
mechanism of prostaglandin synthetase. J. Biol. Chem., 248,
5673 ± 5678.
¯
ammatory drugs. J. Biol. Chem., 268, 6610 ± 6614.
MITCHELL, J.A., AKARASEREENONT, P., THIEMERMANN, C.,
FLOWER, R.J. & VANE, J. (1994). Selectivity of nonsteroidal
antiin¯ammatory drugs as inhibitors of constitutive and
inducible cyclooxygenase. Proc. Natl. Acad. Sci. U.S.A., 90,
1
1693 ± 11697.
NUGTEREN, D.H. & CHRIST-HAZELHOF, E. (1980). Chemical and
2
enzymatic conversions of prostaglandin endoperoxide PGH .
Adv. Prostaglandin Thromboxane Res., 6, 129 ± 137.
XU, X.-M., TANG, J.-L., HAJIBEIGI, A., LOOSE-MITCHELL, D.S. &
WU, K.K. (1996). Transcriptional regulation of endothelial
constitutive PGHS-1 expression by phorbol ester. Am. J.
Physiol., 270 (Cell Physiol, 39), C259 ± C264.
O'BANNION, M.K., WINN, V.D. & YUNG, D.A. (1992). cDNA cloning
and functional activity of a glucocorticoid-regulated in¯amma-
tory cyclooxygenase. Proc. Natl. Acad. Sci. U.S.A., 89, 4888 ±
4
892.
YOKOYAMA, C. & TANABE, T. (1989). Cloning of human gene
encoding prostaglandin endoperoxide synthase and primary
structure of the enzyme. Biochem. Biophys. Res. Commun., 165,
888 ± 894.
O'NEILL, G.P. & FORD-HUTCHINSON, A.W. (1993). Expression of
mRNA for cyclooxygenase-1 and cyclooxygenase-2 in human
tissues. FEBS Letts, 330, 156 ± 160.
O'SULLIVAN, M.G., CHILTON, F.H., HUGGINS, E.M. & MCCALL,
C.E. (1992a). Lipopolysaccharide priming of alveolar macro-
phages for enhanced synthesis of prostanoids involves induction
of novel prostaglandin H synthase. J. Biol. Chem., 267, 14547 ±
ZAVOICO, G.B., EWENSTEIN, B.M., SCHAFER, A.I. & POBER, J.S.
(1989). IL-1 and related cytokines enhance thrombin-stimulated
2
PGI , production in cultured endothelial cells without a€ecting
thrombin-stimulated von Willebrand factor secretion or platelet-
activating factor biosynthesis. J. Immunol., 142, 3993 ± 3999.
1
4550.
O'SULLIVAN, M.G., HUGGINS, E.M., MEADE, E.A., DE WITT, D.L. &
MCCALL, C.E. (1992b). Lipopolysaccharide induces prostaglan-
din H synthase-2 in alveolar macrophages. Biochem. Biophys.
Res. Commun., 187, 1123 ± 1127.
(Received July 23, 1996
Revised January 27, 1997
Accepted February 5, 1997)