FASEB J. Mp Biomedicals
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published as doi: 10.1096/fj.08-106591.
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF) Free
Right arrow All Versions of this Article:
fj.08-106591v1
23/2/605    most recent
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Bolego, C.
Right arrow Articles by Sala, A.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Bolego, C.
Right arrow Articles by Sala, A.
(The FASEB Journal. 2009;23:605-612.)
© 2009 FASEB

Critical role of COX-1 in prostacyclin production by human endothelial cells under modification of hydroperoxide tone

Chiara Bolego*,1, Carola Buccellati*, Alberto Prada{dagger}, Rosa Maria Gaion{ddagger}, Giancarlo Folco* and Angelo Sala*,2

* Department of Pharmacological Sciences, University of Milan, Milan, Italy;

{dagger} Rho Hospital, Rho, Italy; and

{ddagger} Department of Pharmacology and Anesthesiology, University of Padova, Padova, Italy

2Correspondence: Department of Pharmacological Sciences, University of Milan, Via Balzaretti 9, 20133 Milan, Italy. E-mail: angelo.sala{at}unimi.it


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
We aimed at evaluating the relative contribution of cyclooxygenase (COX) -1 and COX-2 to the synthesis of prostacyclin in endothelial cells under static conditions in the presence or absence of exogenous arachidonic acid and/or altered intracellular redox balance. Selective inhibitors of either COX-1 (SC560 and FR122047) or COX-2 (SC236) concentration dependently (1–300 nM) reduced basal and interleukin (IL) -1β-induced prostacyclin production in human umbilical vein endothelial cells by 70% or more; compound selectivity was confirmed using a whole-blood assay (IC50 COX-1/COX-2: 13 nM/930 nM for SC-560; 9 µM/457 nM for SC-236). The observed concomitant formation of isoprostane appeared to be associated with COX enzyme activity, while formation of COX-1/COX-2 heterodimers was detected by immunoprecipitation. In the presence of arachidonic acid and 12-hydroperoxy-eicosatetraenoic acid, either exogenous or provided by platelet activation, or after glutathione depletion, COX-1 inhibition but not COX-2 inhibition concentration dependently decreased prostacyclin production. Both isoforms appear to contribute to basal prostacyclin production by endothelial cells, with COX-2 providing the hydroperoxide tone required for COX-1 activity. Conversely, in the case of intracellular glutathione depletion or enhanced availability of arachidonic acid and hydroperoxides, selective COX-2 inhibition did not significantly affect the production of endothelial prostacyclin. These findings contribute to a better understanding of the effects of cyclooxygenase inhibitors on prostacyclin production.—Bolego, C., Buccellati, C., Prada, A., Gaion, R. M., Folco, G., Sala, A. Critical role of COX-1 in prostacyclin production by human endothelial cells under modification of hydroperoxide tone.


Key Words: selective cyclooxygenase inhibitors • prostaglandin • isoprostanes • COX-2 heterodimers • 12-HpETE


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
CYCLOOXYGENASE (COX), the rate-limiting enzyme in the synthesis of biologically relevant prostanoids, such as thromboxane and prostacyclin, exists as two isoforms encoded by two distinct genes: COX-1 is constitutively expressed in most tissues and mediates basal physiological functions, while COX-2 is induced by various stimuli, such as inflammatory cytokines, and is therefore mostly associated with pathological conditions (1) . In unstimulated human endothelial cells, COX-1 was reported to be mainly responsible for TXB2 production, whereas COX-2, detectable only after stimulation, was solely involved in prostaglandin I2 (PGI2) and prostaglandin E2 (PGE2) synthesis (2) . Traditional nonsteroidal anti-inflammatory drugs (NSAIDs) inhibit both isoforms, while the COXIBs selectively affect the formation of prostanoids by COX-2. The risk-benefit ratio associated with their use is strictly dependent on the distribution of the enzyme isoforms in target organs, such as the gastrointestinal tract, kidneys, platelets, and endothelium (3) . COX-1 and COX-2 activity is also differentially regulated by factors such as the amount of substrate (i.e., arachidonic acid), either released from cellular phospholipids via cytosolic phospholipase A2{alpha} (cPLA2{alpha}; ref. 4 ) or provided by neighboring cells (5 6 7) , or the availability of hydroperoxides (8) . The housekeeping prostanoid produced by the endothelium is PGI2, which notably contributes to the maintenance of the physiological, antiadhesive, and antithrombotic properties of this organ (9) . Although a large body of data suggests that endothelial cells express COX-1, which in turn is the isoform responsible for the constitutive production of PGI2 (2 , 10 , 11) , there is now a general agreement that PGI2 in vascular endothelium may be generated mainly by COX-2, possibly as a result of COX-2 expression induced by laminar flow shear stress (12 , 13) . Nevertheless, this hypothesis has been recently challenged (14) , as the evidence of the in vivo vascular expression of COX-2 under normal conditions is still controversial (15 , 16) . Studies (8) in cells expressing both COX isoforms demonstrated that both enzymes may contribute to prostaglandin synthesis, while important regulatory mechanisms (i.e., hydroperoxide tone and substrate availability) can contribute to control their catalytic activity with striking differences observed between COX-1 and COX-2. Moreover, a recent study (17) disclosed the in vivo formation of COX heterodimers, thus opening the question of their potential role toward the synthesis of prostanoids or the activity of COX inhibitors.

With the present work, we investigated the relative contribution of the two COX isoforms toward prostacyclin production in human endothelium under static conditions, evaluating how selective COX-1 and COX-2 inhibitors affect its production, as well as that of isoprostanes. Furthermore, we studied the formation of COX-1/COX-2 heterodimers and the effect of enhanced substrate availability and alterations of the endothelial peroxide tone on the synthesis of prostacyclin.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 REFERENCES
 
Endothelial cell culture
Human umbilical vein endothelial cells (HUVECs) were grown in medium 199 (M199, Invitrogen, S. Giuliano Milanese, Milan, Italy) supplemented with 15% FCS (Euroclone; Pero, Milan, Italy), gentamicin (40 µg/ml, Invitrogen), endothelial cell growth factor (25 µg/ml; Sigma, St. Louis, MO, USA), and heparin (100 µg/ml, Sigma). Cells were identified as endothelial by their morphology and the presence of CD31-related antigen. All experiments were performed on cells at the first passage. HUVECs were seeded at equal density either in 24-well plates (7x104/well) or in 6-well plates (3x105/well) or in 100-mm culture dishes (1.4x106/dish). Before experiments, cells were maintained overnight in serum-free medium enriched with 0.5% albumin. For experiments, cells were incubated in M199 supplemented with 5% FCS or in PBS as indicated. Selected experiments were carried out in the presence of buthionine sulfoximine (BSO; 1 mM; Sigma), diethyl maleate (DEM; 0.5 mM; Sigma), SC236 (1–100 nM; Searle, St. Louis, MO, USA), arachidonic acid (10–50 µM; Cayman Chemical, Ann Arbor, MI, USA), SC560 (1–100 nM; Cayman Chemical), FR122047 (1–300 nM; Cayman Chemical), and 12-hydroperoxyeicosatetraenoic acid (12-HpETE; 50 nM; Cayman Chemical). Inhibitors were added 10–15 min before the stimulus.

Human platelet isolation and coincubation with HUVECs
Human blood was taken from the antecubital vein of healthy volunteers, anticoagulated with ACD (84 mM trisodium citrate, 41 mM citric acid, and 136 mM glucose, 1:7, v:v) and treated with 1 mM acetylsalicylic acid. Platelet-rich plasma was obtained by centrifugation at 180 g (15 min at room temperature), which was further centrifuged at 650 g (15 min at room temperature); resuspended in a washing buffer (5 mM KCl, 103 mM NaCl, 1 mM MgCl2, 2 mM CaCl2, 39 mM citric acid, and 5 mM glucose and NaOH 10 N to pH 6.5); centrifuged again; and finally resuspended in D-PBS with CaCl2 and MgCl2 (1x108 platelets/ml).

After incubation with inhibitors for 15 min, the HUVEC medium was substituted with a platelet suspension (108 platelets/ml, 2 ml/well) containing the inhibitor used, and the coincubation was stimulated with arachidonic acid (10 µM). Thirty minutes after challenge, the whole medium was collected, centrifuged at 12,000 g (2 min at 4°C), and analyzed by specific enzyme immunoassay (EIA) to quantify 6-keto-PGF1{alpha}.

COX-1/COX-2 whole-blood selectivity assay
Peripheral venous blood was collected from healthy volunteers, divided into two parts, and used for evaluating COX-1/COX-2 selectivity of different compounds according to the method by Patrignani et al. (18) . Briefly, 1-ml aliquot samples of whole-blood samples, containing 50 µM acetylsalycilic acid and 10 IU heparin (Sigma), were incubated with LPS (10 µg/ml; Sigma) for 24 h at 37°C, both in the absence and presence of inhibitors (SC560 and SC236: 1–100 nM). Plasma was separated by centrifugation (10 min, 2500 g) and assayed for PGE2. A second 1-ml aliquot of whole blood was immediately transferred into glass tubes containing either the inhibitors (SC560 and SC236) or their solvents and was allowed to clot at 37°C for 1 h. Serum was separated by centrifugation (10 min at 2500 g) and assayed for TXB2.

Determination of PGI2, PGE2, TXB2, and 8-iso-PGF2{alpha} production
After incubation, the culture medium was collected and centrifuged at 10,000 g for 5 min. PGE2, TXB2, 8-iso-PGF2{alpha}, and 6-keto-PGF1{alpha}, the stable hydrolysis product of PGI2, were measured with specific EIA kits (Cayman Chemical) according to the manufacturer’s instructions.

Immunoprecipitation
We obtained cell lysates at a concentration of 106 cells/ml. For immunoprecipitation, cell lysates were precleared by centrifugation at 10,000 g for 15 min at 4°C and subsequent incubation with protein G-agarose beads (2 h, 4°C). After centrifugation at 10,000 g for 30 s, supernatants were incubated overnight with specific monoclonal antibodies against COX-1 or COX-2 (3 µg/ml) at 4°C (Cayman Chemical). Cell-free supernatants were used as negative controls. The target proteins bound by the specific antibodies were immunoprecipitated by addition of protein G-agarose beads and incubation for 3 h. Finally, the antibody-conjugated beads were washed several times according to the manufacturer’s protocol (Roche Immunoprecipitation Kit, Protein G-agarose; Roche Diagnostics, Basel, Switzerland) and resuspended in SDS-gel loading buffer under reducing conditions for the analysis, as described below.

Western blot analysis
At the end of incubations, cells were harvested in lysis buffer. After quantization by Lowry’s method, equal amounts of cell protein were loaded onto 10% SDS-acrylamide gels together with a definite amount of either COX-1 or COX-2 purified proteins (Cayman Chemical). At the end of the run, proteins were transferred to a nitrocellulose membrane and incubated with polyclonal antibodies against COX-2 (1:400, Cayman Chemical) and COX-1 (1:1000, Cayman Chemical) overnight and then with suitable peroxidase-conjugated secondary antibodies for 1 h. Proteins were detected by chemiluminescence (Amersham Biosciences, Piscataway, NJ, USA). Loading control was performed using actin immunodetection. In the case of immunoprecipitated proteins, COX isoforms immunoprecipitated by COX-1 and COX-2 monoclonal antibodies were then detected using COX-1 and COX-2 polyclonal antibodies.

Statistical analysis
Data were obtained from 3 to 6 independent experiments, each value representing mean ± SE of duplicate or triplicate determinations. Comparison between groups was performed by ANOVA followed by Sheffé’s test for multiple comparisons. Values of P < 0.05 were considered statistically significant.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 REFERENCES
 
Expression of COX-2 in HUVECs
Expression of COX-2 was evaluated in HUVECs under different cellular culture conditions and in agreement with previous studies (19 , 20) increased expression of COX-2 was observed after treatment with FCS 5% either in the presence or absence of interleukin (IL) -1β activation (Fig. 1 ).


Figure 1
View larger version (34K):
[in this window]
[in a new window]

  		Figure 1. Expression of COX-2 in HUVECs. Cells, either resting or treated with IL-1β (2 ng/ml), were grown in medium in the presence or absence of 5% FCS. COX-2 expression was determined in cellular lysates by Western blot of equal amounts of protein and semiquantitated by densitometry. Results are the average of 2 separate experiments expressed as fold increase vs. control. Inset: representative Western blot.

Effect of selective COX-1 and COX-2 inhibitors on the production of prostacyclin
Pretreatment of 5% FCS-treated cells with either a selective COX-1 inhibitor (SC560) or a selective COX-2 inhibitor structurally correlated to celecoxib (21 ; SC236) reduced the production of the hydrolysis product of PGI2 (6-keto-PGF1{alpha}) by 70% or more (Fig. 2A ).


Figure 2
View larger version (13K):
[in this window]
[in a new window]

  		Figure 2. Effect of treatment with selective COX-1 (SC560 and FR122047) or COX-2 (SC236) inhibitors toward PGI2 and isoprostane formation in HUVECs. Cells were incubated in medium containing 5% FCS in the presence of increasing concentrations of SC560 or SC2536 (1–100 nM; A, C) and FR122047 (3–300 nm; B) for 6 h. Isoprostanes (as 8-iso-PGF2{alpha}) and PGI2 (as 6-keto-PGF1{alpha}) were evaluated in aliquots of cell supernatants using specific EIAs. Values are expressed as percentage ± SE (n=3–5) of the metabolite production observed in the absence of inhibitors (control values: 6-keto-PGF1{alpha}, 7.86±1.63 ng/mg protein; 8-iso-PGF2{alpha}, 0.64±0.04 ng/mg protein). *P < 0.05, **P < 0.01 vs. control.

The selectivity of the compounds used was confirmed using the whole-blood assay according to Patrignani et al. (18) , and the results were obtained, taking into account plasma protein binding, and confirmed the expected selectivity in the range of concentrations used (IC50 COX-1/COX-2: 13 nM/930 nM for SC-560; 9 µM/457 nM for SC-236).

The use of a different, structurally unrelated COX-1 selective inhibitor FR122047 (with a reported COX-1/COX-2 selectivity of >2300-fold; ref. 22 ) also significantly decreased the basal PGI2 synthesis, with an EC50 of ~30 nM (Fig. 2B ).

Concomitant formation of isoprostanes, expressed as 8-iso-PGF2{alpha}, was also observed, with absolute values ~10-fold lower than prostacyclin (i.e., 6-keto-PGF1{alpha}: 7.86±1.63 ng/mg protein; 8-iso-PGF2{alpha}: 0.64±0.04 ng/mg protein). Isoprostane formation also was inhibited by COX-1 and COX-2 selective inhibitors, with a potency equal to that shown for the inhibition of prostacyclin, suggesting that, as expected (2) , COX-activity and the synthesis of prostacyclin in endothelial cells is associated with significant lipid peroxidation and formation of isoprostanes (Fig. 2C ).

Similar results were obtained using IL-1β, showing efficient inhibition of PGI2 formation by either the COX-1 or the COX-2 selective compounds (Fig. 3 ). Again, isoprostane formation closely followed the observed production of prostacyclin, confirming that COX activity was linked to significant lipid peroxidation and formation of isoprostanes (data not shown).


Figure 3
View larger version (21K):
[in this window]
[in a new window]

  		Figure 3. Effect of treatment with selective COX-1 (SC560) or COX-2 (SC236) inhibitors toward PGI2 formation in IL-1β-treated HUVECs. Cells were incubated in medium containing 5% FCS and 2 ng/ml of IL-1β in the presence of increasing concentrations (1–100 nM) of SC560 or SC2536 for 6 h. PGI2 (as 6-keto-PGF1{alpha}) was evaluated in aliquots of cell supernatants using a specific EIA. Values are expressed as percentage ± SE (n=3) of the metabolite production observed in the absence of inhibitors. *P < 0.05, **P < 0.01 vs. IL-1β.

COX-1 and COX-2 heterodimer formation
To better understand the mechanisms involved in the synthesis of prostacyclin by endothelial cells and the complex behavior of COX-1 and COX-2 selective inhibitors, we investigated the potential formation of the recently described COX-1/COX-2 heterodimers (17) . Confocal microscopy of endothelial cells colabeled using COX-1 and COX-2 specific antibodies showed significant colocalization, in particular at the perinuclear envelope (data not shown). Western blot analysis of COX-1 selective immunoprecipitates obtained from endothelial cells also showed the presence of minor amounts of COX-2, but the signal obtained by Western blot was clearly lower that that observed using COX-2 selective immunoprecipitates (Fig. 4A ); the same was observed for COX-1 in COX-2 selective immunoprecipitates (Fig. 4B ), suggesting the occurrence of only limited COX-1/COX-2 heterodimer formation in HUVECs.


Figure 4
View larger version (23K):
[in this window]
[in a new window]

  		Figure 4. Representative Western blot analysis of COX-1/COX-2 heterodimers formation in HUVECs. Specific immunoprecipitates obtained with COX-1 (IPP1) or COX-2 (IPP2) monoclonal antibodies in IL-1β-treated cells (2 ng/ml, 6 h) were recognized by polyclonal COX-1 (A) or COX-2 (B) specific antibodies. Arrows indicate the specific bands (corresponding to 70 kDa of the purified proteins) recognized by the antibodies. Nonspecific bands, as in lanes 4 and 5, represent protein A-associated immunoglobulins.

Effect of arachidonic acid in the presence or absence of 12-HpETE or glutathione depletion on prostacyclin formation
In physiopathological conditions where the antiadhesive properties of the endothelium become critical, such as during platelet activation and thrombus formation, large amounts of arachidonic acid together with hyperoxides may be available to endothelial cells. As lower availability of substrate or hydroperoxides may result in the preferential activation of COX-2, whereas a higher "hydroperoxide tone" or higher arachidonic acid concentrations may lead to the preferential arachidonic acid oxygenation by COX-1 (23 , 24) , we compared the effect of the SC236 and SC560 on prostacyclin production in the presence of exogenous arachidonic acid with or without 12-HpETE, a hydroperoxide also formed by activated platelets (25) . In HUVECs exposed to low concentrations of arachidonic acid (10 µM) for 30 min, SC560 as well as SC236 concentration dependently inhibited PGI2 formation (Fig. 5 ). Conversely, the formation of prostacyclin after addition of 10 µM in the presence of hydroperoxides (50 nM 12-HpETE) was concentration dependently inhibited by SC560 (IC50: 5.1 nM) but not by SC236 (Fig. 6 ). Similar results were obtained using platelets as a source of endogenous 12-HpETE, as the observed production of PGI2 resulted effectively inhibited by SC560 but not affected by the inhibition of COX-2 (Fig. 7 ).


Figure 5
View larger version (19K):
[in this window]
[in a new window]

  		Figure 5. Effect of treatment with selective COX-1 (SC560) or COX-2 (SC236) inhibitors toward PGI2 production in HUVECs on incubation with arachidonic acid. Cells were incubated in PBS with calcium and magnesium in the presence of increasing concentrations of SC560 or SC236 (1–100 nM); 10 min after treatment with inhibitors, cells were added with arachidonic acid (10 µM) for 30 min. PGI2 was evaluated as 6-keto-PGF1{alpha} in aliquots of cell supernatants using a specific EIA. Values are expressed as percentage ± SE (n=4) of the 6-keto-PGF1{alpha} production observed in the presence of arachidonic acid (control). *P < 0.05, **P < 0.01 vs. control.


Figure 6
View larger version (26K):
[in this window]
[in a new window]

  		Figure 6. Effect of treatment with selective COX-1 (SC560) or COX-2 (SC236) inhibitors toward PGI2 production in HUVECs on challenge with arachidonic acid and 12-HpETE. Cells were incubated in PBS with calcium and magnesium in the presence of increasing concentration of SC560 or SC236 (1–100 nM); 10 min after treatment with inhibitors, cells were added with arachidonic acid (10 µM) including 12-HpETE (50 nM) for 30 min. PGI2 was evaluated as 6-keto-PGF1{alpha} in aliquots of cell supernatants using a specific EIA. Values are expressed as percent ± SE (n=3) of the 6-keto-PGF1{alpha} production observed in the presence of arachidonic acid and 12-HpETE (control). *P < 0.05, **P < 0.01 vs. control.


Figure 7
View larger version (20K):
[in this window]
[in a new window]

  		Figure 7. Effect of treatment with selective COX-1 (SC560) or COX-2 (SC236) inhibitors toward PGI2 production in HUVECs coincubated with human washed platelets and challenged with arachidonic acid. Cells were incubated in PBS with calcium and magnesium in the presence of increasing concentrations of SC560 or SC236 (1–100 nM); 15 min after treatment with inhibitors, the medium was changed with a suspension of washed human platelets (2 ml, 108 platelets/ml) containing the inhibitor used, and arachidonic acid (10 µM) was added for 30 min. PGI2 was evaluated as 6-keto-PGF1{alpha} in aliquots of cell supernatants using a specific EIA. Values are expressed as percentage ± SE (n=3) of the 6-keto-PGF1{alpha} production observed in the presence of arachidonic acid and platelets (control). *P < 0.05, **P < 0.01 vs. control.

To further confirm these results, we treated cells with DEM (0.5 mM) and BSO (1 mM), a widely recognized method to enhance the hydroperoxide tone by diminishing the availability of reduced form of glutathione (26) . Under these experimental conditions, the formation of prostacyclin was also concentration dependently inhibited by SC560 but only marginally affected by SC236 at the highest concentration tested (Fig. 8 ).


Figure 8
View larger version (27K):
[in this window]
[in a new window]

  		Figure 8. Effect of treatment with selective COX-1 (SC560) or COX-2 (SC236) inhibitors toward PGI2 production in reduced glutathione-depleted HUVECs on incubation with arachidonic acid. Cells were incubated in PBS with calcium and magnesium in the presence of DEM (0.2 mM) and BSO (1 mM) for 20 min. Afterward, increasing concentrations of SC560 or SC236 (1–100 nM) were added for 10 min. Finally, cells were treated with arachidonic acid (10 µM) for 30 min. PGI2 was evaluated as 6-keto-PGF1{alpha} in aliquots of cell supernatants using a specific EIA. Values are expressed as percentage ± SE (n=3) of the 6-keto-PGF1{alpha} production observed in DEM/BSO-pretreated cells in the presence of arachidonic acid (control). *P < 0.05, **P < 0.01 vs. control.


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
REFERENCES
 
In the present study, we provide evidence of a significant effect of both COX-1 and COX-2 selective inhibitors on the production of prostacyclin from endogenous substrate in human umbilical vein endothelial cells. The inhibition of prostacyclin production is associated with a comparable suppression of isoprostane formation, suggesting that COX-2 activity in endothelial cells may provide the hydroperoxide tone necessary for the activity of COX-1 and that COX-2 inhibition may therefore cause a significant, indirect inhibition of COX-1. In support of this hypothesis in the presence of exogenous arachidonic acid and hydroperoxides (12-HpETE) or washed platelets (as an endogenous source of 12-HpETE), as well as after GSH depletion (mimicking a sustained redox imbalance at the cellular level), we provide evidence that COX-1, independently from COX-2 activity, is responsible of the observed production of prostacyclin. This evidence contributes to a better understanding of the potential cardiovascular effects of nonselective NSAIDs and selective COX-2 inhibitors.

Prostacyclin, the main product of COX activity in endothelial cells, possesses antiadhesive, antithrombotic, and antiproliferative properties, thereby being essential to the functional integrity of the endothelium (9) . Until the debate on the cardiovascular safety of COX-2-selective inhibitors, the production of prostacyclin was mainly attributed to the housekeeping enzyme COX-1. Evidence for COX-2 expression as a result of laminar flow shear stress and the effect of selective COX-2 inhibition on prostacyclin formation in vivo suggest a critical contribution of COX-2 to the physiological production of PGI2 (12 , 13) . Although widely accepted, this hypothesis is mainly based on the observed inhibition of the urinary excretion of prostacyclin metabolites on treatment with COX-2 selective inhibitors in humans (14) , and the question as to the relative role of either COX isoform toward the synthesis of prostacyclin in endothelial cells remains an open issue (14 , 27) together with the expression of COX-2 under normal conditions in vivo (15 , 16) .

To address it, we verified that, in agreement with previously reported data (5 , 19) , nonactivated endothelial cells maintained in 5% FCS (basal conditions) express both protein isoforms (28) . As endothelial cells easily change phenotype, we consistently used first passage HUVECs, throughout all the experiments. Selective inhibitors of either COX-1 (SC560 and FR122047) or COX-2 (SC236) concentration dependently reduced basal or IL-1β-induced prostacyclin production by 70% or more. This is clearly in contrast with the involvement of only one of either COXs (COX isoforms) in prostacyclin release by human endothelium. To address the specificity of the compounds at the concentrations used in our experiments, we evaluated their activity in the whole blood selectivity test (18) ; SC560 showed an IC50 of 1 µM with respect to COX-2, while SC236 inhibited COX-1 only at concentrations in the micromolar range, all concentrations well above the ones used in our experiments. Also taking into account the contribution of plasma protein binding to the results obtained (29) , it appears clearly that the compounds were used at concentrations at which their selectivity appeared to be well preserved.

Together with the formation of prostacyclin, consistent amounts of isoprostanes were observed, and either COX-1 or COX-2 inhibition also resulted in a parallel inhibition of the formation of isoprostanes, indicating that active cyclooxygenation of arachidonic acid was also leading to lipid peroxidation and formation of isoprostanes (30) . This effects turns out to be extremely important because of the differences in sensitivity to hydroperoxide activation between COX isoforms, with COX-1 activation requiring higher concentration of peroxides than COX-2. Therefore, the activation of COX-2 may lead to a secondary activation of COX-1, which may indeed represent a significant source of prostacyclin. The significant colocalization of the two isoforms observed by immunohistochemistry provides further support for this kind of close interaction. We therefore propose that the effect of COX-2 selective compounds is mainly resulting from the indirect inhibition of COX-1 activity through the decrease in hydroperoxide availability resulting from COX-2 inhibition, while the observed activity of COX-1 selective inhibitors is the result of the direct inhibition of this isoform of the COX enzyme.

The crystal structures of sheep COX-1 and mouse COX-2 have been elucidated and suggest that COX-1 and COX-2 are homodimeric proteins (31 , 32) . A recent study (17) reporting that COX-1 and COX-2 can heterodimerize in vivo prompted us to evaluate the formation of COX heterodimers as a possible alternative interpretation of the observed overlap between COX-1 and COX-2 selective inhibitors; nevertheless, immunoprecipitation experiments provided evidence of only limited amounts of COX heterodimers in HUVECs. In fact, while the immunoprecipitates obtained by antibodies specifically recognizing one isoform contained also detectable amounts of the other isoform (i.e., immunoprecipitates obtained using an anti-COX-1 antibody also contained COX-2, as detected by COX-2 specific Western blot), the overall immunoprecipitate amount of each isoform (i.e., the amount of COX-2 as detected by COX-2 specific polyclonal antibody Western blot in immunoprecipitates obtained using a different monoclonal anti-COX-2 antibody) was significantly higher than the amount coprecipitating with the heterologous antibody.

Purified COX-1 has shown a sigmoidal dependence on arachidonate substrate, whereas COX-2 displays simple saturable behavior (23) . Indeed, a large variation in Ki values of selective competitive inhibitors was observed at increasing arachidonic acid concentrations, supporting the concept that COX-2 is very efficient in using its substrate at low (<0.5 µM) arachidonic acid concentrations, while COX-1 requires larger amounts of substrate (33) . Based on these considerations, we also studied the effect of the availability of exogenous substrate, a condition occurring for instance when high amounts of arachidonic acid are released by activated platelets (34 , 35) , on endothelial prostacyclin production. In the presence of 10 µM hydroperoxide-free arachidonic acid, the inhibitory profile of selective COX-1 and COX-2 inhibitors was similar to that observed in the absence of exogenous arachidonic acid, suggesting that also in this case COX-2 activity may be providing COX-1 with the necessary hydroperoxides. On the contrary, when the concentrations of hydroperoxides were boosted by providing 10 µM arachidonic acid added with 12-HpETE, which promotes the abstraction of a hydrogen atom from arachidonic acid and the initiation of prostacyclin synthesis by COXs (36) , SC236 turned out to be largely ineffective, while SC560 inhibited prostacyclin formation, with an IC50 of 5.1 nM. Note that 12-HpETE represents one of the major arachidonic acid metabolites synthesized by activated platelets (37) and that NSAIDs have been shown to significantly inhibit the peroxidase activity that normally reduces 12-HpETE to 12-HETE in platelets (38) , therefore leading to increased concentrations of the hydroperoxide. Indeed, coincubation of HUVECs and washed human platelets in the presence of arachidonic acid resulted in the formation of prostacyclin that was not sensitive to COX-2 inhibitors but was efficiently suppressed by the treatment with a specific COX-1 inhibitor. These results are in agreement with the hypothesis that changes in endothelial peroxide tone may cause the activation of the COX-1 (39) and support the hypothesis that the role played by COX-2 in basal prostacyclin formations could be the result of changes in the formation of lipid peroxides associated with its activity.

In further agreement with these results, by inducing a cellular redox imbalance through the decreased availability of the reduced form of glutathione, we observed that prostacyclin was formed preferentially via COX-1 rather than COX-2, again suggesting that the role of COX-2 in endothelial cell prostacyclin formation under basal conditions may be that of providing COX-1 with the required hydroperoxides.

In conclusion, both COX isoforms appear to be important for the production of prostacyclin in human endothelial cells, with COX-2 providing the hydroperoxide tone necessary for COX-1 activity. On the contrary, in conditions such as those defined by intracellular glutathione depletion, or whenever significant amounts of arachidonic acid and hydroperoxides may be made available, such as during platelet adhesion and activation, COX-1 may be freed from its partial dependence on COX-2, becoming the main activity leading to the production of endothelial prostacyclin. These findings may contribute to a better understanding of the effects of cyclooxygenase inhibitors on endothelial cells by pointing to a critical role of COX-1 in prostacyclin generation, at least under conditions where increased substrate and lipid hydroperoxide availability are observed.


   ACKNOWLEDGMENTS
 
We thank Dr. A. Zanini for the assistance in collecting the umbilical cords and Dr. G. Simonutti for confocal microscopy analysis. The study was supported by grants from the European Community (LSHM-CT-2004-005033 to G.F. and A.S.) and from the Italian Ministry of Instruction, University and Research (FIRST to G.F. and A.S., and PRIN to A.S.).


   FOOTNOTES
 
1 Current address: Largo Meneghetti 2, Department of Pharmacology and Anesthesiology, University of Padova, 50131 Padova, Italy.

Received for publication March 10, 2008. Accepted for publication September 4, 2008.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 

  1. Smith, W. L., Song, I. (2002) The enzymology of prostaglandin endoperoxide H synthases-1 and -2. Prostaglandins Other Lipid Mediat. 69,115-128[CrossRef]
  2. Caughey, G. E., Cleland, L. G., Penglis, P. S., Gamble, J. R., James, M. J. (2001) Roles of cyclooxygenase (COX)-1 and COX-2 in prostanoid production by human endothelial cells: selective up-regulation of prostacyclin synthesis by COX-2. J. Immunol. 167,2831-2838[Abstract/Free Full Text]
  3. Mitchell, J. A., Warner, T. D. (2006) COX isoforms in the cardiovascular system: understanding the activities of non-steroidal anti-inflammatory drugs. Nat. Rev. Drug Discov. 5,75-86[CrossRef][Medline]
  4. Shimizu, T., Ohto, T., Kita, Y. (2006) Cytosolic phospholipase A2: biochemical properties and physiological roles. IUBMB Life 58,328-333[Medline]
  5. Karim, S., Habib, A., Levy-Toledano, S., Maclouf, J. (1996) Cyclooxygenase-1 and -2 of endothelial cells utilize exogenous or endogenous arachidonic acid for transcellular production of thromboxane. J. Biol. Chem. 271,12042-12048[Abstract/Free Full Text]
  6. Camacho, M., Lopez-Belmonte, J., Vila, L. (1998) Rate of vasoconstrictor prostanoids released by endothelial cells depends on cyclooxygenase-2 expression and prostaglandin I synthase activity. Circ. Res. 83,353-365[Abstract/Free Full Text]
  7. Folco, G., Murphy, R. C. (2006) Eicosanoid transcellular biosynthesis: from cell-cell interactions to in vivo tissue responses. Pharmacol. Rev. 58,375-388[Abstract/Free Full Text]
  8. Kulmacz, R. J. (1998) Cellular regulation of prostaglandin H synthase catalysis. FEBS Lett. 430,154-157[CrossRef][Medline]
  9. Davidge, S. T. (2001) Prostaglandin H synthase and vascular function. Circ. Res. 89,650-660[Abstract/Free Full Text]
  10. Okahara, K., Sun, B., Kambayashi, J. (1998) Upregulation of prostacyclin synthesis-related gene expression by shear stress in vascular endothelial cells. Arterioscler. Thromb. Vasc. Biol. 18,1922-1926[Abstract/Free Full Text]
  11. James, M. J., Penglis, P. S., Caughey, G. E., Demasi, M., Cleland, L. G. (2001) Eicosanoid production by human monocytes: does COX-2 contribute to a self-limiting inflammatory response?. Inflamm. Res. 50,249-253[CrossRef][Medline]
  12. Topper, J. N., Cai, J., Falb, D., Gimbrone, M. A., Jr (1996) Identification of vascular endothelial genes differentially responsive to fluid mechanical stimuli: cyclooxygenase-2, manganese superoxide dismutase, and endothelial cell nitric oxide synthase are selectively up-regulated by steady laminar shear stress. Proc. Natl. Acad. Sci. U. S. A. 93,10417-10422[Abstract/Free Full Text]
  13. McAdam, B. F., Catella-Lawson, F., Mardini, I. A., Kapoor, S., Lawson, J. A., FitzGerald, G. A. (1999) Systemic biosynthesis of prostacyclin by cyclooxygenase (COX)-2: the human pharmacology of a selective inhibitor of COX-2. Proc. Natl. Acad. Sci. U. S. A. 96,272-277[Abstract/Free Full Text]
  14. Flavahan, N. A. (2007) Balancing prostanoid activity in the human vascular system. Trends Pharmacol. Sci. 28,106-110[CrossRef][Medline]
  15. Zidar, N., Dolenc-Strazar, Z., Jeruc, J., Jerse, M., Balazic, J., Gartner, U., Jermol, U., Zupanc, T., Stajer, D. (2007) Expression of cyclooxygenase-1 and cyclooxygenase-2 in the normal human heart and in myocardial infarction. Cardiovasc. Pathol. 16,300-304[CrossRef][Medline]
  16. Radi, Z. A., Ostroski, R. (2007) Pulmonary and cardiorenal cyclooxygenase-1 (COX-1), -2 (COX-2), and microsomal prostaglandin E synthase-1 (mPGES-1) and -2 (mPGES-2) expression in a hypertension model. Mediators Inflamm. 2007,85091[Medline]
  17. Yu, Y., Fan, J., Chen, X. S., Wang, D., Klein-Szanto, A. J., Campbell, R. L., FitzGerald, G. A., Funk, C. D. (2006) Genetic model of selective COX2 inhibition reveals novel heterodimer signaling. Nat. Med. 12,699-704[CrossRef][Medline]
  18. Patrignani, P., Panara, M. R., Greco, A., Fusco, O., Natoli, C., Iacobelli, S., Cipollone, F., Ganci, A., Creminon, C., 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[Abstract/Free Full Text]
  19. Habib, A., Creminon, C., Frobert, Y., Grassi, J., Pradelles, P., Maclouf, J. (1993) Demonstration of an inducible cyclooxygenase in human endothelial cells using antibodies raised against the carboxyl-terminal region of the cyclooxygenase-2. J. Biol. Chem. 268,23448-23454[Abstract/Free Full Text]
  20. Habib, A., Martinuzzo, M. E., Carreras, L. O., Levy-Toledano, S., Maclouf, J. (1995) Increased expression of inducible cyclooxygenase-2 in human endothelial cells by antiphospholipid antibodies. Thromb. Haemost. 74,770-777[Medline]
  21. Penning, T. D., Talley, J. J., Bertenshaw, S. R., Carter, J. S., Collins, P. W., Docter, S., Graneto, M. J., Lee, L. F., Malecha, J. W., Miyashiro, J. M., Rogers, R. S., Rogier, D. J., Yu, S. S., Anderson, G. D., Burton, E. G., Cogburn, J. N., Gregory, S. A., Koboldt, C. M., Perkins, W. E., Seibert, K., Veenhuizen, A. W., Zhang, Y. Y., Isakson, P. C. (1997) Synthesis and biological evaluation of the 1,5-diarylpyrazole class of cyclooxygenase-2 inhibitors: identification of 4-[5-(4-methylphenyl)-3-(trifluoromethyl)-1H-pyrazol-1-yl]benze nesulfonamide (SC-58635, celecoxib). J. Med. Chem. 40,1347-1365[CrossRef][Medline]
  22. Ochi, T., Motoyama, Y., Goto, T. (2000) The analgesic effect profile of FR122047, a selective cyclooxygenase-1 inhibitor, in chemical nociceptive models. Eur. J. Pharmacol. 391,49-54[CrossRef][Medline]
  23. Chen, W., Pawelek, T. R., Kulmacz, R. J. (1999) Hydroperoxide dependence and cooperative cyclooxygenase kinetics in prostaglandin H synthase-1 and -2. J. Biol. Chem. 274,20301-20306[Abstract/Free Full Text]
  24. Shitashige, M., Morita, I., Murota, S. (1998) Different substrate utilization between prostaglandin endoperoxide H synthase-1 and -2 in NIH3T3 fibroblasts. Biochim. Biophys. Acta 1389,57-66[Medline]
  25. Coffey, M. J., Jarvis, G. E., Gibbins, J. M., Coles, B., Barrett, N. E., Wylie, O. R., O'Donnell, V. B. (2004) Platelet 12-lipoxygenase activation via glycoprotein VI: involvement of multiple signaling pathways in agonist control of HpETE synthesis. Circ. Res. 94,1598-1605[Abstract/Free Full Text]
  26. Kokura, S., Wolf, R. E., Yoshikawa, T., Granger, D. N., Aw, T. Y. (1999) Molecular mechanisms of neutrophil-endothelial cell adhesion induced by redox imbalance. Circ. Res. 84,516-524[Abstract/Free Full Text]
  27. Mitchell, J. A., Lucas, R., Vojnovic, I., Hasan, K., Pepper, J. R., Warner, T. D. (2006) Stronger inhibition by nonsteroid anti-inflammatory drugs of cyclooxygenase-1 in endothelial cells than platelets offers an explanation for increased risk of thrombotic events. FASEB J. 20,2468-2475[Abstract/Free Full Text]
  28. Bolego, C., Buccellati, C., Radaelli, T., Cetin, I., Puglisi, L., Folco, G., Sala, A. (2006) eNOS, COX-2, and prostacyclin production are impaired in endothelial cells from diabetics. Biochem. Biophys. Res. Commun. 339,188-190[CrossRef][Medline]
  29. Warner, T. D., Vojnovic, I., Bishop-Bailey, D., Mitchell, J. A. (2006) Influence of plasma protein on the potencies of inhibitors of cyclooxygenase-1 and -2. FASEB J. 20,542-544[Abstract/Free Full Text]
  30. Pepicelli, O., Fedele, E., Berardi, M., Raiteri, M., Levi, G., Greco, A., Ajmone-Cat, M. A., Minghetti, L. (2005) Cyclo-oxygenase-1 and -2 differently contribute to prostaglandin E2 synthesis and lipid peroxidation after in vivo activation of N-methyl-D-aspartate receptors in rat hippocampus. J. Neurochem. 93,1561-1567[CrossRef][Medline]
  31. Kurumbail, R. G., Stevens, A. M., Gierse, J. K., McDonald, J. J., Stegeman, R. A., Pak, J. Y., Gildehaus, D., Miyashiro, J. M., Penning, T. D., Seibert, K., Isakson, P. C., Stallings, W. C. (1996) Structural basis for selective inhibition of cyclooxygenase-2 by anti-inflammatory agents. Nature 384,644-648[CrossRef][Medline]
  32. Picot, D., Loll, P. J., Garavito, R. M. (1994) The X-ray crystal structure of the membrane protein prostaglandin H2 synthase-1. Nature 367,243-249[CrossRef][Medline]
  33. Swinney, D. C., Mak, A. Y., Barnett, J., Ramesha, C. S. (1997) Differential allosteric regulation of prostaglandin H synthase 1 and 2 by arachidonic acid. J. Biol. Chem. 272,12393-12398[Abstract/Free Full Text]
  34. Camacho, M., Vila, L. (2000) Transcellular formation of thromboxane A2 in mixed incubations of endothelial cells and aspirin-treated platelets strongly depends on the prostaglandin I-synthase activity. Thromb. Res. 99,155-164[CrossRef][Medline]
  35. Riendeau, D., Guay, J., Weech, P. K., Laliberte, F., Yergey, J., Li, C., Desmarais, S., Perrier, H., Liu, S., Nicoll-Griffith, D., and,, et al (1994) Arachidonyl trifluoromethyl ketone, a potent inhibitor of 85-kDa phospholipase A2, blocks production of arachidonate and 12-hydroxyeicosatetraenoic acid by calcium ionophore-challenged platelets. J. Biol. Chem. 269,15619-15624[Abstract/Free Full Text]
  36. Smith, W. L., DeWitt, D. L., Garavito, R. M. (2000) Cyclooxygenases: structural, cellular, and molecular biology. Annu. Rev. Biochem. 69,145-182[CrossRef][Medline]
  37. Nugteren, D. H. (1975) Arachidonate lipoxygenase in blood platelets. Biochim. Biophys. Acta 380,299-307[Medline]
  38. Siegel, M. I., McConnell, R. T., Porter, N. A., Cuatrecasas, P. (1980) Arachidonate metabolism via lipoxygenase and 12L-hydroperoxy-5,8,10,14-icosatetraenoic acid peroxidase sensitive to anti-inflammatory drugs. Proc. Natl. Acad. Sci. U. S. A. 77,308-312[Abstract/Free Full Text]
  39. Kulmacz, R. J., Wang, L. H. (1995) Comparison of hydroperoxide initiator requirements for the cyclooxygenase activities of prostaglandin H synthase-1 and -2. J. Biol. Chem. 270,24019-24023[Abstract/Free Full Text]




This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF) Free
Right arrow All Versions of this Article:
fj.08-106591v1
23/2/605    most recent
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Bolego, C.
Right arrow Articles by Sala, A.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Bolego, C.
Right arrow Articles by Sala, A.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS