Original Contributions

Rate of Vasoconstrictor Prostanoids Released by Endothelial Cells Depends on Cyclooxygenase-2 Expression and Prostaglandin I Synthase Activity

Mercedes Camacho, Jesús López-Belmonte, Luis Vila
https://doi.org/10.1161/01.RES.83.4.353
Circulation Research. 1998;83:353-365
Originally published August 24, 1998

Jump to

Abstract

Abstract—This study was undertaken to investigate the enzymatic regulation of the biosynthesis of vasoconstrictor prostanoids by resting and interleukin (IL)-1β–stimulated human umbilical vein endothelial cells (HUVECs). Biosynthesis of eicosanoids in response to IL-1β, exogenous labeled arachidonic acid (AA), or histamine, as well as their spontaneous release, was evaluated by means of HPLC and RIA. HUVECs exposed to IL-1β produced prostaglandin (PG) I2 for no longer than 30 seconds after the substrate was added irrespective of the cyclooxygenase (COX) activity, whereas the time course of PGE2 and PGD2 formation was parallel to the COX activity. The ratio of PGE2 to PGD2 produced by HUVECs was similar to that obtained by purified COX-1 and COX-2. Production of PGF2α from exogenous AA was limited and similar in both resting and IL-1β–treated cells. PGF2α was the main prostanoid released into the medium during exposure to IL-1β, whereas when HUVECs treated with IL-1β were stimulated with histamine or exogenous AA, PGE2 was released in a higher quantity than PGF2α. PGF2α released into the medium during treatment with IL-1β and the biosynthesis of PGE2 and PGD2 in response to exogenous AA or histamine increased with COX-2 expression, whereas this did not occur in the case of PGI2. We observed that PGI synthase (PGIS) mRNA levels were not modified by the exposure to IL-1β, but the enzyme was partially inactivated. When SnCl2 was added to the incubation medium, the transformation of exogenous AA-derived PGH2 into PGE2 and PGD2 was totally diverted toward PGF2α. Overall, these results support the conclusions that PGE2 and PGD2 (and also probably PGF2α) were nonenzymatically derived from PGH2 in HUVECs. The concept that a high ratio of PGH2 was released by the IL-1β–treated HUVECs and isomerized outside the cell into PGE2 and PGD2 was supported by the biosynthesis of thromboxane B2 by COX-inactivated platelets, indicating the uptake by platelets of HUVEC-derived PGH2. The IL-1β–induced increase in the release of PGH2 by HUVECs was suppressed by the COX-2–selective inhibitor SC-58125 and correlated with both COX-2 expression and PGIS inactivation. An approach to the mechanism of inactivation of PGIS by the exposure to IL-1β was performed by using labeled endoperoxides as substrate. The involvement of HO· in the PGIS inactivation was supported by the fact that deferoxamine, pyrrolidinedithiocarbamate, DMSO, mannitol, and captopril antagonized the effect of IL-1β on PGIS to different degrees. The NO synthase inhibitor NG-monomethyl-l-arginine also antagonized the PGIS inhibitory effect of IL-1β, indicating that NO· was also involved. NO· reacts with O2−· to form peroxynitrite, which has been reported to inactivate PGIS. Homolytic fission of the O-O bond of peroxynitrite yields NO2· and HO·. The fact that 2-(4-carboxyphenyl)-4,4,5,5-tetramethylimidazoline-1-oxyl-3-oxide (carboxy-PTIO), which reacts with NO· to form NO2·, dramatically potentiated the IL-1β effect suggests that NO2· could be a species implicated in the inactivation of PGIS. Cooperation of HO· was supported by the fact that DMSO partially antagonized the effect of carboxy-PTIO. Although our results on the exact mechanism of the inactivation of PGIS caused by IL-1β were not conclusive, they strongly suggest that both NO· and HO· were involved.

Endothelium modulates the response of vascular smooth muscle to hormones, neurotransmitters, and platelet products by releasing relaxing factors, such as NO and prostaglandin (PG) I2 (also termed prostacyclin), and contracting factors, such as endothelin-1, angiotensin II, and vasoconstrictor metabolites of arachidonic acid (AA).1 2 3 4 5 6 7

PGI2, PGF2α, and PGE2 are the main eicosanoids detected in vitro in the incubates of endothelial cells under different experimental conditions, which include exogenously added AA and several agonists. PGD2, 12-hydroxyheptadecatrienoic acid (HHT), 15-HETE, and 11-HETE have also been detected in minor amounts.8 9 10 11 All these eicosanoids are derived from cyclooxygenase (COX) activity in human umbilical vein endothelial cells (HUVECs).12

PGH2 is the common precursor of the prostanoids whose formation from AA is catalyzed by 2 COX isoenzymes encoded by different genes: COX-1 and COX-2. COX-1 is expressed in a constitutive manner, and COX-2 is the isoenzyme inducible by mitogens and is overexpressed in inflammatory processes.13 14 Interleukin (IL)-1, together with tumor necrosis factor, plays a pivotal role in the inflammatory response and orchestrates a pleiad of secondary events that orient endothelial cells toward a proinflammatory and prothrombotic function by inducing adhesion molecules, endothelin-1, and coagulation factors.15 A characteristic activity of IL-1 is to induce COX-2 in several cell types, including endothelial cells.16 17 18 19

PGI synthase (PGIS) transforms PGH2 into what is the most characteristic prostanoid formed by resting endothelial cells, PGI2.8 20 PGI2 is the most potent platelet antiaggregatory agent, and it exhibits antiadhesive and smooth muscle–relaxing properties.2 21 PGH2 also undergoes spontaneous or enzymatic transformation toward PGF2α, PGE2, and PGD2.22 23 A growing amount of data support the concept that untransformed PGH2 is released in vivo by the vascular endothelium under different circumstances.24 25 26 27 28 29 30 The predominance of relaxing prostanoids, such as PGI2, or contracting prostanoids, such as PGH2 and PGF2α, released by endothelial cells will primarily depend on COX activity but also on the secondary pathways that yield the different prostanoids. PGI2 release as a consequence of the exposure of endothelial cells to IL-1 has been widely reported.16 19 31 32 33 34 35 36 37 Nevertheless, when COX activity is overexpressed, PGIS activity could be the limiting step in the biosynthesis of PGI2, and other prostanoids could be more stimulated by IL-1.33 36 37

Although release of PGH2 by endothelium both in vivo and ex vivo has been suggested, its regulation has not been systematically studied, and whether transformation of PGH2 into PGF2α, PGE2, and PGD2 in endothelial cells is spontaneous or enzymatically catalyzed remains unclear. In fact, PGIS is the only enzyme implicated in prostanoid biosynthesis that has been found in endothelial cells in addition to COX. The objective of the present work was to evaluate the hypothesis that the release of vasoconstricting prostanoids is regulated by 2 enzyme activities, COX and PGIS, and that these prostanoids are released predominantly under inflammatory conditions, such as after exposure to IL-1.

Materials and Methods

Cell Culture and Treatment

Endothelial cells were isolated from human umbilical veins and cultured as previously described.35 When HUVECs cultured with 20% of FBS reached confluence, they were seeded into 6-well plates and maintained without heparin and endothelial cell growth factor for 48 hours before the addition (or not) of 10 U/mL human recombinant IL-1β (50000 U/μg; purity, >98%; Boehringer Mannheim S.A.) in medium 199 containing 4% FBS and maintained for the indicated period of time until incubation with [14 C]AA or histamine.

Human dermal fibroblasts were isolated and cultured as described previously.38 Cells cultured in 6-well plates, in passages 4 to 6, were maintained with 1% FBS for 48 hours before the addition of 10 nmol/L phorbol 12-myristate 13-acetate (PMA, Sigma-Aldrich Química S.A.). After 6 hours, COX-2 inhibition experiments were performed as described below.

The permanent endothelial cell line HUV-EC-C was cultured and treated with 10 nmol/L PMA for 6 hours as described previously39 before COX-2 inhibition experiments. Experiments were performed with the HUV-EC-C line at passages 21 to 27.

Cells of the human erythroleukemia cell (HEL) line were obtained from the American Type Culture Collection (CRL 1730) and cultured in RPMI 1640 medium supplemented with 10% FBS, 1 mmol/L sodium pyruvate, 2 mmol/L l-glutamine, 100 U/mL penicillin, and 100 μg/mL streptomycin. Cultures were maintained at 37°C in a humidified atmosphere of 5% CO2 in air and subcultured every 3 to 4 days. To study COX-1 activity, cells were used without further treatment.

Preparation of Washed Platelet Suspensions

Peripheral venous blood was drawn from healthy donors who had received no medication in the 2 weeks before extraction. Washed platelet suspensions were prepared as previously described.40 Platelet density was 2×109 platelets/mL for the PGH2 trapping experiments and 1.7×108 platelets/mL for the study of COX-1 inhibition. COX-inactivated platelet suspensions were obtained by incubating platelet-rich plasma with 200 μmol/L acetylsalicylic acid (ASA) dissolved in ethanol (final concentration of ethanol, 0.5% [vol/vol]) for 10 minutes at room temperature before the washing procedure.

Release of Prostanoids From Endogenous AA in Resting and IL-1β–Treated HUVECs

Cells cultured in 6-well plates were incubated at 37°C in 2.5 mL medium 199 containing 4% FBS and 10 U/mL IL-1β. At the indicated periods of time, prostanoids released into the medium were evaluated after 250-μL aliquots of media were collected and placed in tubes containing indomethacin (final concentration, 10 μmol/L). Samples were kept at −80°C until analysis of PGF2α, PGE2, and 6-keto-PGF1α by RIA (PGF2α was from Amersham Ibérica; PGE2 and 6-keto-PGF1α were from Advanced Magnetics Inc).

Formation of Eicosanoids From Exogenous [14C]AA and Determination of COX Activity

After the indicated period of time of exposure to 10 U/mL IL-1β, cells were incubated at 37°C in 0.5 mL medium 199 containing 10 mmol/L HEPES, and 25 μmol/L [14C]AA (55 to 58 mCi/mmol [1-14C]AA, Amersham Ibérica) was added in 5 μL of ethanol. At the indicated periods of time, the reactions were stopped by adding 1N HCl to yield pH 3 followed by 1 vol of cold methanol. COX activity was evaluated as the sum of all the eicosanoids formed through the COX pathway.

PGH2 is an unstable prostaglandin that is converted into PGF2α by mild reducing agents, such as SnCl2.41 Therefore, to estimate the PGH2 released, we calculated the difference between the PGF2α peak of the samples from cells incubated as previously described and cells incubated in the presence of 200 μg/mL SnCl2. Samples were kept at −80°C until analysis. HPLC analysis of eicosanoids was performed as previously described.42

Release of Prostanoids From Endogenous AA in IL-1β–Treated HUVECs in Response to Histamine

Cells untreated and treated with 10 U/mL IL-1β for the indicated periods of time were incubated at 37°C in the presence of 0.5 mL medium 199 containing 10 mmol/L HEPES and 50 μmol/L histamine. After 10 minutes, supernatants were removed and placed in liquid N2. Samples were then stored at −80°C until analysis of PGF2α, PGE2, and 6-keto-PGF1α by RIA.

COX-1–and COX-2–Specific mRNA Analysis

COX-1–and COX-2–specific mRNA levels were determined as previously described.38 39

PGIS mRNA Analysis

Total RNA was isolated by phenol chloroform extraction according to the protocol described by Chomczynski and Sacchi43 and quantified spectrophotometrically by absorption at 260 and 280 nm. The specific levels of PGIS mRNA were determined by means of a quantitative reverse transcriptase-polymerase chain reaction (RT-PCR). Briefly, 1 μg of total RNA was reverse-transcribed into cDNA in the presence of 50 U reverse transcriptase of murine leukemia virus in a reaction buffer containing 10 mmol/L Tris-HCl (pH 8.3), 50 mmol/L KCl, 5 mmol/L MgCl2, 2.5 μmol/L random hexamers, 20 U of RNAsin (Geneamp RNA PCR kit, Perkin-Elmer), and 1 mmol/L deoxynucleotide triphosphates (Epicentre Technologies) in a final volume of 20 μL. The reaction was stopped by heating for 5 minutes at 99°C and then 5 minutes at 5°C.

The primers used for the amplification of the PGIS-specific cDNA were purchased from Oxford Biomedical Research, Inc. The cDNA for GAPDH was amplified and used as the in- ternal control, and the sense and antisense primers for GAPDH were 5′-CCACCCATGGCAAATTCCATGGCA-3′ and 5′-TCTAGACGGCAGGTCAGGTCCACC-3′, respectively.16 The PCR was carried out in a DNA Thermal Cycler 480 (Perkin-Elmer) with a reaction mixture (100 μL) containing 10 mmol/L Tris-HCl (pH 8.3), 50 mmol/L KCl, 2.5 mmol/L MgCl2, 0.2 mmol/L deoxynucleotide triphosphates, 0.25 μmol/L for PGIS and 1 μmol/L for GAPDH sense or anti-sense primers, 2.5 U Taq polymerase (Perkin-Elmer), and 4 μCi of [3H]deoxycytidine triphosphate (48 to 71 Ci/mmol, Amersham Ibérica). Aliquots were removed after 27, 30, and 35 cycles in order to test the linearity of the amplification. The profile of the amplification cycles was 94°C for 1 minute, 55°C for 2 minutes, and 72°C for 90 seconds. Amplification products were separated in a 1.5% low-melting-point agarose gel (GIBCO BRL) containing ethidium bromide (0.5 μg/mL, Perkin-Elmer). The bands were visualized under UV light and cut with a circular template. Radioactivity associated to the specific band was determined as previously described38 39 and normalized with respect to GAPDH.

Western Blotting of COX-1 and COX-2

Lysates of control cells and HUVECs treated with IL-1β for the indicated period of time were prepared as previously described.19 Total protein equivalents of each sample were submitted to SDS-PAGE and transferred to polyvinylidene difluoride membranes (Immobilon-P, Millipore). After blocking nonspecific binding of antibody, membranes were incubated with a specific polyclonal antibody for COX-1 and COX-2 (Oxford Biomedical Research). Positive controls were performed by using isolated COX-1 from ram seminal vesicles (Oxford Biomedical Research) or COX-2 purified from sheep placenta (Cayman Chemical). Detection was performed by using the ECL System (Amersham) according to the manufacturer’s instructions. To estimate the relative levels of each COX protein, the films were scanned in order to obtain digital images. The digital images of the bands were quantified, with the use of PC-Image analysis software (Foster Findlay Associates Ltd), by determining the sum of the gray density value of all pixels inside a rectangle drawn to encompass each band (the same-sized rectangle for all bands in a particular gel), and a ratio of gray density to rectangular area was obtained for each band. Results were then expressed as arbitrary units or normalized to the darkest band.

Trapping of HUVEC-Derived PGH2 by Platelets

HUVECs in 6-well plates, untreated and treated with 10 U/mL IL-1β for 24 hours, were incubated at 37°C in 0.450 mL medium 199 containing 10 mmol/L HEPES and 25 μmol/L [14C]AA. After 0.5 minutes, 108 COX-inactivated platelets in 50 μL were added to the wells, and the reactions were stopped after another 9.5 minutes, as described previously. Inactivation of COX in the platelet suspensions was monitored previously by incubating 0.5-mL aliquots of the washed platelet suspensions with 25 μmol/L [14C]AA at 37°C for 5 minutes.

Trapping of PGH2 from endogenous AA derived from resting and IL-1β–treated HUVECs by platelets was achieved by adding 50 μL of a platelet suspension containing 108 COX-inactivated platelets to HUVECs 0.5 minutes after cell challenge with 50 μmol/L histamine. After 10 minutes of the addition of histamine, supernatants were removed and placed in liquid N2. Samples were then kept at −80°C until the analysis of 6-keto-PGF1α and thromboxane (Tx) B2 by RIA (TxB2, Advanced Magnetics Inc).

Incubation With Isolated COX-1 and COX-2 and Obtaining of 14C-Labeled Endoperoxides

COX-1 (5 U/0.5 mL) isolated from ram seminal vesicles (Oxford Biomedical Research) or COX-2 purified from sheep placenta (Cayman Chemical) was incubated in 100 mmol/L Tris-HCl (pH 8.5) containing 2 mmol/L phenol and 5 μmol/L [14C]AA. Samples were incubated at 37°C for 5 minutes, after which the reaction was stopped by adding 1N HCl to yield pH 3 followed by 1 vol cold methanol. The products were subjected to HPLC analysis.

For the experiments concerning PGIS inactivation, an incubation of 180 U/mL of isolated COX-1 with 25 μmol/L [14C]AA at 30°C for 1 minute was used as a source of 14C-labeled endoperoxides. The reaction was stopped by adding an excess of indomethacin (final concentration, 10 μmol/L). In these conditions, the remaining untransformed [14C]AA was <10%, and the percentage of nondegraded endoperoxides, evaluated by the difference of PGF2α after addition of 1 vol methanol or 1 vol methanolic solution of 400 μg/mL SnCl2, was always >80%. Hence, the incubation mixture was immediately used as a source of labeled endoperoxides without further manipulation.

Effect of COX-2–Selective Inhibition on the Release of PGH2

IL-1β–treated HUVECs expressed both COX-1 and COX-2 in all experimental conditions. Therefore, in order to determine the concentration of the selective COX-2 inhibitor SC-5812539 44 (Laboratorios Almirall-Prodesfarma) to be used with HUVECs, cultures of PMA-treated human dermal fibroblasts and the HUV-EC-C line were used for COX-2 tests, and cultures of the HEL line and fresh human platelets were used for COX-1 tests. Fibroblasts and HUV-EC-C attached in 6-well plates, a 0.5-mL aliquot of HEL suspension (5×106 cells/mL), or a 0.15-mL platelet suspension were incubated at 37°C in the presence of the indicated concentration of SC-58125 for 5 minutes. [14C]AA (25 μmol/L) was later added, and cells were incubated for another 5 minutes. Reactions were stopped, and COX-derived eicosanoids were analyzed as described above. COX activity was evaluated as the sum of all the chromatographic peaks corresponding to COX-derived eicosanoids.

As a result of these experiments, SC-58125 concentrations of 1 and 10 μmol/L were chosen to study the effect of selective inhibition of COX-2 on the release of PGH2 by HUVECs. Resting HUVECs and HUVECs treated with 10 U/mL IL-1β overnight were incubated with 1 and 10 μmol/L of SC-58125 for 5 minutes. HUVECs were then incubated in the presence of 25 μmol/L [14C]AA alone or in the presence of 200 μg/mL SnCl2 for another 5 minutes as previously mentioned. The untransformed PGH2 was then estimated as described above.

Effect of IL-1β on the Inactivation of PGIS

To determine the PGIS activity after treatment of HUVECs with IL-1β, cells cultured in 6-well plates were treated with 10 U/mL IL-1β for 6 hours, without any additive or in the presence of final concentrations of the following: 1 mmol/L deferoxamine (Sigma), 1% (vol/vol) DMSO (Sigma), 1 mmol/L captopril (Laboratorios CINFA), 200 μmol/L pyrrolidinedithiocarbamate (PDTC, Sigma), 1 mmol/L 1,2,3-oxadiazolium, 5-amino-3-(4-morpholyn)-chloride (SIN-1, Cayman), 100 μmol/L 2-(4-carboxyphenyl)-4,4,5,5-tetramethylimidazoline-1-oxyl-3-oxide (carboxy-PTIO, Cayman), 50 U/mL superoxide dismutase (Sigma), 500 U/mL catalase (Sigma), 1 mmol/L NG-monomethyl-l-arginine (L-NMMA, Cayman), 5 mmol/L mannitol (Merck), or 100 μmol/L tyrosine (Sigma) (all dissolved in culture medium) or 0.5 mmol/L phenylbutazone, 10 μmol/L indomethacin, or 50 μg/mL vitamin E (all from Sigma) dissolved in ethanol (final ethanol concentration, 0.25% [vol/vol]). Parallel incubations of HUVECs with the scavengers and inhibitors in the absence of IL-1β were performed as controls. The culture medium was then removed, and HUVECs were treated with 100 μmol/L ASA in medium 199 for 10 minutes at 37°C. The ASA-containing medium was removed, and 200 μL of the solution containing the 14C-labeled endoperoxides was added. HUVECs were allowed to stand at 37°C for another 5 minutes. The final concentration of 14C-labeled endoperoxides was ≈18 μmol/L. 14C-labeled 6-keto-PGF1α was evaluated by HPLC as described. Resting (control) and IL-1β–treated HUVECs for each condition were always incubated with the same bach solution of 14C-labeled endoperoxides. Negative controls were performed using cells boiled for 15 minutes.

Synthesis of NO· in untreated and in IL-β–treated cells was determined by assay of the culture supernatants for NO2− (nitrite) using the Griess reagent, which contains 1 vol of 10 g/L sulfanilamine in 50 mL/L H3PO4, and 1 vol of 1 g/L N-(1-naphthyl)ethylenediamine dihydrochloride in water (all from Sigma). An aliquot of 500 μL of the culture medium was reacted with an equal volume of the Griess reagent for 10 minutes at room temperature in the dark. The optical densities of the assay samples were measured at 550 nm. Concentrations of NO2− were determined from a standard curve prepared using NaNO2 (Sigma) diluted in culture medium.

Statistics

Sigma-Stat software was used for statistical analysis. Statistical significance between groups was assessed using the Student t test for 2 groups. One-way ANOVA and the Student-Newman-Keus test were used to compare >2 groups. Correlations were evaluated by linear regression. A value of P<0.05 was considered significant.

Results

AA Metabolism in HUVECs

As expected, resting and IL-1β–treated HUVECs incubated with exogenous [14C]AA produced 6-keto-PGF1α, PGE2, and PGF2α as major eicosanoids, and PGD2, HHT, 15-HETE, and 11-HETE were minor compounds.

COX-1 and COX-2 Expression

The time course of mRNA and protein expression of COX-1 and COX-2 was measured by incubating HUVECs with or without 10 U/mL of IL-1β during various time intervals. Specific mRNAs encoding for both COX-1 and COX-2 were induced by IL-1β in a time-dependent manner (Figure 1⇓), although only the increase of COX-2 was statistically significant (P<0.05). The levels of COX-1 mRNA were slightly enhanced, whereas for COX-2 they were enhanced 12.5-fold by IL-1β. The maximum levels of COX-2 mRNA were observed between 3 and 6 hours of exposure to IL-1β.

Figure 1.
Figure 1.

Top, Levels of specific mRNA for COX-1 and COX-2 normalized with respect to GAPDH in HUVECs as a function of time of treatment with 10 U/mL IL-1β. Values are mean+SD (n=4). *P<0.05 compared with time 0. Bottom, Representative photographs of electrophoresis of RT-PCR samples corresponding to the indicated periods of time of exposure to IL-1β. Std indicates standard.

The Western blot analysis of the COX isoforms present in HUVECs is shown in Figure 2⇓. In the absence of IL-1β, the antibody against COX-1 recognized a band in HUVEC samples of ≈70 kDa corresponding to the migration of purified COX-1 from ram seminal vesicles. The antibody against COX-2 did not detect the purified COX-1 from ram seminal vesicles and had a slight reaction with a protein band of the unstimulated cells. In contrast, after 1 hour of IL-1β treatment, the antibody against COX-2 peptide recognized a protein band in HUVEC samples of ≈70 kDa corresponding to the migration of purified COX-2 from sheep placenta. The densitometric analysis of the bands showed that after 3 hours, IL-1β significantly increased the expression of COX-2 in a time-dependent manner, whereas modifications in the protein band corresponding to COX-1 over time were minimal. COX-2 expression reached a maximum level between 6 and 9 hours of IL-1β exposure, followed by a slight decrease at 24 hours. Figure 2⇓ shows that the total COX activity also increased in a time-dependent manner with the IL-1β treatment, although it was only statistically significant after 6 hours of exposure to the cytokine. It was notable that whereas COX-2 protein levels decreased after 24 hours of exposure to IL-1β, the total activity continuously increased until 24 hours.

Figure 2.
Figure 2.

Top, COX activity and protein expression as a function of time of exposure to IL-1β. HUVECs were incubated with 10 U/mL IL-1β for the indicated periods of time and were then incubated with 25 μmol/L [14C]AA for 10 minutes. The COX activity was evaluated as the sum of all HPLC peaks corresponding to eicosanoids formed through the COX pathway. Values are mean±SD (n=3). Cell lysates were subjected to Western blotting analysis as described in Material and Methods. The relative optical density (OD) of the bands, normalized for the darkest band on each gel, is shown to facilitate comparisons (mean±SD, n=4). *P<0.05 compared with time 0. Bottom, Representative photographs of Western blots. Isolated COX-1 (from ram seminal vesicles) and COX-2 (from sheep placenta) were used as positive controls.

Evaluation of Release of Prostanoids by HUVECs in Response to IL-1β

Release of prostanoids by resting and 10 U/mL IL-1β–treated cells was determined by monitoring the most significant prostanoids over 24 hours: PGE2, PGF2α, and 6-keto-PGF1α. Results in Figure 3⇓ show that after 24 hours, treatment with IL-1β caused a higher increment in the quantity of PGF2α (≈20-fold) and PGE2 (≈16.5-fold) than of 6-keto-PGF1α (≈3.5-fold) accumulated in the medium compared with no treatment. Levels of 6-keto-PGF1α in the medium were abruptly enhanced 1 hour after exposure of cells to IL-1β, whereas PGE2 and PGF2α were released into the medium more slowly, following the pattern of COX-2 expression. We should point out that 6-keto-PGF1α was the only compound whose accumulated quantity in the culture medium was significantly enhanced (≈7-fold) 1 hour after exposure to IL-1β, when COX-2 protein and activity was not enhanced (Figure 2⇑). Release of PGE2, PGF2α, and 6-keto-PGF1α by control cells was quantitatively slight, and the 3 progression curves had similar patterns.

Figure 3.
Figure 3.

Time course of IL-1β–stimulated release of prostanoids. HUVECs were incubated with (IL-1β) and without (control) 10 U/mL IL-1β for the indicated periods of time. Prostanoids in the medium were analyzed by RIA (n=3, mean±SD). *P<0.05 for IL-1β–treated cells compared with control cells.

Biosynthesis of Eicosanoids From Exogenous AA

To observe the differences concerning synthesis of the different eicosanoids between those cells that overexpress COX-2 and resting cells, progression curves after the addition of labeled substrate were evaluated, and results are shown in Figure 4⇓. The maximum levels of 6-keto-PGF1α and PGF2α were similar in both resting and IL-1β–stimulated HUVECs (10 U/mL IL-1β for 24 hours). Nevertheless, the amount of PGI2 increased to a maximum 10 minutes after substrate addition in resting HUVECs, whereas the synthesis of PGI2 practically ceased 1 minute after addition in the cells treated with IL-1β. Furthermore, the amount of 6-keto-PGF1α produced after 10 minutes by IL-1β–treated cells was significantly lower than that produced by control cells. HHT was produced in a greater amount in IL-1β–treated than in resting HUVECs. However, production stopped 1 minute after incubation, and the progression curve showed a pattern similar to that of 6-keto-PGF1α. In contrast, an increasingly higher production of PGE2, PGD2, and HETEs was observed in IL-1β–treated cells compared with control cells (Figure 4⇓).

Figure 4.
Figure 4.

Time course of prostanoid formation by HUVECs. Control cells and cells treated with 10 U/mL IL-1β for 24 hours were incubated at 37°C for the indicated periods of time with 25 μmol/L [14C]AA, and eicosanoids were then analyzed by HPLC (n=3, mean±SD). *P<0.05 for IL-1β–treated cells compared with control cells.

Effect of IL-1β on the Expression of PGIS

Results presented in Figure 4⇑ suggest that IL-1β does not exert any effect on PGIS expression. To confirm this, PGIS mRNA levels were determined by RT-PCR, and results are shown in Figure 5⇓. Results from 4 experiments indicated that PGIS mRNA was not increased in HUVECS exposed overnight to IL-1β compared with control cells.

Figure 5.
Figure 5.

Levels of specific mRNA for PGIS expression normalized with respect to GAPDH in HUVECs that were untreated (control) and treated with 10 U/mL IL-1β overnight. Data in the graph (bottom panel) are the mean of 4 separate experiments, and representative photographs (top panel) of electrophoresis of PCR samples corresponding to the indicated number of cycles are also shown. St indicates standard.

Characterization of the Biosynthetic Pathway of PGE2 and PGD2 in HUVECs

The slow time course of PGE2, PGD2, and HETE formation and the fact that they were the only eicosanoids that increased as a consequence of the induction of the COX-2 expression when an excess of exogenous substrate was supplied suggested that all of these eicosanoids could be formed in HUVECs by the exclusive intervention of COX. To verify that transformation of PGH2 into PGE2 and PGD2 was not enzymatic, 2 different experimental approaches were carried out.

Figure 6⇓ shows representative chromatograms corresponding to samples of HUVECs and isolated COX incubated with exogenous [14C]AA. Isolated COX-1 and COX-2 yield a ratio of PGE2 to PGD2 of 4.2±0.8 and 4.4±1.2 (mean±SD, n=5), respectively, similar to the ratio obtained in samples from HUVECs (3.9±0.99) (n=10). No 6-keto-PGF1α was detected in the isolated COX incubations. In addition, the presence of SnCl2 in the incubations of HUVECs with [14C]AA caused a complete diversion of the transformation of HUVEC-derived PGH2 toward PGF2α instead of toward PGE2 plus PGD2 (Figure 6⇓).

Figure 6.
Figure 6.

Representative chromatograms of samples from IL-1β–treated HUVECs (24 hours with 10 U/mL) incubated with 25 μmol/L of [14C]AA for 10 minutes in the absence and in the presence of SnCl2 and from 5 U of COX isolated from ram seminal vesicles incubated with 5 μmol/L [14C]AA for 5 minutes.

Comparison Between the Metabolism of Exogenous and Endogenous AA as a Function of Time of Incubation With IL-1β

After treatment of HUVECs with 10 U/mL IL-1β for different periods of time, cells were incubated in the presence of 25 μmol/L [14C]AA or 50 μmol/L histamine. Histamine-induced release of endogenous AA and the formation of the major PGs were evaluated by RIA. Results in Figure 7⇓ show some differences between the effect of IL-1β on the metabolism of exogenous and endogenous AA. When substrate was supplied exogenously, we observed a slight reduction in the ability of HUVECs to form PGI2 (evaluated as 6-keto-PGF1α) and PGF2α as a function of time of incubation with IL-1β. In contrast, the ability to form PGE2 and PGD2 increased with the time of exposure to the cytokine and correlated with COX-2 expression. Whereas formation of PGI2 and PGF2α from endogenous substrate increased suddenly after 1 hour of incubation with IL-1β and a sustained ability to form these PGs was then observed, statistically significant correlations were observed only between the increment in COX-2 protein expression and the increment in PGE2 and PGD2 formation from exogenous substrate and between the increment in COX-2 protein expression and the increment in PGE2 formation in response to histamine (formation of PGD2 from endogenous substrate was not determined; Figure 7⇓).

Effect of IL-1β on the Release of Untransformed PGH2 by HUVECs From Exogenous and Endogenous Substrate

The amount of untransformed PGH2 released by HUVECs from exogenous [14C]AA was estimated as the difference in the labeled PGF2α between the samples incubated in the presence and absence of SnCl2. After any period of time of exposure to IL-1β, the formation of 6-keto-PGF1α and HETEs from exogenous substrate was only minimally modified by the presence of SnCl2 in the medium. In contrast, PGE2 and PGD2 completely disappeared in the samples containing SnCl2, and the levels of PGF2α increased concomitantly. HHT levels were also substantially reduced in the presence of SnCl2 but were not totally suppressed (not shown). Figure 8⇓ shows the estimated release of PGH2 and PGI2 (evaluated as 6-keto-PGF1α) relative to the resting cells as a function of time of incubation with IL-1β. The increment in the release of PGH2 into the medium was statistically correlated with the increase in COX-2 protein and with the decrement in PGI2 formation (Figure 8⇓). The ratio of PGH2 to PGI2 increased ≈4-fold as a consequence of treatment with IL-1β.

Figure 8.
Figure 8.

Increment of untransformed PGH2 and PGI2 released by HUVECs as a function of time of exposure to IL-1β relative to resting HUVECs (top panel). The PGH2 released by HUVECs was estimated as the difference between the PGF2α produced by HUVECs from exogenous AA in the presence and in the absence of SnCl2. PGI2 was evaluated as 6-keto-PGF1α. Values are mean±SD (n=3). Correlations between the increment in PGH2 released and in COX-2 protein and between the increment in PGH2 released and the decrement in PGI2 are also shown (middle and bottom panels). OD indicates optical density. COX-2 protein after cell treatment with the cytokine for the indicated periods of time was evaluated by Western blotting in independent experiments (n=4).

PGE2, PGF2α and 6-keto-PGF1α released from endogenous sources in HUVECs stimulated with histamine were also evaluated, after several periods of time of exposure to IL-1β, both in the absence and presence of SnCl2. Formation of 6-keto-PGF1α was lower in the presence of SnCl2 than in its absence (Figure 9⇓), in contrast with results from the experiments adding exogenous [14C]AA. As expected, PGE2 levels were substantially lower in the presence of SnCl2 but were not totally suppressed. In addition, the production of PGF2α was similar both in the absence and in the presence of SnCl2 in particular when the activity of COX was maximum. Data from exogenous substrate experiments indicated that SnCl2 did not modify COX and PGIS activities. Hence, the lower levels of prostanoids observed in the presence of SnCl2 should be due to a less effective release of AA. This led to a decreased formation of PGH2 and therefore a decreased production of PGI2. The less effective mobilization of endogenous AA would also explain why PGF2α did not increase as much as expected as a consequence of the diversion of the PGH2 toward formation of PGF2α instead of PGE2 and PGD2. Hence, although the results indicated that untransformed PGH2 was also released from endogenous AA, the estimation of the amount of PGH2 released was not possible when this indirect method was used.

Figure 9.
Figure 9.

Production of the major prostanoids formed by HUVECs from endogenous AA in response to histamine as a function of time of exposure to IL-1β in the presence and in the absence of SnCl2 in the medium. Cells were treated with 10 U/mL IL-1β for the indicated period of time and were later challenged with 50 μmol/L histamine for 10 minutes in the presence and in the absence of SnCl2. Values are mean±SD (n=3). *P<0.05 compared with time 0.

To evaluate the untransformed PGH2 released by HUVECs from endogenous sources, PGH2 was trapped by ASA-treated platelets, and its transformation by Tx synthase was evaluated. First, platelets treated with 200 μmol/L ASA were added to the wells of HUVECs and incubated in the presence of [14C]AA. ASA-treated platelets did not produce any eicosanoid derived from COX activity. Preliminary experiments served to choose the appropriate time to add the platelet suspensions to the HUVECs. COX-inactivated platelets were added to the HUVECs treated with IL-1β at increasing periods of time after the addition of 25 μmol/L [14C]AA to the HUVECs. The maximum conversion of labeled PGH2 into labeled TxB2 and HHT was observed when 108 platelets were added to HUVECs 0.5 minutes later than the addition of [14C]AA (not shown). Therefore, a delay of 0.5 minutes between the addition of [14C]AA and platelet suspensions was chosen for the following experiments.

Results in the Table⇓ show that the presence of platelets in the incubation mixture did not modify the amount of PGI2 produced by HUVECs from endogenous or exogenous AA. ASA-treated platelets trapped HUVEC-derived PGH2 from both exogenous and endogenous AA. HHT in platelets is formed by the action of Tx synthase.20 The ratio of HHT to TxB2 formed in our conditions was evaluated by incubating suspensions of platelets untreated with ASA with 25 μmol/L [14C]AA and determining both compounds by HPLC. This ratio was 1.47±0.18 (mean±SD, n=5), which allowed us to estimate the amount of PGH2 transformed by ASA-treated platelets into TxB2 plus HHT in the experiments using histamine. The estimated amounts of PGH2 trapped by platelets from HUVECs incubated with [14C]AA and those incubated with 50 μmol/L histamine were comparable, indicating a substantial release of PGH2 from endogenous AA.

View this table:
Table 1.

Production of PGI2 (Evaluated as 6-Keto-PGF1α) and Amount of PGH2 Produced by IL-1β–Stimulated HUVECs That Was Trapped by ASA-Treated Platelets

Effect of the Selective Inhibition of COX-2 on the Release of PGH2

To select the most suitable concentration of the selective inhibitor in our experimental system, concentration-dependent inhibition curves on both COX-1 and COX-2 whole-cell systems were obtained by using human platelets and HEL line suspensions for COX-1 and the HUV-EC-C line and human dermal fibroblasts stimulated with PMA for COX-2. The [14C]AA concentration used was the same as that in HUVEC experiments. The COX activity in these systems was measured as the sum of all COX-derived compounds. As a result of these experiments, we chose 1 and 10 μmol/L SC-58125, since at these concentrations, inhibition of COX-2 was high and COX-1 was not substantially inhibited (see Figure 10⇓). As expected, the inhibition of the release of PGH2 evaluated using the SnCl2 indirect method was consistent with a major COX-2 origin.

Figure 10.
Figure 10.

Effect of the selective COX-2 inhibitor SC-58125 on the release of PGH2 (bars). HUVECs treated with 10 U/mL IL-1β overnight were incubated with and without the indicated concentrations of SC-58125 for 5 minutes before the addition of 25 μmol/L [14C]AA. The release of PGH2 was then estimated by the indirect method of the SnCl2. Error bars represent individual values (n=2). Inhibition curves of SC-58125 of COX-2 and COX-1 in whole-cell systems are also shown. The HUV-EC-C line and human dermal fibroblasts treated with PMA were used as COX-2 systems. In the experimental conditions used, COX-1 was slightly expressed in both fibroblasts and the HUV-EC-C line (not shown). The HEL line and human platelets were used as the COX-1 system. COX activity was evaluated as a sum of all COX-derived eicosanoids formed after incubation of cells with [14C]AA. Values are mean±SD (n=4).

Inactivation of PGIS During Incubation of HUVECs With IL-1β

After treatment of HUVECs with 10 U/mL IL-1β for 6 hours, cells were incubated with 14C-labeled endoperoxides, and the formation of PGI2 as 6-keto-PGF1α was measured. A significant decrease in the formation of PGI2 was observed in cells after IL-1β treatment compared with resting cells. This decrease was similar to that observed when cells were incubated with [14C]AA (Figure 11⇓). These results clearly indicate that PGIS was partially inactivated during the treatment with IL-1β.

Figure 11.
Figure 11.

Effect of several radical scavengers and inhibitors on the inactivation of PGIS during the exposure of HUVECs to IL-1β. Cells were treated with 10 U/mL IL-1β for 6 hours in the absence of any drug or in the presence of final concentration of 1 mmol/L captopril, 500 U/mL catalase, 100 μmol/L carboxy-PTIO (C-PTIO), 1 mmol/L deferoxamine, 1% (vol/vol) DMSO, 10 μmol/L indomethacin, 5 mmol/L mannitol, 1 mmol/L L-NMMA, 0.5 mmol/L phenylbutazone, 200 μmol/L PDTC, 50 U/mL superoxide dismutase (SOD), 100 μmol/L tyrosine (L-Tyr), 50 μg/mL vitamin E, or the indicated combinations of substances. Parallel incubations in the absence of IL-1β were performed as controls. a-Tocopherol indicates α-tocopherol. Cells were then incubated with [14C]endoperoxides as described in Materials and Methods. Results are expressed as percentage of increment with respect to their own controls. Values are mean±SD; the numbers in parentheses indicate the number of experiments. *P<0.05 compared with their own controls; #P<0.05 compared with the cells without scavengers or inhibitors (none); and @P<0.05 compared with C-PTIO alone.

To approach the mechanism of inactivation of PGIS in response to IL-1β, cells were incubated with several radical scavengers and inhibitors during exposure to IL-1β. Vitamin E and extracellular addition of catalase and superoxide dismutase did not exert any effect. In contrast, the Fe3+-chelating agent deferoxamine and the free radical scavengers PDTC and DMSO totally reverted the effect of IL-1β. Mannitol and captopril also significantly prevented the effect of the cytokine. Indomethacin and phenylbutazone partially avoided the effect of IL-1β, but no statistical significance was achieved. The effect of IL-1β was totally suppressed by the NO synthase L-NMMA but not by extracellular addition of l-tyrosine. This was consistent with a significant 1.5-fold increase of the NO2− levels in the culture medium of IL-1β–treated cells compared with control cells (492.3±117 and 320.5±63.3 pmol/106 cells for 6 hours, respectively; mean±SD, n=6, P<0.05). In contrast, the presence the NO· scavenger carboxy-PTIO dramatically increased the PGIS inhibitory effect of IL-1β. The effect of carboxy-PTIO was significantly prevented by DMSO but not by l-tyrosine. The inhibitors or scavengers referred in Figure 11⇑ did not exert any statistically significant effect on the PGIS activity in IL-1β–untreated cells. Additional experiments were performed exposing the cells to 1 mmol/L of the NO· and ·O2− (peroxynitrite) donor SIN-1 in the absence or presence of 10 U/mL IL-1β for 6 hours, and a significant inhibition of PGIS was achieved in resting and IL-1β–treated cells (51.51±16.14% and 41.2±19.9%, respectively; mean±SD, n=3).

Discussion

In 1978, Marcus et al8 suggested that PGE2 and PGD2 could be formed nonenzymatically in HUVECs. Their suggestion was based on the observation that PGH2 incubated in the absence of cells yielded a mixture of all the prostanoids except PGI2. Nevertheless, COX-2 had not yet been discovered at that time. We have now extended this research in order to verify the hypothesis that endothelial cells express only 2 enzymes that are involved in prostanoid biosynthesis, COX and PGIS, and that when COX-2 is overexpressed, accumulated PGH2 is effectively released outside the cell.

Our results indicate that HUVECs do not transform PGH2 into PGE2 and PGD2 enzymatically. This concept is supported by the following: (1) Progression curves indicate that the formation of PGE2 and PGD2 is substantially slower than that of the other prostanoids and overlap those of 15-HETE and 11-HETE, whose formation does not require later enzymatic steps.45 (2) The ratio of PGE2 to PGD2 formed by HUVECs from exogenous substrate was almost identical to that yielded by isolated COX. (3) The presence of SnCl2 totally diverted exogenous AA-derived PGH2 toward PGF2α instead of PGE2 and PGD2.

In agreement with other reports,16 33 34 36 46 we observed that release of PGI2 by HUVECs was increased in cells exposed to IL-1β compared with resting cells. However, the release of PGF2α and PGE2 was substantially more induced by IL-1β than was the release of PGI2. The mobilization of AA from cell lipids induced by IL-1β46 47 could account for the fact that the cytokine-induced release of PGI2 occurred (Figure 3⇑) before an appreciable increase in the COX expression and activity (Figure 2⇑). In contrast, a slow PGF2α and PGE2 accumulation in the medium was observed as COX-2 increased. This indicated that a COX-2–dependent excess of PGH2, which was not transformed into PGI2, was formed during the incubation with IL-1β. On the other hand, the ability of cells to form PGE2 and PGD2 from exogenous AA was greatly enhanced with the expression of COX-2 (Figure 7⇓). The same phenomenon was observed when mobilization of endogenous AA was induced by histamine; the ability to synthesize PGE2 (in this case PGD2 was not measured) also correlated with COX-2 expression (Figure 7⇓). This was consistent with the data reported by Bull et al,36 who found that PGE2 was the most enhanced eicosanoid in response to histamine after pretreatment of dermal microvascular endothelial cells with IL-1β. The fact that the increased ability to form PGI2 in response to histamine was already observed after 1 hour of IL-1β-exposure, when COX activity was not appreciably induced, suggests that IL-1β exerted priming in the histamine-induced AA mobilization.

Figure 7.
Figure 7.

Production of the major prostanoids formed by HUVECs from exogenous and endogenous substrate as a function of time of exposure to IL-1β (top panels). Cells were treated with 10 U/mL IL-1β for the indicated periods of time. They were later incubated with 25 μmol/L [14C]AA or with 50 μmol/L histamine for 10 minutes. Eicosanoids were analyzed by HPLC or RIA. Values are mean±SD (n=3). Correlations between the IL-1β–increased prostanoids and the increment in COX-2 are also shown (bottom panels). OD indicates optical density. COX-2 protein expression after cell treatment with the cytokine for the indicated periods of time was evaluated by Western blotting in independent experiments (n=4).

Since biosynthesis of PGE2 and PGD2 are nonenzymatic in HUVECs, PGH2 could be transformed into a mixture of these position isomers outside the cell. The ratio of untransformed [14C]AA-derived PGH2 to PGI2 released by HUVECs as a function of time of exposure to IL-1β overlapped with the increase in the COX activity and COX-2 expression. This indicates that the presence of COX-2 causes a dramatic increase in the ratio of PGH2 to PGI2 released by HUVECs when exogenous substrate is supplied. The fact that the COX-2–selective inhibitor SC-5812539 44 inhibited, in a concentration-dependent manner, the release of PGH2 at concentrations that did not appreciably inhibit COX-1 also supports this conclusion (Figure 10⇑). A substantial release of PGH2 formed from endogenous AA could be demonstrated by trapping with ASA-treated platelets. In contrast to that which occurred in the exogenous substrate experiments, SnCl2 was unable to completely suppress PGE2 formation when IL-1β–treated cells were stimulated with histamine. This may be due to the endogenous and exogenous AA having access to the COX located in different intracellular compartments. PGH2 generated from endogenous sources could be partially converted nonenzymatically into PGE2 inside the cell.

The formation of PGF2α by 2-electron reduction of PGH2 could occur nonenzymatically inside or partially outside the cells. Our results are not conclusive with respect to these points. Nevertheless, unlike 6-keto-PGF1α, formation of PGF2α showed a slow progression curve consistent with a nonenzymatic reduction of PGH2. In contrast with that which occurred with PGE2 and PGD2, PGF2α did not increase with the COX activity when an excess of exogenous substrate was supplied. This was consistent with an enzymatic and/or a electron donor–dependent synthesis. These facts taken together strongly suggest that formation of PGF2α in endothelial cells is limited by the presence of some cell-derived electron donor rather than by an enzymatic activity posterior to COX. The cellular nature of the electron donor is suggested by the fact that the relative quantity of PGF2α formed in the incubations with isolated COX was negligible. We can speculate that a slow turnover of such an electron donor(s) could account for the predominant formation of PGF2α when the slow release of prostanoids caused by the presence of IL-1β was measured (Figure 3⇑). In contrast, when an excess of exogenous AA was supplied, or a high quantity of AA was suddenly released by the action of histamine, a great amount of PGH2 was rapidly formed. As a consequence, this factor(s) could be rapidly depleted and finally exhausted, thus limiting the transformation of PGH2 into PGF2α. Under such conditions, PGE2 and PGD2 would be the predominant nonenzymatic transformations of PGH2 (Figures 4⇑ and 7⇑).

These considerations lead to the idea that COX and PGIS are the only enzymes involved in the biosynthesis of prostanoids in endothelial cells. Moreover, progression curves indicate that PGE2, PGD2, and PGF2α were formed in detectable amounts, mainly when PGIS became inactive, which occurred faster in the IL-1β–treated cells than in the resting HUVECs (Figure 4⇑). The faster inactivation of PGIS under COX-stimulated conditions yielding an excess of PGH2 is consistent with the “suicide” behavior of PGIS.48 The fact that the levels of mRNA for PGIS were not induced by IL-1β indicates that the only enzyme activity regulated at the transcriptional level was COX, particularly COX-2. This was consistent with the fact that the PGI2 that formed when an excess of exogenous substrate was supplied did not increase according to the increment of COX-2 protein and activity. Moreover, PGI2 formation ability was reduced as a function of time of treatment with IL-1β, reaching a minimum 6 to 9 hours after the addition of IL-1β (Figure 8⇑), when the expression of COX was maximum (Figure 2⇑).

The partial inactivation of PGIS as a consequence of exposure of HUVECs to IL-1β was further confirmed by supplying 14C-labeled endoperoxides and analyzing the formation of labeled 6-keto-PGF1α. The inhibition of PGIS observed after exposure to IL-1β using labeled endoperoxides as substrate was comparable to that obtained the addition of [14C]AA (Figure 10⇑). The partial inactivation of PGIS also contributed to the accumulation of an excess of PGH2, and the increase of PGH2 released by HUVECs correlated not only with COX-2 expression but also with PGIS inactivation (Figure 8⇑).

Evidence that the HO· radical was involved in the inactivation of PGIS was provided by the following: (1) The use of DMSO, PDTC, mannitol, and captopril, powerful HO· scavengers,49 50 51 52 partially or totally antagonized the effect of IL-1β. (2) Iron-dependent HO· formation from ·O2−-generating systems such as Fenton or Haber-Weiss reactions was inhibited by deferoxamine, a powerful Fe3+ chelator,49 which also blocked the inactivation of PGIS induced by IL-1β. (3) α-Tocopherol did not antagonize the IL-1β–induced inactivation of PGIS, since α-tocopherol did not react efficiently with HO·.53 Catalase and superoxide dismutase added to the extracellular phase did not exert any effect on the inactivation of PGIS, probably because of the fast intracellular generation of HO· from H2O2 generated by dismutation of ·O2− or from homolytic fission of ONOOH, with the extracellular presence of these enzymes being not effective.

Peroxidase activity as a source of free radicals has been reported to be involved in the inactivation of PGIS.54 Deferoxamine and PDTC are also inhibitors of the peroxidase activity of COX.55 Since most of the COX inhibitors do not exert any effect on peroxidase activity,55 the fact that indomethacin and phenylbutazone partially (but without statistical significance) inhibited the inactivation of PGIS by IL-1β would indicate that PGG2 contributed to the production of the inactivating oxygen-derived free radicals.

Another explanation has been given by Zou and Ullrich,56 who found that peroxynitrite irreversibly inhibits PGIS. Our results from L-NMMA experiments, which inhibited inactivation of PGIS by IL-1β, are consistent with those of these authors, who suggested that NO· formation is the limiting step in the peroxynitrite production in endothelial cells and may be induced by the inflammatory cytokines. Actually, exposition to IL-1β caused a significant 1.5-fold increase in the NO· released by HUVECs in terms of NO2−, although we did not evaluate the expression of the inducible NO synthase. The fact that the NO· scavenger carboxy-PTIO57 significantly increased the inhibitory effect of IL-1β in the present study was in apparent contradiction with this. Nevertheless, carboxy-PTIO equimolarly reacts with NO· to yield 2-(4-carboxyphenyl)-4,4,5,5-tetramethylimidazoline-1-oxyl plus NO2·.57 Homolytic fission of peroxynitrous acid, formed by protonation of peroxynitrite, yields NO2· and HO·.58 59 Zou et al60 recently reported that nitration of an active site–related tyrosine could account for the PGIS inactivation produced by peroxynitrite. These authors proposed an O-O homolytic cleavage in the peroxynitrite molecule caused by the heme iron of the PGIS. This could finally yield NO2+-nitrating species, which may attack tyrosine or tyrosinate.60 Nevertheless, nitration of tyrosyl residues by NO2· is also possible through the formation of tyrosyl radicals,61 62 which could be favored by HO·. Our results from carboxy-PTIO experiments are consistent with an inactivation of the PGIS mediated by NO2·. The fact that the presence of DMSO in addition of carboxy-PTIO decreased the ratio of inactivated PGIS, when compared with carboxy-PTIO alone, suggests that HO· was also involved. Consistently, another mechanism by which deferoxamine could prevent the effect of IL-1β on PGIS inactivation may be the direct reaction with peroxynitrite.59 We also observed that SIN-1, which releases both NO· and ·O2− as a source of peroxynitrite, inactivated PGIS in both resting and IL-1β–treated HUVECs. In summary, our results about the exact mechanism of the partial inactivation of PGIS caused by the exposure of HUVECs to IL-1β are not conclusive, but they strongly suggest that both NO· (which could generate peroxynitrite that undergoes homolytic cleavage of the O-O bond to yield NO2·) and HO· were involved.

Present results provide a biochemical explanation for the production and release of the constrictive prostanoids, PGH2 and PGF2α, by endothelial cells. Predominance of relaxing or constricting prostanoids will depend on both COX and PGIS activities. An impairment of endothelial cell function occurs under inflammatory conditions, and COX activity is increased by means of the induction of COX-2 and by partial inactivation of PGIS. Under such conditions, endothelial cells could produce an excess of PGH2 that could exert proinflammatory and prothrombotic activities before its conversion to other prostaglandins inside and outside the endothelium. PGH2 from endothelial cells enriched in COX-2 could also be converted into TxA2 by platelets and macrophages. Microvessel expression of COX-2 could contribute to the platelet activation observed in the disseminated intravascular coagulation, usually associated with systemic inflammatory responses. By use of an antagonist of the PGH2/TxA2 receptor, it has been shown that endogenous PGH2 decreases the thrombolytic activity of tissue plasminogen activator.61 PGH2 and TxA2 stimulate proliferation of vascular smooth muscle cells in vitro.62 63 64 65 66 Expression of COX-2 could also contribute to the myointimal hyperplasia observed in atherosclerosis by promoting the formation of PGH2 (and transcellular TxA2) and hydroxy linoleic acids,19 which elicit smooth muscle cell mitogenesis.67

In summary, our results show that resting HUVECs and HUVECs exposed to IL-1β produce only PGH2 and PGI2 enzymatically. Production of PGF2α could apparently be linked to the presence of a limiting intracellular electron donor. Overexpression of COX-2 turns the endothelial cell toward a phenotype characterized by an increased production of PGH2, which cannot be totally transformed into PGI2, partly because of its inactivation. Despite the fact that caution and more research is required to assess the biological relevance of our findings in vivo, the present data increase our knowledge about the metabolism of AA by endothelial cells and about the possible biological role of the induction of COX-2 in the vascular endothelium.

Acknowledgments

This study was supported by grants from Institut de Recerca of the HSCSP, FIS 94/1559, and SAF96-0138. The authors wish to thank Dr de Castellarnau for her suggestions and Cristina Gerbolés, Esther Gerbolés, and Aurora Sierra for their technical assistance.

Footnotes

  • Reprint requests to Dr Luis Vila, H.S. Creu i S. Pau (Casa de Convalecencia), S. Antonio Ma Claret 167, 08025-Barcelona, Spain.

  • Preliminary data were presented as a poster at the 9th International Conference on Prostaglandins and Related Compounds, Florence, Italy, June 6–10, 1994.

  • Received March 13, 1998.
  • Accepted May 18, 1998.

References

  1. ↵
    Vane JR, Änggard EE, Botting RM. Regulatory functions of the vascular endothelium. N Engl J Med. 1990;323:27–36.
    OpenUrlCrossRefPubMed
  2. ↵
    Moncada S, Vane JR. Pharmacology and endogenous roles of prostaglandin endoperoxides, thromboxane A2, and prostacyclin. Pharmacol Rev. 1979;30:293–331.
    OpenUrlPubMed
  3. ↵
    Gryglewski RJ, Botting RM, Vane JR. Mediators produced by endothelial cell. Hypertension. 1988;12:530–548.
    OpenUrlAbstract/FREE Full Text
  4. ↵
    Rees DD, Palmer RMJ, Hodson HF, Moncada S. A specific inhibitor of nitric oxide formation from L-arginine attenuates endothelium-dependent relaxation. Br J Pharmacol. 1989;96:418–424.
    OpenUrlCrossRefPubMed
  5. ↵
    Richard V, Tanner FC, Tschudi MR, Lüscher TF. Differential activation of the endothelial L-arginine pathway by bradykinin, serotonin and clonidine in porcine coronary arteries. Am J Physiol. 1990;259:H1433–H1439.
    OpenUrlAbstract/FREE Full Text
  6. ↵
    Tschudi M, Richard V, Bühler FR, Lüscher TF. Importance of endothelium-derived nitric oxide in intramyocardial porcine coronary arteries. Am J Physiol. 1990;160:H13–H20.
    OpenUrl
  7. ↵
    Lüscher TF. The endothelium in hypertension: bystander, target or mediator? J Hypertens. 1994;12(suppl 10):S105–S116.
  8. ↵
    Marcus AJ, Weksler BB, Jaffe EA. Enzymatic conversion of prostaglandin endoperoxide H2 and arachidonic acid to prostacyclin by cultured human endothelial cells. J Biol Chem. 1978;253:7138–7141.
    OpenUrlFREE Full Text
  9. ↵
    Alhenc-Gelas F, Tsai SJ, Callahan KS, Campbell WB, Johnson AR. Stimulation of prostaglandin formation by vasoactive mediators in cultured human endothelial cells. Prostaglandins. 1982;24:723–742.
    OpenUrlCrossRefPubMed
  10. ↵
    Kühn H, Pönicke K, Halle W, Wiesner R, Schewe T, Förster W. Metabolism of [1-14C]-arachidonic acid by cultured calf aortic endothelial cells: evidence for the presence of a lipoxygenase pathway. Prostaglandins Leukot Med. 1985;17:291–303.
    OpenUrlCrossRefPubMed
  11. ↵
    Hopkins NK, Oglesby TD, Bundy GL, Gorman RR. Biosynthesis and metabolism of 15-hydroperoxy-5,8,11,13-eicosatetraenoic acid by human umbilical vein endothelial cells. J Biol Chem. 1984;259:14048–14053.
    OpenUrlAbstract/FREE Full Text
  12. ↵
    López S, Vila L, Breviario F, de Castellarnau C. Interleukin-1 increases 15-hydroxyeicosatetraenoic acid formation in cultured human endothelial cells. Biochim Biophys Acta. 1993;1170:17–24.
    OpenUrlPubMed
  13. ↵
    Smith WL. Prostanoid biosynthesis and mechanisms of action. Am J Physiol. 1992;263:F181–F191.
    OpenUrlAbstract/FREE Full Text
  14. ↵
    Hla T, Ristimäki A, Appleby S, Barriocanal JG. Cyclooxygenase gene expression in inflammation and angiogenesis. Ann N Y Acad Sci. 1993;696:197–204.
    OpenUrlPubMed
  15. ↵
    Dinarello CA. Interleukin-1 and interleukin-1 antagonism. Blood. 1991;77:1627–1652.
    OpenUrlAbstract/FREE Full Text
  16. ↵
    Maier JAM, Hla T, Maciag T. Cyclooxygenase is an immediate-early gene induced by interleukin-1 in human endothelial cells. J Biol Chem. 1990;265:10805–10808.
    OpenUrlAbstract/FREE Full Text
  17. ↵
    Hla T, Neilson K. Human cyclooxygenase-2 cDNA. Proc Natl Acad Sci U S A.. 1992;89:7384–7388.
    OpenUrlAbstract/FREE Full Text
  18. ↵
    Jones DA, Carlton D, McIntyre MT, Zimmerman GA, Prescott SM. Molecular cloning of human prostaglandin endoperoxide synthase type II and demonstration of expression in response to cytokines. J Biol Chem. 1993;268:9049–9054.
    OpenUrlAbstract/FREE Full Text
  19. ↵
    Camacho M, Godessart N, Antón R, García M, Vila L. Interleukin-1 enhances the ability of cultured umbilical vein endothelial cells to oxidize linoleic acid. J Biol Chem. 1995;270:17279–17286.
    OpenUrlAbstract/FREE Full Text
  20. ↵
    Hecker M, Ullrich V. On the mechanism of prostacyclin and thromboxane A2 biosynthesis. J Biol Chem. 1989;264:141–150.
    OpenUrlAbstract/FREE Full Text
  21. ↵
    Honn KV, Cicone B, Skoff A. Prostacyclin: a potent antimetastatic agent. Science. 1981;212:1270–1272.
    OpenUrlAbstract/FREE Full Text
  22. ↵
    Nugteren DH, Christ-Hazelhof E. Chemical and enzymatic conversions of prostaglandin endoperoxide PGH2. Adv Prostaglandin Thromboxane Res. 1980;6:129–137.
    OpenUrlPubMed
  23. ↵
    Smith WL. Prostanoid biosynthesis and mechanisms of action. Am J Physiol. 1992;263:F181–F191.
  24. ↵
    Koga T, Takayata Y, Kobayashi K, Takishita S, Yamashita Y, Fujishima M. Age and hypertension promote endothelium-dependent contractions to acetylcholine in the aorta of the rat. Hypertension. 1989;14:542–548.
    OpenUrlAbstract/FREE Full Text
  25. ↵
    Pagano PJ, Lin L, Sessa WC, Nasjletti A. Arachidonic acid elicits endothelium-dependent release from the rabbit aorta of a constrictor prostanoid resembling prostaglandin endoperoxides. Circ Res. 1991;69:396–405.
    OpenUrlAbstract/FREE Full Text
  26. ↵
    Shimizu K, Muramatsu M, Kakegawa Y, Asano H, Toki Y, Miyazaki Y, Okumura K, Hashimoto H, Ito T. Role of prostaglandin H2 as an endothelium-derived contracting factor in diabetic state. Diabetes. 1993;42:1246–1252.
    OpenUrlAbstract/FREE Full Text
  27. ↵
    Williams SP, Campbell AK, Roscell N, Myatt L, Leikauf GD, Rapoport RM. Modulation of phorbol ester-induced contraction by endogenously released cyclooxygenase products in rat aorta. Am J Physiol. 1994;267:H1654–H1662.
    OpenUrlAbstract/FREE Full Text
  28. ↵
    Fu-Xiang D, Skopec J, Diederich A, Diederich D. Prostaglandin H2 and thromboxane A2 are contractile factors in intrarenal arteries of spontaneously hypertensive rats. Hypertension. 1992;19:795–798.
    OpenUrlPubMed
  29. ↵
    Lin L, Balazy M, Pagano PJ, Nasjletti A. Expression of prostaglandin H2–mediated mechanism of vascular contraction in hypertensive rats: relation to lipoxygenase and prostacyclin synthase activities. Circ Res. 1994;74:197–205.
    OpenUrlAbstract/FREE Full Text
  30. ↵
    Asano H, Shimizu K, Muramatsu M, Iwama Y, Toki Y, Miyazaki Y, Okumura K, Hashimoto H, Ito T. Prostaglandin H2 as an endothelium-derived contracting factor modulates endothelin-1-induced contraction. J Hypertens. 1994;12:383–390.
    OpenUrlPubMed
  31. ↵
    Dejana E, Breviario F, Balconi G, Rossi V, Remuzzi G, de Gaetano G, Mantovani A. Stimulation of prostacyclin synthesis in vascular cells by mononuclear cell products. Blood. 1984;64:1280–1283.
    OpenUrlAbstract/FREE Full Text
  32. ↵
    Rossi V, Breviario F, Ghezzi P, Dejana E, Mantovani A. Prostacyclin synthesis induced in vascular cells by interleukin-1. Science. 1985;229:174–176.
    OpenUrlAbstract/FREE Full Text
  33. ↵
    Albrightson CR, Baenziger NL, Needleman P. Exaggerated human vascular cell prostaglandin biosynthesis mediated by monocytes: role of monokines and interleukin 1. J Immunol. 1985;135:1872–1877.
    OpenUrlAbstract
  34. ↵
    Rustin MHA, Bull HA, Dowd PM. Effect of human recombinant interleukin-1α on release of prostacyclin from human endothelial cells. Br J Dermatol. 1989;120:153–159.
    OpenUrlCrossRefPubMed
  35. ↵
    Bull HA, Dowd PM. Interleukin-1 potentiates histamine-induced release of prostacyclin from human endothelial cells. Br J Pharmacol. 1990;101:703–709.
    OpenUrlPubMed
  36. ↵
    Bull HA, Cohen J, Dowd PM. Responses of human dermal microvascular endothelial cells to histamine and their modulation by interleukin 1 and substance P. J Invest Dermatol. 1991;97:787–792.
    OpenUrlCrossRefPubMed
  37. ↵
    Ristimäki A, Garfinkel S, Wessendorf J, Maciag T, Hla T. Induction of cyclooxygenase-2 by interleukin-1α evidence for post-transcriptional regulation. J Biol Chem. 1994;269:11769–11775.
    OpenUrlAbstract/FREE Full Text
  38. ↵
    Godessart N, Camacho M, López-Belmonte J, Antón R, García M, de Moragas J-M, Vila L. Prostaglandin H-synthase-2 is the main enzyme involved in the biosynthesis of octadecanoids from linoleic acid in human dermal fibroblasts stimulated with IL-1β. J Invest Dermatol. 1996;107:726–732.
    OpenUrlCrossRefPubMed
  39. ↵
    Miralpeix M, Camacho M, López-Belmonte J, Canalías F, Beleta J, Palacios JM, Vila L. Selective induction of cyclooxygenase-2 activity in the permanent human endothelial cell line HUV-EC-C: biochemical and pharmacological characterization. Br J Pharmacol. 1997;121:171–180.
    OpenUrlCrossRefPubMed
  40. ↵
    Vila L, Cullaré C, Solá J, Puig L, de Castellarnau C, de Moragas JM. Cyclooxygenase activity is increased in platelets from psoriatic patients. J Invest Dermatol. 1991;97:922–926.
    OpenUrlCrossRefPubMed
  41. ↵
    Hamberg M, Svensson J, Wakabayashi T, Samuelsson B. Isolation and structure of two prostaglandin endoperoxides that cause platelet aggregation. Proc Natl Acad Sci U S A.. 1974;71:345–349.
    OpenUrlAbstract/FREE Full Text
  42. ↵
    Solá J, Godessart N, Vila L, Puig L, de Moragas JM. Epidermal cell-polymorphonuclear leukocyte cooperation in the formation of leukotriene B4 by transcellular biosynthesis. J Invest Dermatol. 1992;98:333–339.
    OpenUrlCrossRefPubMed
  43. ↵
    Chomczynski P, Sacchi N. Single-step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction. Anal Biochem. 1987;162:156–159.
    OpenUrlCrossRefPubMed
  44. ↵
    Penning TD, Talley JJ, Bertenshaw SR, Carter JS, Collins PW, Docter S, Graneto MJ, Lee LF, Malecha JW, Miyashiro JM, Rogers RS, Rogier DJ, Yu SS, Anderson GD, Burton EG, Cogburn JN, Gregory SA, Koboldt CM, Perkins WE, Seibert K, Veenhuizen AW, Zhang YY, Isakson PC. Synthesis and biological evaluation of the 1,5-diarylpyrazole class of cyclooxygenase-2 inhibitors: identification of 4-[5-(4-methyl-phenyl)-3-(trifluoromethyl)-1H-pyrazol-1-yl]benzenesulfonamide (SC-58635, Celecoxib). J Med Chem. 1997;40:1347–1365.
    OpenUrlCrossRefPubMed
  45. ↵
    Hecker M, Ullrich V, Fischer C, Meese CO. Identification of novel arachidonic acid metabolites formed by prostaglandin H synthase. Eur J Biochem. 1987;169:113–123.
    OpenUrlPubMed
  46. ↵
    Breviario F, Proserpio P, Bertocchi F, Lampugnani MG, Mantovani A, Dejana E. Interleukin-1 stimulates prostacyclin production by cultured human endothelial cells by increasing arachidonic acid mobilization and conversion. Arteriosclerosis. 1990;10:129–134.
    OpenUrlAbstract/FREE Full Text
  47. ↵
    García JGN, Stasek JE, Bahler C, Natarajan V. Interleukin 1-stimulated prostacyclin synthesis in endothelium: lack of phospholipase C, phospholipase D, or protein kinase C involvement in early signal. J Lab Clin Med. 1992;120:929–940.
    OpenUrlPubMed
  48. ↵
    Wade ML, Voelkel NF, Fitzpatrick FA. “Suicide” inactivation of prostaglandin I2 synthase: characterization of mechanism-based inactivation with isolated enzyme and endothelial cells. Arch Biochem Biophys. 1995;321:453–458.
    OpenUrlCrossRefPubMed
  49. ↵
    Halliwell B, Gutteridge JMC. Protection against oxidants in biological systems: the superoxide theory of oxygen toxicity. In: Halliwell B, Gutteridge JMC, eds. Free Radicals in Biology and Medicine. 2nd ed. Oxford, UK: Oxford University Press; 1989:86–187.
  50. ↵
    Tsai J-C, Jain M, Hsieh C-M, Lee W-S, Yoshizumi M, Patterson C, Perrella MA, Cooke C, Wang H, Haber E, Schlegel R, Lee M-E. Induction of apoptosis by pyrrolidinedithiocarbamate and N-acetylcysteine in vascular smooth muscle cells. J Biol Chem. 1996;271:3667–3670.
    OpenUrlAbstract/FREE Full Text
  51. ↵
    Abello PA, Fidler SA, Bulkley GB, Buchman TG. Antioxidants modulate induction of programmed endothelial cell death (apoptosis) by endotoxin. Arch Surg. 1994;129:134–141.
    OpenUrlCrossRefPubMed
  52. ↵
    Bagchi D, Prasad R, Das DK. Direct scavenging of free radicals by captopril, and angiotensin converting enzyme inhibitor. Biochem Biophys Res Commun. 1989;158:52–57.
    OpenUrlCrossRefPubMed
  53. ↵
    Liebler DC. The role of metabolism in the antioxidant function of vitamin E. Crit Rev Toxicol. 1993;23:147–169.
    OpenUrlPubMed
  54. ↵
    Ham EA, Egan RW, Soderman DD, Gale PH, Kuehl FA Jr. Peroxidase-dependent deactivation of prostacyclin synthetase. J Biol Chem. 1979;254:2191–2194.
    OpenUrlAbstract/FREE Full Text
  55. ↵
    Tam SSC, Lee DHS, Wang EY, Munroe DG, Lau CY. Tepoxalin, a novel dual inhibitor of the prostaglandin-H synthase cyclooxygenase and peroxidase activities. J Biol Chem. 1995;270:13948–13955.
    OpenUrlAbstract/FREE Full Text
  56. ↵
    Zou M-H, Ullrich V. Peroxynitrite formed by simultaneous generation of nitric oxide and superoxide selectively inhibits bovine aortic prostacyclin synthase. FEBS Lett. 1996;382:101–104.
    OpenUrlCrossRefPubMed
  57. ↵
    Akaike T, Yoshida M, Miyamoto Y, Sato K, Kohno M, Sasamoto K, Miyazaki K, Ueda S, Maeda H. Antagonistic action of imidazolineoxyl N-oxides against endothelium-derived relaxing factor/·NO through a radical reaction. Biochemistry. 1993;32:827–832.
    OpenUrlCrossRefPubMed
  58. ↵
    Hogg N, Darley-Usmar VM, Wilson MT, Moncada S. Production of hydroxyl radicals from the simultaneous generation of superoxide and nitric oxide. Biochem J. 1992;281:419–424.
  59. ↵
    Beckman JS, Beckman TW, Chen J, Marshall PA, Freeman BA. Apparent hydroxyl radical production by peroxynitrite: implications for endothelial injury from nitric oxide and superoxide. Proc Natl Acad Sci U S A.. 1990;87:1620–1624.
    OpenUrlAbstract/FREE Full Text
  60. ↵
    Zou MH, Martin C, Ullrich V. Tyrosine nitration as a mechanism of selective inactivation of prostacyclin synthase by peroxynitrite. Biol Chem. 1997;378:707–713.
    OpenUrl
  61. ↵
    Prütz WA, Mönig H, Butler J, Land EJ. Reactions of nitrogen dioxide in aqueous model systems: oxidation of tyrosine units in peptides and proteins. Arch Biochem Biophys. 1985;243:125–134.
    OpenUrlCrossRefPubMed
  62. ↵
    Van der Vliet A, Eiserich JP, O’Neill CA, Halliwell B, Cross CE. Tyrosine modification by reactive nitrogen species: a closer look. Arch Biochem Biophys. 1995;319:341–349.
    OpenUrlCrossRefPubMed
  63. ↵
    Golino P, Rosolowsky M, Yao S-K, McNatt J, De Clerck F, Buja LM, Willerson JT. Endogenous prostaglandin endoperoxides and prostacyclin modulate the thrombolytic activity of tissue plasminogen activator: effects of simultaneous inhibition of thromboxane A2 synthase and blockade of thromboxane A2/prostaglandin H2 receptors in a canine model of coronary thrombosis. J Clin Invest. 1990;86:1095–1102.
  64. ↵
    Morinelli TA, Zhang L-M, Newman WH, Meier KE. Thromboxane A2/prostaglandin H2-stimulated mitogenesis of coronary artery smooth muscle cells involves activation of mitogen-activated protein kinase and S6 kinase. J Biol Chem. 1994;25:5693–5698.
    OpenUrl
  65. ↵
    Sachinidis A, Flesch M, Ko Y, Schrör K, Böhm M, Düsing R, Vetter H. Thromboxane A2 and vascular smooth muscle cell proliferation. Hypertension. 1995;26:771–780.
    OpenUrlAbstract/FREE Full Text
  66. ↵
    Jones DA, Benjamin CW, Linseman DA. Activation of thromboxane and prostacyclin receptors elicits apposing effects on vascular smooth muscle cell growth and mitogen-activated protein kinase signaling cascades. Mol Pharmacol. 1995;48:890–896.
    OpenUrlAbstract
  67. ↵
    Rao GN, Alexander RW, Runge MS. Linoleic acid and its metabolites, hydroperoxy-octadecadienoic acids, stimulate c-fos, c-jun, and c-myc mRNA expression, mitogen activated protein kinase activation, and growth in rat aortic smooth muscle cells. J Clin Invest. 1995;96:842–846.
View Abstract

This Issue

Jump to

Article Tools

Share this Article

  • Email
  • Rate of Vasoconstrictor Prostanoids Released by Endothelial Cells Depends on Cyclooxygenase-2 Expression and Prostaglandin I Synthase Activity
    Mercedes Camacho, Jesús López-Belmonte and Luis Vila
    Circulation Research. 1998;83:353-365, originally published August 24, 1998
    https://doi.org/10.1161/01.RES.83.4.353
    del.icio.us logo Digg logo Reddit logo Twitter logo CiteULike logo Facebook logo Google logo Mendeley logo

Related Articles

Cited By...