Cytochrome P4502C9-Derived Epoxyeicosatrienoic Acids Induce the Expression of Cyclooxygenase-2 in Endothelial Cells
http://www.100md.com
《动脉硬化血栓血管生物学》
From the Institut für Kardiovaskul?re Physiologie (U.R.M., R.B., I.F.) and Pharmazentrum Frankfurt, Institut für Klinische Pharmakologie and ZAFES (R.S.), Johann Wolfgang Goethe-Universit?t, Frankfurt am Main, Germany; Department of Biochemistry (J.R.F.), University of Texas Southwestern Medical Center, Dallas, Tex.
Correspondence to Ingrid Fleming, PhD, Vascular Signalling Group, Institut für Kardiovaskul?re Physiologie, Johann Wolfgang Goethe-Universit?t, Theodor-Stern-Kai 7, D-60590 Frankfurt am Main, Germany. E-mail fleming@em.uni-frankfurt.de
Abstract
Objective— Cytochrome P450 (CYP) epoxygenases metabolize arachidonic acid to epoxyeicosatrienoic acids (EETs). CYP2C9-derived EETs elicit endothelial cell proliferation and angiogenesis, but the signaling pathways involved are incompletely understood. Because cyclooxygenase-2 (COX-2) is involved in angiogenesis, we determined whether a link exists between CYP2C9 and COX-2 expression.
Methods and Results— Human umbilical vein endothelial cells were infected with CYP2C9 sense or antisense adenoviral constructs. Overexpression of CYP2C9 increased COX-2 promoter activity, an effect accompanied by a significant increase in COX-2 protein expression and elevated prostacyclin production. The CYP2C9-induced expression of COX-2 was inhibited by the CYP2C9 inhibitor, sulfaphenazole, whereas 11,12-EET increased COX-2 expression. Overexpression of CYP2C9 and stimulation with 11,12-EET increased intracellular cAMP levels and stimulated DNA-binding of the cAMP-response element-binding protein. The protein kinase A inhibitor, KT5720, attenuated the CYP2C9-induced increase in COX-2 promoter activity and protein expression. Overexpression of CYP2C9 stimulated endothelial tube formation, an effect that was attenuated by the COX-2 inhibitor celecoxib. Identical responses were observed in cells preconditioned by cyclic strain to increase CYP2C expression.
Conclusion— These data indicate that CYP2C9-derived EETs induce the expression of COX-2 in endothelial cells via a cAMP-dependent pathway and that this mechanism contributes to CYP2C9-induced angiogenesis.
Overexpression of cytochrome P450 (CYP) 2C9 in endothelial cells increased cAMP levels, stimulated the cAMP-response element-binding protein, and enhanced cyclooxygenase-2 (COX-2) promoter activity, protein expression, and prostacyclin production. CYP2C9 overexpression stimulated endothelial tube formation, which was attenuated by the COX-2 inhibitor celecoxib. Thus, COX-2 contributes to CYP2C9-induced angiogenesis.
Key Words: angiogenesis ? endothelium ? gene expression ? prostacyclin ? cytochrome P450
Introduction
Cytochrome P450 epoxygenases of the 2B, 2C, and 2J subfamilies are expressed in endothelial cells and metabolize arachidonic acid to epoxyeicosatrienoic acids (EETs) (5,6-EET, 8,9-EET, 11,12-EET, and 14,15-EET). Because EETs induce the hyperpolarization and relaxation of vascular smooth muscle cells, these eicosanoids were initially described as endothelium-derived hyperpolarizing factors.1 However, EETs are now appreciated to be more than simple vasodilators and exert numerous membrane potential-independent effects on endothelial cell signaling and vascular homeostasis.2
Whereas EETs can be released from endothelial cells, they also act as intracellular signal transduction molecules and are capable of activating tyrosine kinases and phosphatases, MAP kinases (extracellular signal regulated kinase 1/2, p38 MAP kinase), the phosphatidylinositol-3 kinase, and protein kinase B/Akt.2 Furthermore, EETs have been reported to stimulate the ADP-ribosylation of G-proteins and to prevent the activation of the transcription factor nuclear factor B.3,4 More recently, CYP2C9-derived EETs were found to induce endothelial cell proliferation and angiogenesis5,6 via a mechanism involving the induction of MAP kinase phosphatase-1 and cyclin D1, as well as the transactivation of the epidermal growth factor (EGF) receptor.6,7
Cyclooxygenases (COX) also metabolize arachidonic acid, leading to the production of prostaglandins and thromboxanes. There are 2 isoforms of this enzyme. The first (COX-1) is constitutively expressed in most tissues, whereas the expression of COX-2 can be induced by growth factors, cytokines, and vasoactive peptides such as endothelin.8 COX-derived products have been shown to be key mediators in inflammatory responses, and they also regulate vascular tone, as well as a number of signal transduction pathways affecting cell adhesion, growth, and differentiation. Although COX-2 expression is increased in inflammatory conditions, its products are not always pro-inflammatory. In the vasculature, COX-2 generates mainly the vasodilator and antithrombotic prostanoid, prostacyclin (PGI2),9 and should therefore be considered as vasoprotective.
Because CYP-derived EETs transactivate the EGF receptor,10 which has been linked to an increase in COX-2 expression,11 the aim of the present investigation was to determine whether an increase in CYP2C9 expression and activity could elicit an increase in COX-2 expression and whether COX-2 is involved in CYP2C9-induced angiogenesis.
Methods
Materials
11,12-EET was purchased from Cayman Chemicals (Massy, France), KT 5720 was from Calbiochem (La Jolla, Calif), and thrombin was from Hemochrom Diagnostica GmbH (Essen, Germany). Celecoxib was a generous gift from Dr Gerd Geisslinger (Frankfurt, Germany). Epoxyeicosa-5(Z)-enoic acid (14,15-EEZE) and N-methylsulfonyl-6-(2-propargyloxyphenyl) hexanamide (MS-PPOH) were synthesized as described.12,13 Sulfaphenazole and all other chemicals were from Sigma.
Cell Culture
Human umbilical vein endothelial cells (HUVECs) were purchased from Cell Systems/Clonetics (Solingen) and cultured as described.6 First- or second-passage endothelial cells were used in all experiments.
Adenoviral Infection
HUVECs (80% to 90% confluent) were serum-starved for 24 hours before infection with adenoviral vectors. Cells were incubated with recombinant adenoviruses expressing CYP2C9 sense or antisense cDNA (10 pfU/cell) in medium without antibiotics for 4 hours at 37°C, followed by recovery in the presence of 2% fetal calf serum. The efficiency of infection was between 90% and 100%.
Immunoblotting
For Western blot analysis, cells were lysed in Triton X-100 lysis buffer and separated by SDS-PAGE as described.6 Proteins were detected using antibodies against COX-2 (Santa Cruz Biotechnology, Santa Cruz, Calif) and ?-actin (Sigma). The CYP2C9 antibody used was purified by Eurogenetec (Seraing, Belgium) from rabbits immunized with a CYP2C9 peptide (RRRKLPPGPTPLPIC).
Reporter Gene Assay
Twelve hours after infection with CYP2C9 sense or antisense adenoviral constructs, endothelial cells were transfected with a COX-2 promoter construct (kindly provided by Dr Margarete Goppelt-Struebe, Erlangen, Germany), together with a LacZ construct (Invitrogen, Karlsruhe, Germany). After 24 hours, the cells were lysed, and luciferase activity was assayed according to the manufacturer’s protocol (Promega, Mannheim, Germany). Values were corrected for transfection efficiency by measuring ?-galactosidase activity.
Radioimmunoassays
Commercially available kits (both from Amersham Biosciences, Buckinghamshire, UK) were used to determine 6-keto-prostaglandin (PG) F1 (6-Keto-PGF1) levels in cell supernatants and intracellular cAMP levels in cells treated with 3-isobutyl-1-methylxanthine (30 μmol/L, 30 minutes) as described.14
Electrophoretic Mobility Shift Assay
Double-stranded oligonucleotides containing the sequence of the binding site for cAMP-response element-binding protein (CREB: 5'-AGAGATTCGCTGACGTCAGAGAGCTAG-3') were labeled with 32P-ATP using a Ready-to-go DNA-Labeling Kit (Amersham Biosciences, Piscataway, NJ). Nuclear proteins (6 to 10 μg) were incubated with the 32P-labeled oligonucleotides (10000 counts/sample) for 30 minutes in electrophoretic mobility shift assay-binding buffer (10 mmol/L HEPES, pH 7.5, 100 mmol/L NaCl, 1 mmol/L EDTA, 5% [v/v] glycerol, 1 mmol/L DTT, 1 μL/sample poly-dI/dC). Supershift assays were performed by adding a CREB antibody 30 minutes before incubation with the oligonucleotide. Protein–DNA complexes were resolved by gel electrophoresis, and the gels were fixed and dried before being exposed to x-ray films. Densitometric analysis of the autoradiographs was performed after nonsaturating exposures, and the values were expressed as the percentage of the binding activity detected in extracts from unstimulated cells.
In Vitro Angiogenesis Assay
Fibrin gels were prepared using thrombin (0.5 U/μL)-polymerized fibrinogen (1.5 mg/mL in MCDB 131). HUVECs were infected with CYP2C9 sense or antisense adenoviruses 24 hours after seeding onto the fibrin gels and cultured in MCDB 131, supplemented with 4% fetal calf serum, L-gutamine (10 mmol/L), basic fibroblast growth factor (0.5 ng/mL), epidermal growth factor (0.05 ng/mL), and endothelial cell growth-stimulating factor from bovine brain (ECGS/H, 0.2%), and in the absence or presence of celecoxib (1 μmol/L). After 14 days, photographs were taken and tube formation was assessed and total tube length determined.
Cyclic Strain
To induce CYP2C expression in cultured endothelial cells human umbilical vein endothelial cells were exposed to cyclic strain (12%, 1Hz) for 24 hours as described previously.15 After stimulation, the endothelial cells were recovered and seeded onto Matrigel; tube formation was assessed after 12 hours and quantified by counting the number of branch points.
Isolation of RNA and Reverse-Transcription Polymerase Chain Reaction
Total RNA was isolated from cultured HUVECs using phenol and guanidine isothiocyanate (TriReagenz; Sigma). Random hexanucleotide primers were used for reverse-transcription of equal amounts of RNA. The cDNA was used for real-time polymerase chain reaction (PCR) using Taqman probes for the detection of the specific amplification products. The oligonucleotides used for the PCR were derived from the human CYP2C8 sequence (2C8 forward: ggactttatcgattgcttcctg, reverse: ccatatctcagagtggtgcttg; FAM and dabcyl labeled Taqman probe: ttggcactgtagctgatctatttgttgctgga). To control for the amount of cDNA used, the 18S RNA was amplified by real-time PCR with Primer and detected by Taqman probes (Applied Biosystems). The relative amount of the cDNA in the samples was calculated on the basis of a standard row obtained by a serial dilution of a positive control. The relative amounts of CYP2C were normalized to the relative amounts of 18S. At least 2 reverse-transcription reactions were performed with each RNA sample and at least 2 PCR reactions were performed with each cDNA sample.
Statistics
Data are expressed as the mean±SEM. Statistical analyses were performed by 1way ANOVA followed by Bonferroni multiple comparison test. Values of P<0.05 were considered statistically significant.
Results
Effect of CYP2C9 Overexpression and EET Stimulation on COX-2 Expression
CYP2C protein is expressed in native endothelial cells but its expression rapidly decreases after cell isolation. Therefore, to determine the effect of CYP2C9 on COX-2 expression in endothelial cells, HUVECs were infected with either CYP2C9 sense or CYP2C9 antisense adenoviruses. Endothelial cells infected with the CYP2C9 sense adenovirus generated 2-fold more 11,12-EET and 14,15-EET under basal conditions than cells infected with the control (CYP2C9 antisense) virus and was sensitive to the CYP2C9 inhibitor, sulfaphenazole (see http://atvb.ahajournals.org).
CYP2C9 overexpression markedly increased COX-2 protein levels compared with the control cells, ie, infected with the CYP2C9 antisense virus (Figure 1A). The antisense virus did not affect basal COX-2 expression or that induced by IL-1? (data not shown). Sulfaphenazole abolished the CYP2C9-induced expression of COX-2 (Figure 1B). In contrast, AG 1478, which inhibits the intrinsic tyrosine kinase activity of the EGF receptor, did not affect the CYP2C9-induced increase in COX-2 expression (data not shown).
Figure 1. Effect of CYP2C9 and 11,12-EET on COX-2 expression in endothelial cells. A, Human umbilical vein endothelial cells were infected with CYP2C9 sense (2C9) or antisense (CTL) adenoviruses for 24 or 48 hours, lysed, and subjected to Western blotting. B, Effect of sulfaphenazole (30 μmol/L) on CYP2C9-induced COX-2 expression 24 hours after the infection with CYP2C9 sense (2C9) or antisense (CTL) adenoviruses. Similar results were obtained in 3 additional experiments. C, Effect of solvent (Sol, 0.03% DMSO) or 11,12-EET (1 μmol/L) on the expression of COX-2. To demonstrate equal loading of each lane, membranes were reblotted with an antibody recognizing ?-actin. The bar graphs summarize data obtained in 4 independent experiments. *P<0.05, **P<0.01 vs CTL.
To determine whether a CYP2C9 product also elicits COX-2 protein expression, we investigated the effects of 11,12-EET on COX-2 expression. Exogenous application of 11,12-EET significantly increased COX-2 expression in endothelial cells (Figure 1C).
Effect of CYP2C9 on PGI2 Levels in Endothelial Cells
Levels of 6-keto-PGF1, the main stable metabolite of PGI2, were increased in the supernatant of cells overexpressing CYP2C9, an effect inhibited by sulfaphenazole (Figure 2A). Sulfaphenazole did not affect 6-keto PGF1 production in cells infected with the control virus or the basal and bradykinin-stimulated production of 6-keto PGF1 in noninfected cells expressing only COX-1, indicating that the effects of sulfaphenazole cannot be attributed to a nonspecific action as a COX inhibitor. 6-Keto PGF1 levels were 200±26 and 867±69 versus 222±35 and 942±72 pg/mL in solvent and bradykinin-stimulated (100 nmol/L) cells in the absence and presence of sulfaphenazole, respectively (n=3). Thromboxane B2 could not be detected in cells expressing CYP2C9 (data not shown).
Figure 2. Effect of CYP2C9 overexpression on PGI2 production and intracellular cAMP levels. Human umbilical vein endothelial cells infected with CYP2C9 sense (2C9) or antisense (CTL) adenoviruses were incubated in the absence or presence of sulfaphenazole (Sulfa, 30 μmol/L) for 24 hours. A, The concentration of 6-keto PGF1, the stable metabolite of PGI2, was assessed in the cell supernatant, and (B) intracellular cAMP levels were determined in the absence (solvent) and presence of bradykinin (100 nmol/L, 5 minutes). The bar graphs summarize data obtained in 5 to 6 independent experiments; *P<0.05 vs CTL.
Effect of CYP2C9 on cAMP and CREB
The expression of CYP2C9 was associated with a significant increase in endothelial cAMP levels compared with control (antisense virus-treated) cells, both under basal conditions and after stimulation with bradykinin (Figure 2B). Treatment with sulfaphenazole reduced the cAMP concentration of CYP2C9-expressing cells to control levels.
Overexpression of CYP2C9 induced the activation and DNA binding of CREB, an effect not observed in infected cells treated with sulfaphenazole (Figure 3A). Moreover, the CYP2C9 product, 11,12-EET, also elicited the time-dependent activation of CREB (Figure 3B).
Figure 3. Effect of CYP2C9 overexpression and exogenous application of 11,12-EET on the activity of the transcription factor CREB in cultured human endothelial cells. Human umbilical vein endothelial cells were either (A) infected with CYP2C9 sense (2C9) or antisense (CTL) adenoviruses and incubated in the presence or absence of sulfaphenazole (sulfa, 30 μmol/L) for 24 hours, or (B) stimulated with solvent (Sol, 0.03% DMSO) or 11,12-EET (1 μmol/L) for the times indicated. SS refers to the supershift induced using a specific CREB antibody. The bar graphs summarize data obtained in 3 to 8 independent experiments. *P<0.05 vs CTL or Sol.
Involvement of Protein Kinase A in the CYP2C9-Induced Expression of COX-2
Overexpression of CYP2C9 markedly increased COX-2 promoter activity and protein expression (Figure 4). Both these effects were sensitive to treatment of the cells with sulfaphenazole, as well as with the protein kinase A (PKA) inhibitor, KT 5720. Sulfaphenazole and KT 5720 did not affect COX-2 promoter activity or COX-2 levels in control (antisense virus-treated) endothelial cells.
Figure 4. Effect of the PKA-inhibitor KT 5720 on CYP2C9-induced COX-expression in human endothelial cells. Human umbilical vein endothelial cells were infected with CYP2C9 sense (2C9) or antisense (CTL) adenoviruses and cultured for 24 hours in the absence or presence of sulfaphenazole (Sulfa, 30μmol/L). A, Effect of KT 5720 (KT, 500 nmol/L) on COX-2 expression. B, Effect of sulfaphenazole (Sulfa, 30 μmol/L) and KT 5720 (500 nmol/L) on the CYP2C9-induced increase in COX-2 promoter activity. The bar graphs summarize data obtained in 4 independent experiments. *P<0.05 vs CTL.
Effect of the COX-2 Inhibitor Celecoxib on CYP2C9-Induced and 11,12-EET–Induced Angiogenesis
To determine whether the induction of COX-2 expression contributes to the angiogenic effect of CYP2C9, we investigated the effect of the specific COX-2 inhibitor celecoxib on CYP2C9-induced endothelial tube formation in a fibrin gel.
HUVEC were seeded on fibrin gels, infected with CYP2C9 sense or antisense adenoviruses, and cultured in the absence or presence of celecoxib. After 14 days, endothelial cell tubes were clearly present in endothelial cells expressing CYP2C9, whereas little or no tubes were found in gels containing cells treated with the antisense virus or in gels containing CYP2C9-overexpressing cells treated with celecoxib (Figure 5).
Figure 5. Effect of the selective COX-2 inhibitor celecoxib on CYP2C9-induced endothelial cell tube formation. Human endothelial cells were seeded on fibrin gels, infected with CYP2C9 sense (2C9) or antisense (CTL) adenoviruses, and cultured in the absence or presence of celecoxib (Cele, 1μmol/L). After 14 days, tube formation was observed and total tube length was assessed using a computer-assisted microscope. The bar graph summarizes data obtained in 4 independent experiments. **P<0.01 vs CTL. The insert shows an endothelial cell tube expressing CYP2C9 and GFP within the fibrin gel.
Effect of CYP2C Upregulation by Cyclic Strain on Endothelial Cell Tube Formation.
To study a more physiologically relevant response, tube formation was monitored in endothelial cells in which CYP2C expression was enhanced by pre-exposure to cyclic strain for 24 hours. As previously reported,15 cyclic stretch (12%, 1Hz) for 24 hours significantly increased CYP2C levels (Figure 6A). The cyclic strain-induced increase in CYP2C expression was associated with an increase in COX-2 expression and was sensitive to the epoxygenase inhibitor MS-PPOH (Figure 6B).
Figure 6. Effect of cyclic strain on the induction of CYP2C and COX-2 expression and on subsequent tube formation. Human endothelial cells were either maintained under static conditions or exposed to cyclic strain (12%, 1Hz) for 24 hours. A, Reverse-transcription PCR showing the effect of cyclic stretch on CYP2C expression. The bar graph summarizes data obtained in 3 independent experiments. B, Representative Western blots showing the effect of the epoxygenase inhibitor MS-PPOH on the cyclic strain-induced increase in COX-2 expression. Similar results were obtained in 2 additional experiments. C, Effect of preconditioning with cyclic strain on endothelial cell tube formation in the Matrigel assay. Experiments were performed in the absence and presence of solvent (CTL), sulfaphenazole (Sulfa, 10 μmol/L), 14,15-epoxyeicosa-5(Z)-enoic acid (EEZE, 10 μmol/L), celecoxib (Cele, 1 μmol/L), and MS-PPOH (MS, 10 μmol/L). The bar graph summarizes data obtained in 4 independent experiments. *P<0.05, **P<0.01, *** P<0.001 vs CTL.
Given that CYP2C protein levels decrease in cultured cells maintained under static conditions, it was necessary to monitor tube formation over a relatively short period (12 to 16 hours) using the Matrigel tube formation assay. However, 24 hours after the cessation of strain, CYP2C RNA expression remained 8-fold elevated over levels detected in static cultures. We observed a significant increase in the number of branching points in the Matrigel assay when cells were preconditioned by cyclic strain. This effect was abolished in cells pretreated with sulfaphenazole and the EET antagonist EEZE, and markedly inhibited in cells treated with celecoxib or the epoxygenase inhibitor, MS-PPOH (Figure 6C).
Discussion
The results of the present investigation demonstrate that the overexpression of a CYP epoxygenase (CYP2C9), as well as the application of a CYP epoxygenase product (11,12-EET) to endothelial cells, induce the expression of COX-2 in a cAMP/PKA-dependent manner. This effect contributes to the angiogenic effect of CYP2C9, which could be demonstrated by the use of the selective COX-2 inhibitor celecoxib. The concentration of celecoxib used (1 μmol/L) was low enough to selectively inhibit COX-2 without affecting COX-1 activity or exerting other COX-independent effects.16
Although CYP2C is expressed in native endothelial cells, levels of CYP2C protein and RNA decrease rapidly after cell isolation.17 Therefore, when assessing responses in cultured cells, it is necessary to induce CYP expression using pharmacological agents or a physiological stimulus such as cyclic strain, or to transfect the cells with a CYP2C9 expression vector. To achieve high transfection efficiency, we used adenoviral vectors encoding for CYP2C9. However, because the adenoviral infection of cells can itself elicit an inflammatory response and increase COX-2 expression, we compared responses with those obtained in cells treated with a control (CYP2C9 antisense) virus. Adenoviral infection alone elicited a rapid but transient upregulation of COX-2 (protein levels had returned to basal values 16 hours after infection). The effects we have attributed to the overexpression of CYP2C9 were the consequence of enhanced enzyme activity inasmuch as sulfaphenazole, a CYP2C9 inhibitor, was able to prevent the induction of COX-2 expression observed after infection with the CYP2C9 (sense) virus. We also observed that exposure of endothelial cells to cyclic strain increased both CYP2C and COX-2 expression. COX-2 was previously reported to be induced by fluid shear-stress and to be mediated by PKA and CREB.18 Here, we found that an epoxygenase inhibitor prevented the increase in COX-2 expression. The exogenous application of 11,12-EET was also found to induce the expression of COX-2 in endothelial cells. It is difficult to estimate the concentration of exogenously applied EET that actually enters a cell under culture conditions as the cellular EET content reflects generation, degradation, binding, and re-incorporation of these lipid mediators into phospholipids of the membrane. Although an EET receptor has not yet been identified in endothelial cells, the fact that it is possible to antagonize the effects of 11,12-EET with the EET antagonist, 14,15-epoxyeicos-5(Z)-enoic acid, suggests that a receptor exists. If this receptor is intracellular, it is logical to assume that the concentration of intracellularly generated EET required to elicit a specific effect is much lower than that of the compound added exogenously. Thus, although our studies with 11,12-EET show that a CYP2C9 product can elicit the same effect as overexpression of an EET-generating enzyme, they cannot be used to estimate the concentration of EETs required to elicit a biological effect. However, taken together, our findings implicate the involvement of an EET-dependent and cAMP-dependent process in the regulation of the stretch-induced increase in COX-2 expression and may link our observations with previous reports showing that PKA and CREB are involved in the induction of COX-2 in endothelial cells.18
EETs are now recognized as intracellular signal transduction molecules affecting kinase cascades,19 as well as modulating gene expression in a variety of different cell types. Exactly how EETs are able to initiate such signaling cascades is unclear, but several groups have reported a link between EETs and the activation of adenylyl cyclase and the cAMP/PKA pathway.20,21 Our conclusion that the EET-induced activation of PKA is involved in the induction of COX-2 is based on experiments using the kinase inhibitor KT 5720, which may inhibit kinases other than PKA.22 However, CYP2C9 overexpression was associated with an increase in cAMP levels, and we have previously reported that other PKA inhibitors are also able to interfere with EET signaling.14 Moreover, the CYP2C9-induced increase in COX-2 expression reported here also depends on an increase in the DNA binding of CREB to the COX-2 promoter. CYP 2J2, an EET-generating epoxygenase that is also expressed in endothelial cells, has been linked with the induction of tissue plasminogen activator in endothelial cells via a cAMP-dependent mechanism.23 In the latter study, the EET-induced increase in cAMP was attributed to the activation of the guanine nucleotide-binding protein Gs, suggesting that this mechanism may also underlie the induction of COX-2 described in this report. Some of the consequences of EET production can however be attributed to the transactivation of the EGF receptor after the release of heparin-binding EGF from the endothelial cell surface.6,10 Although COX-2 expression is reported to be modulated by EGF or activation of the EGF receptor in some cells,11 such a mechanism cannot account for the induction of COX-2 in CYP2C9-expressing endothelial cells inasmuch as treatment with an inhibitor of the EGF receptor tyrosine kinase did not influence COX-2 expression.
CYP epoxygenase-derived EETs inhibit the activation of nuclear factor B3 and attenuate the PDGF-induced migration of smooth muscle cells.24 Such observations, together with their well-characterized vasodilator and fibrinolytic properties,23 have led to the classification of EETs as anti-inflammatory compounds. At first sight, the results of the present investigation, showing that CYP2C9-derived EETs induce COX-2 expression, tend to contradict this classification because COX-2 has long been considered as the "bad" COX isoform that is generally induced as a consequence of an inflammatory reaction. On the basis of findings obtained in other cell systems, it has been generally assumed that COX-2 plays a detrimental role in cardiovascular homeostasis. COX-2 expression is however increased in endothelial cells exposed to fluid shear/stress,25 a well-documented anti-atherosclerotic stimulus. Moreover, COX-2 is the major source of the potent anti-aggregatory vasodilator PG, PGI2,9 and exerts anti-apoptotic actions in other cell types.26 Recent findings have also associated enhanced COX-2 activity with the protection of cardiomyocytes against oxidative stress and ischemia/reperfusion injury.27,28 The inhibition of PGI2 production by COX-2 and the loss of its vasoprotective effects may account for the results of a recent meta-analysis of all randomized clinical trials of COX-2 inhibitors that concluded that the inhibition of COX-2 potentially increases the risk of cardiovascular events.29 Although it is clear that COX-2–derived thromboxanes would be expected to induce vasoconstriction and potentiate an inflammatory state, the enzyme generates mainly PGI2 when it is expressed in endothelial cells.9
Increased expression of COX-2 has been linked to angiogenesis,30 and in co-cultures of COX-2–overexpressing colon cancer cells with endothelial cells, enhanced enzyme activity results in endothelial cell migration and angiogenesis.31 EETs have been shown to induce angiogenesis in vitro6 and in vivo,5 and we have previously reported that CYP2C9 induces angiogenesis via the activation of the EGF receptor, which is probably mediated via matrix metalloprotease activation followed by the release of EGF/HB-EGF.6 The results of the present investigation indicate that the induction of COX-2 expression also contributes to the angiogenic effect of CYP2C9, although the 2 pathways do not appear to be directly linked.
Acknowledgments
The authors are indebted to Isabel Winter for expert technical assistance. Research described in this article was supported by Philip Morris Inc, the Deutsche Forschungsgemeinschaft (FI 830/2-1), the National Institutes of Health (NIH GM31278, to J.R.F.), and by a Young Investigators Grant (to U.R.M.) from the Medical Faculty of the Johann Wolfgang Goethe Universit?t.
Received June 2, 2004; accepted November 12, 2004.
References
Campbell WB, Gebremedhin D, Pratt PF, Harder DR. Identification of epoxyeicosatrienoic acids as endothelium-derived hyperpolarizing factors. Circ Res. 1996; 78: 415–423.
Fleming I. Cytochrome P450 and vascular homeostasis. Circ Res. 2001; 89: 753–762.
Node K, Huo Y, Ruan X, Yang B, Spiecker M, Ley K, Zeldin DC, Liao JK. Anti-inflammatory properties of cytochrome P450 epoxygenase-derived eicosanoids. Science. 1999; 285: 1276–1279.
Li PL, Chen CL, Bortell R, Campbell WB. 11,12-Epoxyeicosatrienoic acid stimulates endogenous mono-ADP- ribosylation in bovine coronary arterial smooth muscle. Circ Res. 1999; 85: 349–356.
Zhang C, Harder DR. Cerebral capillary endothelial cell mitogenesis and morphogenesis induced by astrocytic epoxyeicosatrienoic acid. Stroke. 2002; 33: 2957–2964.
Michaelis UR, Fisslthaler B, Medhora M, Harder D, Fleming I, Busse R. Cytochrome P450 2C9-derived epoxyeicosatrienoic acids induce angiogenesis via cross-talk with the epidermal growth factor receptor (EGFR). FASEB J. 2003; 17: 770–772.
Potente M, Michaelis UR, Fisslthaler B, Busse R, Fleming I. Cytochrome P450 2C9-induced endothelial cell proliferation involves induction of mitogen-activated protein (MAP) kinase phosphatase-1, inhibition of the c-Jun N-terminal kinase, and up-regulation of cyclin D1. J Biol Chem. 2002; 277: 15671–15676.
Davidge ST. Prostaglandin H synthase and vascular function. Circ Res. 2001; 89: 650–660.
McAdam BF, Catella-Lawson F, Mardini IA, Kapoor S, Lawson JA, Fitzgerald GA. 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. 1999; 96: 272–277.
Chen J-K, Capdevila J, Harris RC. Heparin-binding EGF-like growth factor mediates the biological effects of P450 arachidonate epoxygenase metabolites in epithelial cells. Proc Natl Acad Sci U S A. 2002; 99: 6029–6034.
Vadlamudi R, Mandal M, Adam L, Steinbach G, Mendelsohn J, Kumar R. Regulation of cyclooxygenase-2 pathway by HER2 receptor. Oncogene. 1999; 18: 305–314.
Gauthier KM, Deeter C, Krishna UM, Reddy YK, Bondlela M, Falck JR, Campbell WB. 14,15-Epoxyeicosa-5(Z)-enoic acid: a selective epoxyeicosatrienoic acid antagonist that inhibits endothelium-dependent hyperpolarization and relaxation in coronary arteries. Circ Res. 2002; 90: 1028–1036.
Wang MH, Brand-Schieber E, Zand BA, Nguyen X, Falck JR, Balu N, Schwartzman ML. Cytochrome P450-derived arachidonic acid metabolism in the rat kidney: characterization of selective inhibitors. J Pharmacol Exp Ther. 1998; 284: 966–973.
Popp R, Brandes RP, Ott G, Busse R, Fleming I. Dynamic modulation of interendothelial gap junctional communication by 11,12-epoxyeicosatrienoic acid. Circ Res. 2002; 90: 800–806.
Fisslthaler B, Popp R, Michaelis UR, Kiss L, Fleming I, Busse R. Cyclic stretch enhances the expression and activity of coronary endothelium-derived hyperpolarizing factor synthase. Hypertension. 2001; 38: 1427–1432.
Gierse JK, Koboldt CM, Walker MC, Seibert K, Isakson PC. Kinetic basis for selective inhibition of cyclo-oxygenases. Biochem J. 1999; 339 (Pt 3): 607–614.
Vernia S, Sanz-Gonzalez SM, Lopez-Garcia MP. Involvement of peroxynitrite on the early loss of p450 in short-term hepatocyte cultures. Adv Exp Med Biol. 2001; 500: 209–212.
Ogasawara A, Arakawa T, Kaneda T, Takuma T, Sato T, Kaneko H, Kumegawa M, Hakeda Y. Fluid shear stress-induced cyclooxygenase-2 expression is mediated by C/EBP ?, cAMP-response element-binding protein, and AP-1 in osteoblastic MC3T3–E1 Cells. J Biol Chem. 2001; 276: 7048–7054.
Fleming I, Fisslthaler B, Michaelis UR, Kiss L, Popp R, Busse R. The coronary endothelium-derived hyperpolarizing factor (EDHF) stimulates multiple signalling pathways and proliferation in vascular cells. Pflugers Arch. 2001; 442: 511–518.
Wong PY, Lai PS, Shen SY, Belosludtsev YY, Falck JR. Post-receptor signal transduction and regulation of 14(R),15(S)-epoxyeicosatrienoic acid (14,15-EET) binding in U-937 cells. J Lipid Mediat Cell Signal. 1997; 16: 155–169.
Imig JD, Inscho EW, Deichamnn PC, Reddy KM, Falck JR. Afferent arteriolar vasodilation to the sulfonimide analog of 11,12-epoxyeicosatrienoic acid involves protein kinase A. Hypertension. 1999; 33: 408–413.
Olsen MK, Reszka AA, Abraham I. KT5720 and U-98017 inhibit MAPK and alter the cytoskeleton and cell morphology. J Cell Physiol. 1998; 176: 525–536.
Node K, Ruan XL, Dai J, Yang SX, Graham L, Zeldin DC, Liao JK. Activation of Gs mediates induction of tissue-type plasminogen activator gene transcription by epoxyeicosatrienoic acids. J Biol Chem. 2001; 276: 15983–15989.
Sun J, Sui XX, Bradbury A, Zeldin DC, Conte MS, Liao JK. Inhibition of vascular smooth muscle cell migration by cytochrome P450 epoxygenase-derived eicosanoids. Circ Res. 2002; 90: 1020–1027.
Topper JN, Cai J, Falb D, Gimbrone MA. 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. 1996; 93: 10417–10422.
Hsu AL, Ching TT, Wang DS, Song X, Rangnekar VM, Chen CS. The cyclooxygenase-2 inhibitor celecoxib induces apoptosis by blocking Akt activation in human prostate cancer cells independently of Bcl-2. J Biol Chem. 2000; 275: 11397–11403.
Adderley SR, Fitzgerald DJ. Oxidative damage of cardiomyocytes is limited by extracellular regulated kinases 1/2-mediated induction of cyclooxygenase-2. J Biol Chem. 1999; 274: 5038–5046.
Bolli R, Shinmura K, Tang XL, Kodani E, Xuan YT, Guo Y, Dawn B. Discovery of a new function of cyclooxygenase (COX)-2: COX-2 is a cardioprotective protein that alleviates ischemia/reperfusion injury and mediates the late phase of preconditioning. Cardiovasc Res. 2002; 55: 506–519.
Mukherjee D, Nissen SE, Topol EJ. Risk of cardiovascular events associated with selective COX-2 inhibitors. JAMA. 2001; 286: 954–959.
Iniguez MA, Rodriguez A, Volpert OV, Fresno M, Redondo JM. Cyclooxygenase-2: a therapeutic target in angiogenesis. Trends Mol Med. 2003; 9: 73–78.
Tsujii M, Kawano S, Tsuji S, Sawaoka H, Hori M, Dubois RN. Cyclooxygenase regulates angiogenesis induced by colon cancer cells. Cell. 1998; 93: 705–716.(U. Ruth Michaelis; John R)
Correspondence to Ingrid Fleming, PhD, Vascular Signalling Group, Institut für Kardiovaskul?re Physiologie, Johann Wolfgang Goethe-Universit?t, Theodor-Stern-Kai 7, D-60590 Frankfurt am Main, Germany. E-mail fleming@em.uni-frankfurt.de
Abstract
Objective— Cytochrome P450 (CYP) epoxygenases metabolize arachidonic acid to epoxyeicosatrienoic acids (EETs). CYP2C9-derived EETs elicit endothelial cell proliferation and angiogenesis, but the signaling pathways involved are incompletely understood. Because cyclooxygenase-2 (COX-2) is involved in angiogenesis, we determined whether a link exists between CYP2C9 and COX-2 expression.
Methods and Results— Human umbilical vein endothelial cells were infected with CYP2C9 sense or antisense adenoviral constructs. Overexpression of CYP2C9 increased COX-2 promoter activity, an effect accompanied by a significant increase in COX-2 protein expression and elevated prostacyclin production. The CYP2C9-induced expression of COX-2 was inhibited by the CYP2C9 inhibitor, sulfaphenazole, whereas 11,12-EET increased COX-2 expression. Overexpression of CYP2C9 and stimulation with 11,12-EET increased intracellular cAMP levels and stimulated DNA-binding of the cAMP-response element-binding protein. The protein kinase A inhibitor, KT5720, attenuated the CYP2C9-induced increase in COX-2 promoter activity and protein expression. Overexpression of CYP2C9 stimulated endothelial tube formation, an effect that was attenuated by the COX-2 inhibitor celecoxib. Identical responses were observed in cells preconditioned by cyclic strain to increase CYP2C expression.
Conclusion— These data indicate that CYP2C9-derived EETs induce the expression of COX-2 in endothelial cells via a cAMP-dependent pathway and that this mechanism contributes to CYP2C9-induced angiogenesis.
Overexpression of cytochrome P450 (CYP) 2C9 in endothelial cells increased cAMP levels, stimulated the cAMP-response element-binding protein, and enhanced cyclooxygenase-2 (COX-2) promoter activity, protein expression, and prostacyclin production. CYP2C9 overexpression stimulated endothelial tube formation, which was attenuated by the COX-2 inhibitor celecoxib. Thus, COX-2 contributes to CYP2C9-induced angiogenesis.
Key Words: angiogenesis ? endothelium ? gene expression ? prostacyclin ? cytochrome P450
Introduction
Cytochrome P450 epoxygenases of the 2B, 2C, and 2J subfamilies are expressed in endothelial cells and metabolize arachidonic acid to epoxyeicosatrienoic acids (EETs) (5,6-EET, 8,9-EET, 11,12-EET, and 14,15-EET). Because EETs induce the hyperpolarization and relaxation of vascular smooth muscle cells, these eicosanoids were initially described as endothelium-derived hyperpolarizing factors.1 However, EETs are now appreciated to be more than simple vasodilators and exert numerous membrane potential-independent effects on endothelial cell signaling and vascular homeostasis.2
Whereas EETs can be released from endothelial cells, they also act as intracellular signal transduction molecules and are capable of activating tyrosine kinases and phosphatases, MAP kinases (extracellular signal regulated kinase 1/2, p38 MAP kinase), the phosphatidylinositol-3 kinase, and protein kinase B/Akt.2 Furthermore, EETs have been reported to stimulate the ADP-ribosylation of G-proteins and to prevent the activation of the transcription factor nuclear factor B.3,4 More recently, CYP2C9-derived EETs were found to induce endothelial cell proliferation and angiogenesis5,6 via a mechanism involving the induction of MAP kinase phosphatase-1 and cyclin D1, as well as the transactivation of the epidermal growth factor (EGF) receptor.6,7
Cyclooxygenases (COX) also metabolize arachidonic acid, leading to the production of prostaglandins and thromboxanes. There are 2 isoforms of this enzyme. The first (COX-1) is constitutively expressed in most tissues, whereas the expression of COX-2 can be induced by growth factors, cytokines, and vasoactive peptides such as endothelin.8 COX-derived products have been shown to be key mediators in inflammatory responses, and they also regulate vascular tone, as well as a number of signal transduction pathways affecting cell adhesion, growth, and differentiation. Although COX-2 expression is increased in inflammatory conditions, its products are not always pro-inflammatory. In the vasculature, COX-2 generates mainly the vasodilator and antithrombotic prostanoid, prostacyclin (PGI2),9 and should therefore be considered as vasoprotective.
Because CYP-derived EETs transactivate the EGF receptor,10 which has been linked to an increase in COX-2 expression,11 the aim of the present investigation was to determine whether an increase in CYP2C9 expression and activity could elicit an increase in COX-2 expression and whether COX-2 is involved in CYP2C9-induced angiogenesis.
Methods
Materials
11,12-EET was purchased from Cayman Chemicals (Massy, France), KT 5720 was from Calbiochem (La Jolla, Calif), and thrombin was from Hemochrom Diagnostica GmbH (Essen, Germany). Celecoxib was a generous gift from Dr Gerd Geisslinger (Frankfurt, Germany). Epoxyeicosa-5(Z)-enoic acid (14,15-EEZE) and N-methylsulfonyl-6-(2-propargyloxyphenyl) hexanamide (MS-PPOH) were synthesized as described.12,13 Sulfaphenazole and all other chemicals were from Sigma.
Cell Culture
Human umbilical vein endothelial cells (HUVECs) were purchased from Cell Systems/Clonetics (Solingen) and cultured as described.6 First- or second-passage endothelial cells were used in all experiments.
Adenoviral Infection
HUVECs (80% to 90% confluent) were serum-starved for 24 hours before infection with adenoviral vectors. Cells were incubated with recombinant adenoviruses expressing CYP2C9 sense or antisense cDNA (10 pfU/cell) in medium without antibiotics for 4 hours at 37°C, followed by recovery in the presence of 2% fetal calf serum. The efficiency of infection was between 90% and 100%.
Immunoblotting
For Western blot analysis, cells were lysed in Triton X-100 lysis buffer and separated by SDS-PAGE as described.6 Proteins were detected using antibodies against COX-2 (Santa Cruz Biotechnology, Santa Cruz, Calif) and ?-actin (Sigma). The CYP2C9 antibody used was purified by Eurogenetec (Seraing, Belgium) from rabbits immunized with a CYP2C9 peptide (RRRKLPPGPTPLPIC).
Reporter Gene Assay
Twelve hours after infection with CYP2C9 sense or antisense adenoviral constructs, endothelial cells were transfected with a COX-2 promoter construct (kindly provided by Dr Margarete Goppelt-Struebe, Erlangen, Germany), together with a LacZ construct (Invitrogen, Karlsruhe, Germany). After 24 hours, the cells were lysed, and luciferase activity was assayed according to the manufacturer’s protocol (Promega, Mannheim, Germany). Values were corrected for transfection efficiency by measuring ?-galactosidase activity.
Radioimmunoassays
Commercially available kits (both from Amersham Biosciences, Buckinghamshire, UK) were used to determine 6-keto-prostaglandin (PG) F1 (6-Keto-PGF1) levels in cell supernatants and intracellular cAMP levels in cells treated with 3-isobutyl-1-methylxanthine (30 μmol/L, 30 minutes) as described.14
Electrophoretic Mobility Shift Assay
Double-stranded oligonucleotides containing the sequence of the binding site for cAMP-response element-binding protein (CREB: 5'-AGAGATTCGCTGACGTCAGAGAGCTAG-3') were labeled with 32P-ATP using a Ready-to-go DNA-Labeling Kit (Amersham Biosciences, Piscataway, NJ). Nuclear proteins (6 to 10 μg) were incubated with the 32P-labeled oligonucleotides (10000 counts/sample) for 30 minutes in electrophoretic mobility shift assay-binding buffer (10 mmol/L HEPES, pH 7.5, 100 mmol/L NaCl, 1 mmol/L EDTA, 5% [v/v] glycerol, 1 mmol/L DTT, 1 μL/sample poly-dI/dC). Supershift assays were performed by adding a CREB antibody 30 minutes before incubation with the oligonucleotide. Protein–DNA complexes were resolved by gel electrophoresis, and the gels were fixed and dried before being exposed to x-ray films. Densitometric analysis of the autoradiographs was performed after nonsaturating exposures, and the values were expressed as the percentage of the binding activity detected in extracts from unstimulated cells.
In Vitro Angiogenesis Assay
Fibrin gels were prepared using thrombin (0.5 U/μL)-polymerized fibrinogen (1.5 mg/mL in MCDB 131). HUVECs were infected with CYP2C9 sense or antisense adenoviruses 24 hours after seeding onto the fibrin gels and cultured in MCDB 131, supplemented with 4% fetal calf serum, L-gutamine (10 mmol/L), basic fibroblast growth factor (0.5 ng/mL), epidermal growth factor (0.05 ng/mL), and endothelial cell growth-stimulating factor from bovine brain (ECGS/H, 0.2%), and in the absence or presence of celecoxib (1 μmol/L). After 14 days, photographs were taken and tube formation was assessed and total tube length determined.
Cyclic Strain
To induce CYP2C expression in cultured endothelial cells human umbilical vein endothelial cells were exposed to cyclic strain (12%, 1Hz) for 24 hours as described previously.15 After stimulation, the endothelial cells were recovered and seeded onto Matrigel; tube formation was assessed after 12 hours and quantified by counting the number of branch points.
Isolation of RNA and Reverse-Transcription Polymerase Chain Reaction
Total RNA was isolated from cultured HUVECs using phenol and guanidine isothiocyanate (TriReagenz; Sigma). Random hexanucleotide primers were used for reverse-transcription of equal amounts of RNA. The cDNA was used for real-time polymerase chain reaction (PCR) using Taqman probes for the detection of the specific amplification products. The oligonucleotides used for the PCR were derived from the human CYP2C8 sequence (2C8 forward: ggactttatcgattgcttcctg, reverse: ccatatctcagagtggtgcttg; FAM and dabcyl labeled Taqman probe: ttggcactgtagctgatctatttgttgctgga). To control for the amount of cDNA used, the 18S RNA was amplified by real-time PCR with Primer and detected by Taqman probes (Applied Biosystems). The relative amount of the cDNA in the samples was calculated on the basis of a standard row obtained by a serial dilution of a positive control. The relative amounts of CYP2C were normalized to the relative amounts of 18S. At least 2 reverse-transcription reactions were performed with each RNA sample and at least 2 PCR reactions were performed with each cDNA sample.
Statistics
Data are expressed as the mean±SEM. Statistical analyses were performed by 1way ANOVA followed by Bonferroni multiple comparison test. Values of P<0.05 were considered statistically significant.
Results
Effect of CYP2C9 Overexpression and EET Stimulation on COX-2 Expression
CYP2C protein is expressed in native endothelial cells but its expression rapidly decreases after cell isolation. Therefore, to determine the effect of CYP2C9 on COX-2 expression in endothelial cells, HUVECs were infected with either CYP2C9 sense or CYP2C9 antisense adenoviruses. Endothelial cells infected with the CYP2C9 sense adenovirus generated 2-fold more 11,12-EET and 14,15-EET under basal conditions than cells infected with the control (CYP2C9 antisense) virus and was sensitive to the CYP2C9 inhibitor, sulfaphenazole (see http://atvb.ahajournals.org).
CYP2C9 overexpression markedly increased COX-2 protein levels compared with the control cells, ie, infected with the CYP2C9 antisense virus (Figure 1A). The antisense virus did not affect basal COX-2 expression or that induced by IL-1? (data not shown). Sulfaphenazole abolished the CYP2C9-induced expression of COX-2 (Figure 1B). In contrast, AG 1478, which inhibits the intrinsic tyrosine kinase activity of the EGF receptor, did not affect the CYP2C9-induced increase in COX-2 expression (data not shown).
Figure 1. Effect of CYP2C9 and 11,12-EET on COX-2 expression in endothelial cells. A, Human umbilical vein endothelial cells were infected with CYP2C9 sense (2C9) or antisense (CTL) adenoviruses for 24 or 48 hours, lysed, and subjected to Western blotting. B, Effect of sulfaphenazole (30 μmol/L) on CYP2C9-induced COX-2 expression 24 hours after the infection with CYP2C9 sense (2C9) or antisense (CTL) adenoviruses. Similar results were obtained in 3 additional experiments. C, Effect of solvent (Sol, 0.03% DMSO) or 11,12-EET (1 μmol/L) on the expression of COX-2. To demonstrate equal loading of each lane, membranes were reblotted with an antibody recognizing ?-actin. The bar graphs summarize data obtained in 4 independent experiments. *P<0.05, **P<0.01 vs CTL.
To determine whether a CYP2C9 product also elicits COX-2 protein expression, we investigated the effects of 11,12-EET on COX-2 expression. Exogenous application of 11,12-EET significantly increased COX-2 expression in endothelial cells (Figure 1C).
Effect of CYP2C9 on PGI2 Levels in Endothelial Cells
Levels of 6-keto-PGF1, the main stable metabolite of PGI2, were increased in the supernatant of cells overexpressing CYP2C9, an effect inhibited by sulfaphenazole (Figure 2A). Sulfaphenazole did not affect 6-keto PGF1 production in cells infected with the control virus or the basal and bradykinin-stimulated production of 6-keto PGF1 in noninfected cells expressing only COX-1, indicating that the effects of sulfaphenazole cannot be attributed to a nonspecific action as a COX inhibitor. 6-Keto PGF1 levels were 200±26 and 867±69 versus 222±35 and 942±72 pg/mL in solvent and bradykinin-stimulated (100 nmol/L) cells in the absence and presence of sulfaphenazole, respectively (n=3). Thromboxane B2 could not be detected in cells expressing CYP2C9 (data not shown).
Figure 2. Effect of CYP2C9 overexpression on PGI2 production and intracellular cAMP levels. Human umbilical vein endothelial cells infected with CYP2C9 sense (2C9) or antisense (CTL) adenoviruses were incubated in the absence or presence of sulfaphenazole (Sulfa, 30 μmol/L) for 24 hours. A, The concentration of 6-keto PGF1, the stable metabolite of PGI2, was assessed in the cell supernatant, and (B) intracellular cAMP levels were determined in the absence (solvent) and presence of bradykinin (100 nmol/L, 5 minutes). The bar graphs summarize data obtained in 5 to 6 independent experiments; *P<0.05 vs CTL.
Effect of CYP2C9 on cAMP and CREB
The expression of CYP2C9 was associated with a significant increase in endothelial cAMP levels compared with control (antisense virus-treated) cells, both under basal conditions and after stimulation with bradykinin (Figure 2B). Treatment with sulfaphenazole reduced the cAMP concentration of CYP2C9-expressing cells to control levels.
Overexpression of CYP2C9 induced the activation and DNA binding of CREB, an effect not observed in infected cells treated with sulfaphenazole (Figure 3A). Moreover, the CYP2C9 product, 11,12-EET, also elicited the time-dependent activation of CREB (Figure 3B).
Figure 3. Effect of CYP2C9 overexpression and exogenous application of 11,12-EET on the activity of the transcription factor CREB in cultured human endothelial cells. Human umbilical vein endothelial cells were either (A) infected with CYP2C9 sense (2C9) or antisense (CTL) adenoviruses and incubated in the presence or absence of sulfaphenazole (sulfa, 30 μmol/L) for 24 hours, or (B) stimulated with solvent (Sol, 0.03% DMSO) or 11,12-EET (1 μmol/L) for the times indicated. SS refers to the supershift induced using a specific CREB antibody. The bar graphs summarize data obtained in 3 to 8 independent experiments. *P<0.05 vs CTL or Sol.
Involvement of Protein Kinase A in the CYP2C9-Induced Expression of COX-2
Overexpression of CYP2C9 markedly increased COX-2 promoter activity and protein expression (Figure 4). Both these effects were sensitive to treatment of the cells with sulfaphenazole, as well as with the protein kinase A (PKA) inhibitor, KT 5720. Sulfaphenazole and KT 5720 did not affect COX-2 promoter activity or COX-2 levels in control (antisense virus-treated) endothelial cells.
Figure 4. Effect of the PKA-inhibitor KT 5720 on CYP2C9-induced COX-expression in human endothelial cells. Human umbilical vein endothelial cells were infected with CYP2C9 sense (2C9) or antisense (CTL) adenoviruses and cultured for 24 hours in the absence or presence of sulfaphenazole (Sulfa, 30μmol/L). A, Effect of KT 5720 (KT, 500 nmol/L) on COX-2 expression. B, Effect of sulfaphenazole (Sulfa, 30 μmol/L) and KT 5720 (500 nmol/L) on the CYP2C9-induced increase in COX-2 promoter activity. The bar graphs summarize data obtained in 4 independent experiments. *P<0.05 vs CTL.
Effect of the COX-2 Inhibitor Celecoxib on CYP2C9-Induced and 11,12-EET–Induced Angiogenesis
To determine whether the induction of COX-2 expression contributes to the angiogenic effect of CYP2C9, we investigated the effect of the specific COX-2 inhibitor celecoxib on CYP2C9-induced endothelial tube formation in a fibrin gel.
HUVEC were seeded on fibrin gels, infected with CYP2C9 sense or antisense adenoviruses, and cultured in the absence or presence of celecoxib. After 14 days, endothelial cell tubes were clearly present in endothelial cells expressing CYP2C9, whereas little or no tubes were found in gels containing cells treated with the antisense virus or in gels containing CYP2C9-overexpressing cells treated with celecoxib (Figure 5).
Figure 5. Effect of the selective COX-2 inhibitor celecoxib on CYP2C9-induced endothelial cell tube formation. Human endothelial cells were seeded on fibrin gels, infected with CYP2C9 sense (2C9) or antisense (CTL) adenoviruses, and cultured in the absence or presence of celecoxib (Cele, 1μmol/L). After 14 days, tube formation was observed and total tube length was assessed using a computer-assisted microscope. The bar graph summarizes data obtained in 4 independent experiments. **P<0.01 vs CTL. The insert shows an endothelial cell tube expressing CYP2C9 and GFP within the fibrin gel.
Effect of CYP2C Upregulation by Cyclic Strain on Endothelial Cell Tube Formation.
To study a more physiologically relevant response, tube formation was monitored in endothelial cells in which CYP2C expression was enhanced by pre-exposure to cyclic strain for 24 hours. As previously reported,15 cyclic stretch (12%, 1Hz) for 24 hours significantly increased CYP2C levels (Figure 6A). The cyclic strain-induced increase in CYP2C expression was associated with an increase in COX-2 expression and was sensitive to the epoxygenase inhibitor MS-PPOH (Figure 6B).
Figure 6. Effect of cyclic strain on the induction of CYP2C and COX-2 expression and on subsequent tube formation. Human endothelial cells were either maintained under static conditions or exposed to cyclic strain (12%, 1Hz) for 24 hours. A, Reverse-transcription PCR showing the effect of cyclic stretch on CYP2C expression. The bar graph summarizes data obtained in 3 independent experiments. B, Representative Western blots showing the effect of the epoxygenase inhibitor MS-PPOH on the cyclic strain-induced increase in COX-2 expression. Similar results were obtained in 2 additional experiments. C, Effect of preconditioning with cyclic strain on endothelial cell tube formation in the Matrigel assay. Experiments were performed in the absence and presence of solvent (CTL), sulfaphenazole (Sulfa, 10 μmol/L), 14,15-epoxyeicosa-5(Z)-enoic acid (EEZE, 10 μmol/L), celecoxib (Cele, 1 μmol/L), and MS-PPOH (MS, 10 μmol/L). The bar graph summarizes data obtained in 4 independent experiments. *P<0.05, **P<0.01, *** P<0.001 vs CTL.
Given that CYP2C protein levels decrease in cultured cells maintained under static conditions, it was necessary to monitor tube formation over a relatively short period (12 to 16 hours) using the Matrigel tube formation assay. However, 24 hours after the cessation of strain, CYP2C RNA expression remained 8-fold elevated over levels detected in static cultures. We observed a significant increase in the number of branching points in the Matrigel assay when cells were preconditioned by cyclic strain. This effect was abolished in cells pretreated with sulfaphenazole and the EET antagonist EEZE, and markedly inhibited in cells treated with celecoxib or the epoxygenase inhibitor, MS-PPOH (Figure 6C).
Discussion
The results of the present investigation demonstrate that the overexpression of a CYP epoxygenase (CYP2C9), as well as the application of a CYP epoxygenase product (11,12-EET) to endothelial cells, induce the expression of COX-2 in a cAMP/PKA-dependent manner. This effect contributes to the angiogenic effect of CYP2C9, which could be demonstrated by the use of the selective COX-2 inhibitor celecoxib. The concentration of celecoxib used (1 μmol/L) was low enough to selectively inhibit COX-2 without affecting COX-1 activity or exerting other COX-independent effects.16
Although CYP2C is expressed in native endothelial cells, levels of CYP2C protein and RNA decrease rapidly after cell isolation.17 Therefore, when assessing responses in cultured cells, it is necessary to induce CYP expression using pharmacological agents or a physiological stimulus such as cyclic strain, or to transfect the cells with a CYP2C9 expression vector. To achieve high transfection efficiency, we used adenoviral vectors encoding for CYP2C9. However, because the adenoviral infection of cells can itself elicit an inflammatory response and increase COX-2 expression, we compared responses with those obtained in cells treated with a control (CYP2C9 antisense) virus. Adenoviral infection alone elicited a rapid but transient upregulation of COX-2 (protein levels had returned to basal values 16 hours after infection). The effects we have attributed to the overexpression of CYP2C9 were the consequence of enhanced enzyme activity inasmuch as sulfaphenazole, a CYP2C9 inhibitor, was able to prevent the induction of COX-2 expression observed after infection with the CYP2C9 (sense) virus. We also observed that exposure of endothelial cells to cyclic strain increased both CYP2C and COX-2 expression. COX-2 was previously reported to be induced by fluid shear-stress and to be mediated by PKA and CREB.18 Here, we found that an epoxygenase inhibitor prevented the increase in COX-2 expression. The exogenous application of 11,12-EET was also found to induce the expression of COX-2 in endothelial cells. It is difficult to estimate the concentration of exogenously applied EET that actually enters a cell under culture conditions as the cellular EET content reflects generation, degradation, binding, and re-incorporation of these lipid mediators into phospholipids of the membrane. Although an EET receptor has not yet been identified in endothelial cells, the fact that it is possible to antagonize the effects of 11,12-EET with the EET antagonist, 14,15-epoxyeicos-5(Z)-enoic acid, suggests that a receptor exists. If this receptor is intracellular, it is logical to assume that the concentration of intracellularly generated EET required to elicit a specific effect is much lower than that of the compound added exogenously. Thus, although our studies with 11,12-EET show that a CYP2C9 product can elicit the same effect as overexpression of an EET-generating enzyme, they cannot be used to estimate the concentration of EETs required to elicit a biological effect. However, taken together, our findings implicate the involvement of an EET-dependent and cAMP-dependent process in the regulation of the stretch-induced increase in COX-2 expression and may link our observations with previous reports showing that PKA and CREB are involved in the induction of COX-2 in endothelial cells.18
EETs are now recognized as intracellular signal transduction molecules affecting kinase cascades,19 as well as modulating gene expression in a variety of different cell types. Exactly how EETs are able to initiate such signaling cascades is unclear, but several groups have reported a link between EETs and the activation of adenylyl cyclase and the cAMP/PKA pathway.20,21 Our conclusion that the EET-induced activation of PKA is involved in the induction of COX-2 is based on experiments using the kinase inhibitor KT 5720, which may inhibit kinases other than PKA.22 However, CYP2C9 overexpression was associated with an increase in cAMP levels, and we have previously reported that other PKA inhibitors are also able to interfere with EET signaling.14 Moreover, the CYP2C9-induced increase in COX-2 expression reported here also depends on an increase in the DNA binding of CREB to the COX-2 promoter. CYP 2J2, an EET-generating epoxygenase that is also expressed in endothelial cells, has been linked with the induction of tissue plasminogen activator in endothelial cells via a cAMP-dependent mechanism.23 In the latter study, the EET-induced increase in cAMP was attributed to the activation of the guanine nucleotide-binding protein Gs, suggesting that this mechanism may also underlie the induction of COX-2 described in this report. Some of the consequences of EET production can however be attributed to the transactivation of the EGF receptor after the release of heparin-binding EGF from the endothelial cell surface.6,10 Although COX-2 expression is reported to be modulated by EGF or activation of the EGF receptor in some cells,11 such a mechanism cannot account for the induction of COX-2 in CYP2C9-expressing endothelial cells inasmuch as treatment with an inhibitor of the EGF receptor tyrosine kinase did not influence COX-2 expression.
CYP epoxygenase-derived EETs inhibit the activation of nuclear factor B3 and attenuate the PDGF-induced migration of smooth muscle cells.24 Such observations, together with their well-characterized vasodilator and fibrinolytic properties,23 have led to the classification of EETs as anti-inflammatory compounds. At first sight, the results of the present investigation, showing that CYP2C9-derived EETs induce COX-2 expression, tend to contradict this classification because COX-2 has long been considered as the "bad" COX isoform that is generally induced as a consequence of an inflammatory reaction. On the basis of findings obtained in other cell systems, it has been generally assumed that COX-2 plays a detrimental role in cardiovascular homeostasis. COX-2 expression is however increased in endothelial cells exposed to fluid shear/stress,25 a well-documented anti-atherosclerotic stimulus. Moreover, COX-2 is the major source of the potent anti-aggregatory vasodilator PG, PGI2,9 and exerts anti-apoptotic actions in other cell types.26 Recent findings have also associated enhanced COX-2 activity with the protection of cardiomyocytes against oxidative stress and ischemia/reperfusion injury.27,28 The inhibition of PGI2 production by COX-2 and the loss of its vasoprotective effects may account for the results of a recent meta-analysis of all randomized clinical trials of COX-2 inhibitors that concluded that the inhibition of COX-2 potentially increases the risk of cardiovascular events.29 Although it is clear that COX-2–derived thromboxanes would be expected to induce vasoconstriction and potentiate an inflammatory state, the enzyme generates mainly PGI2 when it is expressed in endothelial cells.9
Increased expression of COX-2 has been linked to angiogenesis,30 and in co-cultures of COX-2–overexpressing colon cancer cells with endothelial cells, enhanced enzyme activity results in endothelial cell migration and angiogenesis.31 EETs have been shown to induce angiogenesis in vitro6 and in vivo,5 and we have previously reported that CYP2C9 induces angiogenesis via the activation of the EGF receptor, which is probably mediated via matrix metalloprotease activation followed by the release of EGF/HB-EGF.6 The results of the present investigation indicate that the induction of COX-2 expression also contributes to the angiogenic effect of CYP2C9, although the 2 pathways do not appear to be directly linked.
Acknowledgments
The authors are indebted to Isabel Winter for expert technical assistance. Research described in this article was supported by Philip Morris Inc, the Deutsche Forschungsgemeinschaft (FI 830/2-1), the National Institutes of Health (NIH GM31278, to J.R.F.), and by a Young Investigators Grant (to U.R.M.) from the Medical Faculty of the Johann Wolfgang Goethe Universit?t.
Received June 2, 2004; accepted November 12, 2004.
References
Campbell WB, Gebremedhin D, Pratt PF, Harder DR. Identification of epoxyeicosatrienoic acids as endothelium-derived hyperpolarizing factors. Circ Res. 1996; 78: 415–423.
Fleming I. Cytochrome P450 and vascular homeostasis. Circ Res. 2001; 89: 753–762.
Node K, Huo Y, Ruan X, Yang B, Spiecker M, Ley K, Zeldin DC, Liao JK. Anti-inflammatory properties of cytochrome P450 epoxygenase-derived eicosanoids. Science. 1999; 285: 1276–1279.
Li PL, Chen CL, Bortell R, Campbell WB. 11,12-Epoxyeicosatrienoic acid stimulates endogenous mono-ADP- ribosylation in bovine coronary arterial smooth muscle. Circ Res. 1999; 85: 349–356.
Zhang C, Harder DR. Cerebral capillary endothelial cell mitogenesis and morphogenesis induced by astrocytic epoxyeicosatrienoic acid. Stroke. 2002; 33: 2957–2964.
Michaelis UR, Fisslthaler B, Medhora M, Harder D, Fleming I, Busse R. Cytochrome P450 2C9-derived epoxyeicosatrienoic acids induce angiogenesis via cross-talk with the epidermal growth factor receptor (EGFR). FASEB J. 2003; 17: 770–772.
Potente M, Michaelis UR, Fisslthaler B, Busse R, Fleming I. Cytochrome P450 2C9-induced endothelial cell proliferation involves induction of mitogen-activated protein (MAP) kinase phosphatase-1, inhibition of the c-Jun N-terminal kinase, and up-regulation of cyclin D1. J Biol Chem. 2002; 277: 15671–15676.
Davidge ST. Prostaglandin H synthase and vascular function. Circ Res. 2001; 89: 650–660.
McAdam BF, Catella-Lawson F, Mardini IA, Kapoor S, Lawson JA, Fitzgerald GA. 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. 1999; 96: 272–277.
Chen J-K, Capdevila J, Harris RC. Heparin-binding EGF-like growth factor mediates the biological effects of P450 arachidonate epoxygenase metabolites in epithelial cells. Proc Natl Acad Sci U S A. 2002; 99: 6029–6034.
Vadlamudi R, Mandal M, Adam L, Steinbach G, Mendelsohn J, Kumar R. Regulation of cyclooxygenase-2 pathway by HER2 receptor. Oncogene. 1999; 18: 305–314.
Gauthier KM, Deeter C, Krishna UM, Reddy YK, Bondlela M, Falck JR, Campbell WB. 14,15-Epoxyeicosa-5(Z)-enoic acid: a selective epoxyeicosatrienoic acid antagonist that inhibits endothelium-dependent hyperpolarization and relaxation in coronary arteries. Circ Res. 2002; 90: 1028–1036.
Wang MH, Brand-Schieber E, Zand BA, Nguyen X, Falck JR, Balu N, Schwartzman ML. Cytochrome P450-derived arachidonic acid metabolism in the rat kidney: characterization of selective inhibitors. J Pharmacol Exp Ther. 1998; 284: 966–973.
Popp R, Brandes RP, Ott G, Busse R, Fleming I. Dynamic modulation of interendothelial gap junctional communication by 11,12-epoxyeicosatrienoic acid. Circ Res. 2002; 90: 800–806.
Fisslthaler B, Popp R, Michaelis UR, Kiss L, Fleming I, Busse R. Cyclic stretch enhances the expression and activity of coronary endothelium-derived hyperpolarizing factor synthase. Hypertension. 2001; 38: 1427–1432.
Gierse JK, Koboldt CM, Walker MC, Seibert K, Isakson PC. Kinetic basis for selective inhibition of cyclo-oxygenases. Biochem J. 1999; 339 (Pt 3): 607–614.
Vernia S, Sanz-Gonzalez SM, Lopez-Garcia MP. Involvement of peroxynitrite on the early loss of p450 in short-term hepatocyte cultures. Adv Exp Med Biol. 2001; 500: 209–212.
Ogasawara A, Arakawa T, Kaneda T, Takuma T, Sato T, Kaneko H, Kumegawa M, Hakeda Y. Fluid shear stress-induced cyclooxygenase-2 expression is mediated by C/EBP ?, cAMP-response element-binding protein, and AP-1 in osteoblastic MC3T3–E1 Cells. J Biol Chem. 2001; 276: 7048–7054.
Fleming I, Fisslthaler B, Michaelis UR, Kiss L, Popp R, Busse R. The coronary endothelium-derived hyperpolarizing factor (EDHF) stimulates multiple signalling pathways and proliferation in vascular cells. Pflugers Arch. 2001; 442: 511–518.
Wong PY, Lai PS, Shen SY, Belosludtsev YY, Falck JR. Post-receptor signal transduction and regulation of 14(R),15(S)-epoxyeicosatrienoic acid (14,15-EET) binding in U-937 cells. J Lipid Mediat Cell Signal. 1997; 16: 155–169.
Imig JD, Inscho EW, Deichamnn PC, Reddy KM, Falck JR. Afferent arteriolar vasodilation to the sulfonimide analog of 11,12-epoxyeicosatrienoic acid involves protein kinase A. Hypertension. 1999; 33: 408–413.
Olsen MK, Reszka AA, Abraham I. KT5720 and U-98017 inhibit MAPK and alter the cytoskeleton and cell morphology. J Cell Physiol. 1998; 176: 525–536.
Node K, Ruan XL, Dai J, Yang SX, Graham L, Zeldin DC, Liao JK. Activation of Gs mediates induction of tissue-type plasminogen activator gene transcription by epoxyeicosatrienoic acids. J Biol Chem. 2001; 276: 15983–15989.
Sun J, Sui XX, Bradbury A, Zeldin DC, Conte MS, Liao JK. Inhibition of vascular smooth muscle cell migration by cytochrome P450 epoxygenase-derived eicosanoids. Circ Res. 2002; 90: 1020–1027.
Topper JN, Cai J, Falb D, Gimbrone MA. 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. 1996; 93: 10417–10422.
Hsu AL, Ching TT, Wang DS, Song X, Rangnekar VM, Chen CS. The cyclooxygenase-2 inhibitor celecoxib induces apoptosis by blocking Akt activation in human prostate cancer cells independently of Bcl-2. J Biol Chem. 2000; 275: 11397–11403.
Adderley SR, Fitzgerald DJ. Oxidative damage of cardiomyocytes is limited by extracellular regulated kinases 1/2-mediated induction of cyclooxygenase-2. J Biol Chem. 1999; 274: 5038–5046.
Bolli R, Shinmura K, Tang XL, Kodani E, Xuan YT, Guo Y, Dawn B. Discovery of a new function of cyclooxygenase (COX)-2: COX-2 is a cardioprotective protein that alleviates ischemia/reperfusion injury and mediates the late phase of preconditioning. Cardiovasc Res. 2002; 55: 506–519.
Mukherjee D, Nissen SE, Topol EJ. Risk of cardiovascular events associated with selective COX-2 inhibitors. JAMA. 2001; 286: 954–959.
Iniguez MA, Rodriguez A, Volpert OV, Fresno M, Redondo JM. Cyclooxygenase-2: a therapeutic target in angiogenesis. Trends Mol Med. 2003; 9: 73–78.
Tsujii M, Kawano S, Tsuji S, Sawaoka H, Hori M, Dubois RN. Cyclooxygenase regulates angiogenesis induced by colon cancer cells. Cell. 1998; 93: 705–716.(U. Ruth Michaelis; John R)