Abstract
Cytochrome P450 (P450)-dependent metabolites of arachidonic acid, the epoxyeicosatrienoic acids (EETs), are proposed to be endothelium-derived hyperpolarizing factors (EDHF) that affect vascular tone; however, the effects of EDHF on endothelial-derived nitric oxide biosynthesis remain unknown. We examined the regulation of endothelial nitric-oxide synthase (eNOS) by EDHF and investigated the relevant signaling pathways involved. The P450 epoxygenases CYP102 F87V mutant, CYP2C11-CYPOR, and CYP2J2 were transfected into cultured bovine aortic endothelial cells, and the effects of endogenously formed or exogenously applied EETs on eNOS expression and activity were assessed. Transfection with the P450 epoxygenases led to increased eNOS protein expression, an effect that was attenuated by cotreatment with the P450 inhibitor 17-ODYA. Northern analysis demonstrated that P450 transfection led to increased eNOS mRNA levels consistent with an effect at the pretranslational level. P450 epoxygenase transfection resulted in increased eNOS activity as measured by the conversion of l-arginine to l-citrulline. Addition of synthetic EETs (50-200 nM) to the culture media also increased eNOS expression and activity. Treatment with mitogen-activated protein kinase (MAPK), MAPK kinase, and protein kinase C inhibitors apigenin, 2′-amino-3′-methoxyflavone (PD98059), and 1-(5-isoquinolinesulfonyl)-2-methylpiperazine (H-7), respectively, significantly inhibited the effects of P450 transfection on eNOS expression. Overexpression of P450 epoxygenases or addition of synthetic EETs increased Thr495 phosphorylation of eNOS, an effect that was inhibited by both apigenin and PD98059. Overexpression of P450 epoxygenases in rats resulted in increased aortic eNOS expression, providing direct evidence that EDHF can influence vascular eNOS levels in vivo. Based on this data, we conclude that EDHF up-regulates eNOS via activation of MAPK and protein kinase C signaling pathways.
Vascular endothelial cells control vascular tone and modulate blood flow to organs by synthesizing and releasing the vasoactive autocoids endothelium-derived relaxing factor (EDRF), which is synonymous with nitric oxide (NO), and prostacyclin (PGI2) (Furchgott and Zawadzki, 1980; Palmer et al., 1987). Among these, NO probably plays a more important role. In vascular endothelium, NO is produced by a constitutively expressed enzyme known as endothelial nitric-oxide synthase (eNOS), which converts l-arginine to l-citrulline (Vallance et al., 1989; Moncada and Higgs, 1993). In addition to endothelium-dependent vasodilatation, NO also has a number of other critical functions in the vascular system, including inhibition of platelet aggregation, inhibition of endothelial cell adhesion molecule expression, prevention of vascular smooth muscle cell migration and proliferation, and prevention of intravascular coagulation and thrombosis (Azuma et al., 1986; De Caterina et al., 1995; Freedman et al., 1997; Jeremy et al., 1999; Loscalzo, 2001). Therefore, NO is an important factor in the maintenance of normal vascular homeostasis and the protection of vessels from injuries induced by atherogenic processes such as smooth muscle cell proliferation, platelet aggregation, monocyte adhesion, and oxidative modification of low-density protein (Moncada and Higgs, 1993; Cannon, 1998).
In addition to NO and PGI2, endothelial cells synthesize and release a third factor called endothelium-derived hyperpolarizing factor (EDHF), which causes hyperpolarization of underlying vascular smooth muscle cells via activation of Ca2+-activated K+ channels (Cohen and Vanhoutte, 1995; Harder et al., 1995). EDHF-dependent actions may serve as a backup to NO-dependent vasodilatory mechanisms, but EDHF seems to be a more important regulator of vascular tone under certain pathological conditions, and in certain vascular beds such as in the coronary microcirculation (Najibi and Cohen, 1995; Oltman et al., 1998). Epoxyeicosatrienoic acids (EETs), which are cytochrome P450 epoxygenase-derived metabolites of arachidonic acid, display many of the characteristics of EDHF and are thought to be potential candidates for EDHF (Cohen and Vanhoutte, 1995; Harder et al., 1995; Gebremedhin et al., 1998). Indeed, transfection of endothelial cells with an antisense oligonucleotide to a P450 arachidonic acid epoxygenase (CYP2C8/34) results in attenuation of EDHF-mediated vascular responses, and treatment with the P450 inducer β-naphthaflavone enhances EDHF-mediated vasorelaxation (Fisslthaler et al., 1999).
Both NO and EDHF are important vasodilatory autocoids released by the endothelium, but their mechanisms of action are quite different. Remarkably little is known regarding potential interactions between these two important vascular mediators and the enzymes that are responsible for their biosynthesis. A previous study has demonstrated that NO attenuates the synthesis and/or release of EDHF (Bauersachs et al., 1996); however, the effect of EDHF on eNOS expression and activity in endothelial cells has not been investigated. In the present study, we used cultured bovine aortic endothelial cells (BAECs) to study the effects of EDHF on the eNOS pathway. We found that transfection with several different P450 epoxygenases or addition of physiological concentrations of synthetic EETs resulted in increased eNOS expression and activity. Moreover, the effects of EDHF on eNOS expression involve activation of both mitogen-activated protein kinase (MAPK) and protein kinase C (PKC) signal transduction pathways.
Materials and Methods
Materials. All standard cell culture reagents were obtained from Invitrogen (Carlsbad, CA), including Dulbecco's modified Eagle's medium (DMEM), trypsin, and fetal bovine serum (FBS). PD98059, apigenin, and H-7 were supplied by Calbiochem-Novabiochem (Darmstadt, Germany). [3H]l-Arginine was supplied by PerkinElmer Life Sciences (Boston, MA). Anti-eNOS and anti-phospho-eNOS (Thr495) antibodies were purchased from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA). Anti-ERK1/2 (p44/42 MAPK) and anti-phospho-ERK1/2 or were purchased from New England Biolabs (Beverly, MA). Anti-CYP2C11 and anti-CYP102 antibodies were a generous gift from Dr. Jorge Capdevila (Vanderbilt University, Nashville, TN). Anti-CYP2J2 antibodies were prepared as described previously (Wu et al., 1996). Enhanced chemiluminescent substrate (SuperSignal West Pico chemiluminescent substrate) was purchased from Pierce Chemical (Rockford, IL). Hybrisol solution was purchased from Intergen (Purchase, NY). Polyvinylidene difluoride and nylon membranes were purchased from Schleicher & Schuell (Dassel, Germany). The PathDetect signal transduction pathway trans-reporting system and the Primer-It II random primer labeling kit were purchased from Stratagene (Kinsport, TN). [α-32P]dCTP (3000 μCi/mmol) was purchased from YaHui Nuclear (Beijing, China). Superfect cell transfection reagent was supplied by QIAGEN GmbH (Hilden, Germany). The full-length cDNA of human eNOS was a generous gift from Dr. James K. Liao (Harvard University, Boston, MA). The CYP102 F87V mutant cDNA and a cDNA encoding rat CYP2C11 fused to rat NADPH-cytochrome P450 oxido-reductase (CYPOR) (Helvig and Capdevila, 2000) were kindly provided by Dr. Jorge Capdevila (Vanderbilt University). All other chemicals and reagents were purchased from Sigma-Aldrich (St. Louis, MO) unless otherwise specified.
Isolation and Culture of Endothelial Cells. Fresh bovine thoracic aortas were obtained at a local slaughterhouse. BAECs were harvested using trypsin (0.25%) and grown to confluence in a growth medium containing DMEM supplemented with 5 mM l-glutamine, 10% FBS, and an antibiotic mixture of penicillin (100 units/ml) and streptomycin (100 μg/ml). The identification of cultured cells as endothelial cells was performed by examination for the typical morphology using phase contrast microscopy and by immunofluorescent staining for factor VIII-related antigen. All passages were performed using 0.05% trypsin and 0.02% EDTA. Only endothelial cells passaged less than four times were used for experiments.
Construction and Preparation of Recombinant Adeno-Associated Virus. The recombinant adeno-associated virus (rAAV) vector pXXUF1, packaging plasmid pXX2, adenovirus helper plasmid pXX6, and a rAAV plasmid containing the GFP cDNA (GFP-pUF1) were a generous gift from Dr. Xiao Xiao (University of Pittsburgh, Pittsburgh, PA). The coding regions of CYP102 F87V mutant, CYP2C11-CYPOR, and CYP2J2 were subcloned into pXXUF1 downstream from the cytomegalovirus promoter to produce the constructs CYP102 F87V-pUF1, CYP2C11-CYPOR-pUF1, and CYP2J2-pUF1, respectively. The rAAV was made as described previously (Xiao et al., 1996, 1998). Briefly, human 293 cells were grown in DMEM supplemented with 10% FBS and antibiotics. One to 2 h before transfection, each 15-cm-diameter plate of cells (70-80% confluent) was fed with 15 ml of fresh medium. A total of 85 μg of plasmid DNA (molar ratio pUF1/pXX2/pXX6, 1:1:1) was added to a CaCl2 solution (final concentration, 125 mM) and then quickly mixed with BES buffer (final concentration, 25 mM N,N-bis[2-hydroxyethyl]-2-aminoethane sulfonic acid, 140 mM NaCl, 0.75 mM Na2HPO4, pH 7.0). The resultant DNA/calcium/BES mixture was incubated at room temperature for 30 min before adding to the cells. After 8 to 12 h, the medium containing the DNA/calcium/BES mixture was replaced with fresh medium. After 48 to 72 h, the cells were harvested by centrifugation. Cell pellets were resuspended in 1 to 2 ml of 100 mM NaCl/10 mM Tris-HCl (pH 8.5), subjected to four freeze-thaw cycles, and cell debris was removed by centrifugation. For large-scale rAAV preparations, 40 plates, each containing ∼5 × 106 cells, were used and a single-step gravity-flow column purification method was carried out as described previously (Auricchio et al., 2001). The titer of rAAV was determined by dot blot hybridization. The eluted rAAV was aliquoted and stored at -80°C. The resultant rAAVs were designated rAAV·CYP102 F87V, rAAV·CYP2C11-CYPOR, rAAV·CYP2J2, and rAAV·GFP, respectively.
Overexpression of P450 Epoxygenases in Endothelial Cells. The coding regions of CYP102 F87V mutant, CYP2C11-CYPOR,and CYP2J2 were cloned into the KpnI and XbaI, ClaI, and XbaI and HindIII sites of the mammalian expression vector pCB6 to produce the constructs CYP102 F87V-pCB6, CYP2C11-CYPOR-pCB6, and CYP2J2-pCB6, respectively. Bovine rather than human endothelial cells were used for transient transfection because of their higher transfection efficiency (Peng et al., 1995). Cultured endothelial cells were seeded into six-well plates and were grown to ∼60% confluence. Cells were then transfected using Superfect transfection reagent exactly according to the manufacturer's instructions. After 2 to 3 h, the DNA-Superfect mixture was removed and the cells were incubated continuously in the medium containing 10% FBS for 48 h. After that, the cells were processed for immunoblotting, northern analysis, eNOS activity assay, EET measurements, and analysis of signal transduction pathways (see below). In some experiments, BAECs were infected with rAAV·CYP102 F87V, rAAV·CYP2C11-CYPOR, rAAV·CYP2J2, or rAAV·GFP (∼50 virions/cell), and the cells were processed 1 week later.
Animals and Treatment. Male Sprague-Dawley rats (200-250 g) were fed standard laboratory chow ad libitum, given free access to tap water, and maintained on a 12-h light/dark cycle at constant temperature and humidity. All experiments were conducted according to the National Institutes of Health Guide for the Care and Use of Laboratory Animals and approved by the local animal care and use committee. Rats (N = 6/group) were injected intravenously via the tail vein with either vehicle, empty pCB6 vector, CYP102 F87V-pCB6, CYP2C11-CYPOR-pCB6, or CYP2J2-pCB6 (5 mg DNA/kg body weight). After 2 weeks, animals were sacrificed by decapitation, and thoracic aortas were harvested, cleaned in phosphate-buffered saline from adjacent tissues, snap-frozen in liquid nitrogen, and stored at -80°C until processing.
Nitric-Oxide Synthase Activity Assay. NO synthase activity was determined in cell homogenates by measuring the conversion of [3H]l-arginine to [3H]l-citrulline as described previously with minor modifications (Ramasamy et al., 1998). Forty-eight hours after transfection, the cells were washed twice with ice-cold phosphate-buffered saline and harvested by scraping in 200 μl of HEPES buffer (20 mM, pH 7.2) containing 0.2 M sucrose, 1 mM EDTA, 1 mM dithiothreitol, 2 μg/ml leupeptin, 2 μg/ml aprotinin, and 1 mM phenylmethylsulfonyl fluoride. In some studies, endothelial cells were pretreated with the NOS inhibitor l-NMMA (1 mM) for 15 min at room temperature. The cells were then homogenized, with a Kontes MicroUltrasonic cell disrupter (Kontes, Vineland, NJ), and the homogenates were used for NOS activity assay. Each sample (200 μl) was added to 50 μl of a medium containing 20 μM l-arginine, 10 μg/ml calmodulin, 3 mM CaCl2, 1 mM NADPH, and 10 μCi [3H]l-arginine and was incubated at 37°C for 30 min. The reaction was quenched by addition of 500 μl of ice-cold stop buffer (20 mM HEPES, 2 mM EDTA). The total reaction mixture was loaded onto a column containing 2 ml of Dowex 50WX-8 resin (pre-equilibrated with NaOH) followed by elution of [3H]l-citrulline with 2 ml of water. Aliquots (100 μl) of the eluents were counted in a liquid scintillation counter (Beckman Coulter, Inc., Fullerton, CA). The net radioactivity was determined by subtracting the counts per minute observed in the presence of l-NMMA from those observed in the absence of l-NMMA. The protein concentration in the homogenates was measured spectrophotometrically by the method of Bradford. NOS activity was determined as the production of [3H]l-citrulline per minute per milligram of protein.
Western Blotting. Forty-eight hours after plasmid transfection, or 1 week after rAAVs infection, or 24 h after addition of 17-ODYA, PD98059, apigenin, or H-7, or 2 to 12 h after addition of 14,15-EET, 11,12-EET, or 8,9-EET, BAECs were washed twice with ice-cold phosphate-buffered saline and then lysed with an ice-cold buffer containing 50 mM Tris-Cl (pH 8.0), 150 mM NaCl, 0.02% NaNS, 0.1% SDS, 1 μg/ml aprotinin, 100 μg/ml phenylmethylsulfonyl fluoride, 1% Nonidet P-40, and 0.05% deoxycholic acid. Lysates were centrifuged at 12,000g for 2 min, and supernatants were used for Western blot analysis. Frozen aortas were homogenized individually in the same buffer and centrifuged as described above. Protein concentrations of the lysates were determined by the method of Bradford. Ten micrograms of protein per lane and prestained molecular weight markers (Bio-Rad, Hercules, CA) were separated by SDS-PAGE (8% running, 4% stacking gel). The proteins were electrophoretically transferred onto polyvinylidene difluoride membranes overnight at 4°C. The membranes were incubated at room temperature for 2 h with a blocking solution containing 5% nonfat dried milk, 10 mM Tris-Cl (pH 7.5), 100 mM NaCl, and 0.1% Tween 20. The membrane was incubated overnight at 4°C with a 1:500 dilution of polyclonal rabbit anti-eNOS and then treated with a goat anti-rabbit secondary antibody conjugated to horseradish peroxidase (Jackson Immunoresearch Laboratories, West Grove, PA) at room temperature for 2 h. After incubation with each antibody, the membrane was washed four times with 10 mM Tris-Cl (pH 7.5), 100 mM NaCl, and 0.1% Tween 20 at room temperature. Immunodetection was accomplished by using chemiluminescent substrate and then exposure to X-ray film. For some experiments, rabbit anti-phospho-eNOS (1:750 dilution), rabbit anti-phospho-ERK1/2 (1:1000 dilution), rabbit anti-ERK1/2 (1:1000 dilution), rabbit anti-CYP2C11 (1:1000 dilution), rabbit anti-CYP102 (1:1000 dilution) or rabbit anti-CYP2J2 (1:1000 dilution) were used instead of the anti-eNOS primary antibody.
Northern Blotting. Forty-eight hours after transfection, total RNA from endothelial cells was extracted using TRI reagent (Molecular Research Center, Inc., Cincinnati, OH). Equal amounts of total RNA (10 μg/lane) were separated on formaldehyde-1.2% agarose gels, transferred overnight onto nylon membranes by capillary action, and baked for2hat80°C before prehybridization for2hat42°C in Hybrisol solution. Radiolabeling of the human eNOS and human GAPDH cDNAs were performed using a Random Primer labeling kit and [α-32P]dCTP. The membrane was hybridized with the probes overnight at 42°C in Hybrisol solution. Northern blots were subjected to stringent washing conditions (0.1× standard saline citrate, 0.1% SDS at 42°C) before autoradiography with an intensifying screen at -80°C for up to 5 days.
Measurement of Endogenous EET Levels. Forty-eight hours after transfection, endothelial cells were harvested by scraping, rapidly frozen in liquid nitrogen, and maintained at -80°C until analysis. Methods to quantify endogenous EETs in endothelial cells have been reported previously (Node et al., 2001). Briefly, 1.5 to 2.0 × 107 cells were homogenized in phosphate-buffered saline, extracted under acidic conditions with chloroform/methanol, and extracts evaporated in tubes containing [1-14C]EET internal standards (57 μCi/μmol, 30 ng each). Saponification to recover phospholipid-bound EETs was followed by silica column purification, reverse-phase high-pressure liquid chromatography separation of individual EETs, derivatization to corresponding EET-pentafluorobenzyl esters, normal-phase high pressure liquid chromatography purification of derivatized EETs, and gas chromatography/mass spectrometry analysis with selected ion monitoring at m/z 319 and m/z 321.
Analysis of Signal Transduction Pathways. A PathDetect signal transduction pathway trans-reporting system was used according to the manufacturer's instructions to examine whether CYP102 F87V transfection altered MAPK signal transduction pathways. Cultured BAECs were seeded into six-well plates, grown to ∼60% confluence, and cotransfected using Superfect transfection reagent with pFA-ELK1, pFR-Luc, and either pCB6-CYP102 F87V or empty pCB6 vector, or positive and negative control plasmids included in the system. If there is activation of MAPK pathway after pCB6-CYP102 F87V transfection, phosphorylation of ELK1 will activate transcription of the luciferase gene from the reporter plasmid (pFR-Luc). Luciferase levels were determined on a luminometer (Lumat LB 9507; Lumat, Bad Wildbad, Germany) using a luciferase assay kit (Stratagene) and reflect the activation status of the MAPK pathway.
Results
Effect of P450 Transfection on eNOS Expression. Relatively pure (>95%) bovine aortic endothelial cell cultures were confirmed by their morphological features (i.e., cuboidal, cobblestone appearance; contact inhibited) using phase contrast microscopy and by immunofluorescent staining with antibodies to factor VIII-related antigen (data not shown).
Previous studies have demonstrated that CYP102 F87V and CYP2C11-CYPOR are active arachidonic acid epoxygenases that biosynthesize the EETs (Capdevila et al., 1992a; Graham-Lorence et al., 1997; Qu et al., 1998; Helvig and Capdevila, 2000). Transfection of BAECs with CYP102 F87V-pCB6 or CYP2C11-CYPOR-pCB6 resulted in abundant expression of the corresponding recombinant P450 proteins as determined by immunoblotting with specific P450 antibodies (Fig. 1A). To examine the effect of overexpression of these P450s on eNOS protein expression in vitro, BAECs were transfected with either CYP102 F87V-pCB6, CYP2C11-CYPOR-pCB6 or the empty pCB6 vector. Forty-eight hours after transfection, cells were lysed, and eNOS protein levels were determined by immunoblotting using a specific eNOS antibody. As shown in Fig. 2A, transfection of BAECs with either CYP102 F87V-pCB6 or CYP2C11-CYPOR-pCB6 resulted in a marked increase in eNOS protein expression compared with transfection with the empty pCB6 vector. Treatment of the cells with the P450 epoxygenase inhibitor 17-ODYA (50 and 100 μm) significantly attenuated the up-regulatory effect of P450 transfection on eNOS protein expression (Fig. 2A). We have previously shown that CYP2J2 is abundant in endothelial cells and likely contributes significantly to the epoxidation of endogenous arachidonic acid pools (Wu et al., 1996; Node et al., 1999, 2001). To examine the effect of overexpression of this endothelial epoxygenase on eNOS expression, we also transfected BAECs with the CYP2J2-pCB6 vector. Transfection of BAECs with CYP2J2-pCB6 resulted in expression of the recombinant CYP2J2 protein (Fig. 1A). Expression of CYP2J2 was slightly lower than that of CYP2C11 and CYP102 F87V. As shown in Fig. 2B, CYP2J2 transfection also resulted in a significant increase in eNOS protein levels. Densitometric analysis of immunoblots from multiple independent experiments normalized to β-actin expression revealed that CYP102 F87V, CYP2C11-CYPOR, and CYP2J2 transfection were associated with a 5.1-, 5.8-, and 2.8-fold induction of eNOS protein compared with transfection with pCB6 alone (p < 0.05) (Fig. 2C).
To examine whether the effect of P450 transfection on eNOS expression occurred at the pretranslational level, total RNA was extracted from the BAECs and eNOS mRNA levels were assessed by Northern analysis using a specific cDNA probe. As shown in Fig. 3A, eNOS mRNA levels were significantly increased after CYP102 F87V-pCB6, CYP2C11-CYPOR-pCB6, and CYP2J2-pCB6 transfection compared with transfection with the empty pCB6 vector. Densitometric analysis of blots from separate experiments normalized to GAPDH signals to control for loading differences revealed that P450 transfection was associated with a 2- to 3-fold increase in eNOS mRNA levels (p < 0.05) (Fig. 3B). Consistent with the immunoblotting data, CYP2J2 induced eNOS mRNA to a slightly lesser extent than did CYP102 F87V and CYP2C11-CYPOR.
Effect of P450 Transfection on eNOS Activity. To examine whether the changes in eNOS mRNA and protein were accompanied by corresponding changes in eNOS activity, we measured the conversion of [3H]l-arginine to [3H]l-citrulline in P450-transfected BAECs. Compared with cells transfected with the empty pCB6 vector, cells transfected with either CYP102 F87V-pCB6 or CYP2C11-CYPOR-pCB6 exhibited ∼3-fold more eNOS activity (p < 0.05) (Fig. 4).
Effect of Synthetic EETs on eNOS Protein Expression. Cultured endothelial cells were seeded into six-well plates and were grown to ∼80% confluence. The cells were treated with physiologically relevant concentrations (50-200 nM) of 8,9-EET, 11,12-EET, and 14,15-EET or corresponding vehicle (ethanol) for 4 h. As shown in Fig. 5A, addition of exogenous EETs resulted in a significant, dose-dependent increase in eNOS protein levels. Each of the EET regioisomers tested was active in increasing eNOS protein expression. Densitometric analysis of immunoblots from multiple independent experiments normalized to β-actin expression revealed that 8,9-EET, 11,12-EET, and 14,15-EET (200 nM) were associated with a significant 2.6-, 2.5-, and 3.3-fold increase in eNOS protein expression compared with vehicle (p < 0.05) (Fig. 5B). The induction of eNOS by 8,9-EET and 11,12-EET (200 nm) occurred in a time-dependent manner up to 4 h (Fig. 5C). eNOS expression returned to basal levels 12 h after EET treatment (data not shown). In agreement with the immunoblotting data, addition of exogenous EETs resulted in increased conversion of [3H]l-arginine to [3H]l-citrulline (p < 0.05) (Fig. 5D). There were no statistically significant differences in the magnitude of conversion by the three EET regioisomers.
Effect of P450 Transfection on Endothelial EET Levels. We have previously shown that control BAECs rapidly metabolize exogenous arachidonic acid to EETs (Node et al., 1999) and contain 8.0 ng of total endogenous EET per 107 cells (∼50 nM total EET) (Node et al., 2001). These levels are similar to those used in the experiments in Fig. 5, which examine effects of exogenous EETs on eNOS expression and activity. We have also shown that CYP2J2-transfected endothelial cells exhibit a 2-fold increase in arachidonic acid epoxygenase activity (Node et al., 1999) and contain significantly increased levels of EETs (Node et al., 2001) compared with control endothelial cells. For example, 14,15-EET levels increased 5% and 11,12-EET levels increased 35% after transfection with CYP2J2. To document that CYP2C11-CYPOR transfected cells also have increased arachidonic acid epoxygenase activity, we measured levels of EETs by gas chromatography/mass spectometry in BAECs under control conditions and after transfection with CYP2C11-CYPOR. Consistent with our previously published data, control endothelial cells contained 8.7 ng of total endogenous EET per 107 cells (∼55 nM). After transfection with CYP2C11, there was a 9% increase in 14,15-EET and a 46% increase in 11,12-EET. These data are consistent with the known regiochemical selectivity of CYP2C11 (Qu et al., 1998) and the presence of an active endothelial epoxide hydrolase with preference for metabolism of 14,15-EET (Zeldin et al., 1993; Fang et al., 2001). The concentration of endogenous EETs in CYP2J2- and CYP2C11-transfected endothelial cells are also within the range used in the experiments in Fig. 5. Moreover, these data are consistent with the greater degree of eNOS induction by CYP2C11-CYPOR compared with CYP2J2 (Fig. 2, B and C).
Effect of P450 Epoxygenase Products on eNOS Phosphorylation. Previous studies have demonstrated that eNOS is regulated by phosphorylation, which can influence enzyme activity, subcellular trafficking, and interaction with other proteins (Garcia-Cardena et al., 1996; Matsubara et al., 1996; Chen et al., 1999b; Butt et al., 2000; Fleming et al., 2001a; Michell et al., 2001). For example, phosphorylation of Thr495 has been shown to deactivate eNOS by hindering binding to calmodulin (Fleming et al., 2001a; Aoyagi et al., 2003). Hence, we conducted experiments using an antibody that reacts with Thr495-phosphorylated eNOS (p-eNOS). Treatment of BAECs with each of the synthetic EETs (100 nM) significantly increased expression of p-eNOS (Fig. 6A). Densitometric analysis of immunoblots from multiple independent experiments normalized to β-actin expression revealed that addition of 8,9-EET, 11,12-EET, and 14,15-EET was associated with a 1.5-, 2.3-, and 2.8-fold increase in p-eNOS expression compared with addition of vehicle (p < 0.05) (Fig. 6B). Moreover, rAAV-mediated overexpression of CYP2C11-CYPOR, CYP2J2, or CYP102 F87V in BAECs (Fig. 1B) also significantly increased p-eNOS expression (Fig. 7A). Densitometric analysis of immunoblots from multiple independent experiments normalized to β-actin revealed that P450 overexpression was associated with a 1.7- to 1.9-fold increase in p-eNOS expression compared with GFP control (p < 0.05) (Fig. 7B).
Effect of P450 Epoxygenase Overexpression on eNOS and p-eNOS Expression in Vivo. To confirm the physiological relevance of our findings, we conducted in vivo experiments in rats to determine the effects of P450 epoxygenase overexpression on vascular eNOS and p-eNOS expression. Rats were injected intravenously with CYP102 F87V-pCB6, CYP2C11-CYPOR-pCB6, CYP2J2-pCB6, or the empty pCB6 vector. Two weeks later, animals were sacrificed, aortas were harvested, and immunoblots were performed. Injection of the P450 expression vectors resulted in abundant aortic expression of the corresponding recombinant proteins as determined by immunoblotting with specific P450 antibodies (Fig. 1C). Importantly, P450 overexpression was associated with significant increases in aortic eNOS and p-eNOS levels (Fig. 8A). Densitometric analysis of immunoblots from multiple independent experiments normalized to β-actin revealed that P450 overexpression resulted in a significant 1.5- to 2.2-fold increase in aortic eNOS expression and a significant 1.3- to 1.5-fold increase in aortic p-eNOS expression (p < 0.05) (Fig. 8B). These data provide direct evidence that P450 epoxygenases can influence eNOS and p-eNOS expression in vivo.
Effect of P450 Transfection on eNOS Expression Involves MAPK and PKC Signal Transduction Pathways. To investigate the involvement of various signaling pathways in the mechanism of P450-induced up-regulation of eNOS expression, we pretreated BAECs with the MAPK inhibitor apigenin (25 μM), the MAPK kinase inhibitor PD98059 (20 μM), or the PKC inhibitor H-7 (12 μM) and then evaluated eNOS protein levels by immunoblotting. Pretreatment with either PD98059 or apigenin inhibited both CYP102 F87V- and CYP2C11-CYPOR-induced up-regulation of eNOS, thus suggesting involvement of the MAPK signal transduction pathway in this process (Fig. 9, A and B). Pretreatment with H-7 caused a small up-regulation of eNOS, but significantly attenuated both CYP102 F87V- and CYP2C11-CYPOR-induced up-regulation of eNOS, indicating involvement of the PKC signaling pathway as well (Fig. 9C). Importantly, both apigenin and PD98059 significantly limited induction of eNOS by synthetic 14,15-EET (Fig. 10, A and B). Both apigenin and PD98059 also inhibited the increased expression of p-eNOS caused by treatment with 14,15-EET (Fig. 10, A and B). These data provide evidence to support involvement of the MAPK pathway in EET-mediated induction and phosphorylation of eNOS.
The PathDetect signal transduction pathway trans-reporting system was used to confirm the activation status of MAPK pathway. BAECs were cotransfected with pFA-ELK1, pFR-Luc, and either CYP102 F87V-pCB6 or empty pCB6 vector. As shown in Fig. 11A, transfection with CYP102 F87V resulted in a significant increase in luciferase activity compared with transfection with the empty pCB6 vector (p < 0.05). Because activation of mitogen-activated protein kinases occurs through phosphorylation, we also examined the effect of P450 epoxygenase transfection on the phosphorylation status of MAPK. Forty-eight hours after transfection with either CYP102 F87V-pCB6, CYP2C11-CYPOR-pCB6, CYP2J2-pCB6 or empty pCB6 vector, cells were lysed and levels of total or phosphorylated p44/42 MAPK were determined by immunoblotting using antibodies to p44/42 MAPK or phospho-p44/42 MAPK, respectively. As shown in Fig. 11B, transfection of BAECs with the three P450 epoxygenases resulted in increased levels of phosphorylated p44/42 MAPK but no significant change in levels of total p44/42 MAPK. Densitometric analysis revealed a 35 to 45% increase in phosphorylated p44/42 MAPK in P450-transfected cells relative to empty pCB6 vector transfected cells (p < 0.05) (Fig. 11C). Together, these data demonstrate conclusively that P450-induced up-regulation of eNOS involves the MAPK signal transduction pathway.
Discussion
In addition to NO and PGI2, vascular endothelial cells release a third factor termed EDHF, which causes hyperpolarization and relaxation of underlying vascular smooth muscle cells (Cohen and Vanhoutte, 1995; Harder et al., 1995). Recent studies from a number of laboratories have confirmed that cytochrome P450 metabolites of arachidonic acid display many of the characteristics of EDHF and are potential candidates for mediating EDHF vascular responses (Cohen and Vanhoutte, 1995; Harder et al., 1995; Gebremedhin et al., 1998; Lu et al., 2001). Cytochrome P450 epoxygenases convert arachidonic acid into four EET regioisomers (5,6-EET, 8,9-EET, 11,12-EET, and 14,15-EET), which are converted by epoxide hydrolases to corresponding DHET regioisomers (Fitzpatrick and Murphy, 1988; Capdevila et al., 1992a). EETs and/or DHETs have been shown to possess a number of potent biological activities, including effects on peptide hormone secretion, cardiac function, and renal function (Fitzpatrick and Murphy, 1988; Capdevila et al., 1992b; Lu et al., 2001; Zeldin, 2001). The EETs have been reported to be potent mitogens which mediate the effects of epidermal growth factor via a mechanism that involves activation of Src kinase and initiation of a tyrosine kinase phosphorylation cascade (Chen et al., 1999a). Recently, Node et al. (1999) have reported that EETs decrease cytokine-induced endothelial cell adhesion molecule expression via inhibition of nuclear factor-κB and prevent leukocyte adhesion to the vascular wall. EETs also increase tissue plasminogen activator expression and fibrinolytic activity via activation of Gαs (Node et al., 2001). Hence, EETs possess homeostatic properties in the vasculature in addition to their vasodilatory actions.
EDRF, which is synonymous with NO, is produced from l-arginine by eNOS in the vascular endothelium. Like EDHF, it plays a pivotal role in the maintenance of normal vascular homeostasis and the regulation of systemic blood pressure (Vallance et al., 1989; Moncada and Higgs, 1993). In addition, it is clear that NO, like the cytochrome P450 epoxygenase products, has a number of other important functions in the vessel wall, including inhibition of platelet aggregation and adhesion molecule expression, prevention of smooth muscle proliferation and modulation of vascular growth, and prevention of coagulation and thrombosis (Moncada and Higgs, 1993; Cannon, 1998). Thus, NO may be considered an antiatherogenic, antiproliferative, and anti-thrombotic factor. We hypothesized that because both EDHF (EETs and/or DHETs) and EDRF (NO) have similar functions within the vasculature, albeit through different mechanisms, interactions between NO and EDHF in endothelial cells might very well occur. Indeed, a previous study has demonstrated that NO attenuates the synthesis and/or release of EDHF (Bauersachs et al., 1996).
CYP102 (also called P450BM-3), isolated from Bacillus megaterium, is an enzyme that contains both a cytochrome P450 domain and a reductase domain fused into a single polypeptide chain to form a self-contained catalytic unit (Fulco, 1991). The enzyme catalyzes the high turnover, NADPH-dependent metabolism of arachidonic acid to 18-hydroxyeicosatetraenoic acid (Graham-Lorence et al., 1997). The active site replacement of phenylalanine 87 with valine has been shown to convert CYP102 into a highly regio- and stereoselective arachidonic acid 14(R),15(S)-epoxygenase (99% optical purity) (Graham-Lorence et al., 1997). Similarly, we and others have previously reported that rat CYP2C11 is an active arachidonic acid epoxygenase that forms mixtures of 14,15-EET, 11,12-EET, 8,9-EET, and 5,6-EET (Capdevila et al., 1992; Qu et al., 1998; Helvig and Capdevila, 2000). In the present study, these P450s were used as pharmacological tools to investigate the effects of enhanced endogenous EDHF biosynthesis on eNOS expression and activity in cultured bovine endothelial cells. However, neither CYP102 F87V nor CYP2C11-CYPOR is normally present within the vasculature; hence, we also examined the effect of overexpression of the more functionally relevant CYP2J2, an endothelial arachidonate epoxygenase (Node et al., 1999; Node et al., 2001), on eNOS expression and activity. We found that transfection of BAECs with any of these P450s increased the expression of eNOS, both at the mRNA level and the protein level. The effects of CYP102 F87V and CYP2C11-CYPOR were generally more pronounced than that of CYP2J2, consistent with the relative catalytic efficiencies of these P450 enzymes. Moreover, transfection with CYP102 F87V or CYP2C11-CYPOR also increased eNOS activity as measured by the conversion of l-arginine to l-citrulline. Treatment with 17-ODYA, an inhibitor of cytochrome P450s, attenuated the up-regulatory effect of P450 transfection on eNOS expression. This supports the hypothesis that the effects of P450 on eNOS expression are not due to a direct interaction between the two heme-containing proteins, but are rather due to an effect of a P450 metabolite on the relevant signaling pathways involved in regulation of eNOS levels. In this regard, we showed that addition of nanomolar concentrations of exogenous EETs to the BAECs also up-regulated eNOS expression. Importantly, the biological effects of exogenously added, synthetic EETs occurred at concentrations similar to those that we have previously shown (and confirmed herein) to be present endogenously in BAECs (Node et al., 2001). Based on these data, we conclude that the effects of EETs on eNOS occur at physiologically relevant concentrations.
As discussed above, EETs play an important role in the regulation of vascular homeostasis. A number of studies suggest that EDHF may enhance the relaxing actions of NO but may not be the primary mediator of endothelium-dependent relaxation under normal physiological conditions and in most vascular beds (Cohen and Vanhoutte, 1995; Quilley et al., 1997). But, EDHF seems to be a more important regulator in some pathological conditions that are associated with impairment of the nitric oxide biosynthetic pathway such as during hypercholesterolemia (Kilpatrick and Cocks, 1994; Najibi and Cohen, 1995) and in some vascular beds such as the coronary microcirculation (Oltman et al., 1998). Indeed, Bauersachs et al. (1996) found that the production of EDHF is dampened by NO under physiological conditions and suggested that when NO synthesis is impaired, EDHF-dependent vasorelaxation may maintain endothelial vasodilator function, at least in part, by alleviating this intrinsic inhibition (Bauersach et al., 1996). Moreover, several studies have shown that the biosynthesis of EETs is increased under certain pathological conditions such as during hypercholesterolemia in vivo in rabbits and upon exposure of endothelial cells to oxidized low-density protein in vitro (Pritchard et al., 1990; Pfister et al., 1991). Importantly, a growing list of conditions, including those commonly associated with increased risk of atherosclerosis (e.g., hypertension, hypercholesterolemia, and diabetes) are associated with diminished synthesis and/or release of NO (Cannon, 1998). Our finding herein that P450-derived EETs can up-regulate eNOS suggest one potential compensatory mechanism for enhancing the production of NO in these conditions.
The EETs have been shown to play critical roles in regulating a variety of intracellular signaling pathways (Chen et al., 1998; Chen et al., 1999a; Node et al., 1999, 2001; Rzigalinski et al., 1999). The findings herein that 1) the MAPK pathway is activated in CYP102 F87V transfected BAECs using a trans-reporting system; 2) levels of the phosphorylated MAPK are increased in CYP102 F87V-pCB6-, CYP2C11-CYPOR-pCB6-, and CYP2J2-pCB6-transfected cells; and 3) inhibitors of MAPK, MAPK kinase, and PKC can attenuate the up-regulation of eNOS induced by P450 transfection or addition of synthetic EETs suggest that the mechanism of EDHF-mediated up-regulation of eNOS involves activation of both MAPK and PKC signal transduction pathways. In this regard, Fleming et al. (2001b) have recently shown that EETs activate the MAPK pathway in human endothelial and vascular smooth muscle cells. Importantly, activation of the MAPK pathway by growth factors can stimulate eNOS expression (Zheng et al., 1999).
eNOS can be phosphorylated at multiple different sites by different kinases; phosphorylation at some of the sites leads to activation of eNOS and phosphorylation at other sites leads to deactivation (Garcia-Cardena et al., 1996; Matsubara et al., 1996; Chen et al., 1999b; Butt et al., 2000; Fleming et al., 2001a; Michell et al., 2001). Recent studies have focused on the role of Thr495 phosphorylation in regulating eNOS activity after agonist stimulation (Fleming et al., 2001a; Aoyagi et al., 2003). Thr495 has been shown to be phosphorylated both in vitro and in vivo in endothelial cells by protein kinase C, AMP-activated protein kinase, and cyclic-nucleotide-dependent protein kinases (Matsubara et al., 1996; Chen et al., 1999b; Butt et al., 2000; Fleming et al., 2001a). Thr495 phosphorylation deactivates eNOS by hindering the binding of calmodulin (Fleming et al., 2001a; Aoyagi et al., 2003). Our data indicate that P450-derived EETs increase Thr495 phosphorylation of eNOS and suggest that this occurs, at least in part, via the MAPK pathway. Thus, EETs seem to have complex effects on eNOS expression and phosphorylation status. Importantly, the net effect of EETs is to increase eNOS activity. Future studies will examine the effects of P450-derived eicosanoids on phosphorylation of eNOS at other sites.
We present several lines of evidence to support our hypothesis that P450 metabolites of arachidonic acid can influence eNOS expression and activity. First, transfection of endothelial cells with three different P450 arachidonic acid epoxygenases (CYP2C11, CYP102 F87V, and CYP2J2) induces eNOS mRNA, protein and activity. Second, the P450 inhibitor 17-ODYA significantly attenuates the effect of P450 epoxygenase transfection on eNOS expression. Third, treatment of endothelial cells with synthetic EETs, at levels that are physiologically relevant and comparable with those present endogenously in endothelial cells, results in the up-regulation of eNOS protein and activity. Forth, induction of eNOS by P450 epoxygenases is associated with activation of the MAPK signaling pathway and inhibitors of this signaling pathway attenuate eNOS induction by P450. Fifth, inhibitors of MAPK also inhibit eNOS induction by synthetic EETs. Sixth, P450 epoxygenase transfection or treatment with synthetic EETs increase phosphorylation of eNOS at Thr495. Seventh, overexpression of P450 epoxygenases in vivo leads to increased expression of both eNOS and p-eNOS. Together, these data provide compelling evidence that important interactions exist between these two functionally relevant biosynthetic pathways.
In conclusion, this study demonstrates that transfection of BAECs with P450 epoxygenases in vitro or overexpression of P450 epoxygenases in vivo results in increased eNOS expression and increased phosphorylation of eNOS at Thr495. The net effect in BAECs is enhanced eNOS activity. Treatment with a P450 inhibitor attenuates the up-regulatory effect of P450 transfection on eNOS expression suggesting an effect of a P450 metabolite rather than a direct interaction between the two heme-containing proteins. Application of physiological concentrations of synthetic EETs to BAECs also increases eNOS and p-eNOS expression. The up-regulation of eNOS by P450 epoxygenase products involves activation of both MAPK and PKC signaling pathways. Together, these data provide new insight into the regulation of endothelial cell function by EDHF and highlight important interactions between two functionally relevant pathways. Given that cardiovascular and related disorders are frequently associated with endothelial dysfunction and diminished NO production, the present study provides a framework for development of novel therapeutics designed to improve endothelial function by modulating cytochrome P450 epoxygenase levels and/or activity.
Acknowledgments
We are grateful to Dr. Jorge Capdevila for providing the CYP102 F87V and CYP2C11-CYPOR cDNAs and corresponding polyclonal antibodies, to Dr. James K. Liao for providing the eNOS cDNA, and to Dr. Xiao Xiao for providing the rAAV plasmids. We also thank Drs. Joyce Goldstein and Tom Eling for helpful comments during the preparation of this manuscript.
Footnotes
-
This work was supported by the funds from National Natural Science Foundation Committee of China (39870307) and the National Institute of Environmental Health Sciences/National Institutes of Health Division of Intramural Research.
-
DOI: 10.1124/jpet.103.052787.
-
ABBREVIATIONS: EDRF, endothelium-derived relaxing factor; NO, nitric oxide; PGI2, prostacyclin; eNOS, endothelial nitric-oxide synthase; p-eNOS, phosphorylated endothelial nitric-oxide synthase (Thr495); EDHF, endothelium-derived hyperpolarizing factor; P450, cytochrome P450; EET, epoxyeicosatrienoic acids; BAEC, bovine aortic endothelial cell; MAPK, mitogen-activated protein kinase; PKC, protein kinase C; DMEM, Dulbecco's modified Eagle's medium; FBS, fetal bovine serum; 17-ODYA, 17-octadecynoic acid; ERK, extracellular signal-regulated kinase; CYPOR, NADPH-cytochrome P450 oxidoreductase; rAAV, recombinant adeno-associated virus; GFP, green fluorescent protein; l-NMMA, NG-monomethyl-l-arginine; DHET, dihydroxyeicosatrienoic acid; PD98059, 2′-amino-3′-methoxyflavone; H-7, 1-(5-isoquinolinesulfonyl)-2-methylpiperazine.
- Received April 7, 2003.
- Accepted August 8, 2003.
- The American Society for Pharmacology and Experimental Therapeutics