Hydrogen peroxide decreases endothelial nitric oxide synthase promoter activity through the inhibition of AP-1 activity

Abstract

Previously, we have reported that endothelial nitric oxide synthase (eNOS) promoter activity is decreased in pulmonary arterial endothelial cells (PAECs) in response to hydrogen peroxide (H₂O₂). Thus the objective of this study was to identify the cis-element(s) and transcription factor(s) responsible for oxidant-mediated downregulation of the eNOS gene. Initial promoter experiments in PAECs treated with H₂O₂ revealed a significant decrease in activity of a promoter fragment containing 840 bp of upstream sequence of the human eNOS gene fused to a luciferase reporter. However, a promoter construct containing only 640 bp of upstream sequence had a significantly attenuated response to H₂O₂ challenge. As the 840-bp promoter construct had a putative binding site for the transcription factor activator protein-1 (AP-1) that was lacking in the 640-bp construct, we evaluated the effect of H₂O₂ on promoter activity after mutation of the AP-1 binding sequence (TGAGTCA at −661 to TGAGTtg in the 840-bp construct). Similar to the results seen with the 640 bp, the AP-1 mutant promoter had a significantly attenuated response to H₂O₂. EMSA revealed decreased binding of AP-1 during H₂O₂ treatment. Supershift analysis indicated that the AP-1 complex consisted of a c-Jun and FosB heterodimer. Furthermore, in vitro EMSA analysis indicated the c-Jun binding was significantly decreased after H₂O₂ exposure. Using chromatin immunoprecipitation analysis, we demonstrated decreased binding of AP-1 to the eNOS promoter in vivo in response to H₂O₂. These data suggest a role of decreased AP-1 binding likely through c-Jun in the H₂O₂-mediated decrease in eNOS promoter activity.

normal pulmonary vascular tone is regulated by a complex interaction of vasoactive substances produced by the vascular endothelium (12, 16, 20). These include nitric oxide (NO) and endothelin-1 (ET-1). NO is an endothelium-derived relaxing factor synthesized by the oxidation of l-arginine after activation of endothelial nitric oxide synthase (eNOS) (35). Studies using null mice have confirmed the importance of eNOS in the control of blood vessel tone (23), remodeling (37), hemostasis (14), angiogenesis (27), and the mobilization of endothelial progenitor cells (3). NO activates soluble guanylate cyclase, leading to elevation of cGMP and vasorelaxation (15). This is the primary basis for blood-flow and -pressure regulation. Loss of NO bioavailability is a key feature of endothelial dysfunction. Multiple pathways can reduce NO bioavailability by altering its synthesis or biodegradation. Excessive production of reactive oxygen species (ROS) and hydrogen peroxide (H₂O₂) seems to be a major mechanism of reduced vascular NO levels (19). Increased production of ROS has been implicated in the pathogenesis of cardiovascular diseases such as atherosclerosis, restenosis, hypertension, diabetic vascular complication, and heart failure (6, 9, 28, 46). Abnormal regulation of endothelial function has been implicated in the pathophysiology of a number of pulmonary hypertensive disorders. For example, in persistent pulmonary hypertension of the newborn (PPHN), pulmonary vascular resistance does not decrease normally at birth, resulting in pulmonary hypertension, right-to-left shunting, and hypoxemia (43). Newborns who die of PPHN exhibit an increase in the thickness of the smooth muscle layer within small pulmonary arteries and extension of this muscle to nonmuscular arteries (21), and there is also a decrease in eNOS expression (17). Surgical ligation of the ductus arterious in the fetal lamb generates a model of PPHN that resembles the human condition (2, 4, 50). Previously, our laboratory has reported an increase in both superoxide (8) and H₂O₂ (49) in this ovine model of PPHN. We also demonstrated that ET-1-mediated increases in H₂O₂ within fetal pulmonary arterial smooth muscle cells (FPASMC) produced decreases in eNOS promoter activity and eNOS activity in endothelial cells (FPAEC) cocultured with FPASMC (47). A similar effect was observed when FPAEC were challenged directly with H₂O₂ (47).

The eNOS promoter lacks a typical TATA box, but it has potential binding sites for a number of transcription factors including activator protein-1 (AP-1), Sp1, GATA, NF-κB, AP-2, and also shear stress-response elements and sterol-regulatory elements (29, 48). The Sp1 binding site at −104 is involved in the basal expression of eNOS (25, 45, 56), and we (47) have previously reported a decrease in eNOS promoter activity and eNOS expression in response to H₂O₂ treatment. Thus the purpose of this study was to identify the cis-element(s) and transcription factor(s) responsible for H₂O₂-mediated downregulation of the eNOS gene. Here, we demonstrate the involvement of AP-1 in H₂O₂-mediated downregulation of eNOS promoter in pulmonary arterial endothelial cells (PAECs).

MATERIAL AND METHODS

Cell culture.

Primary cultures of PAECs from late-gestation fetal lambs were isolated by the explant technique as we (48) have described previously. Briefly, the heart and lungs were obtained from fetal (138–140 days gestation) lambs after death. These fetal lambs had not undergone previous surgery or study. The main and branching pulmonary arteries were removed, and the exterior of the vessels were rinsed with 70% ethanol prepared using sterile distilled water water. The vessel was opened longitudinally, and the interior was rinsed with PBS to remove blood. The endothelium was lightly scraped away, placed in Dulbecco's modified Eagle's-H16 media (DME-H16) containing 10% FBS and antibiotics (penicillin and streptomycin), and incubated at 37°C in 21% O₂-5% CO₂-balance N₂. After 5 days, islands of endothelial cells (ECs) were cloned to ensure purity. Basic fibroblast growth factor (10 ng/ml) was added to the medium every other day. When confluent, the cells were passaged to maintain them in culture or frozen in liquid nitrogen. EC identity was confirmed by the typical cobblestone appearance, contact inhibition, specific uptake of acetylated low-density lipoprotein labeled with 1,1′-dioctadecyl-3,3,3′,3′-tetramethylindocarbocyanine, and positive staining for von Willebrand factor (DAKO, Carpinteria, CA). Ovine FPAECs were studied between passages 3 and 10.

Cell transfection and promoter assays.

We generated promoter constructs by PCR using human genomic DNA as a template. The reverse primer common to all constructs binds immediately upstream of the ATG initiation codon of eNOS (5′-GTTACTGTG CGT CCA CTCTGCTGCC-3′). Forward primers were 5′-TGTAGTTTCCCTAGTCCCCC-3′ (840-bp fragment) and 5′-GGTGTGGGGGTTCCAGGAAA-3′ (650-bp fragment). A 2-bp mutation was introduced into the AP-1 consensus sequence within the 840-bp construct. The sequence TGAGTCA at −661 in the human eNOS promoter (29) was changed to TGAGTtg by site-directed mutagenesis as we (48) described previously. FPAECs were cotransfected with 1.6 μg of test plasmid and 0.4 μg of β-galactosidase plasmid (as an internal control to normalize for transfection efficiency) on a 10-cm² tissue culture plate at 90% confluency with Effectene Transfection Reagent (Qiagen, Valencia, CA) according to the manufacturer's instructions. On the next day, the cells were split onto six-well plates and allowed to adhere for 36 h. The culture medium was then replaced with fresh DMEM containing 0 to 200 μM H₂O₂ for 16–18 h. The luciferase activity of 20 μl of protein extract was determined using Luciferase Assay System (Promega, Madison, WI), and β-galactosidase of 10 μl of protein extract was determined using β-galactosidase enzyme assay system (Promega) as transfection efficiency control.

Nuclear extract preparation and EMSA.

Nuclear extracts were prepared using the NE-PER nuclear extraction kit (Pierce Biotechnology, Rockford, IL). EMSA were performed using biotinylated double-stranded oligonucleotides corresponding to the AP-1 site at −661 of the eNOS promoter (29). We also used an AP-1 mutant oligonucleotide (AP-1mut) where a 2-bp mutation was introduced into the AP-1 consensus sequence at −661 of the human eNOS promoter. The single-stranded oligonucleotides were biotinylated using Biotin 3′ End DNA Labeling Kit (Pierce Biotechnology) to incorporate 1–3 biotinylated ribonucleotides (Biotin-11-UTP) onto the 3′ end of DNA strands using terminal deoxynucleotidyl transferase (TdT) and then annealed to make it double stranded. The sequence of the different oligonucleotides are AP-1 −677 5′-TTT TGT GTC CCC CAC TTG AGT CAT GGG GGT-3′ −648 and AP-1mut −677 5′-TTT TGT GTC CCC CAC TTG AGT TGT GGG GGT-3′ −648 [Integrated DNA Technologies (IDT), Coralville, IA]. Binding reactions involved incubating 10 μg of nuclear extract with biotinylated oligonucleotide and 1 μg poly(dI-dC) for 20 min at room temperature. In addition, the binding of recombinant purified c-Jun was examined in presence of H₂O₂. Purified c-Jun (1.2 μg, Promega) was treated with 10 μM H₂O₂ for 20 min. Binding reaction was then set up by adding biotinylated AP-1 oligonucleotide and 1 μg poly(dI-dC) for 20 min at room temperature. The DNA-protein complexes were resolved on a 5% nondenaturing polyacrylamide gel in 1× Tris-borate-EDTA (TBE) buffer and then transferred to nylon membrane, and biotinylated oligonucleotide was detected using a LightShift Chemiluminescent EMSA Kit (Pierce Biotechnology). Competition reactions were done with 10-, 50-, and 100-fold excess of unlabeled oligonucleotides. Supershift experiments were conducted by incubating nuclear extracts overnight at 4°C with 4 μg of polyclonal antibodies specific to c-Jun (cat. no. sc-1694X), FosB (cat. no. sc-48X), and Fra-1 (cat. no. sc-605X) all from Santa Cruz Biotechnology (Santa Cruz, CA) followed by the addition of biotinylated oligonucleotide. The complexes were visualized using an Image Station 440 CF (Eastman Kodak).

Chromatin immunoprecipitation.

Chromatin immunoprecipitation (ChIP) assays were performed essentially following the method of Braunstein et al. (7). FPAEC cells (∼1.0 × 10⁶) cultured in 10-cm plates were treated with H₂O₂ (100 μM) for 2 h and then cross-linked with 1% formaldehyde for 10 min, harvested, and sonicated to 200∼1,000-bp fragments. Supernatants were incubated overnight at 4°C with an anti-c-Jun antibody. After ethanol precipitation, DNA was resuspended in 20 μl for each 10⁶ cells, and 1 μl was used as PCR template. The primer pairs used to amplify the regions containing the AP-1-binding site in the −700/−1 promoter region, respectively, were 5′-TGCCTCAGCCCTAGTCTCTC-3′/5′-GATAGAGGCCCAGCAAGGAT-3′ (for the AP-1-binding site at −661 resulting in a fragment of 225 bp). A DNA sample from sonicated lysates that underwent reverse cross-link and phenol/chloroform extraction was used as a positive control (input).

Analysis of cell death.

FPAECs were treated with 0 to 200 μM H₂O₂ overnight. After incubation, the medium was collected and centrifuged for 5 min at 500 g, and the supernatant was stored at 4°C until assay. Relative cytotoxicity was quantified by measurement of release of the soluble cytoplasmic enzyme lactate dehydrogenase (LDH). LDH activity in cell-free supernatant was measured using a commercial kit (Roche Applied Science, Indianapolis, IN). After a given time course, 50-μl aliquots of cell-free medium were transferred from all wells to a fresh 96-well plate, and 50 μl of reconstituted substrate mix were added to each well. The plate was incubated for 30 min at room temperature in the dark, and the absorbance was recorded at 490 nm. Each culture sample was measured in quadruplicate, with a minimum of three samples per experimental group. Relative cytotoxicity was determined by comparison of absorbance of the experimental group with absorbance of a control cell group treated with 2% Triton X-100 cell lysis buffer according to the manufacturer's protocol.

At high concentration, H₂O₂ may be a toxic chemical. Thus we initially performed LDH-release assays to examine potential cytotoxic effects of H₂O₂. However, at all doses examined, H₂O₂ did not induce significant cytotoxicity (untreated = 0.0%; 100 μM for 2 h = 0.86% ± 0.06%; 200 μM overnight = 2.7% ± 0.2%; 2% Triton X-100 = 100% cell death).

Statistical analyses.

The means ± SE were calculated for all samples, and significance was determined either by the unpaired t-test (for 2 groups) or ANOVA with Newman-Keuls post hoc testing (for ≥3 groups). The statistical significance of differences was set at P < 0.05. Statistical analysis was performed using GraphPad Prism version 4.01 for Windows (GraphPad Software, San Diego, CA).

RESULTS

Effect of H₂O₂ on eNOS promoter activity.

We initially examined the effect of H₂O₂ on eNOS transcription using several promoter constructs from the human eNOS gene fused upstream of a luciferase reporter (Fig. 1A). When FPAECs were transfected with constructs containing 840 bp of upstream promoter sequence (−840 eNOS promoter construct) and then treated with 200 μM H₂O₂ for 16–18 h, there was a 65% decrease (P < 0.05) in eNOS promoter activity. However, when FPAECs were transfected with a promoter construct containing only 640 bp of upstream sequence (lacking an AP-1 binding site at −661), the H₂O₂-mediated decrease in eNOS promoter activity was attenuated. To determine the effect of AP-1 on the regulation of eNOS promoter activity, we transfected FPAEC with a promoter construct identical to the −840 eNOS promoter construct except for a 2-bp mutation in the wild-type AP-1 sequence at −661 (from TGAGTCA to TGAGTtg). We were unable to detect any decrease in promoter activity with H₂O₂ (Fig. 1B). The 650-bp construct and AP-1mut construct both lack a functional AP-1 (−661) sequence, suggesting that AP-1 is regulating eNOS promoter activity in response to H₂O₂. In addition, in the absence of H₂O₂ treatment, there was a 30–35% decrease (although these decreases did not reach statistical significance) in promoter activity of the −650 and AP-1mut constructs compared with the −840 construct, suggesting AP-1 may be also be involved in basal regulation of the eNOS promoter.

Fig. 1.

Transcriptional activity of the human endothelial nitric oxide synthase (eNOS) promoter in response to H₂O₂ in fetal pulmonary arterial endothelial cells (FPAEC). A: the constructs used contain −840 or −650 [without the activator protein-1 (AP-1) binding site] of eNOS promoter DNA upstream of the transcription start site. AP-1 mutant oligonucleotide (AP-1mut) is identical to the −840-bp construct except for a 2-bp mutation in the AP-1 consensus sequence at −661. B: FPAECs were transfected either with the −840, the −650, or the AP-1mut eNOS promoter construct fused to a luciferase reporter gene. Cells were then treated with 200 μM H₂O₂ for 16 h. The luciferase activity and β-galactosidase activity were measured in the cell lysate, and the ratio of luciferase activity to β-galactosidase activity was used to determine the normalized eNOS promoter activity. H₂O₂ significantly decreased the activity of the −840 human eNOS promoter fragment but did not affect the AP-1mut and −650 constructs. Values expressed are means ± SE, n = 6. *P < 0.05 vs. untreated.

Effect of H₂O₂ on AP-1 binding.

We next examined the effect of H₂O₂ on binding of AP-1 at −661 of the eNOS promoter. We performed EMSA using nuclear extracts prepared from FPAECs treated with 100 μM H₂O₂ for 2 h. Compared with control extracts, we observed a significant (42%) decrease in the binding of AP-1 (Fig. 2, A and B). We then carried out a study to confirm the specificity of AP-1 binding to an AP-1 oligonucleotide prepared from the eNOS promoter sequence. We mutated two nucleotides from the AP-1 consensus sequence, specifically CA to TG. The binding of the AP-1 complex was significantly decreased with this mutation (Fig. 3, A and B), and the specific AP-1 complex was competed out with 50× and 100× excesses of unlabeled wild-type AP-1 oligonucleotide but not with 10×, 50×, and 100× excesses of AP-1mut (Fig. 3C). To determine the composition of the AP-1 complex, we used anti-c-Jun, -FosB, and -Fra-1 antibodies in supershift analyses. The AP-1 complex was found to supershift with anti-c-Jun and anti-FosB antibodies but not with anti-Fra-1 (Fig. 3, D and E), suggesting that the AP-1 complex is a dimer of c-Jun and FosB. In addition, we found that the intensity of the band shift in c-Jun and FosB lanes was significantly decreased in the presence of the antibodies to c-Jun and FosB but not Fra-1 (Fig. 3, D and E).

Fig. 2.

Effect of H₂O₂ on transcription factor binding to the AP-1 cis-element in the human eNOS promoter. EMSA analyses were performed using a biotinylated AP-1 oligonucleotide corresponding to −661 of the human eNOS promoter. The biotinylated oligonucleotide was incubated for 20 min with nuclear extracts from FPAECs treated with 100 μM H₂O₂ for 2 h. The reaction was run on a 5% native polyacrylamide gel and transferred to a nylon membrane, and the biotinylated oligonucleotide was detected by using a streptavidin-horseradish peroxidase antibody and chemiluminescent substrate. Arrow indicates the position of the AP-1 complex. Lane marked Free is biotinylated oligonucleotide without any extract. The bar graph shows relative changes in band intensity for FPAECs exposed to H₂O₂ for 2 h. Band intensities were calculated by densitometric analysis. Binding is expressed as fold changes relative to untreated control (Cont). The values expressed are means ± SE, n = 6. *P < 0.05 vs. untreated.

Fig. 3.

Determination of the transcription factors constituting the AP-1 complex in FPAEC. A: EMSA using biotinylated probes containing sequences for AP-1 and AP-1mut, where 2 bp of the AP-1 consensus sequence TGAGTCA were changed to TGAGTtg and incubated with 5 or 10 μg of nuclear extracts (NE) from PAECs. Panel is representative of 3 independent experiments. B: densitometric analysis used to quantify the band intensity of the band shift of the AP-1 and AP-1mut oligonucleotide sequences in A identifies a significantly decreased binding to the AP-1mut compared with the wild-type (wt) sequence. *P < 0.05 vs. wild-type AP-1 sequence. C: to check the specificity of AP-1, binding competition was performed with 10×, 50×, and 100× excess of cold AP-1 wild-type or AP-1mut oligonucleotide. There was a decrease in the binding of AP-1 with 50× and 100× excess of wild-type but not with mutant. D: supershift assays were conducted by incubating NE with 4 μg of polyclonal antibodies specific to c-Jun, FosB, and Fra-1 overnight at 4°C. The supershifted complex is shown with an arrow. The complex is found to be composed of c-Jun and FosB but not Fra-1. Panel is representative of 3 independent supershift experiments. E: densitometric analysis was used to quantify the band intensities of the band shift and supershift and determine relative changes in band intensity. Binding is expressed as percentage of total shifted band. The values are expressed as means ± SE, n = 4. *P < 0.05 vs. band shift no antibody, †P < 0.05 vs. supershift no antibody.

It has been established that c-Jun contains conserved redox-sensitive cysteine residues in the DNA-binding domain (1). Previously, it has been reported that c-Jun and AP-1 activity can be altered by S-nitrosylation of cysteine (30) or by chemical oxidation of cysteine residues (52). Thus we analyzed the effect of H₂O₂ on the ability of purified recombinant c-Jun to bind a consensus AP-1 sequence in vitro after challenge with H₂O₂. Our data indicate that H₂O₂ (10 μM for 20 min) induced a significant (50%; P < 0.05 vs. untreated) decrease in c-Jun binding during H₂O₂ treatment (Fig. 4, A and B), suggesting that the oxidation of c-Jun may be responsible for the loss of AP-1 transcription activity after H₂O₂ challenge.

Fig. 4.

Effect of H₂O₂ on the binding activity of recombinant c-Jun. A: recombinant c-Jun (1.2 μg) was treated with 10 μM H₂O₂ (for 20 min) and then used in the binding reaction. There is a decrease in the c-Jun-to-the-AP-1 consensus sequence after H₂O₂ exposure. A representative image from 3 separate experiments is shown. B: the bar graph shows relative changes in c-Jun binding with and without H₂O₂ treatment. Band intensities were calculated by densitometric analysis. Binding is expressed as fold changes relative to untreated control. The values expressed are means ± SE, n = 4. *P < 0.05 vs. untreated.

Effect of H₂O₂ on the binding of AP-1 to the endogenous eNOS promoter in vivo.

To examine the effect of H₂O₂ on the binding of AP-1 to the endogenous promoter in vivo, ChIP analysis was used. FPAECs were treated with H₂O₂ (100 μM, 2 h), and the cells were processed. Similar to the luciferase and EMSA data, we found that H₂O₂ induced a significant decrease in the binding of AP-1 to the eNOS promoter in vivo (Fig. 5).

Fig. 5.

The effect of H₂O₂ on AP-1 binding to the endogenous ovine eNOS promoter in FPAEC. Chromatin immunoprecipitation (ChIP) assays were conducted with FPAECs treated with H₂O₂ (100 μM, 2 h) followed by PCR using specific primers for the eNOS promoter. A: representative gel showing a decrease in PCR product with H₂O₂ treatment. There is a decrease in the binding of AP-1 to the eNOS promoter with H₂O₂ treatment. B: bar graph indicating that there is a significant decrease in the band intensity after H₂O₂ treatment, although there is no change in the levels of input DNA. Data are means ± SE, n = 6. *P < 0.05 vs. untreated. Ip, immunoprecipitation analysis.

DISCUSSION

NO is one of the primary components for blood-flow and -pressure regulation, and loss of NO bioavailability is a key feature of endothelial dysfunction. Multiple pathways can reduce NO bioavailability by altering its synthesis or biodegradation, but excessive production of ROS including H₂O₂ seems to be a major mechanism of reduced vascular NO levels (19). Increased production of ROS has been implicated in the pathogenesis of cardiovascular diseases such as atherosclerosis, restenosis, hypertension, diabetic vascular complication, and heart failure (6, 9, 28, 46). In PPHN, there is a decrease in eNOS expression (17). Surgical ligation of the ductus arterious in the fetal lamb generates a model of PPHN that resembles this condition in humans (2, 4, 50). Previously, our laboratory has reported a decrease in eNOS mRNA and protein and an increase in ET-1 expression in this ovine model of PPHN (4). These changes in gene expression are associated with an increase in both superoxide (8) and H₂O₂ (49), which are associated with increased activity of the NADPH oxidase complex. Our (47) previous in vitro studies demonstrated that exogenous ET-1 stimulated an increase in H₂O₂ within FPASMC, which led to decrease in eNOS promoter activity and eNOS activity in FPAEC cocultured with FPASMC. In addition, eNOS expression decreased when FPAECs were treated with H₂O₂ (47). Together, these data suggest a link between increased ROS production and decreased eNOS expression in PPHN lambs. However, the mechanisms involved are incompletely understood.

The eNOS promoter has binding sites for a variety of transcription factors in the setting of a TATA-less promoter, including Sp1, Ets, GATA, NF-1, AP-1, and shear stress-response element (29). A number of transcription factors including NF-κB, AP-1, Sp1, and GATA are involved in the regulation of eNOS expression basally or in response to exogenous stimuli (18, 48, 56). Of the above-mentioned factors, AP-1 and Sp1 are redox-sensitive transcription factors (31, 32, 40). AP-1 is known to be involved in the upregulation of eNOS transcription by cyclosporine A (33), hypoxia (22), and shear stress (48). Sp1 is known to be involved in the upregulation of eNOS expression by insulin (13), lysophosphatidylcholine (10, 53), and cyanidin-3-glucoside (54), and Sp1 is also involved in the basal transcription of the eNOS gene (45, 56). Recently, we (26) have reported that AP-1 upregulates eNOS expression when ECs were treated with antioxidant nordihydroguaiaretic acid (NDGA). Thus we were interested to study the potential roles of AP-1 in modulating eNOS promoter activity in response to H₂O₂ treatment. We found a decrease in promoter activity when FPAECs, transfected with our −840 promoter construct, were treated with H₂O_2. The −840 construct contains an AP-1 binding site at −661, and mutation of 2 bp within the consensus AP-1 binding sequence specifically disrupts DNA binding (5, 24). Previously, we (48) have shown that introduction of the same mutations within the AP-1 sequence (AP-1mut) prevented shear-induced increase in promoter activity in FPAEC. In this study, the AP-1mut construct did not exhibit H₂O₂-mediated decreases in eNOS promoter activity (Fig. 1B), suggesting the involvement of AP-1 in the redox regulation of eNOS transcription. Similarly the −650 construct, which is truncated just before the AP-1 site at −661, did not appear to respond significantly to H₂O₂ (Fig. 1B), suggesting AP-1 is the major player responsible for the H₂O₂-mediated decreases in eNOS expression.

For EMSA, we used the sequence from −677 to −648 of the human eNOS promoter as a probe. This region includes the AP-1 binding site. AP-1 binding decreased in response to 100 μM H₂O₂ treatment (Fig. 2, A and B). Again, we found that the binding of AP-1 decreased significantly when we mutated the AP-1 consensus sequence (Fig. 3, A and B), suggesting the complex we observed in EMSA was formed at the AP-1 element at −661 of the eNOS promoter. The AP-1 transcription factors consist of Jun (c-Jun, JunB, JunD, and v-Jun) homodimers or heterodimers of Jun with Fos (c-Fos, FosB, Fra-1, and Fra-2), and in previous studies it has been shown that regulation of c-Jun expression in ECs regulates eNOS expression. For example, in cyclosporine A-treated human umbilical vein endothelial cells (33) and shear stress-treated PAECs (48), increased c-Jun/AP-1 binding activity at the eNOS promoter AP-1 site are associated with the upregulation of eNOS mRNA and protein. By contrast, in high glucose-treated human aortic ECs, c-Jun/AP-1 binding to eNOS promoter is accompanied by decreased eNOS mRNA (42). In addition, it has been shown in the human hepatoma cell line HuH-7 there is an increase in c-Jun expression and AP-1 DNA binding with 0.5 to 5 mM H₂O₂ (55). This treatment was also found to be associated with a dose-dependent increase in necrosis (55). However, although our PAECs were treated with 100 to 200 μM H₂O₂, we were unable to detect any increase in necrosis as measured by LDH release. Indeed, we (51) have previously found that doses of H₂O₂ of ≥250 μM are necessary to cause increases in PAEC cell death both by necrosis and apoptosis. Furthermore, we found that the cell death was also associated with increased mitochondrial-derived superoxide (51). However, at the doses we have used in this study, no effects on mitochondrial activity were observed (51).

Typically, AP-1 forms complexes with Jun (c-Jun, JunB, JunD, and v-Jun) and Fos (c-Fos, FosB, Fra-1, and Fra-2). In this study, we found that the AP-1 complex is composed of c-Jun and FosB, since it was supershifted with c-Jun- and FosB-specific antibodies (Fig. 3, D and E). Recently, it has been reported that overexpression of c-Jun increased eNOS expression in uterine artery ECs (36). Both c-Jun and c-Fos contain conserved redox-sensitive cysteine residues in the DNA-binding domains (1), and changes in redox potential can regulate AP-1 activity via these residues. Previously, cellular and in vitro studies using NO donor compounds have demonstrated that NO can inhibit the activity of c-Jun and bind to AP-1 sequence (34, 44). Furthermore, the activity of the AP-1 complex can be inhibited through S-nitrosylation of cysteine residues of c-Jun by NO (30) or by chemical oxidation of c-Jun cysteine residues (52). In confirmation of previous studies, we also observed decreased binding of purified recombinant c-Jun following H₂O₂ treatment (Fig. 4, A and B). For our promoter analysis study, we used a human eNOS promoter transfected into ovine FPAECs. Sequence analysis of the recently cloned ovine eNOS promoter identified a consensus AP-1 binding at −682 to −676 (36). To confirm the effect of H₂O₂ on AP-1 binding to the ovine eNOS promoter in vivo, we performed ChIP assays with FPAECs treated with 100 μM H₂O₂ for 2 h. We demonstrated a decrease in AP-1 binding, indicating that H₂O₂ also regulates AP-1 binding to the eNOS promoter in vivo.

It has been reported that H₂O₂ can increase eNOS promoter activity, eNOS mRNA stability, as well as eNOS activity as measured by the conversion of l-arginine to l-citrulline in bovine aortic endothelial cells (BAECs) (11, 41). Previously, our group (47) has shown that eNOS expression increased at lower H₂O₂ concentration and that BAECs have twofold higher catalase activity than FPAECs. In the same report, it has been shown that when BAECs and FPAECs were transfected with an eNOS promoter construct and treated with the same amount of H₂O₂, BAECs had increased eNOS promoter activity, whereas FPAECs had decreased eNOS promoter activity. Thus the effective concentration of H₂O₂ inside ovine FPAECs may be higher than BAECs as they have lower catalase activity. This suggests that the antioxidant capacity of the cells may also affect the eNOS expression. Recently, we have reported that, in the lamb, lung antioxidant enzymes catalase, SOD1, and SOD2 are regulated both by development and by increased pulmonary blood flow (39). Additional research is therefore warranted to characterize further the complex redox regulation of eNOS expression in ECs.

Overall, our data suggest a major role for AP-1 in the downregulation of eNOS expression in FPAEC treated with H₂O₂. We conclude that H₂O₂ decreases AP-1 binding at the eNOS promoter, resulting in decreased promoter activity. It has been reported previously that AP-1 is involved in modulating eNOS expression in response to anoxia (38) and hypoxia (22). However, we believe this to be the first report suggesting that AP-1 plays a major role in regulating eNOS expression in response to H₂O₂. By characterizing the molecular mechanisms of redox regulation of eNOS expression in vivo, we hope to identify potential therapies to reduce or prevent endothelial dysfunction in diseases such as PPHN where decreased eNOS expression is associated with increased ROS production.

GRANTS

This research was supported, in part, by National Heart, Lung, and Blood Institute Grants HL-60190, HL-67841, HL-72123, and HL-70061 and a Fondation Leducq Grant, all to S. M. Black.