Regulation of endothelial nitric-oxide synthase during hypoxia.

The mechanism by which nitric-oxide (NO) production increases during hypoxia is unknown. To explore the effect of hypoxia upon endothelial nitric-oxide synthase (ecNOS) activity and expression, we exposed bovine aortic endothelial cells (BAEC) to hypoxia (1% O2) for 0-24 h and measured levels of ecNOS mRNA, protein, and activity. The amount of ecNOS mRNA increases to more than twice the basal level after 6 h of hypoxia. Incubation of BAEC with actinomycin D during hypoxia prevents this increase, demonstrating that higher levels of mRNA observed during hypoxia are due to increased synthesis, not to increased stability of ecNOS mRNA. Levels of ecNOS protein increase throughout 24 h of hypoxia to more than twice normoxic levels. Although ecNOS expression increases within 2 h of hypoxia, total activity remains unchanged. To explore the transcriptional regulation of ecNOS, we constructed a reporter plasmid containing the ecNOS promoter region upstream of the luc gene and transfected this reporter plasmid into BAEC. In this system, hypoxia induces a linear increase over time in the expression of luciferase driven by the ecNOS promoter. It is concluded that hypoxia induces an increase in transcription of ecNOS in endothelial cells, activating the regulatory region of ecNOS by undefined transcription factors.

The mechanism by which nitric-oxide (NO) production increases during hypoxia is unknown. To explore the effect of hypoxia upon endothelial nitric-oxide synthase (ecNOS) activity and expression, we exposed bovine aortic endothelial cells (BAEC) to hypoxia (1% O 2 ) for 0 -24 h and measured levels of ecNOS mRNA, protein, and activity. The amount of ecNOS mRNA increases to more than twice the basal level after 6 h of hypoxia. Incubation of BAEC with actinomycin D during hypoxia prevents this increase, demonstrating that higher levels of mRNA observed during hypoxia are due to increased synthesis, not to increased stability of ecNOS mRNA. Levels of ecNOS protein increase throughout 24 h of hypoxia to more than twice normoxic levels. Although ecNOS expression increases within 2 h of hypoxia, total activity remains unchanged. To explore the transcriptional regulation of ecNOS, we constructed a reporter plasmid containing the ecNOS promoter region upstream of the luc gene and transfected this reporter plasmid into BAEC.

In this system, hypoxia induces a linear increase over time in the expression of luciferase driven by the ecNOS promoter. It is concluded that hypoxia induces an increase in transcription of ecNOS in endothelial cells, activating the regulatory region of ec-NOS by undefined transcription factors.
Systemic arteries vasodilate in response to hypoxia, delivering more blood to peripheral tissues. This dilation occurs within seconds after hypoxia, and is maintained for hours. The molecular basis of hypoxic vasodilation is not completely understood. Possible mechanisms include direct relaxation of smooth muscle cells induced by changes in pH, ion channel conductance changes, and decreases in ATP levels. Endothelial cells may also release lower amounts of vasoconstrictors, such as endothelin or thromboxane, or may release increased amounts of vasodilators such as adenosine, prostacyclin, endothelial hyperpolarizing factor, and nitric oxide (NO) 1 (1)(2)(3).
Several studies have shown that hypoxia increases the synthesis of NO from systemic arteries both ex vivo (28 -30) and in vivo (31,32). The effect of hypoxia on NO synthesis from endothelial cells in vitro is less clear; some studies report an increase, decrease, or no change (33)(34)(35). These studies exposed cells isolated from systemic vessels to hypoxia for different periods of time, which may explain the different effects of hypoxia. An increase in NO synthesis during hypoxia could be due to increased levels of calcium prolonging the activation of ecNOS, post-translational modification increasing the activity of ecNOS, or an increase in ecNOS synthesis. One study has shown that hypoxia for 1 day decreases ecNOS levels in vitro (33), and another study has shown that hypoxia for 7 days increases ecNOS in vivo (36).
In this study we tested the hypothesis that hypoxia induces the transcription of ecNOS. We show that hypoxia increases the amount of ecNOS mRNA and protein. Furthermore, we show that hypoxia can activate the transcription of a reporter gene under control of the ecNOS 5Ј-flanking region. However, while ecNOS levels increase, there is no increase in NOS activity in hypoxic cells.

EXPERIMENTAL PROCEDURES
Materials-The bovine ecNOS cDNA and the human ecNOS genomic clone have been previously described (24,37). Actinomycin D was obtained from Sigma; Opti-MEM and Lipofectin were obtained from Life Technologies, Inc. The pGL2-Basic plasmid, luciferase assay system, and ␤-galactosidase enzyme assay system were obtained from Promega. Donkey anti-rabbit Ig antibody conjugated to horseradish peroxidase and the chemiluminescent assay are from Amersham Corp.
Antibody Preparation-Affinity-purified polyclonal antibody against bovine ecNOS was prepared by standard techniques (38). In brief, a peptide composed of the terminal 15 amino acids of bovine ecNOS was synthesized by the Johns Hopkins Biopolymer Laboratory. The peptide was conjugated to thyroglobulin and injected into rabbits (Hazelton Research Products). Immune sera from injected rabbits was affinitypurified on a peptide-bovine serum albumin-agarose column.
Cell Culture-Bovine aortic endothelial cells were prepared from aortas obtained from the slaughterhouse and then cloned as described elsewhere (39). Their endothelial phenotype was verified by demonstrating acetylated LDL uptake and the expression of von Willebrand's factor by immunofluorescence. Cells were grown on 10-cm dishes coated with bovine gelatin in RPMI 1640 medium with 10% fetal bovine serum, 100 units/ml penicillin, and 100 g/ml streptomycin. Cells were used up to passage 19. When cells were approximately 90% confluent, they were refed with media and exposed to normoxia or hypoxia.
Hypoxia-BAEC grown in tissue culture plates were placed in an incubator that was continuously flushed with 1% O 2 , 5% CO 2 , 94% N 2 (Puritan Bennett, Linthicum Heights, MD). Prior measurements with an oxygen electrode showed that the partial pressure of oxygen was 8 * This work was supported by National Institutes of Health Grants HL52315 and K1102451 (to C. J. L.), the Swiss National Foundation, Ciba-Geigy Jubilä umsstiftung, and the Roche Research Foundation, Switzerland (to U. A. A.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
torr. 2 Cells were harvested 0 -24 h after being placed in this chamber.
Construction of Reporter Plasmid--A fragment of DNA containing the 5Ј-flanking region of the human ecNOS gene was amplified by the PCR. The template DNA for the PCR was a BlueScript plasmid (Stratagene) containing 6 kilobase pairs of genomic human DNA, including 1.6 kilobase pairs upstream from the human ecNOS gene (37). The primers for the PCR, designed to amplify DNA from 1.6 kilobase pairs upstream of the ecNOS AUG to the transcriptional start site of ecNOS, were: CGCGGTACCATCTGATGCTGCCTGTCACC (upstream) and GCGCGCAAGCTTGTTACTGTGCGTCCACTCTGC (downstream). The 1.6-kilobase pair PCR product (not shown) was inserted into KpnI and HindIII sites in the pGL2-Basic plasmid (Promega) upstream of the luc gene, and the nucleotide sequence was determined to confirm that no errors had been introduced by PCR amplification.
Transfection of BAEC-The luciferase reporter plasmid and the pSV-␤-galactosidase plasmid (Promega) as an internal control were introduced into BAEC by liposomal-mediated transfection according to the manufacturer's instructions (Life Technologies, Inc.). In brief, BAEC were incubated for 6 h with a mixture of 0.5 g of pSV-␤-galactosidase plasmid, 0.5 g of luciferase reporter plasmid, and 10 l of Lipofectin. BAEC were then fed with RPMI and grown for an additional 16 h. Transfected BAEC were then exposed to hypoxia as above for various times and harvested with lysis buffer within 30 s of removal from hypoxia. Cell lysates were assayed for luciferase activity using a luciferase assay kit (Promega) and a luminometer (Turner Designs 20). Lysates were also assayed for ␤-galactosidase activity with a kit (Promega) and assayed for protein concentration using a kit (Bio-Rad). The luminometer reported relative light units, which were normalized by dividing by galactosidase activity.
NOS Catalytic Assay-After exposure to hypoxia, BAEC were immediately harvested in sample buffer (50 mM Tris-HCl, pH 7.5, 1 mM EDTA, 1 mM EGTA), sonicated, and centrifuged at 1000 ϫ g for 5 min, and the supernatant was used to measure NOS activity by the conversion of [ 14 C]arginine to [ 14 C]citrulline (40). In brief, 75 g of total protein was incubated in 50 mM HEPES, pH 7.4, 10 M CaCl 2 , 3 M tetrahydrobiopterin, 1 M FAD, 1 M FMN, 0.1 M calmodulin, and 1 mM NADPH at 37°C for 10 min. The reaction was stopped by adding 3 ml of 20 mM HEPES, pH 5.5, 2 mM EDTA, and 2 mM EGTA. The resulting samples were applied to 1 ml columns of Dowex AG 50WX-8 (Na ϩ form), and the radioactivity in the flow-through was determined by liquid scintillation counting. Specific ecNOS activity is calculated as citrulline production from the above reaction mixture with 10 M CaCl 2 minus citrulline production from the above reaction mixture with 5 mM EDTA.
Northern Blot Analysis-BAEC were lysed in acidic guanidium isothiocyanate and RNA extracted (41,42). A total of 15 g of RNA was fractionated on an 1% agarose-MOPS gel and transferred to nitrocellulose membrane. Hybridization was conducted as described elsewhere (43). In brief, nitrocellulose membranes were hybridized at 65°C overnight with a bovine ecNOS cDNA probe that had been labeled with [ 32 P]dCTP. Blots were washed once with 2% SDS, 2 ϫ SSC at 22°C for 20 min, followed by 0.1% SDS, 0.1 ϫ SSC at 65°C for 20 min, and then exposed to x-ray film for 24 h.

RESULTS
Hypoxia Increases ecNOS Expression-BAEC were exposed to hypoxia for increasing periods of time, harvested, and assayed for ecNOS mRNA and protein. During hypoxia, the amount of ecNOS mRNA rises from a basal level in normoxic cells. The increase is slight during 2-4 h of hypoxia and peaks after 6 h; levels remain elevated after 24 h of hypoxia (Fig. 1). Re-oxygenation of BAEC for 1 h reduces the mRNA to basal levels. Densitometry shows that mRNA levels of ecNOS after 6 h hypoxia are roughly twice the basal levels (not shown). This experiment was repeated three times with similar results.
Hypoxia also increases the levels of ecNOS protein.  protein, which is expressed at a basal level in normoxic BAEC, begins to increase 2-4 h after hypoxia and continues to rise during 24 h of hypoxia ( Fig. 2A). Densitometry shows that ecNOS protein increases by a factor of 2.2 after 6 h of hypoxia (Fig. 2B). Re-oxygenation for 1 h has little effect on ecNOS protein after 24 h of hypoxia, in contrast to the decline in ecNOS mRNA levels. This experiment was repeated five times with similar results.
Hypoxia Decreases ecNOS Specific Activity-Arginine to citrulline conversion assays were performed to measure the activity of ecNOS in hypoxic BAEC. During 24 h of hypoxia, total NOS activity in BAEC decreases slightly (Fig. 3A), although the amount of ecNOS increases (Fig. 2). When relative NOS specific activity is calculated by dividing NOS activity (Fig. 3A) by NOS protein levels (Fig. 2B), the relative specific activity of ecNOS gradually decreases below that of NOS in normoxic controls (Fig. 3B). This decrease is not due to any increase in general protein synthesis, since the total amount of protein in each dish remains constant during hypoxia (Fig. 3C). These results are the average of three experiments in which NOS activity was measured in triplicate. Thus hypoxia decreases ecNOS relative specific activity.
Hypoxia Does Not Change ecNOS mRNA Stability-Hypoxia could increase ecNOS mRNA levels either by increasing its rate of transcription or by decreasing its destruction. In order to distinguish between these possibilities, actinomycin D was added to BAEC. BAEC were then placed either in an incubator or in an hypoxic chamber. The BAEC were harvested at various times and their RNA analyzed by Northern blotting. Hypoxia does not affect the stability of ecNOS mRNA (Fig. 4). Thus the increase of ecNOS mRNA levels during hypoxia is not due to increased stability.
Hypoxia Increases Transcription of a Reporter Gene Controlled by the ecNOS Regulatory Region-A reporter plasmid was constructed by inserting the 5Ј-regulatory region of the ecNOS gene into a plasmid so that it drives the expression of the luc cDNA (see "Experimental Procedures"). This reporter plasmid was transfected into BAEC, along with a plasmid constitutively expressing ␤-galactosidase as an internal control. Transfected BAEC were exposed to hypoxia, harvested, and assayed for luciferase and ␤-galactosidase activity.
Hypoxia increases expression of luciferase in BAEC transfected with the reporter plasmid containing the ecNOS regulatory region (Fig. 5). Luciferase is expressed in normoxic BAEC, just as ecNOS is normally expressed at a basal level in BAEC. Expression of luciferase relative to ␤-galactosidase rises linearly from 2 to 24 h of hypoxia, increasing 2.1-fold over normoxia after 24 h. This experiment was repeated three times with similar results.

DISCUSSION
Our study provides several lines of evidence that hypoxia regulates ecNOS and its transcription. Hypoxia increases the levels of ecNOS mRNA and protein. The increase in ecNOS mRNA is not due to an increase in its stability, since this does not change during hypoxia. Hypoxia also increases transcription of a reporter gene driven by the ecNOS 5Ј-flanking region. Thus increased transcription produces more ecNOS during hypoxia. FIG. 3. NOS activity in BAEC during hypoxia. BAEC were exposed to 1% O 2 for 0 -24 h. One sample was re-oxygenated for 1 h after exposed to hypoxia for 24 h (R). A, cell lysates were incubated with [ 14 C]arginine, and the amount of [ 14 C]citrulline formed was measured and expressed as picomoles of [ 14 C]citrulline/mg of total cell protein/min reaction. B, ecNOS activity (Fig. 3A) was divided by the relative amount of ecNOS protein ( Fig. 2A). C, the total protein concentration of BAEC from one tissue culture dish resuspended in 300 l of lysis buffer was measured as a function of time. Our data suggest that the influence of NOS on vascular tone during hypoxia is complex. Others have shown that during acute hypoxia, intracellular calcium levels rise within minutes in endothelial cells (35), which can activate ecNOS to synthesize NO. We show that during chronic hypoxia, more ecNOS is produced within hours, although total NOS activity does not change. The amount of NO synthesized during hypoxia in vivo is difficult to predict from our in vitro studies, since intracellular calcium concentrations, amount of ecNOS, and specific activity of ecNOS all influence NO synthesis. Furthermore, low levels of O 2 might prolong the half-life of NO by reducing the rate at which NO and O 2 combine to form NO 2 (44 -47). The net effect of hypoxia upon vascular tone in vivo is the sum of many influences, including ecNOS levels and NO synthesis, other endogenous vasodilators, and vasoconstrictors.
In contrast to our findings, others have reported that ecNOS mRNA levels actually decrease during hypoxia (33). The difference between our findings might be due to the different origins of the endothelial cells used: our study used endothelial cells isolated from aortas instead of from veins, and our cells are cloned. Furthermore, the other study only examined the effect of 48 h of hypoxia and might have missed the peak level of mRNA. We show that after rising initially, mRNA levels begin to decline after 24 h of hypoxia (although protein levels continue to increase), and it is possible that if this trend continues, then after 48 h of hypoxia the levels of mRNA in hypoxia cells are lower than in normoxic cells.
Although ecNOS is expressed constitutively in resting endothelial cells, the level of ecNOS expression can be altered by various stimuli, including shear-stress, estrogen, and tumor necrosis factor-␣ (13)(14)(15)(16)(17)(18)(19)(20)(21)(22)(23)(24)(25)(26)48). The DNA upstream from the ecNOS coding region contains potential regulatory sequences that may mediate control of ecNOS expression. For example, the 5Ј-flanking region includes the shear-stress response element (37,49) and two half-consensus sequences for the estrogen response element (49). We have shown that hypoxia is another stimulus for ecNOS expression. Although we have established that ecNOS transcription is increased during hypoxia, the enhancer elements and transcription factors mediating this increase have not yet been identified. The 5Ј-region flanking the ecNOS gene does not contain an HIF-1 consensus sequence (24,37). None of the consensus sequences in the 5Ј-flanking region are known to mediate hypoxic induction.
The total activity of ecNOS is unchanged after 24 h of hy-poxia, although the amount of enzyme has doubled. Why the activity does not increase along with the amount of enzyme is unclear: it is not due to a depletion of co-factors, since all co-factors are added exogenously for the NOS assay (see "Experimental Procedures"). ecNOS can be myristoylated, palmitoylated, and phosphorylated (50,51). However, it is not known whether these modifications occur during hypoxia. None of these modifications have been shown to change the activity of ecNOS. Thus oxygen regulates the transcription of ecNOS. The transcription factors that sense hypoxia and activate transcription of ecNOS have not yet been identified.