Sp1 Increases Expression of Cyclooxygenase-2 in Hypoxic Vascular Endothelium

Cyclooxygenase-2 (COX-2) catalyzes prostaglandin synthesis from arachidonic acid and is expressed locally in aortic aneurysm and heart failure. Cellular hypoxia is also found in these conditions. We have previously shown that cox-2 is transcriptionally regulated by hypoxia in human umbilical vein endothelial cells (HUVEC) in culture via the transactivation factor NF-κB p65, leading to increased production of prostaglandin E2, an inhibitor of vascular smooth muscle cell proliferation. Sp1 is a transactivation factor known to be important in the regulation of cytokine expression in association with NF-κB. We hypothesized that Sp1 is involved in the induction of cox-2in hypoxic HUVEC. Electrophoretic mobility shift assays with hypoxic HUVEC nuclear protein showed that both Sp1 and the related protein Sp3 specifically bound to the cox-2 promoter. Immunoblotting demonstrated that hypoxia increased the nuclear localization of Sp1 but did not change the Sp3 content in HUVEC. Overexpression of Sp1 through transfection of HUVEC enhancedcox-2 promoter activity as measured by reporter gene expression and by the production of COX-2. The specificity of the results was confirmed by mutation of the Sp1-binding site in thecox-2 promoter construct and by reproducibility in an Sp-deficient Drosophila SL2 cell line. The regulatory role of Sp1 discovered in this work supports the concept that a mechanistic link exists between vascular cellular hypoxia and mediators of inflammation associated with aortic aneurysm and heart failure.

Although the pharmacologic inhibition of cyclooxygenase activity with aspirin is a cornerstone of modern cardiovascular therapy, we do not yet understand how endothelial cyclooxygenases function in health or in disease. The cyclooxygenases, also referred to as prostaglandin endoperoxide synthases or PGH synthases, catalyze the rate-limiting step in prostaglandin synthesis. A constitutive cyclooxygenase (COX-1) and an inducible cyclooxygenase (COX-2) have been identified.
Myocardial hypoxia has been found in animal models of heart failure (3). Both COX-2 expression and NF-B activation appear in heart failure (4). The wall of an aneurysmal human aorta is also hypoxic (5) and has increased cellular COX-2 content and PGE 2 production (6). These findings are not surprising in light of our earlier work, within which we demonstrated that cox-2 is transcriptionally regulated by hypoxia via the transactivation factor NF-B p65 in human vascular endothelial cells (7), leading to increased production of PGE 2 (8).
Nuclear factor-B p65 (RelA) is one of the NF-B family of transcriptional activator proteins. The p65 subunit is known to be responsible for initiating transcription by DNA binding (9), but it is also likely that a number of other proteins bind with dimerized p65-p50 to initiate NF-B-mediated transcription so as to allow a gene-specific response to this ubiquitous transcription factor (10). The intracellular signaling mechanism that leads to induction of cox-2 by hypoxia in human vascular endothelium includes binding of p65 to the NF-B consensus element closest to the transcription start site. However, we learned in earlier deletion experiments that there is a relationship between the length of the region upstream of this NF-B element and the degree of induced transcription. It appeared to us that binding of the NF-B p65 is a necessary but not sufficient step in hypoxic induction of COX-2 (7).
After finding that cytoplasmic NF-B p65 and IB␣ (an inhibitory protein that binds NF-B p65 precursors) levels are not changed by hypoxia, we hypothesized that other factors might play a role in regulating the cox-2 promoter. The HMG I(Y) family of proteins features multiple A⅐T hooks and is associated with NF-B-mediated transactivation. We recently discovered that hypoxia increases expression of HMG I(Y) proteins while facilitating transactivation of the cox-2 promoter (8).
Sp1 is typically a positive-acting transcription factor that is ubiquitously expressed and required for the expression of a variety of genes (11). It is known to be important in the regulation of cytokine and human immunodeficiency virus gene activation in association with NF-B (12). Sp3 is a bi-functional regulator of transcription (repressor/activator, depending on the context) that competes with Sp1 for the same binding site. Sp1 mediates expression of the constitutive cyclooxygenase (COX-1) (13). Transcription of the hypoxia-inducible factor-1␣ (HIF-1␣) gene is initiated just downstream of two Sp1 sites (14). Extension of our previous search for hypoxia-related regulatory elements in the cox-2 promoter (15) revealed an Sp1binding site just upstream of the NF-B-3Ј element that we already know enhances cox-2 transcription in hypoxia. We therefore hypothesized that Sp1 is involved in the induction of cox-2 in hypoxic human vascular endothelial cells, and this hypothesis is supported by the results of experiments detailed herein.

EXPERIMENTAL PROCEDURES
Cell Culture-HUVEC were obtained as cryopreserved primary cultures that demonstrated factor VIII-related antigen and low density lipoprotein uptake (Clonetics). HUVEC were grown in Medium 199 with 2.2 g of NaHCO 3 /liter, 50 units/ml penicillin, and 50 g/ml streptomycin (Life Technologies, Inc.) with 5% fetal bovine serum (HyClone Labs), 50 g/ml endothelial cell growth supplement (Collaborative Biomedical Products), 50 g/ml heparin, and 1.0 g/ml hydrocortisone. The medium was prepared with a pH of 7.3. Incubator conditions were either normoxic (21% O 2 , 5% CO 2 ) or hypoxic (1% O 2 , 5% CO 2 , balance N 2 ) in a humidified incubator with an interior temperature of 37°C. The medium was equilibrated to the environmental gas conditions overnight before cellular exposure. HUVEC cells were studied at the third to fifth passages. Hypoxic stimulation was produced with ambient oxygen concentrations of 1% (using a controlled incubator with CO 2 /O 2 monitoring and CO 2 /N 2 gas sources). Reoxygenation was prevented through immediate replacement of hypoxic medium with lysis buffers while the cells were on ice. The pO 2 was generally 30 -40 mm Hg in hypoxia and 160 mm Hg in normoxia. Hypoxic HUVECs were therefore maintained in an oxygen environment approximating that of desaturated or venous blood. Their appearance and growth were indistinguishable from those of normoxic cells by phase contrast microscopy. Drosophila Schneider SL2 cells were obtained from the American Type Culture Collection and grown at 25°C without CO 2 in Schneider's medium (Life Technologies, Inc.) supplemented with 10% fetal bovine serum, 50 units/ml penicillin, and 50 g/ml streptomycin. SL2 Cells were passaged every 4 days.
Plasmids-The wild type of COX-2 promoter construct pGL2-NF-B-3Ј-WT (containing bases Ϫ531 to ϩ65 relative to the transcriptional start site) was described previously (7) and is referred to as pWT here. The Sp1 mutant of COX-2 promoter construct pSPM was created by mutation of the Sp1-binding site in the pWT from 5Ј-GGGAGG-3Ј to 5Ј-GAAAGG-3Ј (Ϫ245 to Ϫ240) by PCR. To generate pSPM, two sets of PCR primers were used to amplify the COX-2 promoter DNA. Set 1 included forward primer NF-B-P1 (Ϫ531 to Ϫ510) and reverse mutagenic primer CX9LMR (Ϫ221 to Ϫ250, 5Ј-CCCACTCTCCTGTCTGATC-CCTTTCTCTCC-3Ј, mutated bases are italicized and bold). Set 2 included forward mutagenic primer CX9LMF (Ϫ250 to Ϫ221, 5Ј-GGAGAGAAAGGGATCAGACAGGAGAGTGGG-3Ј, mutated bases are italicized and bold) and reverse primer NF-B-P2 (ϩ65 to ϩ46). Both NF-B-P1 and NF-B-P2 were previously described (7). Two fragments generated from these two sets of primers by PCR were gel-purified and combined for 4 cycles of PCR. NF-B-P1 and NF-B-P2 were then added for another 28 PCR cycles. The resulting fragment was then subcloned into pGL2-Basic at KpnI and XhoI in the multiple cloning site. The correctness of the mutation was confirmed by sequencing. The PCR protocol for each cycle was 94°C for 30 s, 65°C for 30 s, and 72°C for 80 s. The pPacSp1 and pPac␤-gal plasmids were kindly provided by Dr. Robert Tjian (16). The pPacSp3, pCMVSp1, and pCMVSp3 plasmids were kindly provided by Dr. Guntram Suske (17).
Western Blots-Cell lysate and nuclear protein extraction were performed as described previously (7). Protein extracts were separated on SDS-polyacrylamide gel, blotted to the membrane, and reacted with polyclonal anti-Sp1 or anti-Sp3 antibody (Santa Cruz Biotechnology) in blocking buffer for 30 min at 37°C and then incubated similarly with horseradish peroxidase-conjugated anti-IgG for 30 min and washed in TBST buffer (20 mM Tris-HCl (pH 7.5), 137 mM NaCl, and 0.1% Tween 20) for 5 min five times. The membrane was then incubated in luminol ECL detection reagents (Amersham Pharmacia Biotech) and exposed to the film. A monoclonal anti-COX-2 antibody (Transduction Laboratories) was used to detect COX-2 protein. Immunoblotting with antichicken ␣-tubulin antibody (Sigma) was performed for comparative purposes.
Electrophoretic Mobility Shift Assay-Nuclear protein extraction from HUVEC and gel shift assay were performed as described previously (7). The duration of HUVEC exposure to hypoxia was 2 h. The DNA probes (20,000 cpm) labeled by [␥-32 P]ATP and 1.0 g of bulk carrier poly(dI-dC) DNA were incubated with 3 g of nuclear protein extract. An oligonucleotide CX9 corresponding to the Sp1-binding site in the COX-2 promoter (5Ј-GGAGAGGGAGGGATCAGACA-3Ј) and a consensus Sp1 (Promega) oligonucleotide c-Sp1 (5Ј-ATTCGATCGGGG-CGGGGCGAGC-3Ј) were used as probes and competitors to perform the EMSA. A CX9 mutant CX9M (5Ј-GGAGAGAAAGGGATCAGACA-3Ј, mutated bases are italicized and bold) was used to confirm the specific Sp1 binding. Anti-Sp1, anti-Sp3, anti-NF-B p65 (Santa Cruz Biotechnology), and anti-␣-tubulin antibodies were used to perform supershift. Briefly, the antibodies were incubated with the nuclear protein for 30 min at room temperature. The probes were then added into the mixture and incubated for another 30 min at room temperature. Blocking with cold DNA or substitution of 5.0 g of bovine serum albumin for nuclear protein was performed as a control. The protein-DNA product was run in a 5% nondenaturing polyacrylamide gel and autoradiographed.
Transfection-We performed transfection of human vascular endothelial cells in 35-mm dishes at 60 -80% confluency in 21% O 2 , 5% CO 2 , at 37°C. For each dish, 12 l of Lipofectin reagent (Life Technologies, Inc.) and 88 l of Opti-MEM I medium (Life Technologies, Inc.) were mixed and incubated at room temperature for 45 min. The mammalian expression plasmid pCMVSp1 or pCMVSp3 was used to perform cotransfection with 3 g of pWT or pSPM. pSV2PAP, containing the alkaline phosphatase reporter gene downstream from an SV40 promoter, was included in each transfection to control the transfection efficiency. The Lipofectin and DNA solution were mixed with another 800 l of Opti-MEM I medium, overlaid on the cells, and incubated at 37°C for 8 h. The mixture was then replaced with the original medium. After recovering overnight, HUVEC cells were treated with hypoxia or continued normoxia for 24 h and lysed with 200 l of Reporter lysis buffer (Promega). Luciferase activity was then determined in cell lysates using a Turner Designs luminometer. Alkaline phosphatase activity was determined as described previously (7). The results of luciferase (measured in light units) and alkaline phosphatase (OD) activity was reported as a ratio, thereby correcting for minor discrepancies in transfection efficiency. The Drosophila SL2 cells were transfected using calcium phosphate precipitation as described previously (18). Cells were transfected at 50 -60% confluence in 35-mm dishes. For each dish, 100 l of 0.25 M CaCl 2 containing the DNA was added dropwise to 100 l of 2ϫ HBS (274 mM NaCl, 9.4 mM KCl, 2.8 mm Na 2 HPO 4 , 11 mM dextrose, 42 mM Hepes, final pH 7.1) and incubated at room temperature for an hour. The Drosophila expression plasmid pPacSp1 or pPacSP3, both of which were under control of the Drosophila actin promoter, was used to co-transfect with pWT. The total amounts of DNA were kept constant with the PUC19 plasmid. The pPac␤-gal (200 ng) plasmid was included in each transfection to control the transfection efficiency. The transfection solution was added dropwise into culture dish medium and incubated for 42 h. Transfected cells were subsequently harvested by pipetting the medium across the dish several times, washed twice with phosphate-buffered saline (PBS, pH 7.4), and lysed by adding 200 l of lysis buffer. Cell lysates were used to perform luciferase assay and ␤-galactosidase assay using commercially available reagents (Promega). The Drosophila results were expressed as a ratio of luciferase activity to ␤-galactosidase activity.

Sp1/Sp3
Specifically Interacts with the Ϫ250/Ϫ231 Sp1 Site in the cox-2 Promoter-Several potential cis-acting elements have been found within the promoter region of cox-2. Our previous transfection analysis using a set of deletion constructs revealed that the NF-B-3Ј element (Ϫ223 to Ϫ214) is critical to the hypoxia-mediated cox-2 induction. Furthermore, when the promoter region is extended from Ϫ225 to Ϫ245 base pairs, reporter gene promotion was increased (7), indicating that other elements in addition to NF-B-3Ј may be involved in regulation of cox-2 transcription by hypoxia. Sequence analysis of the cox-2 promoter revealed an Sp1 consensus element (Ϫ245 to Ϫ240) just upstream from the NF-B-3Ј element that we previously demonstrated to enhance cox-2 transcription in hypoxia (Fig. 1).
To determine if Sp1/Sp3 interacts with this Sp1 consensus element in the cox-2 promoter, we performed electrophoretic mobility shift assays (EMSA) with HUVEC nuclear protein using the oligonucleotide CX9 containing the Sp1 consensus element in the cox-2 promoter as a probe. Probes were described under "Experimental Procedures." As shown in Fig. 2, two major DNA-protein complexes were observed. Hypoxia enhanced the binding intensity. For comparison, a consensus Sp1 oligonucleotide c-Sp1 was also used as a probe to perform EMSA. The binding profile was identical to that observed with CX9. The DNA-protein complexes were competed away with 100-fold excess of unlabeled CX9 or c-Sp1. Antibodies specific for Sp1 and Sp3 supershifted the DNA-protein complexes. Sp1 antibody partially eliminated the upper complex. Sp3 antibody almost completely eliminated the lower complex and dimin-ished the upper complex. As controls, the nonspecific NF-B and ␣-tubulin antibodies were used to perform supershift. None of them had an effect on the DNA-protein complexes, indicating the binding is Sp1/Sp3-specific. The binding specificity was further confirmed by a mutation experiment. We mutated the Sp1 site of CX9 by only two bases and created CX9M. No DNA-protein complex was observed when CX9M was used as a probe, suggesting that the intact Sp1 site is required for interaction of Sp1/Sp3 with the cox-2 promoter.
Hypoxia Increases the Nuclear Localization of Sp1 and Sp1/ Sp3 Ratio-To determine how Sp1 and Sp3 changed in response to hypoxia, Western blotting was performed with normoxic and hypoxic HUVEC protein extracts. As shown in Fig.  3, Sp1 antibody recognized a doublet band migrating at ϳ100 kDa. Sp3 antibody recognized two doublet bands migrating at ϳ110 and 80 kDa, respectively. Sp1 content was increased approximately 2-fold after 2 h hypoxia in HUVEC nuclear protein (but not cytoplasmic, data not shown), suggesting that hypoxia increases the nuclear localization of Sp1. Meanwhile, Sp3 content was not changed by hypoxia. Taken together, hypoxia resulted in an elevated Sp1/Sp3 ratio.
Sp1/Sp3 Regulates cox-2 Promoter Activity in Drosophila SL2 Cells and HUVEC-To determine directly the role of Sp1 and Sp3 on regulation of cox-2 promoter activity, a Drosophila Schneider SL2 cell line that does not express endogenous Sp1 and Sp3 was used to perform co-transfection experiments. The wild type cox-2 promoter luciferase reporter construct pWT was transfected into SL2 cells with the expression vectors pPacSp1 or pPacSp3. As shown in Fig. 4, overexpression of pPacSp1 (over the range 0 -100 ng/well) caused a dose-dependent increase of cox-2 promoter activity as measured by luciferase reporter gene expression, with a maximal 16-fold increase. Overexpression of pPacSp3 also increased the cox-2 promoter activity, with a maximal 3-fold increase.
Similar dose-response co-transfection experiments were also performed in HUVECs except that mammalian expression vectors pCMVSp1 and pCMVSp3 were used instead of pPacSp1 and pPacSp3. As illustrated in Fig. 5, the promoter activity of pWT was enhanced by increasing the amount of pCMVSp1 (over the range 0 -1000 ng/well) with a maximal 5-fold increase but not by pCMVSp3. Increasing amount of pCMVSp3 (over the range 0 -1000 ng/well) actually caused a decrease of the pro- moter activity. Sp1 also increased the promoter activity of pSPM (an Sp1 mutant of cox-2 promoter luciferase reporter construct in which the Sp1 site was mutated by two bases as in EMSA) but to a much less extent than that of pWT. These results indicate that Sp1 specifically activates cox-2 promoter activity in mammalian cells as well as in SL2 cells. Sp3 appears to be a bi-functional regulator of the cox-2 promoter, a repressor in HUVEC and an activator in SL2 cells.
The Sp1 Site Is Required for cox-2 Induction by Hypoxia-To determine if Sp1-site mutation affects hypoxia-mediated cox-2 promoter induction, either the wild type construct pWT or Sp1 mutant construct pSPM was transfected into HUVECs and treated with 24-h hypoxia. Compared with normoxia control, hypoxia significantly increased the cox-2 promoter activity of pWT by 2.2-fold but not that of pSPM (Fig. 6). This suggests that the Sp1-binding site in the cox-2 promoter region is critical to the regulation of cox-2 expression by hypoxia.
Sp1 Enhances cox-2 Promoter Activity in Hypoxic HU-VECs-To determine the effect of Sp1 and Sp3 on cox-2 promoter activity in hypoxic HUVEC, 300 ng of pCMVSp1 or pCMVSp3 along with 3 g of pWT were co-transfected into HUVECs. After 24 h normoxia or hypoxia treatment, cells were harvested and assayed for promoter activity. A blank vector pCMV5 (19) was added as a control and to keep the total amount of transfected DNA constant. Compared with base line (co-transfection with blank vector under normoxia), overexpression of sp1 synergistically increased the cox-2 promoter activity with hypoxia. In contrast, overexpression of sp3 not only caused a lower cox-2 promoter activity than that of base line but also decreased the transactivation of the cox-2 promoter by hypoxia (Fig. 7). It appears that Sp3 represses cox-2 gene expression under both normoxia and hypoxia in HUVECs.
Sp1 Increases COX-2 Protein in Hypoxic HUVECs-To determine the effect of overexpression of sp1 on COX-2 protein expression in hypoxic HUVECs, 300 ng of expression plasmid pCMV-Sp1 was transfected into HUVECs. In the control experiment, 300 ng of blank vector pCMV5 was transfected into HUVECs as we found DNA transfection itself resulted in slightly less COX-2 production. After 24 h normoxia or hypoxia treatment, cell lysates were collected and employed for Western blot analysis using COX-2 antibody. As demonstrated in Fig. 8, Sp1 progressively increased COX-2 immunoreactive protein in hypoxic HUVECs. We did not demonstrate a clear reduction of COX-2 protein by Sp3 (data not presented) as there was very little COX-2 protein production at base-line normoxia. DISCUSSION We previously demonstrated that cox-2 is induced by hypoxia via the NF-B p65 transcription factor in human vascular endothelium, leading to increased production of PGE 2 , and that the expression of the high mobility group protein HMG I(Y), known to be associated with NF-B-mediated transactivation, is increased by hypoxia, facilitating transactivation of the cox-2 promoter. In this study, we demonstrated that Sp1 (and possibly Sp3) also participates in the regulation of cox-2 gene expression by hypoxia. Sp1 content was increased by hypoxia in HUVEC nuclear (but not cytoplasmic) protein, whereas Sp3 was not changed. Both Sp1 and Sp3 bound specifically to a Sp1 consensus sequence just upstream of the NF-B-3Ј element that we previously demonstrated to enhance the COX-2 transcription in hypoxia. Overexpression of sp1 synergistically increased cox-2 promoter activity and protein expression with hypoxia. We conclude that hypoxia increases the nuclear localization of the Sp1 and thereby activates expression of cox-2 in vascular endothelium.
The Sp family includes Sp1, Sp2, Sp3, and Sp4. All four members contain highly conserved DNA-binding zinc finger domains close to the C terminus and glutamine-rich and serine/ threonine-rich domains in the N-terminal region. In most promoters, Sp1 and Sp3 recognize the classical Sp1 consensus element with comparable affinity and specificity (20). In our study, both Sp1 and Sp3 bind to the same Sp1 consensus sequence (Ϫ250 to Ϫ231) in the COX-2 promoter. However, they contribute differently to cox-2 promoter activity. In HU-VEC, Sp1 not only activated cox-2 promoter in a dose-dependent manner but also synergistically enhanced the promoter with hypoxia. In contrast, Sp3 acted as a repressor in HUVECs. Co-transfection of a higher dose of Sp3 resulted in decreased cox-2 promoter activity. It has been reported elsewhere that down-regulation of sp3 in hypoxic C 2 C 12 myocytes removes the associated transcriptional repression of muscle specific pyruvate kinase-M and ␤-enolase, thereby enabling expression of these glycolytic enzyme genes (21). In our experiment, hypoxia activated the COX-2 promoter by increasing the activator Sp1 and not changing the repressor Sp3. In both cases, the Sp1/Sp3 ratio is elevated. It appears that an elevated Sp1/Sp3 ratio is critical to the gene activation involved with Sp1 and Sp3 as both compete for the same binding site. Both Sp1 and Sp3 are expressed at high levels in endothelial cells, and the Sp1/Sp3 ratio in these cells is much higher than that in non-endothelial cells (22). In Sp-deficient Drosophila SL2 cells, we noticed that Sp3 acted as a weak activator of COX-2 promoter instead of a repressor. This is consistent with the concept that Sp3 is a bi-functional regulator (repressor/activator) depending on the promoter or cellular context, as reported previously (20,23). An inhibitory domain has been identified at the 5Ј end of the zinc finger DNA-binding domain of Sp3; our findings support previous data indicating that the inhibitory domain of Sp3 functions only in mammalian cells (24).
HIE-1 is not found adjacent to cox-2 (15,25). Given the sum of our published data, it would appear that there are hypoxiainducible enhancers other than HIE-1 that regulate cox-2 transcription. Recently an endothelial PAS domain protein was identified in embryonic mouse endothelium which appears to be a new basic helix-loop-helix/PAS domain transcription factor (26). This protein can induce transcription of the gene for endothelial tyrosine kinase Tie-2, and its activity is stimulated by hypoxia. It shares a 48% sequence identity with HIF-1␣. The HIF-1␣-like factor (27) and MOP-2 (28) are among other candidate hypoxia-inducible factors that regulate transcription and have been partially characterized. ORP150, an oxygenregulated protein, appears to be regulated by hypoxia and is found in atherosclerotic blood vessels (29). Our studies show that the Sp1 element just upstream the NF-B-3Ј element is critical to the regulation of cox-2 promoter activity by Sp1/Sp3 and hypoxia. The interaction between this element and nuclear protein is Sp1/Sp3-specific. Both Sp1 and Sp3 antibodies supershifted the DNA-protein complex. This element is distinct from the adjacent NF-B-3Ј element as NF-B p65 antibody has no effect on the DNA-protein complex. Sp1-mediated induction of the COX-2 promoter reflects interaction with the general transcription machinery, including the TATA box-binding protein TBP (30) and components of the TFIID complex (31)(32)(33). Recently it was noted that a transcriptional complex cofactor (CRSP) is required for Sp1-mediated transcription (34). Sp1 has been found to cooperate with NF-B during transcription of the human immunodeficiency virus, type I promoter and to provide transcriptional specificity in a number of contexts, as reviewed by Perkins (35). Sp1 can induce a directional bend in DNA toward the major groove upon binding to the consensus GC box (36). Sp1 can self-associate and loop out the intervening DNA to exert its transcription synergism (37). HMG I(Y) is also capable of changing the DNA conformation and regulating the chromatin structure by binding to AT-rich sequence via the minor groove of DNA (38). Therefore, it appears that a multiple protein-DNA transcription complex is involved in hypoxia-mediated cox-2 activation. Further analysis will be required to determine how Sp1 recruits other transcription factors, co-activators, and general transcription machinery to modulate induction of cox-2 promoter by hypoxia.
Sp1 has been previously associated with transcriptional regulation in cardiovascular pathophysiology. Sp1 is essential to activate the human endothelial nitric-oxide synthase promoter (39). The Sp1-binding site of the tumor necrosis factor-␣ promoter has been reported to function as a nitric oxide response element (40). Nitric oxide and tumor necrosis factor-␣ have been implicated in the pathophysiology of chronic heart failure (41,42). Ping et al. (43) reported that Sp1 binding is critical for activation by tumor necrosis factor-␣ of the monocyte chemoattractant protein-1 gene, expressing a chemokine associated with atherosclerosis. Sp1/Sp3 modulates the promoter activity of vascular endothelial growth factor receptor, a mediator of endothelial cell growth and vascular development (22). Sp1/ Sp3 also regulates the transcription of human thrombin receptor (44).
Abdominal aortic aneurysms are believed to be hypoxic based on markers visualized by immunofluorescence (5). Hypoxia induces the p53 protein associated with tumor suppression (45). Apoptosis, or programmed cell death, can be triggered by hypoxia (46). Vascular endothelial cells are stimulated by hypoxia to release mediators of inflammatory signals such as cytokines (47), including interleukin-6 (48) and interleukin-8 (49). The PGE 2 found in abdominal aortic aneurysms (6) is associated with the presence of secondary mediators of inflammation and decreased cell proliferation and deterioration of the vascular wall (50). PGE 2 can also experimentally inhibit vascular smooth muscle cell proliferation (51). COX-2 is found in atherosclerotic lesions and those associated with transplant allograft vasculopathy (52), as well as in cerebrovascular hy-poxia (53). We have published data suggesting that COX-2 enzymatically facilitates endothelial production of PGE 2 in hypoxia (8). Both COX-2 expression and NF-B activation appear in heart failure (4). We believe that NF-B-mediated transcription in hypoxia and other forms of oxidative stress will eventually be shown to be an important pathway in the pro-inflammatory environment that leads to modulation of vascular endothelial and smooth muscle cellular proliferation.
Epidemiologic data suggest that patients with elevated plasma C-reactive protein, a marker for systemic inflammation, are the ones most likely to benefit from aspirin as a preventive therapy for myocardial infarction and stroke (54). We may hope in the not-too-distant future to learn whether COX-2 is a "friend or foe" (55) in cardiovascular disease. This may lead to development of a "better aspirin" (56) that will be useful in cardiovascular disease by learning how to control the regulation of COX-2 in disease states such as those associated with tissue hypoxia. The regulatory role of Sp1 in the transactivation of COX-2 by hypoxia in this cell culture model of human vascular endothelium suggests a link between cellular hypoxia and the mediators of inflammatory responses associated with aortic aneurysm and heart failure.