Superoxide Fluxes Limit Nitric Oxide-induced Signaling*

Independently, superoxide (\batchmode \documentclass[fleqn,10pt,legalpaper]{article} \usepackage{amssymb} \usepackage{amsfonts} \usepackage{amsmath} \pagestyle{empty} \begin{document} \(\mathrm{O}_{2}^{-}\) \end{document}) and nitric oxide (NO) are biologically important signaling molecules. When co-generated, these radicals react rapidly to form powerful oxidizing and nitrating intermediates. Although this reaction was once thought to be solely cytotoxic, herein we demonstrate using MCF7, macrophage, and endothelial cells that when nanomolar levels of NO and \batchmode \documentclass[fleqn,10pt,legalpaper]{article} \usepackage{amssymb} \usepackage{amsfonts} \usepackage{amsmath} \pagestyle{empty} \begin{document} \(\mathrm{O}_{2}^{-}\) \end{document} were produced concomitantly, the effective NO concentration was established by the relative fluxes of these two radicals. Differential regulation of sGC, pERK, HIF-1α, and p53 were used as biological dosimeters for NO concentration. Introduction of intracellular- or extracellular-generated \batchmode \documentclass[fleqn,10pt,legalpaper]{article} \usepackage{amssymb} \usepackage{amsfonts} \usepackage{amsmath} \pagestyle{empty} \begin{document} \(\mathrm{O}_{2}^{-}\) \end{document} during NO generation resulted in a concomitant increase in oxidative intermediates with a decrease in steady-state NO concentrations and a proportional reduction in the levels of sGC, ERK, HIF-1α, and p53 regulation. NO responses were restored by addition of SOD. The intermediates formed from the reactions of NO with \batchmode \documentclass[fleqn,10pt,legalpaper]{article} \usepackage{amssymb} \usepackage{amsfonts} \usepackage{amsmath} \pagestyle{empty} \begin{document} \(\mathrm{O}_{2}^{-}\) \end{document} were non-toxic, did not form 3-nitrotyrosine, nor did they elicit any signal transduction responses. H2O2 in bolus or generated from the dismutation of \batchmode \documentclass[fleqn,10pt,legalpaper]{article} \usepackage{amssymb} \usepackage{amsfonts} \usepackage{amsmath} \pagestyle{empty} \begin{document} \(\mathrm{O}_{2}^{-}\) \end{document} by SOD, was cytotoxic at high concentrations and activated p53 independent of NO. This effect was completely inhibited by catalase, suppressed by NO, and exacerbated by intracellular catalase inhibition. We conclude that the reaction of \batchmode \documentclass[fleqn,10pt,legalpaper]{article} \usepackage{amssymb} \usepackage{amsfonts} \usepackage{amsmath} \pagestyle{empty} \begin{document} \(\mathrm{O}_{2}^{-}\) \end{document} with NO is an important regulatory mechanism, which modulates signaling pathways by limiting steady-state levels of NO and preventing H2O2 formation from \batchmode \documentclass[fleqn,10pt,legalpaper]{article} \usepackage{amssymb} \usepackage{amsfonts} \usepackage{amsmath} \pagestyle{empty} \begin{document} \(\mathrm{O}_{2}^{-}\) \end{document}.

Soluble guanylyl cyclase (sGC), 3 extracellular signal-regulated kinases (ERK), hypoxia-inducible factor 1␣ (HIF-1␣), and p53 are post-translationally regulated by nitric oxide (NO). For example, HIF-1␣, which is normally limited by proteasome degradation, is stabilized in the presence of NO (1). This leads to translocation to the nucleus, where it activates the transcription of various hypoxia survival genes (2). Similarly, p53 becomes phosphorylated in response to various cellular stresses, including NO metabolites and reactive oxygen species (ROS). p53 phosphorylation can result in cell cycle arrest, which in some cases leads to apoptosis (3). Activation of HIF-1␣ or p53 therefore has opposing outcomes with respect to cell survival. We have shown previously how distinct threshold NO concentrations are necessary to elicit post-translational regulation of both p53 and HIF-1␣. High levels of NO (Ն300 nM) cause p53 phosphorylation, whereas intermediate levels (50 -300 nM) induce HIF-1␣ accumulation (1,4). ERK and sGC are similarly activated by discrete low (10 -30 nM) concentrations of NO (5).
It has been proposed that one of the primary reactions of NO in vivo is its reaction with superoxide (O 2 Ϫ ) (6). This diffusioncontrolled reaction between NO and O 2 Ϫ produces a variety of reactive intermediates that can nitrate, nitrosate, and oxidize many biologically important targets (7)(8)(9)(10). Nitrotyrosine formed from this reaction has been associated with functional aberrations in macromolecules. These mechanisms may be important in the etiology of many disease states leading to tissue injury and cell death.
In contrast, a number of studies show that NO is a powerful antioxidant having the ability to abate ROS-mediated oxidative stress (11). The numerous and diametrically opposing outcomes attributed to the reaction of NO with O 2 Ϫ have led to much controversy concerning the role of these radicals in physiology and pathophysiology (12)(13)(14)(15).
Superoxide dismutase (SOD) reacts with O 2 Ϫ in a near diffusion-controlled manner (16). Therefore, cells and tissues with high relative concentrations of this enzyme limit the reaction of O 2 Ϫ with NO. This potentially increases local NO concentrations. Reports have shown that in the presence of O 2 Ϫ , NO-mediated post-translational regulation is abolished and that this can be reversed by the presence of SOD (17).
In cancer biology, the actions of nitric oxide have been shown to have both positive and detrimental effects. Reactive nitrogen oxide species have been implicated in both the initiation and promotion of carcinogenesis, yet these same chemical species have been linked to tumor growth suppression (18). Nitric oxide can have beneficial as well as deleterious effects on tumor growth (19,20). Recent studies have indicated that these contradictory outcomes are linked to different susceptibilities in genotoxicity, cell death, angiogenesis, and metastasis as well as the p53 status of the tumor.
Clinically it has been noted that tumors expressing elevated iNOS and nitrotyrosine indicate a poor prognosis (21). On the other hand, studies using a iNOS transfection xenograft nude mouse model demonstrated an inhibition of tumor growth in a NO concentration-dependent manner (22). Tumor cells transfected with iNOS also resulted in decreased tumorigenesis, which was dependent on both the level of iNOS and the p53 status of the tumors (20,23,24). Other studies have demonstrated NOS expression was significantly higher in both primary tumors and lymph nodes than in normal gastric mucosa (25,26). These and other examples illustrate the dichotomous nature of NO in tumor biology.
We and others (4,27) have shown that part of these discrepancies could be explained by different concentration dependence of specific signaling targets of NO. Higher NO concentrations activate p53 (Ͼ300 nM), while pro-growth mechanisms such as pERK, HIF-1␣, and cGC, are activated at lower nanomolar levels (1,4,7,28,29). Similarly, in endothelial cells, low NO fluxes (Յ1 nM) lead to increased proliferation and a downregulation of the anti-angiogenic and tumor suppressor thrombospondin TSP-1 but higher doses induce MKP1 and phosphorylation of p53 (5). Based on these differences in sensitivities of specific signal transduction responses to NO, it follows that any situation resulting in changes in the steady-state NO concentration, such as its reaction with O 2 Ϫ , would have important implications toward cellular behavior.
Although the simple concentration-dependent effects of NO are not the only explanation, the interaction of NO with ROS may provide important clues as to mechanisms of specific in vivo pathways. Here we have examined the effects of NO-mediated signaling in relation to superoxide and hydrogen peroxide. We demonstrate that the primary consequence of O 2 Ϫ generation concomitant with NO production is not the toxicity associated with the formation of higher nitrogen oxides, but rather the resultant phenotypic cellular changes that occur because of limiting the bioavailability of NO and H 2 O 2 .
Measurement of steady-state concentrations of NO produced from Sper/NO during cell culture treatments was accomplished using a Seivers (Boulder, CO) NO gas analyzer. Aliquots of media (100 l) from Sper/NO-treated cells were immediately injected into the reaction chamber containing 0.5 mM NaOH to stop NO donor decomposition and under anaerobic conditions (purged with helium) to eliminate NO autooxidation. Steady-state molar NO concentrations were calculated from the absolute amounts of NO detected.
SOD Activity Assay-SOD activity was measured by the method of Spitz and Oberley (30). This assay is based upon the reduction of NBT to blue formazan by O 2 Ϫ . One unit of SOD activity was defined as the amount required to yield 50% of sample maximum inhibition of NBT reduction by O 2 Ϫ . The activity of the SOD we used was 178.6 units/g indicating Ͼ90% functional SOD. H 2 O 2 was measured by ferrous iron oxidation in the presence of xylenol orange as previously described (31).
Oxidation Assay-The strong oxidants generated from the NO/O 2 Ϫ reaction were measured by formation of the fluorescent compound rhodamine (RH) via two-electron oxidation of DHR (32). Immediately prior to use, stock solutions of 50 mM DHR (10 mg) were prepared in DMF (0.6 ml) and diluted 1000fold into cell culture plates. After various concentrations of XO and or Sper/NO were added to the cultures, the reactions were allowed to proceed for 2 h at 37°C (5% CO 2 95%, air incubator). Fluorescence was measured at 570 nm following excitation at 500 nm. Measurements were made in the 6-well plates for total oxidation, or the media was removed and measured separately for extracellular oxidation, and PBS was replaced onto the cells (washed ϫ3) and measured to determine intracellular oxidation.
Cell Culture-MCF7 human breast carcinoma cells (ATCC) were grown to 85% confluency in 6-well culture plates with RPMI 1640 medium (Invitrogen) containing 10% fetal bovine serum (Hyclone, Logan, UT). Prior to treatments, the cells were incubated in 2 ml of serum-free RPMI 1640 medium overnight. All treatments were conducted under these culture conditions.
Co-culture experiments: ANA-1 murine macrophages were activated with interferon-␥ (10 ng/ml) for 16 h followed by LPS (10 ng/ml) for 4 h to induce iNOS expression. The cells were trypsinized, counted, and seeded on top of MCF-7 cells at various densities in serum-free media supplemented with either L-arginine (1 mM) or aminoguanidine (5 mM). ANA-1 cells are HIF-1␣ and p53-null and therefore do not interfere with the measurement of these proteins in MCF7 cells. ANA-1 cells only loosely adhers to the culture flask and/or MCF-7 cells during the treatment period and are easily removed by gentle washing of the MCF7 cells with PBS at the termination of the experiment.
Human Umbilical Vein Endothelial Cells (HUVECs, Cambrex, East Rutherford, NJ) were cultured in EGM media supplemented with 5% fetal bovine serum, epidermal growth factor, bovine brain extract, cortisone, gentamycin, penicillin, and streptomycin and maintained at 37°C in an atmosphere of 5% CO 2 and room air. Prior to treatment, cells were trypsinized and plated at a density of 1 ϫ 10 6 cells per 100-mm tissue culture dish and grown for 72 h. The cells were then washed with PBS and incubated in phenol red-free EGM media containing 5% fetal bovine serum, bovine brain extract, heparin, and penicillin.
Western Blot-Protein cell extracts were made by washing cells in cold PBS, scraping plates, centrifuging, and resuspending in lysis buffer (1% Nonidet P-40, 0.5% sodium deoxycholate, 0.1% sodium dodecyl sulfate, and protease inhibitor mixture; Calbiochem, San Diego, CA). Following 30 min of incubation on ice, the samples were centrifuged at 14,000 ϫ g, and the supernatant protein concentration was determined by the bicinchoninic acid method (Pierce). Protein samples of equal amounts were subjected to PAGE on 10% Tris-glycine acrylamide gels (Novex-Invitrogen). Following transfer to polyvinylidene difluoride Immunolon P membranes (Millipore, Bedford, MA), samples were probed with rabbit polyclonal or mouse monoclonal antibodies (HIF-1␣ (Transduction Laboratories), ERK, p53 P-Ser-15, (Cell Signaling), HPRT (Santa Cruz Biotechnology). Bands were visualized with horseradish peroxidase-conjugated secondary antibodies (1:2,000 -10,000; Sigma) and chemiluminescent substrate (Pierce). HPRT protein-loading controls were run for each gel. Gels images were scanned using an AGFA DuoScan hiD scanner (Wilmington, MA) or Alpha Innotech Imaging system, and relevant bands were cropped to size using Photoshop 7.0 with no further manipulation. Densitometry was measured using AlphaEase FC Stand Alone Software. Figures are representative of n Ն 3 individual experiments.
Clonogenic Assay-Cell survival/proliferation was measured by clonogenic assay with plating efficiency ranging between 85 and 95% as previously described (33). Briefly, stock cultures of each cell type were plated (7 ϫ 10 6 cells/100-cm 2 dish) and incubated for 24 h prior to treatment. After treatments, cells were washed, trypsinized, counted, and plated in triplicate for macroscopic colony formation. Following a 10 -14-day incubation period, colonies were fixed, stained, and counted. Colonies containing Ͼ50 cells were scored.
cGMP Measurement-cGMP was measured using the enzyme immunoassay (EIA) Biotrak system (Amersham Biosciences). Cells were grown in 96-well plates, serum-starved overnight and subjected to various treatment conditions. The cells were washed, lysed, and the supernatant transferred to the EIA plate for cGMP measurement.

Threshold NO Concentrations Regulate HIF-1␣ and p53-It
has been demonstrated that NO causes p53 serine 15 phosphorylation and HIF-1␣ accumulation in a concentration-dependent manner (1,4,28). When increasing concentrations of the NO donor Sper/NO (25-100 M) were added to MCF7 cells, both p53 P-Ser-15 and HIF-1␣ increased (Fig. 1.) These changes were compared with the mean steady-state NO concentrations measured in the media of the Sper/NO-treated cells (Fig. 1). The concentration of NO necessary for HIF-1␣ stabilization and p53 P-Ser-15 correlated to the steady-state NO concentrations as previously reported (4).  that ROS differentially modulate the effects of NO on these two proteins.

HIF-1␣ and p53 Regulation by Other Radical/Oxidant
To determine the direct effects of ROS on p53 and HIF-1␣ signaling, MCF7 cells were treated only with XO plus HX. p53 phosphorylation (P-Ser-15) was detected whereas HIF-1␣ was not (lane 3  Fig. 4, A and B demonstrates that when 100 M bolus of H 2 O 2 was added to MCF7 cells in culture, it was metabolized in a first-order manner with a rate of k obs Ϸ Ϫ0.0649 min Ϫ1 , Ϫ1.73 ϫ 10 Ϫ7 min Ϫ1 cell Ϫ1 . Fig. 4C shows that this rate was proportional to the two-dimensional density of the MCF7 cell monolayer. When aminotriazole, a catalase inhibitor, was added, the rate of H 2 O 2 metabolism decreased (Fig. 4C) similar to what has been reported previously for other cell types (34).
Unlike NO, it appears that cell membranes pose important limiting barriers for the partitioning and diffusion of H 2 O 2 (35)(36)(37). To test this, MCF7 cells were harvested and placed in suspension at various concentrations to increase their density. Fig.  3D demonstrates that even in the presence of 1 ϫ 10 7 MCF7 cells/ml, p53 P-Ser-15 was detected after 2 h in the presence of 100 M H 2 O 2 . The half-life of NO at this cell density is calculated to be Ͻ0.5 s (37), far too short to exert an effect on p53 phosphorylation. This demonstrates that the rate of H 2 O 2 metabolism is slow relative to the rate of NO metabolism and therefore the exposure time of cellular targets to H 2 O 2 is greater than that for NO.  It has been shown previously that H 2 O 2 is primarily metabolized in cells by catalase and glutathione peroxidase (34). Pharmacologic catalase inhibition should therefore further extend the half-life (exposure time) of H 2 O 2 and increase its biological effects (Fig. 4C). When MCF7 cells were incubated for 3 h with increasing H 2 O 2 concentrations, a dose-dependent increase in p53 P-Ser-15 was observed (Figs. 3 and 5). In the presence of aminotriazole, a potent catalase inhibitor, the minimum dose of H 2 O 2 necessary to induce p53 phosphorylation was decreased and the response at higher concentrations was markedly more robust (Fig. 5).
Because both NO and H 2 O 2 independently induce p53 phosphorylation, we expected the result to be additive when both molecules were present simultaneously. Remarkably, concurrent NO and H 2 O 2 exposure demonstrated the opposite effect on p53 phosphorylation (Fig. 5)   Ϫ /min), increasing the rate of NO production first increased the level of DHR oxidation and then decreased DHR oxidation when NO was in excess. The curve in Fig. 6 represents the sum of both intra-and extracellular DHR oxidation. After washing to remove extracellular DHR, the shape of the intracellular oxidation curve remained the same, and only its magnitude decreased (data not shown), similar to that previously reported (40,41). When SOD was present, the shape of the curve flattened as O 2 Ϫ was converted to H 2 O 2 before it could react with NO. Because oxidant formation could be accurately determined from the co-generation of NO and O 2 Ϫ , the oxidant profile was correlated with the post-translational regulation of p53 and HIF-1␣ under these conditions. Extracellular O 2 Ϫ Generation Restricts NO Signaling-We treated MCF7 cells with increasing concentrations of the NO donor Sper/NO (1-100 M) for 2 h. Under these conditions, we were able to simultaneously measure both DHR oxidation and the steady-state NO concentration in the media. Fig. 7A demonstrates the dose-dependent increase in HIF-1␣ and p53 P-Ser-15 as the concentration of NO increased, with minimal oxidation of DHR. Under the same conditions of NO exposure as Fig. 7A, introduction of O 2 Ϫ at a constant rate (HX/XO) resulted in minimal relative changes in HIF-1␣ and p53 P-Ser-15 for every condition in Fig. 7B, unlike the dramatic dose-dependent changes observed for every condition in Fig.  7A. With the introduction of O 2 Ϫ , NO levels fell to below the limit of detection, yet oxidation of DHR greatly increased with its shape resembling the bell-shaped curve of Fig. 6. Surprisingly, despite large increases in the formation of oxidants, the p53 P-Ser-15 response was not increased. When SOD was present during cells treated with NO/O 2 Ϫ (identical to Fig. 7B) the steady-state NO levels were re-established, DHR oxidation was suppressed, and both HIF-1␣ and p53 P-Ser-15 protein levels were nearly restored (Fig. 7C).
The temporal effects of NO/O 2 Ϫ on HIF-1␣ and p53 P-Ser-15 regulation were also examined. When MCF7 cells were treated with NO, followed sequentially 2 h later with O 2 Ϫ , it was found that HIF-1␣ required constant NO exposure, whereas p53 P-Ser-15 stabilization, like with H 2 O 2 , was initiated by NO but remained elevated long after exposure (Fig. S1, supplemental data).

Intracellular O 2 Ϫ Generation Restricts NO Signaling-Having established that extracellularly generated O 2
Ϫ can have substantial effects on the magnitude of post-translational regulation of HIF-1␣ and p53 by NO, the effects of intracellularly produced O 2 Ϫ were examined. Exposing DHR-treated MCF7 cells to NO and DMNQ or menadione, both known generators of intracellular O 2 Ϫ (34), resulted in measurable increases in DHR oxidation (Fig. S2, supplemental data). Fig. 8A shows the regulation of HIF-1␣ and p53 P-Ser-15 in the presence of NO alone or NO ϩ menadione. This figure demonstrates that intracellularly generated O 2 Ϫ shifts the threshold for Sper/NO-mediated HIF-1␣ and p53 Ser-15 regulation toward higher concentrations, analogous to the effect of HX/XO. Thus, higher NO production is required to activate HIF-1␣ and p53 in the presence of intracellular O 2 Ϫ production. As a control, CoCl 2 , a potent and NO-independent activator of HIF-1␣, was used. The increase in HIF-1␣ by CoCl 2 treatment was unaffected by menadione. Taken together, these findings imply that decreases in HIF-1␣ levels in the presence of O 2 Ϫ generated

NO-mediated Guanylyl Cyclase Activation Is Limited by O 2
Ϫ and Enhanced by SOD-Because sGC is one of the most important and sensitive biological targets for NO, it follows that O 2 Ϫ should change the threshold concentration for activation by NO. Intracellular cGMP levels in MCF7 cells were determined after 30 min of exposure to NO. We found that maximal sGC activation required Ն5 M Sper/NO (Fig. 8B). However, the presence of extracellular (HX/XO) or intracellular (DMNQ) O 2 Ϫ resulted in higher concentrations of Sper/NO necessary to activate sGS. In the presence of XO, 10 times higher Sper/NO concentration was required to achieve the same level of sGC activation as without. When SOD was present during NO and extracellular O 2 Ϫ production, the sensitivity of sGC to NO was enhanced. The presence of SOD alone increased the sensitivity of sGC to NO indicating that basally produced O 2 Ϫ is modulating NO-mediated sGC activation. Similarly, when O 2 Ϫ was generated intracellularly, the sensitivity of sGC to NO was lost. Unlike extracellularly produced O 2 Ϫ , when it was produced intracellularly, the addition of SOD had a minimal effect on restoring the sensitivity of sGC to NO (Ͻ10% of maximal cGMP produced). Because DMNQ-generated O 2 Ϫ is unaffected by the addition of SOD, this confirms that O 2 Ϫ is predominantly formed intracellularly and reacts with NO.

O 2
Ϫ Antagonizes Signaling Mediated by Macrophage-derived NO-NO-producing cells are common to many inflammatory and malignant conditions. Various studies have used activated macrophages as a biological source of NO production (4,42), and they were used in this study to examine the effects of O 2 Ϫ generation on NO signaling events. When cytokine (IFN␥/ LPS)-activated NO-producing ANA-1 cells were co-cultured with MCF7 cells at 1:4 and 1:8 ratios, both HIF-1␣ and p53 P-Ser-15 were detected in the MCF7 cells (Fig. 9). The detection of either protein was abolished by aminoguanidine, a potent iNOS inhibitor, indicating that the response resulted from NO generation (data not shown). When O 2 Ϫ was introduced by HX/XO during the co-culture period, the levels of both HIF-1␣ and p53 P-Ser-15 were markedly decreased. This again emphasizes the profound influence O 2 Ϫ has on NO signaling, independent of the source of NO production.

O 2 Ϫ Antagonizes NO Signaling in Endothelial Cells-The effect of NO/O 2
Ϫ on p53 P-Ser-15 and HIF-1␣ was also examined in endothelial cells. Fig. 10 illustrates the dramatic influence O 2 Ϫ had on NO-induced HIF-1␣, and p53 in HUVEC cells. In the presence of NO alone, HIF-1␣ and p53 are relatively sensitive to its effects. Yet during concomitant low fluxes of O 2 Ϫ , the rate of NO production necessary to elicit the same signal transduction responses was greatly increased. These results in HUVECs suggest a general mechanism in response to NO/O 2 Ϫ chemistry. In endothelial cells, NO at high concentration reduces pERK through increasing its corresponding phosphatase MKP-1 (5). The concentration of Sper/NO that activates p53 also decreased pERK. However, in the presence of O 2 Ϫ , pERK was maintained, suggesting that NO levels are critical in  regulating pro-and antigrowth mechanisms in endothelial cells.

NO Protects Cells from O 2
Ϫ /H 2 O 2 -induced Cytotoxicity-Because NO and ROS can generate potentially toxic species, clonogenic assays were performed to determine the potential long term cytotoxic/cytostatic effects of our various treatment paradigms. Fig. 11 summarizes our results. NO alone was nontoxic at 100 M Sper/NO consistent with other reports (43,44). The cytoxicity of H 2 O 2 alone was completely prevented by the addition of NO. Low nanomolar concentrations of intra-or extracellular O 2 Ϫ alone, or in combination with NO at varying ratios, were also non-toxic. These results demonstrate that physiologically relevant concentrations of NO production are protective rather than toxic. Even the formation of strong oxidants, as measured by DHR, from the reaction of NO with O 2 Ϫ had no effect on cell survival. Furthermore, the formation of powerful nitrating intermediates from the reaction of NO with O 2 Ϫ is purported to have major biological and pathological consequences (45). We measured the formation of 3-nitrotyrosine in all of the samples in Fig. 7. We were unable to detect 3-nitrotyrosine by Western blot analysis from total cell lysates in any of the samples (data not shown) consistent with previous reports (12). These findings reconfirm the important antioxidant rather than pro-oxidant properties of NO (43).

DISCUSSION
Herein we report an alternative physiologic function for the reaction between NO and O 2 Ϫ with respect to signal transduction. Much debate and controversy has arisen regarding the physiologic and pathologic consequences of the strong nitrative and oxidative intermediates formed from this reaction. Some previous work has emphasized the potential toxic and deleterious nature of the isolated chemical species formed from the NO/O 2 Ϫ reaction. Our data instead demonstrate that this reaction is not a significant source of toxic species, but rather a mechanism to regulate steady-state NO levels and signal transduction events.
NO-regulated proteins have different sensitivities to steadystate concentrations of NO. Therefore, the rate of O 2 Ϫ production is critical in this paradigm because O 2 Ϫ concentrations determine the steady-state concentration of NO. A greater rate of NO production is necessary in the presence of O 2 Ϫ to achieve the same cellular response as in the absence of O 2 Ϫ . Superoxide generation will therefore alter the signaling response to NO by directly altering its steady-state concentration. Both the chemical data and cellular responses suggest that quantitatively the most biologically significant consequence of this reaction is the change in steady-state NO concentration because of its reaction with O 2 Ϫ . The elegance of this reaction lies in its simplicity and the importance, therefore is not a function of the resultant chemistry from the formation of higher nitrogen oxides, but rather the alterations in cellular phenotype because of the attenuation in bioactive NO levels.
One of the earliest examples of this paradigm was the observation that O 2 Ϫ decreased cGMP levels and inhibited NO-mediated vascular smooth muscle relaxation (46). This eventually led to the identity of NO as endothelial-derived relaxation factor. Recently Kohl et al. (47) have demonstrated HIF-1␣ regulation by NO is determined by intracellularly generated O 2 Ϫ . Conversely, just as O 2 Ϫ abates NO signaling, NO had been shown to influence O 2 Ϫ -mediated signaling as well as. For example: activation of Erg1, ICAM-1, and MCP-1 by O 2 Ϫ in endothelial cells is abated by NO (48,49). In summary, NO and O 2 Ϫ work in concert, each co-regulating the concentration of the other.
High concentrations of NO have been shown to inhibit tumor (50,51) and endothelial cell growth (5) through increases in p53 phosphorylation and MKP-1 activation. Low NO concentrations tend to favor pro-growth pathways that protect against apoptosis, whereas at high relative steady-state NO levels, p53 will be phosphorylated resulting in cell cycle arrest, and potentially apoptosis. Intermediate levels of NO will cause ERK phosphorylation and HIF-1␣ accumulation, which are associated with a mitogenic, angiogenic, growth phenotype. Lower nanomolar steady-state NO levels will activate sGC with the subsequent formation of cGMP. Therefore, high NO levels are generally inhibitory whereas lower levels are stimulatory.  Despite the formation of RNOS, the primary effect of O 2 Ϫ on signaling induced by higher levels of NO is to abate anti-proliferative mechanisms and to enhance pro-growth systems.
An important mechanism described here is that not all proteins have the same response to a given redox stimulus. Susceptible proteins respond to NO and/or oxidative stress either transiently or constitutively. For example, the regulation of HIF-1␣ or cGMP by NO parallels the dose and duration of NO exposure. When NO is removed or falls below the concentration threshold necessary for activation, HIF-1␣ levels decline. However, p53 phosphorylation by NO (Ͼ300 nM for over 2 h) or H 2 O 2 (20 -40 M for Ͻ15 min) will remain elevated long after exposure. This becomes important when considering the temporal effects of O 2 Ϫ exposure on various NO-regulated proteins. P53 P-Ser-15, once activated by NO, will remain elevated even during subsequent O 2 Ϫ exposure. However, HIF-1␣, which requires constant NO exposure, will therefore be sensitive to the temporal and NO-scavenging effects of O 2 Ϫ . This suggests that certain cellular proteins will be more affected than others by acute changes in the local redox environment where NO is suspected to be a major driving force. Redox signaling (p53) induced by stress levels of NO or H 2 O 2 are persistent, but progrowth and anti-apoptotic signals (cGMP, HIF-1␣) are more reversible. This suggests that the timing of NO and O 2 Ϫ generation may be equally important for understanding the mechanisms of redox-mediated signaling in relation to tissue growth and injury.
Numerous studies have shown that NO has powerful antioxidant properties serving to suppress chemical and biological mechanisms associated with oxidative stress. Of all the NO/ROS treatment paradigms explored in this study, H 2 O 2 exposure was most toxic, resulting in only Ϸ10% surviving fraction. Interestingly, although both NO and H 2 O 2 induce p53 P-Ser-15, only H 2 O 2 resulted in cell death, indicating that additional signaling targets determine the cellular consequences of p53 activation. Conversely, O 2 Ϫ or NO alone or in combination at various flux ratios was not toxic despite the formation of strong oxidants. Furthermore, the toxicity of H 2 O 2 was completely prevented if NO was present during treatment and the induction of p53 P-Ser-15 was abolished. These results go against the current dogma regarding free radicals and disease by highlighting the antioxidant and protective properties of NO. The presence of nitrotyrosine, as evidence for the reaction of NO with O 2 Ϫ , has been invoked in the etiology of many disease states, and its formation is often considered to be the precipitating or causative event (45,52,53). We were unable to detect 3-nitrotyrosine after co-generation of various low flux ratios of NO and O 2 Ϫ ; nor did these treatments result in decreases in cell survival or viability. Quantitatively, the effects of 3-nitrotyrosine formed from this reaction are negligible compared with the bio-regulatory aspects.
It is important to note that whereas NO affords protection against acute exposure to toxic levels of O 2 Ϫ /H 2 O 2 , the longterm ramifications of these interactions may be of considerable consequence. Deviations from normal NO signaling in response to O 2 Ϫ production can dramatically alter the long term outcome. When p53-inducing levels of NO are reduced to concentrations only capable of activating cGC and HIF-1␣, the cell is essentially rendered p53-null and it is transformed from an inhibitory to a pro-growth phenotype. It should be emphasized that it is not the toxicity of the intermediates from these reactions that are important in disease progression but the potential long term changes in cellular phenotypes.
There is much evidence demonstrating the roles of NO and O 2 Ϫ in various mechanisms of carcinogenesis and tumor biology. Macrophage-derived NO is essential for the suppression of some tumors (54). Recent studies have shown that there is a higher incidence of lymphomas in p53/NOS-2 double knockout mice (55). Yet there was a significant delay in the onset of tumors in p53 Ϫ/Ϫ mice that were fed TEMPOL, an SOD mimetic (56). These findings suggest a link between p53 regulation and redox status in vivo such that higher levels of NO may be important in suppressing tumorigenesis. Abatement of ROS may also be important. Because p53 is elevated by NO (breast, TK6, Ref. 27, etc.), it stands to reason that limiting oxidative stress in part increases NO. Some tumors have been associated with increased oxidative stress, which may serve to abate NOmediated killing. It has been shown that NOS-2 transfection limits tumor growth and is directly related to NO production (20,24,27). Thus, the antitumor activity of NO correlates to high NO and appears to be opposed by oxidative stress.
Angiogenesis is also necessary for tumor progression. Soluble guanylyl cyclase is activated by low nanomolar levels of NO and appears to play a central role in this process. In the presence of O 2 Ϫ , generated intracellularly or extracellularly, the concentration of NO had to be substantially increased to achieve the same level of cGMP formation as in the absence of O 2 Ϫ (17). Low levels of NO can also stimulate proliferation of endothelial cells and reduce the transcription of antiangiogenic factors such as TSP-1. Although HIF-1␣ and subsequent VEGF production are increased by NO in endothelial cells, higher concentrations of NO antagonize this by increasing p53 phosphorylation and MKP-1 as well (5,51). Again we find that O 2 Ϫ generation inverts the NO-linked phenotype. These are effects that were reversed by the addition of SOD. This demonstrates the profound effect low levels of O 2 Ϫ can have on NO signaling. These data also illustrate the pathological consequences of aberrant O 2 Ϫ generation or the absence of SOD. In the face of excess O 2 Ϫ , the induction of cGMP-mediated signal transduction pathways by moderate levels of NO may be abolished.
In conclusion, these results demonstrate an alternate yet fundamental role of O 2 Ϫ in regulating NO pathophysiology. The primary consequence of O 2 Ϫ production may simply be the result of reducing the bioavailability of NO. Because most NOregulated proteins appear to be exquisitely sensitive to concentration effects, alterations in NO levels as a result of excess O 2 Ϫ may have profound phenotypic effects. The generation of NO and O 2 Ϫ appear to be mostly non-toxic, and in all cases tested NO was protective. These results also demonstrate that not only will the actual radical fluxes effect cellular signaling, but indirectly, so will the presence of antioxidant enzymes (SOD, catalase). These results can be extended to many physiologic and pathologic circumstances, and they become especially important when examining conditions attributed to excess radical production or alterations in antioxidant defense systems.