Cytotoxicity and Site-specific DNA Damage Induced by Nitroxyl Anion (NO−) in the Presence of Hydrogen Peroxide

Nitroxyl anion (NO−), the one-electron reduction product of nitric oxide (NO⋅), is formed under various physiological conditions. We have used four different assays (DNA strand breakage, 8-oxo-deoxyguanosine formation in calf thymus DNA, malondialdehyde generation from 2′-deoxyribose, and analysis of site-specific DNA damage using32P-5′-end-labeled DNA fragments of the human p53tumor suppressor gene and the c-Ha-ras-1 protooncogene) to study the effects of NO− generated from Angeli’s salt on DNA damage. It was found that strong oxidants are generated from NO−, especially in the presence of H2O2 plus Fe(III)-EDTA or Cu(II). NO⋅released from diethylamine-NONOate had no such effect. Distinct effects of hydroxyl radical (HO⋅) scavengers and patterns of site-specific DNA cleavage caused by Angeli’s salt alone or by Angeli’s salt, H2O2 plus metal ion suggest that NO− acts as a reductant to catalyze the formation of the HO⋅ from H2O2 plus Fe(III) and formation of Cu(I)-peroxide complexes with a reactivity similar to HO⋅ from H2O2 and Cu(II). Angeli’s salt and H2O2 exerted synergistically cytotoxic effects to MCF-7 cells, determined by lactate dehydrogenase release assay. Thus NO− may play an important role in the etiology of various pathophysiological conditions such as inflammation and neurodegenerative diseases, especially when H2O2 and transition metallic ions are present.

Excess production of nitric oxide (NO ⅐ ) has been implicated as a cause of diverse pathophysiological conditions such as inflammation, neurodegenerative diseases, cardiovascular disorders, and cancer. These detrimental effects of NO ⅐ have been attributed to reactive nitrogen species such as NOx and peroxynitrite (ONOO Ϫ ), which are formed by the reaction of NO ⅐ with oxygen and superoxide, respectively. Reactive nitrogen species can oxidize, nitrate, and nitrosate biomolecules such as proteins, DNA, and lipids, thus altering their functions. We have recently reported that NO Ϫ , which is the one-electron reduction product of NO ⅐ , can also cause strand breakage and oxidative damage in DNA in vitro (1). We have proposed that a highly toxic hydroxyl radical (HO ⅐ ) generated from the reaction between NO Ϫ and NO ⅐ is responsible for the oxidation reactions (Equations 1 and 2).
(Eq. 1) NO Ϫ has been also reported to be cytotoxic, reducing intracellular glutathione levels and causing DNA strand breakage in cultured cells (2). However, it can also be converted under physiological conditions in vitro, as well as in cells, to NO ⅐ and other reactive oxygen and nitrogen species including superoxide, hydrogen peroxide (H 2 O 2 ), and peroxynitrite (3)(4)(5), and the actual mechanisms and reactive species responsible for the cytotoxic effects of NO Ϫ have not been established.
Three recent publications have suggested that NO ⅐ synthase generates NO Ϫ , which can be then converted to NO ⅐ by superoxide dismutase and other electron acceptors (6 -8). NO Ϫ can also be produced from S-nitrosothiols in the presence of thiols (9 -11). It has been reported that, in the absence of oxygen, nitrosylhemoglobin liberates NO Ϫ in a reaction producing methemoglobin (12). Ferrocytochrome c also reacts with NO ⅐ to form ferricytochrome c and NO Ϫ , which may have implications for inhibition of mitochondrial oxygen consumption by NO ⅐ (13). In our previous reports, NO Ϫ was proposed as one of the possible agents responsible for DNA strand breakage induced by NO ⅐ and catechol-type compounds such as catecholamines, catechol-estrogens, and certain flavonoids (14). NO Ϫ can be formed by one-electron reduction of NO ⅐ by the quinone/hydroquinone redox system in a manner similar to that of the formation of O 2 . from oxygen (14).
In the present study, we have studied the effects of NO Ϫ generated from Angeli's salt (sodium trioxodinitrate, Na 2 N 2 O 3 ) on DNA strand breakage and DNA base modifications in vitro mediated by H 2 O 2 in the presence of metallic ions. At physiological pH, Angeli's salt exists predominantly in the form of the monoanion HN 2 O 3 Ϫ , which decomposes to NO Ϫ and nitrite (NO 2 Ϫ ) (Equation 3) (15). As HNO is a weak acid (pK a ϭ 4.7), NO Ϫ is the predominant form in aqueous solution at neutral pH (Equation 4) (15).
We have found that NO Ϫ generated from Angeli's salt dramatically enhances DNA damage mediated by H 2 O 2 in the presence of the ferric ion (Fe(III))-EDTA or copper ion (Cu(II)), indicating that NO Ϫ acts as an endogenous reductant to catalyze formation of strong oxidants. Furthermore, Angeli's salt and H 2 O 2 cooperatively exerted cytotoxic effects toward human breast cancer cells. We discuss possible implications of our * 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. findings as a cause of diverse pathophysiological conditions mediated by activation or overexpression of NO ⅐ synthase.
Measurement of Malondialdehyde (MDA) Produced from Oxidation of Deoxyribose by Angeli's Salt-MDA formed from the oxidation of 2Ј-deoxyribose was measured as a marker of HO ⅐ generation, according to the method of Hogg et al. (16). The reactions were carried out in 100 mM sodium phosphate buffer, pH 7.4, containing 10 M DTPA, 1 mM 2Ј-deoxyribose, 500 M H 2 O 2 , 50 M either FeCl 3 -EDTA or CuCl 2 , an appropriate amount of HCl to neutralize the NaOH present in the Angeli's salt solution, and 200 M Angeli's salt prepared in 0.01 N NaOH at 37°C (final volume, 1 ml; final pH, ϳ7.5). The MDA content was determined after reaction with 2-thiobarbituric acid using HPLC with a fluorescence detector, as reported previously (1,17). All experiments were carried out in triplicate.
Analysis of 8-oxo-2Ј-Deoxyguanosine (8-oxo-dG) in Calf Thymus DNA Incubated with Angeli's Salt-Angeli's salt prepared in 0.01 N NaOH (0 -5 mM, 100 l) was added to a reaction mixture (final volume, 1 ml) containing 0.1 M sodium phosphate buffer, pH 7.5, calf thymus DNA (1 mg), 10 M DTPA, 500 M H 2 O 2 , either 50 M FeCl 3 -EDTA or CuCl 2 , and an appropriate amount of HCl to neutralize the NaOH present in the Angeli's salt solution (final pH ϳ7.5), and the solution was incubated at 37°C for 30 min. After the reaction, ethanol-precipitated DNA was hydrolyzed enzymatically, and 8-oxo-dG and 2Ј-deoxyguanosine were analyzed by HPLC with a Coulochem II electrochemical detector (ESA Inc., Chelmsford, MA) and a Shimadzu UV spectrophotometer (model SPD-2A), respectively, according to a modification of the method of Yamaguchi et al. (18). All experiments were carried out in duplicate or triplicate.
Induction and Analysis of DNA Single Strand Breaks-The experiments were carried out by incubating plasmid pBR322 DNA (100 ng) at 37°C for 30 min in 100 mM sodium phosphate buffer, pH 7.4, containing 10 M DTPA, 500 M H 2 O 2 , either 50 M FeCl 3 -EDTA or CuCl 2 , 200 M Angeli's salt prepared in 0.01 N NaOH, and an appropriate amount of HCl to neutralize the NaOH present in the Angeli's salt solution (final volume, 10 l; final pH, ϳ7.5). After the reaction, electrophoresis was carried out as described previously (1, 14, 19 -21). The average number of single strand breaks/pBR322 DNA molecule was calculated according to Epe and co-workers (22,23), taking into account that the relaxed form (form II) when stained with ethidium bromide gives a fluorescence intensity 1.4-fold higher than the supercoiled form (form I) and that a relaxation is caused by one single strand break/DNA molecule. All experiments were carried out in triplicate, and statistical significance was calculated using the Student's t test.
Site-specific DNA Damage-32 P-5Ј-end-labeled DNA fragments of the human p53 tumor suppressor gene 2 and the c-Ha-ras-1 protooncogene (25) were prepared as previously reported (26,27). The standard reaction mixture in a microtube (1.5-ml, Eppendorf) contained 100 M Angeli's salt, 20 M CuCl 2 , 20 M H 2 O 2 , 20 M/base of sonicated calf thymus DNA, and a 32 P-5Ј-end-labeled DNA fragment in 200 l of 10 mM sodium phosphate buffer (pH 7.8) containing 5 M DTPA. After incubation at 37°C for 30 min, the DNA fragments were heated at 90°C in 1 M piperidine and treated as described previously (26). The treated DNA fragments were electrophoresed on an 8% polyacrylamide/8 M urea gel, and the autoradiogram was obtained by exposing x-ray film to the gel. The preferred cleavage sites were determined by direct comparison of the positions of the oligonucleotides with those produced by the procedure of Maxam and Gilbert (28) using a DNAsequencing system (LKB 2010 Macrohor). A laser densitometer (LKB 2222 UltroScan XL) was used for the measurement of the relative amounts of oligonucleotides from treated DNA fragments.
Lactate Dehydrogenase (LDH) Cytotoxicity Assay-Human breast cancer cells (MCF-7) were cultured in Dulbecco's modified Eagle's medium, (Life Technologies, Inc.) supplemented with 10% fetal bovine serum, 2 mM glutamine, 100 units/ml penicillin, and 100 g/ml streptomycin at 37°C in a humidified atmosphere containing 10% CO 2 . Cells were seeded into 96-well microculture plates one day before the treatment, at a density of 10 5 cells/well with 100 l of Dulbecco's modified Eagle's medium without phenol red containing 5% fetal bovine serum. Cells were treated with Angeli's salt or DEA-NO prepared in 0.01 N NaOH in combination with H 2 O 2 prepared in water (0 -10 mM). The NaOH present in the Angeli's salt or DEA-NO solutions was neutralized with the same volume of 0.01 N HCl. The LDH assay was performed using the CytoTox Non-Radioactive Cytotoxicity Assay (Promega, Madison, WI). To obtain maximal LDH release, nontreated cells were incubated in the presence of lysis solution (9% Triton X-100) for 45 min at 37°C. After 4.5 h of incubation in the presence of the different products, the plate was centrifuged at 250 ϫ g for 4 min, and 50 l of supernatant was used for the LDH assay. The percentage of cytotoxicity was calculated according to the following equation, where exp ϭ experimental value, max ϭ mean of cell maximum LDH release, back 1 ϭ mean of cell culture background, and back 2 ϭ mean of cell culture background plus lysis solution. All experiments were carried out at least in triplicate, and the results are expressed as means Ϯ S.D.

RESULTS
Four different assays were used to study the effects of Angeli's salt on oxidation reactions mediated by H 2 O 2 in the presence of Fe(III)-EDTA or Cu(II).
MDA Production from 2Ј-Deoxyribose-The first assay was based on oxidation of 2Ј-deoxyribose leading to the formation of MDA, which has been measured as a marker of HO ⅐ generation (29). The formation of MDA in the presence of Angeli's salt, H 2 O 2 , and Fe(III)-EDTA was very rapid and reached a plateau in 10 min, whereas H 2 O 2 and Fe(III)-EDTA alone catalyzed the formation of MDA linearly up to 60 min of incubation (Fig. 1A). Lower concentrations of MDA were formed when the reaction was carried out in the presence of H 2 O 2 plus Cu(II) than with H 2 O 2 plus Fe(III)-EDTA. MDA was also formed dose dependently with different concentrations of Angeli's salt in the presence of H 2 O 2 and metallic ions (Fig. 1B). However, its formation was inhibited by a high concentration (2 mM) of Angeli's salt, especially when the reaction was carried out in the presence of Fe(III)-EDTA.  (Table II). As we previously reported (1), incubation of plasmid pBR322 with Angeli's salt alone formed 52.6% of form II, corresponding to ϳ1.25 single strand breaks/10 4 bp. Increased levels of DNA strand breakage were also observed when plasmid DNA was incubated with metallic ions (Fe(III)-EDTA or Cu(II)) alone or in combination with H 2 O 2 and Fe(III)-EDTA or Cu(II) compared with nontreated plasmid. However, the addition of Angeli's salt dramatically enhanced strand breakage induced by H 2 O 2 plus Fe(III)-EDTA or Cu(II). In particular, when the reaction was carried out in the presence of Angeli's salt, H 2 O 2 , and Cu(II), none of forms I, II, and III were clearly detected, indicating that the DNA was completely fragmented. When the plasmid was incubated with Angeli's salt, H 2 O 2 , and Fe(III), only forms II and III were formed, indicating that in addition to single strand breakage, double strand breaks were also induced.
Effects of ⅐ OH Scavengers and Bathocuproine on DNA Dam- Site Preference of DNA Cleavage-The DNA cleavage sites were examined using 32 P-5Ј-end-labeled DNA fragments of the human p53 tumor suppressor gene and the c-Ha-ras-1 protooncogene by the procedure of Maxam and Gilbert (28). As seen in Fig. 5, B and D, Angeli's salt alone caused DNA cleavage at every nucleotide position without marked site preference. On the other hand, Angeli's salt in the presence of H 2 O 2 and Cu(II) induced piperidine-labile sites frequently at thymine residues ( Fig. 5, A and C). The most preferred site was the thymine residue, especially in the 5Ј-CTG-3Ј, 5Ј-GTG-3Ј, and 5Ј-GTA-3Ј sequences.
Cytotoxicity-Human breast cancer cells (MCF-7) were incubated for 4.5 h with various concentrations of H 2 O 2 in the presence or absence of 500 M Angeli's salt or DEA-NO (Fig.  6A). Either 500 M Angeli's salt or DEA-NO or 0 -0.5 mM H 2 O 2 alone did not elicit significant cytotoxic effects in MCF-7 cells. However, in the presence of Angeli's salt or DEA-NO, LDH release was increased by H 2 O 2 dose dependently with the increase reaching 48.6 Ϯ 5.2% and 30.3 Ϯ 0.2%, respectively, at the 500 M concentration. On the other hand, Angeli's salt alone showed weak cytotoxic activity after 4.5 h of incubation. However, increased cytotoxicity was observed when the cells were analyzed after 8 h of incubation (data not shown). The presence of H 2 O 2 increased dramatically the cytotoxicity mediated by Angeli's salt. The increases in LDH release induced by 500 M H 2 O 2 alone or 2 mM Angeli's salt alone were only 9.0 Ϯ 2.0% and 17.0 Ϯ 0.3%, respectively, but up to 72.4 Ϯ 1.7% when both compounds were incubated together. Although no cytotoxic effects were found with even the highest concentration (2 mM) of DEA-NO alone, the presence of H 2 O 2 also enhanced cytotoxicity mediated by DEA-NO (Fig. 6B). However, the cytotoxic effect with DEA-NO plus H 2 O 2 was, in general, weaker than that with Angeli's salt plus H 2 O 2 . It should be noted that Angeli's salt and DEA-NO have similar half-lives (ϳ2.5 min) under physiological conditions. DISCUSSION Four different assays (DNA strand breakage, MDA formation from oxidation of 2Ј-deoxyribose, hydroxylation of 2Ј-deoxyguanosine in DNA, and analysis of site-specific DNA damage using 32 P-5Ј-end-labeled DNA fragments of the human p53 tumor suppressor gene and the c-Ha-ras-1 protooncogene) have been used to study the effects of Angeli's salt, an NO Ϫ generating compound (2-5), on DNA damage. It was found that Angeli's salt alone produced oxidants as previously reported (1), whereas the presence of H 2 O 2 and either Fe(III)-EDTA or   Cu(II) dramatically enhanced the production of oxidants mediated by Angeli's salt. NO ⅐ generated from DEA-NO did not enhance the formation of oxidants even in the presence of H 2 O 2 and metallic ions but rather inhibited it. Electron acceptors, ferricyanide and 4-OH-TEMPO, which have been reported to convert NO Ϫ to NO ⅐ (2), inhibited Angeli's salt-mediated oxidation reactions even in the presence of H 2 O 2 and metallic ions, suggesting that NO Ϫ , but not NO ⅐ , is responsible for the oxidation reactions. As previously reported for other reducing agents such as ascorbic acid and NADH (27), Angeli's salt induced DNA strand breakage and catalyzed the hydroxylation of 2Ј-deoxyguanosine in DNA more efficiently in the presence of Cu(II) than in the presence of Fe(III)-EDTA. Distinct effects of HO ⅐ scavengers on the oxidation reactions mediated by Fe(III)-EDTA and Cu(II) were also similar to those reported for ascorbic acid, glutathione, and NADH (27). The DNA cleavage sites examined using 32 P-5Ј-end-labeled DNA fragments of the human p53 tumor suppressor gene and the c-Ha-ras-1 protooncogene indicate that Angeli's salt alone caused DNA cleavage at every nucleotide position without marked site preference. This cleavage pattern was similar to that reported for DNA damage induced by HO ⅐ (with reductants such as ascorbic acid, glutathione, and NADH in the presence of H 2 O 2 and Fe(III)-EDTA) (27). On the other hand, Angeli's salt in the presence of H 2 O 2 and Cu(II) induced damage frequently at thymine residues, which was also similar to the pattern reported for DNA damage induced by the reductant, H 2 O 2 , and Cu(II) (27). These findings together lead us to conclude that NO Ϫ can act as a reducing agent to generate strong oxidants in the presence of H 2 O 2 and transition metallic ions.
It has been reported that reducing agents such as ascorbic acid and NADH can reduce transition metallic ions (M nϩ1 ) to their reduced forms (M n ), which stimulate production of reactive oxygen species from H 2 O 2 (Equation 6). Similarly NO Ϫ can act as a reducing agent to reduce transition metallic ions (Equation 7).
In the case of Fe(III)-EDTA, the Fenton reaction mediated by Fe(II) forms HO ⅐ , which is responsible for DNA damage (Equation 8), because HO ⅐ scavengers effectively inhibit the oxidation reactions. Conversely, Cu(II)-mediated DNA damage is not inhibited by HO ⅐ scavengers, suggesting that a Cu(I)-peroxide complex, which exhibits HO ⅐ -like activities, may be responsible for the DNA damage (Equation 9) (27, 31-33). In a previous study (1), we demonstrated that Angeli's salt and Piloty's acid (NO Ϫ -generating compounds) can produce strong oxidant(s) capable of inducing DNA strand breakage and oxidizing 2Ј-deoxyribose and calf thymus DNA to form MDA and 8-oxo-dG, respectively. These results led us to propose that NO Ϫ is a possible endogenous source of HO ⅐ , which may be formed either directly from the reaction product of NO Ϫ with NO ⅐ (N 2 O 2 . ) (Equations 1 and 2) or indirectly through H 2 O 2 formation (Equations 10 and 11).
In the present study, we observed that Angeli's salt alone induced DNA damage in the absence of H 2 O 2 and transition metallic ions, confirming our previous findings (1). The inclusion of both H 2 O 2 and transition metallic ions, however, synergistically enhanced DNA damage induced by Angeli's salt. On the other hand, the DNA damage induced by Angeli's salt was suppressed by the addition of either Fe(III)-EDTA or Cu(II) (Fig. 3), suggesting that NO Ϫ was converted to NO ⅐ , which was inactive in generating oxidant(s) through the above mechanisms via either Equations 1 and 2 or Equations 10 and 11. In addition, NO ⅐ possibly formed complexes with reduced transition metallic ions, inhibiting the generation of O 2 . and also the formation of oxidants from H 2 O 2 (19,34). Several groups have recently reported that H 2 O 2 and NO ⅐ cooperatively enhance their cytotoxic activity toward hepatoma cells (35), lymphoma cells (36), ovarian cancer cells (37), and Escherichia coli (38). Using aromatic hydroxylation of salicylate as an indicator, the reaction of H 2 O 2 with NO ⅐ generated from DEA-NO was shown to produce an HO ⅐ -like oxidant (39). Farias-Eisner et al. (37) also reported that NO ⅐ , H 2 O 2 , and ferric ion in combination produce a potent oxidant, which can oxidize benzene to produce phenol, and they proposed the following mechanism for HO ⅐ generation (Equations [12][13][14]. In contrast, our results demonstrated inhibitory effects of NO ⅐ on the Fenton reaction and no production of oxidants, at least under our experimental conditions using DEA-NO, H 2 O 2 , and Fe(III)-EDTA or Cu(II) in vitro. This NO ⅐ -mediated inhibition of the Fenton reaction was in agreement with results from our previous study (19) and others (34,40,41).
On the other hand, Angeli's salt alone at higher concentrations (1 and 2 mM) exerted weak cytotoxicity, as reported for cultured Chinese hamster V79 lung fibroblasts by Wink et al. (2), whereas no cytotoxic effects were observed with even the highest concentration (2 mM) of DEA-NO alone under our experimental conditions. However, the presence of H 2 O 2 increased cooperatively the cytotoxicity mediated by Angeli's salt. These results are in good agreement with those from our present in vitro study. This cytotoxic effect is probably because of the generation of HO ⅐ through the Fenton reaction, which could occur in our cell culture system because Fe(NO 3 ) 3 was present in the medium. Conversely, synergistic cytotoxic effects of DEA-NO and H 2 O 2 against MCF-7 cells were also observed, although the effects were, in general, lower than with Angeli's salt plus H 2 O 2 . Similar cooperative effects of NO ⅐ and H 2 O 2 on cytotoxicity were reported for other types of cells (35)(36)(37)(38). These results are not in agreement with our in vitro study, which showed that DEA-NO did not generate oxidants even in the presence of H 2 O 2 and metallic ions. There are several possible explanations for this discrepancy between in vitro and in vivo results. One possibility is that DEA-NO plus H 2 O 2 induced cytotoxicity by a mechanism independent of HO ⅐ formation. For example, NO ⅐ inhibits the mitochondrial respiratory chain reaction, and H 2 O 2 further enhances its toxicity. Alternatively, NO ⅐ may be converted in cells to NO Ϫ , which exerts toxic effects with H 2 O 2 . NO ⅐ has been reported to be converted to NO Ϫ by ferrocytochrome c (13).
In conclusion, we have demonstrated that the NO Ϫ -releasing compound, Angeli's salt, can catalyze the formation of strong oxidants in the presence of H 2 O 2 and metallic ions, inducing DNA strand breakage, oxidation of DNA to form 8-oxo-dG in vitro, and exerting cytotoxic effects toward human breast cancer cells. Recent studies have demonstrated that NO Ϫ may be formed in vivo under a variety of physiological conditions, including by NO ⅐ synthase (6 -8) and from S-nitrosothiols (9 -11) and nitrosylhemoglobin (12). NO ⅐ can be also converted to NO Ϫ in the presence of biomolecules such as superoxide dismutase (42) and ferrocytochrome c (13) and by the quinone/ hydroquinone redox system in a manner similar to that of the formation of O 2 . from oxygen (14). As stimulated immune cells including neutrophils and macrophages can produce H 2 O 2 , one can expect that, during an inflammatory process, the formation of both NO Ϫ and H 2 O 2 could enhance dramatically the antimicrobial and anti-tumoricidal activity. In addition to the inflammatory process, under a number of pathological conditions (e.g. ischemia reperfusion injury, etc.), increased production of reactive oxygen species and activation of NO ⅐ synthase have been shown to occur (24). Activated NO ⅐ synthase produces NO Ϫ , which may then encounter H 2 O 2 to generate strong oxidants, as shown in this study. Thus NO Ϫ may also play an important role as a cause of diverse pathophysiological conditions such as inflammation, neurodegenerative diseases, and cardiovascular disorders, especially when H 2 O 2 and transition metallic ions are present together.