Correspondence to: Dr. M. B. Grisham, Department of Molecular and Cellular Physiology, LSU Medical Center, PO Box 33932, Shreveport, LA 71130–3932, USA. E-mail: firstname.lastname@example.org
It is well known that chronic inflammation of the digestive tract is associated with an increased risk of malignant transformation. Because phagocytic leukocytes and cytokine-activated parachymal cells produce large amounts of reactive metabolites of oxygen and nitrogen, there has been substantial interest in ascertaining whether these reactive intermediates may mediate mutagenesis and malignant transformation in vivo. However, very little information is available regarding the basic chemistry of how these oxygen and nitrogen-derived species may interact to yield potentially carcinogenic agents. This review will discuss our present understanding of the chemical and biochemical interactions between superoxide and nitric oxide and provide a model by which these reactive species may damage DNA and mediate mutagenesis.
A growing body of clinical and experimental data suggest that chronic inflammation of the digestive tract is associated with an increased risk of developing malignancies. Patients with chronic oesophageal reflux, Helicobacter pylori-induced peptic ulcer disease, chronic pancreatitis, inflammatory bowel disease or hepatitis are known to be at significant risk for developing cancer of the oesophagus, stomach, pancreas, colon and liver, respectively. However, the mechanisms by which chronic inflammation promotes tumour formation remain poorly understood. It has been suggested that certain leucocyte and/or parenchymal cell-derived products may act as endogenous carcinogens or tumour promoters in vivo. 1 Several recent studies have demonstrated that tissue obtained from patients with Barrett's oesophagus, chronic H. pylori infection, chronic pancreatitis, inflammatory bowel disease and viral hepatitis exhibits upregulation of the inducible isoform of nitric oxide synthase (iNOS) and enhanced production of the free radical nitric oxide (NO). 2–9 Nitric oxide is unstable in the presence of molecular oxygen (O2) and will rapidly auto-oxidize to yield a variety of nitrogen oxide intermediates, some of which are potent nitrosating agents that will produce mutagenic and carcinogenic nitrosamines. 10–11 In addition to iNOS-derived NO, chronic inflammation is known to be associated with enhanced production of reactive oxygen species such as superoxide (O2–), hydrogen peroxide (H2O2) and hypochlorous acid (HOCl). 12–15 Secondary radicals produced from metal-catalysed degradation of H2O2, such as hydroxyl radical (OH·), are well-known mutagenic species capable of oxidatively modifying DNA. 16 The objectives of this discussion are to: (1) review the physiological chemistry of NO and oxygen-derived free radical interactions, and (2) examine how these interactions may be involved in inflammation-induced DNA damage and mutagenesis.
NO and N- and S-NITROSATION REACTIONS
Exposure of critical genes to mutagenic conditions increases the probability of tumour development. Chronic inflammation is one such environment that may promote malignant transformation. First, tissues neighbouring inflammatory foci undergo increased cell division. Consequently, mutagenic effects associated with chronic inflammation can become multiplicative, as the chance of DNA modifications and misrepair increase. 17 Indeed, cell proliferation is in itself a risk for malignant transformation. 17 Secondly, certain leukocyte- and tissue-derived metabolites may cause genomic damage, thereby increasing the probability of nicks, deletions and point mutations. 18 During chronic inflammation of the GI tract, for example, epithelial cells may be exposed to large amounts of NO (as much as 104 molecules/cell/s). 19 Although NO is chemically rather unreactive toward most bio-organic compounds, it will rapidly and spontaneously autoxidize to yield a variety of nitrogen oxide intermediates:
2NO + O2→ 2NO2
2NO + 2NO2→ 2N2O3
2N2O3 + 2H2O → 4NO2–+ 4H+
where NO2·, N2O3, and NO2– represent nitrogen dioxide, dinitrogen trioxide, and nitrite, respectively. Of these, N2O3 has drawn particular interest due to its ability to N- and S-nitrosate certain nucleophilic substrates such as primary and secondary amines as well as thiol-containing peptides and proteins. 20
Dinitrogen-trioxide has been shown to promote the nitrosative deamination of primary aromatic amines, including purines and pyrimidines, via the formation of nitrosamine and diazonium ion intermediates.
ArNH2 + N2O3→ ArNH – NO + NO2–+ H+
ArNH − NO + H+→ ArN2 + H2O
ArN2 + H2 O → ArOH + N2 + H+
Deamination of cytosine, methyl cytosine, adenine or guanine results in the formation of uracil, thymine, hypoxanthine and xanthine, respectively ( Fig. 1). Base conversion of cytosine and methyl cytosine can lead ultimately to base pair substitution mutations, while deamination of adenine and guanine will produce transversion mutations. 21 Moreover, the instability of hypoxanthine and xanthine in the DNA structure leads to rapid depurination and consequent single strand breaks. Even cross-linking with other nucleic acids or proteins has been suggested via reaction of a nucleophilic site on an adjacent macromolecule with the diazonium ion of the modified base. 22
NO-dependent N-nitrosation of secondary aliphatic and aromatic amines can also produce potentially carcinogenic nitrosamines. Secondary nitrosamines (R2NNO) are more stable than their primary amine counterparts:
R2NH + XNO → R2NNO + HX
Such nitrosamines, like many chemical carcinogens, are thought to promote mutagenesis and carcinogenesis via their ability to alkylate specific sites in DNA. For example, these types of nitrosamines undergo enzymatic (or nonenzymatic) α-hydroxylation. The α-hydroxy nitrosamine decomposes to form the alkyl diazonium ion and free alkyl carbocation ( Fig. 2). The alkyl diazonium salt or carbocation can then react with nucleophilic sites in DNA. To date, alkylation of DNA has been noted on the ring-nitrogen positions in the bases (adenine, guanine, cytosine, thymine) and the oxygen atoms of hydroxyl or carbonyl groups (guanine, thymine and cytosine), as well as on the phosphate groups. Thus, the limitation of NO-mediated genomic damage rests primarily on the localized diffusion of the small molecule, the degree of reactivity and nitrosation, and ultimately the cells' replicative and DNA repair machinery. However, even in the latter case, DNA repair proteins such as O 6-methylguanine-DNA-methyltransferase, formamidopyrimidine-DNA glycosylase and DNA ligase (Fpg), have been shown to be inactivated by NO-derived nitrosating agents such as N2O3in vitro and in vivo. 23–24 These DNA-repair proteins are thought to be inhibited by the NO-dependent S- or N-nitrosation of thiols, zinc finger moieties or amine groups of lysine residues at the active sites of these repair enzymes ( Fig. 3). 23–24
EFFECTS OF SUPEROXIDE ON NO-DEPENDENT N- and S-NITROSATION
Coincident with the sustained overproduction of NO, inflammatory foci are also sites of enhanced production of reactive oxygen species, such as O2– and H2O2, both of which are produced by activated phagocytic leukocytes. Because O2– is known to rapidly react with NO, it is of interest to determine whether this reactive oxygen species may modulate NO-dependent N- and S-nitrosation of primary aromatic amines and thiols, respectively, and thus modulate potentially mutagenic reactions. We have demonstrated that the addition of an O2– and H2O2 generator such as hypoxanthine/xanthine oxidase virtually eliminates the NO-dependent N- and S-nitrosation of model aromatic amines and thiols, respectively. 25–26 Inhibition was maximal when equimolar fluxes of NO and O2– were produced. We also noted that this inhibition was reversed by the addition of superoxide dismutase but not by catalase addition, suggesting that O2–, not H2O2, was responsible for the inhibition. We proposed that at equimolar fluxes, O2– reacts rapidly with NO to generate products that possess only limited ability to N- and/or S-nitrosate amino and/or thiol-containing compounds. Although we found that O2– inhibits the potentially mutagenic N-nitrosation of primary amines, interaction of O2– and NO yields the potent oxidant peroxynitrite (ONOO–), which could conceivably promote oxidative (and nitrative) modifications of DNA bases, thereby switching NO-mediated DNA damage from a nitrosative to a more oxidative pattern of mutagenic reactions ( Fig. 4). 27
Superoxide–no interactions: oxidative reactions
NO reacts with O2– to yield ONOO– and its conjugate acid peroxynitrous acid (ONOOH; pKa 6.6) in a second-order reaction with a rate constant of 6.7 × 109/M/s: 28
NO + O2–→ ONOO–+ H+ <==> ONOOH →
Peroxynitrous acid is very unstable and at physiological pH decomposes to yield nitrate (NO3–) via the intermediate formation of an excited isomer of ONOOH (ONOOH*). This electronically excited intermediate is thought to be the potent oxidizing agent that may homolyse to yield nitrogen dioxide radical (NO2·) and hydroxyl radical (OH·) within a solvent cage, the two free radicals diffusing out of the solvent cage to mediate oxidation reactions. 29 However, this mechanism has been shown to be thermodynamically unfavourable. A more probable mechanism to account for the potent oxidizing activity of ONOO–/ONOOH is that ONOOH* does not homolyse but does possess both NO2·- and OH·-like properties. 30 From the standpoint of biologically relevant reactions, ONOO– is an oxidizing, hydroxylating and nitrating agent that will oxidize and nitrate isolated DNA, resulting in DNA strand breaks. 31
Little is known regarding the detailed reactions of ONOO– with DNA. Because of the multiplicity of DNA modifications produced during oxidative reactions, it has been difficult to establish the specificity of mutations engendered by individual oxidants such as ONOO–. Oxidant-mediated DNA base modifications to produce 8-hydroxydeoxyguanosine (8-OHdG) has been suggested as a major product. 32 The occurrence of this alteration has been associated with a number of conditions leading to increased oxidative stress, including higher basal metabolic rate, gamma-irradiation and hydrogen peroxide-mediated oxidative stress. 33 NO and iron have also been implicated in the formation of 8-OHdG in asbestos-treated human lung epithelial cells. 34 Peroxynitrite also mediates the oxidation of deoxyguanosine. 35 In all of those conditions, the formation of 8-OHdG might lead to mutations by inducing misreading of the base itself and of the adjacent bases, 32 which may represent an important source of mutations. 36 Peroxynitrite induces G:C to T:A mutations for the supF gene in Escherichia coli and in human AD293 cells. 37 In addition to oxidative reactions, recent data suggest that the interaction of ONOO– with DNA results in the nitration of guanine to form 8-nitroguanine ( Fig. 4). 38 This modification is potentially mutagenic, the depurination of 8-nitroguanine yielding apurinic sites with the resultant possibility of G:C to T:A transversions. 39 However, whether ONOO– mediates such DNA damage in cells and tissues is yet to be determined. One would have to envision that O2– and NO are produced in equimolar amounts within the nucleus, since O2– will not diffuse away from its source of production to any significant extent. Although the identification of 8-nitroguanine may be used as a marker of ONOO–-induced DNA modification, it should be noted that other NO-derived nitrating agents such as NO2·, derived from myeloperoxidase-catalysed oxidation of nitrite (NO2–), or OClNO, produced from the interaction between HOCl and NO, 40–42 would also be expected to nitrate DNA bases.
It is important to note that the evaluation of ONOO–-mediated mutagenic properties has been assessed in vitro using bolus amounts of chemically synthesized oxidant. However, it is becoming increasingly apparent that the formation of ONOO– at sites where both O2– and NO are produced may depend upon the relative fluxes of NO and O2. We have found that the simultaneous production of equimolar fluxes of O2– and NO dramatically increased the oxidation of the oxidant-sensitive probe dihydrorhodamine (DHR). 27 This oxidation was inhibited by superoxide dismutase but not by catalase, suggesting that O2–, not H2O2, interacted with NO to form ONOO–/ONOOH. 27 As the flux of one radical exceeded the other, oxidation of DHR was inhibited, suggesting that excess production of either radical may act as an endogenous modulator of ONOO–/ONOOH formation. Subsequent experiments by our laboratory, as well as others, have demonstrated that NO (or O2–) interacts with and decomposes ONOO–/ONOOH. 27–43 Although this hypothesis suggests that NO and O2– may modulate steady state concentrations of ONOO–, there has yet to be any direct evidence demonstrating such modulation of ONOO–-mediated oxidative and/or nitrating reactions under physiological conditions.
FENTON CHEMISTRY and NO
It has been known for many years that certain reactive oxygen species generated by external radiation, activated leucocytes or specific enzymes may modify and/or damage DNA. Neither O2– nor H2O2 are potent enough oxidizing agents to damage or modify DNA. However, O2– and H2O2 may produce potent secondary radicals such as OH· via metal (Mn)-catalysed reactions. It is known that O2– and H2O2 interact with chelates of iron or copper to yield the potent oxidant OH· or OH·-like species via the superoxide-driven Fenton reaction:
The interaction of free radicals derived from superoxide-driven Fenton reactions with DNA has received considerable interest over the years. 44 Recent reports have focused on the ability of copper to participate in mutagenic reactions in vivo via Fenton-catalysed reactions. Copper is an important structural metal in chromatin 45 that in fact induces more DNA bases damage in the presence of H2O2 than does iron. 46 There is now evidence to suggest that NO may modulate Fenton-driven oxidative reactions. We have recently investigated the ability of different fluxes of NO to modulate iron complex- 27 and haemoprotein-catalysed oxidative reactions. 47 We found that generation of O2– and H2O2 in the presence of 5 μm Fe+3-EDTA stimulated dramatically the hydroxylation of benzoic acid. Catalase and superoxide dismutase were both effective at inhibiting this classic Fenton-driven reaction. Addition of NO inhibited this reaction in a concentration-dependent manner such that a ratio of NO/O2–/H2O2 of 1:1:1 inhibited hydroxylation of benzoic acid by 90%. 27
We have recently assessed the ability of a model haemoprotein (myoglobin) to oxidize DHR in the presence or the absence of O2–, H2O2 and/or NO. 47 In the presence of equimolar fluxes of H2O2 and O2–, the addition of metmyoglobin (Mb-Fe+3) dramatically enhanced DHR oxidation via the formation of ferryl myoglobin (Mb-Fe+4). This oxidative reaction was, as expected, inhibited by catalase but not superoxide dismutase. Addition of NO to this system further enhanced DHR oxidation, which was inhibited by superoxide dismutase, suggesting that O2– reacted with NO to yield ONOO–/ONOOH in addition to Mb-Fe+4· Further increases in NO flux dramatically inhibited DHR oxidation, which was found to be due to the NO-mediated reduction of Mb-Fe+4 to Mb-Fe+3· Taken together, these data suggest that NO may modulate iron complex- or haemoprotein-catalysed oxidative reactions depending upon the relative fluxes of O2–, H2O2 and NO. In accordance with our results, Pacelli et al. have shown that NO can inhibit DNA strand breaks induced by H2O2 and certain transition metals. 48
It is well known that chronic inflammation of the digestive tract is associated with an increased risk of malignant transformation. The phagocytic leucocytes as well as cytokine-activated parenchymal cells have the potential to produce large amounts of reactive metabolites of oxygen and nitrogen, which may in turn mediate mutagenesis and possibly tumour transformation. Thus, a fundamental understanding of the interplay between O2– and NO may provide new insight into the role that these reactive species play in DNA damage and mutagenesis.
Some of this work was supported by the Crohn's and Colitis foundation of America, The Feist Foundation of LSU Health Sciences Center and the NIH (DK 47663 and DK 43785).