DNA strand breaks produced by oxidative stress in mammalian cells exhibit 3'-phosphoglycolate termini

DNA-targeted 2-nitroimidazoles: studies of the influence of the phenanthridine-linked nitroimidazoles, 2-NLP-3 and 2-NLP-4, on DNA damage induced by ionizing radiation

The nitroimidazole-linked phenanthridines 2-NLP-3 (5-[3-(2-nitro-1-imidazoyl)-propyl]-phenanthridinium bromide) and 2-NLP-4 (5-[3-(2-nitro-1-imidazoyl)-butyl]-phenanthridinium bromide) are composed of the radiosensitizer, 2-nitroimidazole, attached to the DNA intercalator phenanthridine by a 3- and 4-carbon linker, respectively. Previous in vitro assays showed both compounds to be 10-100 times more efficient as hypoxic cell radiosensitizers (based on external drug concentrations) than the untargeted 2-nitroimidazole radiosensitizer, misonidazole (Cowan et al., Radiat. Res. 127, 81-89, 1991). Here we have used a (32)P postlabeling assay and 5'-end-labeled oligonucleotide assay to compare the radiation-induced DNA damage generated in the presence of 2-NLP-3, 2-NLP-4, phenanthridine and misonidazole. After irradiation of the DNA under anoxic conditions, we observed a significantly greater level of 3'-phosphoglycolate DNA damage in the presence of 2-NLP-3 or 2-NLP-4 compared to irradiation of the DNA in the presence of misonidazole. This may account at least in part for the greater cellular radiosensitization shown by the nitroimidazole-linked phenanthridines over misonidazole. Of the two nitroimidazole-linked phenanthridines, the better in vitro radiosensitizer, 2-NLP-4, generated more 3'-phosphoglycolate in DNA than did 2-NLP-3. At all concentrations, phenanthridine had little effect on the levels of DNA damage, suggesting that the enhanced radiosensitization displayed by 2-NLP-3 and 2-NLP-4 is due to the localization of the 2-nitroimidazole to the DNA by the phenanthridine substituent and not to radiosensitization by the phenanthridine moiety itself.

Folate deficiency disturbs hepatic methionine metabolism and promotes liver injury in the ethanol-fed micropig

Alcoholic liver disease is associated with abnormal hepatic methionine metabolism and folate deficiency. Because folate is integral to the methionine cycle, its deficiency could promote alcoholic liver disease by enhancing ethanol-induced perturbations of hepatic methionine metabolism and DNA damage. We grouped 24 juvenile micropigs to receive folate-sufficient (FS) or folate-depleted (FD) diets or the same diets containing 40% of energy as ethanol (FSE and FDE) for 14 wk, and the significance of differences among the groups was determined by ANOVA. Plasma homocysteine levels were increased in all experimental groups from 6 wk onward and were greatest in FDE. Ethanol feeding reduced liver methionine synthase activity, S-adenosylmethionine (SAM), and glutathione, and elevated plasma malondialdehyde (MDA) and alanine transaminase. Folate deficiency decreased liver folate levels and increased global DNA hypomethylation. Ethanol feeding and folate deficiency acted together to decrease the liver SAM/S-adenosylhomocysteine (SAH) ratio and to increase liver SAH, DNA strand breaks, urinary 8-oxo-2'-deoxyguanosine [oxo(8)dG]/mg of creatinine, plasma homocysteine, and aspartate transaminase by more than 8-fold. Liver SAM correlated positively with glutathione, which correlated negatively with plasma MDA and urinary oxo(8)dG. Liver SAM/SAH correlated negatively with DNA strand breaks, which correlated with urinary oxo(8)dG. Livers from ethanol-fed animals showed increased centrilobular CYP2E1 and protein adducts with acetaldehyde and MDA. Steatohepatitis occurred in five of six pigs in FDE but not in the other groups. In summary, folate deficiency enhances perturbations in hepatic methionine metabolism and DNA damage while promoting alcoholic liver injury.

Inhibition of cell proliferation and fibronectin biosynthesis by Na ascorbate

BACKGROUND

The importance of ascorbate on the production of extracellular matrix proteins (as elastin and collagens) is now well documented, but no studies have been published concerning its effects on fibronectin biosynthesis. Fibronectin is important for cell attachment and for proliferation.

MATERIALS AND METHODS

The effects of Na ascorbate were investigated on cell attachment, proliferation, viability and fibronectin biosynthesis by human skin fibroblasts in vitro. Proliferation was followed by the monitoring of [(3)H]-thymidine incorporation; viability by the MTT-test, cell adherence by counting adherent and nonadherent cells and fibronectin biosynthesis by immunoprecipitation of biosynthetically labelled fibronectin.

RESULTS

In the presence of ascorbate, the fibroblasts showed a biphasic growth pattern. At 500 microM ascorbate, [(3)H]-thymidine incorporation was stimulated by 15% as compared to the controls. Higher concentrations gradually decreased proliferation up to 36% of the control value at 5 mM. These effects of ascorbate on DNA synthesis were followed to > 1.25 mM by a strong inhibition, cytotoxic effect and cell death. The non-adherent cell count increased to 10% of the total population at 2.5 mM and to 31% at 5.0 mM ascorbate.Increasing concentrations of ascorbate resulted in a dose-dependent decrease of fibronectin biosynthesis, both in the culture supernates and cell extracts. This inhibition mainly concerned cell membrane-associated fibronectin.Superoxide-dismutase or catalase could inhibit Na ascorbate-induced cytotoxicity and partially re-establish fibronectin biosynthesis. Desferrioxamine, ergothionein and vitamin E were inefficient.

CONCLUSIONS

Our results indicate that ascorbate decreases fibronectin biosynthesis of cultured human skin fibroblasts, thereby producing cell detachment and decreased proliferation. This effect is mainly mediated by the reactive oxygen species and can be inhibited by superoxide-dismutase and catalase.

The role of oxidative stress in mechanisms of metal-induced carcinogenesis

Metals are necessary for the normal functioning of cells and the survival of organisms. However, exposure to higher than the physiological levels of several metals may lead to tumor development. Although the exact molecular mechanism(s) of metal-induced carcinogenesis is not clear, a vast body of evidence indicates that metal-induced generation of reactive oxygen species (ROS) may play a central role in this process. Two main pathways of ROS-induced effects are discussed in this chapter: (i) increased DNA damage induced either directly or indirectly by impeding DNA repair, and (ii) modulation of nuclear transcriptional factor activities, such as NF-kappaB and AP-1, through mitogen-activated protein kinases signal transduction mechanisms.

Evaluation of phosphodiesterase I-based protocols for the detection of multiply damaged sites in DNA: the detection of abasic, oxidative and alkylative tandem damage in DNA oligonucleotides

It has been proposed that DNA multiply damaged sites (MDS), where more than one moiety in a local region ( approximately 1 helical turn, 10 bp) of the DNA is damaged, are lesions of enhanced biological significance. However, other than indirect measures, there are few analytical techniques that allow direct detection of MDS in DNA. In the present study we demonstrate the potential of protocols incorporating an exonucleolytic snake venom phosphodiesterase (SVPD) digestion stage to permit the direct detection of certain tandem damage, in which two lesions are immediately adjacent to each other on the same DNA strand. A series of prepared oligonucleotides containing either single or pairs of tetrahydrofuran moieties (F), thymine glycol lesions (T(g)) or methylphosphotriester adducts (Me-PTE) were digested with SVPD and the digests examined by either (32)P-end-labelling or electrospray mass spectrometry. The unambiguous observation of SVPD-resistant 'trimer' species in the digests of oligonucleotides containing adjacent F, T(g) and Me-PTE demonstrates that the SVPD digestion strategy is capable of allowing direct detection of certain tandem damage. Furthermore, in studies to determine the specificity of SVPD in dealing with pairs of lesions on the same strand, it was found mandatory to have the two lesions immediately adjacent to each other in order to generate the trimer species; pairs of lesions separated by as few as one or two normal nucleotides behave principally as single lesions towards SVPD.

Characterization of a multi-functional metal-mediated nuclease by MALDI-TOF mass spectrometry

Mass spectrometric analysis of reaction products allows simultaneous characterization of activities mediated by multifunctional enzymes. By use of MALDI-TOF mass spectrometry, the relative influence of magnesium and manganese promoted exonuclease and phosphatase activities of Esherichia coli exonuclease III have been quantitatively measured, offering a rapid and sensitive alternative to radioactivity quantification and gel electrophoresis procedures for determination of reaction rate constants. Manganese is found to promote higher levels of exonuclease activity, which could be a source of mutagenic effects if this ion were selected as the natural cofactor. Several potential applications of these methods to quantitative studies of DNA repair chemistry are also described.

In vitro repair of complex unligatable oxidatively induced DNA double-strand breaks by human cell extracts

We describe a new assay for in vitro repair of oxidatively induced DNA double-strand breaks (DSBs) by HeLa cell nuclear extracts. The assay employs linear plasmid DNA containing DNA DSBs produced by the radiomimetic drug bleomycin. The bleomycin-induced DSB possesses a complex structure similar to that produced by oxidative processes and ionizing radiation. Bleomycin DSBs are composed of blunt ends or ends containing a single 5'-base overhang. Regardless of the 5'-end structure, all bleomycin-induced DSBs possess 3'-ends blocked by phosphoglycolate. Cellular extraction and initial end joining conditions for our assay were optimized with restriction enzyme-cleaved DNA to maximize ligation activity. Parameters affecting ligation such as temperature, time, ionic strength, ATP utilization and extract protein concentration were examined. Similar reactions were performed with the bleomycin-linearized substrate. In all cases, end-joined molecules ranging from dimers to higher molecular weight forms were produced and observed directly in agarose gels stained with Vistra Green and imaged with a FluorImager 595. This detection method is at least 50-fold more sensitive than ethidium bromide and permits detection of </=0.25 ng double-stranded DNA per band in post-electrophoretically stained agarose gels. Consequently, our end-joining reaction requires </=100 ng substrate DNA and >/=50% conversion of substrate to product is achieved with simple substrates such as restriction enzyme-cleaved DNA. Using our assay we have observed a 6-fold lower repair rate and a lag in reaction initiation for bleomycin-induced DSBs as compared to blunt-ended DNA. Also, end joining reaction conditions are DSB end group dependent. In particular, bleomycin-induced DSB repair is considerably more sensitive to inhibition by increased ionic strength than repair of blunt-ended DNA.

Iron and its sensitive balance in the cell

Iron is vital in life because it is an important component of molecules that undergoes redox reactions or transport oxygen. However, the existence of two stable and inter-convertible forms of iron, iron(III) and iron(II), makes possible one electron being transferred to or captured from other species to form radicals. In particular, superoxide and hydroxyl radicals may be formed in these reactions, both with capacity of attacking other molecules. DNA is one important target and a vast literature exists showing that attack of hydroxyl radical to DNA leads to cell death cellular necrosis, apoptosis, mutation and malignant transformation. Therefore, a fine balance must exist at various levels of an organism to maintain iron concentration in a narrow range, above and below which deleterious effects of distinct nature occur. This review will deal with the formation of oxygen reactive species in iron participating reactions, defenses in the organism against these species, the different mechanisms of iron homeostasis and iron deficiency and iron overload related diseases.

Retinol supplementation induces oxidative stress and modulates antioxidant enzyme activities in rat sertoli cells

Recent intervention studies revealed that supplementation with retinoids resulted in a higher incidence of lung cancer. Recently the causal mechanism has begun to be clarified. We report here that retinol caused cellular oxidative stress and modulated superoxide dismutase, catalase and glutathione peroxidase activities. Retinol (7 microM) significantly increased TBARS, conjugated dienes, and hydroperoxide-initiated chemiluminescence in cultured Sertoli cells. In response to retinol treatment superoxide dismutase, catalase and glutathione peroxidase activities increased. TBARS content and catalase activities were decreased by a free radical scavenger. These findings suggest that retinol may induce oxidative stress and modulate antioxidant enzyme activities in Sertoli cells.

Retinol supplementation induces DNA damage and modulates iron turnover in rat Sertoli cells

Recent intervention studies revealed that supplementation with retinoids resulted in a higher incidence of lung cancer. Recently the causal mechanism has begun to be clarified. We report here that retinol caused cellular DNA damage probably involving cellular iron accumulation. Retinol (7 microM) significantly induced DNA single strands breaks, DNA fragmentation and production of 8-oxo-7, 8-dihydro-2'-deoxyguanosine in cultured Sertoli cells. In contrast, lower doses seemed not to induce single-strands break in this experimental model. The breaks in DNA were inhibited by an iron scavenger; and 7 microM retinol treatment modulated iron turnover leading to iron accumulation, suggesting that iron ions were required for the retinol cellular effects. These findings suggest that retinol-induced DNA damage was associated with the modulation of iron turnover, and these characteristics could be responsible for the increased incidence of lung cancer associated with retinoids supplementation.

Stannous chloride mediates single strand breaks in plasmid DNA through reactive oxygen species formation

Stannous ion (Sn) has been employed in nuclear medicine and in food industry. We described that Stannous Chloride (SnCl2) inactivation effect in Escherichia coli is mediated by a Fenton-like reaction. The effect of SnCl2 was studied through: (i) the alteration of plasmid topology in neutral and acidic pH by gel electrophoresis; and (ii) the transformation efficiency of an wild type E. coli strain. Treatment of plasmid DNA pUC 9.1 with SnCl2, at pH 7.4, results in DNA single-strand breaks (SSB), in a dose-dependent manner. Addition of sodium benzoate partly inhibited the DNA damage, while EDTA completely abolishes DNA-SSB. Furthermore, the ability of the plasmid to transform E. coli was reduced. At pH 1.3, SnCl2 exerts a protective effect on plasmid against HCI depurination. Our results suggest the generation of ROS, such as *OH by a Fenton-like reaction, close to the site of the lesions due to a possible complexation of stannous ion to DNA.

Determination of human DNA polymerase utilization for the repair of a model ionizing radiation-induced DNA strand break lesion in a defined vector substrate

Human DNA polymerase and DNA ligase utilization for the repair of a major class of ionizing radiation-induced DNA lesion [DNA single-strand breaks containing 3'-phosphoglycolate (3'-PG)] was examined using a novel, chemically defined vector substrate containing a single, site-specific 3'-PG single-strand break lesion. In addition, the major human AP endonuclease, HAP1 (also known as APE1, APEX, Ref-1), was tested to determine if it was involved in initiating repair of 3'-PG-containing single-strand break lesions. DNA polymerase beta was found to be the primary polymerase responsible for nucleotide incorporation at the lesion site following excision of the 3'-PG blocking group. However, DNA polymerase delta/straightepsilon was also capable of nucleotide incorporation at the lesion site following 3'-PG excision. In addition, repair reactions catalyzed by DNA polymerase beta were found to be most effective in the presence of DNA ligase III, while those catalyzed by DNA polymerase delta/straightepsilon appeared to be more effective in the presence of DNA ligase I. Also, it was demonstrated that the repair initiating 3'-PG excision reaction was not dependent upon HAP1 activity, as judged by inhibition of HAP1 with neutralizing HAP1-specific polyclonal antibody.

The ATM protein is required for sustained activation of NF-kappaB following DNA damage

Cells lacking an intact ATM gene are hypersensitive to ionizing radiation and show multiple defects in the cell cycle-coupled checkpoints. DNA damage usually triggers cell cycle arrest through, among other things, the activation of p53. Another DNA-damage responsive factor is NF-kappaB. It is activated by various stress situations, including oxidative stress, and by DNA-damaging compounds such as topoisomerase poisons. We found that cells from Ataxia Telangiectasia patients exhibit a defect in NF-kappaB activation in response to treatment with camptothecin, a topoisomerase I poison. In AT cells, this activation is shortened or suppressed, compared to that observed in normal cells. Ectopic expression of the ATM protein in AT cells increases the activation of NF-kappaB in response to camptothecin. MO59J glioblastoma cells that do not express the DNA-PK catalytic subunit respond normally to camptothecin. These results support the hypothesis that NF-kappaB is a DNA damage-responsive transcription factor and that its activation pathway by DNA damage shares some components with the one leading to p53 activation.

Crosslinking of DNA and proteins induced by protein hydroperoxides

Exposure of DNA to several proteins peroxidized by radiation-generated hydroxyl free radicals resulted in formation of crosslinks between the macromolecules, detected by retardation and broadening of DNA bands in agarose gels. This technique proved suitable for the study of crosslinking of DNA with peroxidized BSA, insulin, apotransferrin and alpha casein, but not with several other proteins, including histones. The crosslinking depended on the presence of intact hydroperoxide groups on the protein, on their number, and on the duration of the interaction with DNA. All DNA samples tested, pBR322, pGEM, lambda/HindIII and pUC18, formed crosslinks with the peroxidized BSA. Sodium chloride and formate prevented the crosslinking if present during incubation of the peroxidized protein and DNA, but had no effect once the crosslinks had formed. The gel shift of the crosslinked DNA was reversed by proteolysis, indicating that the DNA mobility change was due to attachment of protein and that the crosslinking did not induce DNA strand breaks. The metal chelators Desferal and neocuproine reduced the extent of the crosslinking, but did not prevent it. Scavengers of free radicals did not inhibit the crosslink formation. The DNA-protein complex was not disrupted by vigorous agitation, by filtration or by non-ionic detergents. These observations show that the crosslinking of DNA with proteins mediated by protein hydroperoxides is spontaneous and probably covalent, and that it may be assisted by transition metals. It is suggested that formation of such crosslinks in living organisms could account for some of the well-documented forms of biological damage induced by reactive oxygen species-induced oxidative stress.

SVPD-post-labeling detection of oxidative damage negates the problem of adventitious oxidative effects during 32P-labeling

The exploitation of oxidative DNA lesions as biomarkers of oxidative stress in vivo requires techniques that allow for the precise and valid measurement of oxidative damage to DNA. Previously, endogenous levels of the oxidative lesion 8-hydroxy-2'-deoxyguanosine (8-HO-dG) in rat tissues determined by a micrococcal nuclease/calf spleen phosphodiesterase-based 32P-post-labeling protocol were found to be at least 10-fold higher than those determined by HPLC with electrochemical detection. This was attributed to the adventitious oxidation of the normal nucleotides (dGp) occurring during the labeling stage of the postlabeling protocol, which could only be prevented by the introduction of additional chromatographic steps to remove the unmodified species prior to labeling. In the present study we report that an alternative snake venom phosphodiesterase-based 32P-post-labeling procedure (SVPD-postlabeling) negates the problem of adventitious oxidative damage during labeling by virtue of a unique digestion strategy. In SVPD-post-labeling, digestion yields certain lesions (thymine glycols, phosphoglycolates and abasic sites) as damage-containing dimer species which are ready substrates for labeling. In contrast, the undamaged DNA is recovered as mononucleoside species (dN) which are not substrates for labeling and so remain undetected. Furthermore, even if the mononucleosides are oxidized during labeling, they will not contribute to the level of damage detected. Indeed, we demonstrate that neither the external gamma-irradiation of the digested DNA samples nor increasing the incubation time of the labeling reaction alters the levels of damage detected by SVPD-post-labeling. The negation of adventitious oxidative effects during labeling deems that an optimized SVPD-post-labeling procedure should be well-suited for the biomonitoring of endogenous oxidative stress in vivo.

Sequence-specific DNA cleavage by Fe2+-mediated fenton reactions has possible biological implications

Preferential cleavage sites have been determined for Fe2+/H2O2-mediated oxidations of DNA. In 50 mM H2O2, preferential cleavages occurred at the nucleoside 5' to each of the dG moieties in the sequence RGGG, a sequence found in a majority of telomere repeats. Within a plasmid containing a (TTAGGG)81 human telomere insert, 7-fold more strand breakage occurred in the restriction fragment with the insert than in a similar-sized control fragment. This result implies that telomeric DNA could protect coding DNA from oxidative damage and might also link oxidative damage and iron load to telomere shortening and aging. In micromolar H2O2, preferential cleavage occurred at the thymidine within the sequence RTGR, a sequence frequently found to be required in promoters for normal responses of many procaryotic and eucaryotic genes to iron or oxygen stress. Computer modeling of the interaction of Fe2+ with RTGR in B-DNA suggests that due to steric hindrance with the thymine methyl, Fe2+ associates in a specific manner with the thymine flipped out from the base stack so as to allow an octahedrally-oriented coordination of the Fe2+ with the three purine N7 residues. Fe2+-dependent changes in NMR spectra of duplex oligonucleotides containing ATGA versus those containing AUGA or A5mCGA were consistent with this model.

60 Hz magnetic field exposure induces DNA crosslinks in rat brain cells

In previous research, we found an increase in DNA strand breaks in brain cells of rats acutely exposed to a 60 Hz magnetic field (for 2 h at an intensity of 0.5 mT). DNA strand breaks were measured with a microgel electrophoresis assay using the length of DNA migration as an index. In the present experiment, we found that most of the magnetic field-induced increase in DNA migration was observed only after proteinase-K treatment, suggesting that the field caused DNA-protein crosslinks. In addition, when brain cells from control rats were exposed to X-rays, an increase in DNA migration was observed, the extent of which was independent of proteinase-K treatment. However, the X-ray-induced increase in DNA migration was retarded in cells from animals exposed to magnetic fields even after proteinase-K treatment, suggesting that DNA-DNA crosslinks were also induced by the magnetic field. The effects of magnetic fields were also compared with those of a known DNA crosslink-inducing agent mitomycin C. The pattern of effects is similar between the two agents. These data suggest that both DNA-protein and DNA-DNA crosslinks are formed in brain cells of rats after acute exposure to a 60 Hz magnetic field.

Levels of DNA strand breaks and superoxide in phorbol ester-treated human granulocytes

Phorbol ester treatment of granulocytes triggers release of superoxide (O2.-) and a concomitant burst of DNA strand breaks. The relationship between the amount of O2.- and the number of DNA breaks has not previously been explored. To quantify the relatively large amount of O2.- generated over a 40-min period by 1 x 10(6) granulocytes/mL, a discontinuous "10-min pulse" method employing cytochrome c was used; 140 nmol O2.- per 1 x 10(6) cells was detected. DNA strand breaks were quantified by fluorimetric analysis of DNA unwinding (FADU). To vary the level of O2.- released by cells, inhibitors of the respiratory burst were used. Sodium fluoride (1-10 mM) and staurosporine (2-10 nM) both inhibited O2.- production. In both cases, however, inhibition of strand breakage was considerably more pronounced than inhibition of O2.-. Zinc chloride (50-200 microM) inhibited both O2.- and DNA breaks, approximately equally. Dinophysistoxin-1 (okadaic acid) inhibited O2.- production more effectively than it inhibited DNA breaks. O2.- dismutes to H2O2, a reactive oxygen species known to cause DNA breaks. The addition of catalase to remove extracellular H2O2 had no effect on DNA breakage. Using pulse field gel electrophoresis, few double-stranded breaks were detected compared to the number detected by FADU, indicating that about 95% of breaks were single-stranded. The level of DNA breaks is not directly related to the amount of extracellular O2.- or H2O2 in PMA-stimulated granulocytes. We conclude that either an intracellular pool of these reactive oxygen species is involved in breakage or that the metabolic inhibitors are affecting a novel strand break pathway.

Iron homeostasis, oxidative stress, and DNA damage

Cellular DNA damage under prooxidant conditions has been shown to be mediated by iron. In fact, iron is an important element in the establishment of a prooxidant status in the cell. It is discussed that there exists a mutual dependence between iron metabolism and oxidative stress. Changes in the former by means of genetic manipulation bring about modification in the redox status as judge by oxidative damage in DNA. On the other hand, the induction of a cellular prooxidative condition activates the protein IRP (iron regulatory protein) in a way that renders the cell more able to take up iron. The possible implications of these results is discussed in the light of recent findings reported in the literature on hydrogen peroxide as a signaling species for cell proliferation. The question of DNA strand break formation under prooxidant conditions is reviewed from the viewpoint of which agent is more important: an oxidant generated by Fenton type reaction or Ca2+-activated nucleases. The presence of iron in the nucleus is reviewed. Results have been produced indicating that the larger concentration of this metal in the nucleus, as compared to the cytosol, seems to be explained by an iron-type P-ATPase. There is no explanation, presently, for iron presence in the nucleus, but it certainly imposes a prooxidant trend that needs to be counterbalanced in some way, and evidence is reviewed that nuclear metallothionein plays a role in this regard.

Chinese hamster fibroblasts overexpressing CuZn-superoxide dismutase undergo a global reduction in antioxidants and an increasing sensitivity of DNA to oxidative damage

Transfection of a CuZn-superoxide dismutase (SOD) expression vector into V79 Chinese hamster cells produced clones in which CuZn-SOD activities were 2.2-3.5-fold higher than in the parental cells. An overall moderate reduction of glutathione peroxidase, glutathione reductase and glucose-6-phosphate dehydrogenase activities and of both GSSG and total glutathione levels was found. In one particular clone the catalase activity was also reduced. The pro-oxidant status established by the lower level of antioxidant defence rendered the SOD-overexpressing cells more sensitive to the production of 8-oxo-2'-deoxyguanosine by hydrogen peroxide. The data are discussed in terms of a model resembling the bacterial sox RS system.