BioFactors 33 (2008) 33–48 IOS Press 33 Effect of quercetin on oxidative nuclear and mitochondrial DNA damage Lucia Potenzaa,∗ , Cinzia Calcabrinia , Roberta De Bellisa , Umberto Mancinia, Luigi Cucchiarinia and Marina Dachàb a Dipartimento b Centro di Scienze Biomolecolari, Universit à degli Studi di Urbino “Carlo Bo”, Urbino, Italy Integrato di Ricerche, Universit à Campus Bio-Medico, Rome, Italy Received 5 May 2008 Revised 11 August 2008 Accepted 12 August 2008 Abstract. Quercetin is a well-investigated antioxidant known to protect cells against oxidative nuclear DNA damage. There is no knowledge regarding its effect on oxidative mitochondrial DNA damage. In this study we investigated the effect of quercetin on oxidatively-injured DNA. Cell-free and cell studies were performed. Cell-free analyses carried out on plasmidic DNA showed that quercetin protects from all oxidative challenges used. Cellular studies were carried out on NCTC 2544 cells which were insulted with hydrogen peroxide and UVC radiations. Nuclear and mitochondrial DNAs were analysed by measuring DNA damage with a quantitative polymerase chain reaction. Quercetin supplementation showed significant genoprotective activity on mitochondrial DNA when hydroperoxide was used. The evidence of the protection afforded by quercetin suggests that this flavonoid may play an important role on mitochondrial genome stability. Keywords: Quercetin, oxidative damage, mitochondrial DNA, nuclear DNA, NCTC 2544 Abbreviations: CCC, covalently closed circular; DMSO, dimetylsulfoxide; DTT, DL-dithiothreitol; LDH, lactate dehydrogenase; MFO, mixed-function oxidase; mtDNA, mitochondrial DNA; nDNA, nuclear DNA; OC, open circular; PBS, phosphate-buffered saline; PCR, polymerase chain reaction; ROS, reactive oxygen species; QPCR, quantitative polymerase chain reaction. 1. Introduction Oxidative stress, associated with the formation of reactive oxygen species (ROS), plays an important role in the pathogenesis of various physiological and pathological processes harmful to humans such as inflammation, immunosuppression, aging and carcinogenesis. It is generally believed that oxidative stress-induced damage may be prevented through the intake of antioxidants, molecules which can react with biologic targets breaking chain reactions or blocking highly reactive oxygen metabolites. Among them are flavonoids, a group of low molecular weight benzo-γ pyrone derivatives, present in fruit, vegetables, grains, bark, roots, stems, flowers, tea and wine [36]. The protective effects of flavonoids in biological systems are due to their ability to transfer electrons to the ∗ Address for correspondence: Lucia Potenza, Dipartimento di Scienze Biomolecolari, Via A. Saffi, 2 Università degli Studi di Urbino “Carlo Bo”, 61029 Urbino, Italy. Tel.: +39 0722 305232; Fax: +39 0722 305324; E-mail: lucia.potenza@uniurb.it. 0951-6433/08/$17.00  2008 – IUBMB/IOS Press and the authors. All rights reserved 34 L. Potenza et al. / Effect of quercetin on oxidative nuclear and mitochondrial DNA damage free radicals, chelate metals, activate antioxidant enzymes, reduce the radical α-tocopherol and inhibit oxidases [26]. Quercetin (3,3’,4’,5,7-pentahydroxyflavone), one of the most common flavonoid, has been placed at the center of scientific investigation due to its physiological and biochemical properties. It belongs to the flavonol family and occurs in nature in the glycosidic form; onions, broccoli, tomatoes, lettuce, tea, berries, red wine, olive oil and apple peel represent the main dietary sources. Catechol group and hydroxyl group at the central ring are responsible for its free radical-scavenging activity. Fluorescence microscopy has demonstrated that quercetin is localized in membranes, cytoplasm and inside the nucleus of the cell [41]. The same authors also reported the mechanism by which quercetin, at physiological pH, modifies membrane phospholipids probably because of its incorporation in the polar head groups, with a consequent influence on membrane-dependent regulation mechanisms [41]. Experimental evidences proved the role of quercetin in modifying eicosanoid biosynthesis (antiprostanoids and anti-inflammatory responses), in protecting lipoproteins from oxidation (prevention in the atherosclerotic plaque formation), in preventing platelet aggregation (antithrombic effects) and in promoting the cardiovascular smooth muscle relaxation (antihypertensive and antiarrhythmics effects) [22]. Furthermore, quercetin exerts antiviral and carcinostatic effects [27,39]. In the present study we have investigated the effect of quercetin against oxidative damage on nuclear (nDNA) and mitochondrial DNA (mtDNA) with the aim of filling existing gaps in the literature of flavonoids. Up today, in fact, it is available papers on the protection of quercetin only against nuclear– injured DNA mainly investigated by comet assay [53–55]. While it is not known whether quercetin protects from functionally relevant oxidative-induced mutations of mtDNA. In this study molecular approaches different from comet assay have been applied in cell-free and cell models in order to evaluate the possible genoprotective effect of quercetin using conditions of oxidative stress induced by various systems such as thiol/Fe 3+ /O2 mixed-function oxidase (MFO), hydrogen peroxide and UVC radiations. 2. Materials and methods 2.1. DNA and chemicals The plasmid pGEM-T (3000 bp) was purchased from Promega. All reagent grade chemicals were obtained from Sigma–Aldrich Inc. All primers were obtained from Sigma–Genosys Inc. and DL-dithiothreitol (DTT) from Clontech. 6-Hydroxy-2,5,7,8-tetramethylchroman-2-carboxylic acid (trolox) obtained from Aldrich, H2 O2 , acquired from Sigma, Fe3+ and quercetin, purchased from Fluka, were freshly prepared before each experiment. 2.2. Plasmid DNA study pGEM-T DNA (0.2 µg) was mixed with freshly prepared DTT (10 mM) and FeCl 3 (3 µM) (thiol/Fe3+ /O2 mixed-function oxidase, MFO system) in the presence or absence of different quercetin concentrations (from 0.01 µM to 100 µM of quercetin dissolved in DMSO), in a final volume of 20 µl of 40 mM HEPES (pH 7). Reaction mixtures were incubated for 200 min at 37 ◦C. Same quantities of plasmid DNA in a final volume of 20 µl of 40 mM HEPES (pH 7) were UVC irradiated for 15 min, 30 min and 40 min. pGEM-T DNA was also exposed to MFO or UVC, at the same conditon described above, in the presence of different trolox concentrations (from 0.01 µM to 100 µM in ethanol) how positive control. L. Potenza et al. / Effect of quercetin on oxidative nuclear and mitochondrial DNA damage 35 DNA samples were applied to 0.8% agarose/Tris Borate EDTA (TBE) gel, stained with ethidium bromide (0.3 µg/ml) and visualized under UV light. Quantification was made by densitometric analysis using Quantity One Software 4.01 (Bio-Rad). 2.3. Cell culture and treatment conditions Cellular experiments were carried out on normal human keratinocytes (NCTC 2544). Cells were cultured at 37◦ C in an atmosphere of 95% air and 5% CO 2 in DMEM medium containing 1% streptomycin/penicillin antibiotics, 2% glutamine, 7% fetal bovine serum and were seeded at an appropriate density 24 h before treatments. At the oxidative challenge stage, the cell number was about 6 × 10 5 cells/well. Quercetin was dissolved in dimetylsulfoxide (DMSO), which never exceeded 0.5%. Quercetin solution was added to cells in Hank’s solution (0.015 g/l CaCl 2 , 8 g/l NaCl, 0.4 g/l KCl, 1.6 g/l glucose, 0.06 g/l KH2 PO4 , 0.048 g/l Na2 HPO4 , 0.35 g/l NaHCO3 , 0.2 g/l MgSO4 · 7H2 O, pH 7.4) and incubated for 1 h at 37◦ C. After this time the medium was removed and cells washed twice with phosphate-buffered saline solution (PBS, 8 g/l NaCl, 1.44 g/l Na 2 HPO4 , 0.2 g/l KH2 PO4 , 0.2 g/l KCl, pH 7.4) before the challenge with H2 O2 or UVC. Oxidative challenge consisted in: i) 60 min incubation of quercetin-free or quercetin-supplemented cells with 500 µM H 2 O2 at 37◦C in 2 ml of Hank’s solution; ii) 30 sec or 1 min exposition to UVC. After oxidative treatments, cells were washed with PBS, harvested by gentle scraping and processed for DNA analyses. 2.4. Cytotoxicity assay by LDH release The release of lactate dehydrogenase (LDH; EC 1.1.1.27) in the incubation medium is an index of membrane integrity leakage. The LDH activity in the medium and in the cell lysate was evaluated by the disappearance of NADH during LDH-catalysed conversion of pyruvate to lactate as a decrease in absorbance at 340 nm [6]. The release of LDH was evaluated after cell incubation with quercetin (0–400 µM) for 1 h at 37◦ C or after treatment with H2 O2 (1 h at 37◦C) and UVC (30 sec and 1 min). It was calculated by dividing the activity in the media samples by total LDH activity (medium + lysate) of each sample. Cell viability was determined by dividing the release of LDH in treated samples by the release of LDH in control sample expressed as percentage [30]. 2.5. Isolation of total DNA High molecular weight DNA was isolated with the QIAamp DNA mini kit (Qiagen) according to the manufacturer’s instructions. Total cellular DNA concentration was determined at 260 nm using spectrophotometer (Beckman DU-640). 2.6. Long PCR and Quantitative PCR (QPCR) Long PCR was performed in a final volume of 25 µl using a Thermocycler model 9700 (Perkin-Elmer). Specific primers were used to amplify a 16.2 kb fragment of the mitochondrial DNA (primers Af-Ar, long mtDNA PCR, Table 1) and a 13.5 kb fragment of the nuclear gene β globin (primers Bf-Br, long nDNA PCR, Table 1). The reaction mixture contained 50 ng or 100 ng template total DNA (to amplify mtDNA and nDNA respectively), 2.5 µl buffer 1, 200 µM dNTPs, 0.5 µM of each primer and 0.5 µl of Advantage 2-polymerase (Clontech). PCR parameters are reported in Table 2. 36 L. Potenza et al. / Effect of quercetin on oxidative nuclear and mitochondrial DNA damage Table 1 Primers employed 16.2 kb, mitochondrial fragment Af 5’-TGAGGCCAAATATCATTCTGAG Ar 5’-TTTCATCATGCGGAGATGTTGG 13.5 kb, nuclear DNA of the β globin gene Bf 5’-CGAGTAAGAGACCATTGTGGC Br 5’-GCACTGGCTTAGGAGTTGGAC 100 bp, mitochondrial fragment Cf 5’-ACGCCATAAAACTCTTCACCA AAG-3’ Cr 5’-TAGTAGAAGAGCGATGGTGAGAGCTA-3’ Table 2 Thermal cycling parameters in QPCR 16.2 kb mitochondrial fragment 13.5 kb nuclear fragment Cycles 1x 24x 1x 1x 1x 30x 1x 1x Temperature 95◦ C 94◦ C 68◦ C 68◦ C 4◦ C 95◦ C 94◦ C 68◦ C 68◦ C 4◦ C Time 1 min 30 sec 16 min 7 min ∞ 1 min 30 sec 13 min 12 min ∞ The amplification products obtained by long PCR were electrophoresed on 0.8 % agarose/TBE gel, stained with ethidium bromide (0.3 µg/ml) and quantified by densitometric analyses of the intensity of bands using Quantity One Software 4.01 (Bio-Rad). Treated samples were then compared with controls and the relative amplification was calculated according to Santos et al. [42]. Results presented herein are the mean of two sets of PCR for each target of at least three different biological experiments. 2.7. Lesion frequency calculation f(x) = e−λ λx /x! Poisson expression Zero Class : f(0) = eλ λ = lesion frequency AD = Amplification of damaged template AC = Amplification of non-damaged template Lesion frequency/genomic strand: λ = −ln A D / AC 2.8. Quantitative real-time PCR Long mtDNA and PCR products, obtained from keratinocytes DNA, were also quantified by Sybr Green Real-Time PCR, using primers Cf-Cr, localized in the middle of the long PCR fragments (Table 1). After long amplification, samples were diluted 10 −3 , while the corresponding genomic DNA samples were diluted 10−2 . Quantitative Real-Time PCR was performed in a Bio-Rad iCycler iQ Multi-Color Real-Time PCR Detection System using 2X Quantitect SYBR Green PCR kit (Qiagen). The quantitative PCR reaction was performed at 95◦ C for 10 min to activate HotStart DNA polymerase followed by 50 cycles of the L. Potenza et al. / Effect of quercetin on oxidative nuclear and mitochondrial DNA damage 37 Table 3 Thermal cycling parameters in Real-Time PCR 100 bp, mitochondrial Cycles 1x 40x 1x Temperature 95◦ C 95◦ C 60◦ C 4◦ C Time 10 min 30 sec 30 sec ∞ two-step at 95◦ C for 30 sec and at 60 ◦C for 30 sec (Table 3). The specificity of the amplification products obtained was confirmed by examining thermal denaturation plots and by sample separation in a 3% DNA agarose gel. Results were normalized by quantitating each sample for the amount of initial genomic DNA without previous long PCR amplification, in the same Real-Time PCR conditions. Each sample was tested in triplicate, and the experimental groups (control DNA, H 2 O2 treated DNA with or without quercetin) consisted of at least three independent experiments. 2.9. Statistical analysis For statistical analysis software SPSS12.0 for Windows package was used. Wilcoxon signed-rank test was applied for comparisons among the data obtained from controls, H 2 O2 treated and H2 O2 treated + quercetin groups. A < 5% probability was considered as significant. 3. Results 3.1. Plasmid relaxation assay We investigated the effect of quercetin on DNA in cell-free systems. Untreated plasmid contained 80–98%, 20–2%, and 0% of the supercoiled (CCC covalently closed circular), open circular (OC), and linear forms, respectively (Fig. 1A and B). Thiol/Fe 3+ /O2 mixed-function oxidase (MFO) and UVC rays were used to produce oxidative DNA damage. In this type of experiments MFO oxidative system was preferred to H 2 O2 challenge because it quickly generates oxygen radical species in vitro. Aliquots of 0.2 µg pGEM-T plasmid were used in order to investigate and assess the induction of strand breaks in the supercoiled form, which in DNA preparations from prokaryotic cells is the prevalent one. Control DNA, MFO and UVC treated samples, in the absence or presence of different concentration of quercetin, were assayed through 0.8% agarose gel electrophoresis to quantify the plasmid topoforms. Electrophoresis analysis showed that both oxidation systems perturbed plasmid stability with a drastic reduction of the supercoiled form and an increase of open circular (OC) form, indicating the development of single strand breaks. In the experiments herein presented did not appear linear form (L), index of double strand breaks, evident after a more prolungated exposition time (data not shown). Conversely all the concentrations of quercetin herein used without oxidant systems did not perturb plasmid stability and results did not vary from control suggesting that this concentrations does not act as pro-oxidant (data not shown). The addition of quercetin in MFO system resulted in the protection of the prevalent topoform CCC starting from 0.5 µM and in a dose-dependent manner. Almost total protection was obtained at 100 µM 38 L. Potenza et al. / Effect of quercetin on oxidative nuclear and mitochondrial DNA damage M 1 2 3 4 5 6 7 8 OC CCC A CCC, OC DNA (%) 100 CCC 80 OC 60 40 20 50 M FO +Q 10 0 FO +Q M M FO +Q 25 15 5 FO +Q M 0. 4 FO +Q M FO FO +Q M M Co nt ro l 0 B Fig. 1. Protective effect of quercetin at different concentrations on plasmid DNA strand breaks induced by the MFO system (A, B). (A) Example of electrophoretic pattern of total plasmid DNA (CCC + OC) obtained from MFO and quercetin treatment. Lane M: DNA molecular weight λ/Hind III; lane 1: control DNA; lane 2: 200 min MFO-treated DNA; lane 3: 200 min MFO and 100 µM quercetin treated DNA; lane 4: 200 min MFO and 50 µM quercetin treated DNA; lane 5: 200 min MFO and 25 µM quercetin treated DNA; lane 6: 200 min MFO and 15 µM quercetin treated DNA; lane 7: 200 min MFO and 5 µM quercetin treated DNA; lane 8: 200 min MFO and 0.4 µM quercetin treated DNA; (B) Quantitation of quercetin protection against DNA strand breaks induced by MFO system. Results are expressed as the percentage of the two forms on total plasmid DNA (CCC + OC). Values are means ± S.D. of results from at least three replicates. (Fig. 1B). Instead when UVC system was used partial inhibition of the conversion of supercoiled to open circular forms was observed when concentrations grater than 0.1 µM quercetin were used. Total protection of the CCC form was never obtained in our conditions but the highest one allowed a yield of 50% of this form (Fig. 2B). In order to understand how strong is the effect of quercetin, experiments with trolox (25 µM, 50 µM and 100 µM), a well-known radical scavenger, were also carried out using both oxidative systems. When DNA was injuried with MFO system quercetin resulted more protective than trolox (Figs 1B and 3). Instead, when the oxidative system was UVC the protective effect of trolox on DNA was similar to quercetin, but it showed pro-oxidant properties over 25 µM. 3.2. Cell assays Results obtained in acellular assays did not consider the intracellular conditions. Therefore we evaluated the effect of quercetin on nuclear and mitochondrial DNAs in a cellular system, consisting of human keratinocytes (NCTC 2544), and hydrogen peroxide or UVC as sources of oxidative damage. L. Potenza et al. / Effect of quercetin on oxidative nuclear and mitochondrial DNA damage M 1 2 3 4 5 6 7 39 8 OC CCC A CCC, OC DNA (%) 100 CCC OC 80 60 40 20 Co nt ro l UV C 15 ’ U VC +Q 5 UV C+ Q 10 UV C+ Q 15 UV C+ Q 20 UV C+ Q 25 U VC +Q 50 UV C+ Q 10 0 0 B Fig. 2. Protective effect of quercetin at different concentrations on plasmid DNA strand breaks induced by 15 min UVC (A, B). (A) Electrophoretic pattern of total plasmid DNA (CCC + OC) obtained from 15 min UVC and quercetin treatment. Lane M: DNA molecular weight λ/Hind III; lanes 1: control DNA; lane 2: 15 min UVC treated DNA; lane 3: 15 min UVC and 100 µM quercetin treated DNA; lane 4 :15 min UVC and 50 µM quercetin treated DNA; lane 5: 15 min UVC and 25 µM quercetin treated DNA; lane 6: 15 min UVC and 20 µM quercetin treated DNA; lane 7: 15 min UVC and 15 µM quercetin treated DNA; lane 8: 15 min UVC and 5 µM quercetin treated DNA. (B) Quercetin protection against DNA strand breaks was deduced by densitometric analysis of bands from electrophoretic gels. Results are expressed as the percentage of the two forms on total plasmid DNA (CCC + OC). Values are means ± S.D. of results from at least three replicates. 3.2.1. Cytotoxicity assay by LDH release The cytotoxicity of quercetin on NCTC 2544 was evaluated as described in Materials and Methods section. It was assayed in a concentration range of 0–400 µM (Fig. 4), in which the viability of keratinocytes varied from 100% to 62% at the minimum and the maximum quercetin concentrations respectively. Cell viability was also investigated after treatment with hydrogen peroxide and UVC. Hydrogen peroxide led viability to about 80%, the UVC to about 60% and 80% respectively for treatment of 1 min and 30 sec. 3.2.2. QPCR The effect of quercetin on DNA extract from oxidatively injuried keratinocytes with hydrogen peroxide and UVC was investigated. The effect of this molecule was evaluated on both types of human DNAs: nuclear DNA (nDNA) and mitochondrial DNA (mtDNA), which have different sizes and conformations, being double strand linear and double strand circular molecules, respectively. Total genomic DNA was extracted from NCTC 2544 preincubated for 1 h with or without quercetin and then treated for 1 h with 500 µM H 2 O2 or exposed for 30 sec and 1 min to UVC. Quercetin was L. Potenza et al. / Effect of quercetin on oxidative nuclear and mitochondrial DNA damage 100 CCC 80 OC 60 40 20 50 FO +T r FO +T r A M M M FO +T r 25 FO M Co nt 10 0 0 ro l CCC, OC DNA (%) 40 CCC, OC DNA (%) 100 CCC OC 80 60 40 20 VC +T r1 00 B U +T r5 0 VC U U VC +T r2 5 15 ’ U VC C on tr ol 0 Fig. 3. Effect of trolox, a reference of radical scavenger, at different concentrations on plasmid DNA strand breaks induced by MFO (A) and 15 min UVC (B). 120 % Viability 100 80 60 40 20 0 0 50 100 150 200 300 400 Quercetin ( M) Fig. 4. Cell viability following quercetin incubation. NCTC 2544 cells were incubated with quercetin (0–400 µM) for 1 h at 37◦ C. LDH activity was determined in media and lysates as described in Materials and Methods section. Each point represents mean ± S.D. of five independent experiments. used at concentrations of 25, 50, 100 µM. Two long fragments of nDNA and mtDNA were analysed through QPCR. This technique is based on the premise that DNA lesions, including oxidative damage such as strand breaks, abasic sites and some base modifications (for example, 8oxodA) block the progression of any thermostable polymerase on the L. Potenza et al. / Effect of quercetin on oxidative nuclear and mitochondrial DNA damage M 1 2 3 4 5 6 7 8 9 10 11 41 12 A M 1 2 3 4 5 6 7 8 9 10 11 12 B Fig. 5. Examples of electrophoretic patterns showing protective effect of quercetin on nuclear DNA (A) and mitochondrial DNA (B). Keratinocytes were pre-incubated for 1 h in the absence or presence of quercetin, treated for 60 min in Hank’s solution with 500 µM H2 O2 and immediately assayed for QPCR. (A) Lane M: DNA molecular weight λ/Hind III; lanes 1,2: control amplified products; lanes 3,4: amplified products from H2 O2 (500 µM) treated cells; lanes 5-12: amplified products from H2 O2 treated cells in presence of quercetin (50 µM).; (B) Electrophoretic profile of a 16.25 kb mtDNA region PCR.; Lane M: DNA molecular weight λ/Hind III; lanes 1,2,7,8: control amplified products; lanes 3,4,9,10: amplified products from H2 O2 (500 µM) treated cells; lanes 5,6,11,12: amplified products from H2 O2 treated cells in presence of quercetin (25 µM). template [43,45]. The reduction in the relative yield of the long PCR product reflects the presence of blocking DNA polymerase lesions [9,42] rather than the exhaustion of a critical reagent, such as dNTPs, primers or enzyme. Thus, amplification is inversely proportional to DNA damage: greater is the lesion on DNA target, lesser is the amplification. The usefulness of QPCR assay for the detection of DNA damage requires that amplification yields are directly proportional to the starting amount of template, condition met by keeping the PCR in the exponential phase when the other components of the reaction (dNTPs, primers and Taq polymerase), are not limiting. Furthermore, in order to obtain quantitative PCR (QPCR) the amplification products from long PCR must be quantified by densitometric analysis, after gel electrophoresis migration. The values utilised for statistical analyses were the relative amplifications, obtained dividing fluorescence values of the treated samples by the control (non treated samples). Results of the experiments with hydrogen peroxide are shown in Figs. 5 and 6. In Fig. 5 electrophoretic gel represents an example of the analyses carried out. Treatment with 500 µM H 2O2 resulted in significant reductions in the amplification of both the β -globin gene and mtDNA; quercetin supplementation reduced the amount of damage. The analyses of the two genomes showed that oxidative challenges induce more extensive damage in mtDNA (Fig. 6A) than in nDNA (Fig. 6B). The protective effect of quercetin was statistically significant only on mtDNA at each used concentration and in a dose-dependent manner. The values of the relative amplifications were also used to determine the lesion frequency per fragment at a particular dose (Table 4), such that lesions/strand (average for both strands) at dose D = −ln (AD /AC ). This equation is based on the “zero class” of Poisson expression, where Poisson distribution requires an assumption that DNA lesions are randomly distributed [3]. The analyses after cells irradiation with UVC show that the damage was too severe on both genomes and consequently there was no protection by quercetin. Electrophoretic patterns showed no amplification product in absence or presence of quercetin (not shown). L. Potenza et al. / Effect of quercetin on oxidative nuclear and mitochondrial DNA damage Relative Amplification 42 H2O2 H2O2/Q 1,2 1 * * 0,8 0,6 0,4 * 0,2 0 25 50 100 A Quercetin concentration ( M) H2O2 H2O2/Q Relative Amplification 1,2 1 0,8 0,6 0,4 0,2 0 25 50 Quercetin concentration ( M) 100 B Fig. 6. Densitometric analysis of the electrophoretic profiles of mitochondrial (A) and nuclear (B) Long PCR from quercetin-supplemented (25, 50, 100 µM) and unsupplemented keratinocytes. The decrease in amplification was calculated comparing treated samples to undamaged control (dashed line). Data are expressed as the mean ± S.E.M. of at least five independent determinations in which two Long PCRs were performed per experiment. Wilcoxon signed-rank test was performed comparing controls to H2 O2 treated groups or quercetin untreated to treated groups. P < 0.05 when controls and H2 O2 treated groups were compared; * P < 0.05 comparing DNA from H2 O2 treated cells to DNA from quercetin supplemented H2 O2 treated cells. 3.2.3. Real-Time PCR after hydroperoxide challenge We also applied a more sophisticated protocol previously developed [25], in order to assess whether the QPCR technique provides reliable results. We used a two-step strategy, the first of which consists of a Long PCR while the second of a Real-Time PCR. In this latter a sequence of 100 bp was amplified with primers designed in the central portion of the Long segment. As amplification targets were used the amplification product and genomic DNA from the same sample used for Long to normalize the results. This two-step assay was selected because a standard Real-Time PCR using total DNA, without previous long PCR amplification, produces overlapping curves either with or without induced DNA damages, since the amplification product is too short to show DNA polymerase proofreading blocking. Only mtDNA from control cells and H2 O2 injured cells, without or with 50 µM quercetin, was analysed with this strategy. Primers Af-Ar were used for Long PCR and primers Cf-Cr were selected from this fragment (Table 1). Standard curve has been established between DNA quantities used as templates in the long PCR and PCR amplicons detected by Real-Time PCR. Results were normalized by quantifying each sample for the amount of initial DNA without previous long PCR amplifications. L. Potenza et al. / Effect of quercetin on oxidative nuclear and mitochondrial DNA damage 43 Table 4 Lesion frequency calculation mtDNA Treated samples H2 O2 H2 O2 /Q25 µM H2 O2 H2 O2 /Q50 µM H2 O2 H2 O2 /Q100 µM Relative amplification 0.1293 0.2192 0.326 0.687 0.23 0.766 Lesions/16.2 Kb 2.05 1.52 1.12 0.38 1.47 0.27 Lesions/10 Kb 1.26 0.94 0.69 0.23 0.91 0.16 nDNA Treated Samples H2 O2 H2 O2 /Q25 µM H2 O2 H2 O2 /Q50 µM H2 O2 H2 O2 /100 µM Relative Amplification 0.527 0.5019 0.592 0.687 0.53 0.7265 Lesions/13.5 Kb 0.64 0.70 0.52 0.38 0.63 0.32 Lesions/10 Kb 0.47 0.51 0.39 0.28 0.47 0.24 Relative Amplification 1,2 * 1 0,8 0,6 0,4 0,2 0 H2O2 H2O2/Q50 Fig. 7. Analysis of quercetin effect on mtDNA damage obtained by combining Long-PCR with Real-Time PCR from control cells, 500 µM H2 O2 treated in absence or presence of 50 µM quercetin (H2 O2 /Q50). Results were from 5 independent experiments and are expressed as relative PCR amplification. Wilcoxon signed-rank test was performed comparing controls to H2 O2 treated groups or quercetin untreated to treated groups. P < 0.05 when controls and H2 O2 treated groups were compared; ∗ P < 0.05 comparing mtDNA from H2 O2 treated cells to mtDNA from quercetin supplemented H2 O2 treated cells. Using this strategy, the amplification curves of samples treated with hydrogen peroxide moved to the right compared to the curves of the control samples, reflecting the least amount of product obtained with the Long PCR and therefore the presence of injuries that inhibit DNApol; curves obtained from quercetin supplemented samples shifts to controls reflecting its protection and minor DNA damage on mitochondrial DNA with respect to H2 O2 treated samples. The results obtained with this technique (Fig. 7) agreed with those obtained with QPCR (Fig. 6A), showing the same level of statistical significance and demonstrating that the QPCR is an equally reliable technique. 4. Discussion In the present work the effect of quercetin on nDNA and mtDNA in different conditions of oxidative stress has been assessed. Both types of DNA were investigated in that they coexist in eukaryotic cells 44 L. Potenza et al. / Effect of quercetin on oxidative nuclear and mitochondrial DNA damage and have different sensitivity, perhaps due to different location and conformation [25]. Nuclear genome consists of linear molecules located in the nucleus, while the human mitochondrial genome, completely sequenced in 1981 [2], is a 16.569 bp closed circular double strand molecule, present in a high copy number per cell and widely varying among cell types. mtDNA has been observed to be more susceptible to damage than nuclear DNA because of several factors, such as exposure to high levels of reactive oxygen species (ROS) produced during oxidative phosphorilation [57], lack of protective histones, and limited DNA repair pathways [7] having a robust base excision repair (BER) system but not a nucleotide excision repair (NER) one [44]. Oxidative damage to mtDNA may lead to loss of membrane potential, reduced ATP synthesis and cell death [52]. Literature data regarding the effects of quercetin on mtDNA are not currently available. In order to fill this gap, cell-free and cellular assays were performed in this study. Plasmid pGEM-T was choosen like cell-free model, human keratinocytes NCTC 2544 were selected as cell line and MFO (Thiol/Fe 3+ /O2 mixed-function oxidase), hydrogen peroxide and UVC rays as oxidative systems. Keratinocytes were selected in that they represent the first line of defence against environmental stress, such as chemical pollution and solar radiation and their responsiveness, assume a decisive role in skin and body homeostasis. Keratinocytes are in fact actively involved in regulating electrolytic balance and thermoregulation. Therefore, their inability to appropriately respond to the chemical and physical environment can induce to physiological disorders. In the MFO system, autoxidation of thiols (DTT) in the presence of iron generates ROS such as superoxide anion, H 2 O2 and hydroxyl radical (OH·) [31]. In particular, hydroxyl radical generated in close proximity to nucleic acid molecules can add hydrogen atoms to DNA bases leading to modified bases, DNA strand breaks or abasic sites [12]. The hydrogen peroxide was chosen as oxidative system since it generates the same ROS formed in physiological or stress conditions. Keratinocytes exposed to hydrogen peroxide are often used as a model to simulate the damage generated by the ROS and by UV rays to the skin [47]. UVC were selected as further oxidative system in order to achieve more information about the possible protection of quercetin, which is able to protect from UVA and UVB damages [10,11,21,58]. In fact in recent decades, the gradual depletion of the ozone layer due to the increasing of environmental pollution, sparked alarmism and suggest incomplete fulfillment of its functions to filter the sun’s rays, which cause the harmful effects of UVC. In cell-free assays performed on plasmid DNA the protective effect of quercetin from oxidative injuries was shown. Results achieved are in agreement with other data on the assessment of DNA strand breaks obtained using either gel electrophoresis or other techniques that measure the average value of the extent of the lesions [1,5,24,33,35,46]. Cellular experiments showed that quercetin exerts genoprotective activity on mitochondrial DNA when exposed to oxidative damage by hydrogen peroxide. To date there is no evidence in literature that investigate the effect of quercetin on both genomes. A lot of data are available regarding the quercetin effect on nuclear DNA by measuring nonspecific DNA damage by comet assay [8,14–16,18,20,29,32,40,48–51,54–56,59]. This technique, however, though sensitive, fast and easy, highlights only damages in terms of break of single or double strand of nuclear DNA. The strategies herein used (QPCR and Real-Time PCR) instead, allowed us to detect, in addition to single and double strands breaks, a wider range of lesions caused by oxidative stress, such as modified bases or sites AP. Furthermore these techniques present the main advantage to allow the monitoring of the integrity of mtDNA directly from total cellular DNA avoiding processes which can increase base oxidation such as isolation of mitochondria or separate mtDNA purification steps. L. Potenza et al. / Effect of quercetin on oxidative nuclear and mitochondrial DNA damage 45 The results obtained quantifying the oxidative damage of mtDNA and nDNA from NCTC 2544 cells were consistent with previous researches performed on different human and rodent cell types, pointing out that mtDNA is much more prone to oxidative damage than nuclear DNA [4,15,25,34]. The protective effect of quercetin on mtDNA supports the possibility that this molecule enters into the mitochondria, capacity not given to most conventional antioxidants. Previous works carried out on rat liver [17] and bovine heart mitochondria [23] have demonstrated that quercetin reaches these organelles interacting with their membranes [17] and the F(1)-ATPase, enzymatic complex with mitochondrial localization [23]. In the present work we carried out the analyses on cells pre-treated with quercetin before exposure to the oxidant systems without the recovery period and found that quercetin prevents DNA damage, with statistical significance at mitochondrial level. The potential mechanism of action through which the quercetin inhibits the DNA damage has been reported by Kanakis et al. [28] which studied the interaction of quercetin with DNA in aqueous solution at physiological conditions and found from the structural analysis that this molecule intercalates DNA. The intercalated quercetin can make it strong antioxidant to protect DNA from harmful free radical reactions. Concerning the damage caused by UV rays, quercetin has already proven to increase the stability of β caroten preventing the prooxidant effect that would lead, after exposure to UVA, genomic damage [55]. Moreover, in a recent study, the application of topical formulations containing quercetin on hairless skin of mice, showed its effectiveness inhibiting UVB oxidative insults on the skin [10]. However the mechanism that prevents the quercetin to protect from UVC rays is not known. In our cellular assays quercetin was ineffective in preventing UVC damage in both types of genomes. A possible explanation could be found in a study on A549 cells carried out in order to assess the role of overexpression of Heat Shock Proteins (Hsp70) in the protection against UVC induced DNA damage. In this study quercetin inhibited the expression of the above-mentioned genes with a consequent accumulation of genomic damage due to start failure of repair mechanisms normally carried on by the same Hsp [13]. It is important to underline that the quercetin concentration used in our cell studies to obtain protection, are far above physiological plasmatic levels where quercetin is mostly present in conjugated form (glucoronide and sulphate conjuagates) [8]. However it has been reported that such quercetin conjugates can be deconjugates by enhanced beta-glucuronidase activity during inflammation [37,38] and such information might support the possibility that quercetin aglycone acts as an antioxidants on genomes as above explained. In conclusion, the results herein reported highlight the protective effect of quercetin on circular molecule of DNA under MFO and hydroperoxide challenges. The findings here reported give an important and innovative input to scientific field about mtDNA studies and its protection from oxidative damages and make quercetin a promising molecule for pharmacological studies for the wide range of pathologies where mitochondrial oxidative damage is known to play an etiological role. Further in vivo studies, using animal models or human volunteers are, however, needed in order to evaluate whether quercetin intake will ever lead to a significant similar effect in humans. References [1] [2] W. Adam, J. Hartung, H. Okamoto, C.R. Saha-Moller and K. Spehar, N-hydroxy-4-(4-chlorophenyl)thiazole-2(3H)thione as a photochemical hydroxyl-radical source: photochemistry and oxidative damage of DNA (strand breaks) and 2’-deoxyguanosine(8-oxodG formation), Photochem Photobiol 72 (2000), 619–624. S. Anderson, A.T. Bankier, B.G. Barrell, M.H.L. deBruijn, A.R. Coulsen, J. Drouin, I.C. Eperon, B.A. Roe, F. Sanger, P.H. Scheir, R. Staden and I.G. Young, Sequence and organization of the human mitochondrial genome, Nature 290 (1981), 457–46. 46 [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] L. Potenza et al. / Effect of quercetin on oxidative nuclear and mitochondrial DNA damage S. Ayala-Torres, Y. Chen, T. Svoboda, J. Rosenblatt and B. Van Houten, Analysis of gene-specific DNA damage and repair using quantitative polymerase chain reaction, Methods 22 (2000), 135–147. S.W. Ballinger, B. Van Houten, G.F. Jin, C.A. Conklin and B.F. Godley, Hydrogen peroxide causes significant mitochondrial DNA damage in human RPE cells, Exp Eye Res 68 (1999), 765–772. I. Banmeyer, C. Marchand, A. Clippe and B. Knoops, Human mitochondrial peroxireodin 5 protects from mitochondrial DNA damages induced by hydrogen peroxide, FEBS Lett 579 (2005), 2327–2333. E. Beutler, Lactate Dehydrogenase (LDH), in: Red Cell Metabolism, A Manual of Biochemical Methods, (3rd ed.), Grune and Stratton, eds, Harcourt B. Jovanovich, New York, 1957, pp. 65–66. V.A. Bohr, Repair of oxidative DNA damage in nuclear and mitochondrial DNA, and some changes with aging in mammalian cells, Free Radical Biol Med 32 (2002), 804–812. A.W. Boots, H. Li, R.P.F. Schino, R. Duffin, J.W.M. Heemskerk, A. Bast and G.R.M.M. Haenen, The quercetin paradox, Toxicol Appl Pharm 222 (2007), 89–96. J. Cao, L. Jia, H.M. Zhou, Y. Liu and L.F. Zhong, Mitochondrial and nuclear DNA damage induced by curcumin in human hepatoma G2 cells, Toxicol Sci 91 (2006), 476–483. R. Casagrande, S.R. Georgetti, W.A. Jr Verri, D.J. Dorta, A.C. dos Santos and M.J. Fonseca, Protective effect of topical formulations containing quercetin against UVB-induced oxidative stress in hairless mice, J Photoch and Photobio B 84 (2006), 21–27. R. Casagrande, S.R. Georgetti, W.A. Jr Verri, D.J. Dorta, A.C. dos Santos and M.J. Fonseca, In vitro evaluation of quercetin cutaneous absorption from topical formulations and its functional stability by antioxidant activity, J Photoch and Photobio B 328 (2007), 183–190. M.S. Cooke, M.D. Evans, M. Dizdaroglu and J. Lunec, Oxidative DNA damage: mechanisms, mutation, and disease, FASEB J 17 (2003), 1195–1214. N.A. Cridland, M.C. Martin, K. Stevens, C.A. Baller, A.J. Pearson, C.M. Driscoll and R.D. Saunders, Role of stress responses in human cell survival following exposure to ultraviolet C radiation, Int J Radiat Biol 77 (2001), 365–374. S. De, C. Ganguly and S. Das, Natural dietary agents can protect against DMBA genotoxicity in lymphocytes as revealed by single cell gel electrophoresis assay, Teratog Carcinog Mutagen Suppl 1 (2003), 71–78. G. Deng, J.H. Su, K.J. Ivins, B. Van Houten and C.W. Cotman, Bcl-2 facilitates recovery from DNA damage after oxidative stress, Exp Neurol 159 (1999), 309–318. M.M. Dobrzyńska, A. Baumgartner and D. Anderson, Antioxidants modulate thyroid hormone- and noradrenalineinduced DNA damage in human sperm, Mutagenesis 19 (2005), 325–330. D.J. Dorta, A.A. Pigoso, F.E. Mingatto, T. Rodrigues, I.M. Prado, A.F. Helena, S.A. Uyemura, A.C. Santos and C. Curti, The interaction of flavonoids with mitochondria: effects on energetic processes, Chem Biol Interact 152 (2005), 67–78. R. Edenharder, J.W. Sager, H. Glatt, E. Muckel and K.L. Platt, Protection by beverages, fruits, vegetables, herbs, and flavonoids against genotoxicity of 2-acetylaminofluorene and 2-amino-1metil-6-phenylimidazo [4,5-b] pyridine (PhIP) in metabolically competent V79 cells, Mutat Res 521 (2002), 71–78. I. Erlund, Review of the flavonoids quercetin, hesperetin and narigenin. Dietary sources, bioactivities, bioavality and epidemiology, Nutr Res 24 (2004), 851–874. R. Fagiani, A. De Bartolomeo, P. Rosignoli and G. Morozzi, Antioxidants prevent the lymphocyte DNA damage induced by PMA-stimulated monocytes, Nutr Cancer 39 (2002), 284–291. P. Filipe, J.N. Silva, J. Haigle, J. Freitas, A. Fernandes, R. Santus and P. Molière, Contrasting action of flavonoids on phototoxic effects induced in human skin fibroblast by UVA alone or UVA plus cyamemazine, a phototoxic neuroleptic, Photoch Photobio Sci 4 (2005), 420–428. J.V. Formica and W. Regelson, Review of the Biology of quercetin and related bioflavonoids, Food Chem Toxicol 33 (1995), 1061–1080. J.R. Gledhill, M.G. Montgomery, A.G. Leslie, J.E. Walker, Mechanism of inhibition of bovine F1-ATPase by resveratrol and related polyphenols, Proc Natl Acad Sci USA 104 (2007), 13632–13637. B.F. Godley, F.A. Shamsi, F.Q. Liang, S.G. Jarrett, S. Davies and M. Boulton, Blue light induces mitochondrial DNA damage and free radical production in epithelial cells, J Biol Chem 280 (2005), 21061–21066. C. Guidi, L. Potenza, P. Sestili, C. Martinelli, M. Guescini, L. Stocchi, S. Zeppa, E. Polidori, G. Annibalini and V. Stocchi, Differential effect of creatine on oxidatively-injured mitochondrial and nuclear DNA, Biochim Biophys Acta 1780 (2008), 16–26. K.E. Heim, A.R. Tagliaferro and D.J. Bobilya, Flavonoid antioxidants: chemistry, metabolism and structure-activity relationships, J Nutr Biochem 13 (2002), 572–584. K. Igura, T. Ohta, Y. Kuroda and K. Kaji, Resveratrol and quercetin inhibit angiogenesis in vitro, Cancer Lett 171 (2001), 11–16. C.D. Kanakis, P.A. Tarantilis, M.G. Polissiou, S. Diamantoglou, H.A. Tajmir-Riahi, An overview of DNA and RNA bindings to antioxidant flavonoids, Cell Biochem Biophys 49 (2007), 29–36. L. Potenza et al. / Effect of quercetin on oxidative nuclear and mitochondrial DNA damage [29] [30] [31] [32] [33] [34] [35] [36] [37] [38] [39] [40] [41] [42] [43] [44] [45] [46] [47] [48] [49] [50] [51] [52] [53] [54] 47 M. Kapiszewska, A. Cierniak, M. Elas and A. Lankoff, Lifespan of etoposide-treated human neutrophils is affected by antioxidant ability of quercetin, Toxicol In Vitro 21 (2007), 1020–1030. D.M. Kending and J.B. Tarloff, Inactivation of lactate dehydrogenase by several chemicals: implications for in vitro toxicology studies, Toxicol In Vitro 21 (2007), 125–132. K. Kim, I.H. Kim, K.Y. Lee, S.G. Rhee and E.R. Stadtman, The isolation and purification of a specific “protector” protein which inhibits enzyme inactivation by a thiol/Fe(III)/O2 mixed-function oxidation system, J Biol Chem 263 (1988), 4704–4711. N. Llópiz, F. Puiggròs, E. Céspedes, L. Arola, A. Ardévol and C. Bladé, Antigenotoxic effect of grape seed procyanidin extract in Fao cells submitted to oxidative stress, J Agr Food Chem 52 (2004), 1083–1087. G.M. Makrigiorgos, E. Bump, C. Huang, J. Baranowska-Kortylewicz and A.I. Kassis, A fluorimetric method for the detection of copper-mediated hydroxyl free radicals in the immediate proximity of DNA, Free Radical Biol Med 18 (1995), 669–678. B.S. Mandavilli, S.F. Ali and B. Van Houten, DNA damage in brain mitochondria caused by aging and MPTP treatment, Brain Res 885 (2000), 45–52. M. Meriyani Odyuo and R.N. Sharan, Differential DNA strand breaking abilities of OH· and ROS generating radiomimetic chemicals and gamma-rays: study of plasmid DNA, pMTa4, in vitro, Free Radical Res 39 (2005), 499–505. E.J. Middleton, Effect of plant flavonoids on immune and inflammatory cell function, Adv Exp Med Biol 439 (1998), 175–182. M. Mochizuki, K. Kajiya, J. Terao K. Kaji, S. Kumazawa, T. Shimoi, Effect of quercetin conjugates on vascular permeability and expression of adhesion molecules, Biofactors 22 (2004), 201–204. A. Murakami, H. Ashida and J. Terao, Multitargeted cancer prevention by quercetin, Cancer Lett, 2008 May 6, [Epub ahead of print]. N. Neznanov, A. Kondratova, K.M. Chumakov, L. Neznanova, R. Kondratov, A.K. Banerjee and A.V. Gudkov, Quercetinase Pirin Makes Poliovirus Replication Resistant to Flavonoid Quercetin, DNA Cell Biol 27 (2008), 191–198. P. Niu, L. Liu, Z. Gong, H. Tan, F. Wang, J. Yuan, Y. Feng, Q. Wei, R.M. Tanguay and T. Wu, Overexpressed heat shock protein 70 protects cells against DNA damage caused by ultraviolet C in a dose-dependent manner, Cell Stress Chaperon 11 (2006), 162–169. B. Pawlikowska-Pawlega, W.I. Gruszecki, L. Misiak, R. Paduch, T. Piersiak, B. Zarzyka, J. Pawelec and A. Gawron, Modification of membranes by quercetin, a naturally occurring flavonoid, via its incorporation in the polar head group, Biochim Biophys Acta 1768 (2007), 2195–2204. J.H. Santos, B.S. Mandavilli and B. Van Houten, Measuring oxidative mtDNA damage and repair using quantitative PCR, Methods Mol Biol 197 (2002), 159–176. J.H. Santos, L. Hunakova, Y. Chen, C. Bortner and B. Van Houten, Cell sorting experiments link persistent mitochondrial DNA damage with loss of mitochondrial membrane potential and apoptotic cell death, J Biol Chem 278 (2003), 1728– 1734. D.E. Sawyer and B. Van Houten, Repair of DNA damage in mitochondria, Mutat Res 434 (1999), 161–176. J.A. Sikorsky, D.A. Primerano, T.W. Fenger and J. Denvir, Effect of DNA damage on PCR amplification efficiency with the relative threshold cycle method, Biochem Biophys Res Commun 323 (2004), 823–830. Y.Y.M. Su and G. Yang, Quantitative measurement of hydroxyl radical induced DNA double-strand breaks and the effect of N-acetyl-L-cysteine, FEBS Lett 580 (2006), 4136–4142. A. Svobodová, D. Walterová and J. Psotová, Influence of Silymarin and its flavolignans on H2 O2 -induced oxidative stress in human keratinocytes and mouse fibroblast, Burns 32 (2006), 973–979. Y.T. Szeto and I.F. Benzie, Effects of dietary antioxidants on human DNA damage in lysed cells using a modified comet assay procedure, Free Radic Res 500 (2002), 31–38. Y.T. Szeto and I.F. Benzie, Effects of dietary antioxidants on human DNA ex vivo, Free Radic Res 36 (2002), 113–118. X.L. Tang, X.J. Liu, W.M. Sun, J. Zhao and R.L. Zheng, Oxidative stress in Graves’disease patients and antioxidant protection against lymphocytes DNA damage in vitro, Pharmazie 60 (2005), 696–700. J. Tieppo, R. Vercelino, A.S. Dias, M.F. Silva Vaz, T.R. Silveira, C.A. Marroni, N.P. Marroni, J.A. Henriques and J.N. Picada, Evaluation of the protective effects of quercetin in the hepatopulmonary syndrome, Food Chem Toxicol 45 (2007), 1140–1146. V.W.B. Van Houten and J.H. Santos, Role of mitochondrial DNA in toxic response to oxidative stress, DNA Repair 5 (2006), 145–152. L.C. Wilms, P.C. Hollman, A.W. Boots and J.C. Kleinjans, Protection by quercetin and quercetin-rich fruit juice against induction of oxidative DNA damage and formation of BPDE-DNA adducts in human lymphocytes, Mutat Res 582 (2005), 155–162. L.C. Wilms, T.A. Claughton, T.M. de Kok and J.C. Kleinjans, GSTM1 and GSTT1 polymorphism influences protection against induced oxidative DNA damage by quercetin and ascorbic acid in human lymphocytes in vitro, Food Chem Toxicol 45 (2007), 2592–2596. 48 [55] [56] [57] [58] [59] L. Potenza et al. / Effect of quercetin on oxidative nuclear and mitochondrial DNA damage L.C. Wilms, A.W. Boots, V.C. de Boer, L.M. Maas, D.M. Pachen, R.W. Gottschalk, H.B Ketelslegers, R.W. Godshalk, G.R. Haenen, F.J. van Schooten and J.C. Kleinjans, Impact of multiple genetic polymorphisms on effects of a 4-week blueberry juice intervention on ex vivo induced lymphocytic DNA damage in human volunteers, Carcinogenesis 28 (2007), 1800–1806. L.C. Wilms, J.C. Kleinjans, E.J. Moonen and J.J. Briedé Discriminative protection against hydroxyl and superoxide anion radicals by quercetin in human leucocytes in vitro, Toxicol In Vitro 22 (2008), 301–307. F.M. Yakes and B. Van Houten, Mitochondrial DNA damage is more extensive and persists longer than nuclear DNA damage in human cells following oxidative stress, Proc Natl Acad Sci USA 94 (1997), 514–519. S.L. Yeh, W.Y. Wang, C.H. Hang and M.L. Hu, Pro-oxidative effect of beta-carotene and the interaction with flavonoids on UVA-induced DNA strand breaks in mouse fibroblast C3H10T1/2 cells, J Nutr Biochem 16 (2005), 729–735. J. Zhao and X.J. Liu, Antioxidative and immunomodulatory role of melatonin, sodium selenite, N-acetyl-L-cysteine and quercetin on human umbelical blood, Pharmazie 60 (2005), 683–688.