Oxidative DNA Damage Induced by S ( IV ) in the Presence of Cu ( II ) and Cu ( I ) Complexes

O dano ao DNA induzido por S(IV) na presença de alguns complexos de Cu(II) em soluções saturadas com ar foi investigado. A adição de S(IV) a uma solução saturada com ar contendo CuGGA (GGA = glicilglicil-L-alanina), CuG 3 (G 3 = triglicina) ou CuG 4 (G 4 = tetraglicina) e traços de Ni(II) origina a formação rápida do respectivo complexo de Cu(III), com o simultâneo consumo de oxigênio e a oxidação de S(IV). SO 3 •− e HO foram detectados por experimentos de EPR-spin trapping. As quebras das fitas de DNA foram atribuídas aos radicais de óxido de enxofre formados. Na redução de Cu(II)/BCA (BCA = 4,4’ dicarboxi-2-2’-biquinolina) por S(IV), com a formação do complexo CuBCA, há a possível formação de um radical centrado em carbono do BCA ou um radical peróxido (ROO), capazes de oxidar as bases de DNA. A intensidade do dano ao DNA na presença desses complexos de Cu(II) e S(IV) (10-300 μmol L) seguiu a ordem: CuBCA ∼ CuG 4 ∼ Cu(II) (adicionado como Cu(NO 3 ) 2 ) > CuG 3 ∼ CuGGA. Especialmente para o CuBCA, o dano ocorreu mesmo em concentrações baixas de S(IV) (0,1 μmol L). Para os complexos de Cu(II) com glicilglicilhistidina, glicilhistidilglicina, glicilhistidillisina e glicilgliciltirosilarginina a formação de Cu(III) e o dano do DNA não foram observados.


Introduction
Copper peptide complexes, with low reduction potentials and high stability in aqueous solution, are of special interest in biological redox processes 1 due to the probable participation of Cu(III) in the activity of some enzymes and as an intermediate in the enzymatic DNA cleavage mediated by metalloproteins.
3][4][5][6] In these studies the generation of reactive oxygen species in a Fenton type mechanism was proposed, where Cu(II) is reduced to Cu(I) which reacts with H 2 O 2 to generate HO • .However, the intermediate that causes DNA cleavage has not been identified.
Nowadays it is widely known that S(IV) (SO 2 , HSO 3 ) autoxidation is catalyzed by transition metal Vol. 20, No. 7, 2009   ions, such as Cu(II), Ni(II), Mn(II) and Co(II), where free oxy sulfur radicals (SO 3 •− , SO 4 •− or SO 5 •− 0][21][22][23][24] Previously, we showed that the oxidation of S(IV) in air saturated solution ([O 2 ] = 0.25 mmol L −1 , pH 7), in the presence of the Cu II G 4 (G 4 = tetraglycine) complex, occurs with Cu III G 4 formation and oxygen consumption. 13,14,16,17In addition, we verified that Cu II G 4 complexes interact with DNA producing strand breaks in significant yields in the presence of S(IV) without or with a second metal ion (traces of Ni(II)). 22u II G 4 alone induced little or no DNA damage in the absence of S(IV). 19,21The oxidation of 2´deoxyguanosine to 8-oxodGuo and both single and double-strand breaks in DNA were observed under the same conditions. 19,21n the present work, the DNA damage induced by S(IV) in the presence of some Cu(II) peptide complexes in air saturated solution was investigated.6][27] The peptides: glycylglycylhistidine (GGH), glycylhistidyllysine ( G H K ) , g l y cy l g l y cy l t y r o s y l a rg i n i n e ( G G Y R ) , glycylhistidylglycine (GHG), glycylglycyl-L-alanine (GGA), triglycine (G 3 ) and tetraglycine (G 4 ) were selected for this study in order to evaluate the probable role of trivalent copper in the DNA damage mechanism.
Due to the different protonation degree of the coordinated ligand, in this study the representations Cu II L and Cu III L refer to all complexes species present in solution (at pH 7.5) formed with the metal ions. 28,29n addition, DNA damage was also investigated in the presence of Cu(II) and BCA (4,4' dicarboxy-2-2'biquinoline), a specific chelator for Cu(I).It was observed that Cu(II) is reduced by S(IV) to form the Cu(I)/BCA complex. 30

Reagents
All reagents were of analytical grade.All solutions were prepared by using deionized water purified with a Milli-Q Plus Water System (Millipore).
Supercoiled pUC-19 DNA and electrophoresis grade agarose were purchased from MBI Fermentas.
Stock solutions of S(IV) (0.010 mol L −1 , Merck) were fresh prepared by dissolving Na 2 S 2 O 5 salt in water previously purged with nitrogen.Deionised water was flushed with nitrogen for at least half an hour to remove dissolved oxygen.To prepare diluted solutions of S(IV), small volumes of the stock solutions were added to air saturated water.
Cu(II) and Ni(II) (0.2 mol L −1 ) stock solutions were prepared from the direct reaction of Cu (wire, 99.99%) and Ni (powder, 99.99%) with double distilled nitric acid followed by standardization with EDTA by a conventional procedure. 31n the experiments fresh Cu(II) complex solutions were prepared by dissolving the appropriate peptide in water (pH 7.5) followed by the addition of Cu(II) solution (solutions were prepared to have 10% excess of peptide to restrain any Cu(OH) 2 precipitation).In some experiments, aliquots of Ni(II) solution was added to Cu(II) complex solution in order to study the synergistic effect.The final pH was adjusted with 0.1 mol L −1 NaOH or 0.1 mol L −1 HClO 4 solutions.The ionic strength, I, was kept at 0.05 mol L −1 with NaClO 4 only for the spectrophotometric studies.0.2 mol L −1 BCA solution was prepared by dissolution of Na 2 BCA in water.Cu II BCA solutions was prepared few minutes before the experiments by mixing Cu(NO 3 ) 2 and BCA solutions, such as the final working solution was 3 mmol L −1 BCA and 1 mmol L −1 Cu(II).As Cu II BCA precipitates after a few minutes, the solution must be freshly prepared followed by the fast addition of S(IV).Diluted solution of Cu II BCA was employed in gel electrophoresis experiments (see Figure 2).
Air saturated solutions were employed in all the experiments for which the dissolved oxygen concentration can be considered to be 0.25 mmol L −1 .A pH meter Metrohm 713 with a glass electrode (filled with sat.NaCl) was used in the pH measurements.

Spectrophotometric measurements
Equals volumes of S(IV) and metal ion complex solutions were mixed.The concentrations of each reagent just after the mixture are indicated in the Figure 1.
The kinetic runs were followed at the wavelength of maximum absorption of each Cu(III) complex by using a HP8453A diode array spectrophotometer coupled to a Pro-K.2000Stopped-Flow Mixing Accessory (Applied Photophysics).

Gel electrophoresis experiments
The DNA strand break efficiencies induced by the copper complexes, in the presence or absence of S(IV), were determined by mixing the Cu(II) complex solution with 100 ng of pUC 19 plasmid DNA followed by the addition of S(IV) in a total volume of 50 µL.The final concentrations after mixing are indicated in Figures 2 and 3.
The separation of the different conformations of pUC 19 plasmid DNA (supercoiled, open circular and linear) was performed by gel electrophoresis, using 0.8% agarose/1.8µmol L −1 ethidium bromide in a horizontal gel electrophoresis chamber at 30 mA for 120 min in 90 mmol L −1 tris-borate /2 mmol L −1 EDTA buffer (pH 8.0).The bands were visualized under UV light and quantified with the ImageMaster VDS densitometer (Pharmacia Biotech, San Francisco, CA, USA).

EPR experiments Direct EPR of copper(II) complexes
EPR spectra were recorded in a Bruker EMX EPR (Electron Paramagnetic Resonance) spectrometer equipped with a standard cavity, operating at X-band frequency, using standard Wilmad quartz tubes, at −196 ºC or at  [pUC19] = 5 µg mL −1 , pH 7.2, air saturated solutions.The data points represent the average of three trials (limit of error between 2 and 13%).room temperature using a flat quartz cell.DPPH (α,α´diphenyl-β-picrylhydrazyl) was used as frequency calibrant (g = 2.0036) with samples in frozen aqueous solution, at −196 ºC.Usual conditions used in these measurements were 10 G modulation amplitude and 50 mW power (spectra at room temperature) or 20 mW power (at −196 ºC).
In a typical spin trapping experiment, a solution of copper complex was previously mixed with DMPO (2,2'dimethyl-pyrroline-N-oxide, from Aldrich) followed by the addition of S(IV) solution, and 200 µL of the mixture was transferred to a flat quartz cell and the EPR spectra recorded during the time.The concentrations of the final solutions are indicated in Figures 4 and 5.
The working solutions (at pH 7.5) were treated with Chelex 100 to remove metal ion contaminants.DMPO was vacuum distilled previously to use. 32The stock aqueous solution of DMPO 2 mol L -1 was maintained at 6 o C.

Results and Discussion
The DNA damage was investigated in the presence of Cu(II) complexes (with GHK, GGYR, GGH, GHG, GGA, G 3 , G 4 , and BCA), S(IV) and dissolved oxygen.The ability of each complex to cleave DNA was verified by gel electrophoresis experiments, as will be further discussed.DNA strand breaks occurred with high efficiency only in the presence of free Cu(II), Cu II GGA, Cu II G 3 , Cu II G 4 or Cu II BCA and dissolved oxygen after the addition of S(IV).Therefore, to gain a better understanding of the reactions involved, most of the EPR and UV-Vis experiments were carried out with these complexes.

Spectrophotometric studies of the oxidation of Cu(II) complexes in the presence of S(IV) and dissolved oxygen
The addition of S(IV) (a reducing agent) to an air saturated solution containing Cu II GGA, Cu II G 3 or Cu II G 4 According to Anast and Margerum, 27 the oxidation of Cu II G 4 to Cu III G 4 in aqueous medium, by dissolved oxygen, is strongly accelerated in the presence of S(IV) with simultaneous oxidation of Cu(II) and S(IV).However, our studies [12][13][14] showed that in fact, this reaction is very slow in the presence of S(IV).The Cu III G 4 formation is efficient and fast only in the presence of S(IV) and trace concentrations of nickel(II) ion, present as impurity in the copper (II) salts (CuClO 4 , Sigma) reagents.These studies were carried out at pH 9.
In our previous work [12][13][14] the oxidation of Cu II G 4 induced by S(IV) at pH 7, in the presence and absence of Ni(II), was described.The oxidation of 1 mmol L −1 Cu II G 4 , after addition of S(IV), is relatively slow with an induction period, characteristic of autocatalytic reactions.In the presence of 10 µmol L −1 Ni(II), the induction period decreases with a slight increase in the rate and effectiveness of Cu III G 4 formation, which could be followed by its characteristic absorbance peak at 365 nm. 13,14,26,27The effectiveness of the Cu III G 4 formation and the synergistic effect of Ni(II) were more pronounced at pH 9, which can be explained by the different reactivity of Cu II G 4 complexes due to the different protonation degrees of the coordinated ligand. 28,33n the present work, similar experiments were carried out focusing on the oxidation of Cu(II) complexes with GGA, G 3 , GGH, GHG, GHK and GGYR.
The spectral changing of an air saturated solution of Cu II GGA 1.0 mmol L −1 and Ni(II) 10 µmol L −1 at pH 7.5 after S(IV) addition are shown in Figure 1a inset, with solution colors changing from purple to yellow.The two new peaks at 250 (not shown) and 385 nm are attributed to Cu III GGA formation.The absorbance changes at 385 nm (Figure 1a), can be followed since the absorbance of Ni II GGA 10 µmol L −1 does not interfere. 20The oxidation of Cu II GGA is not efficient in the absence of traces of Ni(II) ion (Figure 1a,  B).However. the synergistic effect of this second metal ion can be seen by the fast Cu III GGA formation (Figure 1a, C), followed by its decomposition with probable ligand oxidation by Cu(III).The absorbance of Cu II GGA (555 nm), of the final solution, is about the same, showing that only a small percentage of the initial Cu II GGA was consumed.The same behavior was observed for the oxidation of Cu II G 3 under similar experimental conditions (Figure 1b).The existence of Cu III G 3 was previously confirmed by electrochemical experiments. 34nder the same experimental conditions the addition of S(IV) to air saturated solutions of Cu II GHG, Cu II GHK, Cu II GGYR and Cu II GGH (in the absence or presence of Ni(II) traces) led to no spectral changes showing that the respective Cu(III) complexes, if formed, were not at detectable amounts.
In order to better evaluate the oxidation of Cu II GGA and Cu II G 3 complexes the S(IV) (reducing agent) was replaced by HSO 5 − (strong oxidant) (data not shown).In the presence of Ni(II) and HSO 5 − the oxidation of Cu II GGA is more efficent than with S(IV), with the appearance of the same absorption peak at 385 nm.After the decomposition of these unstable intermediates , the peak at 555 nm (Cu II GGA) shifts to 575 nm, problably due to the formation of a new complex Cu II GGA' (where GGA' is an oxidized form of the ligand).
On the other hand, the addition of S(IV) to a solution containing Cu(II) and BCA, the involved mechanism is very different.Cu(II) is reduced, in less than 1 minute, to the stable Cu I BCA complex, which has two absorption peaks 558 nm (weak) and 330 nm (intense). 35

Agarose gel electrophoresis experiments. DNA damage
Agarose gel electrophoresis was carried out using pUC-19 plasmid DNA in the presence of air-saturated solutions of copper(II) complexes (with GGH, GHK, GGA, GHG, GGYR, G 3 , G 4 and BCA) and S(IV) to provide evidence of DNA strand breaks.The effect of incubation time and the synergistic effect of Ni(II) were also evaluated.
The ability of each copper complex to cleave DNA was verified through the conversion of the supercoiled form of pUC-19 plasmid DNA (SC, native conformation) to open circular (OC) and linear (L) forms.DNA damage was quantified based on the ratio of the total amount of OC and L forms produced (normalized with respect to the background produced by DNA alone) to the total amount of DNA present.We observed that the order of addition of the reagents is very important in the extension of the damage.The higher ratio of (OC + L) : (SC) percentage was observed when the order of addition of the reagents was: DNA + Cu II complexes (or Cu(NO 3 ) 2 ) + S(IV).
In the present study, no DNA damage was observed in the presence of S(IV) alone (1-2000 µmol L −1 ) even after incubation with DNA for 2 h.
When the experiments were carried without incubation, Cu(II) ion (1-500 µmol L −1 ), added as Cu(NO 3 ) 2 , little or no damage occurred in the absence or presence of S(IV) (lower than 10 µmol L −1 ).However, in the presence of higher concentrations of S(IV) (10-500 µmol L −1 ) the formation of the OC form was observed.With incubation for 2 h (37 o C) the DNA strand breaks were more efficient and the linear form was also observed.SC form became quantitatively converted to OC and L forms when S(IV) was used in the range 300-1000 µmol L −1 (Figure 2b).Therefore, the extent of DNA damage depends on the incubation and S(IV) and Cu(II) concentrations.In addition, an optimum condition to induce 100% DNA damage may be 300 µmol L −1 S(IV) and 50 µmol L −1 Cu(II) (Figure 2b, with incubation).
Figure 2a shows that considerable DNA damage (only with formation of OC form) also occurs when the experiments were carried out without incubation in the presence of Cu(II) complexes over the entire S(IV) concentration range.The intensity of DNA damage in the presence of the different Cu(II) complexes and for S(IV) (10-300 µmol L −1 ) followed the order: Cu II BCA ∼ Cu II G 4 ∼ Cu(II) > Cu II G 3 ∼ Cu II GGA.Specifically for Cu II BCA the damage occurred even at a lower S(IV) concentration (0.1 µmol L −1 , data not shown).
By comparing Figure 2a and b, one can conclude that the damage in the presence of Cu II BCA did not depend on the incubation time (for 2 h at 37 o C), which indicates that the species responsible for the DNA damage is produced at the initial stage of the reduction reaction of Cu II BCA by S(IV).It is important to mention that no DNA damage was observed when the order of addition of the reagents was Cu II BCA + S(IV) + DNA, such as Cu I BCA was formed before DNA addition, which shows that Cu I BCA alone does not induce DNA damage.
DNA damage increased with incubation time in the presence of S(IV) and Cu(II), Cu II G 4 , especially with Cu II G 3 and Cu II GGA.For the Cu(II) peptide complexes it can be related to the rate of the oxidation of the Cu(II) complex induced by S(IV) in the presence of dissolved oxygen, as shown by the spectrophotometric data (Figure 1) and previous work. 12,16,19,21,34ue to the synergistic effect of Ni(II) on the oxidation of Cu II G 4 in the presence of S(IV) and dissolved oxygen, 12,16,17,21 some experiments were carried out to evaluate the DNA strand breakage as shown in Figure 3.
After the treatment of DNA with 50 µmol L −1 Cu II GGA and different concentrations of S(IV), in the presence and absence of 0.5 µmol L −1 Ni(II), the results showed that DNA strand breakage increased with S(IV) concentration (up to 500 µmol L −1 ) and incubation time.In the presence of traces of Ni(II) the increment was higher especially at lower S(IV) concentrations (10-100 µmol L −1 ).

EPR experiments Direct EPR of copper complexes
In order to verify the variation of the oxidation state of copper(II) complexes in the presence of S(IV) and dissolved oxygen, direct EPR studies were performed at room temperature.
The EPR spectra of the complex Cu II BCA is represented in Figure 4A-C.The EPR signal characteristic of Cu(II) d 9 paramagnetic environment (Figure 4A, before S(IV) addition) disappeared after S(IV) addition (Figure 4B), remaining as a stable copper(I) species (diamagnetic) for 2 h.These results are in agreement with the spectrophotometric ones.
Regarding the Cu II GGA complex, there is no report on the formation of Cu(III) species in the literature.EPR spectra of Cu II GGA solution before (Figure 4D) and after (Figure 4E) S(IV) addition at room temperature showed no changes, indicating that Cu(II) with GGA remains relatively stable after S(IV) addition, in the presence or absence of Ni(II).
In order to verify slight changes in the intensity of Cu(II) signal (in CuGGA complex) during the S(IV) reaction, low temperature EPR spectra were performed with time (Figure 4F).A contineous increase in the EPR Cu(II) signal up to 3 min, with a decay after 5 min was observed.This EPR spectrum of Cu II GGA showed, even after the addition of S(IV), a characteristic profile of an axial environment around copper(II) ion 36 with EPR hyperfine parameter g // > g ⊥ .No change in the oxidation state of Cu II GGA was verified.The increase in the intensity of the EPR signal in the first minutes of reaction indicated an increase of Cu(II) moiety in solution, however, the signal shape did not change, signifying that a similar environment around the copper(II) (in the complex that originates such increase) was maintained during the reaction.The increase in the intensity of EPR signal was accompanied by a color change of the solution from purple to yellow, as described in the spectrophotometric studies (Figure 1a), indicating that the species that produces such an increase is the same as that absorbing at 385 nm.The reduction of the EPR signal (after 5 min), also accompanied by a color change from yellow to purple, indicates that the concentration of Cu(II) complex decreases probably due to either return to its initial form (Cu II GGA) or decomposition via ligand oxidation.The same experiments, carried out in the absence of Ni(II), showed similar results.
All facts appear to indicate that the copper GGA complexes are initially present mainly as Cu II GGA and may be some small amounts as Cu III GGA and/or Cu I GGA, formed by oxidation of Cu II GGA by dissolved oxygen (equation 4) or disproportionation of this complex (equation 3), respectively.After the S(IV) addition, a small amount of Cu III GGA should be initially reduced (equation 5), by this way increasing the total amount of the Cu(II) complex, witch may be also formed by the oxidation of the initial Cu I GGA if it exists.
Very similar results were obtained using the Cu II G 3 complex (data not shown).

EPR spin trapping
To verify the formation of free radicals during the oxidation of S(IV) in the presence of dissolved oxygen and Cu II BCA, Cu II GGA or Cu II G 3 , EPR spin trapping experiments were carried out using 5,5-dimethyl-1pyrroline-N-oxide (DMPO) which forms the DMPO/ SO 3 •− 37 and DMPO/HO • 38 spin adducts in the presence of sulfite and hydroxyl radicals, respectively.
Air saturated solutions containing only 0.5 mmol L −1 S(IV) or Cu II GGA or Cu II G 3 generated no EPR signal with DMPO (Figure 5A, B and C, respectively).
The EPR spectra obtained from S(IV) oxidation in the presence of Cu II GGA or Cu II G 3 are shown in Figure 5D and E, respectively.Both spectra showed the presence of the DMPO/SO 3 •− spin adduct, with a N = 14.6 G and a H = 16.0G. 19,37 For both complexes, the signal intensity reached saturation after 6 minutes of reaction followed by a slow decay, indicating that the spin adduct is not stable over longer time under these experimental conditions.In the presence of Cu II G 3 (0.5 mmol L −1 ) and Ni(II) (5 µmol L −1 ), sulfite radical generation was more efficient than with Cu II GGA (0.5 mmol L −1 ) and Ni(II) (5 µmol L −1 ) (data not shown).
Copper(II) nitrate was used instead of copper(II) peptide to compare the sulfite radical generation (Figure 5F).The DMPO/SO 3 •− spin adduct signal was more intense than with copper peptide complexes (Figure 5F), showing that free copper(II) ions enhanced the sulfite radical production in the medium.
The addition of DMPO to the Cu II BCA solution (in the absence of S(IV)) generated the radical adduct DMPO/HO • (data not shown), suggesting that DMPO was oxidized to its radical DMPO •+ which could undergo addition of water to yield DMPO/HO • (equations 1 and 2), 39 with a N = a H = 14.9.
Cu II BCA + DMPO → Cu I BCA + DMPO •+ (1) The Cu II BCA complex in the presence of S(IV) and DMPO also did not generate sulfite radicals, but the radical adducts DMPO/HO • (via equations 1 and 2) and DMPOX (5,5-dimethylpyrrolidone-(2)-oxy-(1), with a N = 7.1 G and a H = 4.2 G) were detected (data not shown).So, in this case, a strong oxidant capable of oxidizing DMPO (1.87 V vs. NHE) must be formed. 40In addition, there is a possible formation of a carbon centered radical of BCA not detected in this case.According to Rosen and Rauckman, 41 the DMPOX signal is an indirect evidence for peroxyl radical (ROO • ) formation, which is trapped by DMPO.
Taken together, these results proved that the mechanism of free radical generation and consequently DNA damage caused by sulfite oxidation depends on the type of copper(II) complex used, as will be further discussed.
Reactions 3-6 (Scheme 1) show the main steps that can be involved in the Cu II L (L = G 3 , G 4 or GGA) oxidation in the absence of S(IV).Cu III L can be produced by disproportion of Cu II L to Cu I L and Cu III L (equation 3) 12,27 or by the slow oxidation of Cu II L by oxygen (equation 4) to produce H 2 O 2 , which can subsequently oxidize Cu II L to Cu III L (equation 5).The HO • produced from the slow Cu II G 4 oxidation, in absence of S(IV), can also oxidize Cu II G 4 to Cu III G 4 (equation 6) at high enough levels to initiate the process.Reactions 3 and 4 are more probable to occur for Cu II GGA and Cu II G 3 complexes since HO • radical formation was not observed by EPR.However, reactions 3-6 can occur in the case of Cu II G 4 since in the absence of S(IV), HO • was detected. 19,21ccording to the proposed mechanism, in the presence of S(IV), some initial Cu III L is necessary at zero time to initiate the process (the formation of sulfite radical, equation 7)).It could be produced by reactions 3-6 or by low Fe(III) concentration (10 -8 -10 -7 mol L −1 ), present as impurity. 42Thus, the Cu III L formed reacts with sulfite to generate SO 3 •− radical (equation 7).In the autocatalytic process, Cu II L is oxidize by SO 5 •− (equation 11), produced by oxidation of SO 3 •− by dissolved oxygen (equation 10); HSO 5 − and SO 4 •− can also oxidize Cu II L in subsequent steps (equations 13-15).Cu III L can then be reduced by SO 3 •− and SO 5 4][45][46][47] These studies showed that Cu(III) complexes with tripeptides or tetrapeptides decompose with rapid oxidative degradation of the peptides.In addition, some of the authors 29 suggested Cu(III) as an intermediate in the oxidative degradation and cleavage of DNA in reactions involving Cu II GGHG/H 2 O 2 /ascorbic acid (GGHG = glycylglycylhistidylglycine).
Our spectrophotometric and EPR studies also suggest the involvement of Cu(III) intermediates, which decompose rapidly during the reaction between Cu II G 3 (or Cu II GGA) and S(IV), in the presence of oxygen.The decomposition should regenerate some of the initial Cu(II) complex and partially oxidize the ligand.
After the addition of S(IV) to an air saturated solution of Cu II G 3 , a characteristic absorption band of Cu III G 3 (λ max = 375 nm 34 which rapidly shifts to 385 nm, Figure 1b) was observed, followed by an absorption decrease, indicating that an instable complex is formed.In the case of Cu II GGA, only the absorption band at 385 nm is observed (Figure 1a), which also decays with time.As these Scheme 1. Mechanism of Cu II L (L = G 3 , G 4 or GGA) oxidation in the presence of S(IV) and dissolved oxygen.,34 Initiation in the absence of S(IV) Cu II  (15) 2Cu III L → Cu II L + Cu II L' (16) L' is the oxidized form of L intermediate complexes are instable, they could not be detected by EPR (item "Direct EPR of copper complexes"), which showed no formation of Cu(III).If some Cu(III) is formed during the first second, the life time is so short that it could not be detected by EPR (Figure 4D, E and F).
The increase in the EPR signal after addition of S(IV), (Figure 4F, I-IV ) seems to indicate that initially the most of the total Cu(II) is present as Cu II GGA and a small portion forms other species.After addition of S(IV), the species Cu III GGA and/or Cu I GGA could likely form Cu II GGA' (GGA' is the oxidized form of GGA), but always keeping the same coordination geometry.The oxidation of the ligand could occur in some region of the molecule such as the initial coordination to the Cu(II) by the GGA peptide was not modified.
In the case of Cu II G 4 , EPR studies 21 showed the oxidation of DMPO by some species resulting from the decomposition of Cu III G 4 , probably a peroxyl radical.According to Kurtz et al., 29 after decomposition of Cu III G 4 , generated by oxidation of Cu II G 4 by oxygen, some reactive intermediate (a carboncentered free radical, a Cu(I) complex, a peroxyl radical or a copper(III) peroxide) may be formed.

3 Figure 3 .
Figure 3. Formation of DNA strand breaks upon exposure of plasmid pUC 19 to Cu II GGA and different concentrations of S(IV) with (A) no incubation (T = 25 o C) and (A') incubation for 2 h (T = 37 o C). (B) and (B'): results obtained without and with incubation respectively in medium containing 50 µmol L -1 Cu II GGA / 0.5 µmol L −1 Ni(II) and different concentrations of S(IV).[pUC19] = 5 µg mL −1 ; pH 7.5; air saturated solutions.% DNA strand breaks represent the sum of OC and L forms.The percentages are the average of three trials.Lane 1: DNA alone.

2 − 5 −
(equation 7) to continue the chain reaction.This redox cycling is active as long as sulfite and oxygen are present in solution to generate the SO 5 •− , HSO 5 − and SO 4 •− species.The Cu III L complex can also decompose to Cu II L' (equation 16, where L' is the ligand in the oxidized form).The synergistic effect of traces of Ni(II) can be explained by the faster oxidation of Ni II L by O 2 (equation 8), producing Ni(III), which rapidly reacts with sulfite to form the SO 3 •− radical (equation 9).Ni II L can also be oxidized to Ni III L by SO 5 and SO 4 •− , and participate in the redox cycling similarly to the Cu(II) reactions (see Scheme 1).The chain propagation, product formation and termination reactions involving sulfite, HSO 5 − , SO 3