DNA Lesions Derived from the Site Selective Oxidation of Guanine by

James S. Stover, Madalina Ciobanu, David E. Cliffel, and Carmelo J. Rizzo .... Guimarães , Mauro Geller , Flavia de Paoli , Adenilson de Souza da Fon...
1 downloads 0 Views 164KB Size
1528

Chem. Res. Toxicol. 2003, 16, 1528-1538

DNA Lesions Derived from the Site Selective Oxidation of Guanine by Carbonate Radical Anions Avrum Joffe, Nicholas E. Geacintov, and Vladimir Shafirovich* Chemistry Department and Radiation and Solid State Laboratory, 31 Washington Place, New York University, New York, New York 10003-5180 Received July 8, 2003

Carbonate radical anions are potentially important oxidants of nucleic acids in physiological environments. However, the mechanisms of action are poorly understood, and the end products of oxidation of DNA by carbonate radicals have not been characterized. These oxidation pathways were explored in this work, starting from the laser pulse-induced generation of the primary radical species to the identification of the stable oxidative modifications (lesions). The cascade of events was initiated by utilizing 308 nm XeCl excimer laser pulses to generate carbonate radical anions on submicrosecond time scales. This laser flash photolysis method involved the photodissociation of persulfate to sulfate radical anions and the one electron oxidation of bicarbonate anions by the sulfate radicals to yield the carbonate radical anions. The latter were monitored by their characteristic transient absorption band at 600 nm. The rate constants of reactions of carbonate radicals with oligonucleotides increase in the ascending order: 5′-d(CCATCCTACC) [(5.7 ( 0.6) × 106 M-1 s-1] < 5′-d(TATAACGTTATA), self-complementary duplex [(1.4 ( 0.2) × 107 M-1 s-1] < 5′-d(CCATCGCTACC [(2.4 ( 0.3) × 107 M-1 s-1] < 5′-d(CCATC[8-oxo-G]CTACC) [(3.2 ( 0.4) × 108 M-1 s-1], where 8-oxo-G is 8-oxo-7,8-dihydroguanine, the product of a two electron oxidation of guanine. This remarkable enhancement of the rate constants is correlated with the presence of either G or 8-oxo-G bases in the oligonucleotides. The rate constant for the oxidation of G in a singlestranded oligonuclotide is faster by a factor of ∼2 than in the double-stranded form. The site selective oxidation of G and 8-oxo-G residues by carbonate radicals results in the formation of unique end products, the diastereomeric spiroiminodihydantoin (Sp) lesions, the products of a four electron oxidation of guanine. These lesions, formed in high yields (40-60%), were isolated by reversed phase HPLC and identified by matrix-assisted laser desorption/ ionization time-of-flight mass spectrometry. These assignments were supported by the characteristic circular dichroism spectra of opposite signs of the two lesions. The oxidation of guanine to Sp diastereomers occurs, at least in part, via the formation of 8-oxo-G lesions as intermediates. The Sp lesions can be considered as the terminal products of the oxidation of G and 8-oxo-G in DNA by carbonate radical anions. The mechanistic aspects and biological implications of these site selective reactions in DNA initiated by carbonate radicals are discussed.

Introduction A variety of endogenous and exogenous reactive species such as free radicals produced by metabolic and other chemical reactions, or ionizing radiation, can cause diverse, potentially mutagenic oxidative modifications (lesions) of the nucleic acid bases in DNA (1). The integrity of the genetic apparatus is maintained by base excision and other repair enzymes that remove and repair the damaged bases or nucleotide sequences containing the lesions (2). However, the overproduction of DNA lesions in tissues under conditions of oxidative stress accompanying chronic inflammation and various degenerative diseases increases the risk of malignant cell transformation leading to the development of cancer (3, 4). Guanine is the most easily oxidizable nucleic acid base (5) and is therefore a primary target of attack in DNA (6). * To whom correspondence should be addressed. Tel: (212)998 8456. Fax: (212)998 8421. E-mail: [email protected].

The focus of this work is on the oxidation of guanine in DNA by CO3•-,1 one of the important free radicals in biological systems (7). There are a number of different pathways in vivo that can result in the generation of CO3•- radical anions. These include one electron oxidation of bicarbonate anions at active sites of copper-zinc superoxide dismutase (8-10) and the reaction of peroxynitrite with carbon dioxide (11-14). The CO3•- radical with a reduction potential (15), E°(CO3•-/CO32-) ) 1.59 V vs NHE, is a strong one electron oxidant that can rapidly abstract electrons from appropriate donors (16). In contrast, hydrogen atom abstraction by the CO3•- radicals is generally slow (16). In living systems, the CO3•- radical is known to contribute to the development of DNA lesions 1 Abbreviations: CO •-, carbonate radical anion; G(-H)•, neutral 3 guanine radical; Sp, spiroiminodihydantoin; 8-oxo-G, 8-oxo-7,8-dihydroguanine; 8-nitro-G, 8-nitroguanine; 8-oxo-G(-H)•, neutral 8-oxo-G radical; oligo, oligonucleotide; MALDI-TOF, matrix-assisted laser desorption/ionization with time-of-flight detection; MS, mass spectrometry.

10.1021/tx034142t CCC: $25.00 © 2003 American Chemical Society Published on Web 10/28/2003

DNA Lesions Derived from Oxidation of Guanine

Chem. Res. Toxicol., Vol. 16, No. 12, 2003 1529

Scheme 1. Site Selective Oxidation of Guanine by CO3•- in DNA

(17) and the post-translational modification of proteins (18-20). Recently, we have investigated reactions of CO3•radicals with double-stranded DNA by laser flash photolysis techniques (21). In these time-resolved experiments, the CO3•- radicals were derived from the one electron oxidation of HCO3- anions by photochemically generated sulfate radical anions, SO4•-. The CO3•- radicals induce the site selective oxidation of guanine residues in DNA, a reaction that can be monitored by the rise of the characteristic transient absorption band of the guanine neutral radicals, G(-H)• thus generated. The oxidation of guanine in DNA by CO3•- radicals results in the formation of alkali-labile lesions. However, the structures of the alkali-labile guanine oxidation products by CO3•- radicals, as well as the mechanism of the associated reactions, remain unknown. In this work, we explore the nature of the oxidative processes in DNA that are initiated by CO3•- radicals generated by 308 nm nanosecond excimer laser pulses. The site selective oxidation of guanine by CO3•- radicals in single- and double-stranded oligos results predominantly in the formation of diastereomeric Sp lesions, the products of a four electron oxidation of guanine (Scheme 1). This complex oxidation process occurs, at least in part, via the formation of an intermediate that is identified as the 8-oxo-G residue, the product of two electron oxidation of guanine. Consistent with this conclusion, small concentrations of 8-oxo-G lesions were detected during the course of oxidation of guanine bases in DNA by CO3•- radicals. In parallel, we also studied the oxidation of DNA sequences containing site specifically incorporated 8-oxo-G lesions and found that the oxidation of 8-oxo-G lesions by CO3•- radicals is significantly faster than the oxidation of guanine in the identical sequence context. The end products of reaction, Sp diastereomers, are the same in both cases (Scheme 1). The mechanistic aspects of these site selective reactions initiated by CO3•radicals are discussed.

Experimental Procedures Materials. All chemicals (analytical grade) were used as received. Snake venom phosphodiesterase was purchased from Pharmacia (Piscataway, NJ); bovine spleen phosphodiesterase, nuclease P1, and alkaline phosphatase were from Sigma Chemical (St. Louis, MO). The oligos were synthesized by standard automated phosphoramidite chemistry techniques. Phophoramidites and other chemicals required for oligo synthesis were obtained from Glen Research (Sterling, VA). The tritylated oligos

were deprotected overnight at 55 °C employing either concentrated aqueous ammonia solutions (oligos containing only normal DNA bases) or concentrated aqueous ammonia solutions with 0.25 M mercaptoethanol (8-oxo-G-modified oligos) to prevent the oxidation of 8-oxo-G residues. The crude oligonuclotides were purified by reversed phase HPLC, detritylated in 80% acetic acid according to standard protocols, and desalted using reversed phase HPLC. The integrity of the oligos was confirmed by MALDI-TOF MS. Laser Flash Photolysis. The kinetics of oxidative reactions initiated by free radicals were monitored directly using a fully computerized kinetic spectrometer system (∼7 ns response time) described elsewhere (22). Briefly, 308 nm nanosecond XeCl excimer laser pulses were used to photolyze the sample solutions (0.25 mL) contained in a 0.4 cm × 1 cm quartz cell. The transient absorbance was probed along a 1 cm optical path by a light beam from a 75 W xenon arc lamp with its light beam oriented perpendicular to the laser beam. To maximize the signal/noise ratio, the photomultiplier output was terminated by a 3 kΩ resistor and recorded by a Tektronix TDS 5052 oscilloscope operating in its high-resolution mode. The second-order rate constants of the oxidative reactions initiated by free radicals were determined by least-square fits of the appropriate kinetic equations to the transient absorption profiles that were based on five different experiments with five different samples. Photochemical Synthesis of the Oxidatively Modified Oligos and Nucleosides. Samples containing 10 nmol of the oligos or 30 nmol of 2′,3′,5′-tri-O-acetylguanosine, 300 µmol of NaHCO3, and 10 µmol of Na2S2O8 in 1 mL of air-equilibrated solutions adjusted to pH 7.5 by the addition of appropriate amounts of a 1 M solution of NaH2PO4 were prepared. These samples were excited by an excimer laser (15 mJ/pulse/cm2) operating at 10 Hz for fixed time intervals (1-30 s), controlled by a computer-interfaced electromechanical light shutter. The light beam incident on the sample was defocused in order to attenuate the fluence rate. The authentic Sp nucleosides were prepared by the photooxidation of 2′,3′,5′-tri-O-acetylguanosine in the presence of the photosensitizer methylene blue (23). Samples containing 2 µmol of 2′,3′,5′-tri-O-acetylguanosine and 1 µmol of methylene blue in 2 mL of water were irradiated by a 100 W xenon arc beam filtered by a 500 nm cutoff glass filter for 1 h. The oxidatively modified oligos and nucleosides were isolated, purified, and desalted by reversed phase HPLC methods. HPLC Isolation, Purification, and Desalting. A crude separation of the oligo reaction products was performed on an analytical (250 mm × 4.6 mm i.d.) Microsorb-MV C18 column (Varian, Walnut Creek, CA). An 11-20% linear gradient of acetonitrile in 50 mM triethylammonium acetate in water (pH 7) was utilized for 60 min at a flow rate of 1 mL/min (detection of products at 260 nm). The HPLC fractions were evaporated under vacuum to remove acetonitrile and purified by a second

1530 Chem. Res. Toxicol., Vol. 16, No. 12, 2003

Joffe et al.

run of HPLC. The diastereoisomeric Sp adducts were separated using an analytical (250 mm × 4.6 mm i.d.) Hypersil ODS column (Varian) employing an isocratic mixture of acetonitrile and 50 mM triethylammonium acetate (8:92) as the mobile phase at a flow rate of 1 mL/min. The purified adducts were desalted by reversed phase HPLC using the following mobile phases: 10 mM ammonium acetate (10 min), deionized water (10 min), and an isocratic 50:50 acetonitrile and H2O mixture (15 min). The diastereoisomeric Sp nucleosides were separated using an analytical Microsorb-MV C18 column employing the following protocol: an isocratic mixture of methanol and 50 mM ammonium acetate (5:95) for 10 min, a 5-40% linear gradient of methanol in 50 mM ammonium acetate for 20 min, and an isocratic phase of methanol and 50 mM ammonium acetate (40: 60) for 30 min, all at a flow rate of 1 mL/min (detection of products at 230 nm). Detection of 8-oxo-G Lesions. Samples containing 10 nmol of the oligo in 150 µL of 100 mM Tris-HCl buffer (pH 8.8) and 20 mM MgCl2 were incubated with 2 units of snake venom phosphodiesterase and 5 units of alkaline phosphatase at 37 °C overnight. In some experiments, the oligos (10 nmol) were digested with nuclease P1 (2 units) in 35 µL of 30 mM acetate buffer (pH 5.2) for 1.5 h at 37 °C; following digestion, the samples were treated with 2 units of alkaline phosphatase in a Tris-HCl (∼100 mM) buffer solution (pH 8.2) for 1 h at 37 °C. The incubation mixtures were passed through a Millipore centrifugal filter to remove the enzymes. The nucleosides were separated by reversed phase HPLC using an isocratic phase (8:92) of methanol and 50 mM ammonium acetate (adjusted to pH 5.2 by acetic acid) at a flow rate of 1 mL/min. The amperometric detection (24) of 8-oxo-7,8-dihydro-2′-deoxyguanosine (8-oxo-dG) was performed employing a model 800A electrochemical analyzer (CHI Instruments, Austin, TX). For detection of 8-oxo-dG, the potential of the glassy carbon working electrode was set at 0.6 V vs Ag/AgCl electrode. Quantification of 8-oxo-dG was performed by injecting known quantities of 8-oxo-dG. The recovery of 8-oxo-dG, determined by enzymatic digestion of the 8-oxo-G-modified oligos, was about 75-80%. Both enzymatic digestion procedures resulted in the same yields of the 8-oxo-G lesions. MALDI-TOF MS Assay. The mass spectra were acquired using a Bruker OmniFLEX instrument. The matrix was a 2:1 mixture of 2′,4′,6′-trihydroxyacetophenone methanol solution (30 mg/mL) and ammonium citrate aqueous solution (100 mg/mL). The aliquots (1-2 µL) of the 10 pmol/µL desalted samples and the matrix solution were spotted on a MALDI target and air-dried before analysis. The mass spectrometer equipped with a 337 nm nitrogen laser was operated in the negative linear mode (accelerating voltage, 19 kV; extraction voltage, 92.7% of the accelerating voltage; ion focus, 9 kV; and delay time, 250 ns). Each spectrum was obtained with an average of 50-100 laser shots. The mass spectra were internally calibrated by using synthetic oligos of known molecular weights. Exonuclease Digestion for Oligo Sequencing. The method described previously (25) was adapted for the DNA adducts used here. Dry adducts (150 pmol) were dissolved in 6 µL of water. For digestion from the 3′-end, 6 µL of 100 mM ammonium citrate (pH was adjusted to 9.4 by NH4OH) and 0.5 µL of snake venom phosphodiesterase (0.004 units/µL) were added to the sample solution. The digest solution was incubated at 37 °C. For digestion from the 5′-end, 7 µL of water and 1 µL of bovine spleen phosphodiesterase (0.005 units/µL) were added to the sample solution. Aliquots (1 µL) of the digested solutions were removed at fixed periods of time and placed on dry ice to interrupt the digestion reaction. Aliquots (1-2 µL) of the samples and the trihydroxyacetophenone matrix solution were immediately spotted on a MALDI target and air-dried before analysis.

Table 1. Oligos Studied designation

oligodeoxyribonucleotide

1 2 3 4

5′-d(CCATCGCTACC) 5′-d(CCATC[8-oxo-G]CTACC) 5′-d(CCATCCTACC) 5′-d(TATAACGTTATA) 3′-d(ATATTGCAATAT)

Table 2. Oxidation of the Oligos by the CO3•- Radicals; Reaction Scheme and Rate Constants N

kn (M-1 s-1)a

reaction

CO3 + 5′-dCCATCGCTACC f CO3•- + 5′-dCCATC[8-oxo-G]CTACC f CO3•- + 5′-dCCATCCTACC f CO3•- + 5′-d(TATAACGTTATA) f 3′-d(ATATTGCAATAT) 5 CO3•- + 5′-dCCATC[Sp]CTACC f 6 S2O82- + hν f 2SO4•7 SO4•- + SO4•- f S2O828 SO4•- + HCO3- f SO42- + CO3•9 CO3•- + CO3•- f CO42- + CO2 10 SO4•- + 5′-dCCATCGCTACC f 1 2 3 4

•-

(2.4 ( 0.3) × 107 (3.2 ( 0.4) × 108 (5.7 ( 0.6) × 106 (1.4 ( 0.2) × 107b (5.8 ( 0.6) × 107 φ308 ) 0.55 (ref 26) (1.1 ( 0.1) × 109 (ref 21) (4.6 ( 0.5) × 106 (ref 21) (1.3 ( 0.1) × 107 (ref 21) (3.2 ( 0.3) × 109

a The rate constants were measured in air-equilibrated buffer solutions (pH 7.5), and uncertainties are given as standard errors for the best least-squares fits of the appropriate kinetic equations to the transient absorption profiles for decay of CO3•- (600 nm) and SO4•- (445 nm). b Measured at 15 °C.

Results The objectives of this work were to explore the oxidative chemistry of guanine in DNA starting from the laser pulse-induced generation of carbonate radical anions to the identification of the final end products of the oxidation. Oligo Design. Experiments were performed with several different oligos (Table 1). Sequence 1 contains a single guanine and three other normal DNA bases (A, C, and T), while sequence 2 is identical to 1 except that G is replaced by a single 8-oxo-G residue. Sequence 2 was selected for study because 8-oxo-G was suspected to be a potential intermediate in the oxidation of guanine initiated by CO3•- radicals. Oligo 3 is also identical to sequence 1, except that the single guanine is missing; sequence 3 was selected to determine the reactivities of CO3•- radicals toward DNA bases other than guanines. The self-complementary sequence 4 was selected to explore guanine oxidation in double-stranded DNA. At pH 7.5 and an ionic strength of 0.3 M (close to the conditions utilized for our studies of the DNA oxidation by CO3•- radicals), this duplex exhibits a single, welldefined cooperative melting curve with a melting temperature of 40.8 ( 0.7 °C and a hyperchromicity of 26%. The oxidation of oligo 4 in its double-stranded form was studied at 15 °C to further minimize the partial dissociation of the duplexes. Studies with the other singlestranded oligos were performed at room temperature (23 ( 2 °C). Kinetics of DNA Oxidation by CO3•- Radicals. Photochemical techniques for generating CO3•- radical ions from bicarbonate anions are well-documented (16). We have previously shown that these methods can be utilized to generate oxidative damage of guanine residues in oligo sequences and that this damage can be revealed in the form of strand cleavage after treatment of the irradiated sequences with hot piperidine (21). Briefly, the cascade of events (Table 2) is triggered by 308 nm excimer laser pulses that induce photodissociation (26) of persulfate anions, S2O82-, into SO4•- radical anions (reaction

DNA Lesions Derived from Oxidation of Guanine

Chem. Res. Toxicol., Vol. 16, No. 12, 2003 1531

these profiles reveals that the lifetimes of the CO3•radicals are reduced in the order: pure buffer solution > 5′-d(CCATCCTACC) > 5′-d(TATAACGTTATA) duplex > 5′-d(CCATCGCTACC > 5′-d(CCATC[8-oxo-G]CTACC). The effects of oligos 1-4 on the lifetimes of the CO3•radicals have been analyzed in detail. The decay of CO3•radicals occurs via two competitive reactions: (i) oxidation of DNA (reactions 1-4) and (ii) bimolecular recombination of CO3•- radicals (reaction 9) (21). The rate of the CO3•- decay can be expressed as:

- d[CO3•-]/dt ) kn[CO3•-][oligo] + 2k9[CO3•-]2

(1)

If [oligo] > [CO3•-], the integration of eq 1 yields the following expression for the time dependence of the CO3•radical concentration:

[CO3•-]t ) kn[oligo]0[CO3•-]0/{(kn[oligo]0 + 2k9[CO3•-]0) exp(kn[oligo]0t) - 2k9[CO3•-]0} (2)

Figure 1. Kinetics of oligo oxidation by CO3•- radicals in airequilibrated buffer solutions (10 mM Na2S2O8 and 300 mM NaHCO3, pH 7.5). (A) Transient absorption spectra recorded at fixed time intervals after the 308 nm laser pulse excitaion (60 mJ/pulse/cm2) of 0.08 mM solutions of 5′-d(CCATCGCTACC). (B) Transient absorption profiles recorded at 600 nm and attributed to the decay of the CO3•- radicals in pure buffer solution (0) and in the presence of single-stranded oligos 1, 3 (0.08 mM), 2 (0.04 mM), and the self-complementary duplex 4 (0.08 mM).

6). The abrupt rise in the SO4•- radical anion concentrations is evident from the prompt rise of a characteristic transient absorption band at 445 nm with an extinction coefficient (27) of 1600 M-1 cm-1. The SO4•- radical anions cause one electron oxidation of bicarbonate anions generating CO3•- radicals (reaction 8). The CO3•- radicals thus formed are detectable because of their characteristic absorption band with an extinction coefficient of 1970 M-1 cm-1 at 600 nm (28). The quantum yield of SO4•- radicals, the known rate constants of reaction of SO4•- radicals with bicarbonate anions, and the bimolecular recombination rates of SO4•- and CO3•- radicals are summarized in Table 2. Typical transient absorption spectra of the 5′-d(CCATCGCTACC) oligo solution containing S2O82- and HCO3ions, recorded at different time intervals after the incidence of a nanosecond excimer laser pulse, are shown in Figure 1A. The decay of the CO3•- transient absorption band at 600 nm correlates with the growth of the narrow absorption band near 315 nm due to guanine neutral radicals (21). Similar results were obtained in the presence of 5′-d(CCATC[8-oxo-G]CTACC), but the decay of the CO3•- transient absorption is accompanied by the rise of another, narrow absorption band with a maximum situated at 325 nm (data not shown), attributed to the 8-oxo-G radicals (29, 30). Thus, the transient absorption spectra and kinetics of the oxidation of the single G or 8-oxo-G bases by the CO3•- radicals can be monitored because of the clearly defined and well-separated transient absorption spectra of these intermediates. Typical decay kinetics of the CO3•- radicals recorded at 600 nm in the absence of oligos or in the presence of oligos 1-4 are depicted in Figure 1B. An examination of

where [oligo]0 and [CO3•-]0 are the initial concentrations of oligo and carbonate radical concentrations, respectively. We have verified that the decay kinetics of CO3•radicals indeed conform to eq 2. In the first set of these experiments, the concentration [oligo]0 was varied from 0.03 to 0.1 mM, and the concentration [CO3•-]0 was about 0.01 mM; in the second set, [CO3•-]0 was varied from 0.005 to 0.02 mM, and [oligo]0 was kept at 0.1 mM (data not shown). The values of kn (n ) 1-4) obtained by fitting eq 2 to the experimental data points are summarized in Table 2. The value of k1 for the oxidation of the 5′-dCCATCGCTACC sequence is ∼4 times greater than the value of k3 for the reaction of CO3•- radicals with the 5′-d(CCATCCTACC) sequence that does not contain any guanines. In the double-stranded form of sequence 4, guanine oxidation is not as efficient and the value of k4 is ∼2 times smaller than the value of k1. In contrast, the value of k2, defining the reaction rate with 5′-dCCATC[8-oxo-G]CTACC, is ∼60 times greater than the value of k3 and ∼13 times greater than the value of k1. The net rate constants of oxidation of G or 8-oxo-G in the oligos 1 or 2 can be estimated as k1 - k3 or k2 - k3, respectively. The rate constant k3 is a measure of the reaction rates of the CO3•- radicals with bases other than G and 8-oxo-G. Thus, the net rate constant associated with the presence of either G or 8-oxo-G bases in oligos 1 and 2, as compared to the analogous rate constant in sequence 3, is a direct kinetic indication that the CO3•radicals are selective oxidants of G and 8-oxo-G in DNA. Optimization of Reaction Conditions to Reduce the Oxidation of DNA by SO4•- Radicals. The site selective oxidation of guanine by CO3•- radicals competes with the direct oxidation of all four bases by SO4•radicals (31-33). The interaction of SO4•- radicals with our oligo sequence 1 (reaction 10) was explored in the absence of bicarbonate anions. The value of k10 calculated by fitting eq 2 to the SO4•- decay profiles recorded at 445 nm in the presence of different concentrations (0.03-0.1 mM) of oligo 1 is close to the diffusion-controlled limit (Table 2). To minimize the direct oxidation of oligos 1-4 by SO4•- radicals, the concentration of HCO3- must be much greater than the DNA concentration since k8, the rate constant of reaction of SO4•- radicals with HCO3-, is much smaller than k10 (Table 2). With [HCO3-] ) 300 mM and a 0.01 mM concentration of the 5′-dCCATCGC-

1532 Chem. Res. Toxicol., Vol. 16, No. 12, 2003

Joffe et al.

Figure 2. Reversed phase HPLC elution profiles of the oxidation products generated by 7 s laser pulse excitation (15 mJ/pulse/cm2, 10 pulse/s) of 5′-d(CCATCGCTACC) in airequilibrated buffer solution (10 mM Na2S2O8 and 300 mM NaHCO3, pH 7.5). The unmodified sequence 5′-d(CCATCGCTACC) is eluted at 20.4 min, the diastereomeric Sp adducts are eluted at 15.9-16.4 min, the guanidinohydantoin adduct is eluted at 15.3 min, and the abasic sequence is eluted at 19.3 min. The inset shows the reversed phase HPLC elution profiles of purified diastereomeric Sp adducts.

TACC sequence, the SO4•- radicals are scavenged by HCO3- with a characteristic time of 0.7 µs. Under these conditions, the contribution of the direct oxidation of the oligo 1 by SO4•- radicals (reaction 10), estimated as the ratio of the pseudo first-order rate constants, k10[G-oligo]/ (k3[HCO3-] + k10[G-oligo]), does not exceed 0.02. The concentration of S2O8- anions used in our system was 10 mM in order to yield a 1 µM initial concentration of CO3•- after a single laser flash with an energy of ∼15 mJ/pulse/cm2. At this concentration of CO3•-, the contribution of the CO3•- radical recombination (reaction 9) is negligible, and the oxidation of the oligos (reactions 1-4) by CO3•- radicals is dominant. Chemical End Products of the Oxidation of Guanine by CO3•- Radicals. Following laser irradiation and oxidation of the oligos by CO3•- radicals, the reaction solutions were analyzed by reversed phase HPLC methods. Typical HPLC elution profiles of an irradiated solution (7 s excimer laser excitation) containing the 5′-d(CCATCGCTACC) sequence is depicted in Figure 2. The unmodified oligo 1 elutes at 20.4 min, and two major reaction products eluting close to one another in time are observed at 15.9-16.4 min. The MALDI-TOF mass spectrum of the 15.9-16.4 min fraction recorded in the negative mode exhibits a major signal at m/z 3268.3, whereas the unmodified sequence 1 shows a signal at m/z 3236.5 (M). Therefore, the oxidized oligo has a mass of M + 32. Other reaction products eluted at 15.3 and 19.3 min (Figure 2) and exhibited characteristic masses of m/z 3241.1 (M + 6) and m/z 3101.9 (M - 133), respectively. The masses of these products are consistent with that of an adduct containing guanidinohydantoin residue (M + 6) (34-37) and a depurinated sequence 1, i.e., the abasic sequence 5′-d(CCATC_CTACC) missing the guanine base moiety (M - 133). The fraction eluting at 15.9-16.4 min (Figure 2) was further separated into two products using an isocratic elution regime. The chromatograms of the two pure adducts obtained after further HPLC separation are shown in the inset of Figure 2. The existence of two closely eluting compounds, their similar yields, and their identical masses (M + 32) suggest that the oxidized base in 1 is the diastereomeric form of Sp (34-46), initially designated as the 4-hydroxy-7,8-dihydro-7,8-oxo-guanine (23, 47).

Figure 3. Reversed phase HPLC elution profile of the products derived from the oxidation of 2′,3′,5′-tri-O-acetylguanosine by CO3•- radicals (A) or by the photosensitization of methylene blue (B). The Sp diastereomers are indicated as SpA and SpB.

To obtain more insight into the nature of the guanine oxidation by CO3•-, we explored the oxidation of the model compound, 2′,3′,5′-tri-O-acetylguanosine, in which three sugar hydroxyls are protected from reactions with electrophiles by acetate groups. Because of their enhanced hydrophobic character, the acylated nucleoside products are more easily separated by reversed phase HPLC. This approach has been extensively exploited to investigate the products of oxidative DNA damage (23, 34, 36, 37, 40). We compared the reaction products of 2′,3′,5′-tri-O-acetylguanosine oxidized either by CO3•radicals or by the photosensitization of methylene blue that is known to produce the pair of diastereomeric Sp molecules (23, 40, 46, 48). In both cases, the major reaction products eluted as two closely spaced fractions (Figure 3), in agreement with results obtained in the case of the Sp diastereomers (23, 34, 36, 37, 40). The HPLC elution orders and retention times of these two samples derived from the oxidation of 2′,3′,5′-tri-O-acetylguanosine by CO3•- radicals (Figure 3A) and by the methylene blue photosensitization (Figure 3B) are identical, as are the CD spectra (Figure 4A,B). The circular dichroism (CD) spectra of the two products of the oxidation of oligo 1 by CO3•- radicals (Figure 2, inset) are shown in Figure 4C. These CD spectra exhibit the prominent positive band associated with DNA at about 280 nm but otherwise resemble in shape and signs of the CD spectra of the two pairs of mononucleoside Sp adducts derived from the oxidation of 2′,3′,5′-tri-Oacetylguanosine by CO3•- radicals and by the methylene blue photosensitization (Figure 4A,B). Additional proof that the modified oligos 1 eluting at 28.7 and 31.7 min are the diastereomeric Sp adducts was obtained by synthesizing standards of the latter by the oxidation of oligo 2 containing 8-oxo-G instead of G utilizing Na2IrCl6 and the protocol described by Burrows and co-workers (35, 38). We found that the adducts obtained by oxidation of oligo 2 by Na2IrCl6 yielded the same MALDI-TOF mass spectra (m/z 3268.2), had identi-

DNA Lesions Derived from Oxidation of Guanine

Chem. Res. Toxicol., Vol. 16, No. 12, 2003 1533

Figure 5. MALDI-TOF mass spectra of oligo 1, containing a single Sp lesion (A), digested by snake venom phosphodiasterase after a 5 or 30 min digestion time (B and C, respectively), or by bovine spleen phosphodiesterase after a 5 or 30 min digestion time (D and E, respectively).

Figure 4. CD spectra of the diastereomeric Sp adducts derived from the oxidation of 2′,3′,5′-tri-O-acetylguanosine by CO3•radicals (A), by the photosensitization of methylene blue (B), and by the oxidation of the single guanine residue in 5′-d(CCATCGCTACC) by CO3•- radicals (C).

cal HPLC retention times, and coeluted with the adducts obtained by the oligos 1 oxidation by CO3•- radicals (see Supporting Information for additional information). Therefore, we identify the modified oligos 1 eluting at 28.7 and 31.7 min, respectively (Figure 2, inset), as containing the single diastereomeric Sp lesions. To test the possible contribution of SO4•- radicals to the oxidation of DNA, solutions of oligo 1, entirely lacking HCO3- anions, were photoexcited as before, and the nature of the reaction products was analyzed by reversed phase HPLC. On the basis of MALDI-TOF analysis, the major end products isolated by HPLC were tentatively identified as the imidazalone or parabanic acid lesions, both known to result from the complete oxidation of guanine (48-50). However, nucleobases other than guanine are also oxidized by SO4•- radicals since a very fast and total degradation of the oligos is observed. Another control reaction performed was the oxidation of sequence 3 by CO3•- radicals. In agreement with the kinetic parameters (Table 2), oxidation of this sequence containing no guanines occurs much more slowly than the oxidation of oligo 1, also resulting in the formation of lesions that have not been identified. In summary, the site selective oxidation of guanine in intact DNA by CO3•radicals is highly favored, but other bases are also oxidized with lower efficiencies. Digestion of the Modified Oligo 1 by Phosphodiesterases I and II. The sites of oxidation by CO3•radicals in oligo 1 were further verified by digestion with phosphodiesterases. Typical exonuclease digestion ladders of this modified oligo (the 31.7 min-eluting fraction

SpB; Figure 2, inset) obtained by MALDI-TOF/MS techniques (negative mode) are depicted in Figure 5. The undigested, oxidized oligo 1 exhibits a signal at m/z 3268 (Figure 5A). Employing snake venom phosphodiesterase I, oligos with 3′-OH ends are digested nucleotide by nucleotide from the 3′-end. After a 5 min digestion time, only the first five nucleotides counted from the 3′-side are partially released and signals of the 6-11-mer sequences containing the Sp residue are observed (Figure 5B). The masses of these sequences correspond to Mi + 32, where Mi is the mass of the 6-11-mer sequences containing unmodified guanine. When the digestion time is increased to 30 min, only a signal of the 6-mer 5′-CCATC[Sp] is observed at m/z 1783.3 (Figure 5C), which indicates that the C-[Sp] phosphodiester bond is resistant to digestion by phosphodiesterase I. Digestion of the same oxidized oligo 1 by bovine spleen phosphodiesterase II from the 5′-end yield indicates that only the first four phosphodiester bonds are cleaved (Figure 5D) and that the same C-[Sp] phosphodiester bond is strongly resistant to digestion by this enzyme as well, even after a 30 min reaction time (Figure 5E). Similar exonuclease digestion ladders (data not shown) were observed in the case of the stereoisomeric-modified oligo eluting at 28.7 min (Figure 1, inset). Note that the C-[Sp] phosphodiester bond was also resistant to digestion with nuclease P1, and we were thus unable to generate the free Sp nucleosides utilizing enzymatic digestion techniques. An analogous resistance to exonuclease digestion has been recently reported for the processing of site specifically modified oligos with single 5-hydroxy-5-methylhydantoin lesions (51). Chemical Yield of Sp Lesions in Single-Stranded DNA. The major products of the site selective oxidation of guanine by CO3•- radicals in the single-stranded sequence 1 are the diastereomeric Sp lesions. The yield of these lesions is defined as:

1534 Chem. Res. Toxicol., Vol. 16, No. 12, 2003

Joffe et al.

Y(Sp) ) [Sp-oligo]t/[G-oligo]0 where [Sp-oligo]t is the concentration of the Sp adduct after overall irradiation time, t, and [G-oligo]0 is the initial concentration of the unmodified oligo. To determine Y(Sp), the samples containing the sequences 1 were irradiated for fixed periods of time and subjected to analysis by reversed phase HPLC. The value of Y(Sp) increases with increasing irradiation time and attains a maximum of Y(Sp) ∼ 0.6 after a 10 s exposure to the laser pulse train (Figure 6). However, further irradiation results in a gradual reduction of the Y(Sp) value, suggesting that a slow degradation of the Sp lesions is occurring. The slow degradation of the Sp lesions is consistent with the rate constant of their oxidation by CO3•- radicals (reaction 5). The value of k5 was calculated by fitting eq 2 to the decay profiles of the CO3•- radicals recorded at different concentrations (0.03-0.1 mM) of the oxidized oligo 1 containing the diastereomeric Sp lesions. The value of k5 thus obtained is very close to k3, the rate constant for the nonspecific oxidation of oligo 3, which does not contain either G or Sp (Table 2). The 8-oxo-G Lesion as an Intermediate in the Generation of Sp from the Oxidation of Guanine in Oligo 1. The possibility that 8-oxo-G base is an intermediate in the site selective oxidation of guanine to Sp by CO3•- radicals was investigated. Exposure of sequence 2 with a 8-oxo-G residue by CO3•- radicals employing the same irradiation protocols as described above for sequence 1 also results in the formation of Sp lesions (Figure 7). The value of Y(Sp) first increases rapidly with increasing irradiation time and attains a maximum of Y(Sp) ∼ 0.4 at t ∼ 2 s; further irradiation results in a gradual decrease in the Y(Sp) value. Thus, the oxidation of G or 8-oxo-G by CO3•- radicals generates the same unique end products, the Sp lesions that are resistant to further attack by CO3•- radicals. These findings stimulated efforts to explore whether 8-oxo-G is an intermediate in the oxidative reaction pathway in oligos associated with the conversion of G to Sp. To detect the presence of 8-oxo-G during the course of oxidation of G to Sp, the irradiated solutions of oligo 1 were analyzed for the presence of oligo 2. After irradiation of sequences 1 for fixed periods of time, thus allowing for the oxidation of the guanine residue by CO3•- radicals, the irradiated solutions were quenched with an excess of sodium dithionite dissolved in deoxygenated buffer solution (pH 7) to remove unreacted persulfate (without quenching. The yields of 8-oxo-G lesions in sequence 1 were lower by a factor of ∼2 because S2O82- ions induce the former’s slow degradation; in contrast, control experiments showed that treatment of unirradiated oligo 1 with sodium dithionite did not generate 8-oxo-G lesions). The 20.4 min HPLC fractions (Figure 2) containing both sequences 1 and 2 were collected, desalted, evaporated to dryness, and digested by snake venom phosphodiesterase and alkaline phosphatase. Using HPLC-amperometric detection, we found that the nucleoside mixtures obtained after enzymatic digestion of the irradiated oligos 1 contain 8-oxo-dG. The time-dependent yields of 8-oxo-G lesions determined as Y(8-oxo-G) ) [8-oxo-dG]t/[G-oligo]0 rapidly increase up to ∼0.02 within the first 2 s of irradiation and then decrease with increasing irradiation time as the 5′-d(CCATCGCTACC) sequence is consumed by reactions with CO3•- radicals (Figure 2). Such plots are typical for

Figure 6. Time-dependent yields of the Sp and 8-oxo-G lesions generated by 308 nm laser pulse excitation (15 mJ/pulse/cm2, 10 pulse/s) of air-equilibrated buffer solution (300 mM NaHCO3 and 10 mM Na2S2O8, pH 7.5) containing 0.01 mM of the singlestranded oligo 5′-d(CCATCGCTACC).

Figure 7. Time-dependent yields of the Sp lesions generated by 308 nm laser pulse excitation (15 mJ/pulse/cm2, 10 pulse/s) of air-equilibrated buffer solution (300 mM NaHCO3 and 10 mM Na2S2O8, pH 7.5) containing 0.01 mM of the single-stranded oligo 5′-d(CCATC[8oxo-G]CTACC).

consecutive reactions in which the intermediate 8-oxo-G lesions in sequence 1 are oxidized more rapidly than guanines in the same sequence 1. Oxidation of Guanine by CO3•- Radicals in DoubleStranded DNA. The Sp adducts were also found to be the end products of oxidation of guanine by the CO3•radicals in the duplex formed by the self-complementary sequence 4 (data not shown). The MALDI-TOF mass spectrum (negative mode) of the oxidized oligo 4 isolated by reversed phase HPLC shows a major signal at m/z 3674.5 that is consistent with the formation of the Sp adduct because its mass is greater by 32 amu (M + 32) than the mass (m/z 3642.5) of the unmodified oligo 4 with an intact guanine. However, the formation of the Sp lesion in this duplex (Figure 8) occurs with lower rates than in the case of the single-stranded oligo 1 (Figure 6). The slower rates of formation of the Sp lesion are consistent with the smaller rate constant, k4, of oxidation of the double-stranded oligo 4 by CO3•- radicals than of the single-stranded sequence 1 (Table 2). The value of k4 was calculated by fitting eq 2 to the experimentally observed CO3•- decay profiles recorded at different concentrations of the duplex in the range of 0.03-0.1 mM. Although the formation of the Sp lesions in the duplex occurs with lower rates than in the case of the singlestranded oligo (compare data in Figures 6 and 8), the formation of the intermediate 8-oxo-G lesions in the double-stranded oligo 4 is still observed (Figure 8). The yields of 8-oxo-G lesions rapidly increase to a level of ∼0.002 within the first 2 s of irradiation and then gradually decrease as a function of t beyond 2 s.

DNA Lesions Derived from Oxidation of Guanine

Figure 8. Time-dependent yields of the Sp and 8-oxo-G lesions generated by 308 nm laser pulse excitation (15 mJ/pulse/cm2, 10 pulse/s) of air-equilibrated buffer solution (300 mM NaHCO3 and 10 mM Na2S2O8, pH 7.5) containing 0.005 mM of the selfcomplementary 5′-dTATAACGTTATA duplex.

Figure 9. Time-dependent evolution of the stoichiometric coefficients of Sp lesion formation in the oxidation of the singlestranded oligos 5′-d(CCATCGCTACC) and 5′-d(CCATC[8oxoG]CTACC) by CO3•- radicals. The solid lines show the best fits of the quadratic polynomial equation to the data points.

Discussion Sp Is a Terminal Product of the Site Selective Oxidation of Guanine in Oligos by CO3•- Radicals. The CO3•- radical, an important reactive intermediate in biological systems (7, 20), can initiate the site selective oxidation of guanine leading to hot alkali-labile DNA lesions (21). Here, we found that 8-oxo-G is even more susceptible to oxidation by CO3•- radicals than guanine and that Sp is the ultimate stable oxidation product of both G and 8-oxo-G in single- and double-stranded oligos (Figures 6-8). The rates of G oxidation by CO3•- radicals in single-stranded oligo are typically greater than those in DNA duplexes (Table 2). This apparent decrease in the reactivity of guanines in the duplexes is probably associated with a decreased accessibility of guanines to the CO3•- radicals in accord with our previous observations (21). The site selective oxidation of G and 8-oxo-G in oligos 1 and 2 can be characterized by the following stoichiometric coefficients:

nCR ) f × t × [CO3•-]0/[Sp-oligo]t where f is the laser pulse frequency (10 Hz), t is the irradiation time, and [CO3•-]0 is the concentration of CO3•- generated by a single laser shot. This coefficient denotes the minimum number of CO3•- radicals needed for the formation of a single Sp lesion. The plots of nCR as a function of the irradiation time (Figure 9) were obtained from the experimental data points shown in Figures 6 and 7. These values extrapolated to t ) 0 are

Chem. Res. Toxicol., Vol. 16, No. 12, 2003 1535

nCR ) 5.4 ( 0.7 and nCR ) 1.6 ( 0.7 for oligos 1 and 2, respectively. These numbers are close to the values of nCR ) 4k1/(k1 - k3) ≈ 5.2 (sequence 1) and nCR ) 4k2/(k2 - k3) ≈ 2 (sequence 2) calculated from the kinetic parameters shown in Table 2. It is evident that the oxidation of oligo 1 is associated with the consumption of ∼5 CO3•- radicals, while four CO3•- radicals are required to transform G to Sp; thus, one CO3•- radical is consumed in nonspecific oxidative processes. The oxidation of oligo 2 requires two CO3•- radicals, and the rates of 8-oxo-G oxidation are much greater than the rates of other processes (Table 2). The increase in the values of nCR as a function of increasing irradiation time (Figure 9) suggests that CO3•- radicals are also consumed in side reactions that do not lead to the production of Sp lesions. The contribution of these side reactions appears to grow as the concentrations of unreacted G and 8-oxo-G in oligos 1 and 2 diminish. The further oxidation of Sp lesions by CO3•- radicals is relatively inefficient, as is evident from a comparison of the rate constants of oxidation (Table 2) of the oligos containing either a single Sp residue (sequence 5), a single G (sequence 1), or no G at all (sequence 3). Therefore, the Sp lesion can be considered as the terminal product of the oxidation of G and 8-oxo-G by CO3•- radicals. Is 8-oxo-G a Unique Intermediate in the Site Selective Oxidation of Guanine by CO3•- Radicals? Our experimental observations of the formation of 8-oxoG in the course of oxidation of G bases by CO3•- radicals in both single- and double-stranded oligos (Figures 6 and 8) suggest that the 8-oxo-G lesion can be an intermediate in the oxidation of guanine to Sp. Although the intermediate concentrations of the 8-oxo-G lesions are very low, the 8-oxo-G could be a unique intermediate of G oxidation to Sp because the rates of oxidation of the 8-oxo-G lesions are greater than those of the G bases (Table 2). Figure 6 shows that the maximum yield of 8-oxo-G in oxidation of oligo 1 is about 0.02, which corresponds to an overall ∼0.2 µM 8-oxo-G concentration. This concentration is less than the initial concentration of CO3•- radicals (∼1 µM) detected after each laser pulse, and thus, the complete oxidation of 8-oxo-G is feasible even after a single laser shot. Figure 6 shows that the concentration of 8-oxo-G rapidly increases within the first 2 s of photoexcitation with a train of laser pulses and then slowly decays as sequence 1 is consumed by oxidative reactions. Hence, after an initial 2 s irradiation interval, the rate of change of the concentration of 8-oxo-G can be considered negligible:

- d[8-oxo-G-oligo]/dt ) [CO3•-]t{(k1 - k3)[G-oligo]t (k2 - k3)[8-oxo-G-oligo]t}/2 ≈ 0 (3) and eq 3 can be simplified to

[8-oxo-G-oligo]t/[G-oligo]t ) (k1 - k3)/(k2 - k3) On the basis of the measured rate constants summarized in Table 2, we obtain a value of (k1 - k3)/(k2 k3) ≈ 0.075. On the other hand, typical values of [8-oxoG-oligo]t/[G-oligo]t in the quasi-steady state regime (t g 2 s) of oxidation of oligo 1 utilizing the laser pulse train to inject CO3•- radicals are in the range of ∼0.035 (Figure 6). Thus, the experimentally measured 8-oxo-G concentrations relative to the level of the undamaged oligo concentrations (Figure 6) are similar in value to the

1536 Chem. Res. Toxicol., Vol. 16, No. 12, 2003

relative concentrations estimated from the rate constants (Table 2). However, the difference in this ratio may be indicative of a deviation from the simple consecutive reaction model shown in Scheme 1. Nevertheless, this rather crude analysis suggests that a significant portion of the Sp lesions is formed via the intermediate formation of 8-oxo-G lesions. A more detailed kinetic investigation of possible alternate reaction pathways of the formation of Sp was beyond the scope of this work but is presently being investigated in our laboratory. Generation of the Sp Lesions by Other Oxidants: Comparison with Oxidation by CO3•- Radicals. The 8-oxo-G lesions are more easily oxidizized than G residues as suggested by the difference in the midpoint potentials at pH 7 of the 8-oxo-dG(-H)• radical, E7 ) 0.74 V vs NHE (29), and dG(-H)• radical, E7 ) 1.29 V vs NHE (5). In agreement with these thermodynamics considerations, many oxidants are able to oxidize 8-oxoguanine faster than guanine with the concomitant formation of Sp end products. Burrows et al. reported that the one electron oxidant, IrCl62-, with the reduction potential, E° ) 0.89 V vs NHE (52), oxidizes 8-oxo-G nucleosides (34, 36) or 8-oxo-G incorporated into oligos (35, 38, 53) resulting in the formation of the Sp diastereomers; however, IrCl62- does not oxidize guanine. The Sp lesions were also detected after oxidation of 8-oxo-G in oligos by high-valence chromium complexes (39). The Sp diastereomers were identified among the end products of oxidation of 8-oxo-G nucleosides by peroxynitrite in the presence of thiols (40) and of hypochlorous acid (41-43). Oxidation of guanine in DNA to Sp by other oxidants has been also investigated. For instance, Sp diastereomers are the major products of the oxidation of G nucleosides by singlet oxygen (23). In contrast, singlet oxygen generated only traces of spiroiminodihydandoin upon the oxidation of G in DNA (23, 46). The oxidation of the G nucleosides by hypochlorous acid (42, 43), peroxyl radicals, and the triplet excited states of acetophenones (44, 45) typically lead to lower yields of Sp diastereomers than the oxidation of the 8-oxo-G nucleosides by these same oxidants. In contrast, the CO3•- radicals oxidize G residues resulting in the formation of spiroiminodihydandoin lesions as the major oxidation end products. Other products, such as guanidinohydantoin lesions, were also detected upon the oxidation of 8-oxo-G oligos by Na2IrCl6 (35, 38); gunidinohydantoin lesions and abasic sequences are also formed but in smaller amounts (Figure 2). Our results suggest that CO3•- radicals selectively oxidize both the G and the 8-oxo-G residues to the Sp lesions in DNA. The high selectivity of oxidation of the G bases by the CO3•- radicals is in accord with the midpoint potential at pH 7 of the CO3•- radical, E7 ∼ 1.7 V vs NHE (21), which is greater than E7 ) 1.29 V vs NHE for G(-H)• radicals (5) but is not sufficient in magnitude to efficiently oxidize the other natural DNA bases (A, C, and T) that are characterized by higher redox potentials (5). Biological Implications. The kinetic parameters characterizing the site selective oxidation of guanine by CO3•- radicals have been determined here for the first time. Evidence has been presented that 8-oxo-G lesions are formed as intermediates that are further oxidized by carbonate radical anions to 8-oxo-G radicals and to spiroiminodihydantoins. Because there is ample evidence that 8-oxo-G is present in cellular DNA (3, 6), only a two electron oxidation pathway is necessary to generate Sp

Joffe et al.

lesions from 8-oxo-G. Either one of these steps could involve either CO3•- radicals as the one electron oxidant as demonstrated here or other oxidizing species present in cellular environments. We have shown that the lifetimes of guanine radicals (21) and 8-oxo-G radicals (data not shown) in double-stranded oligos in vitro are of the order of seconds. While the lifetimes of the same radicals in cellular DNA are not known, the possibility that their lifetimes in vivo are also in the range of seconds suggests that two electron oxidation of guanine to 8-oxo-G and 8-oxo-G to Sp lesions is entirely possible. Burrows and co-workers have shown that Sp lesions are removed by prokaryotic (35, 53) and eukaryotic (54) glycosylases. These lesions are highly miscoding in polymerase primer extension assays in vitro (38) catalyzed by Escherichia coli DNA polymerase I (Klenow fragment) and are mutagenic in E. coli cells with mutation frequencies at least 1 order of magnitude higher than those caused by 8-oxo-G (55). The observed formation of Sp lesions from the oxidation of G and 8-oxo-G in DNA by CO3•- radicals, oxidizing species that are known to be present in cellular environments (7, 20), as well as their mutagenic properties (55), suggest that Sp lesions could play an important role in adverse effects associated with oxidative stress in vivo, unless they are removed by DNA repair enzymes before replication occurs.

Acknowledgment. This work was supported by the National Institutes of Health, Grant 5-R01-ES11589-02, and by a grant from the Kresge Foundation. Supporting Information Available: MALDI-TOF negative ion spectra of the Sp adducts prepared by the oxidation of the 5′-d(CCATC[8-oxo]CTACC) sequence by Na2IrCl6 and the reversed phase HPLC elution profiles of the Sp adducts prepared by the different methods. This material is available free of charge via the Internet at http://pubs.acs.org.

References (1) Burrows, C. J., and Muller, J. G. (1998) Oxidative nucleobase modifications leading to strand scission. Chem. Rev. 98, 11091151. (2) Cadet, J., Bourdat, A. G., D’Ham, C., Duarte, V., Gasparutto, D., Romieu, A., and Ravanat, J. L. (2000) Oxidative base damage to DNA: specificity of base excision repair enzymes. Mutat. Res. 462, 121-128. (3) Beckman, K. B., and Ames, B. N. (1997) Oxidative decay of DNA. J. Biol. Chem. 272, 19633-19636. (4) Grisham, M. B., Jourd’heuil, D., and Wink, D. A. (2000) Review article: chronic inflammation and reactive oxygen and nitrogen metabolismsimplications in DNA damage and mutagenesis. Aliment. Pharmacol. Ther. 14 (Suppl. 1), 3-9. (5) Steenken, S., and Jovanovic, S. V. (1997) How easily oxidizable is DNA? One-electron reduction potentials of adenosine and guanosine radicals in aqueous solution. J. Am. Chem. Soc. 119, 617-618. (6) Cadet, J., Bellon, S., Berger, M., Bourdat, A. G., Douki, T., Duarte, V., Frelon, S., Gasparutto, D., Muller, E., Ravanat, J. L., and Sauvaigo, S. (2002) Recent aspects of oxidative DNA damage: guanine lesions, measurement and substrate specificity of DNA repair glycosylases. Biol. Chem. 383, 933-943. (7) Augusto, O., Bonini, M. G., Amanso, A. M., Linares, E., Santos, C. C., and De Menezes, S. L. (2002) Nitrogen dioxide and carbonate radical anion: two emerging radicals in biology. Free Radical Biol. Med. 32, 841-859. (8) Goss, S. P., Singh, R. J., and Kalyanaraman, B. (1999) Bicarbonate enhances the peroxidase activity of Cu, Zn-superoxide dismutase. Role of carbonate anion radical. J. Biol. Chem. 274, 28233-28239. (9) Zhang, H., Joseph, J., Felix, C., and Kalyanaraman, B. (2000) Bicarbonate enhances the hydroxylation, nitration, and peroxidation reactions catalyzed by copper, zinc superoxide dismutase.

DNA Lesions Derived from Oxidation of Guanine

(10)

(11)

(12)

(13)

(14)

(15)

(16)

(17)

(18)

(19)

(20)

(21)

(22)

(23)

(24)

(25)

(26)

(27)

(28)

(29)

(30)

(31)

Intermediacy of carbonate anion radical. J. Biol. Chem. 275, 14038-14045. Zhang, H., Joseph, J., Gurney, M., Becker, D., and Kalyanaraman, B. (2002) Bicarbonate enhances peroxidase activity of Cu, Znsuperoxide dismutase. Role of carbonate anion radical and scavenging of carbonate anion radical by metalloporphyrin antioxidant enzyme mimetics. J. Biol. Chem. 277, 1013-1020. Lymar, S. V., and Hurst, J. K. (1995) Rapid reaction between peroxonitrite ion and carbon dioxide: Implications for biological activity. J. Am. Chem. Soc. 117, 8867-8868. Lymar, S. V., and Hurst, J. K. (1998) CO2-Catalyzed One-Electron Oxidations by Peroxynitrite: Properties of the Reactive Intermediate. J. Am. Chem. Soc. 37, 294-301. Bonini, M. G., Radi, R., Ferrer-Sueta, G., Ferreira, A. M., and Augusto, O. (1999) Direct EPR detection of the carbonate radical anion produced from peroxynitrite and carbon dioxide (published erratum appears in J. Biol. Chem. 1999, 274 (27), 19508). J. Biol. Chem. 274, 10802-10806. Squadrito, G. L., and Pryor, W. A. (2002) Mapping the reaction of peroxynitrite with CO2: energetics, reactive species, and biological implications. Chem. Res. Toxicol. 15, 885-895. Huie, R. E., Clifton, C. L., and Neta, P. (1991) Electron-transfer reaction rates and equilibria of the carbonate and sulfate radical anions. Radiat. Phys. Chem. 38, 477-481. Neta, P., Huie, R. E., and Ross, A. B. (1988) Rate constants for reactions of inorganic radicals in aqueous solution. J. Phys. Chem. Ref. Data 17, 1027-1284. Burney, S., Caulfield, J. L., Niles, J. C., Wishnok, J. S., and Tannenbaum, S. R. (1999) The chemistry of DNA damage from nitric oxide and peroxynitrite. Mutat. Res. 424, 37-49. Tien, M., Berlett, B. S., Levine, R. L., Chock, P. B., and Stadtman, E. R. (1999) Peroxynitrite-mediated modification of proteins at physiological carbon dioxide concentration: pH dependence of carbonyl formation, tyrosine nitration, and methionine oxidation. Proc. Natl. Acad. Sci. U.S.A. 96, 7809-7814. Davis, K. L., Martin, E., Turko, I. V., and Murad, F. (2001) Novel effects of nitric oxide. Annu. Rev. Pharmacol. Toxicol. 41, 203236. Ischiropoulos, H., and Beckman, J. S. (2003) Oxidative stress and nitration in neurodegeneration: cause, effect, or association? J. Clin. Invest. 111, 163-169. Shafirovich, V., Dourandin, A., Huang, W., and Geacintov, N. E. (2001) Carbonate radical is a site-selective oxidizing agent of guanines in double-stranded oligonucleotides. J. Biol. Chem. 276, 24621-24626. Shafirovich, V., Dourandin, A., Huang, W., Luneva, N. P., and Geacintov, N. E. (1999) Oxidation of guanine at a distance in oligonucleotides induced by two-photon photoionization of 2-aminopurine. J. Phys. Chem. B 103, 10924-10933. Ravanat, J. L., and Cadet, J. (1995) Reaction of singlet oxygen with 2′-deoxyguanosine and DNA. Isolation and characterization of the main oxidation products. Chem. Res. Toxicol. 8, 379-388. Berger, M., Anselmino, C., Mouret, J.-F., and Cadet, J. (1990) High performance liquid chromatography-electrochemical assay for monitoring the formation of 8-oxo-7,8-dihydroadenine and its related 2′-deoxyribonucleoside. J. Liq. Chromatogr. 13, 929-940. Zhang, L. K., Ren, Y., Rempel, D., Taylor, J. S., and Gross, M. L. (2001) Determination of photomodified oligodeoxynucleotides by exonuclease digestion, matrix-assisted laser desorption/ionization and post-source decay mass spectrometry. J. Am. Soc. Mass Spectrom. 12, 1127-1135. Ivanov, K. L., Glebov, E. M., Plyusnin, V. F., Ivanov, Y. V., Grivin, V. P., and Bazhin, N. M. (2000) Laser flash photolysis of sodium persulfate in aqueous solution with additions of dimethylformamide. J. Photochem. Photobiol., A 133, 99-104. McElroy, W. J. (1990) A laser photolysis study of the reaction of sulfate(1-) with chloride and the subsequent decay of chlorine(1-) in aqueous solution. J. Phys. Chem. 94, 2435-2441. Lymar, S. V., Schwarz, H. A., and Czapski, G. (2000) Medium effects on reactions of the carbonate radical with thiocyanate, iodide, and ferrocyanide ions. Radiat. Phys. Chem. 59, 387-392. Steenken, S., Jovanovic, S. V., Bietti, M., and Bernhard, K. (2000) The trap depth (in DNA) of 8-oxo-7,8-dihydro-2′deoxyguanosine as derived from electron-transfer equilibria in aqueous solution. J. Am. Chem. Soc. 122, 2373-2374. Shafirovich, V., Cadet, J., Gasparutto, D., Dourandin, A., Huang, W., and Geacintov, N. E. (2001) Direct spectroscopic observation of 8-oxo-7,8-dihydro-2′-deoxyguanosine radicals in double-stranded DNA generated by one-electron oxidation at a distance by 2-aminopurine radicals. J. Phys. Chem. B 105, 586-592. O’Neill, P., and Davies, S. E. (1987) Pulse radiolytic study of the interaction of SO4•- with deoxynucleosides. Possible implications

Chem. Res. Toxicol., Vol. 16, No. 12, 2003 1537

(32)

(33)

(34)

(35)

(36)

(37)

(38)

(39)

(40)

(41)

(42)

(43)

(44)

(45)

(46)

(47)

(48)

(49)

for direct energy deposition. Int. J. Radiat. Biol. Relat. Stud. Phys. Chem. Med. 52, 577-587. Wolf, P., Jones, G. D., Candeias, L. P., and O’Neill, P. (1993) Induction of strand breaks in polyribonucleotides and DNA by the sulphate radical anion: role of electron loss centres as precursors of strand breakage. Int. J. Radiat. Biol. 64, 7-18. Candeias, L. P., and Steenken, S. (1993) Electron transfer in di(deoxy)nucleoside phosphates in aqueous solution: rapid migration of oxidative damage (via adenine) to guanine. J. Am. Chem. Soc. 115, 2437-2440. Luo, W., Muller, J. G., Rachlin, E. M., and Burrows, C. J. (2000) Characterization of spiroiminodihydantoin as a product of oneelectron oxidation of 8-oxo-7,8-dihydroguanosine. Org. Lett. 2, 613-616. Leipold, M. D., Muller, J. G., Burrows, C. J., and David, S. S. (2000) Removal of hydantoin products of 8-oxoguanine oxidation by the escherichia coli DNA repair enzyme, FPG. Biochemistry 39, 14984-14992. Luo, W., Muller, J. G., Rachlin, E. M., and Burrows, C. J. (2001) Characterization of hydantoin products from one-electron oxidation of 8- oxo-7,8-dihydroguanosine in a nucleoside model. Chem. Res. Toxicol. 14, 927-938. Luo, W., Muller, J. G., and Burrows, C. J. (2001) The pHDependent Role of Superoxide in Riboflavin-Catalyzed Photooxidation of 8-Oxo-7,8-dihydroguanosine. Org. Lett. 3, 2801-2804. Kornyushyna, O., Berges, A. M., Muller, J. G., and Burrows, C. J. (2002) In Vitro Nucleotide Misinsertion Opposite the Oxidized Guanosine Lesions Spiroiminodihydantoin and Guanidinohydantoin and DNA Synthesis Past the Lesions Using Escherichia coli DNA Polymerase I (Klenow Fragment). Biochemistry 41, 1530415314. Sugden, K. D., Campo, C. K., and Martin, B. D. (2001) Direct oxidation of guanine and 7,8-dihydro-8-oxoguanine in DNA by a high-valent chromium complex: a possible mechanism for chromate genotoxicity. Chem. Res. Toxicol. 14, 1315-1322. Niles, J. C., Wishnok, J. S., and Tannenbaum, S. R. (2001) Spiroiminodihydantoin Is the Major Product of the 8-Oxo-7,8dihydroguanosine Reaction with Peroxynitrite in the Presence of Thiols and Guanosine Photooxidation by Methylene Blue. Org. Lett. 3, 963-966. Suzuki, T., Masuda, M., Friesen, M. D., and Ohshima, H. (2001) Formation of spiroiminodihydantoin nucleoside by reaction of 8-oxo-7,8-dihydro-2′-deoxyguanosine with hypochlorous acid or a myeloperoxidase-H2O2-Cl- system. Chem. Res. Toxicol. 14, 1163-1169. Suzuki, T., and Ohshima, H. (2002) Nicotine-modulated formation of spiroiminodihydantoin nucleoside via 8-oxo-7,8-dihydro-2′deoxyguanosine in 2′-deoxyguanosine-hypochlorous acid reaction. FEBS Lett. 516, 67-70. Suzuki, T., Friesen, M. D., and Ohshima, H. (2003) Identification of Products Formed by Reaction of 3′,5′-Di-O-acetyl-2′-deoxyguanosine with Hypochlorous Acid or a Myeloperoxidase-H2O2-ClSystem. Chem. Res. Toxicol. 16, 382-389. Adam, W., Arnold, M. A., Grune, M., Nau, W. M., Pischel, U., and Saha-Moeller, C. R. (2002) Spiroiminodihydantoin is a major product in the photooxidation of 2′-deoxyguanosine by the triplet states and oxyl radicals generated from hydroxyacetophenone photolysis and dioxetane thermolysis. Org. Lett. 4, 537-540. Adam, W., Arnold, M. A., Nau, W. M., Pischel, U., and SahaMoeller, C. R. (2002) A Comparative Photomechanistic Study (Spin Trapping, EPR Spectroscopy, Transient Kinetics, Photoproducts) of Nucleoside Oxidation (dG and 8-oxodG) by TripletExcited Acetophenones and by the Radicals Generated from a-Oxy-Substituted Derivatives through Norrish-Type I Cleavage. J. Am. Chem. Soc. 124, 3893-3904. Martinez, G. R., Medeiros, M. H., Ravanat, J. L., Cadet, J., and Di Mascio, P. (2002) [18O]-labeled singlet oxygen as a tool for mechanistic studies of 8-oxo-7,8-dihydroguanine oxidative damage: detection of spiroiminodihydantoin, imidazolone and oxazolone derivatives. Biol. Chem. 383, 607-617. Adam, W., Saha-Moeller, C. R., and Schoenberger, A. (1996) Photooxidation of 8-Oxo-7,8-dihydro-2′-deoxyguanosine by Thermally Generated Triplet-Excited Ketones from 3-(Hydroxymethyl)-3,4,4-trimethyl-1,2-dioxetane and Comparison with Type I and Type II Photosensitizers. J. Am. Chem. Soc. 118, 92339238. Ravanat, J. L., Remaud, G., and Cadet, J. (2000) Measurement of the main photooxidation products of 2′-deoxyguanosine using chromatographic methods coupled to mass spectrometry. Arch. Biochem. Biophys. 374, 118-127. Cadet, J., Berger, M., Buchko, G. W., Joshi, P. C., Raoul, S., and Ravanat, J.-L. (1994) 2,2-Diamino-4-[(3,5-di-O-acetyl-2-deoxy-βD-erythro-pentofuranosyl)amino]-5-(2H)-oxazolone: a Novel and

1538 Chem. Res. Toxicol., Vol. 16, No. 12, 2003 Predominant Radical Oxidation Product of 3′,5′-Di-O-acetyl-2′deoxyguanosine. J. Am. Chem. Soc. 116, 7403-7404. (50) Vialas, C., Claparols, C., Pratviel, G., and Meunier, B. (2000) Guanine Oxidation in Double-Stranded DNA by Mn-TMPyP/ KHSO5: 5,8-Dihydroxy-7,8-dihydroguanine Residue as a Key Precursor of Imidazolone and Parabanic Acid Derivatives. J. Am. Chem. Soc. 122, 2157-2167. (51) Gasparutto, D., Bourdat, A. G., D’Ham, C., Duarte, V., Romieu, A., and Cadet, J. (2000) Repair and replication of oxidized DNA bases using modified oligodeoxyribonucleotides. Biochimie 82, 19-24. (52) Margerum, D. W., Chellappa, K. L., Bossu, F. P., and Burce, G. L. (1975) Characterization of a readily accessible copper(III)peptide complex. J. Am. Chem. Soc. 97, 6894-6896.

Joffe et al. (53) Hazra, T. K., Muller, J. G., Manuel, R. C., Burrows, C. J., Lloyd, R. S., and Mitra, S. (2001) Repair of hydantoins, one electron oxidation product of 8-oxoguanine, by DNA glycosylases of Escherichia coli. Nucleic Acids Res. 29, 1967-1974. (54) Leipold, M. D., Workman, H., Muller, J. G., Burrows, C. J., and David, S. S. (2003) Recognition and Removal of Oxidized Guanines in Duplex DNA by the Base Excision Repair Enzymes hOGG1, yOGG1, and yOGG2. Biochemistry 42, 11373-11381. (55) Henderson, P. T., Delaney, J. C., Gu, F., Tannenbaum, S. R., and Essigmann, J. M. (2002) Oxidation of 7,8-Dihydro-8-oxoguanine Affords Lesions That Are Potent Sources of Replication Errors in Vivo. Biochemistry 41, 914-921.

TX034142T