Document 91V0bDB6vDnkOqrdvaYpGwkpL

et al. Biochemistry 1991, 30, 6283-6289 6283 ristry nger- Xodo, L. E., Manzini, G., Quadrifoglio, F,. van der Marel, G. A., & van Boom, J. 14. (1988a) Biochemistry 27, 632 1-6326. akyo. Xodo, L. E., Manzini, G.,Quadrifoglio, F., van der Marel arns, G . A., & van Boom, J. H. (1988b) Biochemistry 27, 6327-6331. ZoCloallnegd-z, D.(1986) B.S.. Che.mistry Honors T . 77, 289. man, - 1 13. s 26, I 28, man, 26, K. J. auer, 298, 289, ight, hatt, vmp. 122, 978) . 13, Exp. Proc. Ikryf 1. J . , 13, :o, I. Iein, . K., Ncleic Res. S. R. v 28, v 28, le, J. 1967. arel, . 14, Formation of 8-Hydroxy(deoxy)guan osine and Generation of Strand Br Guanine Residues in DNA by Singlet Oxygen+ Thomas P. A. Devasagayam,t Steen Steenken,s Maik S. W. Obendorf,t Wolfgang A. Schulz,*and Helmut Sies*J Institut fiir Physiologische Chemie I , Universitat Diisseldor- Moorenstrasse 5. I)-4000 Dusseldor- FRG, and Max-Planck-lnstitut fiir Strahlenchemie, Stijtslrasse 34-36. D-4330 Mulheim. FUG Received Nouember 19. 1990; Ueuised Manuscript Receiued February 14, 1991 35\ :I..+3%' - -.i; I ' ABSTRACT: Singlet molecular oxygen (IO2) was generated in aqueous solution (H,O or D 2 0 ) at 37 "C by the thermal dissociation of the endoperoxide of 3,3'-( 1,Cnaphthylidene) dipropionate (NDPO,). Guanosine and deoxyguanosine quench '0, with overall quenching rate constants of 6.2 X lo6 M-'s-l and 5.2 X lo6 M-' s-l, respectively. Reaction with '0, results in the formation of 8-hydroxyguanosine (8-OH-Guo) and 8-hydroxydeoxyguanosine (8-OH-dGuo), respectively, with a yield of 1.5% at 1 mM substrate with an NDPOz concentration of 40 mM; a corresponding 8-hydroxy derivative is not formed from deoxyadenosine. In D 2 0 the yield of 8-OH-Guo is 1.5-fold that in H20. Sodium azide suppresses 8-OH-Guo and 8-OH-dGuo production. In contrast, the hydroxyl radical scavengers, terf-butanol, 2-propanol, or sodium formate, do not decrease the production of the 8-OH derivatives. The formation of 8-OH derivatives is significantly increased (2-5-fold) by thiols such as dithiothreitol, glutathione, cysteine, and cysteamine. With use of a plasmid containing a fragment of the mouse metallothionein I promoter (pMTP3') and a novel end-labeling technique, the position of '0,-induced single-strand breaks in DNA was examined. Strand breaks occur selectively at dGuo; no major differences (hot spots) were observed between individual guanines. 8 g# 2t I 53 A A",. --1 - 3.;$- . ~ b$ \ '25 .. 1 clectronically excited molecular oxygen (singlet oxygen, I2)l may be generated by photochemical reactions through insfer of excitation energy to ground-state oxygen (30,fr)om suitable excited triplet-state sensitizer (photoexcitation). It .n also be produced in biological systems by dark reactions hemiexcitation), e.g., in lipid peroxidation, and by enzyme actions such as those catalysed by lactoperoxidase, lipLygenase, and chloroperoxidase (Cadenas & Sies, 1984; anofsky, 1989). IO2is produced during photooxidation of variety of biological compounds and xenobiotics. Since '0, relatively long-lived, with half-times in the range of mioseconds, considerable diffusion of singlet oxygen is possible ih a radius estimated to be in the range of LOO A (Schnuriger Bourdon, 1968; Moan, 1990). Singlet oxygen has been shown to be capable of inducing NA damage and to be mutagenic (for review, see Piette 990)). The guanine moiety has been observed to become cdroxylated at C8 on photolysis of oxygenated DNA solutions the presence of the sensitizer methylene blue (Floyd et al., j89). It was concluded that IO, is the species responsible ir this reaction, in agreement with the fact that IO, reacts .eferentially with free guanine nucleotides (Piette & Moore, 382; Cadet et al., 1983; Kawanishi et al., 1986). Hence, it of interest to ascertain whether IO, is capable of forming 'Supported by National Foundation for Cancer Research, Bethesda. :D. T.P.A.D. was financially supported by Kernforschungsanlage ilich, F.R.G..and the Bhabha Atomic Research Centre, Bombay, India. *To whom correspondence should be addressed. Ilnstitut fiir Physiologische Chemie 1, Universilat DUsseldorf. 'Max-Planck-lnstitut fur Strahlenchemie. 8-OH derivatives with guanosines and study possible factors influencing their formation. 8-Hydroxylation of guanine has also been identified as an important process in OH' radical induced damage to DNA (Floyd et al., 1986, 1988; Kasai et al., 1986; Aruoma et al., 1989). 8-OH-dGuo is a reaction product that can be easily measured and that has therefore been used as an indicator of oxidative DNA damage in vivo (Floyd et al., 1986; Kasai et al., 1986). l h e '0,-induced damage to DNA also leads to strand breaks (Wefers et al., 1987; Di Mascio et al., 1989a, 1990). Decuyper-Debergh et al., (1987) have shown that '0,-induced mutagenicity results from single nucleotide substitutions occurring predominantly at the guanosine residues. However, the method is indirect since it involves measuring base substitutions after repair. Using a newly developed method that allows the detection of end-labeled nicked fragments on sequencing gels, we assign here the base position at which lO,-induced single-strand breaks occur in plasmid DNA and check the possibility of the occurrence of "hot spots". Recent studies (Rougee et al., 1988; Kaiser et al., 1989; Devasagayam et al., 1991a) have shown that thiols and sulfur-containing amino acids quench IO2. Interestingly, if performed in the presence of DNA, this quenching is accompanied by a large increase in the number of strand breaks -- ' Abbreviations: IO,, singlet molecular oxygen; NDP, 3.3'-( I .4- naphthylidene) dipropionate; NDPO,, endoperoxide of 3,3'-( I,4naphthylidene) dipropionate; EDTA, ethylenediaminetetraaceticacid; DETAPAC, diethylenetriaminepentaaceticacid; OH'. hydroxyl radical; 8-01 I-Ciuo, 8-lrydroxyguanoine; 8-01i-dGuo. 8-hydroxydeoxyguanosine; MTP, metallothionein promoter. 0006-2960/9 I /0430-6283$02.50/0 0 1991 American Chemical Society 6284 Biochemistry, Vol. 30. No. 25, 1991 Scheme I: Plasmid phfTP3' Used for Detection o f Single-Slrntid Breaks by End Labeling" C' "Thc l'ragnicnts resulting rrom pMTP3' after cleavage with Kpnl and Clol are shown. Only 3'-recessed ends can be labeled by Sequenase 2 . 0 C' indicates [a-'*PJdCTP. The thick line indicates the strand of the metallothionein promoter in which single-strand breaks arc dckctablc over a range of about 300 b (see Figure 6). (Devasagayam et al., 1991b). On the other hand, thiols were reported to decrease oxidative damage to DNA induced by ionizing radiation (Fahey, 1988;Held, 1988) and mutagenicity induced by oxygen radicals (De Flora et al., 1989). Conversely, thiols can damage DNA in the presence of metal ions (Reed & Douglas, 1989) or can be mutagenic under certain conditions (Glatt et al., 1983; Ross et al., 1986; Glatt, 1989). Therefore, in view of this complex effect of thiols on the results of oxidation reactions it was of interest to study '02-induced changes in the guanine (deoxy)nucleosides. EXPERIMENTAL PROCEDURES Reagents. Deuterium oxide (99.8%), DL-dithiothreitol, DL-CySteine, glutathione, deoxyguanosine, deoxyadenosine,and diethylenetriaminepentaacetic acid (DETAPAC) were from Sigma. Guanosine was from Ega-Chemie. Cysteamine hydrochloride was purchased from Schuchardt. All other chemicals were from Merck. Water used in our studies was purified by Millipore filtration. The endoperoxide of the disodium salt of 3,3'-( 1,4-naphthylidene) dipropionate (NDPO2) was prepared as described (Di Mascio & Sies, 1989). The product was identified by 'H NMR and IR spectroscopy. Restriction enzymes were from Boehringer Mannheim and were regularly checked for endonuclease activity. Plasmids. Plasmid pBR322 DNA was prepared by use of the Qiagen Plasmid kit, and the effect of N D P 0 2 on singlestrand break formation was studied as described previously (Di Mascio et al., 1989a). The 850-bp EcoRl-Bglll fragment from pMTSV (kindly provided by Dr. R. D. Palmiter, Seattle) comprising the murine metallothionein 1 promoter (Stuart et al., 1984) was cloned into the EcoRI and BamHI site of the polylinker of pBluescript KS+ to yield pMTP. pMTP3' was obtained from pMTP by deleting the E~oRI--BstEll-(5') fragment and religating the fragments after filling in with Klenow polymerase. This plasmid t h u s contains a fragment of the mouse MT-I promoter flanked by several additional restriction sites (thick line in Scheme I). Generation and Quenching of IO2 by Guanosine and Deoxyguanosine. Singlet oxygen was generated by the thermal dissociation of the endoperoxide of 3.3'-( 1,4-naphthylidene) dipropionate (NDPO,) yielding 3,3'-( 1,4-naphthylidene) dipropionate (NDP), IO2, and triplet molecular oxygen (Di Mascio & Sies, 1989) (reaction I ) . COONa Devasagayam et al. COONa COONa PNDPO? COONa PNDP The solvent used to estimate quenching constants was 50 +mM sodium phosphate buffer in D20,pD 7.4, pD being taken as the pH measured with a glass electrode 0.4 pH units (Salomaa et al., 1964). At 37 'C, 3 mL of the buffer was placed in a thermostated, closed glass cuvette. A total of 15-30 PLof an NDP02 stock solution in D 2 0was added, to give a final concentration of 5 mM NDP02. The resulting infrared - +photoemission of IO2(reaction 2) was monitored at 1270 nm O2(IAJ 02(3A-8) hv( 1270 nm) (2) with use of a germanium diode photodetector (Di Mascio & Sies, 1989). At the maximum of the monomol emission intensity, achieved within 5-6 min, 10-50 pL of the freshly prepared quencher (guanosine or deoxyguanosine) in deaerated 50 mM sodium phosphate buffer in D,O, pD 7.4, was added and the resulting intensity was recorded. The overall +quenching constant k, k, was calculated from Stern-Volmer plots of the dependence on guanosine or deoxyguanosine concentrations of the emission intensities at 1270 nm (see Figure 1 ) and the IO2 decay constant of 1.9 X IO4 s-l, as determined according to Valduga et al. (1988) using pho- toexcitation of methylene blue as IO2source and time-resolved spectroscopy (courtesy of Professor S. Braslavsky and Dr. D. Mlrtire, Mulheim, F.R.G.). Formation of 8-OH-Guo and 8-OH-dGuo with IO2. The reaction mixture for the assay contained 1 mM of guanosine or deoxyguanosine in 50 mM sodium phosphate buffer in 40, pD 7.4, at 37 "C. The reaction was started by adding NDPOl at a final concentration of 40 mM and stopped by adding IO mM sodium azide. The 8-OH derivatives formed were iden- tified and quantitated by use of high-pressure liquid chromatography on a Nucleosil-5-Ct8(4.6 X 125 mm) column with optical and electrochemical detection. The mobile phase used for the separation of 8-OH-Guo was 2 mM KH2P0.,/20 m M NaC104/4% methanol, pH 4.0, at a flow rate of 0.8 mL/min; for 8-OH-dGuo methanol was 10%. Under these conditions, the retention times for 8-OH-Guo and 8-OH-dCuo were 17 and 21 min, respectively. The absorption spectra and retention times of the 8-OH-Guo and 8-OH-dGuo produced by IO2were compared to those obtained by reaction with radiation-chemically produced hydroxyl radicals (Dizdaroglu, 1985;Candeias and S.S., unpublished results) and found to be the same and identical, respectively. The absorption spectrum of 8-OHdGuo produced by IO2 was also compared to that of the compound as described by Culp et al. ( I 989) and to that of the authentic 8-OH-dGuo kindly provided by Dr. H . Kasai, National Cancer Research Center, Tokyo, Japan. Calibration of the yield of 8-OH-dGuo was performed by measuring the absorption of 8-OIl-dGuo at 293 nm and taking ~ ( 2 9 n3m) = 9700 M-' cm-' (Culp et al., 1989), using 8-bromoguanine as an internal standard and taking e(250 nm) = 13940 M-I cm-l and ~ ( 2 9 3nm) = 4850 M-' cm-I. The electrochemical detector was operated at 0.7 V. Generation 01Single-Strand Breaks in Plasmid DNA, Single-strand breaks were generated by incubating 200 pg of plasmid pMTP DNA/mL with 40 mM NDPO, in 50 m M 8-Hy cj r( sodiuni I M asci o were tak addit ion ples i n i l phoresis around 1 coriipare Analy inductio dialyzed mM Tri changes. preci pita enzymes specifical I n the pl clal/Pg with 3 ut polylinke of the pls with 5 p( (USB, C 27-base sulting fs moter in exonucle themselvc of EDTA directly traction polyacry 44.5 mM tivity Ioai samples. were run Sequenci with use phosphat exposed I were scat G S X L s( RESULsT Single guanosin by guanc Stern-Vc constant, physical action ral constant is almost 2'-deoxyl '0,emis Forma tracings measure( Figure 2. to NDP( buffer in by a fact to a largc '0, also guanosin ever, whc In et al. 1 8-Hydroxyguanosine Formation By Singlet Oxygen 30* (1) was 50 g taken H units fer was f 15-30 ) give a nfrared 270 nm (2) scio & ,ion infreshly ierated added overall Jolmer nosine n (see s-l, as 5 Pho:solved Dr. D. . The iosine ' D20, DP02 ng 10 idenchroI with :used ) mM /min; tions, re 17 :ntion were hemdeias :and OHf the at of asai, stion ;the nm) nine M-1 iical NA. :g of niM sodium phosphate buffer, pD 7.4 in 80% D20,at 37 O C (Di ~ Mascio et al., 1989a). DNA samples incubated with NDP were taken as controls. Reactions were terminated by the addition of 10 mM sodium azide and by immersing the samples in ice. Single-strand breaks were detected by electrophoresis on 0.8% agarose gels. I n a typical 2-h incubation, around 50% of supercoil was converted to relaxed circle as :ompared to 5 1 0 % in controls. Analysis for Position of Single-Strand Breaks. After the induction of single-strand breaks, the reaction mixture was dialyzed against STE buffer (100 mM sodium chloride, I O mM Tris, I mM EDTA, pH 7.5) for 72 h with several changes. In some experiments, DNA was concentrated by precipitation with ethanol before digestion with restriction enzymes. To label the 3' end of the nicked fragment of DNA specifically, two consecutive restriction digests were performed. In the plasmid pMTP3' (Scheme I), digestion with 3 units of Clal/fig of DNA produces two 5' protruding ends. Digestion with 3 units of KpnI/jtg of DNA cleaves off a short piece of polylinkcr leaving behind a 3' protruding end on the remainder of the plasmid. End labeling at 3' recessed ends by incubation with 5 pCi of [a-32P]dCTP/pgand 2 units of Sequenase 2.0 (USB, Cleveland, OH) therefore occurs on a short specific 27-base fragment from the polylinker and on fragments resulting from single-strand breaks in the metallothionein promoter insert (thick line in Scheme 1). Sequenase 2.0 lacks exonuclease activity and does not label at the strand breaks themselves. The labeling reaction was terminated by addition of EDTA (20 mM), and the reaction mixture was either used directly or further purified and concentrated by phenol extraction and ethanol precipitation before loading onto a 6% polyacrylamide/5M urea gel in TBE buffer (44.5 mM Tris, 44.5 mM boric acid, 1 rnM EDTA, p F 1 8.3). Total radioactivity loaded was similar for NDP control and NDP0,-treated samples. On every gel, sequencing lanes of the promotor insert were run in parallel for the identification of the nicked sites. Sequencing reactions were performed by the Sanger method with use of Sequenase and %-labeled deoxynucleotide triphosphates and MI 3 reverse primer. Gels were dried and exposed for various times for Kodak films. Autoradiographs were scanticd by employing an LKB laser densitometer with GSXL software. RESULTS Singlet Oxygen Quenching by Guanosine and Deoxyguanosine. The quenching of monomol photoemission by '0, by guanosine in shown in Figure IA. Figure 1 B shows the +Stern-Volmer plot, the overall singlet oxygen quenching rate constant, k, k,, being 6.2 X lo6 M-' s-', where k, is the physical quenching rate constant and k, is the chemical reaction rate constant. The value for the overall quenching rate constant of deoxyguanosine is 5.2 X IO6 M-' s-l. This number isalmobt identical with that (5.3 X IO6 M-'s-') measurcd for 2'-deoxyguanosine 5'-monophosphate by use of time-resolved '0,emission decay techniques (Lee & Rogers, 1987). Formation of 8-OH Deriuatives by Singlet Oxygen. HPLC tracings of the formation of 8-hydroxy(deoxy)guanosine as measured by electrochemical and UV detection are shown in Figure 2. 8-OW-Guo formation due to exposure to guanosine to NDP0,-generated IO, in D 2 0buffer is compared to H20 buffer in Figure 2 and Table I, the yield being higher in D,O by a factor of 1.5. Enhanced formation of D 2 0 is attributed to a larger half-life of singlet oxygen in D 2 0(Monroe, 1985). '0, also induced the formation of 8-OH-dGuo from deoxyguanosine (Figure 2 , right-hand traces, and Table I). However, when deoxyadenosine was exposed to IO2,there was no Biochernislry. Vol. 30. No. 25. I991 6285 A NDPO2 Guonosine (5mMI t2MI I1 B 7 J1.3 a0 1 2- 1.1- 0 5 10 Tirne(rninl /I Guonosine 7 0 0 2 04 0 6 08 10 QuencherConcentrotion ImMI FIGURE 1: (A) Quenching of NDP0,-generated singlet oxygen monomol photoemission at 1270 nm by guanosine. (B) Stern-Volmer plots for the quenching of singlet oxygen by guanosine ( 0 )and deoxyguanosine (A) in 50 m M sodium phosphate buffer in D 2 0 , pD 1.4. Table 1: Formation of 8-OH-Guo by NDPO,-Generated Singlet Oxygen" additions time 8-OH-Guo 8-OH-dGuo (mid (rM) (PM) none 60 nil nil NDP 60 nil N Db NDPO,, D,O buffer 0 2.0 nil 60 15.6 13.6 NDPO,, HZObuffer 0 I .6 ND 60 10.3 ND ~~ " A I mM solution of guanosine or deoxyguanosine was exposed% 40 iiiM NDP or NI)f'02 in 50 m M sodium phosphate buffer in D,O (pD 7.4) or H20(pH 7.4). The reaction mixture was incubated at 37 "C,and the reaction was stopped by adding 10 mM sodium azide. The 8-OH-Guo or 8-OH-dGuo formed was determined as described in Ex- pcriinental Proccdures. Not determined. formation of the corresponding 8-OH product, in agreement with similar results of Cadet et ai. (personal communication). Production of 8-hydroxydeoxyadenosine, however, occurs readily by radiation-chemically generated OH' radicals (Cadet & Berger, 1985; Dizdaroglu, 1985; Dizdaroglu & Bergtold, 1986;Steenken, 1989; Dizdaroglu & Gajewsky, 1990). Table I also shows that there is no formation of 8-OH-Guo when guanosine is incubated at 37 "C in the presence of NDP. 6286 Biocheinistry, Vol. 30. No. 25, 1991 Devasagayarn et al. 8-1 I\ BC Table Form; +]N 0 9 0 I-n4 c! 4 d FIGURE 2: Detection by HPLC of the 8-OH-Guo (A, B) and 8-OH-dGuo (C) formation from Guo or dGuo. Singlet oxygen was generated from NDPOz (40 mM) in 50 m M sodium phosphate buffer, pD 7.4 in DzO (A, C) or H 2 0 (B). Detection was by UV absorbance (top) and electrochemistry (bottom). a I 200 I7 250 300 Wavelengthlnm) 350 FIGURE 3: Absorption spectrum of 8-OHdGuo (b) produced by singlet oxygen from N D P 0 2 as compared to authentic 8-OH-dGuo (a). Figure 3 shows that the UV absorption spectrum of 8-OHdGuo produced by '0,and that of the authentic reference compound are practically identical. The absorption spectra of 8-OH-dGuo and 8-OH-Guo produced by IO2and OH' are also similar (not shown). On the basis of 8-bromoguanine as internal standard and with use of optical detection, the concentration of 8-OH-dGuo formed from the incubation of 1 mM deoxyguanosine with 40 mM NDPO, was determined to be 15 pM, corresponding to a yield of 1.5% on the basis of deonyguanosine. Sodium azide (2 mM) significantly decreases the yield of 8-OH-Guo as well as that of 8-OM-dGuo whereas the OH' scavengers terr-butanol (100 mM), 2-propanol (100 mM), or sodium formate (10 mM) are ineffective. The metal chelators EDTA and DETAPAC likewise have no effect, suggesting that metal ions are not involved in the formation of the 8-OH derivatives. The time course of '0,-induced formation of 8-011-dGuo is similar to that of single-strand breaks i n plasmid pBR322 DNA (Di Mascio et al., 1989a; Devasagayam et ai., 199lb) (Figure 4). The formation of 8-014-Guo observed decreases from a plateau at pD 6-7 to <IO% at pD 9 (Figure 5 ) . Modulation of Singlet Oxygen Induced 8-OH-Guo Formation by Thiols. Thiols significantly increase the '02-induced formation of 8-OH-Guo or 8-OH-dGuo (Table 11). The enhancing effect of thiols, as exemplified by dithiothreitol, was not significantly altered by the OH' scavengers rert-butanol O b , Lo 0 30 60 90 120 lncubolion Timelrninl FIGURE 4: Time course of formation by singlet oxygen of 8-OH-dGuo and single-strand breaks. Deoxyguanosine(1 mM) or plasmid pBR322 DNA (2 pg) was exposed to 4 0 mM N D P 0 2 in 50 mM sodium phosphate buffer in D 2 0 (pD 7.4) at 37 'C. 6 0 70 80 PD 90 FIGURE 5: pD dependence of singlet oxygen induced (A)and single1 oxygen/thiol (0)induced 8-hydroxyguanosine formation. Conditions are as in Figure 4;incubation was for 1 h at 37 "C. The value obtained with IOz at pD 7.4 (15.2 pM)was taken as 100%. Dithiothreitol (DTT)concentration was 10 mM. ~ no dil glt CY, __cy! "Th guanos m M so -37 "C. (IO0r EDTF involvt that as (Tamt a 1mos I induce of the Cha Plasnzi were f i methoc single-s end lab is prop gel res Single-: compar control method stricti01 activity to a mi1 be iden! in parall and orig longer I correspc Figure ( The r selective detecta t: similar treatmei react ivi t promote densiton vidual G spots" CC (Nehls e Lhionein rich regi Drscussi Reacri guanine I ward sin) react ion surement luminesci Bengal, u I-' respec we have o M-I s-I ar tm et al, d.Hydroxyguanosine Formation By Singlet Oxygen I Table II: Modulating Effect of Thiols on Singlet Oxygen Induced Formation of 8-OH-(d)Guoa thiol 8-OH-Guo (rM) I-OH-dGuo(rM) no additions d i t hiot hreitol glutathione cysteine cysteamine 14.7 69.7 59 I 41.3 31 5 15.2 64.4 66.3 ND N I) 'The incubation medium contained I m M guanosine or deoxy- panosine. 40 mM NDPO,, and IO m M of the respective thiols in 50 mM sodium phosphate buffer in D,O, pD 7.4. Incubation was for I at 17 QC. (100mM) or sodium formate (10 mM) or by metal chelators EDTA ( I mM) or DETAPAC ( 1 mM). Evidence for noninvolvement of radicals in the sensitization by thiols is the fact ihat ascorbate, a scavenger of thiol peroxyl radicals (RSOO') lTamba et ai., 1986), left the enhancing effect of dithiothreitol almost unchanged. The modulating effect of thiols on IO2induced 8-OH guanosine production is dependent on the pD ' of the buffer used (Figure 5). Characterization of 0,-Induced Single-Strand Breaks in I' 2neratdPlasmid DNA. Single-strand breaks induced in plasmid DNA iop) and were further characterized by a newly developed end-labeling method described in Experimental Procedures. Since each 4 single-strand fragment carries the same amount of label after 1 md labeling, the intensity of the bands on the sequencing gels LO is proportional to their overall abundance (Figure 6). The 'I $ gel resolves fragments up to 300 bases 5' to the Cia1 site. 2 Single-strand breaks induced by NDPO, can be identified by d :comparison to the single-strand break background in the NDP 1 control lane. This background level proved critical for the s'0 i method. All reagents employed, particularly batches of re- f ' iiriction enzymes, had to be carefully checked for nicking np activity. For the same reason, precipitation steps were kept io a minimum. The positions where nicks have occurred can be identified by comparison to the sequencing reactions run in parallel on the gel. The sequencing products contain primer and originate outside the polylinker and are therefore 96 bases longer (when the Clal site is used for labeling) than the corresponding single-strand fragments (see dotted lines in Figure 6). The result is that lO,-induced single-strand breaks occur selectively at guanines. At 120 min of treatment, all Gs were detectable on the gels and no major variation was visible. A similar pattern was already obtained after 10 min of '0, ireatment (Figure 7), indicating no major differences in the reactivity of individual Gs or parts of the metallothionein - promoter sequence. However, a close comparison of the densitometer scans shown in Figure 7 reveals that one individual G reacts slowly (arrow). By the present method, "hot spots" comparable to those obtained with alkylating agents (Nehls et al., 1984) were not detected within the metallothionein promoter sequence although it contains several CGrich regions as well as parts with lower CG content. DlscuSSlON Reactivity of '0, with dGuo and Guo. I n DNA only the guanine moiety appears to have an appreciable reactivity toward singlet oxygeii. The rate constants of singlet oxygen reaction with DNA and dGMP, determined from the measurements of the first-order decay constants of the infrared luminescence of IO,, generated by photooxidation of Rose Bengal, were found to be 5.1 X IO5 M-l s-l and 5.3 X lo6 M-l s-I respectively (Cadet et al., 1985). With the nucleosides, we have obtained the overall quenching constants of 6.2 X IO6 M-l s-' and 5.2 X lo6M-' s-l for Guo and dGuo, respectively. Biochemistry, Vol. 30, No. 25, I991 6287 AB n E C3 160- b- -I ** , I - -300 -290 -. "I 1-280 _-.. -270 . -260 e -- -250 * -240 "S *- .. -230 FIGURE 6: Autoradiograph of pMTP3' DNA end labeled after treatment with NDPO, and NDP for 120 min (central lanes) as compared to dideoxyiiucleotidesequencing products. The number of bases indicates the length of fragments. Since the Sanger sequencing products represent the strand opposite from the 3'-end-labeled NDP0,-induced fragments, the complementary bases (T', G', C', A', corresponding to the end-labeled strand) are indicated. The sequencing mixtures were loaded with the NDP0,-treated DNA (A) and 30 min earlier (B). Since the Sanger products are 96 bases longer, the fragments in B run approximately alongside the correspondingnicked fragments (matched by dotted lines). The sequence between 226 and 330 indicated in the sequencing lanes in B corresponds to bases 215 to I I 1 upstream of the MT-I transcriptional start site. These values are similar to that for dGMP but are much lower than those of carotenoids or tocopherols, which are on the order of IOio and 108 M-' s-', respectively (Foote et ai., 1968, 1974; Yamauchi & Matsushita 1977; Di Mascio el al., 1989b;Kaiser et al., 1990). The present study shows that guanosine and deoxyguanasine when exposed to NDP0,-generated singlet oxygen produce significant amounts of 8-OH derivatives, indicators of oxidative DNA damage (Floyd et al., 1986, 1988). The formation of lO,-induced 8-OH-Guo does not seem to involve metal ions as shown by lack of inhibition by metal chelators. A similar observation also has been made with singlet oxygen induced strand breaks in plasmid DNA (Devasagayam et al., 1991b). OH' radicals are also not involved, as evidenced by the lack of effect of the OH' scavengers tert-butanol. 2-propano1, and sodiuin formate. Thc enhancing effect of deuterated buffer as well as inhibition by azide points to '0,as the species responsible. This conclusion is in agreement with the interpretation by Floyd et al. (1989) and Ravanat et al. (1991), who have shown that methylene blue or phthalocyanine plus light induces the formation of 8-hydroxydeoxyguanosine. Ravanat et al. (1991) also studied the specificsinglet oxygen oxidation products of 3',5'-di-Oacetyl-2'-deoxyguanasine using 6288 Biochemistry, Vol. 30, No. 25, I991 Devasagayam et al. 8-Hyd 160 1LO II I1 120 I Number of Bases 100 90 80 70 I-_L-.- . I- .L0- c P0 NDP02-120 min nIn 4 0, -2 c 0 aQ, NDP - 120 min 0 100 2c Distance (mm) FIGURE 7: Laser densitometer scans of the autoradiograph of the gel with pMTP3' DNA treated with NDPO, for 10 and 120 rnin and NDP for 120 min. The G signal at position I17 differs significantly between the 10- and 120-min lanes (arrow). '*02.One of the main products characterized was a 4hydroxy-8-oxoguanosine derivative, which was postulated to be formed via an unstable endoperoxide as an intermediate. It is possible that such a type of endoperoxide is an intermediate also in the formation of 8-OH-Guo or 8-OH-dGuo. For the endoperoxide to give the 8-hydroxypurine, two reducing equivalents (2H) have to be supplied. I -liZO 'Idlribose Strand Break Formation. The mechanism of '02-induced strand break formation is unknown. Formation of single-strand breaks and formation of 8-hydroxydeoxyguanosine initially occur with similar kinetics (Figure 4). Schneider et ai. ( I 990) have reported that single-strand nicking occurs approximately 17-fold less frequently than does formation of 8-hydroxyguanine. These results are inconsistent with the idea that 8-hydroxydeoxyguanosine causes strand breaks. It may, however, be possible that 8-hydroxydeoxyguanosineand strand breaks are derived from a common precursor, possibly the endoperoxide. Thiol Effects. Thiols enhance the '0,-induced formation of 8-OH derivatives significantly. This enhancement was seen with dithiothreitol (Figure 5 ) and with glutathione, cysteine, and cysteamine, which occur at up to millimolar concentrations in biological tissues. This effect is i n contrast to the normally found and well-documented antioxidant and radioprotective role of thiols (Fahey, 1988; Sies, 1989). It is interesting that thiols have also an enhancing effect on the formation of strand breaks (Devasagayam et ai., 1991b). I t may be argued that the observed enhancement by thiols results from the formation of reactive radical species. However, our studies using scavengersdo not support this argument. Sevilla et al. (1990). who studied the reaction of thiols with molecular oxygen, have postulated the sequential formation of the thiyl radical (RS'), the thiol peroxyl radical (RSOO'), the sulfonyl radical (RSO,'), and the sulfonyl peroxyl radical (RS0,OO'). Since these radicals are likely to be scavengeable by ascorbate (Tamba et al., 1986), the observation (vide supra) that the sensitizing effect of thiols on the formation or 8hydroxy(deoxy)guanosine cannot be inhibited by ascorbate indicates that these radicals are riot involved in the formation of the '0,-induced damage. This conclusion is supported by the observation (S.S., unpublished results) that on production of thiol peroxyl radicals by radiation-chemical methods hydroxylation of the 8-position of (deoxy)guanosine does not occur. It is therefore tentatively suggested that the sensitization is by covalent as opposed to radical pathways. One possibility would be by interaction with the peroxide suggested by Cadet et ai. (1990), as shown below. 00 (d)ribose I (@ribose However, regarding single-strand break formation, free radicals could be involved in the extra damage generated by thiols, because this effect was partially counteracted in pBR322 by radical scavengers (Devasagayam et al., I99 I b). To summarize, it appears that of the two effects of IO2on DNA studied here, i.e., strand break formation and hydroxylation at C8 or guanine, the latter is not the cause of the former, althouf studies of thest ACKNO We I Paolo C competc analysi! Regis1 dGuo, 8: DL-di t hit 70-18-8; REFERE Aruoma Chen Cadena 221-; Cadet, 127-I Cadet, . ( 1 983 Cadet, J J. E., Culp, s. Chem Decuype EMBt De Flora Cam0 Devasag Kaiser B 9, I Devasagi (19911 Di Masc 2909-1 Di Masci & Sies Di Masci Biophj Di Masci & Sies Dizdarog Dizdarogl 182-18 Dizdarog 186, 5: Fahey, K Floyd, R. & Rick 163-17 Floyd, R. Tingey, 266-27 Floyd, R. ( 1989) Glatt, If. yam et dl. I-Hydroxyguanosine Formation By Singlet Oxygen although the strand breaks do occur at the guanines. Further studies are necessary in order to understand the mechanism of these and related phenomena. ACKNOWLEDGMENTS We gratefully acknowledge fruitful discussions with Dr. Paolo Di Mascio and his help in the preparation of NDP and competent help by H. Selbach and by H. Wuttke in the HPLC analysis. Registry No. NDP,I 18071- 16-2; 8-OH-Guo, 3868-31-3; 8-OH- dGuo. 88847-89-6;GUO1, 18-00-3;dGuo, 961-07-9; 02,7782-44-7; DI -dithiothreitol, 27565-41-9; oL-cysteine, 3374-22-9; glutathione, 70- 18-8; cysteamine, 60-23-1. and NDP ting that of strand by thiols 5. HOW'gument. 01s with rmation :soo'), 1 radical mgeable e supra) In of 8corbale rrnation rted by d uction ods hyoes not sensitis. One ggested ]ti (4) os8 REFERENCES Aruoma, 0.I., Halliwell, B., & Dizdaroglu, M. (1989) J. Biol. Chem. 264, 13024-1 3028. Cadenas, E., & Sies, H. (1984) Methods Enzymol. 105, 221-23 1. Cadet, J., & Berger, M. (1985) l n f . J . Radial. Biol. 47, 127-1 43. Cadet, J., Decarroz, C., Wang, S. Y., & Midden, W . R. (1983) Isr. J. Chem. 23, 420-429, Cadet, J., Berger, M., Decarroz, C., Mouret, J. F., Van Lier, J. E., Wagner, R. J. (1991) J. Chim. Phys. (in press). Culp, S. J., Cho, B. P., Kadlubar, F. F., & Evans, F. E. (1989) Chem. Res. Toxicol. 2, 4 14-42 1 . Decuyper-Debergh, D., Piette, J., & Van de Vorst, A. (1987) EMBOJ. 6 , 3155-3161. De Flora, S., Bennicelli, C., Zanacchi, P., D'Agostini, F., & Carnoirano, A. (1989) M u f a f .Res. 214, 153-158. Devasagayam, T. P. A., Sundquist, A. R., Di Mascio, P., Kaiser, S., & Sies, H. (1991a) J . Photochem. Photobid. n 9, 105-116. Devasagayam, T. P. A., Di Mascio, P., Kaiser, S., & Sies, H. ( I99 1b) Biochim. Biophys. Acta 1088, 409-41 2. Di Mascio, P., & Sies, H. (1989) J. Am. Chem. Soc. 1 I I , 2909-2914. Di Mascio, P., Wefers, H., Do-Thi, If.-P., Lafleur, M. V. M., & Sies, H.(1 989a) Biochim. Biophys. Acta 1007, 151-157. Di Mascio, P., Kaiser, S., & Sies, H. (1989b) Arch. Biochem. Biophys. 274, 532-538. Di Mascio, P., Menck, C. F. M., Nigro, R. G., Sarasin, A., & Sies, 1. (1990) Photochem. Photobiol. 51, 293-298. Dizdaroglu, M. (1985) Biochemistry 24, 4476-4481. Dizdaroglu, M., & Bergtold, D. S. (1986) Anal. Biochem. 156, 182-188. Dizdaroglu, M., & Gajewski, E. (1990) Methods Enzymol. 186, 530-544. ?ahey, R. C. (1988) Pharmacol. Ther. 39, 101-108. ;loyd, R. A,, Watson, J. J., Wong, P. K., Altmiller, D. H., & Rickard, R. C. (1 986) Free Radical Res. Commun. I , 163-172. h y d , R. A., West, M. S., Eneff, K. L., Hogselt, W.E., & Tmgey, D. T. ( 1988) Arch. Bioi.heni. Biophj-s. 262, 266-272. :loyd, R. A., West, M. S., Eneff, K. L., & Schneider, J. E. (1989) Arch. Biochem. Biophys. 273, 106-1 1 1 . ilatt, H. (1989) Mutagenesis 4, 221-227. n, free Ited by BR322 o sum- DNA ylation ormer, Biochemimy. Vol. 30. No. 25. 1991 6289 Glatt, H., Protic-Sabljic, M., & Oesch, F. (1983) Science 220, 961-963. Held, K . D. (1988) in Oxygen Radicals in Biology and Medicine (Simic, M. G., Taylor, K. A., Ward, J. F., & von Sonntag, C., Eds.) pp 441-448, Plenum Press, New York. Kaiser, S., Di Mascio, P., & Sies, H. (1989) in Thioctsaure (Borbe, H. O., & Ulrich, H., Eds.) pp 69-76, pmi Verlag GmbH, Frankfurt. Kaiser, S., Di Mascio, P., Murphy, M. E., & Sies, H. (1990) Arch. Biochem. Biophys. 277, 101-108. Kanofsky, J. R. (1989) Chem.-Biol. Interactions 70, 1-28. Kasai, H.,Crain, P. F., Kuchino, Y., Nishimura, A., Ootsu- yama, A., & Tanooka, H. (1986) Carcinogenesis 7 , 1949-1 851. Kawanishi, S.,Inoue, S., Sano, S., & Aiba, H. (1986) J. Biol. Chem. 261, 6090-6095. Lee, 1'. C. C., & Kodgers, M. A. J. (1987) Photochem. Phofobiol. 45, 79-86. Moan, J. (1990) J. Photochem. Photobiol. B: Biol. 6, 343-344. Monroe, B. M.(1985) in Singlet 0,(Frimer, A.A., Ed.) Vol. I pp. 177-224, CRC Press, Boca Raton. Nehls, P., Rajewsky, M. F., Spies, E., & Werner, D. (1984) EMBO J. 3, 327-332. Piette, J. (1990) J. Phorochem. Photobiol. B 4, 335-339. Piette, J. & Moore, P. D. (1982) Phorochem. Phorobiol. 35, 705-708. Ravanat, J.-L., Berger, M., Buchko, G. W., Benard, J.-F., van Lier, J., & Cadet, J. (1991) J. Chim. Phys. (in press). Reed, C. J., & Douglas, K. T. (1 989) Biochem. Biophys. Res. Commun. 162, llll-1117. Ross, D., Moldeus, P., Sies,H., & Smith, M. T. (1986) Mutat. Res. 175, 127-131. Rougee, M., Bensasson, R. V., Land, E. J., & Pariente, R. ( 1 988) Photochem. Photobiol. 47, 485-489. Salomaa, P., Schaleger, L. L., & Long, F. A. (1964) J. Am. Chenr. SOC8. 6, 1-7. Schneider, J. E., Price, S., Maidt, L., Gutteridge, J. M. C., & Floyd, R. A. (1990) Nucleic Acids Res. 18, 631-635. Schnuriger, B., & Bourdon, J. (1968) Photochem. Photobiol. 8, 361-368. Sevilla, M.D.,Becker, D., & Yan, M. (1990) Int. J.Radiat. Biol. 57, 65-81. Sies, H . ( 1989) Naturwissenschafen 76, 57-64. Steenken, S. (1989) G e m . Rev. 89, 503-520. Stuart, G. W., Searle, P. F., Chen, 1.Y.,Brinster, R. L., & Palmiter, R. D. (1984) Proc. Natl. Acad. Sci. U.S.A. 81, 73 18-7322. Tamba, M., Simone, G., & Quintiliani, M. (1986) Inr. J. Radial. Biol. 50, 595-600. Valduga, G., Nonell, S.,Reddi, E., Jori, G., & S. E. Braslavsky (1988) Phorochem. Photobiol. 48, 1-5. Wdssernian, 1. H.,& Murray, W. A., Eds. (1979) Singlet orygeti, Academic Press, New York. Wefers, I.,Schulte-Frohlinde, D., & Sies, 1. (1987) FEBS Lett. 21 I , 49-52. Yamauchi. K.,& Matsushita, S. ( I 977) Agric. Biol.Chem. 41, 1425-1430.