Methylene blue plus light mediates 8-hydroxyguanine formation in DNA

Exposure to methylene blue (MB) plus light mediates formation of large levels of 8-hydroxyguanine in DNA. The amount of 8-hydroxy-2'-deoxyguanosine (8-OHdG) present in DNA increased as the amount of MB concentration increased throughout the 2 to 200 microM range studied and was dependent on light exposure. As the time of light exposure increased so did the 8-OHdG content to levels of about 750 8-OHdG/10(5) deoxyguanosine after 15 min of light exposure when MB was at 20 microM. Even though previous research has demonstrated that hydroxyl free radicals formed from a variety of sources mediate 8-OHdG formation in DNA, inclusion of mannitol, superoxide dismutase, catalase, and desferal in the MB plus light experiments demonstrated that these scavengers of oxygen free radical intermediates or precursors caused either no change or an increase in the 8-OHdG content of DNA exposed to MB plus light. These results appear to rule out the direct role of oxygen free radical intermediates in the primary events involved in the MB plus light mediated formation of 8-OHdG in DNA. Oxygen was essential to cause MB plus light mediated 8-OHdG formation in DNA. It was noted that when the reaction was carried out where the deuterium oxide content had been increased to 100%, the amount of 8-OHdG formed in DNA increased about threefold over that observed when comparable reactions were carried out in pure H2O. Use of the singlet oxygen scavenger 2,5-dimethylfuran has yielded variable results on the MB plus light mediated formation of 8-OHdG in DNA. The data taken collectively clearly indicate that MB plus light mediates 8-OHdG formation in DNA. The D2O data and the requirement for oxygen suggest that singlet oxygen may be an intermediate.     

Mediation of 8-hydroxy-guanine formation in DNA by thiazin dyes plus light

We have discovered that methylene blue plus light mediates the formation of 8-OHdG in DNA. Methylene blue is one of several thiazin dyes and we report here that the other thiazin dyes tested, in combination with white light, are effective in mediating 8-OHdG formation in DNA. The effectiveness of light plus the thiazin dyes in forming 8-OHdG in DNA were as follows: methylene blue greater than azure B greater than azure A greater than toluidine blue greater than thionin. Two other compounds tested; riboflavin and fuschin acid, in combination with light, caused formation of very little, if any, 8-OHdG in DNA. Thiazin dye mediated formation of 8-OHdG in DNA was not inhibited by the spin trap alpha-phenyl-t-butyl nitrone, which supports our previous observations that oxygen free radical scavengers did not inhibit methylene blue plus light mediated 8-OHdG formation in DNA. Ascorbate addition to methylene blue plus DNA, in the absence of light, was ineffective in mediating 8-OHdG formation in DNA.     

Serendipitous findings while researching oxygen free radicals

This review is based on the honor of receiving the Discovery Award from the Society of Free Radical Biology and Medicine. The review is reflective and presents our thinking that led to experiments that yielded novel observations. Critical questioning of our understanding of oxygen free radicals in biomedical problems led us to use and develop more direct and extremely sensitive methods. This included nitrone free radical spin trapping and HPLC–electrochemical detection. This technology led to the pioneering use of salicylate to trap hydroxyl free radicals and show increased flux in ischemia/reperfused brain regions and also to first sensitively detect 8-hydroxyl-2-deoxyguanosine in oxidatively damaged DNA and help assess its role in cancer development. We demonstrated that methylene blue (MB) photoinduces formation of 8-hydroxyguanine in DNA and RNA and discovered that MB sensitively photoinactivates RNA viruses, including HIV and the West Nile virus. Studies in experimental stroke led us serendipitously to discover that α-phenyl-tert-butylnitrone (PBN) was neuroprotective if given after the stroke. This led to extensive commercial development of NXY-059, a PBN derivative, for the treatment of stroke. More recently we discovered that PBN nitrones have potent anti-cancer activity and are active in preventing hearing loss caused by acute acoustical trauma.     

Spontaneous and photosensitiser-induced DNA single-strand breaks and formamidopyrimidine-DNA glycosylase sensitive sites at nucleotide resolutionin the nuclear and mitochondrial DNA of Saccharomyces cerevisiae.

A system is described for mapping oxidative DNA damage (sites sensitive to formamidopyrimidine-DNA glycosylase and single-strand breaks) at nucleotide resolution in the nuclear and mitochondrial DNA of Saccharomyces cerevisiae. Our 3' end labelling method is sensitive and was first developed using the well-studied inducer of oxidative DNA damage, methylene blue (MB) plus light. We treated yeast DNA in vitro with this so as to maximise levels of damage for assay development. Unfortunately, MB does not remain in yeast cells and yeast DNA repair mutants sensitive to active oxygen species are not sensitive to this agent, thus for in vivo experiments we turned to a polycyclic aromatic, RO 19-8022 (RO). This resulted in oxidative DNA damage when light was applied to yeast cells in its presence. The spectra of enzyme-sensitive sites and single-strand breaks induced by MB in vitro or by RO plus light in vivo or in vitro were examined in two yeast reporter genes: the nuclear MFA2 and the mitochondrial OLI1. The experiments revealed that most of the enzyme-sensitive sites and single-strand breaks induced by MB or RO plus light are at the same positions in these sequences, and that these are guanines.     

The effect of single-stranded DNA binding proteins on template/primer-independent DNA synthesis in the presence of nicking endonuclease Nt.BspD6I

Nt.BspD6I nicking endonuclease stimulates template/primer-independent DNA synthesis by Bst DNA polymerase. Template/primer-independent DNA synthesis may be one of the reasons for the formation of nonspecific products in certain DNA amplification reactions, especially those involving nicking endonucleases. Expansion of the range of DNA amplification procedures performed in the presence of nicking endonucleases makes the search for template/primer-independent DNA synthesis inhibitors highly relevant. The present work has shown that a single-strand DNA binding protein from E. coli does not affect template/primer-independent DNA synthesis regardless of the presence or absence of Nt.BspD6I. A single-stranded DNA-binding protein coded by gene 32 from bacteriophage T4 completely inhibits template/primer-independent DNA synthesis in the absence of nicking endonuclease. If nicking endonuclease is present, the protein does not suppress the synthesis of the specific product but causes a significant decrease of the amount of template/primer-independent DNA synthesis products.     

Substrate specificity of the Escherichia coli Fpg protein (formamidopyrimidine-DNA glycosylase): excision of purine lesions in DNA produced by ionizing radiation or photosensitization.

We have investigated the excision of a variety of modified bases from DNA by the Escherichia coli Fpg protein (formamidopyrimidine-DNA glycosylase) [Boiteux, S., O'Connor, T. R., Lederer, F., Gouyette, A., & Laval, J. (1990) J. Biol. Chem. 265, 3916-3922]. DNA used as a substrate was modified either by exposure to ionizing radiation or by photosensitization using visible light in the presence of methylene blue (MB). The technique of gas chromatography/mass spectrometry, which can unambiguously identify and quantitate pyrimidine- and purine-derived lesions in DNA, was used for analysis of hydrolyzed and derivatized DNA samples. Thirteen products resulting from pyrimidines and purines were detected in gamma-irradiated DNA, whereas only the formation of 2,6-diamino-4-hydroxy-5-formamidopyrimidine (FapyGua) and 8-hydroxyguanine (8-OH-Gua) was observed in visible light/MB-treated DNA. Analysis of gamma-irradiated DNA after incubation with the Fpg protein followed by precipitation revealed that the Fpg protein significantly excised 4,6-diamino-5-formamidopyrimidine (FapyAde), FapyGua, and 8-OH-Gua. The excision of a small but detectable amount of 8-hydroxyadenine was also observed. The detection of these products in the supernatant fractions of the same samples confirmed their excision by the enzyme. Nine pyrimidine-derived lesions were not excised. The Fpg protein also excised FapyGua and 8-OH-Gua from visible light/MB-treated DNA. The presence of these products in the supernatant fractions confirmed their excision.(ABSTRACT TRUNCATED AT 250 WORDS)     

Iron ion-dependent modification of bases in DNA by the superoxide radical-generating system hypoxanthine/xanthine oxidase

Damage to the bases in DNA produced by the hypoxanthine/xanthine oxidase system in the presence of iron ions was studied. The base products in DNA were measured using gas chromatography-mass spectrometry with selected ion monitoring after acidic hydrolysis of DNA and trimethylsilylation. Products identified were cytosine glycol, thymine glycol, 5,6-dihydroxycytosine, 4,6-diamino-5-formamidopyrimidine, 8-hydroxyadenine, 2,6-diamino-4-hydroxy-5-formamidopyrimidine, and 8-hydroxyguanine. These are typical hydroxyl radical-induced products of the bases in DNA. 2,6-Diamino-4-hydroxy-5-formamidopyrimidine was the major product, followed by 8-hydroxyguanine, in DNA treated with hypoxanthine/xanthine oxidase/Fe3+-EDTA. The use of Fe3+ did not cause as much damage to the bases in DNA as did the use of Fe3+-EDTA. In both systems, the formation of the products was inhibited by superoxide dismutase, catalase, dimethyl sulfoxide, mannitol, and desferrioxamine, but inhibitions were much stronger in the systems containing EDTA. Hence formation of hydroxyl radicals by a superoxide radical-assisted Fenton reaction is proposed to account for the results obtained. 2,6-Diamino-4-hydroxy-5-formamidopyrimidine, 5,6-dihydroxycytosine, 4,6-diamino-5-formamidopyrimidine, and 8-hydroxyguanine were proposed as the products in DNA to measure if one aims to measure DNA products as indices of oxidative DNA damage involving hydroxyl radicals in vivo.     

Effect of single-stranded DNA binding proteins on template/primer-independent DNA synthesis in the presence of nicking endonuclease Nt.BspD6I

In the presence of the Nt.BspD6I nicking endonuclease DNA polymerase Bst stimulates intensive template/primer-independent DNA synthesis. Template/primer-independent DNA synthesis could be the reason for appearing nonspecific DNA products in many DNA amplification reactions particularly in the reactions with using nicking endonucleases. Search of the modes for inhibition template/primer-independent DNA synthesis becomes an urgent task because of broadening the DNA amplification methods with using nicking endonucleases. We report here that the E. coli single-stranded DNA binding protein has no effect on the template/primer-independent DNA synthesis. In the absence of the nicking endonuclease the single-stranded DNA binding protein encoded by bacteriophage T4 gene 32 completely inhibits template/primer-independent DNA synthesis. This protein does not inhibit synthesis of specific DNA product in the presence of nicking endonuclease but remarkably decreases the amount of nonspecific products.     

Characterization of hydroxyl free radical mediated damage to plasmid pBR322 DNA

We have investigated hydroxyl free radical mediated damage to pBR322 DNA produced by ascorbate/iron and oxygen in a phosphate-buffered in vitro system. An observed lag phase in DNA nicking suggests a multi-target model of hydroxyl free radical attack on DNA. In the present report we further examine the model system and show that there is a "heat labile" component of the ascorbate/iron system which can be completely restored by the readdition of ascorbate. These observations have allowed us to rule out the possibility that intermediates build up in the reaction and act independently of ascorbate to increase the reaction rate. We have investigated the initial rate of OH production with two OH trapping agents, salicylate and deoxyguanosine, and find that the lag in DNA nicking is not due to a corresponding lag in the production of OH as assessed by formation of the products, dihydroxybenzoic acids and 8-hydroxydeoxyguanosine, respectively. We have found that the energy of activation for DNA supercoiled nicking is 13.9 kcal/mole and for OH trapping by salicylate is 21.1 kcal/nmole. These two activation energies are sufficiently different to suggest that the rate-limiting steps of these two reactions are different. Investigation of the rate of oxygen consumption during the ascorbate/iron-mediated DNA damage showed that oxygen was not a limiting component at any point in the reaction. The addition of catalase slowed down oxygen consumption by 31% and this data taken together with our previous observations on the model implicate hydrogen peroxide as a key intermediate in DNA damage caused by hydroxyl free radical.     

Signal amplified strategy based on target-induced strand release coupling cleavage of nicking endonuclease for the ultrasensitive detection of ochratoxin A

In this work, a new signal amplified strategy based on target-induced strand release coupling cleavage of nicking endonuclease for the ultrasensitive detection of ochratoxin A (OTA) is reported. OTA aptamer (DNA1) and OTA aptamer complementary (DNA2) were immobilized onto a magnetic bead (MB). In the presence of OTA, DNA2 was dissociated and released from the MB. The released DNA2 then hybridized with DNA3, which was linked at the 5' terminus of the amplification template and can extend along the template in the presence of Phi 29 DNA polymerase. The formed double-stranded DNA was cleaved by nicking endonuclease Nb.BbvCI and produced a short single-stranded DNA. The cleaved DNA strand generated a new site by Phi 29 DNA polymerase and the process of extension and cleavage was cyclical. Thus, a amount of the short single-stranded DNA were produced. Using DNA and ABEI labeled carboxylic silica nanoparticles chemiluminescence (CL) probe, the short single-stranded DNA could be sensitively detected. The CL intensity (ΔI) versus the concentration of OTA was linear in the range from 1.0×10(-12) to 5.0×10(-8)gmL(-1). The detection limit was 3.0×10(-13)gmL(-1), and the RSD was 3.4% at 1.0×10(-10)gmL(-1) (n=7). The developed method has been applied to detect OTA in naturally contaminated wheat samples. Due to its simplicity, sensitivity and no need of specific recognition of aptamer for cleavage, this CL bioassay offers a promising approach for the detection of OTA and other biomolecules.     

Ascorbate/iron mediation of hydroxyl free radical damage to PBR322 plasmid DNA

Plasmid PBR322 DNA has been exposed to hydroxyl free radicals generated from an ascorbate/Fe system. Hydroxyl free radical scavengers as well as the iron chelator desferroxamine and catalase inhibit the DNA nicking which occurs, but superoxide dismutase had no effect. The DNA nicking was temperature dependent, occuring more rapidly at higher temperatures. The rate of DNA nicking was accelerated by the addition of hydrogen peroxide. There was an early lag phase in DNA nicking, even though the rate of hydroxyl free radical generation, as assessed by salicylate hydroxylation, showed no lag phase. It is considered that the early hydroxyl free radical damage to DNA may be biologically very important in mutagenic and carcinogenic processes.     

Strand breakage in solutions of DNA and ethidium bromide exposed to visible light.

Breaks are introduced into DNA strands when DNA solutions containing ethidium bromide (EB) are exposed to incandescent light. The nicking rate is sensitive to the concentration of EB and the light intensity. At short exposure times, this rate is limited by photon capture and formation of an intermediate capable of nicking DNA and zero-order nicking kinetics are observed. If the EB is pre-irradiated, the nicking rate is limited by DNA concentration and first-order nicking kinetics are observed. The nicking rate is not greatly affected by the presence of a low frequency of ribonucleotides in the duplex structure. The nicking reaction produces neither double-strand breaks nor interstrand crosslinks. The nicks produced cannot be closed by DNA ligase. The fluorescent light intensities under normal laboratory conditions are insufficient to induce significant nicking.     

Photodynamic guanine modification by hematoporphyrin is specific for single-stranded DNA with singlet oxygen as a mediator

Photodynamic modification of DNA by hematoporphyrin (Hp) was characterized by the DNA sequencing technique using 32P-labeled DNA fragments, and the reaction mechanism was investigated by ESR spectroscopy. Mild photodynamic treatment of single-stranded DNA with Hp induced an alteration of guanine residues, and subsequent treatment with piperidine led to chain cleavages at each guanine residue. On the other hand, methylene blue plus light modified the guanine residues in both single-stranded and double-stranded DNA. ESR studies using 2,2,6,6-tetramethylpiperidine and 2,2,6,6-tetramethyl-4-piperidone as singlet oxygen traps demonstrated that Hp plus light produced almost the same amount of singlet oxygen as methylene blue plus light and that the photochemically generated singlet oxygen reacts significantly with guanylate but only slightly with other mononucleotides. An ESR spin destruction method revealed that photoexcited Hp generated porphyrin radical, but guanylate did not react with this radical. These results indicate that photoexcited Hp reacts with oxygen to generate singlet oxygen which oxidizes the guanine residues of single-stranded DNA and that the difference in photoreactivities of DNA with Hp and methylene blue may be explained in terms of the structural difference in their intercalating abilities.     

Generation of superoxide free radical by neocarzinostatin and its possible role in DNA damage

Spectroscopic analysis of the reduction of both nitro blue tetrazolium and ferricytochrome c induced by neocarzinostatin shows that superoxide free radical is produced during the spontaneous degradation of the antibiotic. The amount of superoxide free radical produced from neocarzinostatin is not affected by the presence of thiol, although earlier work has shown that DNA damage is stimulated at least 1000-fold by thiol. Transition metals are not involved in this reaction. Although superoxide dismutase inhibits the reduction of nitro blue tetrazolium and cytochrome c induced by neocarzinostatin, neither it nor catalase interferes with the action of neocarzinostatin on DNA, whether or not drug has been activated by thiol. The pH profiles for spontaneous base release and alkali-labile base release (a measure of nucleoside 5'-aldehyde formation at a strand break) do not correspond with that for the generation of superoxide free radical from neocarzinostatin. The same holds for supercoiled DNA cutting by neocarzinostatin chromophore in the absence of a thiol, which is an acid-favored reaction. These results indicate that the generation of superoxide free radical by the drug does not correlate with DNA damage activity, whether or not thiol is present. Furthermore, the failure of hydroxyl free-radical scavengers to inhibit drug-induced single-strand breaks in supercoiled DNA in the absence of thiol also indicates that a diffusible hydroxyl free radical is most probably not involved in this reaction.     

Hydroxyl free radical mediated formation of 8-hydroxyguanine in isolated DNA

Formation of 8-hydroxyguanine within calf thymus DNA has been studied after exposure to uv-H2O2 as a hydroxyl free radical generating system. Using high-pressure liquid chromatography with electrochemical detection, we measured the amount of 8-hydroxy-2-deoxyguanosine (8-OHdG) in the enzymatically digested DNA. The 8-OHdG content of uv-exposed DNA increased linearly with increasing H2O2 levels up to 0.03%, above which the rate of increase was less than linear. All hydroxyl free radical scavengers studied (mannitol, ethanol, thiourea, and salicylate), if present in the system when DNA was exposed to uv-H2O2, caused a decrease in the amount of 8-OHdG formed. Thiourea when incubated with damaged DNA caused a loss of 8-OHdG when it was an integral part of DNA. In contrast, thiourea did not react with the nucleoside free in solution. Reduced glutathione did not cause a decrease of 8-OHdG, either when it was an integral part of DNA, or, as the free nucleoside in solution.     

DNA damage induced by catechol derivatives

We investigated the effect of catechol derivatives, including dopa, dopamine, adrenaline and noradrenaline, on DNA damage and the mechanisms of DNA strand breakage and formation of 8-hydroxyguanine (8HOG). The catechol derivatives caused strand breakage of plasmid DNA in the presence of ADP-Fe(3+). The DNA damage was prevented by catalase, mannitol and dimethylsulfoxide, suggesting hydroxyl radical (HO..)-like species are involved in the strand breakage of DNA. Iron chelators, such as desferrioxamine and bathophenanthroline, and reduced glutathione also inhibited the DNA damage. Deoxyribose, a molecule that is used to detect HO,, was not degraded by dopa in the presence of ADP-Fe(3+). By adding EDTA, however, dopa induced the marked deoxyribose degradation in the presence of ADP-Fe(3+), indicating that EDTA may extract iron from ADP-Fe(3+) to catalyze HO. formation by dopa. Thus, EDTA was a good catalyst for HO.-generation, whereas it did not promote the strand breakage of DNA. However, calf thymus DNA base damage, which was detected as 8-HOG formation, was caused by dopa in the presence of EDTA-Fe(3+), but not in the presence of ADP-Fe(3+). The 8HOG formation was also inhibited by catalase and HO. scavengers, indicating that HO&z.rad; was involved in the base damage. These results suggest that DNA strand breakage is due to ferryl species rather than HO., and that 8HOG formation is due to HO. rather than ferryl species.     

Enhancement of oxidative DNA degradation by histidine: the role of stereochemical parameters.

The nicking of supercoiled DNA by H2O2 and ferrous iron has been studied in a variety of environmental conditions. The replicative form of phage fd DNA (fd RF DNA) was used for investigating the phenomenon. The rate of nicking was measured in 10 mM NaCl. The addition of 1 mM Tris-HCl buffer (pH 7.5) slowed down the rate of nicking, the addition of 0.1 mM histidine enhanced it. The simultaneous presence of 1 mM Tris-HCl buffer and of 0.1 mM histidine further enhanced the rate of nicking of fd RF DNA. Increasing the concentration of NaCl dramatically reduced the rate of the reaction. The degradation of fd RF DNA was determined as a function of the concentration of histidine (0-5 mM): the rate increases with concentration, reaches a maximum and then decreases. In the presence of histidine, increasing the concentration of Tris leads to a similar phenomenon. In the absence of histidine, Tris always quenches the degradation of DNA. Electron spin resonance measurements failed to detect an enhancement of the signal characteristic for the hydroxyl radical when histidine was added to the solution containing hydrogen peroxide and ferrous iron. When the nicking of DNA is achieved via the process of auto-oxidation of ferrous iron (i.e., in the absence of added H2O2), histidine only reduces the rate of reaction in a dose-dependent manner, in the explored range of concentrations. In the presence of H2O2 and ferrous iron, histidine enhances the rate of nicking of double-stranded DNA in its supercoiled as well as in its relaxed state, but fails to modify the rate of nicking of fd DNA when it is in its vegetative, single-stranded form.     

Linear nicking endonuclease-mediated strand-displacement DNA amplification

We describe a method for linear isothermal DNA amplification using nicking endonuclease-mediated strand displacement by a DNA polymerase. The nicking of one strand of a DNA target by the endonuclease produces a primer for the polymerase to initiate synthesis. As the polymerization proceeds, the downstream strand is displaced into a single-stranded form while the nicking site is also regenerated. The combined continuous repetitive action of nicking by the endonuclease and strand-displacement synthesis by the polymerase results in linear amplification of one strand of the DNA molecule. We demonstrate that DNA templates up to 5000 nucleotides can be linearly amplified using a nicking endonuclease with 7-bp recognition sequence and Sequenase version 2.0 in the presence of single-stranded DNA binding proteins. We also show that a mixture of three templates of 500, 1000, and 5000 nucleotides in length is linearly amplified with the original molar ratios of the templates preserved. Moreover, we demonstrate that a complex library of hydrodynamically sheared genomic DNA from bacteriophage lambda can be amplified linearly.     

ATP-independent renaturation of complementary DNA strands by the mutant recA1 protein from Escherichia coli

In an effort to clarify the requirement for ATP in the recA protein-promoted renaturation of complementary DNA strands, we have analyzed the mutant recA1 protein which lacks single-stranded DNA-dependent ATPase activity at pH 7.5. Like the wild type, the recA1 protein binds to single-stranded DNA with a stoichiometry of one monomer per approximately four nucleotides. However, unlike the wild type, the mutant protein is dissociated from single-stranded DNA in the presence of ATP or ADP. The ATP analogue adenosine 5'-O-3' (thiotriphosphate) appears to stabilize the binding of recA1 protein to single-stranded DNA but does not elicit the stoichiometry of 1 monomer/8 nucleotides or the formation of highly condensed protein-DNA networks that are characteristic of the wild type recA protein in the presence of this analogue. The recA1 protein does not catalyze DNA renaturation in the presence of ATP, consistent with the dissociation of recA1 protein from single-stranded DNA under these conditions. However, it does promote a pattern of Mg2+-dependent renaturation identical to that found for wild type recA protein.     

Conditions influencing yield and analysis of 8-hydroxy-2'-deoxyguanosine in oxidatively damaged DNA

We have conducted studies to obtain practical knowledge regarding the stability, digestion, and analytical determination of the content of 8-hydroxy-2-deoxy-guanosine (8-OHdG) in oxidatively damaged DNA. Utilizing H2O2 plus uv light to form oxidatively damaged DNA, we found that storage of the DNA at -20 degrees C at alkaline pH caused a significant loss of 8-OHdG, whereas storage at -20 degrees C at neutral or acidic pH prevented loss of 8-OHdG. The 8-OHdG within DNA is stable at 100 degrees C for at least 15 min. Formation of 8-OHdG within DNA using uv light and H2O2 as a hydroxyl free radical-generating system yields the highest amounts when low levels of phosphate buffer are used; but the use of Tris or citrate buffers causes a lower yield of 8-OHdG because these buffers act as scavengers for the hydroxyl free radicals. Independent assessment of hydroxyl free radical flux by the use of salicylate trapping allows assessment of competitive radical reactions. Ethanol washing of plastic microfuge tubes prior to DNA enzymatic digestion improved the yield of 8-OHdG and reduced the variability between samples. Digestion of the oxidatively damaged DNA by the use of a method involving DNase I, endonuclease, phosphodiesterase, and alkaline phosphatase produced the highest yield of 8-OHdG.