The hydantoin lesions formed from oxidation of 7,8-dihydro-8-oxoguanine are potent sources of replication errors in vivo

Single-stranded DNA genomes have been constructed that site-specifically contain the 7,8-dihydro-8-oxo-2'-deoxyguanine (8-oxoG) oxidation products guanidinohydantoin (Gh) and the two stable stereoisomers of spiroiminodihydantoin (Sp1 and Sp2). The circular viral genomes were transfected into wild-type AB1157 Escherichia coli, and the efficiency of lesion bypass by DNA polymerase(s) was assessed. Viral progeny were analyzed for mutation frequency and type using the recently developed restriction endonuclease and postlabeling (REAP) assay. Gh was bypassed nearly as efficiently as the parent 8-oxoG but was highly mutagenic, causing almost exclusive G --> C transversions. The stereoisomers Sp1 and Sp2 were, in comparison, much stronger blocks to DNA polymerase extension and caused a mixture of G --> T and G --> C transversions. The ratio of G --> T to G --> C mutations for each Sp lesion was dependent on the stereochemical configuration of the base. All observed mutation frequencies were at least an order of magnitude higher than those caused by 8-oxoG. Were these lesions to be formed in vivo, our data show that they are absolutely miscoding and may be refractory to repair after translesion synthesis.     

Removal of hydantoin products of 8-oxoguanine oxidation by the Escherichia coli DNA repair enzyme, FPG

An intriguing feature of 7,8-dihydro-8-oxo-2'-deoxyguanosine (OG) is that it is highly reactive toward further oxidation. Indeed, OG has been shown to be a "hot spot" for oxidative damage and susceptible to oxidation by a variety of cellular oxidants. Recent work has identified two new DNA lesions, guanidinohydantoin (Gh) and spiroiminodihydantoin (Sp), resulting from one-electron oxidation of OG. The presence of Gh and Sp lesions in DNA templates has been shown to result in misinsertion of G and A by DNA polymerases, and therefore, both are potentially mutagenic DNA lesions. The base excision repair (BER) glycosylases Fpg and MutY serve to prevent mutations associated with OG in Escherichia coli, and therefore, we have investigated the ability of these two enzymes to process DNA duplex substrates containing the further oxidized OG lesions, Gh and Sp. The Fpg protein, which removes OG and a variety of other oxidized purine base lesions, was found to remove Gh and Sp efficiently opposite all four of the natural DNA bases. The intrinsic rate of damaged base excision by Fpg was measured under single-turnover conditions and was found to be highly dependent upon the identity of the base opposite the OG, Gh, or Sp lesion; as expected, OG is removed more readily from an OG:C- than an OG:A-containing substrate. However, when adenine is paired with Gh or Sp, the rate of removal of these damaged lesions by Fpg was significantly increased relative to the rate of removal of OG from an OG:A mismatch. The adenine glycosylase MutY, which removes misincorporated A residues from OG:A mismatches, is unable to remove A paired with Gh or Sp. Thus, the activity of Fpg on Gh and Sp lesions may dramatically influence their mutagenic potential. This work suggests that, in addition to OG, oxidative products resulting from further oxidation of OG should be considered when evaluating oxidative DNA damage and its associated effects on DNA mutagenesis.     

Oxidised guanidinohydantoin (Ghox) and spiroiminodihydantoin (Sp) are major products of iron- and copper-mediated 8-oxo-7,8-dihydroguanine and 8-oxo-7,8-dihydro-2'-deoxyguanosine oxidation

8-Oxo-7,8-dihydroguanine (8-oxoGua), an important biomarker of DNA damage in oxidatively generated stress, is highly reactive towards further oxidation. Much work has been carried out to investigate the oxidation products of 8-oxoGua by one-electron oxidants, singlet oxygen, and peroxynitrite. This report details for the first time, the iron- and copper-mediated Fenton oxidation of 8-oxoGua and 8-oxo-7,8-dihydro-2'-deoxyguanosine (8-oxodGuo). Oxidised guanidinohydantoin (Gh(ox)) was detected as the major product of oxidation of 8-oxoGua with iron or copper and hydrogen peroxide, both at pH 7 and pH 11. Oxaluric acid was identified as a final product of 8-oxoGua oxidation. 8-oxodGuo was subjected to oxidation under the same conditions as 8-oxoGua. However, dGh(ox) was not generated. Instead, spiroiminodihydantoin (Sp) was detected as the major product for both iron and copper mediated oxidation at pH 7. It was proposed that the oxidation of 8-oxoGua was initiated by its one-electron oxidation by the metal species, which leads to the reactive intermediate 8-oxoGua (+), which readily undergoes further oxidation. The product of 8-oxoGua and 8-oxodGuo oxidation was determined by the 2'-deoxyribose moiety of the 8-oxodGuo, not whether copper or iron was the metal involved in the oxidation.     

Effect of the oxidized guanosine lesions spiroiminodihydantoin and guanidinohydantoin on proofreading by Escherichia coli DNA polymerase I (Klenow fragment) in different sequence contexts

Oxidative damage to DNA by endogenous and exogenous reactive oxygen species has been directly linked to cancer, aging, and a variety of neurological disorders. The potential mutagenicity of the primary guanine oxidation product 8-oxo-7,8-dihydroguanine (Og) has been studied intensively, and much information is available about its miscoding potential in vitro and in vivo. Recently, a variety of DNA lesions have been identified as oxidation products of both guanine and 8-oxoguanine, among them spiroiminodihydantoin (Sp) and guanidinohydantoin (Gh). To address questions concerning the mutagenic potential of these secondary products of guanine oxidation, the effect of the lesions on proofreading by DNA polymerase was studied in vitro using the Klenow fragment of Escherichia coli polymerase I (Kf exo+). For the first time, k(cat)/K(m) values were obtained for proofreading of the X:N mismatches (X = Og, Gh, or Sp; N = A, G, or C). Proofreading studies of the terminal mismatches demonstrated the significance of the sequence context flanking the lesion on the 3' side. In addition, a sequence dependence was observed for Gh based on the identity of the base on the 5' side of the lesion providing evidence for a primer slippage mode if N was complementary to the 5' base. Internal mismatches were recognized by Kf exo+ resulting in the excision of the correct base pairs flanking mismatches from the 5' side. The absence of a sequence effect for the Gh- and Sp-containing duplexes can be attributed to the severe destabilization of the lesion-containing duplexes that promotes interaction with the exonuclease domain of the Klenow fragment.     

Recognition and removal of oxidized guanines in duplex DNA by the base excision repair enzymes hOGG1, yOGG1, and yOGG2

8-Oxo-7,8-dihydroguanine (OG) is susceptible to further oxidation in vitro to form two secondary oxidation products, guanidinohydantoin (Gh) and spiroiminodihydantoin (Sp). Previous work from this laboratory has shown that OG, Gh, and Sp are recognized and excised from duplex DNA substrates by the Escherichia coli DNA repair enzyme Fpg. In this report, we extend these studies to the functionally related eukaryotic OG glycosylases (OGG) from yeast and humans: yOGG1, yOGG2, and hOGG1. The hOGG1 enzyme was active only toward the removal of 8-oxoguanine, exhibiting a 1000-fold faster rate of removal of 8-oxoguanine from OG.C-containing duplexes relative to their OG.A counterparts. Duplexes containing Gh or Sp opposite any of the four natural bases were not substrates for the hOGG1 enzyme. In contrast, both yOGG1 and yOGG2 enzymes removed Gh and Sp in a relatively efficient manner from an 18 bp duplex. No significant difference was observed in the rate of reaction of Gh- and Sp-containing duplexes with yOGG1. However, yOGG2 removed Sp at a faster rate than Gh. Both yOGG enzymes exhibit a negligible dependence on the base opposite the lesion, suggesting that the activity of these enzymes may be promutagenic. Surprisingly, in the 18 bp sequence context, both yOGG enzymes did not exhibit OG removal activity. However, both removed OG in a 30 bp duplex with a different sequence surrounding the OG. The wide range of repair efficiencies observed by these enzymes with different substrates in vitro suggests that this could greatly affect the mutagenicity of these lesions in vivo. Indeed, the greater efficiency of the yOGG proteins for removal of the further oxidized products, Gh and Sp, over their 8-oxoguanine parent, suggests that these lesions may be the preferred substrates in vivo.     

The substrate specificity of MutY for hyperoxidized guanine lesions in vivo

The DNA damage product 7,8-dihydro-8-oxo-2'-deoxyguanine (8-oxoG) is a commonly used biomarker of oxidative stress. The mutagenic potential of this DNA lesion is mitigated in Escherichia coli by multiple enzymes. One of these enzymes, MutY, excises an A mispaired with 8-oxoG as part of the process to restore the original G:C base pair. However, numerous studies have shown that 8-oxoG is chemically labile toward further oxidation. Here, we examine the activity of MutY on the 8-oxoG oxidation products guanidinohydantoin (Gh), two diastereomers of spiroiminodihydantoin (Sp1 and Sp2), oxaluric acid (Oa), and urea (Ur). Single-stranded viral genomes containing a site-specific lesion were constructed and replicated in E. coli that are either proficient in DNA repair or that lack MutY. These lesions were found previously to be potently mutagenic in repair competent bacteria, and we report here that these 8-oxoG-derived lesions are equally miscoding when replicated in E. coli lacking MutY; no significant change in mutation identity or frequency is observed. Interestingly, however, in the presence of MutY, Sp1 and Sp2 are more toxic than in cells lacking this repair enzyme.     

In vitro ligation of oligodeoxynucleotides containing C8-oxidized purine lesions using bacteriophage T4 DNA ligase

Ligases conduct the final stage of repair of DNA damage by sealing a single-stranded nick after excision of damaged nucleotides and reinsertion of correct nucleotides. Depending upon the circumstances and the success of the repair process, lesions may remain at the ligation site, either in the template or at the oligomer termini to be joined. Ligation experiments using bacteriophage T4 DNA ligase were carried out with purine lesions in four positions surrounding the nick site in a total of 96 different duplexes. The oxidized lesion 8-oxo-7,8-dihydroguanosine (OG) showed, as expected, that the enzyme is most sensitive to lesions on the 3' end of the nick compared to the 5' end and to lesions located in the intact template strand. In general, substrates containing the OG.A mismatch were more readily ligated than those with the OG.C mismatch. Ligations of duplexes containing the OA.T base pair (OA = 8-oxo-7,8-dihydroadenosine) that could adopt an anti-anti conformation proceeded with high efficiencies. An OI.A mismatch-containing duplex (OI = 8-oxo-7,8-dihydroinosine) behaved like OG.A. Due to its low reduction potential, OG is readily oxidized to secondary oxidation products, such as the guanidinohydantoin (Gh) and spiroiminodihydantoin (Sp) nucleosides; these lesions also contain an oxo group at the original C8 position of the purine. Ligation of oligomers containing Gh and Sp occurred when opposite A and G, although the overall ligation efficiencies were much lower than those of most OG base pairs. Steady-state kinetic studies were carried out for representative examples of lesions in the template. Km increased by 90-100-fold for OG.C-, OI.C-, OI.A-, and OA.T-containing duplexes compared to that of a G.C-containing duplex. Substrates containing Gh.A, Gh.G, Sp.A, and Sp.G base pairs exhibited Km values 20-70-fold higher than that of the substrate containing a G.C base pair, while the Km value for OG.A was 5 times lower than that for G.C.     

Formation of spiroiminodihydantoin nucleoside from 8-oxo-7,8-dihydro-2′-deoxyguanosine by nitric oxide under aerobic conditions

When 8-oxo-7,8-dihydro-2′-deoxyguanosine in potassium phosphate buffer of pH 7.4 was bubbled by nitric oxide at room temperature under aerobic conditions, two major products were formed. They were identified as the diastereomers of spiroiminodihydantoin deoxyribonucleoside on the basis of their identical ESI-MS and UV spectra and HPLC retention times with those of the major products in reaction of 8-oxo-7,8-dihydro-2′-deoxyguanosine with hypochlorous acid. A 1000-fold excess of 2′-deoxyguanosine did not inhibit the reaction of 8-oxo-7,8-dihydro-2′-deoxyguanosine with nitric oxide. The results suggest that an 8-oxo-7,8-dihydroguanine moiety formed in DNA may react with nitric oxide in the presence of oxygen molecule generating spiroiminodihydantoin in humans.     

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)

The low redox potential of 8-oxo-7,8-dihydroguanine (OG), a molecule regarded as a marker of oxidative damage in cells, makes it an easy target for further oxidation. Using a temperature-dependent method of synthesis, the oxidation products of OG, guanidinohydantoin (Gh) and/or its isomer iminoallantoin (Ia) as well as spiroiminodihydantoin (Sp), have been site-specifically incorporated into DNA oligomers. Single nucleotide insertion and primer extension experiments using Escherichia coli Kf exo(-) DNA polymerase were carried out under "standing start" and "running start" conditions in various sequence contexts. dAMP and dGMP were found to be inserted opposite these OG oxidation products. Steady-state kinetic studies show that the Gh/Ia.G base pair yields a lower K(m) value compared to the Sp.G pair or X.A (X = Gh/Ia or Sp). Running start experiments using oxidized and unoxidized OG-containing templates showed enhanced full extension in the presence of all four dNTPs. A sequence preference for efficiency of extension was found when Gh/Ia and Sp are present in the DNA template, possibly leading to primer misalignment. Full extension is more efficient for the templates containing two Gs immediately 3' to the lesions compared to two As. Although these lesions cause a significant block for DNA elongation, results show that they are more easily bypassed by the polymerase when situated in the appropriate sequence context. UV melting studies carried out on duplexes mimicking the template/primer systems were used to characterize thermal stability of the duplexes. These experiments suggest that both Gh/Ia and Sp destabilize the duplex to a much greater extent than OG, with Sp being most severe.     

Cellular 8-oxo-7,8-dihydro-2'-deoxyguanosine 5'-triphosphate pyrophosphohydrolase activity of human and mouse MTH1 proteins does not depend on the proliferation rate

Mammalian MTH1 proteins, homologs of Escherichia coli MutT, are enzymes decomposing 8-oxo-7,8-dihydro-2'-deoxyguanosine 5'-triphosphate (8-oxo-dGTP) to 8-oxo-7,8-dihydro-2'-deoxyguanosine 5'-monophosphate and inorganic pyrophosphate. They play an antimutagenic role by preventing the incorporation of promutagenic 8-oxo-dGTP into DNA. MTH1 gene expression is higher in some physiological types of mammalian cells and in numerous cancer cells, but the mechanism of that upregulation still remains unclear. It has been hypothesized that MTH1 expression might be associated with a proliferation rate of the cells. Therefore, we tested this hypothesis by comparing the functional levels of MTH1 gene expression measured as the 8-oxo-dGTPase activity of its protein products in normal mouse livers and hepatectomized regenerating livers. Although the proliferation rate of the hepatocytes in the regenerating livers was much higher than that in control livers, as confirmed by immunohistochemical assay of proliferating cell nuclear antigen, the 8-oxo-dGTPase activity was not different. In a second approach, we used 57 lines of human cancer cells in which 8-oxo-dGTPase activity was measured and confronted with cell population doubling time. No significant correlations between 8-oxo-dGTPase activity and proliferation rate were observed within groups of six leukemia, eight melanoma, nine lung, seven colon, six central nervous system, six ovarian, eight renal, and seven breast cancer cell lines. Thus, we conclude that the MTH1 expression manifested as the 8-oxo-dGTPase activity of its protein products in mammalian cells is not associated with proliferation rate. Our results will help in further testing of the hypothesis that MTH1 overexpression may be a specific marker of carcinogenesis and/or oxidative stress.     

2-Hydroxy-2'-deoxyadenosine 5'-triphosphate enhances A.T --> C.G mutations caused by 8-hydroxy-2'-deoxyguanosine 5'-triphosphate by suppressing its degradation upon replication in a HeLa extract

The coexistence effects of multiple kinds of oxidized deoxyribonucleotides were examined using an SV40 origin-dependent in vitro replication system with a HeLa extract. Oxidized dGTP and dATP, 8-hydroxy-2'-deoxyguanosine 5'-triphosphate (8-OH-dGTP) and 2-hydroxy-2'-deoxyadenosine 5'-triphosphate (2-OH-dATP), were used in this study. The mutation frequency synergistically increased when the two oxidized deoxyribonucleotides were together in the reaction. 2-OH-dATP enhanced the mutagenicity of 8-OH-dGTP, since the induced mutations were A.T --> C.G transversions. The contribution of the highly error-prone DNA polymerase eta was unlikely, since similar results were observed with an XP-V cell extract. The possible involvement of 2-hydroxyadenine in the complementary (template) strand was excluded on the basis of experiments using plasmids containing 2-hydroxyadenine as templates in the reactions with 8-OH-dGTP. 2-OH-dATP suppressed hydrolysis of 8-OH-dGTP, suggesting that the inhibition of the MTH1 protein played the major role in the enhancement. These results highlight the importance of specific hydrolysis of 8-OH-dGTP for the suppression of its induced mutation.     

Cellular effects of 5-formyluracil in DNA.

5-Formyluracil is a major oxidation product of thymine, formed in DNA in yields comparable to that of 8-oxo-7,8-dihydroguanine by exposure to gamma-irradiation. Whereas the repair pathways for removal and the biological effects of persisting 8-oxo-7,8-dihydroguanine are much elucidated, much less attention has been paid to the cellular implications of 5-formyluracil in DNA. Here we review the present state of knowledge in this important area within research on oxidative DNA damage.     

Repair and mutagenic potency of 8-oxoG:A and 8-oxoG:C base pairs in mammalian cells.

Replication of the oxidative lesion 8-oxo-7,8-dihydroguanine (GO) leads to the formation of both 8-oxo-7,8-dihydroguanine:adenine (GO:A) and 8-oxo-7,8-di-hydroguanine:cytosine (GO:C) pairs. The repair and mutagenic potency of these two kinds of base pairs were studied in simian COS7 and human MRC5V1 cells using the shuttle vector technology. Shuttle vectors carrying a unique GO residue opposite either a C or an A were constructed, then transfected into recipient mammalian cells. DNA repair resulting in G:C pairs and mutation frequency, were determined using resistance to digestion by the Ngo MI restriction enzyme for screening and DNA sequencing of suspect mutants. Results showed that the GO:C mismatch was well repaired since almost no mutations were detected in the plasmid progeny obtained 72 h after cell transfection. The GO:A pair was poorly repaired since only 32-34% of the plasmid progeny contained G:C whereas two thirds contained A:T at the original site. Repair kinetics measured with a non-replicating vector deleted by 13 bp at the SV40 replication origin, showed that GO:A was slowly repaired. Only 30% of the mispairs were corrected in 12 h. During this time 100% of the plasmids containing GO:A pairs were replicated as seen by the replication kinetics in a vector with an intact SV40 replication origin. These results show that, under our experimental conditions, replication is occurring before completion of DNA repair which explains the high mutagenic potency of the GO:A mispair.     

Mutation spectra induced by replication of two vicinal oxidative DNA lesions in mammalian cells

Ionizing radiations often induce multiple and clustered DNA lesions at the site of DNA interaction. As a model, we have studied the toxicity and the mutagenicity of two adjacent oxidative bases as clustered DNA lesions in mammalian cells using shuttle vectors. The chosen oxidative lesions were 8-oxo-7,8-dihydroguanine, the formylamine residue resulting from the oxidation of a pyrimidine base and the tandem lesion 8-oxo-7,8-dihydroguanine/formylamine where both modifications are located at a vicinal position. A single-stranded DNA shuttle vector carrying a unique DNA lesion was constructed, transfected into simian COS7 cells and mutations induced after replication in mammalian cells were screened in bacteria. 8-oxo-7,8-dihydroguanine, as expected, does not affect greatly survival (70% bypass) whereas formylamine and the tandem lesions are blocking alterations, DNA polymerase bypass being of 45% and 17%, respectively. Base insertion opposite the lesion was studied. Under our experimental conditions, replication of 8-oxo-7, 8-dihydroguanine finally gives rise to guanine:cytosine pairing, rendering this lesion only slightly mutagenic. This is not the case for the formylamine that codes preferentially for adenine (71%). In addition, one-base deletions were observed targeted to the site to the lesion. Cytosine and thymine were inserted opposite the lesion with similar but low frequencies. Thus, coding properties of the formylamine render this residue very mutagenic when coming from the oxidative alteration of a cytosine. The coding properties of the tandem damage are a combination of the contribution of the two isolated lesions with a very high percentage of adenine insertion (94%) opposite the formylamine residue of the tandem lesion. The toxicity as well as the mutation spectrum of the tandem lesion allow us to speculate about the molecular mechanism with which the DNA polymerase replicates these two lesions.     

Effects of base excision repair proteins on mutagenesis by 8-oxo-7,8-dihydroguanine (8-hydroxyguanine) paired with cytosine and adenine

8-Oxo-7,8-dihydroguanine (8-oxo-Gua, also known as 8-hydroxyguanine) is a major base lesion that is generated by reactive oxygen species in both the DNA and nucleotide pool. The role of DNA glycosylases, which initiate base excision repair, in the mutagenic processes of 8-oxo-Gua in DNA and 8-oxo-7,8-dihydro-2′-deoxyguanosine 5′-triphosphate (8-oxo-dGTP, also known as 8-hydroxy-2′-deoxyguanosine 5′-triphosphate) were investigated using supF shuttle plasmids propagated in human cells. The DNA glycosylases, OGG1, MUTYH, NTH1, and NEIL1, in 293T cells were individually knocked-down by siRNAs and plasmid DNAs containing an 8-oxo-Gua:C/8-oxo-Gua:A pair, and 8-oxo-dGTP plus unmodified plasmid DNA were then introduced into the knocked-down cells. The knock-down of OGG1, MUTYH, NTH1, and NEIL1 resulted in a significant increase in G:C → T:A transversions caused by the 8-oxo-Gua:C pair in the shuttle plasmid. The knock-down of MUTYH resulted in a reduction in A:T → C:G transversions induced by 8-oxo-dGTP and the 8-oxo-Gua:A pair, but the knockdown of OGG1, NTH1, and NEIL1 had no effect on mutagenesis. These results indicate that all of the above DNA glycosylases suppress mutations caused by 8-oxo-Gua:C in DNA. In contrast, it appears that MUTYH enhances A:T → C:G mutations caused by 8-oxo-dGTP.     

A DNA oligomer containing 2,2,4-triamino-5(2H)-oxazolone is incised by human NEIL1 and NTH1

The nucleobase derivative, 2,2,4-triamino-5(2H)-oxazolone (Oz), is an oxidation product of guanine or of 8-oxo-7,8-dihydroguanine that causes G-to-C transversions in DNA. Human NEIL1 (hNEIL1) and NTH1 (hNTH1) are homologues of two prokaryotic base excision repair enzymes, FPG/NEI and NTH, respectively. Here, we demonstrated that hNEIL1 and hNTH1 cleave Oz sites as efficiently as 5-hydroxyuracil sites. Thus, hNEIL1 and hNTH1 can repair Oz lesions. Furthermore, the nicking activities of these enzymes are largely independent of nucleobases opposite Oz; this finding indicates that removing Oz from Oz:G and Oz:A base pairs might cause an increase in the rate of point mutations in human cells.     

Unique response of double-stranded oligonucleotides containing a single 8-oxo-7,8-dihydroguanine to gamma rays in the frozen aqueous state at 77 K

Two kinds of double-stranded oligonucleotides containing a single 8-oxo-7,8-dihydroguanine were labeled with (32)P at their 5' ends and exposed to gamma rays in the frozen aqueous state at 77 K, where both direct and quasi-direct effects of ionizing radiation predominate. Analysis of the oligonucleotides with 20% denaturing polyacrylamide gel electrophoresis revealed no difference in the immediate induction of strand breaks between oligonucleotides containing 8-oxo-7,8-dihydroguanine and their corresponding oligonucleotides with normal guanine, but piperidine-sensitive damage was induced more frequently in the former than in the latter. Sequence analysis of irradiated oligonucleotides showed that not only 8-oxo-7,8-dihydroguanine but also its neighboring bases and the cytosine residue that is paired to it became piperidine-sensitive in both oligonucleotides. These results suggest that 8-oxo-7,8-dihydroguanine, its neighboring bases and the opposite cytosine are candidates for radiation damage hot spots.     

Substantial decrease of urinary 8-oxo-7,8-dihydroguanine, a product of the base excision repair pathway, in DNA glycosylase defective mice

Genome integrity is maintained via removal (repair) of DNA lesions and an increased load of such DNA damage has been linked to numerous pathological conditions, including carcinogenesis and ageing. 8-Oxo-7,8-dihydroguanine is one of the most critical lesions of this type. The free 8-oxo-7,8-dihydroguanine produced by the action of a specific DNA glycosylase is a potential source of this compound in urine. To date, there has been no direct, experimental evidence demonstrating that urinary 8-oxo-7,8-dihydroguanine is produced by the base excision repair pathway. For clarification of this issue, we applied a recently developed methodology which involved high performance liquid chromatography pre-purification followed by gas chromatography with isotope dilution mass spectrometric detection to compare the urinary excretion rate of 8-oxo-7,8-dihydroguanine in wild type and OGG1 glycosylase knock out mice. Our study revealed a 26% reduction in urinary level of 8-oxo-7,8-dihydroguanine in OGG1 deficient mice in comparison with the wild type strain. This clearly indicates that the mouse OGG1 glycosylase contributes significantly to the generation of urinary 8-oxo-7,8-dihydroguanine. Therefore, urinary measurements of 8-oxo-7,8-dihydroguanine may be attributed to DNA damage and repair, which in turn suggests that they may be useful in studying associations between DNA repair and disease.     

8-Oxo-7,8-dihydroguanine and 8-oxo-7,8-dihydro-2'-deoxyguanosine levels in human urine do not depend on diet

In the present study, we used the method involving HPLC pre-purification followed by gas chromatography with isotope dilution mass spectrometric detection for the determination of 8-oxo-7,8-dihydro-2'-deoxyguanosine (8-oxodGuo) and 8-oxo-7,8-dihydroguanine (8-oxoGua) in human urine. The mean levels of 8-oxoGua and 8-oxodGuo in the urine samples of the subjects on unrestricted diet were respectively 1.87 nmol/kg 24 h (±0.90) and 0.83 nmol/kg 24h (±0.49), and in the case of the groups studied, they did not depend on the applied diet. The sum of the amounts of both compounds in urine can give information about the formation rate of 8-oxoGua in cellular DNA. It is also likely that the levels of modified nucleo-base/side in urine sample are reflective of the involvement of different repair pathways responsible for the removal of 8-oxodGuo from DNA, namely base excision repair (BER) and nucleotide excision repair (NER).