Method development for analysis of 8-oxodG as a biomarker for oxidative stress

Abstract: Oxidative damage to DNA gives oxidation of DNA bases and strand breaks that can cause mutations leading to cancer. Various forms of agents are capable of oxidising DNA in vivo including peroxides, singlet oxygen and UV- or gamma-irradiation. Oxidation of DNA bases occurs mainly at dG, giving stable DNA adducts such as 8-oxodG and FapydG. The use of 8-oxodG as a biomarker (expressed as the ratio 8-oxodG/dG) requires well- controlled workup conditions because 8-oxodG can be formed during the workup procedure. This may result in false high levels with difficulties in determining the true background levels, and differences between control and exposed tissue as a result. Existing methods require time consuming and warm (37°C) workup steps to extract and hydrolyse DNA before analysis. Since natural levels of 8-oxodG are low, a sensitive analytical tool is required when working with low mg amounts of cells. In this thesis, methods for workup procedures and analysis of 8-oxodG in DNA using HPLC/EC/UV and 32P-HPLC were developed. 32p was found to strongly oxidise dG into 8- oxodG and a HPLC pre-separation step was developed before 32p- postlabelling of 8-oxodG. Factors such as the purity of solutions, temperature, incubation time, peroxide removal and amount of cells during the sensitive workup procedure were found important, affecting the 8- oxodG/dG ratio. A fast (=10 min) cold (0°C), high salt non-phenol DNA extraction method procedure was developed to reduce the artifactual 8-oxodG formation. In addition, the use of catalase and the electron acceptor TEMPO was found to be protective against artifactual oxidation. A new hydroxylation mechanism of carbon compounds by peroxides is suggested, where transition metals mediate the two-electron reduction of H202, with one-electron oxidation of the reducing agent and compound respectively. This produces carbon radical cations, which are hydroxylated in water. For dG oxidation in a H202ascorbate system, this mechanism is shown to be thermodynamically favourable to a one-electron reduction of H202, producing OH° and, in addition, added OH° scavengers were found to be ineffective. Peroxide oxidation of DNA is likely, if forming DNA radical cations, to give mainly oxidation of G due to electron transfer in DNA, compared to systems generating the more randomly adding OH° (such as gamma- radiolysis of water). Background 8-oxodG/dG ratios in human lymphocytes were found to be much lower than previously reported. Oxidation of dG to 8-oxodG during workup was found to be relatively constant and to fit a mathematically defined curve that can help in estimating the true background level and the artifactual formation of 8-oxodG. As the degree of workup formation of 8-oxodG can vary on different days, it is important to include control samples during each round of workup and analysis.