Oxidative DNA modifications.

Oxidative DNA modifications are frequent in mammalian DNA and have been suggested an important mechanism in carcinogenesis, diabetes and ageing. The foundations for this suggestion are: Evidence for the importance of oxidative DNA modifications in cancer development is: high levels of oxidative lesions in cancer tissue; highly conserved and specific DNA repair systems targeting oxidative lesions; high levels of oxidative DNA lesions in oxidative DNA repair knock-out animals; defective repair of oxidative lesions in cancer-prone progeria syndromes; reduced cancer incidence in populations with high dietary antioxidant intake; and increased oxidative stress to DNA in tobacco smokers. Conflicting evidence for a relation between oxidative stress to DNA and cancer is: disagreement about the true levels and occurrence of the oxidative lesions in vivo; failure to identify the localization of oxidative lesions in important genes, e.g. tumor suppressor and oncogenes; lack of evidence that the oxidative lesions induce mutations in vivo; no cancer development in animals knocked-out for specific DNA repair enzymes in spite of high tissue levels of oxidative lesions; and unchanged cancer rates after antioxidant interventions in large clinical controlled and randomized trials. The rate of DNA oxidation has been estimated from urinary excretion of repair products and it is evident that if these lesions were not repaired, a large part of DNA would be oxidized to a degree not compatible with living. The methodologies by which oxidative DNA modifications are measured cover a wide and different range, advantages and disadvantages will be presented. One particular problem is artificial oxidation, and methods to prevent such artifacts will be presented together with results from a large interlaboratory standardization program. The methodology by which the lesions can be measured is complicated and prone to artifacts during DNA isolation, digestion, derivatization and maybe even during the separation procedure proper prior to detection. A large effort from 20+ laboratories supported by a grant from the EU has reduced artifacts considerably and work towards interlaboratory standardization of the methodology is in progress. The presently agreed "normal" levels of the most frequent known lesion 8-oxodG is about 5 per million dG's in DNA. A comprehensive evaluation of the evidence, from chemistry to clinical and epidemiological trials, linking oxidative modifications to cancer will be given. Finally, an estimate of the quantitative role oxidative DNA modifications play among the multiplicity of other insults is given. While there is no question that all of these oxidative mechanisms do exist, quantitative data on their importance for the human situation do not exist. Prospective human studies that can provide such quantitative data on different mechanisms are underway.