Nitrosated glycine derivatives react with DNA to form O6-carboxymethyl-2'-deoxyguanosine (O6-CMdG) and O6-methyl-2'-deoxyguanosine (O6-MedG) adducts concurrently. O6-CMdG is not repaired by O6-alkylguanine alkyltransferases and might be expected to lead to mutations via a similar mechanism to O6-MedG. Potassium diazoacetate (KDA) is a stable form of nitrosated glycine and its ability to induce mutations in the p53 gene in a functional yeast assay was studied. Treatment of a plasmid containing the human p53 cDNA sequence with KDA afforded readily detectable levels of O6-CMdG and O6-MedG. The treated plasmid was used to transform yeast cells and coloured colonies harbouring a p53 sequence with functional mutations were detected. Recovery of the mutated plasmids followed by DNA sequencing enabled the mutation spectrum of KDA to be characterised. The most common mutations induced by KDA were substitutions with >50% occurring at GC base pairs. In contrast to the methylating agent methylnitrosourea which gives predominantly (>80%) GC-->AT transitions, KDA produced almost equal amounts of transitions (GC-->AT) and transversions (GC-->TA and AT-->TA). This difference is probably due to a different mode of base mispairing for O6-CMdG compared with O6-MedG. The pattern of mutations induced by KDA was very similar to the patterns observed in mutated p53 in human gastrointestinal tract tumours. These results are consistent with the hypothesis that nitrosation of glycine (or glycine derivatives) may contribute to characteristic human p53 mutation profiles. This conclusion is borne out by recent observations that O6-CMdG is present in human DNA both from blood and exfoliated colorectal cells and is consistent with recent epidemiological studies that have concluded that endogenous nitrosation arising from red meat consumption is related to an increased risk of colorectal cancer.

Immunoslot blot assays have been used for the analysis of many DNA adducts but problems are frequently encountered in achieving reproducible results. Each step of the assay has been systematically examined and it was found that the major problems are in the DNA fragmentation step and the use of the manifold apparatus. Optimisation was performed upon both the malondialdehyde-deoxyguanosine adduct (M 1 dG) and the O 6 -carboxymethyldeoxyguanosine adduct (O 6 CMdG) to demonstrate the applicability to other DNA adducts.Blood samples from the EPIC study (n = 162) were analysed for M 1 dG adducts and the data showed no correlation with adduct levels in other tissues indicating that the EPIC blood samples were not useful for studying M 1 dG adducts. Blood samples from a processed meat vs vegetarian diet intervention (n = 6) were analysed for O 6 CMdG and many were below the limit of detection. The reduction of background adduct levels in standard DNA was investigated using chemical and whole-genome amplification approaches. The latter gave a sensitivity improvement of 2.6 adducts per 10 7 nucleotides for the analysis of O 6 CMdG.Subsequent reanalysis for O 6 CMdG showed a weakly significant increase in O 6 CMdG on the processed meat diet compared with the vegetarian diet, demonstrating that further studies are warranted.