Quantitation, comparison, and persistence of the diazonium ion derived purine adducts in DNA of three cyclic nitrosamines in cells.

The three cyclic nitrosamines N-nitrosomorpholine (NMOR), N-nitrosopyrrolidine (NPYR), and N-nitrosopiperidine (NPIP) are potent carcinogens and mutagens that cause tumors in many experimental animal models including rats and hamsters at various bodily sites including the liver, respiratory tract, and esophagus. The mutagenic and carcinogenic potential of these compounds, like nitrosamines in general, follows from metabolic -hydroxylation, typically mediated by P450 enzymes. The -hydroxylated compounds spontaneously decompose, with concomitant rearrangement and opening of the cyclic ring, to a reactive diazonium ion on one end of the molecule and a reactive aldehyde on the other end. The presence of this bifunctional reactivity may underlie the carcinogenic potency of these cyclic compounds. The DNA damage spectrum may include not only mono-alkylation products from irreversible diazonium ion chemistry (forming a pendent aldehyde lesion, PAL) but also the more severe cross-linking products following PAL reaction with nearby amines on DNA. Diazonium ions are very reactive electrophiles capable of irreversible reaction with heteroatoms in DNA and nucleosides. Previous researchers from our lab initially studied the test-tube reactions of the corresponding primary diazonium ion of NMOR with nucleosides and DNA to quantitate the yields of alkylation at various purine heteroatoms. Of the initial purine adducts in DNA the greatest amounts of alkylation were seen at the N7-Gua, O6-Gua, and N3-Ade heteroatoms. Following additional experiments utilizing the primary diazonium ions of NPYR and NPIP with nucleosides and DNA a relative comparison of alkylation levels between the three carcinogens was then possible. NPIP derived alkylations are mysteriously 50-80 fold lower than the qualitatively similar NMOR and NPYR yields following direct comparison under identical experimental conditions with nucleosides. The test-tube reactions with DNA also yielded puzzlingly low relative levels of NPIP alkylation, in one case as low as 100-300 fold lower than the corresponding NMOR alkylation level. The major aim of the research presented here was to extend the comparisons of NMOR, NPYR, and NPIP primary diazonium ion alkylation levels from test-tube conditions to the cellular environment where the presence of repair processes, differences in rates of diffusion, and other biologically relevant conditions may be allowed to impact on the alkylation efficiency, relative yields, and persistence of DNA damage. GM2E1 cells are a human fibroblast cell line that express rat P450 2E1, an isoform of P450 capable of metabolically activating NMOR, NPYR, and NPIP. The cells were grown in tissue culture and treated with carcinogen before isolation of the DNA and quantitation of the DNA alkylation via comparison to authentic standards on HPLC and LC-MS/MS. The first specific aim of the research, to quantitate and compare the relative alkylation levels in DNA at N7-Gua, O6-Gua, and N3-Ade, following continuous exposure of the cells to 4mM carcinogen for 24hr, 48hr, and 72hr showed that all three carcinogens have levels of alkylation that are qualitatively similar in cells relative to the differences observed in test-tube reactions. In general, yields were 0.0003-0.00003% of unadducted nucleobase (qualitatively similar to the mean of yields from test-tube experiments). Relative yields were in the order: N7-Gua ≥ O6-Gua > N3-Ade (NMOR, N3-Ade not detected); N7-Gua > O6-Gua > N3-Ade (NPYR, N3-Ade not detected), and N7-Gua > N3-Ade > O6-Gua (NPIP). In general, NMOR gives greater yields than NPIP which gives greater yields than NPYR, a stark contrast from the relative yields in test-tube reactions. The second specific aim of the research was to quantitate and compare the relative persistence of the alkylation damage mentioned above. Following 24hr exposure to carcinogen and then replacement of carcinogen containing media with carcinogen free media the cells were allowed to recover for 0hr, 24hr, and 48hr before quantitation as above. Relative yields were the same order of magnitude as the continuous exposure experiments even following cessation of carcinogen exposure. Persistence of adducts for 24hr was observed for all adducts detected and for 48hr in the case of all N7-Gua adducts, O6-Gua (NMOR), and N3-Ade (NPIP). Despite the structural similarities of the carcinogens their relative levels of persistence vary from heteroatom to heteroatom. For all adducts detected, at least 24hr of persistence was observed, potentially of mutagenic consequence. In addition to the first two specific aims, an additional set of experiments was performed to assay the presence or absence of repair of O6-Gua adducts by a specific repair protein (alkylguanine DNA alkyltransferase, AGT). Despite levels of AGT activity within GM2E1 cells being sufficient to repair methylation damage that is 200 fold higher than yields from the cyclic carcinogens, no repair of any pendent aldehyde lesions was observed under all of the experiments conducted here