The effect of haem in red and processed meat on the endogenous formation of N-nitroso compounds in the upper gastrointestinal tract.

Red and processed meat (PM) consumption increases the risk of large bowel cancer and it has been demonstrated that haem in red meat (RM) stimulates the endogenous production of N-nitroso compounds (NOCs) within the human intestine. To investigate whether N-nitrosation occurs in the upper gastrointestinal tract, 27 ileostomists were fed diets containing no meat, or 240 g RM or 240 g PM in a randomly assigned crossover intervention design carried out in a volunteer suite. Endogenous NOC were assessed as apparent total N-nitroso compounds (ATNC) in the ileostomy output. ATNC concentration in the diets was 22 microg ATNC/kg (RM) and 37 microg ATNC/kg (PM), and 9 microg ATNC/kg in the no meat diet. Levels significantly increased to 1175 microg ATNC/kg SEM = 226 microg ATNC/kg) following the RM (P=0.001) and 1832 microg ATNC/kg (SEM=294 microg ATNC/kg) following PM (P<0.001) compared to the no meat diet (283 microg ATNC/kg, SEM=74 microg ATNC/kg). ATNC concentrations in the ileal output were equivalent to those measured in faeces in similarly designed feeding studies. Supplementation with either 1 g ascorbic acid or 400 IU alpha-tocopherol had no effect on the concentration of ATNC detected in the ileal output. In in vitro experiments, N-nitrosomorpholine (NMor) was formed in the presence of nitrosated haemoglobin, at pH 6.8 but not in the absence of nitrosated haemoglobin. These findings demonstrate that haem may facilitate the formation of NOC in the absence of colonic flora in the upper human gastrointestinal tract.

Lack of correlation between the observed stability and pharmacological properties of S-nitroso derivatives of glutathione and cysteine-related peptides.

S-Nitrosothiols (RSNOs) have been widely studied as donors of nitric oxide. In general, RSNOs are considered to be somewhat unstable; however, they are both potent vasodilators and inhibitors of platelet aggregation. In order to improve our understanding of the factors that determine the biological activity of RSNOs, the chemical stability and pharmacological activity of a series of RSNOs was determined. Results show that millimolar solutions of S-nitrosocysteine (SNOCys) and S-nitroso-L-cysteinylglycine (SNOCysGly) were the least stable, whereas S-nitroso-3-mercaptopropionic acid (SNOPROPA) and S-nitroso-N-acetyl-L-cysteine (SNONAC) were the most stable of the compounds tested. Recent evidence suggests that RSNOs, such as SNONAC, are as unstable as SNOCys at micromolar concentrations. The decomposition of certain RSNOs is catalysed by trace amounts of copper (II) ions, with this phenomenon being particularly evident for SNOCys and SNOCysGly. The decomposition of the more stable RSNOs, including S-nitroso-L-glutathione (SNOGSH) and L-gamma-glutamyl-L-cysteine (SNOGluCys), were not as sensitive to copper ions. The decomposition of the stable RSNO, SNOGSH, was more rapid in the presence of excess thiol, whereas the decay of the unstable RSNO, SNOCys, was reduced with added thiol. All RSNOs tested inhibited platelet aggregation, relaxed vascular smooth muscle, and inhibited cell growth in the nanomolar range, but their order of potency did not correlate with their chemical stability of millimolar solutions. It is apparent that the potency of an RSNO in a physiological situation will depend on the concentration of the compound present, the presence of trace metal ions such as copper, and the occurrence of transnitrosation reactions.

Lobe-specific increases in malondialdehyde DNA adduct formation in the livers of mice following infection with Helicobacter hepaticus.

Helicobacter hepaticus infection is associated with chronic hepatitis and the development of liver tumours in mice. The underlying mechanism of this liver carcinogenesis is not clear but the oxidative stress associated with H. hepaticus infection may result in induction of lipid peroxidation and the generation of malondialdehyde. Malondialdehyde can react with deoxyguanosine in DNA resulting in the formation of the cyclic pyrimidopurinone N-1,N(2) malondialdehyde-deoxyguanosine (M1dG) adduct. This adduct has the potential to cause mutations that may ultimately lead to liver carcinogenesis. The objective of this study was to determine the control and infection-related levels of M1dG in the liver DNA of mice over time, using an immunoslot-blot procedure. The level of M1dG in control A/J mouse livers at 3, 6, 9 and 12 months averaged 37.5, 36.6, 24.8 and 30.1 adducts per 10(8) nucleotides, respectively. Higher levels of M1dG were detected in the liver DNA of H. hepaticus infected A/JCr mice, with levels averaging 40.7, 47.0, 42.5 and 52.5 adducts per 10(8) nucleotides at 3, 6, 9 and 12 months, respectively. There was a significant age dependent increase in the level of M1dG in the caudate and median lobes of the A/JCr mice relative to control mice. A lobe specific distribution of the M1dG adduct in both infected and control mice was noted, with the left lobe showing the lowest level of the adduct compared with the right and median lobes at all time points. In a separate series of mice experimentally infected with H. hepaticus, levels of 8-hydroxy-deoxyguanosine were significantly greater in the median compared with the left lobe at 12 weeks after treatment. In conclusion, these results suggest that M1dG occurs as a result of oxidative stress associated with H. hepaticus infection of mice, and may contribute to liver carcinogenesis in this model.

Characterization of DNA damage at purine residues in oligonucleotides and calf thymus DNA induced by the mutagen 1-nitrosoindole-3-acetonitrile.

N-Nitrosoindoles can efficiently transfer the nitroso group to nucleophilic targets in isolated purine nucleotides, causing depurination, deamination, and the formation of a novel guanine analogue, oxanine [Lucas, L. T., Gatehouse, D., and Shuker, D. E. G. (1999) J. Biol. Chem. 274, 18319-18326]. To determine the likely biological relevance of these modification pathways, the reactivity of 1-nitrosoindole-3-acetonitrile (NIAN), a model 3-substituted N-nitrosoindole, with oligonucleotides and calf thymus DNA was examined at physiological pH and temperature. Reaction of NIAN with single-stranded oligonucleotides containing various guanine motifs resulted in the production of single-strand break products at guanine sites due to the formation of alkali-labile lesions. The number of lesions increased with NIAN concentration and incubation time. Modification of calf thymus DNA by NIAN resulted in depurination, which gave the corresponding purine bases, deamination coupled with depurination, which gave xanthine, and the formation of oxanine. The former pathway was clearly the most important, and all reaction products exhibited a dose-response relationship. Cytosine and thymine residues were inactive toward NIAN. Further studies revealed an additional product in NIAN-treated duplex DNA containing a CCGG motif that was characterized as an interstrand cross-link, the yield of which increased with increasing NIAN concentration. These results indicate that the transnitrosating ability of NIAN to modify purine residues is preserved at the macromolecular level, with guanine residues appearing to be a primary site of reaction. All of these modification processes are potentially mutagenic events if they occur in vivo.

IPCS guidelines for the monitoring of genotoxic effects of carcinogens in humans. International Programme on Chemical Safety.

The purpose of these guidelines is to provide concise guidance on the planning, performing and interpretation of studies to monitor groups or individuals exposed to genotoxic agents. Most human carcinogens are genotoxic but not all genotoxic agents have been shown to be carcinogenic in humans. Although the main interest in these studies is due to the association of genotoxicity with carcinogenicity, there is also an inherent interest in monitoring human genotoxicity independently of cancer as an endpoint. The most often studied genotoxicity endpoints have been selected for inclusion in this document and they are structural and numerical chromosomal aberrations assessed using cytogenetic methods (classical chromosomal aberration analysis (CA), fluorescence in situ hybridisation (FISH), micronuclei (MN)); DNA damage (adducts, strand breaks, crosslinking, alkali-labile sites) assessed using bio-chemical/electrophoretic assays or sister chromatid exchanges (SCE); protein adducts; and hypoxanthine-guanine phosphoribosyltransferase (HPRT) mutations. The document does not consider germ cells or gene mutation assays other than HPRT or markers of oxidative stress, which have been applied on a more limited scale.

Evaluation of the alkaline comet assay and urinary 3-methyladenine excretion for monitoring DNA damage in melanoma patients treated with dacarbazine and tamoxifen.

PURPOSE: To develop, using dacarbazine as a model, reliable techniques for measuring DNA damage and repair as pharmacodynamic endpoints for patients receiving chemotherapy. METHODS: A group of 39 patients with malignant melanoma were treated with dacarbazine 1 g/m(2) i.v. every 21 days. Tamoxifen 20 mg daily was commenced 24 h after the first infusion and continued until 3 weeks after the last cycle of chemotherapy. DNA strand breaks formed during dacarbazine-induced DNA damage and repair were measured in individual cells by the alkaline comet assay. DNA methyl adducts were quantified by measuring urinary 3-methyladenine (3-MeA) excretion using immunoaffinity ELISA. Venous blood was taken on cycles 1 and 2 for separation of peripheral blood lymphocytes (PBLs) for measurement of DNA strand breaks. RESULTS: Wide interpatient variation in PBL DNA strand breaks occurred following chemotherapy, with a peak at 4 h (median 26.6 h, interquartile range 14.75-40.5 h) and incomplete repair by 24 h. Similarly, there was a range of 3-MeA excretion with peak levels 4-10 h after chemotherapy (median 33 nmol/h, interquartile range 20.4-48.65 nmol/h). Peak 3-MeA excretion was positively correlated with DNA strand breaks at 4 h (Spearman's correlation coefficient, r=0.39, P=0.036) and 24 h (r=0.46, P=0.01). Drug-induced emesis correlated with PBL DNA strand breaks (Mann Whitney U-test, P=0.03) but not with peak 3-MeA excretion. CONCLUSIONS: DNA damage and repair following cytotoxic chemotherapy can be measured in vivo by the alkaline comet assay and by urinary 3-MeA excretion in patients receiving chemotherapy.

DNA adducts in relation to lung tumour outcome are not markers of susceptibility following a single dose treatment of SWR, BALB/c and C57BL/6J mice with N-nitrosodiethylamine.

The formation of DNA adducts by the covalent binding of genotoxic chemicals to DNA represents a valuable marker for assessing exposure to carcinogens but as yet the role of DNA adducts as a biomarker of carcinogenic susceptibility still needs to be clearly ascertained. To address this question an animal study was instigated using mice (SWR (high), BALB/c (intermediate) and C57BL/6J (low)) varying in their susceptibility to lung carcinogenesis. Groups of animals from each strain were dosed with a single intraperitoneal injection of saline or N -nitrosodiethylamine (NDEA) at 15 or 90 mg kg(-1) body weight. Lung and liver tissues were removed at different time points following dosing. Further groups of mice dosed with the same regime had urine samples collected 24 h post dosing and were then left up to 18 months to allow for the development of tumours. Immunoslot-blot analysis was used for the determination of N-7 ethylguanine (N-7EtG) and O(6) ethylguanine (O(6)EtG) adduct levels in the DNA from the tissues and gas chromatography-mass spectrometry (GC-MS) was used to determine N-3 ethyladenine (N-3EtA) adduct levels in the urine samples. Levels of alkyltransferase (ATase) were also determined in the tissues. The results showed that the DNA adduct levels and persistence were similar across the three strains of mice following dosing with 15 and 90 mg kg(-1) NDEA. High levels of adducts were observed in the urine of the BALB/c strain, implying an increased metabolic or repair capacity in this strain. However there were no differences in the levels of ATase in the lung and liver of the three strains of mice following dosing with 15 mg kg(-1) NDEA. The incidence of tumours in C57BL/6J mice was lower compared with the other two strains and showed a dose dependent increase. The results from this study show that the differences in susceptibility to lung carcinogenesis between the three strains of mice do not appear to be linked to the formation of the two adducts detected. These results imply that dosing with NDEA resulted in toxicity which may have led to cell death and induction of tumours by compensatory cell proliferation. Although these results do not allow decisive conclusions to be drawn concerning the relationship between total levels of DNA adducts and differences in carcinogenic susceptibility for the three strains of mice it is clear that the increased presence of a DNA adduct in the target tissue increases the likelihood of tumour development.

Concentration of nitric oxide products in bronchoalveolar fluid obtained from infants who develop chronic lung disease of prematurity.

AIMS: To determine if nitric oxide (NO) products (nitrate and nitrite) are increased in bronchoalveolar lavage (BAL) fluid obtained from infants who develop chronic lung disease of prematurity (CLD). METHODS: One hundred and thirty six serial bronchoalveolar lavages were performed on 37 ventilated infants (12 with CLD, 18 with respiratory distress syndrome (RDS), and seven control infants) who did not receive inhaled NO. RESULTS: During the first week of life nitrate concentration was between 25-31 micromol/l in all three groups. Thereafter, the concentration of BAL fluid nitrate decreased to 14 micromol/l and 5.5 micromol/l, respectively in the RDS and control groups by 14 days of age. In contrast, nitrate in the CLD infants remained constant until 28 days of age (31.3 micromol/l at day 14; p<0.05). In all BAL fluid samples the mean concentration of nitrite was <1.2 micromol/l throughout the first 28 days with no significant differences noted among the three groups. CONCLUSION: The similar concentration of BAL fluid nitrate in all groups during the first week of life suggest that NO may be important in the adaptation of the pulmonary circulation after birth. However, persistence of nitrate in the BAL fluid of infants with CLD during the second week may reflect pulmonary maladaptation, or, more likely, persisting pulmonary inflammation.

Efficient nitroso group transfer from N-nitrosoindoles to nucleotides and 2'-deoxyguanosine at physiological pH. A new pathway for N-nitrosocompounds to exert genotoxicity.

The endogenous formation of N-nitrosoindoles is of concern since humans are exposed to a variety of naturally occurring and synthetic indolic compounds. As part of a study to evaluate the genotoxicity of N-nitrosoindoles, the reactions of three model compounds with purine nucleotides and 2'-deoxyguanosine at physiological pH were investigated. The profiles of reaction products were identical for each of the N-nitrosoindoles and three distinct pathways of reaction could be discerned. These pathways were: (i) depurination to the corresponding purine bases, (ii) deamination, coupled with depurination, to give hypoxanthine and xanthine, and (iii) formation of the novel nucleotide 2'-deoxyoxanosine monophosphate and its corresponding depurination product oxanine in reactions with 2'-deoxyguanosine monophosphate. 2'-Deoxyoxanosine and oxanine were observed in reactions with 2'-deoxyguanosine. Further studies showed that formation of all of these products could be rationalized by an initial transnitrosation step. These results suggest that, in contrast to many other genotoxic N-nitrosocompounds which are known to alkylate DNA, the genotoxicity of N-nitrosoindoles is likely to arise through transfer of the nitroso group to nucleophilic sites on the purine bases. All of the products resulting from transnitrosation by N-nitrosoindoles are potentially mutagenic. These findings reveal a new pathway for N-nitrosocompounds to exert genotoxicity.

The effect of potassium diazoacetate on human peripheral lymphocytes, human adenocarcinoma Colon caco-2 cells, and rat primary colon cells in the comet assay.

In previous studies, N-(N'-acetyl-L-propyl)-N-nitrosoglycine (APNG) has been shown to be a potent mutagen in a variety of genotoxicity assays and a carcinogen in a limited cancer study. APNG decomposes to a carboxymethyldiazonium ion, which can also be generated from potassium diazoacetate (KDA). KDA is particularly interesting because it is a stable nitrosated derivative of glycine, one of the most common dietary amino acids. KDA has been shown to produce more O6 carboxymethyl- and O6 methyl-adducts than APNG, so it was anticipated that it might also be a potent genotoxic agent. Thus in the present study KDA has been investigated in the single cell gel electrophoresis (Comet) assay, which primarily measures DNA strand breakage. Since KDA has been shown to be formed in the gut, the genotoxic effects of KDA were investigated in vitro in human adenocarcinoma colon Caco-2 cells, and in rat primary colon cells and compared to responses from human peripheral lymphocytes. KDA induced DNA damage in the three cell types, confirming that KDA is genotoxic in a range of mammalian cells.

What is the significance of increases in background levels of carcinogen-derived protein and DNA adducts? Some considerations for incremental risk assessment.

Improvements in analytical methodology have led to the detection and quantification of 'background' levels of a number of DNA and protein adducts. Many of these adducts are derived from 'low molecular weight' reactive species which may be generated during normal physiological processes, metabolic pathways or inflammatory processes. The adducts have been detected using gas chromatography-mass spectrometry, HPLC in combination with various detection systems, 32P-postlabelling and immunoassay methods. The reliability and accuracy of many widely used methods for adduct measurements are discussed with reference to several examples where human data is available, namely 4-aminobiphenyl, malondialdehyde, methylating agents, ethylene oxide and hydroxyl radical damage. The accurate and specific quantitation of 'background' levels of damage is essential if reliable estimates of increases in risk associated with incremental increases in exposure to exogenous agents are to be calculated. In experimental studies using low dose exposures to carcinogens, such as N-nitrosodimethylamine, adduct levels in liver correlate closely with tumour incidence. In all likelihood, such relationships need to be established for each exposure and, in order to be relevant to human risk assessment, need to take into account factors such as DNA repair and mutagenic efficiency. Finally, in order to estimate the increase in cancer attributable to a given level of external exposure, it is clearly important to establish background levels of corresponding DNA damage so that the scale of the incremental increase can be calculated.

Detection of concomitant formation of O6-carboxymethyl- and O6-methyl-2'-deoxyguanosine in DNA exposed to nitrosated glycine derivatives using a combined immunoaffinity/HPLC method.

A previous observation that an N-nitroso-N-carboxymethyl derivative reacts with DNA to give both O6-carboxymethyl-2'-deoxyguanosine (O6-CMdGuo) and O6-methyl-2'-deoxyguanosine (O6-MedGuo) [Shuker, D. E. G., and Margison, G. P. (1997) Cancer Res. 57, 366-369] has been confirmed using a range of nitrosated glycine derivatives [N-acetyl-N'-nitroso-N'-prolylglycine (APNG), azaserine (AS), and potassium diazoacetate (KDA)]. In addition, mesyloxyacetic acid (MAA) was also found to give both O6-adducts in DNA. O6-CMdGuo and O6-MedGuo were assessed in enzymatic hydrolysates of treated calf thymus DNA using a combined immunoaffinity/HPLC/fluorescence procedure. The ratio of O6-CMdGuo to O6-MedGuo varied somewhat between the different compounds with APNG giving the most methylation (O6-CM:O6-Me ratio of 10) and AS the least (39), with KDA and MAA giving intermediate amounts (16 and 18, respectively). The formation of O6-MedGuo by the four compounds probably arises through decarboxylation at various stages in the decomposition pathways, but the exact mechanisms remain to be clarified. The formation of O6-MedGuo from reactions of nitrosated glycine derivatives with DNA in vitro may explain the frequent detection of this adduct in human gastrointestinal DNA, as nitrosation of dietary glycine may occur. O6-CMdGuo is likely to be a useful biomarker of this pathway in vivo and has been detected in human tissues.

A sensitive immunoslot-blot assay for detection of malondialdehyde-deoxyguanosine in human DNA.

As part of a large programme on food risk assessment, we have become Interested in the endogenous production of genotoxic agents from dietary precursors. Malondialdehyde (MDA), a product of lipid peroxidation and prostaglandin biosynthesis, is mutagenic in bacterial and mammalian systems. MDA reacts with DNA, and the major adduct (M1-dG) has been detected in healthy human liver and leukocyte DNA. Analytical methods used so far for the detection of M1-dG have not been applied to large numbers of individuals or a large variety of samples. Often, only a few micrograms of DNA from human tissues are available for analysis, and a very sensitive assay is needed to detect background levels of M1-dG in very small amounts of DNA. In this paper, we describe the development of an immunoslot-blot (ISB) assay for the measurement of M1-dG in 1 microgram of DNA. The limit of detection of the assay is about 5 adducts per 10(8) bases. The advantages of ISB over other assays for DNA adduct detection, such as the possibility of analysing 1 microgram DNA per sample and the fact that it is less time-consuming and laborious, mean that it can be more easily used for routine analysis of large numbers of samples in biomonitoring. A series of human samples was analysed, and levels of 0.3-6.43 M1-dG per 10(7) normal bases were detected in 42 gastric biopsy samples and 0.7-16.65 M1-dG per 10(7) normal bases in 28 samples of leukocyte DNA. In an initial study in five human volunteers on standardized diets, the levels of M1-dG in leukocyte DNA changed in relation to meat, vegetable and tea intake.

Determination of malondialdehyde-induced DNA damage in human tissues using an immunoslot blot assay.

Malondialdehyde (MDA) is a product of lipid peroxidation and prostaglandin biosynthesis. It is mutagenic and carcinogenic and the major adduct formed by reaction with DNA, a highly fluorescent pyrimidopurinone (M1-dG), has been detected in healthy human liver and leukocyte DNA. Analytical methods used so far for the detection of M1-dG have not been applied to a large number of individuals or variety of samples. Often, only a few microg of DNA from human tissues are available for analysis and a very sensitive assay is needed in order to detect background levels of M1-dG in very small amounts of DNA. In this paper, the development of an immunoslot blot (ISB) assay for the measurement of MI-dG in 1 microg of DNA is described. The limit of detection of the assay is 2.5 adducts per 10(8) bases. A series of human samples were analysed and levels of 5.6-9.5 (n = 8) and 3.1-64.3 (n = 42) of M1-dG per 10(8) normal bases were detected in white blood cell and gastric biopsy DNA, respectively. Results on four human samples were compared with those obtained using an HPLC/32P-post-labelling (HPLC/PPL) method previously developed and indicated a high correlation between M1-dG levels measured by the two assays. The advantages of ISB over other assays including HPLC/PPL, such as the possibility of analysing 1 microg DNA/sample and the fact that it is less time-consuming and laborious, means that it can be more easily used for routine analysis of a large number of samples in biomonitoring studies.

Stable acetaldehyde--protein adducts as biomarkers of alcohol exposure.

The consumption of alcoholic beverages has been associated with increased risks of a number of chronic disorders including cancers. It is still not clear whether ethyl alcohol or other components such as metabolites are directly involved in the carcinogenic process or whether the effects are due to the modulation of metabolism of other carcinogens. At present, there is no good biomarker of alcohol intake, particularly at low or moderate levels of consumption. A number of studies have shown the ability of the major metabolite acetaldehyde to react with proteins in vitro to give stable and unstable adducts. The interaction of acetaldehyde with model peptides, which correspond to N-terminal globin sequences, was studied. The major stable adduct was identified by mass spectrometry and NMR as a diastereoisomeric mixture of imidazolidinones. This is believed to be formed by reaction and cyclization of the initial Schiff base adduct with the N-terminal valine. Incubation of human globin with acetaldehyde (0-2 mM) yielded products which were identified as the N-terminal adducts by electrospray ionization mass spectrometry (ESI-MS) of proteolytic digests. The specificity and sensitivity of the analysis was improved by the use of on-line HPLC-ESI-MS. Tryptic digests of the modified globin which contained both the N-terminal acetaldehyde adducts of alpha-globin (heptapeptide) and beta-globin (octapeptide) were resolved. These results suggest that analysis of stable imidazolidinone adducts is a promising approach to estimation of alcohol exposure.

Analytical methods for 3-nitrotyrosine as a marker of exposure to reactive nitrogen species: a review.

Nitric oxide (NO*) is a diatomic free radical which has recently been found to have a key role in both normal physiological processes and disease states. The presence of NO in biological systems leads to the formation of reactive nitrogen species (RNS) such as peroxynitrite which reacts avidly with tyrosine residues in proteins to form nitrotyrosine (NTYR). Since peroxynitrite has a very short half-life at neutral pH, the presence of NTYR has been used as a marker of RNS production in various tissues. A number of methods for separation, detection, and quantitation of NTYR in biological samples have been developed. These methods include immunochemical techniques such as immunhistochemistry, ELISA, and Western blotting, high-performance liquid chromatography (HPLC) in combination with various detection systems including UV and electrochemical detection (ECD), gas chromatography (GC), gas chromatography-mass spectrometry (GC-MS), and electrospray mass spectrometry. In terms of sensitivity and specificity, it would appear that methods based on combinations of HPLC and various types of ECD are very versatile giving a limit of detection of 20 fmol per injection of protein hydrolysate. They are only limited by the sample quantity and the preparation that is required to achieve acceptable chromatograms. In addition to the detection of NTYR as a marker of RNS, its role in biological systems may be more subtle with nitration of key tyrosine residues likely to profoundly affect cellular function such as signaling cascades. Further advances are likely to be made in the localization of NTYR residues in peptide fragments using mass spectrometry.

Synthesis, characterization, and immunochemical detection of O6-(carboxymethyl)-2'-deoxyguanosine: a DNA adduct formed by nitrosated glycine derivatives.

O6-(Carboxymethyl)-2'-deoxyguanosine (O6-CMdG) is formed in DNA by nitrosated glycine derivatives and appears to be nonrepairable by O6-alkylguanine transferases. O6-CMdG has been synthesized by an unambiguous route involving the introduction of a methyl glycolate moiety at C6 of a 3',5'-bis-O-(methoxyacetyl)dGuo derivative by displacement of a quinuclidinium ion. Methanolysis of the methoxyacetyl groups and calcium hydroxide-mediated hydrolysis of the methyl ester afforded the calcium salt of O6-CMdG in good yield. A similar route was used to synthesize O6-(carboxymethyl)guanosine (O6-CMGuo), which was used to prepare protein conjugates for immunization. Rabbit antisera were prepared, and a quantitative competitive ELISA was developed which showed 50% inhibition at 2 pmol of O6-CMdG/ well. O6-CMGuo was 30 times less cross-reactive (50% inhibition at 60 pmol/well), and normal nucleosides and carboxymethylated purines did not cross-react to any significant extent. The antiserum was also used to prepare reusable immunoaffinity columns which retained O6-CMdG. The binding of O6-CMdG was so strong that conditions used to elute the adduct (1 M trifluoroacetic acid) resulted in partial hydrolysis (becoming quantitative on heating) of the glycosidic bond to give O6-CMguanine which was detected by HPLC with fluorescence detection. DNA treated with azaserine (5 mmol), N-(N'-acetyl-L-prolyl)-N-nitrosoglycine (5 mmol), and potassium diazoacetate (5 mmol) afforded O6-CMdG at levels of 7.3, 393.9, and 496 mumol of O6-CMdG/mol of dG. The antiserum also recognized O6-CMdG in intact DNA.

Urinary excretion of 3-methyladenine after consumption of fish containing high levels of dimethylamine.

The urinary excretion of the DNA alkylation product, 3-methyladenine (3-MeAde), was measured in human volunteers who were on controlled diets and consumed fresh fish, or frozen-stored fish that contained 50-fold higher levels of dimethylamine (DMA), with or without ingested nitrate. DMA potentially could react with nitrosating agents in the diet or within the body, and produce the potent carcinogen N-nitrosodimethylamine (NDMA), which can then react with DNA to form several adducts including 3-MeAde. Our findings show that there was no increase in urinary levels of 3-MeAde after consumption of fish preserved by frozen storage relative to levels after consumption of fresh fish. Furthermore, consumption of 225 mg sodium nitrate (equal to the nitrate content in a large glass of beet juice) at 1 h prior to consumption of the frozen-stored fish did not increase urinary 3-MeAde levels as would be expected if nitrate enhanced endogenous nitrosation of DMA. In contrast, urinary excretion of 3-MeAde from a volunteer who was a moderate cigarette smoker (11 cigarettes per day) was approximately 3- to 8-fold higher than dietary 3-MeAde intake. These findings indicate that consumption of high levels of DMA in fish does not result in detectable levels of NDMA formation and genetic damage as measured by the urinary biomarker 3-MeAde.