Search

In order to examine the effects of sodium saccharin on in vitro cell transformation, neonatal human diploid fibroblasts were treated with various levels of the compound during specific phases of a synchronized cell cycle, then subcultured for a suitable expression period, and assayed for anchorage-independent growth in soft agar. The dose-relatedness and cell-cycle-dependency of the effects were determined both in cells treated with sodium saccharin and in cells co-treated with saccharin and the direct-acting carcinogen N-methyl-N-nitrosourea (MNU). Cell treatment with saccharin at ng/ml levels for 10 hr following release from a metabolic block (G(1) phase), with or without MNU treatment (70 mug/ml) at 10 hr, was an effective transforming regimen. Continuous saccharin treatment for 12 days (from the end of the synchronized cell cycle until seeding into soft agar) was often as effective as G(1) exposure, but treatment during the metabolic block or during S-phase and mitosis was always less effective. The dose-relatedness of the effects was similar both for saccharin-induced transformation at ng/ml treatment levels, and for co-transformation with MNU. Maximum colony growth (>100/50,000 seeded cells) was observed at a concentration of 50 ng/ml, and a variable, but decreased, response was seen at higher and lower exposures. MNU co-treatment served to maintain the transformation response at higher (mug/ml) saccharin levels at which saccharin alone was not effective. Thus, sodium saccharin demonstrated significant transforming activity over a wide dose range, both when used alone, and in combination with MNU.

Our method to evaluate acetylator phenotype and N-oxidation phenotype with a single caffeine dose and a single urine specimen collection has enabled us to examine acetylation and N-oxidation by two enzyme systems that have a known genetic polymorphism and are important in aromatic amine metabolism. Based on available data, we have hypothesized patients with urinary bladder cancer will have a higher frequency of rapid N-oxidation and slow acetylation phenotypes when compared to controls. On the other hand, patients with colorectal cancer should contain a higher proportion of both rapid N-oxidation phenotype and rapid acetylation phenotype when compared to the control group. In contrast, lung cancer patients, should contain an increased frequency of rapid arylamine N-oxidation phenotypes with the frequency of acetyltransferase phenotype being greater or perhaps the same as in the control group.

The 32P-postlabeling technique was used to investigate the efficiency of phosphorylation reaction by T4 polynucleotide kinase using seven synthetic adducted deoxyguanosine 3'-monophosphates. The adducts included cyclic N1,N2 derivatives and the C8 adduct of 4-aminobiphenyl. The adducted substrates were detected at a subfemtomole sensitivity except for one of the diastereomeric propanoguanine derivatives. In general, the recommended conditions were found to be proper for an efficient phosphorylation of the adducts studied. Sensitivity of the adducts to the 3'-dephosphorylation reaction of nuclease P1 was also tested. All the complex cyclic adducts were resistant towards P1. However, the ethenoguanine and 4-aminobiphenyl adducts were relatively sensitive towards P1. No differences were noted between diastereomers.

In 100 healthy volunteers, we have studied the relationship between the type (air- or flue-cured) and number of cigarettes smoked and different biomarkers relevant to the risk of bladder cancer, including the levels of 4-aminobiphenyl (ABP) hemoglobin adduct (a marker of internal dose), urinary mutagenicity in Salmonella typhimurium TA98, and the N-acetylation phenotype (a marker of susceptibility). ABP is a potent bladder carcinogen that is N-acetylated as an overall detoxification step. Levels of the ABP hemoglobin adduct were higher in smokers of black tobacco (air-cured) than in smokers of blond tobacco (flue-cured), confirming our earlier study. In addition, "slow" acetylators had higher levels of the ABP hemoglobin adduct for the same type and quantity of cigarettes smoked. Urinary mutagenicity was also associated with quantity of cigarettes but not with the acetylation phenotype. Convex dose-response relationships were found between the amount smoked and ABP hemoglobin adduct levels or urinary mutagenicity. In 15 nonsmokers who reported exposure to environmental tobacco smoke, ABP hemoglobin adduct levels, unlike urinary mutagenicity, were found to be an aspecific exposure indicator.

The favored tautomeric and ionic structures were examined for the oxidative DNA damage adduct 8-hydroxy-2'-deoxyguanosine and its RNA analogue 8-hydroxyguanosine by 15N NMR spectroscopy. In addition, 15N chemical shifts and coupling constants from 13 different guanine nucleosides, including a wide variety of C8 substitutions (OH, SH, Br, OCH2C6H5, OCH3, SCH3, and SO2CH3), have been analyzed with respect to their tautomeric structures. A -98.5-Hz proton-nitrogen coupling constant observed for the N7 resonance of 8-hydroxyguanosine in dimethyl sulfoxide was evidence for 8-keto substitution, which is contrary to the structure implied by the generally used nomenclature. The pH dependence of 15N NMR spectra of 8-hydroxyguanosine in aqueous solution showed downfield shifts of the N1 and N7 resonances that were greater than 50 ppm, which indicated the conversion from a neutral 6,8-diketo to a 6-enolate-8-keto (pKa1 = 8.6) and finally to a 6,8-dienolate structure (pKa2 = 11.7). There was no evidence of an 8-enol substituent in the absence of ionization. It is proposed that the syn conformation of these oxidized bases in duplex DNA and RNA can be further stabilized by abnormal hydrogen bonding or mispairing that involves N7-H. The combined data show that 15N NMR is a sensitive probe to examine tautomerism of the guanine ring system. The analysis indicates that the change from a single to a double bond for the C8 substituent, and the accompanying removal of the normal double bond between N7 and C8 on the imidazole ring system, has no detectable effect on the tautomerism at the N1-O6 site of the pyrimidine ring system for both the 8-keto and 8-thio substitutions. In addition, large differences in electronegativity of the C8 substituents do not alter the N1-O6 tautomerism.

In this pilot, case-controlled investigation of 43 colorectal and 41 control male patients, we compared associations of colorectal cancer with the aromatic amine acetyltransferase polymorphism, nutritional and demographic characteristics, medical histories, industrial and occupational histories, and exposures from home environments and personal habits. Persons with the "fast" acetylator trait were at greater risk of colorectal cancer (odds ratio: 2.48; 95% confidence interval: 1.02, 6.03). Results that differed from previous reports were positive associations of colorectal cancer with agricultural and manufacturing industries and with consumption of meats prepared by smoking, curing, and barbecueing. As expected, exercise frequency, cruciferous vegetables, and dietary fiber served as protective factors.

Levels of 4-aminobiphenyl-hemoglobin adducts in smokers of blonde (flue-cured) and black (air-cured) tobacco have been found to be proportional to bladder cancer risk. In addition, risk of bladder cancer due to exposure to occupational carcinogens is elevated in genetically determined slow acetylators. In this study of normal male volunteers, 4-aminobiphenyl-hemoglobin adducts were found to be related to both the quantity and the type of tobacco smoked, as well as to the acetylator phenotype (independently of smoking habits). The demonstration that both the genetically determined slow acetylator phenotype and tobacco smoking are independently associated with levels of the carcinogen 4-aminobiphenyl in adducted hemoglobin suggests a single mechanism to explain the contribution of genetic susceptibility and environmental exposure in bladder carcinogenesis.

The recent development of sensitive methods to detect carcinogen-DNA adducts offers a useful biochemical approach to human risk assessment. However, a major obstacle to developing a human biomonitoring method for carcinogen-DNA adducts has been the problem of obtaining target tissue DNA samples by non-invasive means. This work describes a method for the isolation of nanogram quantities of DNA from urothelial cells exfoliated into urine and the detection of carcinogen-DNA adducts from that DNA by 32P-postlabelling methods. Urine samples were collected on ice from dogs treated with 4-aminobiphenyl (ABP) over a 2-week period and pooled according to an experimental plan that involved analysis of cumulative 48- or 72-h samples. The pooled samples were sieved and then washed repeatedly with a sucrose buffer to dissolve contaminating triple phosphate (MgNH4PO4), calcium oxalate and uric acid crystals. DNA was isolated using an enzyme-solvent extraction method with the DNA being co-precipitated from ethanol with glycogen. The DNA was hydrolysed and postlabelled with 32P under conditions of excess ATP so that nucleotides were labelled quantitatively. Adducts observed on the resulting thin-layer chromatograms were identical to those obtained from DNA modified in vitro with N-hydroxy-4-aminobiphenyl and from dog bladder urothelial DNA isolated from the ABP-dosed animals at termination of the experiment. Furthermore, a dose-related increase in ABP-DNA adduct formation was demonstrated. Thus, it appears that the carcinogen-DNA adduct levels in the exfoliated bladder cells are reflective of the levels in the intact urothelium once steady-state levels have been achieved. To establish the identity of the major ABP-urothelial DNA adduct in chronically-treated dogs, the predominant 32P-postlabelled adduct was eluted from the thin-layer chromatograms and co-injected on an HPLC system with a synthetic [3H]N-(deoxyguanosin-8-yl)-4-aminobiphenyl-3',5'-bisphosphate standard. Dual-label analysis of 3H and 32P indicated that both eluted from the column in the same fraction, which coincided with the UV absorbance peak of the synthetic marker. Preliminary experiments with exfoliated urothelial cells from human urines indicate that these methods should have general utility for monitoring humans exposed to urinary bladder carcinogens and for investigations of the biochemical mechanisms by which such adducts are formed in the urothelium.

Human exposure to carcinogenic aromatic amines can occur from a variety of sources, including cigarette smoke and cooked foods. Initial metabolism occurs primarily in the liver and involves N-hydroxylation by cytochrome P-450IA2 and N-acetylation by acetyltransferase, which serve as competing activation and detoxification reactions, respectively. Both enzymes are polymorphic in humans and individual phenotype can be readily assessed using caffeine ingested in coffee followed by urinary metabolite analysis. N-hydroxy arylamines can then be transported through the blood where they can form protein adducts with hemoglobin or undergo renal filtration into the urinary bladder lumen where, at acidic pH, they can react covalently with urothelial DNA. Carcinogen-DNA adduct levels in human bladder are smoking-related and the C8-deoxyguanosine derivative of 4-amino-biphenyl (ABP) is a major product. Alternatively, N-hydroxy arylamines can be conjugated in the liver by glucuronyl transferases which provides a mechanism for biliary transport to the colon lumen where beta-glucuronidases can regenerate the aglycone. In the colon mucosa, acetyltransferases can further activate the N-hydroxy metabolite by O-acetylation; and persons with the rapid acetylation phenotype are known to be at higher risk for colorectal cancer. Detection of specific arylamine-DNA adducts in human colon should provide direct evidence for the role of these carcinogens in the etiology of this disease.

The dog is an animal model for assessing aromatic amine-induced bladder cancer. For this reason, metabolism and disposition of benzidine in dog was assessed. Dogs were administered a 1 mg/kg i.v. dose of [3H]benzidine (16.4 mCi/mmol). The plasma t1/2 of the radiolabeled material (benzidine plus metabolites) was significantly longer (approximately 3 h) than authentic benzidine (less than 30 min). During the 5 h experiment, the majority of radiolabel was associated with bile, urine and carcass. Bladder transitional epithelium exhibited a consistently higher concentration of bound radioactivity than bladder muscle. A significant amount of binding was observed in DNA from liver, kidney and bladder. DNA from bladder transitional epithelium exhibited the highest concentration of radioactivity. Approximately 30% of the radioactivity recovered following HPLC of urine or bile was identified as unmetabolized benzidine. 3-Hydroxybenzidine was a major metabolite identified in bile (8%) but not urine. Urine samples treated with acid, base or sulfatase yielded 3-hydroxybenzidine (6%) as a major hydrolysis product. Similar treatment of bile samples did not result in increased amounts of 3-hydroxybenzidine. Neither N-acetylated nor N-methylated metabolites of benzidine were observed in urine or bile. Thus, considerable metabolism of benzidine occurs in dogs by pathways that are yet to be determined.

Benzidine, a potent hepatocarcinogen in rodents, is readily metabolised to acetylated derivatives. In this study, the covalent binding of [3H-acetyl]N-acetylbenzidine and [3H-acetyl]N,N'-diacetylbenzidine to liver DNA in rats and hamsters was investigated. Binding to liver DNA of rats at 1 or 7 days after i.p. injection of N-acetylbenzidine was 2-fold higher than that observed in the liver DNA of hamsters which had been similarly treated. Analysis of enzymically hydrolysed DNA from both species indicated the presence of a single adduct which co-eluted with N-(deoxyguanosin-8-yl)-N'-acetylbenzidine. In vitro treatment of rat or hamster liver slices with N-acetylbenzidine also resulted in covalent binding to hepatic DNA and the identical DNA adduct was detected at levels comparable to that observed in vivo. When N,N'-diacetylbenzidine was injected i.p. into rats, a comparatively low level of binding to liver DNA was observed. Following enzymic hydrolysis, the major DNA adduct detected by h.p.l.c. analysis was again N-(deoxyguanosin-8-yl)-N'-acetylbenzidine accompanied by a small amount of N-(deoxyguanosin-8-yl)-N,N'-diacetylbenzidine. In vitro incubation of N,N'-diacetylbenzidine with rat liver slices resulted in DNA binding levels similar to that observed with N-acetylbenzidine. In contrast to what was found in vivo, N-(deoxyguanosin-8-yl)-N,N'-diacetylbenzidine was the major adduct detected in DNA from rat liver slices. These data suggest that both N-hydroxy-N'-acetylbenzidine and N-hydroxy-N,N'-diacetylbenzidine are proximate carcinogenic species of benzidine, with N-hydroxy-N'-acetylbenzidine the more important. The low level of N-(deoxyguanosin-8-yl)N,N'-diacetylbenzidine observed in vivo may be due to its rapid repair. Alternatively, N-sulphonyloxy-N,N'-diacetylbenzidine, which would produce this adduct on reaction with DNA, may be efficiently detoxified in vivo.

Hepatic cytosols catalyzed a 3'-phosphoadenosine 5'-phosphosulfate (PAPS)-dependent O-sulfonation of N-hydroxy-N-methyl-4-aminoazobenzene (N-HO-MAB) and of several other N-hydroxy arylamines. The presumed product from N-HO-MAB, N-methyl-4-aminoazobenzene-N-sulfate, reacted with added guanosine to yield N-(guanosin-8-yl)-N-methyl-4-aminoazobenzene, with methionine to form a sulfonium derivative that decomposed to yield 3-methylmercapto-N-methyl-4-aminoazobenzene, and with ribosomal RNA to give a bound derivative. N-Methyl-4-aminoazobenzene was converted to N-(guanosin-8-yl)-N-methyl-4-aminoazobenzene in concerted N-oxidation and O-sulfonation reactions conducted aerobically with a fortified 10,000 X g rat liver supernatant. In the absence of an added nucleophile, metabolically formed N-methyl-4-aminoazobenzene-N-sulfate (or the nitrenium ion from this unstable ester) was reduced by N-HO-MAB to form N-methyl-4-aminoazobenzene; the N-HO-MAB was oxidized, probably through a nitrone intermediate, to yield products that included N-hydroxy-4-aminoazobenzene and formaldehyde. An analogous reaction was noted between N-benzoyloxy-N-methyl-4-aminoazobenzene and N-HO-MAB in the absence of cytosol and PAPS. Hepatic N-HO-MAB sulfotransferase activities were in the order: male rat greater than female rat, male rabbit, male guinea pig, male mouse greater than male hamster. Male rat kidney and small intestine cytosols had low activities; the other tissues studied had little or no activity. Hepatic sulfotransferase activities for N-HO-MAB and N-hydroxy-N-acetyl-2-aminofluorene displayed different pH optima and inhibitor and activator responses. The rates of PAPS-dependent rat liver cytosol-catalyzed esterification of N-hydroxy-N-ethyl-4-aminoazobenzene, N-hydroxy-4-aminoazobenzene, and N-hydroxy-1- and 2-naphthylamine were 20 to 50% of that for N-HO-MAB. Activities for trans-N-hydroxy-4-aminostilbene, N-hydroxy-2-aminofluorene, N-hydroxyaniline, and N-hydroxy-N-methyl-N-benzylamine were not detected. No microsomal reduced nicotinamide adenine dinucleotide-dependent reduction or reduced nicotinamide adenine dinucleotide phosphate-dependent oxidation or cytosolic transferase reactions for N-HO-MAB, except the above-described PAPS-dependent reaction, were detected in rat liver.

The carcinogen, 2-naphthylamine (2-NA), induces tumor formation in the urinary bladder but not the liver of several species, including humans and dogs. Since its proximate carcinogenic metabolite, N-hydroxy-2-NA, was known to react directly with DNA in vitro to give specific carcinogen-base adducts, we investigated the in vivo formation and persistence of (2-NA)-DNA adducts in the bladder and liver and attempted to determine whether or not these lesions correlated with tissue susceptibility. Male beagle dogs were administered [3H]2-NA and sacrificed after 2 or 7 days. The DNA was isolated from the liver and urothelium and hydrolyzed enzymatically. The (2-NA)-deoxyribonucleoside adducts, which were quantitated by high pressure liquid chromatographic analysis, were the same as those found in vitro, namely 1-(deoxyguanosin-N2-yl)-2-NA, 1-(deoxyadenosin-N6-yl)-2-NA, and an imidazole ring-opened derivative of N-(deoxyguanosin-8-yl)-2-NA. The major difference detected between target and non-target tissues was in the total level of binding to DNA, which was 4-fold higher in the urothelium at 2 days and 8-fold higher at 7 days after 2-NA dosing. Analysis of specific adducts suggested that this difference may be due to the relative persistence of the C-8-guanine adduct in the urothelium as compared to the liver. Similar experiments with the non-carcinogen, 1-naphthylamine, failed to reveal binding in urothelial DNA and indicated a 20-fold lower binding level in hepatic DNA. Evidence for binding of 2-NA to glycogen is also presented and problems associated with measuring total radioactivity in glycogen-contaminated DNA fractions are discussed. The data obtained in this study, through from a necessarily limited number of animals, are consistent with the hypothesis that the formation and persistence of DNA-carcinogen adducts may be important in the initiation of the neoplastic process.

N-Methyl-4-aminoazobenzene (MAB) is believed to be metabolized in the liver to an electrophilic N-sulfonyloxy ester which binds covalently to cellular macromolecules, resulting in the induction of hepatic neoplasia. Previous in vivo studies in the rat detected only two hepatic MAB-DNA adducts, 3-(deoxyguanosin-N2-yl)-MAB(N2-dG) and N-(deoxyguanosin-8-yl)-MAB(C8-dG), which respectively accounted for 25% and 70% of the total MAB bound to DNA at 8 h after a single dose of the carcinogen. Subsequently, the C8-dG adduct was shown to be rapidly lost from the DNA while the N2-dG adduct was a persistent lesion. Since a single dose of MAB is not sufficient for complete carcinogenic activity, we sought to identify the MAB-DNA adducts present in rat liver after multiple oral doses of [3H]MAB. The MAB was administered by intubation at a level of 0.2 mmol/kg for 1, 3 or 4 doses and animals were sacrificed at 8 h after the last dose. Hepatic DNA was isolated by extraction and hydroxylapatite chromatography and was enzymatically hydrolyzed to MAB-mononucleoside adducts, which were quantitated by high pressure liquid chromatography (HPLC). After 3 doses, N2-dG, C8-dG, and an unknown adduct were detected. By 4 doses, these accounted for 51%, 25% and 23% of the total adducts. This data is consistent with rapid removal of the C8-dG derivative and the relative persistence of the N2-dG and the unknown adduct. The latter was shown to exhibit chromatographic and pH-dependent solvent partitioning properties that were identical to a product also present in DNA treated with the synthetic ultimate carcinogen, N-benzoyloxy-MAB. Analysis of this adduct by field desorption mass spectrometry (M+ = 460) and, after perdeuteromethylation, by electron impact mass spectrometry (M+ = 528; M-N(CH3)(CD3) = 481) indicated the structure to be a deoxyadenosin-N6-yl derivative substituted through an aromatic ring of MAB. Further analysis by 270 MHz 1H-NMR spectroscopy allowed complete assignment of the MAB and adenyl resonances and was uniquely consistent with a 3-(deoxyadenosin-N6-yl)-MAB structure. Since this persistent adduct is potentially mutagenic due to possible tautomeric equilibria between the N6-amino and N6-imino structures, it may represent an initiating lesion in MAB hepatocarcinogenesis.

Estimation of urinary levels of N-hydroxyarylamines, which are believed to be ultimate carcinogens for the bladder epithelium, has been complicated by the rapid resorption of these metabolites across the bladder wall and by their formation in the urine upon hydrolysis of N-hydroxyarylamine N-glucuronide conjugates. Therefore, a three-compartment pharmacokinetic model was developed with an analog-digital hybrid computer to separate absorption, metabolism, and distribution in the whole animal from events in the bladder lumen (deposition, hydrolysis, and resorption). The total bladder exposure to the carcinogen, N-hydroxy-2-naphthylamine, was then simulated from experimental data; and urine pH and voiding interval, both of which differ widely among species, were tested as variables. For humans and dogs, urine pH was found to be considerably more acidic than monkeys or rats; while voiding intervals in rats were considerably shorter than in monkeys, dogs, or humans. The relative species susceptibilities, indicated by the model as total bladder exposure to the N-hydroxyarylamine metabolite under conditions of normal urine pH and frequency of micturition for each species, were in the order: human greater than or equal to dog greater than monkey greater than rat. This is consistent with the reported carcinogenic potency of 2-naphthylamine in these species and suggests that urine pH and voiding interval are important determinants in aromatic amine-induced bladder cancer.

Purine ring-opened 7-methylguanine, prepared in vitro by alkaline treatment of 7-methylguanosine or of methylated calf thymus DNA, was extensively characterized by chromatographic and spectral techniques as N5-methyl-N5-formyl-2,5,6-triamino-4-hydroxypyrimidine. This modified base chromatographed as an early-eluting peak on an ion-exchange column but separated into two interconvertible components after reversed-phase or porous-resin h.p.l.c. The two components were analyzed by thermal desorption mass spectrometry and 500 MHz 1H-n.m.r. spectroscopy. Their mass spectra were identical (M+ at m/z 183) and their n.m.r. spectra each exhibited the same two sets of resonances whose relative intensities were solvent-dependent. Analysis by h.p.l.c. showed interconversion of the two components and kinetic studies demonstrated that this reaction was a reversible first-order process. At equilibrium, k1 = k2 = 0.334 h-1 and delta G = 22.9 kcal/mol. These data indicated that the ring-opened 7-methylguanine exists as cis/trans isomers with restricted rotation about the amide bond. Treatment of rats with an intraurethral initiating dose of the carcinogen N-methylnitrosourea resulted in a high level of bladder epithelial DNA modification with 7-methylguanine, O6-methylguanine, and methyl phosphotriesters as major adducts at 2 h after instillation. Purine ring-opened 7-methylguanine, chromatographically identical to the in vitro products, was initially a minor adduct. However, it was the only persistent modification in the bladder epithelial DNA and eventually accounted for 72% of the total carcinogen binding after 21 days. A tumor-promoting regimen, involving dietary sodium saccharin, did not alter the repair or persistence of any of the methylated adducts. These data demonstrate that purine ring-opened 7-methylguanine, previously reported to exist in liver DNA after N,N-dimethylnitrosamine or 1,2-dimethylhydrazine treatment, is present in a carcinogen-target tissue and is considerably more persistent than O6-methylguanine or other DNA methylation products. The possible role of this adduct as a promutagenic lesion initiating urinary bladder carcinogenesis is discussed.

The conversion of urinary N-hydroxy arylamines to carcinogenic electrophiles under mildly acidic conditions in the bladder lumen has been proposed as an essential step in arylamine-induced urinary bladder carcinogenesis. To test the hypothesis that extracellular generation of an ultimate carcinogenic species can initiate a neoplastic event normal human fibroblasts were exposed to N-hydroxy-1- and 2-naphthylamine (N-HO-1-NA and N-HO-2-NA) at pH 5 and pH 7. With both compounds, anchorage independent growth of transformed cells in soft agar were enhanced 3- to 7-fold in the pH 5 incubations. Injection of the N-HO-1-NA and N-HO-2-NA-transformed cells into nude mice resulted in tumors in 1/8 and 2/7 animals, respectively. In a control experiment, no differences in transformation of these cells by aflatoxin B1 were observed between pH 5 and pH 7 exposures. Thus, the results are consistent with the hypothesis that ultimate carcinogenic electrophiles, generated extracellularly, can enter an intact cell and induce neoplasia. Alternatively, the possibility of a local intracellular acidic environment near the cell surface and its role in the generation of a reactive electrophile leading to urinary bladder carcinogenesis is discussed.

Benzidine, a human urinary bladder carcinogen, induces hepatic tumors in mice and rats. In this study, [3H]benzidine was administered in drinking water to mice for 1 week, and the covalent binding of the carcinogen to hepatic DNA was then determined. A single carcinogen:DNA adduct was detected which decreased in concentration by approximately 50% at 1 day after treatment and then remained at a nearly constant level for at least 7 days. Injection of radiolabeled benzidine or N-acetyl-benzidine into rats also resulted in a single carcinogen:DNA adduct that was chromatographically identical to that obtained in mouse liver. While administration of benzidine and N-acetylbenzidine resulted in high levels of the adduct in rat hepatic DNA, injection of N,N'-[ring-14C]diacetylbenzidine did not give detectable binding (less than 0.3 residue/mg DNA). The same carcinogen:DNA adduct found in rat and mouse liver was prepared synthetically by: (a) hydrolysis of calf thymus DNA reacted with N-hydroxy-N'-acetylbenzidine at pH 5; and (b) reaction of N-acetoxy-N,N'-diacetylbenzidine with deoxyguanosine and subsequent selective deacetylation of the product with methanolic ammonia. The in vitro and in vivo products were found to have identical high-pressure liquid chromatography retention times and to exhibit similar pH-dependent solvent partitioning characteristics. Mass and nuclear magnetic resonance spectral data on the synthetic products established the structure of the hepatic adduct as N-(deoxyguanosin-8-yl)-N'-acetylbenzidine. The structural isomer, N-(deoxyguanosin-8-yl)-N-acetylbenzidine, was synthesized by treatment of N-(deoxyguanosin-8-yl)-N,N'-diacetylbenzidine (the intermediate in b) with carboxylesterase and was shown to be chromatographically distinct from the in vivo adduct. Similarly, the nonacetylated derivative, N-(deoxyguanosin-8-yl) benzidine, was synthesized by carboxylesterase treatment of N-(deoxyguanosin-8-yl)-N'-acetylbenzidine and was shown not to occur in rat and mouse liver DNA. These data indicate that the metabolic activation of benzidine to an ultimate carcinogen in rats and mice does not involve N-hydroxybenzidine or sulfotransferase-catalyzed activation of N-hydroxy-N,N'-diacetylbenzidine. The remaining pathways for metabolic conversion of benzidine to an ultimate carcinogenic species are discussed in relation to liver and urinary bladder carcinogenesis.

The metabolic activation of several carcinogenic N-hydroxy (N-OH)-arylamines by cytosolic S-acetyl coenzyme A (AcCoA)-dependent enzymes was examined in tissues and species susceptible to arylamine carcinogenesis. Comparisons of the AcCoA-dependent activity were also made with known cytosolic arylhydroxamic acid-dependent acyltransferases and with the ability of different acyl donors to mediate the binding of N-OH-arylamines to DNA. With rat hepatic cytosol, AcCoA-dependent DNA binding was demonstrated for several [3H]N-OH-arylamines, in the order: N-OH-3,2'-dimethyl-4-aminobiphenyl (N-OH-DMABP), N-OH-2-aminofluorene (N-OH-AF) greater than N-OH-4-aminobiphenyl greater than N-OH-N'-acetylbenzidine greater than N-OH-2-naphthylamine; N-OH-N-methyl-4-amino-azobenzene was not a substrate. No activity was detected in dog hepatic or bladder cytosol with any of the N-OH-arylamines tested. Using either N-OH-DMABP or N-OH-AF and rat hepatic cytosol, activation to DNA-bound products was also detected with acetoacetyl- and propionyl-CoA but not with folinic acid or six other acyl CoA's. However, p-nitrophenyl acetate which is known to generate acetyl-enzyme intermediates effectively replaced AcCoA. Subcellular fractionation of rat liver showed that the AcCoA-dependent DNA-binding of N-OH-DMABP with cytosol was 5 times greater than that obtained with the microsomal or mitochondrial/nuclear fractions. Furthermore, the cytosolic activity was insensitive to inhibition by the esterase/deacetylase inhibitor, paraoxon; while the activity of the other subcellular fractions was completely inhibited (greater than 95%). AcCoA-dependent activation of N-OH-DMABP was also detected with rat tissue cytosols from intestine, mammary gland and kidney, which like the liver, are targets for arylamine-induced tumorigenesis. Using N-OH-DMABP, AcCoA-dependent DNA-binding activity was also detected in the hepatic cytosols from several species in the order: rabbit greater than hamster greater than rat, human greater than guinea pig greater than mouse. In contrast, the arylhydroxamic acid, N-OH-N-acetyl-DMABP, was not activated to a DNA-binding metabolite by the hepatic cytosolic N,O-acyltransferase of any of these species, thus suggesting that the AcCoA-mediated binding of N-OH-DMABP results from the direct formation of N-acetoxy-DMABP. With N-OH-AF as the substrate, the AcCoA-dependent activation was in the order: rabbit greater than guinea pig, hamster greater than mouse greater than human, rat.(ABSTRACT TRUNCATED AT 400 WORDS)

Since the susceptibility of specific tissues to tumor formation has been correlated with the persistence of DNA-carcinogen adducts, the identity and persistence of DNA adducts formed from the hepatocarcinogen N-methyl-4-aminoazobenzene (MAB) has been determined. The synthetic ultimate carcinogen N-benzoyloxy-N-methyl-4-aminoazobenzene (N-BxO-MAB) was reacted in vitro with either calf thymus or rat liver DNA to yield approx. 1 bound residue per 1000 nucleotides. After enzymatic hydrolysis of the DNA and high pressure liquid chromatographic analysis, at least six MAB adducts were detected. Two of the products cochromatographed with MAB-DNA adducts formed in rat liver in vivo following oral administration of the precarcinogen MAB. These two adducts were identified by mass, UV and nuclear magnetic resonance (NMR) spectroscopy as N-(deoxyguanosin-8-yl)- and 3-(deoxyguanosin-N2-yl)-MAB. The former adduct was initially the predominant product in vivo, but it could not be detected 7 days following treatment. The latter adduct remained at a constant level for 14 days and therefore appears to be a persistent lesion.

N-Benzoyloxy-N-methyl-4-aminoazobenzene (N-BzO-MAB) is believed to be an analogue of the ultimate carcinogenic form of N,N-dimethyl-4-aminoazobenzene (DAB). The reaction of N-BzO-MAB with glutathione in vitro yielded one major and two minor aminoazo dye-glutathione adducts. After purification by ion exchange chromatography and high pressure liquid chromatography, analysis of chemical properties, and the measurement of ultraviolet, visible, proton magnetic resonance, and mass spectra, the major and one minor adduct were identified as 3-(glutathion-S-yl)-N-methyl-4-aminoazobenzene (3-GS-MAB) and 2'-(glutathion-S-yl)-N-methyl-4-aminoazobenzene (2'-GS-MAB) respectively. The other minor adduct was tentatively identified as 4'-(glutathion-S-yl)-N-methyl-4-aminoazobenzene (4'-GS-MAB). Fractionation and analyses of biliary metabolites from rats given DAB revealed the presence of two aminoazo dye-glutathione adducts. One of these was identical to 3-GS-MAB in its chromatographic and chemical properties and its visible and ultraviolet spectra. The other adduct was partially characterized and judged to be a 4-aminoazobenzene-glutathione adduct. The role of glutathione in the detoxification of carcinogenic aminoazo dyes is discussed.

Hepatic arylamine acetyltransferase phenotype has been suggested to be an important risk factor for urinary bladder carcinogenesis in individuals with known exposure to aromatic amines. This study was performed to evaluate the relative distribution of fast- and slow-acetylator phenotypes both in a population of men, 45 to 75 years of age, with a history of colorectal cancer and in a matched control group. Acetyltransferase activity was determined by administration of sulfamethazine and by subsequent analysis of blood and urine samples for N-acetylsulfamethazine and sulfamethazine using high-pressure liquid chromatography. The control group was composed of 28 slow-, two intermediate-, and 11 fast-acetylator individuals, while the group of patients with a history of cancer consisted of 20 slow-, three intermediate-, and 20 fast-acetylator phenotypes. This higher relative proportion of fast acetylators in the patients with a cancer history was highly significant and is consistent with the hypothesis that aromatic amines could play a role in the etiology of human colorectal cancer.

The probable ultimate urinary bladder carcinogen, N-hydroxy-2-naphthylamine (N-HO-2-NA), reacted with nucleic acids and proteins under mildly acidic conditions (pH 5) to form covalently bound derivatives. The extent of reaction was in the order: polyguanylic acid > DNA > or = protein > rRNA > tRNA > polyadenylic acid, polyuridylic acid > polycytidylic acid. At pH 7, appreciable reaction occurred only with protein. Enzymatic hydrolyses of the DNA, which contained 1.5 naphthyl residues/1000 nucleotides, yielded 3 nucleoside-arylamine adducts. From chemical, u.v., n.m.r., and mass spectrometric analyses, the adducts were identified as 1-(deoxyguanosin-N2-yl)-2-naphthylamine, 1-(deoxyadenosin-N6-yl)-2-naphthylamine, and a purine ring-opened derivative of N-(deoxyguanosin-8-yl)-2-naphthylamine, tentatively identified as 1-[5-(2,6-diamino-4-oxopyrimidinyl-N6-deoxyriboside)]-3-(2-naphthyl)urea. The properties of these adducts and their possible role in the initiation of carcinogenesis are discussed.

The duplex genome of Escherichia coli virus M13mp10 was modified at a unique site to contain N-(deoxyguanosin-8-yl)-4-aminobiphenyl (dG8-ABP), the major carcinogen-DNA adduct of the human bladder carcinogen 4-aminobiphenyl. A tetradeoxynucleotide containing a single dG8-ABP residue was synthesized by reacting 5'-d(TpGpCpA)-3' with N-acetoxy-N-(trifluoracetyl)-4-aminobiphenyl, followed by high-performance liquid chromatography purification of the principal reaction product 5'-d(TpG8-ABPpCpA)-3' (yield 15-30%). Characterization by fast atom bombardment mass spectrometry confirmed the structure as an intact 4-aminobiphenyl-modified tetranucleotide, while 1H nuclear magnetic resonance spectroscopy established the site of substitution and the existence of ring stacking between the carcinogen residue and DNA bases. Both 5'-d(TpG8-ABPpCpA)-3' and 5'-d(TpGpCpA)-3' were 5'-phosphorylated by use of bacteriophage T4 polynucleotide kinase and were incorporated into a four-base gap uniquely positioned in the center of the recognition site for the restriction endonuclease PstI, in an otherwise duplex genome of M13mp10. In the case of the adducted tetranucleotide, dG8-ABP was located in the minus strand at genome position 6270. Experiments in which the tetranucleotides were 5' end labeled with [32P]phosphate revealed the following: the adducted oligomer, when incubated in a 1000-fold molar excess in the presence of T4 DNA ligase and ATP, was found to be incorporated into the gapped DNA molecules with an efficiency of approximately 30%, as compared to the unadducted d(pTpGpCpA), which was incorporated with 60% ligation efficiency; radioactivity from the 5' end of each tetranucleotide was physically mapped to a restriction fragment that contained the PstI site and represented 0.2% of the genome; the presence of the lesion within the PstI recognition site inhibited the ability of PstI to cleave the genome at this site; in genomes in which ligation occurred, T4 DNA ligase was capable of covalently joining both modified and unmodified tetranucleotides to the gapped structures on both the 5' and the 3' ends with at least 90% efficiency. Evidence also is presented showing that the dG8-ABP-modified tetranucleotide was stable to the conditions of the recombinant DNA techniques used to insert it into the viral genome.(ABSTRACT TRUNCATED AT 400 WORDS)

A genetic polymorphism in the enzymatic N-acetylation of sulfamethazine and other drugs in humans is well known and has been related to differential susceptibility to drug toxicities. Carcinogenic aromatic amines such as 2-aminofluorene also undergo N-acetylation, and phenotypic slow acetylator individuals have been suggested to be at increased risk to arylamine-induced urinary bladder cancer. However, acetyltransferases have also been shown to catalyze a final metabolic activation step in the conversion of both hydroxamic acid (e.g. N-hydroxy-N-acetyl-2-aminofluorene N,O-acyltransferase) and N-hydroxy-arylamine (e.g. N-hydroxy-2-aminofluorene O-acetyltransferase) metabolites to DNA-bound adducts. In this regard, rapid acetylators have recently been reported to be at higher risk for colorectal cancer. In this study, we examined the enzymatic activity of 35 human liver cytosol samples (obtained surgically from organ donors) for sulfamethazine and 2-aminofluorene N-acetyltransferase activities, N-hydroxy-N-acetyl-2-aminofluorene N,O-acyltransferase activity, and the acetyl coenzyme A (CoA)-dependent O-acetylation of N-hydroxy-2-aminofluorene to form DNA-bound products. The results with sulfamethazine indicated that about two-thirds of the human liver samples were of the slow acetylator phenotype; the same individuals also exhibited levels of 2-aminofluorene N-acetylation that were consistent with their respective sulfamethazine-N-acetylation activity. N-Hydroxy-N-acetyl-2-aminofluorene N,O-acyltransferase activity was not detected. However, the acetyl CoA-dependent activation of N-hydroxy-2-aminofluorene was observed for nearly all of the samples and was consistently higher in the fast acetylator group. These data support the hypothesis that phenotypic rapid acetylator individuals are likely to be at higher risk to aromatic amine-induced cancers in those tissues containing appreciable levels of N-hydroxy arylamine O-acetyltransferase.

To assess target-tissue exposure to the human urinary bladder carcinogen 4-aminobiphenyl (ABP), we have developed a sensitive immunochemical method for measuring the major arylamine-DNA adduct formed, N-(guan-8-yl)-ABP (Gua-C8-ABP). High-affinity polyclonal antisera from rabbits immunized with N-(guanosin-8-yl)-ABP coupled to keyhole limpet hemocyanin were characterized and shown to have high specificity for antigenic determinants on the purine and biphenyl rings of Gua-C8-ABP and minimal cross-reactivity with ABP, deoxyguanosine, or hydrolyzed DNA. Assay standards containing ABP-modified DNA were prepared by reacting [3H]N-hydroxy-ABP with calf thymus DNA. DNA samples were hydrolyzed with trifluoroacetic acid and dried under vacuum, and the residues were dissolved in dimethyl sulfoxide under argon. Using a streptavidin-biotin amplified enzyme-linked immunosorbent assay, DNA hydrolysates competed at 25 micrograms DNA/microtiter well for a limiting amount of anti-keyhole limpet hemocyanin-(Gua-C8-ABP) in the presence of excess solid-phase bovine serum albumin-(Gua-C8-ABP) coating antigen. The limit of sensitivity for this assay using 25 micrograms DNA was 2 adducts/10(8) nucleotides. Gua-C8-ABP adducts in liver and bladder epithelial DNAs were readily quantified after p.o. administration of 5 mg/kg ABP to dogs. This methodology is capable of detecting adducts at levels of biological significance and should be applicable to human target-tissue dosimetry.

As a consequence of human exposure to carcinogenic aromatic amines, biochemical approaches to risk assessment have emphasized metabolic determinants of individual susceptibility and quantification of arylamine-macromolecular adducts. A known genetic polymorphism in humans, hepatic arylamine acetyltransferase activity, has been associated with differences in individual susceptibility to urinary bladder (slow acetylators) and colorectal (rapid acetylators) cancers. Similarly, the high specificity of an inducible human cytochrome P450 towards the N-oxidation of 4-aminobiphenyl and other aromatic amines is consistent with metabolic differences that can be used to predict relative human risk. Exposure to aromatic amines has also been documented, primarily by quantification of adducts with protein or DNA. Using 32P-postlabelling methods and a competitive avidin/biotin-amplified enzyme-linked immunoassay, we have estimated 4-aminobiphenyl-DNA adduct levels in surgical samples of human peripheral lung and urinary bladder epithelium and report values ranging from 2 to 97 adducts per 10(8) nucleotides.

Serum albumin was isolated from rats at 27 h after administration of the carcinogen [2,2'-3H]-4-aminobiphenyl. Pronase digestion of the purified albumin yielded a mixture of radiolabeled materials which was resolved into 5 major components by reverse-phase liquid chromatography. From detailed UV, 1H-NMR, and mass spectral analyses, four of these were determined to be 4-aminobiphenyl, 4'-hydroxy-4-acetylaminobiphenyl, and two other metabolites, all of which are presumed to be non-covalently associated with the serum albumin. The fifth component, however, resulted from covalent bond formation and was identified as a tetrapeptide containing 3-tryptophanyl-4-acetylaminobiphenyl, the amino acid sequence of which was H2N-ala-trp-ala-val. Since rat serum albumin contains only a single tryptophan residue in a hydrophobic drug binding site, its high selectivity for carcinogen binding suggests a unique role for this protein in the detoxification and/or transport of ultimate carcinogenic metabolites.

N-Hydroxy-N,N'-diacetylbenzidine (N-HO-DABZ) has been shown to be an in vitro metabolite of benzidine in several rodent species and may represent the proximate form of the carcinogen. Like other arylhydroxamic acids, N-HO-DABZ may be converted to an ultimate carcinogenic electrophile by metabolic O-sulfonation in hepatic cytosol. To investigate this possibility, liver cytosols from rats, mice, and hamsters were assayed for their ability to catalyze the 3'-phosphoadenosine 5'-phosphosulfate-dependent metabolism of N-HO-DABZ and the formation of an adduct with methionine. For comparative purposes, sulfotransferase activity for N-hydroxy-2-acetylaminofluorene (N-HO-AF) was also measured. In the rat, N-HO-DABZ and N-HO-AAF were metabolized at rates of 2.5 and 4.3 nmol of arylhydroxamic acid lost per in per mg of protein, respectively. In the mouse, these rates were 0.5 nmol for N-HO-DABZ and less than 0.05 nmol for N-HO-AAF. Sulfotransferase activity for these substrates in hamster liver cytosol could not be detected (less than 0.05 nmol/min/mg). The inclusion of methionine in sulfotransferase incubation mixtures and subsequent heating resulted in the formation of methylmercapto arylamides from both N-HO-DABZ and N-HO-AAF. From 20 to 40% of the N-HO-DABZ metabolized was trapped and recovered as an adduct, while 80 to 100% of the N-HO-AAF metabolized was similarly obtained. A methylmercapto-N,N'-diacetylbenzidine derivative was also obtained by reaction of N-acetoxy-N,N'-diacetylbenzidine with methionine. Its identity to the adduct formed in the sulfotransferase incubation mixture was established by high-pressure liquid chromatography, ultraviolet light, and mass spectroscopic analyses. By comparing the 13C nuclear magnetic resonance spectra of the synthetic methylmercapto derivative with several model compounds and using chemical shift additivity relationships, the adduct was identified as 3-methylmercapto-N,N'-diacetylbenzidine. Since the yield of the product from N-acetoxy-N,N'-diacetylbenzidine and methionine did not vary appreciably with pH (4 to 8), a reaction mechanism involving an electrophilic carbocation at position 3 is proposed. These studies demonstrate that N-HO-DABZ can be esterified to an electrophilic reactant by hepatic sulfotransferases in the rat and the mouse and suggest the involvement of this metabolite in the hepatocarcinogenicity of benzidine.

It has previously been shown that 6-nitrochrysene can be activated to electrophilic species capable of reacting with DNA through metabolic pathways that form N-hydroxy-6-aminochrysene or trans-1,2-dihydroxy-1,2-dihydro-6-aminochrysene as critical intermediates. Since the lung is a known target tissue for the carcinogenic action of polycyclic nitroaromatic hydrocarbons, we investigated the metabolism and DNA binding of [3H]6-nitrochrysene in 11 specimens of human bronchus. Analysis of medium from [3H]6-nitrochrysene-treated explants indicated the presence of trans-9,10-dihydroxy-9,10-dihydro-6-nitrochrysene (0.04-330 pmol/mg epithelial DNA), trans-1,2-dihydroxy-1,2-dihydro-6-nitrochrysene (12-1700 pmol/mg epithelial DNA), 6-aminochrysene (1.6-2200 pmol/mg epithelial DNA), and trans-1,2-dihydroxy-1,2-dihydro-6-aminochyrsene (3.6-610 pmol/mg epithelial DNA). Both the levels and the relative proportions of these metabolites varied widely in explants from different individuals. The amount of DNA recovered and the level of DNA modification were sufficient for adduct analysis in eight of the 11 cases for which metabolite data were obtained. Five additional bronchial specimens for which metabolite data were not obtained were also analyzed for carcinogen-DNA adducts. The levels of binding varied from 0.06 to 30.5 pmol [3H]6-nitrochrysene bound/mg DNA (two adducts per 10(8) nucleotides-10 adducts per 10(6) nucleotides). HPLC analyses of enzymatic hydrolysates of the explant DNA indicated that 11 of 13 cases contained adducts with retention times identical to those of adducts derived from trans-1,2-dihydroxy-1,2-dihydro-6-aminochrysene or N-hydroxy-6-aminochrysene. The adduct derived from trans-1,2-dihydroxy-1,2-dihydro-6-aminochrysene was the major adduct detected in eight of 13 cases. The reasons for the variation in metabolism and adduct formation observed in [3H]6-nitrochrysene-treated explants of bronchus from different donors are not known but may reflect differences in the activities of enzymes responsible for the metabolism of this compound. The influence of induction of drug metabolizing enzymes on the activation pathway of 6-nitrochrysene in an intact cell system was tested using rat hepatocytes. 6-Nitrochrysene was incubated with freshly isolated hepatocytes from rats that were either untreated or pretreated with phenobarbital, 3-methylcholanthrene or Aroclor 1254. Although the levels of adducts were similar in all cases, the pattern of DNA adducts formed in these hepatocytes was dependent on the nature of the pretreatment of the rats. As previously reported, hepatocytes from untreated rats contained adducts derived from N-hydroxy-6-aminochrysene.(ABSTRACT TRUNCATED AT 400 WORDS)

The arylamine carcinogen 3,2'-dimethyl-4-aminobiphenyl (DMABP) has been proposed to be metabolically activated to DNA-binding derivatives through the formation of an N-hydroxy intermediate. In this study, the subsequent activation of N-hydroxy-DMABP through acid catalysis or enzymatic esterification was examined. [Ring-3H]N-hydroxy-DMABP was reacted with calf thymus DNA at pH 4.6 for 15 min to yield 370 arylamine residues per 10(6) nucleotides, while at pH 7.4 the binding was only two residues per 10(6) nucleotides. The DNA modified under acidic conditions was enzymatically hydrolyzed and analyzed by h.p.l.c. which indicated the presence of three major adducts. The products were identified by spectral and chemical properties as N-(deoxyguanosin-8-yl)-DMABP (60-70%), 5-(deoxyguanosin-N2-yl)-DMABP (2-3%) and N-(deoxyadenosin-8-yl)-DMABP (1-3%). The same adducts have previously been detected in the liver and colon of rats administered DMABP or its hydroxamic acid. Incubation of rat hepatic or intestinal cytosol at pH 7.4 for 15 min with [ring-3H]N-hydroxy-DMABP in the presence of S-acetyl coenzyme A (AcCoA) and calf thymus DNA resulted in DNA binding at levels of 30-80 arylamine residues per 10(6) nucleotides. H.p.l.c. analysis of the DNA modified in the presence of AcCoA indicated the formation of the same adducts detected in the acid-catalyzed reactions. When arylhydroxamic acid N,O-acyltransferase assays were conducted with rat liver cytosol and N-acetyl-N-hydroxy-DMABP as the substrate, binding to nucleic acids was not observed. Similarly, 3'-phosphoadenosine-5'-phosphosulfate-dependent sulfotransferase-mediated DNA binding could not be demonstrated. These data indicate that in a suitable acidic environment, N-hydroxy-DMABP will react with DNA to yield the same adducts found in vivo. Under neutral conditions, however, N-hydroxy-DMABP appears to undergo AcCoA-dependent transacetylation to an electrophilic acetoxy ester which will spontaneously react with DNA.

Microsomal enzyme preparations from dog liver, kidney, and bladder were used to assess the prostaglandin H synthase-catalyzed activation of carcinogenic aromatic amines to bind covalently to proteins and nucleic acids. Benzidine, a urinary bladder carcinogen, bound to protein of bladder transitional epithelial and renal inner and outer medullary microsomes and was dependent upon addition of arachidonic acid, but not upon reduced nicotinamide adenine dinucleotide phosphate. Bladder transitional epithelial microsomes also activated o-dianisidine, 4-aminobiphenyl, and 2-naphthylamine to bind to protein and transfer RNA and benzidine and O-dianisidine to bind DNA. Cosubstrate and inhibitor specificities were consistent with activation by prostaglandin H synthase. Binding of benzidine to protein was not observed with either hepatic or renal cortical microsomes upon addition of arachidonic acid or reduced nicotinamide adenine dinucleotide phosphate. Prostaglandin H synthase and mixed-function oxidase-catalyzed bindings of 2-naphthylamine to protein and to transfer RNA were compared using liver and bladder microsomes. Only mixed-function oxidase-catalyzed binding was observed in liver, and only prostaglandin H synthase-catalyzed binding was observed in bladder. The rate of binding catalyzed by bladder microsomes was considerably greater than that catalyzed by hepatic microsomes. In addition, the bladder content of prostaglandin H synthase activity was approximately 10 times that of kidney inner medullary, a tissue reported to have a relatively high content of this enzyme in other species. These results are consistent with involvement of bladder transitional epithelial prostaglandin H synthase in the genesis of primary aromatic amine-induced bladder cancer.

The N-oxidation of the hepatocarcinogen N-methyl-4-aminoazobenzene (MAB) was catalyzed by hepatic microsomes in a reduced pyridine nucleotide- and oxygen-dependent reaction. The initial N-oxidation product, N-hydroxy-N-methyl-4-aminoazobenzene (N-HO-MAB), was readily oxidized to a second product that yielded N-hydroxy-4-aminoazobenzene upon subsequent acid treatment. The secondary N-oxidation product may be formed nonenzymatically and is presumed to be N-HO-MAB N-oxide or its dehydrated derivative, N-(p-phenylazophenyl)nitrone. Under the same conditions, MAB was also oxidatively N-dealkylated to 4-aminoazobenzene, which was N-oxidized to N-hydroxy-4-aminoazobenzene. Unlike the latter reactions, the microsomal N-oxidation of MAB was independent of cytochrome P-450, as shown by its lack of sensitivity to inhibition by 2-[(2,4-dichloro-6-phenyl)phenoxy]ethylamine and its inability to utilize cumene hydroperoxide in place of reduced pyridine nucleotides and oxygen. The N-oxidation of MAB was also catalyzed by the purified microsomal flavoprotein mixed-function amine oxidase of Ziegler et al. The noncarcinogenic dye N-ethyl-4-aminoazobenzene was metabolized similarly to MAB. For male animals the hepatic levels of MAB N-oxidase activity were in the order: rat greater than hamster, guinea pig greater than mouse, rabbit. Little or no MAB N-oxidase activity was present in several extrahepatic rat tissues. N-HO-MAB, N-hydroxy-N-ethyl-4-aminoazobenzene, and N-hydroxy-4-aminoazobenzene catalyzed the aerobic oxidation of cysteine and glutathione. These hydroxylamines also bound covalently to proteins. The binding of N-HO-MAB with nucleic acids was only 3 to 6% that observed with serum albumin. Under anhydrous conditions the nitrone generated aerobically from N-HO-MAB reacted with carbon-carbon or carbon-nitrogen double bonds, or both, in fatty acids, retinol, purines, and pyrimidines to yield isoxazolidine and/or oxadiazolidine addition products. The nitrone from N-hydroxy-N-ethyl-4-aminoazobenzene was much less reactive under these conditions. Syntheses of N-HO-MAB and N-hydroxy-N-ethyl-4-aminoazobenzene are reported.

Although the peroxidative metabolism of benzidine and other carcinogenic arylamines has been regarded as a significant pathway in extrahepatic tissues, the mechanism of peroxidase-mediated covalent binding to DNA has remained unclear. In this study, we have compared the metabolic activation of benzidine by prostaglandin H synthase, horseradish peroxidase, chloroperoxidase, and lactoperoxidase. All four peroxidases mediated the binding of benzidine to DNA and equimolar amounts of hydrogen peroxide (or arachidonic acid) and benzidine were required for the maximal binding in the system with either horseradish peroxidase or prostaglandin H synthase. In reactions containing both synthetic [3H]benzidine diimine and [14C]benzidine, rapid equilibration of both compounds was evident through the formation of its charge-transfer complex and the 14C-associated binding to DNA. However, the total binding (3H and 14C) correlated with the concentration of benzidine diimine rather than that of the charge-transfer complex. Two major and one minor deoxyguanosine adducts (P-I, P-III and P-IV, respectively) were isolated after the enzymatic hydrolysis of the benzidine-modified DNA. P-I was identified as N-(deoxyguanosin-8-yl)-benzidine, which was the major adduct formed by reaction of benzidine diimine with DNA. P-IV, which was also formed on reaction of benzidine diimine with DNA, was consistent with an N-(deoxyguanosin-N2-yl)-benzidine structure. P-III, which was formed only in the peroxidase incubations with DNA, was characterized as a novel N,3-(deoxyguanosin-N7,C8-yl)-benzidine derivative. Furthermore, this DNA adduct was shown to arise by the action of the peroxidase on DNA that had been previously modified by benzidine diimine. These results indicate that the two-electron oxidation product of benzidine, benzidine diimine, is the predominant reactive intermediate for the DNA binding mediated by peroxidases.

Six cytochrome P-450 (P-450) isozymes were purified to electrophoretic homogeneity from the livers of four human organ donors, with three of these isozymes purified from a single individual. Differences were noted between all six P-450s for some or all of the parameters determined by the techniques of sodium dodecyl sulfate-polyacrylamide gel electrophoresis, peptide mapping, spectral analysis of ferrous-carbon monoxide complexes, double-diffusion immunoprecipitin analysis or crossed immunoelectrophoresis (sodium dodecyl sulfate-polyacrylamide gel electrophoresis/peroxidase-coupled staining) with rabbit antisera raised to five of the P-450s, or catalytic activity toward d-benzphetamine, benzo[a]pyrene, acetanilide, debrisoquine, (R)- and (S)-warfarin, and 1-naphthylamine. While NADPH-fortified human liver microsomal preparations showed catalytic activity toward trichloroethylene, 7-ethoxycoumarin, 2-naphthylamine, and 2-aminofluorene in addition to the other substrates mentioned, none of the P-450s which we purified from these microsomes catalyzed the oxidation of these compounds in reconstituted enzyme systems containing purified rat liver NADPH-P-450 reductase. Antibodies raised against one of the purified P-450s inhibited d-benzphetamine N-demethylase activity in microsomal incubations but did not inhibit the metabolism of 7-ethoxycoumarin, acetanilide, benzo[a]pyrene, or debrisoquine. The data provide a strong biochemical basis for the view that distinct isozymes of P-450 exist in humans and that these isozymes differ in catalytic activity toward drugs and carcinogens.

The carcinogen N-hydroxy-1-naphthylamine reacted with nucleic acids and protein under slightly acidic conditions (pH 5) to form covalently bound derivatives with 3 to 20 naphthyl residues/1000 monomer units. The level of binding was in the following order: DNA greater than polyguanylic acid greater than denatured DNA and ribosomal RNA greater than serum albumin greater than transfer RNA greater than polyadenylic acid. Reactions with nucleosides and nucleotides were not detected, and the binding of N-hydroxy-1-naphthylamine to DNA was not inhibited by the addition of nucleosides, nucleotides, methionine, or glutathione. The reaction rates were first order with respect to both DNA and N-hydroxy-1-naphthylamine concentrations. Enzymatic hydrolysis of the DNA containing naphthyl residues yielded 3 nucleoside-arylamine adducts. The major adduct was identified by chemical, ultraviolet, nuclear magnetic resonance, and mass spectrometric analyses as N-(deoxyguanosin-O6-yl)-1-naphthylamine. The other two adducts were identified as 2-(deoxyguanosin-O6-yl)-1-naphthylamine and its decomposition product. Direct evidence for acid-dependent arylnitrenium ion formation was obtained by isotope exchange upon solvolysis of N-hydroxy-1-naphthylamine in acidic H2 18O, and carbocation formation was indicated by the formation of the solvolysis products, 1-amino-2-naphthol and 1-amino-4-naphthol. These studies demonstrated the conversion of a carcinogenic N-hydroxy arylamine to electrophilic arylnitrenium ion and carbocation species that display high selectivity toward macromolecules. The roles of these electrophiles and their macromolecular adducts in the initiation of urinary bladder carcinogenesis through formation of promutagenic lesions in DNA are suggested.

A major and previously undetected carcinogen-DNA adduct was found in the livers of rats given N,N-dimethylnitrosamine or 1,2-dimethylhydrazine. This adduct, which accounted for 55% of the total methyl residues in DNA at 72 hours after carcinogen treatment, was chromatographically identical to a synthetic purine ring-opened derivative of 7-methylguanine and could be released from the isolated hepatic DNA by a specific E. coli glycosylase. The synthetic ring-opened adduct was characterized by mass and NMR spectroscopy as N5-methyl-N5-formyl-2,5,6-triamino-4-hydroxypyrimidine and appears to exist in two rotameric forms.

Human peripheral lung tissue samples were obtained at autopsy from 17 individuals of known occupational and smoking histories. A spectrum of different carcinogen-DNA adducts was detected using a variety of sensitive techniques. High-pressure liquid chromatography-linked synchronous fluorescent spectrophotometry and an ultrasensitive enzyme radioimmunoassay detected adducts derived from benzo[a]pyrene diol epoxide and other apparent polycyclic aromatic hydrocarbons. An amplified enzyme-linked immunosorbent assay demonstrated the presence of 4-aminobiphenyl-DNA adducts in many of these samples. A number of these specimens also contained O6-alkyldeoxyguanosine as measured by 32P-postlabeling techniques. Thus this pilot study indicates not only that human lung contains a spectrum of carcinogen-DNA adducts, but also that a full scale molecular dosimetry study of human exposure to both aryl and alkyl chemical carcinogens is warranted.

The N-hydroxy derivatives of 1- and 2-naphthylamine (NA) are directly carcinogenic at sites of application. In this study, the carcinogenicity of these two compounds at s.c. injection sites was compared with their relative rates of absorption, with the extent and persistence of their binding to protein, RNA, and DNA in the skin-subcutis, and with acute histopathological changes observed after local application. Male Sprague-Dawley rats were given injections of the N-hydroxy derivatives in the right rear leg at 16 mumol/dose. When administered twice weekly for 12 weeks, N-hydroxy-1-NA caused a 100% incidence (30 of 30) of poorly differentiated sarcomas at the injection site. N-Hydroxy-2-NA administered in a similar manner resulted in a low yield of tumors (7%; 2 of 30). Injection of N-hydroxy-1-NA once weekly for 12 weeks or twice weekly for 6 weeks also induced a high incidence of sarcomas (93 to 97%), but the time to tumor formation was significantly longer (p less than 0.0001) than in animals treated twice weekly for 12 weeks. The tumors were classified as malignant fibrous histiocytomas. Possible antagonistic or synergistic effects between the two compounds were also investigated. A sequential 6-week treatment with each of the N-hydroxy derivatives did not alter the expected tumor yields. However, alternating injections over 12 weeks caused a significant lengthening in the time to tumor formation (p less than 0.05). N-Hydroxy-1-NA bound covalently to protein, RNA, and DNA to a much greater extent than did N-hydroxy-2-NA. Protein binding with both derivatives decreased by 80 to 90% by 7 days after treatment. RNA binding in N-hydroxy-1-NA-treated rats markedly decreased, while N-hydroxy-2-NA-bound residues diminished only slightly. During this period, the extent of DNA binding with both derivatives remained fairly constant. When N-hydroxy-2-NA was injected 3 days after N-hydroxy-1-NA, there was a marked reduction in the apparent levels of N-hydroxy-1-NA bound to RNA and DNA. The greater tumorigenicity of N-hydroxy-1-NA versus N-hydroxy-2-NA correlated with its greater extent of macromolecular binding. Examination of acute histopathological changes occurring after single injections of N-hydroxy-1-NA and/or N-hydroxy-2-NA indicated that both compounds caused extensive necrosis in tissues at the injection site.(ABSTRACT TRUNCATED AT 400 WORDS)

Since the N-oxidation of several carcinogenic arylamines has been shown to be catalyzed preferentially by cytochrome P-450IA2 in several species, homologous ethynyl-substituted aromatic hydrocarbons, 2-ethynylnaphthalene, 1-ethynylnaphthalene, and 2-ethynylfluorene, were synthesized and examined as potential mechanism-based inactivators of this monooxygenase. By use of 2-naphthylamine, whose N-oxidation was known to be selectively catalyzed by rat cytochrome P-450ISF-G (P-450IA2), and hepatic microsomes from isosafrole-treated rats, each of these ethynyl derivatives was found to be strongly inhibitory at concentrations of 1 and 10 microM. However, only inhibition by 2-ethynylnaphthalene was significantly enhanced by prior incubation with the microsomal system. The inactivation of 2-naphthylamine N-oxidation was found to be NADPH- and time-dependent and to follow pseudo-first-order kinetics, demonstrating that 2-ethynylnaphthalene is a potent mechanism-based inactivator of the enzymatic activity. The extrapolated kinactivation and KI were 0.23 min-1 and 8 microM, respectively. By use of 2-aminofluorene, whose N-oxidation was known to be catalyzed by both cytochromes P-450ISF-G and P-450 beta NF-B (P-450IA1), and the purified enzymes in a reconstituted system, both 2-ethynylnaphthalene and 1-ethynylnaphthalene were found to be strongly inhibitory. However, 2-ethynylnaphthalene was a more potent inhibitor of the purified P-450ISF-G than of P-450 beta NF-B; and it was also found to be a more potent inhibitor of P-450ISF-G than was 1-ethynylnaphthalene.(ABSTRACT TRUNCATED AT 250 WORDS)

Formation of DNA adducts in various tissues of dogs fed a single dose of the carcinogen 2-aminofluorene was investigated. Adduct analysis was performed using a technique that allows measurement of both N-(deoxyguanosin-8-yl)-2-amino-2-aminofluorene-DNA adduct formed by reaction of N-hydroxy-2-aminofluorene with DNA, as well as the polar 2-aminofluorene-DNA adducts formed when 2-aminofluorene is activated by prostaglandin H synthase-peroxidase in vitro. Two male beagle (A and B) dogs were examined and a different DNA adduct profile was observed with each dog. For the dog A, N-(deoxyguanosin-8-yl)-2-aminofluorene was the major adduct found in hepatic DNA; no peroxidase-derived adducts were detected in this tissue. In contrast, adducts eluting similarly to peroxidase-derived adducts were found in urinary tract tissues of this dog with the relative abundance of these adducts in the order urothelium greater than renal medulla greater than renal cortex, which correlates with the respective tissues' prostaglandin H synthase activity. N-(Deoxyguanosin-8-yl)-2-aminofluorene was detected in the renal tissues, but not in urothelium. For dog B, only the N-(deoxyguanosin-8-yl)-2-aminofluorene adduct was observed in all tissues examined, including the urothelium. However, total binding to liver, kidney, and bladder were two-, two-, and four-fold lower, respectively, than dog A. These data indicate that both prostaglandin H synthase-mediated activation and N-hydroxylation of 2-aminofluorene occur in vivo and may be subjected to pharmacodynamic considerations. Furthermore, the tissue distribution of the peroxidase-mediated 2-aminofluorene adducts suggests this process may also be of importance in the bladder-specific carcinogenicity of aromatic amines.