The role of DNA adducts in smoking-related carcinogenesis

Epidemiological studies indicate a close association between smoking and cancer. Biological activity of many chemical carcinogens and of their metabolites is induced by covalent binding of their reactive derivatives to DNA, which consequently causes mutations in critical genes. Carcinogen-DNA adducts formed by exposure to tobacco smoke have a key role in the initiation of various types of cancer including lung cancer. Presence of tobacco smoke-related carcinogen-DNA adducts in various tissues of smokers proves the DNA damaging effect of smoking. DNA adducts are important biomarkers for the biomonitoring of human genotoxic exposures to tobacco smoke. The paper gives a short overview on the role of smoking-related DNA adducts in carcinogenesis.     

Application of new techniques for the detection of carcinogen adducts to human population monitoring

Several techniques have recently been developed for the detection and quantitation of carcinogen-DNA or -protein adducts without the requirement for radioactive carcinogens. These assays can be used to detect adducts in animals or cultured cells exposed to test compounds or in humans exposed to environmental carcinogens. Immunologic, 32P-postlabeling and fluorescence techniques, used on human samples for DNA adduct measurement, are reviewed here. Methods for the detection of carcinogen-protein adducts on human samples are also summarized.     

Preferential carcinogen-DNA adduct formation at codons 12 and 14 in the human K-ras gene and their possible mechanisms

In the ras gene superfamily, codon 12 (-TGGTG-) of the K-ras gene is the most frequently mutated codon in human cancers. Recently, we have found that bulky chemical carcinogens preferentially form DNA adducts at codons 12 and 14 (-CGTAG-) in the K-ras gene in normal human bronchial epithelial (NHBE) cells. Furthermore, DNA adducts formed at codon 12 of the K-ras gene are poorly repaired compared with those at other codons including codon 14. These results suggest that targeted carcinogen-DNA adduct formation is a major reason for the observed high mutation frequency at codon 12 of the K-ras gene in human cancers. This preferential carcinogen-DNA adduct formation at codons 12 and 14 could result from effects of (1) primary sequences of these codons and their surrounding codons in the K-ras gene, (2) the chromatin structure, and/or (3) epigenetic factors such as C5 cytosine methylation or other DNA modifications at these codons and their surrounding codons. To distinguish these possibilities, we have introduced modifications with benzo[a]pyrene diol epoxide, N-hydroxy-2-aminofluorene, and aflatoxin B1 8,9-epoxide in (1) naked intact genomic DNA isolated from NHBE cells, (2) fragmented genomic DNA digested by restriction enzymes, and (3) in vitro synthesized DNA fragments containing the K-ras gene exon 1 sequence with or without methylation of the cytosines at CpG sites and the cytosines pairing with the guanines of codons 12 and 14. The distribution of carcinogen-DNA adducts in the K-ras gene was mapped at the nucleotide sequence level using the UvrABC nuclease incision method with or without the ligation-mediated polymerase chain reaction technique. We have found that carcinogens preferentially form adducts at codons 12 and 14 in the K-ras gene exon 1 in intact as well as in fragmented genomic DNA. In contrast, this preferential DNA adduct formation at codons 12 and 14 was not observed in PCR-amplified DNA fragments containing the K-ras gene exon 1 sequence. Methylation of the cytosine at the CpG site of codon 14, or the cytosine pairing with guanine of codon 14, greatly enhanced carcinogen-DNA adduct formation at codon 14 but did not affect carcinogen-DNA adduct formation at codon 12. Methylation of the cytosine pairing with the guanine of codon 12 also did not enhance carcinogen-DNA adduct formation at codon 12. Furthermore, we found that the cytosine at the CpG site of codon 14 is highly methylated in NHBE cells. These results suggest that cytosine methylation at the CpG site is the major reason for the preferential DNA damage at codon 14 and that epigenetic modification(s) other than cytosine methylation may contribute to the preferential DNA damage at codon 12 of the K-ras gene.     

DNA adducts, DNA repair genotype/phenotype and cancer risk

It is well established that the initiating event in chemical carcinogenesis is the binding of reactive carcinogens to DNA. Thus, a number of analytic methods have been developed for determining levels of carcinogen-DNA adducts in humans as a marker of individual exposure and, potentially, of risk for cancer development. We have developed monoclonal and polyclonal antibodies to carcinogen-DNA adducts and highly sensitive ELISA and immunohistochemical assays for determining levels of adducts in human tissues. These methods have been combined with genotyping and phenotyping methods for DNA repair to study gene-environment interactions in cancer risk. Recent studies on breast cancer have utilized two large biorepositories. The first is blood and tumor tissues collected as part of the Long Island Breast Cancer Study Project, a population-based case-control study of environmental risk factors. The second is the Metropolitan New York Registry of Breast Cancer Families, one of six sites funded by the National Cancer Institute as a part of the Breast Cooperative Family Registry (CFR). Analysis of samples from these two studies have demonstrated the utility of measurement of DNA adducts as biomarkers of exposure and that DNA repair capacity, measured by genotyping or phenotyping, can influence risk.     

Understanding the role of xenobiotic-metabolism in chemical carcinogenesis using gene knockout mice

Most chemical carcinogens require metabolic activation to electrophilic metabolites that are capable of binding to DNA and causing gene mutations. Carcinogen metabolism is carried out by large groups of xenobiotic-metabolizing enzymes that include the phase I cytochromes P450 (P450) and microsomal epoxide hydrolase, and various phase II transferase enzymes. It is extremely important to determine the role P450s play in the carcinogenesis and to establish if they are the rate limiting and critical interface between the chemical and its biological activities. The latter is essential in order to validate the use of rodent models to test safety of chemicals in humans. Since there are marked species differences in expressions and catalytic activities of the multiple P450 forms that activate carcinogens, this validation process becomes especially difficult. To address the role of P450s in whole animal carcinogenesis, mice were produced that lack the P450s known to catalyze carcinogen activation. Mouse lines having disrupted genes encoding the P450s CYP1A2, CYP2E1, and CYP1B1 were developed. Mice lacking expression of microsomal epoxide hydrolase (mEH) and NADPH-quinone oxidoreductase (NQO1) were also made. All of these mice exhibit no gross abnormal phenotypes, suggesting that the xenobiotic-metabolizing enzymes have no critical roles in mammalian development and physiological homeostasis. This explains the occurrence of polymorphisms in xenobiotic-metabolizing enzymes among humans and other mammalian species. However, these null mice do show differences in sensitivities to acute chemical toxicities, thus establishing the importance of xenobiotic metabolism in activation pathways that lead to cell death. Rodent bioassays using null mice and known genotoxic carcinogens should establish whether these enzymes are required for carcinogenesis in an intact animal model. These studies will also provide a framework for the production of transgenic mice and carcinogen bioassay protocols that may be more predictive for identifying the human carcinogens and validate the molecular epidemiological studies ongoing in humans that seek to establish a role for polymorphisms in cancer risk.     

Quantitation of benzo[a]pyrene-DNA adducts by postlabeling with 14C-acetic anhydride and accelerator mass spectrometry

Quantitation of carcinogen-DNA adducts provides an estimate of the biologically effective dose of a chemical carcinogen reaching the target tissue. In order to improve exposure-assessment and cancer risk estimates, we are developing an ultrasensitive procedure for the detection of carcinogen-DNA adducts. The method is based upon postlabeling of carcinogen-DNA adducts by acetylation with 14C-acetic anhydride combined with quantitation of 14C by accelerator mass spectrometry (AMS). For this purpose, adducts of benzo[a]pyrene-r-7,t-8-dihydrodiol-t-9,10-epoxide (BPDE) with DNA and deoxyguanosine (dG) were synthesized. The most promutagenic adduct of BPDE, 7R,8S,9R-trihydroxy-10S-(N(2)-deoxyguanosyl)-7,8,9, 10-tetrahydrobenzo[a]pyrene (BPdG), was HPLC purified and structurally characterized. Postlabeling of the BPdG adduct with acetic anhydride yielded a major product with a greater than 60% yield. The postlabeled adduct was identified by liquid chromatography-mass spectrometry as pentakis(acetyl) BPdG (AcBPdG). Postlabeling of the BPdG adduct with 14C-acetic anhydride yielded a major product coeluting with an AcBPdG standard. Quantitation of the 14C-postlabeled adduct by AMS promises to allow detection of attomolar amounts of adducts. The method is now being optimized and validated for use in human samples.     

High-performance liquid chromatographic analysis of 32P-postlabeled DNA-aromatic carcinogen adducts

The technique of 32P postlabeling of DNA-carcinogen adducts is a useful and extremely sensitive method of detecting and quantitating DNA damage by carcinogens. We have adapted the 32P method to analysis by high-pressure liquid chromatography, making the procedure more rapid and convenient than when thin-layer chromatography is used. Following DNA isolation and hydrolysis, nucleotide-carcinogen adducts are enhanced relative to normal nucleotides by solvent extraction and then labeled with high-specific-activity [gamma-32P]ATP. The resulting 32P-postlabeled nucleotides are resolved by reverse-phase ion-pair HPLC. After as little as 3 h of exposure to carcinogens, DNA adducts can be demonstrated from 1 microgram or less of mouse hepatic DNA. Acetylated and nonacetylated adducts can be resolved from hepatic DNA of mice treated with 2-aminofluorene. Differences in DNA damage as measured by adduct formation were demonstrated between "rapid" and "slow" acetylator mouse strains. Rapid-acetylator C57BL/6J mice had three times the amount of hepatic DNA adducts as slow-acetylator A/J mice 3 h after a 60 mg/kg dose of 2-aminofluorene. 4-Aminobiphenyl and 2-naphthylamine each showed an adduct peak with retention time similar to that of the nonacetylated 2-aminofluorene adduct, while benzidine gave a major adduct that eluted somewhat earlier as would be expected for an acetylated adduct. The alkenylbenzenes, safrole and methyleugenol, also formed DNA adducts detectable by this method. DNA prepared from skin of mice painted with benzo[a]pyrene also contained carcinogen-DNA adducts detectable and resolvable by HPLC analysis following 32P postlabeling. The combination of HPLC with 32P postlabeling appears to be a useful technique for the rapid detection and quantitation of DNA damage caused by several classes of aromatic carcinogens.     

Antibodies to DNA alkylation addcuts as analytical tools in chemical carcinogenesis

The availability of antibodies recognising specific carcinogen-induced alkylated DNA adducts has contributed, and will continue to contribute, to the study of the biological significance of these adducts with respect to their toxic, mutagenic and carcinogenic properties. Whilst the antibodies have undoubtedly facilitated the study of such mechanistic aspects in animal and cell-culture systems, their availability has also had a crucial and central role in the area of quantitating human exposure to environmental alkylating agents. This latter field of research was initially established and developed as a result of the availability of suitable antibodies. This review will only briefly cover the range of adducts for which antibodies and immunoassays are available, indicating some of the relevant properties of the applied techniques, and will focus on the areas of research which can be furthered specifically by these immunoanalytical tools. Previous extensive reviews of antibodies to carcinogen-DNA adducts are available (Muller and Rajewsky, 1981; Poirier, 1981; Strickland and Boyle, 1984; Kriek et al., 1984).     

Endogenous mutagens and the causes of aging and cancer

A very large oxidative damage rate to DNA occurs as part of normal metabolism. In each rat cell the steady-state level is estimated to be about 10⁶ oxidative adducts and about 10⁵ new adducts are formed daily. It is argued that this endogenous DNA damage is a major contributor to aging and the degenerative diseases of aging, such as cancer. The oxidative damage rate in mammalian species with a high metabolic rate, short life span, and high age-specific cancer rate is much higher than the rate in humans, a long-lived creature with a lower metabolic rate and a lower age-specific cancer rate. It is argured that deficiency of micronutrients, such as dietary antioxidants or folate, is a major contributor to human cancer and degenerative diseases.

Understanding the role of mitogenesis in mutagenesis is critical for clarifying the mechanisms of carcinogenesis and interpreting high-dose animal cancer tests. High-dose animal cancer tests have been done mainly on synthetic industrial chemicals, yet almost all of the chemicals humans are exposed to are natural. About half of natural chemicals tested in high-dose animal cancer tests are rodent carcinogens, a finding that is consistent with the view that high-dose tests frequently increase mitogenesis rates. Animals have numerous defenses against toxins that make them very well buffered against low doses of almost all toxins, whether synthetic or natural.     

Flavonoids and alkenylbenzenes: mechanisms of mutagenic action and carcinogenic risk

The present review focuses on the mechanisms of mutagenic action and the carcinogenic risk of two categories of botanical ingredients, namely the flavonoids with quercetin as an important bioactive representative, and the alkenylbenzenes, namely safrole, methyleugenol and estragole. For quercetin a metabolic pathway for activation to DNA-reactive species may include enzymatic and/or chemical oxidation of quercetin to quercetin ortho-quinone, followed by isomerisation of the ortho-quinone to quinone methides. These quinone methides are suggested to be the active alkylating DNA-reactive intermediates. Recent results have demonstrated the formation of quercetin DNA adducts in exposed cells in vitro. The question that remains to be answered is why these genotoxic characteristics of quercetin are not reflected by carcinogenicity. This might in part be related to the transient nature of quercetin quinone methide adducts, and suggests that stability and/or repair of DNA adducts may need increased attention in in vitro genotoxicity studies. Thus, in vitro mutagenicity studies should put more emphasis on the transient nature of the DNA adducts responsible for the mutagenicity in vitro, since this transient nature of the formed DNA adducts may play an essential role in whether the genotoxicity observed in vitro will have any impact in vivo. For alkenylbenzenes the ultimate electrophilic and carcinogenic metabolites are the carbocations formed upon degradation of their 1'-sulfooxy derivatives, so bioactivation of the alkenylbenzenes to their ultimate carcinogens requires the involvement of cytochromes P450 and sulfotransferases. Identification of the cytochrome P450 isoenzymes involved in bioactivation of the alkenylbenzenes identifies the groups within the population possibly at increased risk, due to life style factors or genetic polymorphisms leading to rapid metaboliser phenotypes. Furthermore, toxicokinetics for conversion of the alkenylbenzenes to their carcinogenic metabolites and kinetics for repair of the DNA adducts formed provide other important aspects that have to be taken into account in the high to low dose risk extrapolation in the risk assessment on alkenylbenzenes. Altogether the present review stresses that species differences and mechanistic data have to be taken into account and that new mechanism- and toxicokinetic-based methods and models are required for cancer risk extrapolation from high dose experimental animal data to low dose carcinogenic risks for man.     

Studies on biomarkers in cancer etiology and prevention: a summary and challenge of 20 years of interdisciplinary research

Sensitive, specific methods have been developed that allow quantitative measurements of the metabolites of carcinogen metabolites and of DNA and protein adducts in humans exposed occupationally, environmentally and endogenously to genotoxic agents. The interrelationship between exposure to carcinogens, host risk factors and the responses of biomarkers has been examined in cross-sectional, ecological and case-control studies which provided new insights into the causes of cancer and the mechanisms of carcinogenesis. The identification of hitherto unknown DNA-reactive chemicals formed in the human body from dietary precursors and of carcinogenic components of complex mixtures has increased the possibility of establishing causal relationships in etiology. The identification of individuals and subgroups heavily exposed to carcinogens has led to the development of measures for avoiding or decreasing exposure to carcinogenic risk factors. New, ultrasensitive methods for measuring DNA adducts allow the quantification and structural elucidation of specific DNA damage in humans arising from oxidative stress and lipid peroxidation (LPO), which have been found to be the driving forces in several human malignancies. Background DNA damage in "unexposed" individuals has been shown unequivocally to be due to LPO products, and a significant interindividual variation in adduct levels has been shown in individuals with comparable exposure to carcinogens. Thus, pharmacogenetic variants with higher susceptibility to carcinogenic insults, due to genetic polymorphism in xenobiotic-metabolizing enzymes, have been characterized by a combination of genotyping and measurements of macromolecular adducts. Dosimetry has been used in human studies to evaluate the efficacy of interventions with chemopreventive agents like ascorbic acid, dietary phenols and green tea. Advances in the application of selected biomarkers in human studies are reviewed and illustrated by examples from the author's research conducted during the past two decades.     

DNA adducts as markers of exposure and risk

Many carcinogens exert their biological effects through the formation of DNA adducts by metabolically activated intermediates. Detecting the presence of DNA adducts in human tissues is, therefore, a tool for molecular epidemiological studies of cancer. A large body of evidence demonstrates that DNA adducts are useful markers of carcinogen exposure, providing an integrated measurement of carcinogen intake, metabolic activation, and delivery to the target macromolecule in target tissues. Monitoring accessible surrogate tissues, such as white blood cells, also provides a means of investigating occupational or environmental exposure in healthy individuals. Such exposure to carcinogens, e.g. to polycyclic aromatic hydrocarbons, has been demonstrated in several industries and in defined populations, respectively, by the detection of higher levels of adducts. Adducts detected in many tissues of smokers are at levels significantly higher than in non-smokers, although the magnitude of the elevation does not predict the magnitude of the risk. While such associations do not demonstrate causality, they do, importantly, lend plausibility to observed associations between smoking and cancer. However, there is still resistance to the notion that such monitoring can inform, rather than merely confirm, epidemiological investigations of cancer causation. Interestingly, smoking was recently causally linked to cervical cancer after years of being considered a confounding factor; yet smoking-related adducts have been known to be present in cervical epithelium for some time. In the few prospective studies thus far, elevated adduct levels have been found in individuals who subsequently developed cancer compared with individuals who did not. The potential for biomarker measurements, such as DNA adducts, to provide answers to the origin of many cases of human cancer for which an environmental cause is suspected, needs to be exploited more fully in future epidemiological studies.     

Forging the links between metabolism and carcinogenesis

Metabolism plays important roles in chemical carcinogenesis, both good and bad. The process of carcinogen metabolism was first recognized in the first half of the twentieth century and developed extensively in the latter half. The activation of chemicals to reactive electrophiles that become covalently bound to DNA and protein was demonstrated by Miller and Miller [Cancer 47 (1981) 2327]. Today many of the DNA adducts formed by chemical carcinogens are known, and extensive information is available about pathways leading to the electrophilic intermediates. Some concepts about the stability and reactivity of electrophiles derived from carcinogens have changed over the years. Early work in the field demonstrated the ability of chemicals to modulate the metabolism of carcinogens, a phenomenon now described as enzyme induction. The cytochrome P450 enzymes play a prominent role in the metabolism of carcinogens, both in bioactivation and detoxication. The conjugating enzymes can also play both beneficial and detrimental roles. As an example of a case in which several enzymes affect the metabolism and carcinogenicity of a chemical, aflatoxin B1 (AFB1) research has revealed insight into the myriad of reaction chemistry that can occur even with a 1s half-life for a reactive electrophile. Further areas of investigation involve the consequences of enzyme variability in humans and include areas such as genomics, epidemiology, and chemoprevention.     

P-Postlabelling and mass spectrometric methods for analysis of bulky, polyaromatic carcinogen—DNA adducts in humans

There has been significant recent progress toward the development of human carcinogen—DNA adduct biomonitoring methods. ³²P-Postlabelling is a technique which has found wide application in human studies. ³²P-Postlabelling involves enzymatic preparation and labelling of DNA samples, followed by chromatographic separation of carcinogen—nucleotide adducts from unadducted nucleotides. Thin-layer ion-exchange and high-performance liquid chromatography (HPLC) have been utilized. This paper critically reviews ³²P-postlabelling methods for analysis of bulky, polyaromatic carcinogen—DNA adducts and details a strategy to optimize this technique for monitoring human samples. Development of a human carcinogen biomonitoring method requires that the biomarker meet certain criteria: that the biomarker be responsive to exposures known to increase human cancer risk, to reductions in those exposures, and to the influence of metabolic differences. In addition, reliable samples must be available by non-invasive means. The ability of ³²P-postlabelling to meet these criteria is traced in the literature and discussed. Identification of specific carcinogen—DNA adducts is a difficult task due to the low (femtomole) levels in human target tissues. Because co-chromatography in thin-layer chromatography (TLC) is generally not considered to be proof of chemical identity, both synchronous fluorescence and HPLC in conjunction with ³²P-postlabelling and TLC are used to confirm the identity of specific carcinogen-DNA adducts in human samples. Mass spectrometry is a highly specific method, the sensitivity of which has been improved to the point which may allow its use to confirm the identity of carcinogen—DNA adducts isolated by ³²P-postlabelling and other methods. The literature relating to the use of mass spectral techniques in carcinogen—DNA adduct analysis is reviewed.     

Human hepatic cell cultures: in vitro and in vivo drug metabolism

Drug metabolism is the major determinant of drug clearance, and the factor most frequently responsible for inter-individual differences in drug pharmacokinetics. The expression of drug metabolising enzymes shows significant interspecies differences, and variability among human individuals (polymorphic or inducible enzymes) makes the accurate prediction of the metabolism of a new compound in humans difficult. Several key issues need to be addressed at the early stages of drug development to improve drug candidate selection: a) how fast the compound will be metabolised; b) what metabolites will be formed (metabolic profile); c) which enzymes are involved and to what extent; and d) whether drug metabolism will be affected directly (drug-drug interactions) or indirectly (enzyme induction) by the administered compound. Drug metabolism studies are routinely performed in laboratory animals, but they are not sufficiently accurate to predict the metabolic profiles of drugs in humans. Many of these issues can now be addressed by the use of relevant human in vitro models, which speed up the selection of new candidate drugs. Human hepatocytes are the closest in vitro model to the human liver, and they are the only model which can produce a metabolic profile of a drug which is very similar to that found in vivo. However, the use of human hepatocytes is restricted, because limited access to suitable tissue samples prevents their use in high throughput screening systems. The pharmaceutical industry has made great efforts to develop fast and reliable in vitro models to overcome these drawbacks. Comparative studies on liver microsomes and cells from animal species, including humans, are very useful for demonstrating species differences in the metabolic profile of given drug candidates, and are of great value in the judicious and justifiable selection of animal species for later pharmacokinetic and toxicological studies. Cytochrome P450 (CYP)-engineered cells (or microsomes from CYP-engineered cells, for example, Supersomes) have made the identification of the CYPs involved in the metabolism of a drug candidate more straightforward and much easier. However, the screening of compounds acting as potential CYP inducers can only be conducted in cellular systems fully capable of transcribing and translating CYP genes.     

DNA adducts in humans as dosimeters of exposure to environmental, occupational, or dietary genotoxins.

The quantitation of adducts of genotoxins with DNA is probably one of the best indicators of genetic damage due to exposure to toxins or carcinogens. It is generally believed that such adducts can lead to mutations, which in turn can trigger the initiation of the carcinogenic process. DNA adducts have been quantitated in white blood cells and in various tissues of smokers, persons in certain high-exposure occupations, and persons consuming foods contaminated with certain carcinogens. The feasibility of this approach for biochemical epidemiologic studies has been demonstrated using methods such as 32P-postlabeling, enzyme-linked immunosorbent assay, and synchronous fluorescence spectrophotometry. Relatively large interindividual differences in DNA adducts have been observed in both exposed and nonexposed persons. As a result, there are only a few studies in which clear quantitative and qualitative differences between these two groups have been observed. In addition, it appears that in some studies the 32P-postlabeling method does not detect the presence of the polycyclic aromatic hydrocarbon DNA adducts that are detectable by immunoassays. More extensive studies in additional populations at risk should shed further light on the utility of DNA adduct analysis in biochemical monitoring, especially if further refinements in methodology would result in increased sensitivity and specificity.     

Increased DNA alterations in atherosclerotic lesions of individuals lacking the GSTM1 genotype.

Reduced glutathione (GSH) plays a critical role as an intracellular defense system providing detoxification of a broad spectrum of reactive species and their excretion as water-soluble conjugates. Conjugation of GSH with electrophiles is catalyzed by GSH S-transferases (GST), which constitute a broad family of phase II isoenzymes. Two of the GST encoding genes, GSTM1 (mu) and GSTT1 (theta), have a null genotype due to their homozygous deletion that results in lack of active protein. Polymorphisms within GSTT1 and especially GSTM1 have often been associated with cancer in various organs as well as with elevated levels of DNA adducts in various cell types. We recently demonstrated that DNA adducts are consistently detectable in smooth muscle cells (SMC) of human abdominal aorta affected by atherosclerotic lesions. Here we provide evidence that levels of adducts to SMC DNA from atherosclerotic lesions are consistently increased in individuals having the null GSTM1 genotype, whereas no association was established with the GSTT1 polymorphism. The influence of GSTM1 deletion was better expressed in never-smokers and ex-smokers than in current smokers. These findings bear relevance to the epidemiology of atherosclerosis and suggest that metabolic polymorphisms may contribute to the interindividual variability in susceptibility not only to carcinogens, but also to DNA binding atherogens.     

The 32P-postlabeling assay for DNA adducts

32P-postlabeling analysis is an ultrasensitive method for the detection and quantitation of carcinogen-DNA adducts. It consists of four principal steps: (i) enzymatic digestion of DNA to nucleoside 3'-monophosphates; (ii) enrichment of the adduct fraction of the DNA digest; (iii) 5'-labeling of the adducts by transfer of 32P-orthophosphate from [gamma-32P]ATP mediated by polynucleotide kinase (PNK); (iv) chromatographic or electrophoretic separation of the labeled adducts or modified nucleotides and quantitation by measurement of their radioactive decay. The assay requires only microgram quantities of DNA and is capable of detecting adducts at frequencies as low as 1 in 10(10) nt, making it applicable to the detection of events resulting from environmental exposures, or experiments using physiological concentrations of agents. It has a wide range of applications in human, animal and in vitro studies, and can be used for a wide variety of classes of compound and for the detection of adducts formed by complex mixtures. This protocol can be completed in 3 d.