Lipid peroxidation-induced DNA damage in cancer-prone inflammatory diseases: a review of published adduct types and levels in humans

Persistent oxidative stress and excess lipid peroxidation (LPO), induced by inflammatory processes, impaired metal storage, and/or dietary imbalance, cause accumulations and massive DNA damage. This massive DNA damage, along with deregulation of cell homeostasis, leads to malignant diseases. Reactive aldehydes produced by LPO, such as 4-hydroxy-2-nonenal, malondialdehyde, acrolein, and crotonaldehyde, react directly with DNA bases or generate bifunctional intermediates which form exocyclic DNA adducts. Modification of DNA bases by these electrophiles, yielding promutagenic exocyclic adducts, is thought to contribute to the mutagenic and carcinogenic effects associated with oxidative stress-induced LPO. Ultrasensitive detection methods have facilitated studies of the concentrations of promutagenic DNA adducts in human tissues, white blood cells, and urine, where they are excreted as modified nucleosides and bases. Thus, immunoaffinity-(32)P-postlabeling, high-performance liquid chromatography-electrochemical detection, gas chromatography-mass spectrometry, liquid chromatography-tandem mass spectrometry, immunoslotblot assay, and immunohistochemistry have made it possible to detect background concentrations of adducts arising from endogenous LPO products in vivo and studies of their role in carcinogenesis. These background adduct levels in asymptomatic human tissues occur in the order of 1 adduct/10(8) and in organs affected by cancer-prone inflammatory diseases these can be 1 or 2 orders of magnitude higher. In this review, we critically discuss the accuracy of the available methods and their validation and summarize studies in which measurement of exocyclic adducts suggested new mechanisms of cancer causation, providing potential biomarkers for cancer risk assessment in humans with cancer-prone diseases.     

Comparison of DNA adduct levels associated with oxidative stress in human pancreas

DNA adducts associated with oxidative stress are believed to involve the formation of endogenous reactive species generated by oxidative damage and lipid peroxidation. Although these adducts have been reported in several human tissues by different laboratories, a comparison of the levels of these adducts in the same tissue samples has not been carried out. In this study, we isolated DNA from the pancreas of 15 smokers and 15 non-smokers, and measured the levels of 1,N⁶-etheno(2′-deoxy)guanosine (edA), 3,N⁴-etheno(2′-deoxy)cytidine (edC), 8-oxo-2′-deoxyguanosine (8-oxo-dG), and pyrimido[1,2-]purin-10(3H)-one (m₁G). Using the same DNA, the glutathione S-transferase (GST) M1, GSTT1, and NAD(P)H quinone reductase-1 (NQO₁) genotypes were determined in order to assess the role of their gene products in modulating adduct levels through their involvement in detoxification of lipid peroxidation products and redox cycling, respectively. The highest adduct levels observed were for m₁G, followed by 8-oxo-dG, edA, and edC, but there were no differences in adduct levels between smokers and non-smokers and no correlation with the age, sex or body mass index of the subject. Moreover, there was no correlation in adduct levels between edA and eC, or between edA or edC and m₁G or 8-oxo-dG. However, there was a significant correlation (r=0.76; p<0.01) between the levels of 8-oxo-dG and m₁G in human pancreas DNA. Neither GSTM1 nor NQO₁ genotypes were associated with differences in any of the adduct levels. Although the sample set was limited, the data suggest that endogenous DNA adduct formation in human pancreas is not clearly derived from cigarette smoking or from (NQO₁)-mediated redox cycling. Further, it appears that neither GSTM1 nor GSTT1 appreciably protects against endogenous adduct formation. Together with the lack of correlation between m₁G and edA or edC, these data indicate that the malondialdehyde derived from lipid peroxidation may not contribute significantly to m₁G adduct formation. On the other hand, the apparent correlation between m₁G and 8-oxo-dG and their comparable high levels are consistent with the hypothesis that m₁G is formed primarily by reaction of DNA with a base propenal, which, like 8-oxo-dG, is thought to be derived from hydroxyl radical attack on the DNA.     

Correlation of DNA adduct levels with tumor incidence: carcinogenic potency of DNA adducts

The quantitative relationship between DNA adducts and tumor incidence is evaluated in this review. All available data on DNA adduct levels determined after repeated administration of a carcinogen to rats or mice have been compiled. The list comprised 27 chemicals, of all major structural classes of carcinogens. For the correlation with tumor incidence, the DNA adduct levels measured at the given dose were normalized to the dose which resulted in a 50% tumor incidence under the conditions of a 2-year bioassay (TD50 dose). In rat liver, the calculated adduct concentration 'responsible' for a 50% hepatocellular tumor incidence spanned from 53 to 2083 adducts per 108 nucleotides, for aflatoxin B1, tamoxifen, IQ, MeIQx, 2,4-diaminotoluene, and dimethylnitrosamine (in this order). In mouse liver, the respective figures were 812 to 5543 adducts per 108 nucleotides, for ethylene oxide, dimethylnitrosamine, 4-aminobiphenyl, and 2-acetylaminofluorene. The observed span (40-fold in rats, 7-fold in mice) reflects differences between the various DNA adducts to lead to critical mutations. If additional carcinogens fit in with this astonishingly narrow range, the measurement of DNA adduct levels in target tissue has the potential to be not only an exposure marker but an individual cancer risk marker. For toremifen and styrene, low levels of DNA adducts were detected in rat liver at the end of a negative long-term bioassay. This shows that the limit of detection of DNA adducts can be well below the limit of detection of an increased tumor incidence. For a cancer risk assessment at low levels of DNA damage, treatment-related adducts must be discussed in relation to the background DNA damage and its inter- and intraindividual variability.     

Stability of acetaldehyde-derived DNA adduct in vitro

Acetaldehyde (AA) derived from alcoholic beverages is a confirmed carcinogen for esophageal and head and neck cancers. AA forms various DNA adducts and is thought to play a crucial role in carcinogenesis. Transient DNA adducts are usually repaired, but the stability of AA-derived DNA adducts has not been elucidated. We investigated the stability of N(2)-ethylidene-2'-deoxyguanosine (N(2)-ethylidene-dG), a major AA-derived DNA adduct, in cultured cells. First, to determine the optimal concentration of AA for detecting N(2)-ethylidene-dG in cell culture, a dose-response study was performed using HL60 cells of the human promyelocytic leukemia cell line. An AA concentration ≥ 0.01% (1.8 mM) was required to detect N(2)-ethylidene-dG in vitro. We next examined the stability of N(2)-ethylidene-dG. After a 1 or 2h exposure to 0.01% of AA in a tightly sealed bottle, N(2)-ethylidene-dG content was measured by sensitive liquid chromatography tandem mass spectrometry immediately, 24h, and 48 h after exposure. After the 1h exposure, the mean (± SD) N(2)-ethylidene-dG contents were 12.1 ± 1.28, 8.20 ± 0.64, and 6.70 ± 0.52 adducts per 10(7) bases at each postexposure time. After the 2h exposure, N(2)-ethylidene-dG content increased to 21.4 ± 7.50, 10.5 ± 3.61, and 9.83 ± 3.90 adducts per 10(7) bases at each postexposure time. The half-life of this adduct was calculated as ～35 h in independent experiments. These results indicate that AA exposure from daily alcohol consumption may cause DNA damage and may increase the risk of alcohol-related carcinogenesis.     

The effects of putative DNA repair inhibitors on DNA adduct levels and unscheduled DNA synthesis in rat hepatocytes exposed to 2-acetylaminofluorene

The enzymology of DNA repair is currently under active investigation. The purpose of the present study was to examine the involvement of a number of enzymes (DNA polymerase alpha and beta, DNA topoisomerase II and ribonucleotide reductase) in the repair of chemically induced DNA damage in a mammalian cell system. This was done by studying the effects of inhibitors of these enzymes on the levels of 2-acetylaminofluorene (2-AAF)-DNA adducts and on the induction of UDS in primary cultures of rat hepatocytes exposed to the carcinogen in vitro. The results obtained with aphidicolin (an inhibitor of DNA polymerase alpha) show that the binding of 2-AAF to cellular DNA was significantly higher in samples exposed to this compound. Moreover, induction of UDS by 2-AAF was completely blocked in the presence of this compound. Dideoxythymidine, a DNA polymerase beta inhibitor, led to complex results. It produced a reduced DNA-specific activity due to [3H]2-AAF adduct formation as well as a diminished but still detectable UDS response in the presence of 2-AAF. Inhibitors of DNA topoisomerase II (nalidixic acid) and ribonucleotide reductase (hydroxyurea) did not cause any statistically significant change in the accumulation of 2-AAF adducts nor did they affect the induction of UDS. The data clearly suggest that DNA polymerase alpha participates in the repair of 2-AAF adducts in hepatocytes. In addition, neither DNA topoisomerase II activity, nor limitations in the precursor nucleotide pools appear to be critical factors in this process.     

Anatomical variations of the abducent nerve in humans

Anatomical variation of the nervus abducens in human encephali were found and described. They consisted of (1) an unusual trifurcation of the abducent nerve, limited to the extradural portion of the neural trunk (1.4% of the cases) and (2) the duplicity (11.1%) of the neural trunk, starting before reaching the orbit and ending before reaching the m. rectus lateralis. The possibility of correlating these variations with clinical aspects and forensic interpretations is mentioned.     

DNA adduct formation by alachlor metabolites

The extent of DNA adduct formation by alachlor [ArN(CH2OCH3)C(O)CH2Cl wherein Ar is 2,6-diethylphenyl] and its metabolites is used as a guide to deduce the causal agent(s) in the carcinogenicity of this major herbicide. [14C-phenyl]Alachlor is compared to its two metabolic cleavage products, [14C-phenyl]2-chloro-N-(2,6-diethylphenyl)acetamide (CDEPA) [ArNHC(O)CH2Cl] and [14C-phenyl]2,6-diethylaniline (DEA) (ArNH2), and to [14C-methoxy]alachlor in various in vitro and in vivo systems. Horseradish peroxidase and hydrogen peroxide activate DEA, but not CDEPA or alachlor, for formation of adducts with calf thymus DNA, which probably involves 2,6-diethylnitrosobenzene (ArNO) as an intermediate. Mouse liver microsomes and NADPH are both required to enhance the binding from each labeled preparation to calf thymus DNA; 4-fold higher labeling is observed from [14C-methoxy]- than from [14C-phenyl]alachlor. This 4-fold preferential DNA labeling from the 14C-methoxy compound is likewise found in the liver of mice treated intraperitoneally. Mouse liver protein and hemoglobin are also labeled, in vivo, with [14C-phenyl]alachlor, -CDEPA and -DEA, and, as with the DNA, the labeling of these proteins is 1.5- to 2-fold higher with [14C-methoxy]alachlor. Metabolic studies indicate that ArN(CH2OCH2OH)C(O)CH2Cl is an intermediate in forming CDEPA and presumably formaldehyde in the mouse liver microsomal mixed-function oxidase system and in yielding the O-glucuronide of ArN(CH2OH)C(O)CH2Cl in the urine of alachlor-treated mice. These findings point to the N-CH2OCH2OH metabolite or formaldehyde as a reactive intermediate in forming a DNA-adduct and as a candidate proximate carcinogen.     

Relationship between DNA adduct levels, repair enzyme, and apoptosis as a function of DNA methylation by azoxymethane.

DNA alkylating agent exposure results in the formation of a number of DNA adducts, with O6-methyl-deoxyguanosine (O6-medG) being the major mutagenic and cytotoxic DNA lesion. Critical to the prevention of colon cancer is the removal of O6-medG DNA adducts, either through repair, for example, by O6-alkylguanine-DNA alkyltransferase (ATase) or targeted apoptosis. We report how rat colonocytes respond to administration of azoxymethane (a well-characterized experimental colon carcinogen and DNA-methylating agent) in terms of O6-medG DNA adduct formation and adduct removal by ATase and apoptosis. Our results are: (a) DNA damage is greater in actively proliferating cells than in the differentiated cell compartment; (b) expression of the DNA repair enzyme ATase was not targeted to the proliferating cells or stem cells but rather is confined primarily to the upper portion of the crypt; (c) apoptosis is primarily targeted to the stem cell and proliferative compartments; and (d) the increase in DNA repair enzyme expression over time in the bottom one-third of the crypt corresponds with the decrease in apoptosis in this same crypt region.     

Association of tocopherols with circulating autoantibody levels against an oxidized DNA nucleoside in humans

Autoantibodies against oxidized DNA bases are found in vivo and have been used as an indicator of oxidative damage, yet little is known concerning their individual variation and relation to serum micronutrients. Human plasma anti-5-hydroxymethyl-2'-deoxyuridine (HMdU) autoantibody (aAb) levels were repeatedly determined in 41 women and 11 men, and found to have small within-individual variation over time, but large between-individual differences. A positive association in both women (r = .5762, p = .0001) and men (r = .415, p = .2) between plasma total tocopherols and antibody levels was observed. Autoantibody levels were lower in postmenopausal women (8.37 +/- 1.61 vs. 17.18 +/- 2.85 in premenopausal women, p < .01), independently of plasma tocopherol. However, aAb titers in postmenopausal women were still significantly associated with plasma tocopherol levels and adjustment for menopausal status in women yielded a highly significant correlation between HMdU aAb levels and total tocopherol (r = .7342, p = .0001). Plasma malondialdehyde equivalents (MDA), a measure of lipid peroxidation, were also higher in individuals with either high plasma alpha-tocopherol or high beta+gamma-tocopherol levels. The positive association of tocopherols with markers of oxidative damage may reflect a response to the generation of endogenous oxidants associated with enhanced immune function. The decrease in aAb level in postmenopausal women may similarly reflect decreased immune function associated with decreased estrogen levels.     

Hypoxia and oxidation levels of DNA and lipids in humans and animal experimental models

The objective of this review was to evaluate the association between hypoxia and oxidative damage to DNA and lipids. Evaluation criteria encompassed specificity and validation status of the biomarkers, study design, strength of the association, dose-response relationship, biological plausibility, analogous exposures, and effect modification by intervention. The collective interpretation indicates persuasive evidence from the studies in humans for an association between hypoxia and elevated levels of oxidative damage to DNA and lipids. The levels of oxidatively generated DNA lesions and lipid peroxidation products depend on both the duration and severity of the exposure to hypoxia. Largest effects are observed with exposure to hypoxia at high altitude, but other factors, including ultraviolet light, exercise, exertion, and low intake of antioxidants, might contribute to the effect observed in subjects at high altitude. Most of the animal experimental models should be interpreted with caution because the assays for assessment of lipid peroxidation products have suboptimal validity.     

Comparison of DNA adduct levels associated with exogenous and endogenous exposures in human pancreas in relation to metabolic genotype

Recently, we examined normal human pancreas tissue for DNA adducts derived from either exogenous chemical exposure and/or endogenous agents. In an effort to explain the different types and levels of DNA adducts formed in the context of individual susceptibility to cancer, we have focused on gene-environment interactions. Here, we report on the levels of hydrophobic aromatic amines (AAs), specifically those derived from 4-aminobiphenyl (ABP), and DNA adducts associated with oxidative stress in human pancreas. Although these adducts have been reported in several human tissues by different laboratories, a comparison of the levels of these adducts in the same tissue samples has not been performed. Using the same DNA, the genotypes were determined for N-acetyltransferase 1 (NAT1), the glutathione S-transferase (GST) M1, GSTP1, GSTT1, and NAD(P)H quinone reductase-1 (NQO1) as possible modulators of adduct levels because their gene products are involved in the detoxification of AAs, lipid peroxidation products and in redox cycling. These results indicate that ABP-DNA adducts, malondialdehyde-DNA adducts, and 8-oxo-2'-deoxyguanosine (8-oxo-dG) adducts are present at similar levels. Of the metabolic genotypes examined, the presence of ABP-DNA adducts was strongly associated with the putative slow NAT1*4/*4 genotype, suggesting a role for this pathway in ABP detoxification.     

Compositional variations in DNA sequences

Biologically occurring nucleotide sequences differ from randomlygenerated ones. Here we describe general patterns found in prokaryoticand in eukaryotic DNA. In the accompanying paper (Nussinov,1991) we also describe DNA signals recognized by their correspondingprotein factors. In particular, we focus on modes of searchesfor such patterns and signals and on the potential propertiessuch sequences may possess.     

Moderate variations in CDC25B protein levels modulate the response to DNA damaging agents

CDC25B, one of the three members of the CDC25 dual-specificity phosphatase family, plays a critical role in the control of the cell cycle and in the checkpoint response to DNA damage. CDC25B is responsible for the initial dephosphorylation and activation of the cyclin-dependent kinases, thus initiating the train of events leading to entry into mitosis.1 The critical role played by CDC25B is illustrated by the fact that it is specifically required for checkpoint recovery2, 3 and that unscheduled accumulation of CDC25B is responsible for illegitimate entry into mitosis.3-5 Here, we report that in p53-/- colon carcinoma cells, a moderate increase in the CDC25B level is sufficient to impair the DNA damage checkpoint, to increase spontaneous mutagenesis, and to sensitize cells to ionising radiation and genotoxic agents. Using a tumour cell spheroid assay as an alternative to animal studies, we demonstrate that the level of CDC25B expression modulates growth inhibition and apoptotic death. Since CDC25B overexpression has been observed in a significant number of human cancers, including colon carcinoma, and is often associated with high grade tumours and poor prognosis1, our work suggests that the expression level of CDC25B might be a potential key parameter of the cellular response to cancer therapy.     

Mutagenic specificity of 2-acetylaminonaphthalene-derived DNA adduct in mammalian cells

2-Acetylaminonaphthalene (2-AAN) has been recognized as a urinary bladder carcinogen in humans. The deacetylated form, 2-aminonaphthalene (2-AN), is metabolized in vivo and reacts primarily with guanine residues in DNA, resulting in the formation of dG-N(2)-aminonaphthalene (dG-N(2)-AN) adduct. Phosphoramidite chemical procedure has recently been established in our laboratory to prepare oligodeoxynucleotides containing a single dG-N(2)-acetylaminonaphthalene (dG-N(2)-AAN) adduct. Oligodeoxynucleotides ((5')TCCTCCTNXCCTCTC, where X is dG or dG-N(2)-AAN and N is C, A, T or G) with different bases 5' flanking to the lesion were prepared and were inserted into a single-strand shuttle vectors and used to establish the mutational frequency and specificity of dG-N(2)-AAN adduct in simian kidney cells. dG-N(2)-AAN adduct promoted preferential incorporation of dCMP, the correct base, opposite the lesion. When the 5' flanking base to the lesion was C, A or T, the mutational frequency was under 2.1%. When G flanked to the lesion, the mutational frequency was slightly increased to 4.2%. Misincorporation of dAMP, dTMP, and/or dGMP varied depending on the 5' flanking base. When dG-N(2)-AAN was positioned at codon 61 of noncoding strand of human c-Ha-ras1 gene ((5')TCCTCCTXGCCTCTC, where X is dG-N(2)-AAN), the mutational frequency was 6.7%; G-->T transversions (4.7%), followed by G-->A transition (2.0%), were observed. These results demonstrated that dG-N(2)-AAN is a weak mutagenic lesion in mammalian cells. The influence of 5' flanking sequence context was observed on the mutational frequency and specificity of this adduct.     

Escherichia coli responses to a single DNA adduct

To study the mechanisms by which Escherichia coli modulates the genotoxic effects of DNA damage, a novel system has been developed which permits quantitative measurements of various E. coli pathways involved in mutagenesis and DNA repair. Events measured include fidelity and efficiency of translesion DNA synthesis, excision repair, and recombination repair. Our strategy involves heteroduplex plasmid DNA bearing a single site-specific DNA adduct and several mismatched regions. The plasmid replicates in a mismatch repair-deficient host with the mismatches serving as strand-specific markers. Analysis of progeny plasmid DNA for linkage of the strand-specific markers identifies the pathway from which the plasmid is derived. Using this approach, a single 1, N(6)-ethenodeoxyadenosine adduct was shown to be repaired inefficiently by excision repair, to inhibit DNA synthesis by approximately 80 to 90%, and to direct the incorporation of correct dTMP opposite this adduct. This approach is especially useful in analyzing the damage avoidance-tolerance mechanisms. Our results also show that (i) progeny derived from the damage avoidance-tolerance pathway(s) accounts for more than 15% of all progeny; (ii) this pathway(s) requires functional recA, recF, recO, and recR genes, suggesting the mechanism to be daughter strand gap repair; (iii) the ruvABC genes or the recG gene is also required; and (iv) the RecG pathway appears to be more active than the RuvABC pathway. Based on these results, the mechanism of the damage avoidance-tolerance pathway is discussed.     

DNA damage and DNA adduct formation in rat tissues following oral administration of acrylamide

Acrylamide is present as a contaminant in the human diet in heated food products. It has been found to be carcinogenic in laboratory rats and has been classified as probably carcinogenic in humans. In order to clarify the possible involvement of a primary genotoxic mechanism in acrylamide-induced carcinogenicity, both the presence of DNA damage, measured by the comet assay, and the formation of N7-(2-carbamoyl-2-hydroxyethyl)guanine (N7-GA-Gua) and N3-(2-carbamoyl-2-hydroxyethyl)adenine (N3-GA-Ade), derived from reaction of the active metabolite glycidamide (GA) with the DNA, analyzed by LC/MS/MS, were assessed in selected rat tissues. Rats were administered with single oral doses of acrylamide (18, 36 or 54 mg/kg body weight (b.w.) and the organs (blood leukocytes, brain, bone marrow, liver, testes and adrenals) were sampled at different times after treatment. Results from GA-induced DNA adduct measurements indicated a relatively even organ distribution of the adducts in brain, testes and liver. Organ-specificity in acrylamide carcinogenesis can therefore not be explained by a selective accumulation of GA-DNA adducts in the target organs, at least not after a single dose exposure. The DNA adduct profiles and half-lives were similar in the different organs; except that the N3-GA-Ade adduct was more rapidly removed from tissues than the N7-GA-Gua adduct. Increased extent of DNA migration, as measured by the in vivo rat comet assay, was found in brain and testes, and these specific results seem to be in accordance with the known organ-specificity in acrylamide carcinogenesis in rat. Only weak and transient DNA damage was recorded in the liver, bone marrow and adrenals. The DNA-damaging effect of the compound observed in the blood leukocytes could be a simple biomarker of acrylamide exposure and genotoxicity.     

Tissue dependent variations of DNA methylation and endoreduplication levels during tomato fruit development and ripening

Tomato fruit cells are characterized by a strong increase in nuclear ploidy during fruit development. Average ploidy levels increased to similar levels (above 50C) in two distinct fruit tissues, pericarp and locular tissue. However, ploidy profiles differed significantly between these two tissues suggesting a tissue-specific control of endoreduplication in tomato fruit. To determine possible relationships between endoreduplication and epigenetic mechanisms, the methylation status of genomic DNA from pericarp and locular tissue of tomato fruit was analysed. Pericarp genomic DNA was characterized by an increase of CG and/or CNG methylation at the 5S and 18S rDNA loci and at gyspsy-like retrotransposon sequences during fruit growth. A sharp decrease of the global DNA methylation level together with a reduction of methylation at the rDNA loci was also observed in pericarp during fruit ripening. Inversely, no major variation of DNA methylation either global or locus-specific, was observed in locular tissue. Thus, tissue-specific variations of DNA methylation are unlikely to be triggered by the induction of endoreduplication in fruit tissues, but may reflect tissue-specific ploidy profiles. Expression analysis of eight putative tomato DNA methyltransferases encoding genes showed that one chromomethylase (CMT) and two rearranged methyltransferases (DRMs) are preferentially expressed in the pericarp during fruit growth and could be involved in the locus-specific increase of methylation observed at this developmental phase in the pericarp.     

Diurnal variations of serum GM-CSF levels

We aimed to investigate the daily variations of serum granulocyte-macrophage colony-stimulating factor (GM-CSF) levels and to correlate them with peripheral blood cells counts. Venous blood samples from eleven healthy volunteers were taken four times a day, being at 08:00, 14:00, 20:00 and 02:00h and serum GM-CSF levels measured by ELISA. We could not find a significant overall difference among GM-CSF levels at four different times of the day using the Friedman test. On the other hand, serum GM-CSF levels at night (20:00h) were found to be significantly increased when compared to the morning levels (08:00h) using the Wilcoxon test (P=0. 022). The levels of lymphocytes and white blood cells (WBCs) at 20:00h were also higher than the morning levels (08:00h) as expected. While there was a strong relationship between the morning levels of GM-CSF (08:00h) and all measurements of peripheral blood cells during the day, the levels of GM-CSF measured at 02:00, 14:00 and 20:00h were found to be significantly correlated with only the WBC levels. It was concluded that there may be a significant difference between morning and night levels of GM-CSF and morning levels of GM-CSF may be more important in the regulation of WBC counts during the day. These variations warrant further studies about diurnal rhythms of haematopoiesis chronotherapy with CSFs.