Oxidative stress and chronic diseases : exocyclic dna adducts as markers for disrupted genomic integrity and risk

To study whether lesions other than O-alkyl-pyrimidine-adducts or DNA cross-links, two action principles which display mutagenic activity at both post- and pre-meiotic germ cell stages, are responsible for the pre-meiotic activity of vinyl carbamate (Vcar) we studied the nature of DNA sequence changes induced by Vcar in these cell stages. The most important difference with mutations induced in post-meiotic germ cells is the absence of multi locus deletions in pre-meiotic germ cells. In contrast to the Vcar molecular spectrum, chlorambucil, a cross-linking agent, still induced MLD in pre-meiotic male germ cells. This suggests that etheno-adducts may be responsible for the pre-meiotic activity of Vcar and adds another action principle to the above mentioned group of lesions which generate heritable genetic damage in pre-meiotic cells, most probably related to high risk principles for cancer induction. The preference for mutations at AT positions by Vcar nicely fits the relatively high induction of gamma-dA. The adduct distribution after VBr exposure is more random which also correlates with the different types of base-pair substitutions of the molecular spectrum.

Tumors and normal tissues from 50 colon cancer patients were collected in Norway (Telemark Central Hospital, Skien, Norway). Disturbances in either stability of microsatellites (MSI) or mismatch repair (MMR) proteins expression were noted in tumor tissue of 6 out of 50 patients. Lack of MMR proteins expression was observed in tumor tissue of only 5 patients, and was exclusively connected with MLH1 protein. Only in one case microsatellite instability was connected to MMR dysfunction. In the other 4 cases there was no correlation between microsatellite instability and expression of MMR proteins, suggesting that other factors than MMR can also contribute to genomic instability, and that MMR deficiency is not a necessary prerequisite for microsatellite instability. This is exemplified by two patients with no MLH1 expression, which was not accompanied by MSI. Other factors contributing to microsatellite instability and the rate of cancer progression should be identified.

Lipid peroxidation (LPO) generates a large variety of carbonyl compounds, including hydroxyalkenals, which form a variety of exocyclic adducts to DNA bases. The major hydroxyalkenal is trans-4-hydroxynonenal (HNE), which reacts with all four deoxynucleosides forming hexyl-substituted propano-adducts and heptyl-substituted etheno-adducts. HNE interacts with DNA in a sequence-dependent manner, and DNA synthesis by T7 DNA polymerase is inhibited at G>C>A and T sites on DNA pretreated with HNE. HNE triggers base substitutions (mainly C®T transitions), frameshift mutations and recombination in M13 phage model system. Two repair systems were found to eliminate HNE adducts from DNA in E.coli. Nucleotide excision repair (NER) system is excising HNE-adducts from double stranded DNA. However, when present in single-stranded DNA or in NER-deficient E.coli, recombination is the major repair pathway. Further research is necessary to investigate the role of LPO in tissue abnormalities in individuals with affected repair systems - NER and recombination, which occurs in several human diseases, e.g. Xeroderma pigmentosum, breast cancer.

We studied the excision by E.coli Mug and AlkA DNA-glycosylases of the simplest types of 3,N4-exocyclic cytosine adducts, i. e. a-hydroxyethano (HEC), etheno (aaC), a-hydroxypropano (HPC) and a-hydroxy-aa-methyl-propano (mHPC) adducts. HEC is a relatively stable adduct of chloroacetaldehyde, a vinyl chloride metabolite, which spontaneously undergoes dehydration to aaC. HPC and mHPC are adducts of acrolein and crotonaldehyde, respectively. Mug glycosylase, in addition to the known removal of aaC, excises all studied adducts albeit with efficiency decreasing in order aaC >HEC>HPC>>mHPC. AlkA glycosylase excises only HPC with very low efficiency. HPC is excised much more efficiently by Mug than by AlkA. The very low efficiency of excision of mHPC, which is bigger only by one methyl group than HPC, leads to the conclusion that probably Mug glycosylase would not excise the bulky adducts formed by hydroxynonenal or its epoxide, the main lipid peroxidation products.The results make an input into understanding of the mechanisms of elimination of DNA damage triggered by lipid peroxidation, which is accompanying inflammations and infections.

Buffy coat and serum samples were collected from premenopausal adult female volunteers. Samples were analyzed for etheno-DNA adducts by immunoaffinity-32P-postlabelling. 1,N6-ethenodeoxyadenosine (etheno-dA) and 3,N4-etheno-deoxycytidine (etheno-dC) levels were highly correlated, showing large inter-individual variation. A GC-MS method was established for the analysis of metabolites of estradiol (E2), 2- and 4- hydroxy-E2 and 2- and 4-methoxy-E2 in serum of the same subjects. The data indicated that the level of 4-hydoxy-E2, the major catechol estrogen implicated in redox-cycling was highly variable. These results support that metabolic redox-cycling of 4-hydrox-estradiol leads to ROS and DNA damage which may be involved in breast cancer etiology.

The effect of XPG deficiency on induction of Ring X-chromosomal loss is higher than on homologous mitotic recombination in somatic cells of Drosophila exposed to vinyl carbamate or ethyl carbamate. The somatic mutation and recombination test (SMART) was used to study the effect of DmXPG, DmXPC or DmRAD54 deficiency on induction of homologous mitotic recombination (HMR) after larvae exposure to Vcar. In all three genotypes a clear enhancement in eye spot induction was observed compared to the homo- and heterozygous wild type genotypes but the hypermutability effects were mainly seen at the higher dose levels. For mutagens like MMS, ENU, MEC and UV-B, which represent different action principles, similar studies were performed with the XPG mutant. All agents, except of ENU, showed a hypermutability effect in the XPG deficient genotype which was larger than seen for Vcar. The highest effect was observed for UV-B light the model mutagen when studying NER mechanisms. In a modified version of the SMART assay HMR can be studied in parallel with loss of a Ring X-chromosome (clastogenic event). In contrast to MMS, MEC and UV-B, vinyl carbamate and ethyl carbamate showed a much stronger XPG effect on clastogenicity than on HMR. This suggests that lesions causing chromosomal loss are more efficiently removed by the NER mechanism than lesions leading to HMR.

Lack of alkylpurine-DNA-N-glycosylase in mice does not confer increased susceptibility to vinyl carbamate induced hepatocarcinogenesis. In collaboration with Partner 2 (A. Barbin), we first confirmed that 1,N6-ethenoadenine (etheno-A), but not 3,N4-ethenocytosine (etheno-C), is released by alkylpurine-DNA-N-glycosylase (APNG) in vivo. Thus, 6h after the last dose of vinyl carbamate (5 x 250nmol/g), levels of etheno-A were 1.6-fold higher in hepatic DNA from APNG-/- mice and subsequently persisted at higher levels for longer than in DNA from wild-type animals. In contrast, ~14-fold lower levels of etheno-C were induced by Vcar and the kinetics of formation and persistence of etheno-C was similar in the two strains. The carcinogenicity of Vcar was compared in wild-type and APNG-/- suckling mice given a single dose of Vcar (30 or 150nmol/g). After one year, only mice in the high dose group developed hepatocellular carcinoma, however, the incidence was not higher in APNG-/- mice. Although higher levels and increased persistence of etheno-A was observed in hepatic DNA from APNG-/- mice at 150nmol/g Vcar, apoptosis and cell proliferation levels were similar in both strains of mice. This may explain why differences in etheno-A formation / persistence observed here did not result in higher susceptibility of APNG-/- mice to hepato-carcinogenesis.

1,N6-ethenoadenine (etheno-A) is a major promutagenic DNA adduct formed by vinyl carbamate in mice. The dosimetry data obtained could be useful for risk assessment in humans (exposure to urethane present in certain fermented food and beverages). A preliminary QSAR between level of etheno-A in hepatic DNA and carcinogenicity in rodents of five etheno adduct-forming chemicals has been provided. Young animals are more susceptible to DNA damage and tumor development than adult animals. Etheno DNA adducts are repaired through at least two pathways in mice: base excision repair and mismatch repair. This points to potential susceptibility factors in humans such as polymorphisms and mutations in DNA repair genes that affect protein activity. It would be of interest to examine the relationships between polymorphisms/mutations in these genes, levels of etheno DNA adducts and cancer risk. Chemicals that produce etheno DNA adducts also affect cell cycle (apoptosis, cell proliferation) in mice. These biological effects are chemical-, age- and dose-dependent. They are major factors to be considered in risk assessment for this class of chemicals.

We have demonstrated, for the first time, that two structurally unrelated proteins, the Escherichia coli mismatch-specific uracil-DNA glycosylase (MUG) and the human alkylpurine-DNA-N glycosylase (ANPG) can excise 1,N2-ethenoguanine, a major product formed in DNA exposed to lipid peroxidation-derived aldehydes in vitro. The comparison of the kinetic constants for the excision of the adduct indicates that the bacterial enzyme is the most efficient and that the efficiency of repair by the ANPG proteins is low. Using cell-free extracts from genetically modified organisms we further substantiated the role of MUG and ANPG proteins in the repair of 1,N2-ethenoguanine. Moreover, it was shown that the nonconserved, N-terminal 80 amino acids of the ANPG protein are essential for 1,N2-aaG-glycosylase activity, in contrast to other known substrates of this enzyme. In conclusion, the results obtained in this study provide new insights into the molecular mechanisms involved in the repair of an oxidant-derived modified base. Such studies show how a cell reduces the mutagenic load and may ultimately aid the design of new pharmacological drugs.

The aim of this work was to generate single chain antibody fragments (scFv) that specifically recognise etheno adducts (1,N6-ethenoadenine, 3,N4-ethenocytosine, 1,N2-ethenoguanine) in DNA. To generate such clones, we used the Griffin (6.5x109clones) and Nissim (1x108 clones) semi-synthetic libraries, which were provided by the MRC Centre for Protein Engineering, Cambridge, UK. Clones were selected by covalently linking single-stranded oligonucleotides containing the etheno adduct flanked by two random bases to ELISA plates and incubation with the entire phage library. Specific scFvs were obtained by 3 rounds of enrichment, performed with increasing stringency. After the final round of enrichment, individual clones were induced to produce soluble scFv and these were retested by ELISA. This procedure produced two eA specific clones, one eC specific clone and one clone that recognises eA, eC and eG-containing oligonucleotides. While the two latter scFvs were shown to specifically inhibit cleavage of oligonucleotides containing the appropriate etheno adduct when incubated with cell-free extracts, interestingly no such inhibition was observed with the two eA specific clones. Following these analyses of oligonucleotide binding, the effectiveness of the scFvs in detecting etheno adducts in genomic DNA will be assessed.

Sites of HNE-induced lesions within exons 5-8 of p53 gene were studied by DNA polymerase fingerprint analysis, and were found mainly between codons 161-301 that is at sites frequently mutated in colon cancer as well as in all cancers. Interestingly at sites not mutated, or scarcely mutated, in colon cancer no HNE-induced modifications were found. HNE-induced modifications were identified at two hot-spot codons, namely: at codon 248 (CGG), C residue was modified, and at codon 273 (TGT) G was modified. Preference for cytosine and guanine modification by HNE, as well as modification of C and G in mutational hot-spots, correlates with types of mutations GC->AT prevalent in colon cancer and can suggest that lipid peroxidation product - HNE can contribute to pathogenesis of colon cancer.

We studied degradation of 8-OH-dGTP and etheno derivatives - edCTP and edATP by homogenates from human tissues (tumour and healthy surrounding of the tumour tissue from patients suffering from lung cancer) and we have observed the sequential formation of dNDP, dNMP and finally dN. In all tested patients the rate of dephosphorylation of modified dNTPs by tumour tissue homogenates was higher than by healthy tissue ones whereas the rate of dephosphorylation of unmodified dNTP by both types of homogenates was similar. This might be due to increased expression of proteins sanitising nucleotide pool in tumor tissue. We have observed that edCTP is dephosphorylated much more efficiently than any other modified and unmodified dNTPs studied with fast formation of the final product, edC. We suppose that in human cells some enzymatic activities exist, specifically dephosphorylating edCTP to give adducted nucleoside which can be efficiently eliminated from the cell.

The impact of nucleotide excision repair deficiency on induction of gene mutations by vinyl carbamate is significantly higher for damage induced in post-meiotic male germ cells than for pre-meiotic cells. Continuous repair of DNA damage in Drosophila male germ cells takes place until the formation of mid-spermatids and further in the egg after fertilization. We have shown that the effect of DmXPC, DmXPF and DmXPG mutant females on mutation induction after exposure of post-meiotic male germ cells to vinyl carbamate is identical and a factor 3 to 5 higher compared to wild type females. The paternal nucleotide excision repair (NER) effect on DNA lesions induced in repair deficient pre-meiotic cells by Vcar is smaller and only visible at the extreme, almost toxic, high dose levels. Effect of female NER deficiency on mutation induction was a more general phenomenon, and can only be explained when Vcar induces persistent pre-mutagenic lesions. The overall data suggest that NER is not the major DNA repair pathway in removal of mutagenic etheno DNA adducts.

Oxidative DNA damage have been postulated to be major type of endogenous damage leading to human degenerative disorders including cancer, cardiovascular disease and brain dysfunction. Oxygen radicals generate mostly non bulky DNA lesions which are substrates for base excision repair (BER) pathway. However, despite the progress in understanding BER it is still unclear why mice deficient in DNA glycosylases that remove oxidized bases are not sensitive to oxidizing agents. The BER system presents serious drawback, it generates genotoxic intermediates (abasic sites and/or blocking 3'-termini groups) that must be eliminated by additional steps before starting DNA repair synthesis. Therefore, we searched for an alternative, back-up, repair pathway. Here, we show that Nfo-like endonucleases nick DNA on the 5' side of various oxidatively damaged bases, generating 3'-hydroxyl (OH) and 5'-phosphate (P) termini, next to all the modified bases tested, including alpha-2'-deoxyadenosine, alpha-thymidine, 5,6-dihydrothymine (DHT), 5,6-dihydrouracil (DHU), 5-hydroxyuracil (5ohU) and 2,6-diamino-4-hydroxy-5-N-methylformamidopyrimidine (Fapy) residues. The former provides the proper end for DNA repair synthesis. The dangling damaged nucleotide on the 5'-side is then a good substrate for human flap-structure endonuclease (FEN-1) and for DNA polymerase I (pol I) of E.coli. The DNA glycosylase-independent incision of oxidatively damaged DNA by Nfo/Apn1-like enzymes provides an alternative pathway to classic BER. We propose to name it the nucleotide incision repair (NIR) pathway. The main advantages of NIR pathway are to: - Avoid the genotoxic intermediates generated in the BER pathway; - Provide an explanation for the DNA repair-proficiency of DNA glycosylase-deficient mutants and supports the existence of a back-up repair pathway in E. coli, yeast and humans; - Provide a new physiological target, a modified base rather than an artificial reduced a basic site, for the long-patch repair pathway described in human cells. The identification of this new general activity of the Nfo protein family could be of interest, since: - It is evolutionary conserved from E. coli to human; - By its mechanistic characteristics since it generates directly proper ends for DNA replication on one side and on the other side for elimination of the lesion by specific nucleases; - This new pathway has implications in the fields of DNA replication, DNA repair, pharmacology: it is a new target for specific drugs, in radiobiology and oncology, it opens the possibility to identify a new tumour-related genes.

Tumor and normal (with no histological changes) colon tissues from colon cancer patients were collected in Norway (Telemark Central Hospital, Skien, Norway) and in Poland (Central Clinical Hospital of Warsaw Medical University, Warsaw, Poland). High individual variations in enzymatic activities eliminating from DNA lipid peroxidation-derived etheno-DNA adducts were observed. A positive linear association in etheno adducts repair activities were found between normal and tumor colon tissues. High percentage (10-30%) of individuals with extremely low or no activity for excision of etheno-A or etheno-C in colon tissues were found among colon cancer patients. Repair activity for both eA and eC had a tendency to increase in tumor in comparison to normal tissue. However, due to high individual variations and a small number of cases these differences were statistically significant only for etheno-A-glycosylase in polish population. No differences in etheno-A- and etheno-C-repair activities were found in relation to age, gender, smoking habit, and the stage of tumor development. Further research on repair activities in blood leukocytes of cancer patients and healthy controls, will enable to find the possible relation between individual susceptibility to colon cancer and repair efficiency of oxidative DNA damage.

The aim of this work was to elucidate mechanisms of repair of eA in either a transcriptionally active or inactive context in embryonic fibroblasts from normal, or alkylpurine-DNA-N-glycosylase deficient mice. Non-replicating variants of a shuttle vector (pS189) containing a single eA adduct were constructed: pSDori for the study of transcription-coupled repair (deletion of SVori) and pSDT3 for transcription independent repair (both pSVori and promoter regions deleted). However, although pSDori and pSDT3 were successfully transfected into APNG proficient and deficient cells, highly variable recovery of the vector prevented us from using this method to measure eA repair in vivo. However, we were able to show that eA does not block RNA polymerase II directed transcription: RNA was recovered from pSDori and pSDT3 transfected cells and RT-PCR carried out using primers designed to amplify a region, either spanning (251bp), or upstream (300bp) of the eA adduct present in the shuttle vector. As expected, no specific signal was obtained from cells transfected with pSDT3, which lacks the TAg promoter. However, RT-PCR fragments were obtained from both APNG proficient and deficient cells transfected with the transcription competent pSDori. Thus, we can conclude that eA does not block RNA Pol II dependent transcription.

We investigated the genetic activity profile (GAP) of etheno-adduct forming chemicals by estimating forward mutation spectra, structural chromosome aberrations and mitotic recombination, utilizing germ cells and somatic cells of Drosophila melanogaster as in vivo targets. These genetic endpoints were correlated with adduct levels of etheno-dA and etheno-dC in DNA of larvae tissue measured by the immunoaffinity- 32P-postlabelling method. Etheno DNA-adduct forming compounds are bifunctional by definition, and therefore, formation of DNA cross-links and contribution of these adducts to mutation induction cannot be excluded. However, GAP data of bifunctional agents, known to form cross-links within the DNA, revealed that etheno-adducts and not DNA cross-links are involved in mutation induction by e-adduct forming chemicals. Furthermore, induction of gene mutations by vinyl carbamate seems to be associated with the formation of etheno-dA predominantly, while for vinyl bromide etheno-dC and etheno-dG might contribute as well. Nucleotide excision repair deficiencies have only a minor effect on germ cell mutations and induction of homologous mitotic recombination while the effect on chromosomal aberrations in somatic cells is more pronounced.

Trans-4-hydroxynonenal has been synthesized in the form of dimethylacetal according to Chandra and Srivastava (Lipids, (1997) 32: 779-782). The reaction of four deoxynucleosides with trans-4-hydroxynonenal (HNE, mutagenic product of lipid peroxidation) was studied. HPLC analysis of reaction mixtures showed the formation of several products in each case, and reactivity followed the order dG>dC>dT»dA. Mass spectrometry (MS) of the whole reaction mixtures showed peaks corresponding to HNE-dN 1:1 adducts, and also those corresponding to 2:1 and 3:1 adducts. In dA, dC and dG reactions peaks corresponding to heptyl-substituted etheno-adducts were detected, which indicates that during reaction HNE undergoes oxidation to 2,3-epoxide. The four most abundant products of the HNE-dC reaction mixture were isolated by TLC and tentatively characterised on the basis of MS, UV and pKa evaluation: A is N3-substituted HNE-dC cyclic or linear, B and C are N4-substituted HNE-dC, whereas D is a dehydrated heptyl-substituted etheno-dC adduct. Additionally, on the basis of NMR, C was identified as the product of addition of N4 of dC to C(3) of HNE with concomitant formation of cyclic hemiacetal in HNE part of the molecule. Identification of other products of reaction of HNE with dC is in progress.

In addition to hydroxy-nonenal (HNE), lipid peroxidation (LPO) yield malondialdehyde (MDA) that react with deoxyguanosine to form M1dG adduct. In order to compare the LPO induced DNA-adduct abundance by HNE and MDA in tissues we have developed an immuno-enriched 32P-postlabelling-HPLC method for M1dG. The method can be applied to small DNA samples for the measurement of etheno-DNA adducts and M1dG simultaneously. Total white blood cell DNA was analyzed for etheno-DNA and M1dG. M1dG adducts were one order magnitude higher than etheno DNA adducts. A significant positive correlation between etheno-DNA adducts and M1dG adducts was found in the DNA samples analysed. This ultrasensitive detection method is thus suitable for applications in human biomonitoring and molecular epidemiology studies.

Leistungen

Diese Seite teilen Diese Seite in sozialen Netzwerken teilen

Herunterladen Den Inhalt der Seite herunterladen