Effect of dna repair gene polymorphisms on BPDE-DNA adducts in human lymphocytes
Article first published online: 13 MAY 2002
Copyright © 2002 Wiley-Liss, Inc.
International Journal of Cancer
Volume 100, Issue 1, pages 9–13, 1 July 2002
How to Cite
Pastorelli, R., Cerri, A., Mezzetti, M., Consonni, E. and Airoldi, L. (2002), Effect of dna repair gene polymorphisms on BPDE-DNA adducts in human lymphocytes. Int. J. Cancer, 100: 9–13. doi: 10.1002/ijc.10463
- Issue published online: 29 MAY 2002
- Article first published online: 13 MAY 2002
- Manuscript Accepted: 8 APR 2002
- Manuscript Revised: 15 MAR 2002
- Manuscript Received: 21 DEC 2001
- XPD polymorphism;
- XRCC1 polymorphism;
- BPDE-DNA adducts;
- high-resolution gas chromatography-negative ion chemical ionization-mass spectrometry
To determine whether variations in DNA repair genes are related to host DNA damage, we investigated the association between polymorphism in the XPD gene (codon 199, 312, 751) and the XRCC1 gene (codon 194, 399) and the presence of benzo(a)pyrene diolepoxide adducts to lymphocyte DNA (BPDE-DNA) in a group of male patients with incident lung cancer, all current smokers. BPDE-DNA adducts were analyzed by high-resolution gas chromatography-negative ion chemical ionization-mass spectrometry. XPD and XRCC1 genotypes were identified by PCR-RFLP. XRCC1 and XPD genotypes did not affect the levels and proportion of detectable BPDE-DNA adducts. The patients were also genotyped for the GSTM1 polymorphism, given its role in the detoxification of BPDE. Individuals with the GSTM1 deletion had significantly higher levels of BPDE-DNA adducts when they were XPD-Asp312Asp+Lys751Lys than carriers of at least one variant allele. No such association was found with the XRCC1 genotypes. Because of the small study population (n = 60), further statistical analysis of possible gene-gene and gene-environment would not be informative. This is the first study analysing the specific BPDE-DNA adduct in vivo with regard to polymorphic repair genes (XPD, XRCC1) and xenobiotic metabolizing gene (GSTM1). Our results raise the possibility that the XPD-Asp312Asp+Lys751Lys genotype may increase BPDE-DNA damage; this effect might be evident in individuals who are especially likely to have accumulated damage, probably because of lower detoxification capacity and high environmental exposure. © 2002 Wiley-Liss, Inc.
The ability of benzo(a)pyrene (BP), a representative of the pulmonary carcinogens known as polycyclic aromatic hydrocarbons (PAH), to induce lung tumors is attributable to its ability to damage DNA. BP bioactivation in humans generates the highly toxic electrophilic metabolite (+)anti-benzo(a)pyrene-7,8-dihydrodiol-9,10 epoxide (BPDE), which can damage DNA by forming DNA adducts (BPDE-DNA) through covalent binding to the amino group of guanine or adenine.1, 2
An efficient network of complementary DNA repair mechanisms is required to prevent the detrimental consequences of DNA damage, but individuals differ in their DNA repair capacity, and this may be a risk factor for cancer.3 For example, DNA repair capacity was significantly lower in lung cancer cases than in healthy controls, the difference being especially pronounced among younger patients and smokers.4
Polymorphisms in several DNA repair genes involved in nucleotide excision repair (NER) and base excision repair pathways (BER) have been described, but their functional activity remains to be clearly determined. However, their impact in the human population may be important because of the high frequencies of allelic variants.5 In particular XPD and XRCC1 have been proposed as polymorphic genes that might modify the risk of lung cancer.6, 7, 8, 9 The XPD protein is involved in the NER, which recognizes and repairs bulky adducts and helix distortion lesions. XPD functions as an ATP-dependent 5′-3′ helicase joint to the basal transcription factor IIH complex (TFIIH).10 The XRCC1 protein acts as a scaffold protein in the BER, facilitating the repair by binding DNA ligase III at its carboxy and DNA polymerase β to its amino terminus.11
Three polymorphisms that induce amino acid change have been described in the XPD gene at codon 199 (IleMet), codon 312 (AspAsn) and codon 751 (LysGln),5 but there is still no direct biochemical evidence of their functional consequences and some divergent results have been reported on the role of the variant alleles in DNA repair.9, 12 Similarly, the functional consequences of the 3 polymorphisms in the XRCC1 gene that lead to amino acid substitution at codon 194 (ArgTrp), codon 280 (ArgHis) and codon 399 (ArgGln) are still unknown.
We previously reported measurements of the specific BPDE-DNA adduct in lymphocytes of lung cancer patients, showing that metabolic polymorphisms may modulate its presence.13 Our methodology is highly specific, allowing the detection of the chemical species (BPDE) responsible for the DNA adduct. Most likely, this specific BPDE-DNA adduct indicates cumulative exposure to BP not only after the action of metabolizing enzymes but also after the intervention of DNA repair enzymes. Therefore, the BPDE-DNA adduct might serve as a marker of cumulative unrepaired DNA damage and represents a possible DNA damage phenotype.14 Consequently, it might be useful for shedding light on the phenotypic effect of DNA repair polymorphisms on this specific adduct in vivo.
It has been shown that BPDE-DNA adducts are mainly repaired by NER in mammals,15 although some unstable BPDE adducts might promote depurination and elicit the BER pathway.15, 16 Thus we explored the possible association of polymorphisms in XPD and XRCC1 genes with the presence of BPDE-DNA adducts in lymphocytes from lung cancer patients, all active smokers, prototype of subjects heavily exposed to tobacco-derived BP.
Because part of the study population was previously genotyped for the glutathione S-transferase M1 (GSTM1) polymorphism13 and given the role of this enzyme in detoxifying BP metabolites, including BPDE, we also investigated the possible interaction of polymorphisms in XPD and XRCC1 genes with the GSTM1 polymorphism and BPDE-DNA adducts presence.
MATERIAL AND METHODS
Lymphocyte DNA from 60 Caucasian males with incident, histologically confirmed lung cancer was used for genotyping and BPDE-DNA adduct analysis. All patients were current smokers and had not had any pharmacologic treatment. They signed an informed consent form and provided 40 ml of blood. Table I summarizes selected characteristics of the subjects. Forty-three of the 60 patients have been used for previous metabolic genotyping and BPDE-DNA adduct measurements.13
|Characteristics||BPDE-DNA adducts/109 nucleotides mean ± SE|
|Overall subjects (n = 60)||2.26 ± 0.34|
|Smokers (n = 60)|
|Smokers of <30 cigarettes/day1||1.34 ± 0.28|
|Smokers of ≥30 cigarettes/day||3.15 ± 0.573|
|Smokers of <56.5 pack-years2||1.57 ± 0.36|
|Smokers of ≥56.5 pack-years||3.13 ± 0.634|
|Median age (62 years)|
|Subjects aged < 62 years||2.38 ± 0.52|
|Subjects aged ≥ 62 years||2.15 ± 0.46|
|Adenocarcinoma (39%)||1.75 ± 0.37|
|Squamous cell (27%)||2.17 ± 0.60|
|Small cell (7%)||4.95 ± 2.7|
|Large cell (5%)||4.36 ± 3.08|
|Other (22%)||2.10 ± 0.64|
|1 (36%)||2.35 ± 0.72|
|2 (35%)||2.92 ± 0.65|
|3 (25%)||0.89 ± 0.19|
|4 (4%)||1.69 ± 0.06|
Determination of BPDE-DNA Adducts
BPDE adducts were analyzed as benzo(a)pyrene tetrols (BPT) released from DNA (200–300 μg) after acid hydrolysis and quantitated by high-resolution gas chromatography-negative ion chemical ionization-mass spectrometry (HRGC-NlCl-MS) with selected ion recording after immunoaffinity purification to achieve high specificity and sensitivity.13 The detection limit was ≤1.4 adducts/109 nucleotides.
Individuals were screened for XPD polymorphisms at codons 199 (IleMet), 312 (AspAsn) and 751 (LysGln) using PCR-restriction fragment length polymorphism (PCR-RFLP) analysis as described by Lunn et al.12
Genotypic analysis for the XRCC1 polymorphisms at codons 194 (ArgTrp) and 399 (ArgGln) was done using a multiplex PCR assay, followed by restriction fragment length polymorphism (RFLP).17 The subjects were not genotyped at codon 280, due to the very low frequency of this polymorphism in the Caucasian population.5
Forty-three of the 60 patients have been previously genotyped for GSTM1 deletion. The GSTM1 genotyping for the remaining 17 patients was done as previously described.13
To compute statistics for adduct values, subjects with unmeasurable levels were considered as having half the minimum detectable value.
Group differences in BPDE-DNA adduct levels were tested using the Mann-Whitney 2-tailed U-test or Kruskal-Wallis test, as appropriate. Fisher's exact test was used to test the association between genotypes and adduct frequency dichotomized in undetectable and detectable levels.
The XPD and XRCC1 genotype distributions were in Hardy-Weinberg equilibrium. We found no variants of the XPD gene at codon 199 and no homozygotes for the XRCC1-194Trp variant allele. XRCC1-399Gln allele frequency (0.36) was consistent with previous studies.6, 7 In our population, the XRCC1-194Trp variant allele was rare (0.09). The XPD-751Gln variant allele occurred with a frequency of 0.47, which is slightly higher than that reported in lung cancer patients by Spitz et al.9 and by Butkiewicz et al.,6 but compatible with that reported by David-Beabes et al.18 The XPD-312Asn variant allele frequency (0.42) was similar to that of Butkiewicz et al.6 but higher than that given by Spitz et al.9 The polymorphisms in codons 312 and 751 of the XPD gene appeared to be in linkage disequilibrium in our study (χ2 34.85; p ≤ 0.001), consistent with recent reports.6, 9
Thirty-four patients (56.6%) were GSTM1 null genotype, which results from the total absence of the gene. The frequency of the homozygous deletion in this gene was in accordance with that reported for Caucasian lung cancer patients.19
The distribution of BPDE-DNA adducts was skewed (skewness 1.76; kurtosis 2.5) with only 35% of subjects (21/60) having detectable adducts. The average, in the whole study population, was 2.26 ± 0.34 adducts/109 nucleotides (mean ± SE).
Mean adduct levels were higher among those who smoked more than 30 cigarettes/day and reported a greater cumulative smoking history (≥56.5 pack-years), as shown in Table I. A significant doubling of detectable DNA adducts (51.85%) was observed in those who had smoked for ≥56.5 pack-years compared to those who had smoked for <56.5 pack-years (22.22%) (Fisher's exact test: p = 0.04). Age at diagnosis, histology and grade did not affect BPDE-DNA adduct distribution (Table I).
Figure 1 shows the frequency of detectable BPDE-DNA adducts in relation to each genotype. No evidence of association was found between the presence of BPDE-DNA adducts and the XRCC1-399 genotypes or XPD-312 and XPD-751 genotypes, even when heterozygous were combined with homozygous variant genotypes. Similarly, the levels of BPDE-DNA adducts were not significantly affected by XRCC and XPD genotypes (data not shown). Again, the difference in the distribution of BPDE-DNA adducts observed in subjects who had smoked for ≥56.5 pack-years compared to those who had smoked for <56.5 pack-years was not modulated by XRCC1 and XPD genotypes (data not shown).
As shown in Table II, we found no association between the GSTM1 polymorphism and adduct levels or proportion of detectable adducts. However, stratified analysis for the GSTM1 genotype indicated that the XPD polymorphisms had an effect on the presence of adducts, while the XRCC1 genotype did not. Individuals with the GSTM1 deletion had an average level of adducts 2.5 times higher when they were XPD-Asp312Asp+Lys751Lys than carriers of at least 1 variant allele (p = 0.02). The percentage of measurable adducts was only marginally affected by the GSTM1 deletion (p = 0.07). Among those with an active GSTM1, XPD polymorphisms did not have any effect on adduct detection.
|GSTM1 null||GSTM1 active|
|BPDE-DNA adducts/109 nucleotides mean ± SE||Detectable adducts %||BPDE-DNA adducts/109 nucleotides mean ± SE||Detectable adducts %|
|All patients (60)||2.19 ± 0.40 (34)||32.35%||2.35 ± 0.59 (26)||38.46%|
|Asp312Asp+Lys751Lys||4.23 ± 1.18 (6)||66.67%||2.75 ± 1.60 (6)||33.33%|
|1 + variant alleles1||1.75 ± 0.38 (28)||25.00%||2.23 ± 0.63 (20)||40.00%|
|p = 0.022||p = 0.073|
Because of the small study population, further statistical analysis of gene-gene and gene-environment interactions was not informative.
Since peripheral blood DNA adducts have been reported to be possibly predictive of lung cancer risk,25 many recent studies evaluated the possible effect of DNA repair polymorphisms on DNA damage, mainly by detecting unspecific bulky DNA adducts by 32P-postlabelling techniques.26, 27, 28, 29 Given the nature of this method, the exact composition of adducts is usually unknown. This may represent sometimes a limitation when the effect of a single gene in a highly complex pathway such as NER, responsible for the removal of different types of DNA damage, has to be inferred from adduct presence.
A unique aspect of our investigation is the use of the specific BPDE-DNA adduct measured by HRGC-NICI-MS as an in vivo end point for DNA repair polymorphisms. This adduct has the advantage of reflecting a chemical-specific genetic damage (BPDE) that is mechanistically relevant to the carcinogenesis of BP.2
We hypothesized that if there were functional relevances in the polymorphic DNA repair enzymes involved in the removal of the BPDE lesion, we might detect differences in BPDE-DNA adducts presence in individuals exposed to BP. Thus we selected lung cancer patients with a long history of smoking and still active smokers at the time of blood sampling as the prototype of subjects heavily exposed to the tobacco carcinogen BP.
Our result did not show any appreciable association between XRCC1 and/or XPD polymorphisms and the levels or frequency of measurable BPDE-DNA adducts.
Apparently, these findings are not in accord with recent studies, which associated increased levels of bulky DNA adducts with at least one variant allele for XRCC1-399 and XPD-751 polymorphism in healthy populations.27, 28, 29
Various explanations are possible for the divergence between these studies and ours. The controversial results might depend partly on differences in the type of DNA damage detected when compared to the specific BPDE-DNA adduct we have measured. It has been shown that depending on the particular PAH-DNA adduct, the contribution of NER and/or BER pathways can differ in the overall repair process.15 This scenario becomes even more complex when considering many genes involved in different repair pathways with potential divergent effects on different lesions.
We cannot exclude the effect of cancer itself on the DNA repair capacity in our subjects. Its role in terms of repair in lymphocytes is unclear at present. It is possible to hypothesize that it might influence DNA repair activity through, for instance, excessive endogenously generated oxidative stress, which in turn might affect lymphocytes and their repair values,30, 31 flattening the differences among the analyzed genotypes with possibly different repair efficiency.
Alteration of DNA-repair efficiency might also be caused by active smoking, such as that experienced by our subjects. It has been recently reported that cigarette smoking may, in fact, stimulate DNA repair capacity in response to heavy DNA damage caused by tobacco carcinogen,4 although reverse effects have also been reported.32 Nevertheless such an adaptation would mask different repair capacities and may explain why we did not observe any appreciable association between BPDE-DNA adducts formation and XPD genotypes.
We found that the dependence of BPDE-DNA adduct levels on XPD Asp312Asp+Lys751Lys genotype was evident only in GSTM1-deficient smokers, suggesting that the XPD common allele might become a susceptibility factor in the presence of certain gene-gene and environment combinations. The inherited absence of the GSTM1 gene has been associated with a greater likelihood of having detectable PAH-DNA adducts both in the lungs20 and in white blood cells,21 although several studies report no such correlation.22, 23 A potential explanation for the interactions we found could be that impaired BP detoxification capacity coupled with an inefficient repair response or inability to boost the repair response might lead to accumulation of genetic damage, revealed by increased levels of BPDE-DNA adducts.
To date, the association of the XPD-751 polymorphisms with differences in DNA repair is still puzzling and controversial.
Our result concerning the positive association of the XPD Asp312Asp+Lys751Lys genotype with increased DNA damage is not in line with the recent study of Spitz et al.,9 who reported lower DNA repair capacity in XPD Gln751Gln carriers among lung cancer cases. Matullo et al.29 also reported that the XPD751Gln variant allele was associated with higher DNA adducts in never-smokers. On the other hand, Duell et al.26 did not find any correlation between Lys751Gln polymorphism and the presence of polyphenol DNA adducts.
We found some support for the association between the XPD-Asp312Asp+Lys751Lys genotype and enhanced BPDE-DNA damage with a recent report that the Lys751Lys genotype is related to an increased risk of suboptimal DNA repair (as reflected in the number of X-ray-induced lymphocyte chromatid aberrations). Individuals with the Asp312Asp genotype also had heavier DNA damage than 312Asn carriers, although the difference was not significant.12 Increased DNA adducts in carriers of Asp312 allele might be in accord with a recent study associating the presence of Asp/Asp at codon 312 with diminishing apoptotic response.33
The interpretation of the data is limited by our lack of knowledge about the functional significance of the variant 751.33 The Lys allele may have divergent effects in different repair pathways and with different lesions.
As mentioned earlier, our results may be biased by the small numbers of subjects in the various subgroups that could lead to overinterpretation of the nominal p-values. Therefore, our findings should be viewed with caution.
Nevertheless, this evidence does suggest that the XPD-Asp312Asp+Lys751Lys genotype might contribute to increasing the presence of BPDE-DNA adducts, possibly due to reduced DNA repair function, and that this effect might be evident in individuals likely to have accumulated damage as a consequence of low detoxification capacity and high environmental exposure.
In conclusion, this is the first study analysing the specific BPDE adduct to lymphocyte DNA in vivo with regard to polymorphic repair genes (XPD, XRCC1) and xenobiotic metabolizing gene (GSTM1). It suggests that the XPD-Asp312Asp+Lys751Lys genotype may play a role in determining BPDE-DNA adduct levels in the presence of GSTM1 deficiency, probably contributing to vulnerability to BP's effect.
We thank the clinical staff of Clinica Chirurgica, Chirurgia II, Ospedale San Paolo, Milan, Italy, for their cooperation. The editorial assistance of Ms. J. Baggott is gratefully acknowledged.
- 19The glutathione S-transferases: influence of polymorphism on cancer susceptibility. no. 148. Lyon: IARC, 1999. 231–49., .
- 34The World Health Organization histological typing of lung tumours. 2nd ed. Am J Clin Pathol 1982; 77: 123–36.