Micronuclei Related to Anti–B[a]PDE-DNA Adduct in Peripheral Blood Lymphocytes of Heavily Polycyclic Aromatic Hydrocarbon–Exposed Nonsmoking Coke-Oven Workers and Controls

In the theoretical continuum from carcinogen exposure to cancer, DNA adduct formation and other chromosomal anomalies are key initial events in carcinogenesis. They provide a measure of alterations of some important genetic targets and, in the case of indicators of chromosomal damage such as micronuclei (MN) level, the link between the measured variable and the development of tumors is closer (1).

Benzo[a]pyrene (B[a]P), an important contributor to the carcinogenicity of polyclyclic aromatic hydrocarbons (PAH) mixtures, forms damaging DNA adducts after metabolic activation (2). Persistent binding of anti-benzo[a]pyrene diolepoxide (anti–B[a]PDE the ultimate metabolite), causing mutations, including mutational “hotspots” in the pulmonary P53 gene, is a critical step in the carcinogenic process induced by B[a]P (3). Molecular epidemiology studies have shown that PAH-(B[a]P)-DNA adduct formation in the peripheral blood lymphocytes (PBL) is an index of exposure also of the target tissues (i.e., lung; refs. 4, 5). Anti–B[a]PDE-DNA adduct detection in the accessible tissues (PBLs) has allowed us to show genotoxic exposure to PAHs in occupational (cokeries) and environmental (smoking, diet, and environmental) situations (6, 7). Using a large database (HUmMicroNuclei) assembled from many countries (8), an increase in MN frequency has been detected in PBLs of smokers, whereas the increase in coke-oven workers is still uncertain (9 and references inside, 10-12).

In this study, by excluding tobacco smoking as the main confounding factor in PAH exposure, we detected both MN frequency and the biologically effective dose of PAH(B[a]P) exposure estimated using anti–B[a]PDE-DNA within PBLs of nonsmoking Polish coke-oven workers and matched controls verified for PAH exposure by urinary 1-pyrenol. The objectives of the study were to (a) verify the presence of DNA damage as MN in cytokinesis-blocked PBLs of nonsmoking coke-oven workers and (b) examine the relationship between MN formation and levels of anti–B[a]PDE-DNA in PBLs and urinary 1-pyrenol (a recommended bioindicator of exposure to PAH). This is the first study to estimate B[a]P-DNA damage by measuring the specific anti–B[a]PDE-DNA adduct and relating this damage to MN formation within PBLs of the same subject. Assessment of this relationship is valuable as the formation of PAH-DNA adducts in surrogate tissue (i.e., PBL) is also an index of exposure of target tissues (i.e., lung; refs. 4, 5), whereas high adduct and MN levels in PBLs are predictive of cancer risk (13, 14).

Nonsmoking coke-oven workers from three Polish cokeries (n = 49) and nonsmoking Polish controls (n = 45) matched for age and gender were enrolled in this study (January to May 2006). Coke-oven workers had job classifications (i.e., charging, coking, and pushing operations at the coke-oven battery section) that exposed them to high levels of PAHs. We excluded individuals also exposed to benzene (through a by-product operation). Controls were clerks involved in a health check-up program at the Institute of Occupational Medicine and Environmental Health in Sosnowiec (Poland). Nonsmokers were those who, at a minimum, had quit smoking 1 y before sample collection. Analysis of nicotine plus metabolites in the urine of all subjects was used as an objective control of nonexposure to tobacco smoke, as previously described (7). Subjects with nicotine concentrations of ≥0.01 mg/mmol creatinine were considered smokers and were therefore excluded from the study.

Trained interviewers (a) informed subjects of the purpose of the study and asked them to sign an informed consent form; (b) collected data by means a questionnaire, on PAH exposure as previously described (i.e., consumption of charcoaled meat more than weekly was considered a high dietary intake of PAHs; indoor and environmental exposure to PAHs was the use of wood or coal to heat the home; and an index of location of residence in town that included intense local traffic and presence of industries in the residential area; refs. 6, 7); (c) collected urine (50 mL) from workers at the end of their work-shift and after at least three consecutive working days, and also from controls in the late afternoon. At the same time that urine was sampled, blood was collected by a phlebotomist into EDTA-containing tubes (20 mL) and evacuated heparin-containing tubes (10 mL) for adduct and MN analysis, as previously described (6, 7, 11). This study was reviewed by the appropriate Ethics Committee of the Institute of Occupational Medicine and Environmental Health in Sosnowiec. All participants gave their informed consent.

Urinary 1-pyrenol, adjusted by creatinine, was determined as previously described (6). Within 4 h after blood collection, PBLs for adduct analysis were isolated in Ficoll separating solution (Seromed) and kept frozen at −80°C until their transport to the Department of Environmental Medicine and Public Health in Padova, and cultures of PBLs for MN analysis were set up. Anti–B[a]PDE-DNA adduct formation was detected after DNA isolation with a Promega Wizard genomic DNA purification kit (Promega) by high performance liquid chromatography/fluorescence analysis of B[a]P-tetrol-I-1 (see abbreviations) released after acid hydrolysis of DNA samples, as previously described (7). The detection threshold of B[a]P-tetrol-I-1 was 0.25 pg (signal/noise, >3) so that with 100 μg DNA, this assay can measure 0.25 adducts/10⁸ nucleotides (1 fmol/μgDNA = 30 adducts/10⁸ nucleotides; ref. 7). MN analysis was done on coded slides scored by light microscopy at ×400 magnification, as previously described (11). To exclude artifacts, the identification of MN was confirmed at ×1,000 magnification in 10% samples. The scoring of binucleate, trinucleate, and tetranucleate cells and MN was done and cytokinesis block proliferation index was then calculated (15). MN levels were verified on 20 slides (two slides per subject) previously scored at the Institute of Occupational Medicine and Environmental Health in Sosnowiec (Poland) by blind scoring at the Environmental Carcinogenesis Unit, Istituto Nazionale per La Ricerca sul Cancro (IST). The results of this intercalibration exercise showed a good correlation between MN levels scored in the two laboratories (test ANOVA; F = 20.91; P < 0.001).

Statistical comparisons were made between various groups using the nonparametric Mann-Whitney U test or the χ² test. Simple linear regression analysis estimated the relationship between internal (urinary 1-pyrenol) and biological effective dose (anti–B[a]PDE-DNA adduct) indicator levels, on MN formation. Multiple linear regression analysis was used to assess the influence of exposure to PAHs (evaluated by urinary levels of 1-pyrenol μmol/mmol creatinine) and anti–B[a]PDE-DNA adduct levels on MN formation. All statistical tests were two-sided and were done with Statsdirect Statistical software (Ashwell).

The study population consisted of 49 workers and 45 controls who were all nonsmoking males and in the same age range (P = 0.855; Table 1). The diet was very similar in the exposed and control groups as they live in the same area of Poland [56% workers and 49% controls declared a daily consumption of fruits and vegetables (P = 0.550)]; specifically for PAHs exposure with diet, 8% workers and 2% controls declared a consumption of charcoaled meat more than once per week (P = 0.1998). The study cannot however exclude the possibility that variability in MN frequency data may have been affected by dietary factors such as folate and vitamin B12 intake. Alcohol consumption may be also relevant in this regard. Coke-oven workers were exposed to very high levels of PAHs: urinary 1-pyrenol excretion was ∼30 times higher in workers than controls (P < 0.001), with 80% of workers versus 0% controls (P < 0.001) exceeding 2.28 μmoles/mol creatinine, a biological exposure index proposed by Jongeneelen (16). Levels of anti–B[a]PDE-DNA and MN were also significantly higher in workers relative to controls (P < 0.001). Cytokinesis block proliferation index was equal in both groups (P = 0.599; data not shown). All workers (100%) were positive for adduct and MN formation, whereas only 36% (adduct) and 51% (MN) of controls were positive (for adduct and for MN; P < 0.001). In all workers (100%), adduct levels (an indicator closer to 1-pyrenol than MN) exceeded the 95th percentile control value (0.44 adducts/10⁸ nucleotides; P < 0.001), whereas only 61% of the workers exceeded the 95th percentile control value for MN levels (3 MN/1,000 BN cells; P < 0.001). Indoor exposure (i.e., presence of wood or coal heating in the home) was related to a significant increase in adduct levels among controls [n = 27 without versus n = 18 with wood or coal heating, median (range), 0.125 (0.125-0.39) versus median (range), 0.125 (0.125-5.56) adducts/10⁸ nucleotides; P < 0.001; data not shown] but not among workers. There was no effect of environmental exposure (i.e., location of residence in town, intense local traffic, and presence of industries in the residential area) on adduct formation among workers or controls. Low indoor and environmental exposure to PAHs had no effect on MN levels in workers or controls.

Anti–B[a]PDE-DNA adduct and 1-pyrenol levels were significantly correlated with MN frequency in all subjects (n = 94; P < 0.001; Table 2). There was also a significant correlation between 1-pyrenol and anti–B[a]PDE-DNA levels (r = 0.67; P < 0.001; data not shown). There was a stronger correlation between adducts and MN, thereby improving the dose-response relationship, when subjects with adduct levels above the 3rd tertile (≥4.35 adducts/10⁸nucleotides) were excluded from the analysis (n = 61; r = 0.69; P < 0.001; Table 2). Saturation of adduct/MN formation at high levels may disturb the underlying relationship. There was no correlation between adducts and MN at adduct levels of ≥4.35 adducts/10⁸nucleotides (i.e., 3rd and higher tertile adduct levels; n = 33; r = 0.09; P = 0.615; data not shown). The relationship (Table 3) between adducts and MN was similar even if individuals belonging to the 1st and lower tertile adduct levels were excluded (n = 30; r = 0.67; P < 0.001). Linear multiple regression analysis, excluding subjects within the 3rd tertile adduct level (n = 61), revealed that adduct formation (P < 0.001), but not 1-pyrenol, was the significant determinant in increasing MN.

Incorporation of biomarkers into epidemiologic studies is important because biomarkers will improve assessment of exposure, document early changes preceding disease, and identify subgroups in the population with greater susceptibility to cancer, thereby increasing the ability of epidemiologic studies to identify causes and elucidate mechanisms in carcinogenesis. We found that coke-oven workers, which were heavily exposed to PAHs (80% of workers exceeded the urinary 1-pyrenol biological exposure index value), presented significantly higher MN frequency in PBLs relative to control subjects. Substantial variation was also found for adduct levels in PBLs of the same individual. To date, no study has unequivocally established the presence of chromosomal damage in coke-oven workers. Smoking habits have always seemed as a confounding factor (11, 12), or PAH exposure of workers was lower (with <10% exceeding biological exposure index value; ref. 9 and references therein) than in our study. At this value, corresponding to the postshift excretion value with environmental exposure to the airborne threshold limit value of coal tar pitch volatiles (i.e., 0.2 mg/m3 of “benzene soluble matter,” ACGIH; ref. 17), a relative risk of 1.3 of lung cancer for coke-oven workers was estimated (16). Our results indicate that workers are exposed to very high levels of PAHs and this is the main reason for our unequivocal results, showing all indicators were significantly altered by this genotoxic exposure. Our results also substantiate, with the use of an early indicator of biological effect, that workers are at higher cancer risk than controls and confirm the predictive role of this indicator of genotoxic risk for cancer risk. In addition, the cause of an early biological effect due to occupational exposure to PAHs is clearly established excluding other possible elevated nonoccupational genotoxic exposure (smoke, diet, and ambient).

The influence of indoor exposure due to the use of coal (very common in Poland; ref. 18) or wood-burning fireplaces and stoves as residential heating could be the main nonoccupational cause of increased adduct levels in our nonsmoking controls. These results are in concordance with our previous work (7), where an influence of indoor exposure (mainly due to the use of coal or wood in residential heating) on anti–B[a]PDE-DNA in Italian nonsmokers was found. Above all, the overlap of high PAH exposure in coke-oven workers may explain the negligible effect of PAH by indoor exposure.

Coke-oven workers were all responsive for DNA-adduct and, to a lower extent, MN formation. Our results suggest that it is easier to detect PAH exposure with an indicator of the biological effective dose (DNA-adducts) compared with an early biological effect indicator (MN). In the theoretical continuum from exposure to cancer, DNA adduct, rather than MN formation, is closer to the initial step of PAH(B[a]P) exposure. Our study, following the same route of exposure, again supports the original model (19), that a continuum of molecular alterations leading to cancer occurs and that it can be measured and traced using a molecular bioindicator of genotoxic risk.

Increase in MN levels was significantly related to anti–B[a]PDE-DNA formation, a key adduct derived from the ultimate carcinogenic metabolite of B[a]P. The dose-response relationship was improved by excluding subjects with the highest adduct levels. No further increase in the MN-adduct relationship was found for DNA damage higher than 4.35 adducts/10⁸ nucleotides, suggesting that a “plateau” was reached in MN formation as an outcome of anti–B[a]PDE-DNA damage in heavily PAH-exposed coke-oven workers. Anti–B[a]PDE-DNA was the significant determinant for MN formation in individuals with adduct levels lower than 4.35 adducts/10⁸ nucleotides. B[a]P is a strong clastogen, widely used as a positive control reference in experimental treatments. In particular, of the four metabolically formed optical isomers of B[a]PDE, the stereoselective (+)-anti–B[a]PDE binding to the exocyclic N2 of guanine is considered the primary essential damaging event in B[a]P genotoxicity and carcinogenicity (20, 21). A significant role of DNA adducts on MN formation is addressed in our work. Nevertheless, we cannot exclude the influence of genotoxic coexposure with PAHs, or a mode of action other than DNA adduct formation, on MN levels. Kinetochore and spindle protein alterations by PAHs, and/or anti-BPDE itself, could be alternate hypotheses for the increase of MN frequencies.

In conclusion, we clearly showed the influence of occupational exposure to PAHs on an early biological effect excluding other high extraoccupational genotoxic exposure. The early biological effect was mainly related to specific anti–B[a]PDE-DNA formation within PBLs of the same subject. With the use of an early indicator of biological effect, our results substantiate that workers are at higher cancer risk than controls, and confirm the ability of this genotoxic risk indicator to predict cancer risk. The results obtained from this study are important because the formation of adducts and MN is, the former, an index of exposure in the target tissue, whereas the increase in both measures is predictive of higher cancer risk.

No potential conflicts of interest were disclosed.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.