Sister Chromatid Exchange and Oxidative DNA Damage in Paving Workers exposed to PAHs

Ann. Occup. Hyg., Vol. 50, No. 3, pp. 211 218, 2006 # 2005 British Occupational Hygiene Society Published by Oxford University Press doi:10.1093/annhyg/mei072 Sister Chromatid Exchange and Oxidative DNA Damage in Paving Workers exposed to PAHs D. CAVALLO 1 *, C. L. URSINI 1, P. BAVAZZANO 2, C. CASSINELLI 2, A. FRATTINI 3, B. PERNICONI 1, A. DI FRANCESCO 1, A. CIERVO 1, B. RONDINONE 1 and S. IAVICOLI 1 1 Department of Occupational Medicine, ISPESL Monteporzio Catone, Rome, Italy; 2 Laboratory of Public health, ASL-Florence, Italy; 3 Department of Prevention ASL RMG, Guidonia, Rome, Italy Received 22 July 2005; in final form 18 October 2005; published online 12 December 2005 Paving workers are exposed during road paving to several polycyclic aromatic hydrocarbons (PAHs) contained in asphalt fumes. In this study early genotoxic and oxidative effects of exposure to bitumen fumes were evaluated in 19 paving workers and 22 controls. Environmental and biological monitoring of exposure was carried out, measuring, on personal air samples from exposed workers collected during three working days, the concentration of 14 PAHs and urinary OH-pyrene at the end of each of the three working days. Genotoxic effect was evaluated analysing sister chromatid exchange (SCE) frequency and direct-oxidative DNA damage by formamido pyrimidine glycosylase (Fpg)-modified comet assay on lymphocytes. Tail moment values from Fpg-enzyme treated cells (TMenz) and from untreated cells (TM) were used as parameters of direct and oxidative DNA damage, respectively. For each subject, the TMenz/TM ratio >2.0 was used to indicate the presence of oxidative damage. DNA damage was also evaluated analysing comet percentage. Personal air samples showed low level of total PAHs (2.843 mg m 3 ) with prevalence of 2 3 ring PAHs (2.693 mg m 3 ). Urinary OHpyrene after work-shift of the three working days was significantly higher than that found at the beginning of the working week. SCE analysis did not show any difference between two groups while an oxidative DNA damage was found in 37% of exposed with respect to the absence in controls. Comet percentage was significantly higher (P = 0.000 ANOVA) in the exposed than in controls. The results demonstrate the high sensitivity of comet assay to assess early oxidative effects induced by exposure to bitumen fumes at low doses and confirm the suitability of urinary OH-pyrene as a biomarker of PAH exposure. In conclusion the study suggests the use of Fpg-modified comet test as a biomarker of early genotoxic effects and that of urinary OH-pyrene as a biomarker of PAH exposure to furnish indications in terms of characterization, prevention and management of risk in occupational exposure to mixtures of potentially carcinogenic substances. Keywords: genotoxic effects; paving workers; polycyclic aromatic hydrocarbons; urinary 1-OH pyrene INTRODUCTION Occupational exposure of paving workers is still poorly characterized. This category of workers performs different tasks, including bitumen asphalt preparation and road paving, that chronically expose, both by inhalation and dermal contamination, workers to polycyclic aromatic hydrocarbons and heterocyclic compounds containing sulphur, nitrogen and oxygen present in high amount in asphalt fumes *Author to whom correspondence should addressed. Tel: +39-6-94181409; fax: +39-6-94181410; e-mail: cavallo.d@tiscali.it (McClean et al., 2004b). Asphalt fumes are complex mixtures of aerosols and vapours containing various organic compounds including polycyclic aromatic hydrocarbons (PAHs) and heterocyclic compounds (King et al., 1984). Many of the organic compounds found in asphalt fumes have been shown to be mutagenic and/or carcinogenic (Machado et al., 1993; Sivak et al., 1997), and the mutagenic activity seems to be correlated to the amount of 3 7 ring PAHs in the fumes (Machado et al., 1993). For many PAHs contained in asphalt fumes a certain carcinogenicity has been shown, particularly for compounds with more than four aromatic rings. 211

212 D. Cavallo et al. Epidemiological studies have indicated an association between an excess of cancer risk (lung, stomach, bladder, leukaemia and non-melanoma skin cancer) and exposure to asphalt fumes. In particular a higher cancer risk has been found in roofers than in pavers (Bender et al., 1989; Partanen and Boffetta, 1994; Boffetta et al., 1997, 2003), probably due to higher mutagenic/carcinogenic PAH content in the roofing asphalt fumes generated at higher temperatures. The road pavers are exposed to asphalt fumes containing relatively low levels of PAHs, but there is a possibility of long-term health effects following chronic exposure by inhalation or skin contamination. Several epidemiological studies on pavers showed the presence of respiratory diseases and possible association between lung cancer risk and asphalt fume exposure (Partanen et al., 1995; Boffetta et al., 1997), as reported also by IARC (2001), which confirmed an excess of lung cancer in this occupational category. Mutagenic/genotoxic effects of roofing asphalt fumes have been shown through both in vitro and in vivo studies (Bender et al., 1989; Partanen et al., 1995; Qian et al., 1996; Sivak et al., 1997; Toraason et al., 2001), whereas studies on genotoxic effects of paving asphalt fumes contrast with this. In particular an increase in sister chromatid exchange (SCE) and MN formation was reported by Burgaz et al. (1998) in lymphocytes of road paving workers, mainly rakermen, in Turkey and a slight increase of DNA strand breaks was reported by Fuchs et al. (1996) in Germany. A lack of SCE and MN induction was reported by Jarvholm et al. (1999) for road pavers in Sweden. Contrasting results were shown also in studies on animal models (De Meo et al., 1996; Reinke et al., 2000; Zhao et al., 2004). Though some studies have described an excess risk of cancer among asphalt-exposed workers, there is currently insufficient evidence to establish a causal relationship between occupational asphalt exposure and cancer risk. The principal limitation is the lack of a correct exposure assessment by quantitative measurements (Chiazze et al., 1991; NIOSH, 2000) of exposure to asphalt or its constituents. Contrasting results could also be due to the high variability of exposure conditions characterizing the paving activity influenced by environmental factors (wind, heat, etc.) associated with different climatic and meteorological conditions. Despite the low levels of PAHs, paving asphalt fumes are complex mixtures with unclear effects. The carcinogenic potency of asphalt fumes cannot be predicted only on the content of any single known carcinogenic PAH. The existence of other components of asphalt fumes that may not themselves be carcinogenic but act as co-carcinogens could contribute to the development of carcinogenic effects. In addition, PAH-mediated genotoxicity is not limited to the formation of active PAH metabolites. Other mechanisms such as ROS-mediated events may also play a role in paving asphalt fume-induced effects (Kumagai et al., 1997; Takano et al., 2002; Zhao et al., 2004). The aim of our study was to evaluate, in 19 paving workers exposed to complex PAH mixture, early genotoxic and oxidative effects induced by asphalt fumes by SCE and formamido pyrimidine glycosylase (Fpg)-modified comet test. In fact SCEs are promptly induced after exposure to genotoxic chemicals, and Fpg-modified comet assay is a sensitive and rapid technique that shows evidence of early and still repairable oxidative DNA damage. The pavers exposure has been well characterized by environmental and biological monitoring and effects were compared with exposure data. In particular we analysed on personal air samples, collected from May to June 2003, the concentrations of 14 out of the 16 priority PAHs published by the US environmental protection agency (EPA) and the urinary OH-pyrene as a measure of total absorbed dose (by inhalation and dermal exposure). The identification of biomarkers of early biological effect, such as genotoxic and oxidative damage, could contribute to the clarification of the mechanisms of action of complex PAH mixtures such as asphalt fume containing potentially carcinogenic substances at low doses. Moreover the study could furnish useful indications also in terms of risk prevention and management. MATERIALS AND METHODS Study population and study design The study was performed on 19 paving workers (9 smokers and 10 non-smokers) with mean age of 39.3 10.8 years and a mean job seniority in the specific task of 11 years. A control group of 22 healthy subjects working in the administrative area was selected to match exposed group for age and lifestyle (mean age 40.5 9.8 years, 11 smokers and 11 non-smokers). The hospital Ethical Committee and the ISPESL Scientific Committee approved the human study. All subjects gave informed consent. Anagraphic, clinical, working information including protective equipments, lifestyle habits (smoking, dietary habit, alcohol consumption) were obtained from a questionnaire administered by specialized medical personnel. All pavers wore protective equipment: all our subjects reported that they used gloves and safety shoes, while 7/19 reported using disposable respirators, but we are not sure that they always used them during the whole working shift. The exposure assessment was carried out in different paving sites near Rome, from May to June 2003. Environmental and biological monitoring of the exposure and evaluation of effects in paving workers

Sister chromatid exchange and oxidative DNA damage in paving workers 213 were performed on samples of air, urine and blood collected simultaneously in the same working week. Also urine and blood samples from controls were collected simultaneously on the third day of week. The subjects worked with concrete asphalt with a temperature of 160 C, with a shift length of 8 h. Exposure assessment Personal air samples. Personal air samples were obtained from each worker in accordance with NIOSH Method 5506 (NIOSH, 1998). The air samples were collected during full work shift (8 h)of each of three consecutive days (three samples for each subject). For air samples analysis the method of Perico et al. (2001) was used. The air sampling system consisted of a Teflon filter and a cassette holder to collect particulate PAHs, connected in series with an XAD-2 adsorbent tube to collect the organic vapour. Using a personal sampling pump operating for 8 h at 2 l min 1, a 37 mm diameter filter (laminated PTFE of 2 mm pore size) was placed in a cassette and attached to each worker s lapel near the breathing zone and the adsorbent tube was attached in line and downstream from each filter cassette. Samples were transported in coolers and stored at 20 C until analysis. The stored filters (particulate) and tube (vapour) samples were differently treated and analysed. Particulate was extracted from filters with 2 ml of cyclo-hexane (Sigma-Aldrich, Italy) and ultrasonicated for 10 min. Extracts were dried under nitrogen flux, dissolved in 1 ml of acetonitrile (Sigma-Aldrich, UK) and directly injected in high pressure liquid chromatography (HPLC) system. The organic vapour was extracted from adsorbent materials contained in tubes using 5 ml of acetonitrile, ultrasonicated for 10 min and directly injected in HPLC system. The concentrations of 14 PAHs [naphthalene, acenaphthene, fluorene, phenanthrene, anthracene, fluoranthene, pyrene, benzo(a)anthracene, chrysene, benzo(b)fluoranthene, benzo(k)fluoranthene, benzo(a)pyrene, dibenzo(ah)anthracene and benzo(ghi)perylene] listed by the US EPA, were measured. Samples and standards were analysed by HPLC (Alliance, Waters, USA) with fluorometric detector and diode array. The limit of detection (LOD) for particulate samples was 0.010 mg m 3 for naphthalene, 0.005 mg m 3 for acenaphthene, fluorene and phenantrene and 0.001 mg m 3 for the other PAHs analysed, with a limit of quantification (LOQ) of three times the LOD. The LOD for vapour samples was 0.052 mg m 3 for naphthalene and acenaphthene and 0.005 mg m 3 for the other PAHs analysed with an LOQ of 0.156 mg m 3 and 0.015 mg m 3, respectively. Measurements below the detection limit were replaced by values of half of the LOD. Total air exposure estimate was calculated, for each worker, on each sampling day by adding the particulate and vapour measurements and is reported as mean concentration (mg m 3 )of three working days. Urine samples. The urine samples of the paving workers were collected after work-shift of the first three consecutive working days of week. Also a baseline sample was collected on Monday morning before the work-shift. All urine samples were collected in sterilized polypropylene specimen containers and frozen at 20 C until analysis. Concentrations of 1-hydroxypyrene (1-OHP), used as measure of total PAH absorbed dose, were evaluated by HPLC. Urine samples were diluted 1:1 with 5 ml of acetate buffer ph 5.0, and 100 ml ofb-glucuronidase (1.000 units) (Sigma, Aldrich, Italy) were added. The mixture was incubated for 12 h at 37 C in a bath and, after reaching environmental temperature, 200 ml of phenanthrene (250 mgml 1 ), used as internal standard, was added. A calibration curve was constructed using 1-OHP in the range 100 800 ng l 1. Samples and standards were purified on SPE Bond Elut (R) C18 (Varian Inc., USA) columns primed with 3 ml of methanol and 3 ml of H 2 O, were washed with H 2 O/methanol 1:1 and eluted with 2 ml of methanol. An aliquot of 50 ml of eluate was injected in a HPLC VP Series system equipped by fluorometric detector RF10AXL (l Ex 242 nm; l Em 388 nm) (Shimadzu Corp., Japan) with a flow rate of 1.2 ml min 1.A column Lichrosphere C18 (125 mm 4.0 mm ID; 5 mm granulometry) and a specific gradient were used. The LOD was 25 ng l 1 and the linearity interval was between 100 and 3200 ng l 1. The LOD correlates with a value of 0.015 mg g 1 of creatinine considering a mean of 1.65 g l 1 for the creatinine. The 1-OHP concentrations were corrected for creatinine prior to statistical analysis. Biological effects Specialized medical personnel collected whole venous blood samples from exposed (n = 19) and controls (n = 22), by venipuncture in sterile heparinized disposable syringes. From each subject, 5 ml of blood was collected and transferred within 2 h, at room temperature and in the dark, to the laboratory where the analyses were performed. Blood from paving workers was collected on the third working day. Sister chromatid exchange. Cultures were established by adding 0.5 ml of blood to 5 ml complete medium RPMI 1640 containing 20% fetal calf serum, 2% phytohemagglutinin, 100 IU ml 1 penicillin, 100 mg ml 1 streptomycin and 2 mm L-glutamine, and incubating for 24 h at 37 C. A 5-Bromodeoxyuridine (Sigma, USA) solution at a final concentration of 10 mgml 1 was added. Lymphocytes were cultured in the dark for 48 h and metaphases were blocked during the last 2 h with Colcemid (Gibco Technology, USA) at a final concentration of 0.2 mgml 1. Further

214 D. Cavallo et al. processing included hypotonic treatment, fixation, slide preparation and fluorescein plus Giemsa staining for detection of SCE (Latt and Schreck, 1980). Fifty second-division metaphases were scored on coded slides by a single observer as the number of SCE/cell per subject. Comet assay. We used Fpg-modified Comet assay to evaluate oxidative DNA damage, which uses Fpg enzyme a bacterial (Escherichia coli) glycosylase. This enzyme recognizes and specifically cuts the oxidized bases, principally 8-oxoguanine, on DNA, producing apurinic sites converted in breaks by the associated AP-endonuclease activity. Therefore, these breaks can be detected by comet assay and give a measure of DNA oxidative damage, enabling us to detect moderate but still appreciable damage even at low exposures. We followed the procedure of Collins et al. (1999), with minor modifications. Two coded gel bond films were prepared for each case (one to be treated without and the other with Fpg) allowing the detection of direct DNA lesions (single double strand breaks and alkali-labile sites) and oxidative DNA damage, respectively (Collins et al., 1999). On top of each film about 20 ml of mononuclear cells (isolated by Ficoll from 4.5 ml of peripheral blood and suspended in 1 ml of PBS) mixed with 70 ml of low-melting agarose 0.7% in PBS at 37 C were layered. Then they were bathed in lysis solution (2.5 M NaCl, 100 mm Na 2 EDTA, 10 mm Tris with 1% Triton X-100 and 10% DMSO added fresh) and kept in the dark for 1 h at 4 C. The coded films were supported by slides, washed three times in enzyme buffer (50 mm Na 3 PO 4,10mM EDTA, 100 mm NaCl, ph 7.5), drained and incubated with 50 ml of either buffer or Fpg (Sigma Aldrich, USA) (1 mg ml 1 in enzyme buffer) in the dark for 30 min at 37 C. The slides were placed in a horizontal gel electrophoresis tank filled with fresh alkaline buffer (1 mm Na 2 EDTA and 300 mm NaOH, ph 13) for 40 min at 4 C to allow denaturing and unwinding of the DNA and the expression of alkaline-labile sites. Electrophoresis was done in the same buffer at 25 V and 300 ma for 30 min to allow the fragments of damaged DNA to migrate towards the anode. The slides were then washed three times with 0.4 M Tris HCl for 5 min and stained with 50 ml ethidium bromide (10 mg ml 1 ). Slides were examined at 200 magnification under a fluorescence microscope. An undamaged cell appeared as a nucleoid and a cell with damaged DNA as a comet. Images of 50 randomly selected comets stained with ethidium bromide were acquired and analysed from each sample with specific image analyser software (Delta Sistemi, Roma, Italy) to evaluate the tail moment (the product of relative intensity of the tail and the distance from the centre of the head to the centre of gravity of the tail) used as a measure of DNA damage. For each subject we calculated the mean value of the tail moment for 50 comets from Fpg-treated cells (TMenz), which furnishes an indirect measurement of oxidative DNA damage, and from Fpg-untreated cells (TM), which indicates direct DNA damage. For each subject the TMenz/TM ratio >2.0 was used to indicate the presence of oxidative damage. About 1000 cells from each slide were examined for presence of comets (cells with a detectable tail) by an experienced observer and the percentage of comets was calculated. Statistical analysis We analysed preliminarily the presence of possible significant differences between exposed and controls by one-way ANOVA for age, by chi square test for smoking habits, alcohol intake, drug assumptions and dietary habit. For statistical analysis of results ANOVA multivariate, non-parametric tests for nonnormal distribution (Kruskal Wallis, Mann Whitney U-test, c 2 test) and Student s t-test were used and a P < 0.05 was considered significant. Pearson s correlation coefficients (r) were calculated to evaluate the relationship between PAH concentrations in the air samples and urinary 1-OHP. RESULTS Statistical analysis of exposed and controls for age, smoking habits, alcohol intake, drug assumptions and dietary habit did not show any statistically significant difference (Table 1). The results of environmental monitoring are shown in Table 2. The PAH concentrations are expressed as average amount of the 14 PAHs analysed (including particles and vapours) and detected in the breathing zones of workers during paving of three working days. The measured PAHs are also presented as 2 3 rings, 4 6 rings and total PAHs amount and show a prevalence of 2 3 ring PAHs with highest levels for phenanthrene and naphthalene in vapour phase. Urinary OH-pyrene, used as a marker of total PAHs absorbed dose, was evaluated before work-shift at the beginning of the working week and at the end of each work-shift of three consecutive working days (Table 3). The post-shift urinary OH-pyrene level of each working day and the mean 1-OHP post-shift of the three working days were significantly higher than that found at the beginning of the working week (P < 0.05; t-test for paired samples). Moreover the mean 1- OHP post-shift of the three working days and the mean 1-OHP post-shift of 2 working day were significantly higher than that found in controls by Mann Whitney U test (P = 0.036 and 0.011, respectively) (Table 3). Two-way ANOVA performed to

Sister chromatid exchange and oxidative DNA damage in paving workers 215 Table 1. Distribution of studied subjects for age, smoking habits, alcohol intake, drug assumptions, dietary habit and statistical analysis of differences Variable Modality Exposed group (n = 19) Control group (n = 22) Statistical significance of differences among groups Age 39.3 10.8 40.5 9.8 P-value ANOVA = 0.9 Smoking No smokers 5 9 Ex smokers 5 2 P-value c 2 = 0.3 Smokers 9 11 Alcohol assumption (1 missing data) No 3 3 P-value c 2 = 0.9 Yes 16 18 Usual drugs intake No 14 15 Yes 5 7 P-value c 2 = 0.7 Dietary habit (fruits and vegetables intake) Sometime 7 6 Often 6 6 P-value c 2 = 0.6 Usually 6 10 Dietary habit (grilled and smokedfood intake) Sometime 6 3 Often 10 11 P-value c 2 = 0.2 Usually 3 8 Table 2. Mean PAH concentrations (mg m 3 ) in air samples (n = 57) collected in the breathing zone of all paving workers (n = 19) during three working days Number of aromatic rings Arithmetic mean particulate (n =57) Arithmetic mean vapour (n = 57) Arithmetic mean sum of particulate and vapour measurements (n = 57) Naphthalene 2 0.008 0.889 0.897 0.311 3.286 Acenaphthene 2 0.002 0.081 0.083 0.028 0.321 Fluorene 2 0.029 0.151 0.180 0.036 0.853 Phenanthrene 3 0.016 1.317 1.33 0.005 9.288 Anthracene 3 0.002 0.035 0.037 0.003 0.247 Fluoranthene 3 0.046 0.118 0.164 0.016 0.917 Pyrene 4 0.027 0.032 0.059 0.003 0.452 Benzo(a)anthracene 4 0.013 0.006 0.019 0.003 0.136 Chrysene 4 0.014 0.009 0.023 0.003 0.146 Benzo(b)fluoranthene 4 0.006 0.003 0.009 0.003 0.045 Benzo(k)fluoranthene 4 0.003 0.002 0.005 0.003 0.027 Benzo(a)pyrene 5 0.003 0.002 0.005 0.003 0.020 Dibenzo(ah)anthracene 5 0.004 0.003 0.007 0.003 0.027 Benzo(ghi)perylene 6 0.002 0.003 0.005 0.003 0.022 2 3 ring PAHs 0.103 2.591 2.693 0.401 14.912 4 6 ring PAHs 0.072 0.061 0.150 0.033 0.945 Total PAHs 0.175 2.652 2.843 0.434 15.858 For each subject three air samples were collected (one of each day during full work-shift). Range Table 3. Mean urinary 1-OHP concentrations (mg g 1 of creatinine) in paving workers and controls Pre-shift of working week Post-shift 1 day (P paired t-test) Post-shift 2 day (P paired t-test) Post-shift 3 days (P paired t-test) Post-shift 1 3 days mean value (P paired t-test) Exposed (n = 19) 0.27 0.79 (0.010) 0.85* (0.000) 0.77 (0.001) 0.81* (0.000) Controls (n = 22) 0.49 From each paving worker n = 4 urine samples were collected (a baseline sample collected on Monday morning before work-shift and one sample collected after work-shift of each working day). P paired test between post-shift and pre-shift 1-OHP concentrations. *P-value < 0.05 in respect to control group (Mann Whitney U test).

216 D. Cavallo et al. Table 4. Mean urinary 1-OHP concentrations (mg g 1 of creatinine) in smoking versus non-smoking paving workers (post-shift 2 working day values) and controls Mean 1-OHP (mg g 1 of creatinine) smokers Mean 1-OHP (mg g 1 of creatinine) non-smokers Mean 1-OHP (mg g 1 of creatinine) Total Exposed (n = 19) 0.86 (n = 9) 0.82 (n = 10) 0.85 (n = 19) Controls (n = 22) 0.73 (n = 11) 0.15 (n = 11) 0.49 (n = 22) Total 0.79 (n = 20) 0.39 (n = 21) 0.65 (n = 41) Two-way ANOVA for smoke and exposure on 1-OHP urinary levels showed a statistically significance for exposure (P = 0.014) and smoke (P = 0.003) and a non-significant result for smoke exposure interaction (P = 0.148). Excluding smoke effect ANOVA analysis showed a statistically significance (P = 0.031) for exposure effect. Table 5. Correlation coefficients (r) and probability (P) and Partial correlation (excluding smoke effect) coefficients (r) and probability (P) between PAH exposure and urinary 1-hydroxypyrene (1-OHP) of post-shift working days PAH exposure 1-OHP 1 day 1-OHP 2 day 1-OHP 3 day Mean 1-OHP of the three days Correlation Pyrene r = 0.209 (P = 0.390) 2 3 ring PAH r = 0.443 (P = 0.057) 4 6 ring PAH r = 0.162 (P = 0.506) Total PAH r = 0.424 (P = 0.070) *Statistical significant P < 0.05. Partial correlation r = 0.199 (P = 0.429) r = 0.443 (P = 0.066) r = 0.154 (P = 0.542) r = 0.424 (P = 0.080) Correlation r = 0.391 (P = 0.098) r = 0.487 (P = 0.035 * ) r = 0.455 (P = 0.057) r = 0.482 (P = 0.037 * ) Partial correlation r = 0.478 (P = 0.045 * ) r = 0.551 (P = 0.018 * ) r = 0.448 (P = 0.063) r = 0.548 (P = 0.019 * ) Correlation r = 0.262 (P = 0.278) r = 0.374 (P = 0.114) r = 0.310 (P = 0.197) r = 0.372 (P = 0.116) Partial correlation r = 0.407 (P = 0.093) r = 0.490 (P = 0.039 * ) r = 0.440 (P = 0.067) r = 0.489 (P = 0.039 * ) Correlation r = 0.458 (P = 0.049 * ) r = 0.559 (P = 0.013 * ) r = 0.430 (P = 0.066) r = 0.552 (P = 0.014 * ) Partial correlation r = 0.476 (P = 0.046 * ) r = 0.573 (P = 0.013 * ) r = 0.444 (P = 0.065) r = 0.567 (P = 0.014 * ) evaluate the effect of smoke on 1-OHP urinary levels was carried out on 1-OHP urinary levels of 2 working day since they were the highest values (Table 4). This analysis showed a statistical significance for the effects exposure (P = 0.014) and smoke (P = 0.003) on 1-OHP urinary levels while a non-significant result was found for smoke exposure interaction (P = 0.148). The exposure effect was still statistically significant (P = 0.031) after exclusion of smoke effect (Table 4). Correlation analysis between PAH exposure and urinary 1-OHP post-shift working days showed a statistical significance between 1-OHP of 2 day and mean concentrations of 2 3 ring PAH and total PAHs (Table 5). A significant correlation was also found between mean 1-OHP value of the 3 days and pyrene, 2 3 ring PAH and total PAH exposure. The partial correlation (excluding smoke effect) analysis showed a statistical significance between 1-OHP of 2 and 3 day and mean concentrations of 2 3 ring PAH and total PAHs and for the second day a significant partial correlation also for pyrene exposure. Similar to 2 working day a significant partial correlation was found between mean 1-OHP value of the 3 days and pyrene, 2 3 ring PAH and total PAH exposure. The results did not show differences in the frequency of SCE between exposed and controls (4.1 versus 4.0 ) c 2 test P > 0.05. Comet results are shown in Table 6. Analysis by student s t-test for independent samples showed in the exposed group a significant (P = 0.008) increase of comet tail moment from lymphocytes treated with Fpg enzyme (TMenz) with respect to controls (39.0 14.3 versus 28.8 5.0). Moreover a significant (P = 0.000) increase of comet percentage both for cells untreated with Fpg enzyme and treated with Fpg was found in exposed group with respect to controls. For each subject the TMenz/TM ratio >2.0 was used to indicate the presence of oxidative damage. On the basis of this cut-off value 37% of subjects in the exposed group were found with oxidative DNA damage with the respect to lack of damage in control group. Two-way ANOVA was carried out to evaluate the interfering of smoke on DNA damage and did not show any statistical significance for smoke and exposure smoke interaction; whereas, a significance was found for exposure relative to % comets Fpg-untreated cells (P = 0.000), % comets Fpgtreated cells (P = 0.000) and TMenz (P = 0.004) (Table 6). DISCUSSION Environmental monitoring of exposure in studied paving workers shows low exposure levels for the analysed PAHs compared with those of some other PAH-exposed industrial workers, which ranged from 0.4 to 35 mg m 3 for pyrene levels (Elovaara, 1995; Kuljukka, 1996; Pyy, 1997). Our exposure levels are also lower than those reported in other studies that found for paving workers total PAH levels ranging

Sister chromatid exchange and oxidative DNA damage in paving workers 217 Table 6. Comet test results on lymphocytes of paving workers and controls. For each case the mean TM (mean tail moment value of 50 comets from cells untreated with Fpg enzyme) and TMenz (tail moment of 50 comets from Fpg-treated cells) were calculated Group TM Mean SD TMenz Mean SD % Comets enzymeuntreated cells Mean SD % Comets enzymetreated cells Mean SD Subjects with oxidative DNA damage (%) Exposed Smokers (n = 9) 24.7 10.2 38.2 14.2 33.1 16.4 35.8 14.9 1/19 No-smokers (n = 10) 20.9 5.8 39.7 15.1 39.3 14.2 43.0 11.9 6/19 Total (n = 19) 22.7 7.2 39.0 14.3 36.3 15.2 39.6 13.5 7/19 (37) Controls Smokers (n = 11) 18.7 6.2 27.8 4.6 9.8 3.5 10.9 2.5 No-smokers (n = 11) 20.3 4.8 29.8 5.4 8.9 5.3 10.4 6.1 Total (n = 22) 19.5 6.0 28.8 5.0 9.4 4.4 10.6 4.5 0/22 (0) t-test P values 0.144 0.008 0.000 0.000 Total exposed versus controls All subjects Smokers (n = 20) 21.4 8.6 32.5 11.1 20.3 16.2 22.1 16.1 No-smokers (n = 21) 20.6 5.2 34.5 12.0 23.4 18.6 25.9 19.0 Total (n = 41) 21.0 6.9 33.5 11.5 21.9 17.3 24.0 17.5 Two-way ANOVA for smoke and exposure effect on DNA damage showed a statistically significance for exposure relatively to % comets Fpg-untreated cells (P = 0.000), % comets Fpg-treated cells (P = 0.000) and TMenz (P = 0.004) while no significance was found for smoke and exposure smoke interaction. from 4 to8mg m 3 (Vaananen 2003; McClean 2004b) and pyrene levels ranging from 0.18 to 0.3 mg m 3 (McClean, 2004a, b). In particular our results on personal air samples show a prevalence of 2 3 rings PAH with the highest levels for phenanthrene and naphthalene, which is classified as a possible carcinogen for humans (group 2B) by IARC (2002). Although the exposure levels are low, our results evidence a correlation between urinary 1- OHP and occupational exposure to PAH, even after the exclusion of smoke effect. These findings confirm the suitability of urinary OH-pyrene as a biomarker of PAH absorbed dose also for low exposures. This study, even though performed on a restricted number of subjects, evaluates, after accurate measurements of exposure by environmental and biological monitoring, early genotoxic effect of asphalt fume. The fume is a complex mixture of potentially carcinogenic substances whose overall effect and mechanism are still not assessed. The evaluation of possible genotoxic effects induced by exposure to asphalt fume shows lack of SCE induction in exposed workers. This finding is in disagreement with Burgaz et al. (1998), who reported in Turkish road paving workers (mainly rakermen) an induction of SCE in lymphocytes. The different results could be explained by the lack of any protective equipment during paving operations in the Burgaz study and probably by different asphalt composition and exposure levels in the two studies. On the other hand the lack of SCE induction found in our study confirms the results reported by Jarvholm et al. (1999), in a similar study on Swedish paving workers showing PAH exposure levels comparable with ours. The lack of clear genotoxicity might be related to the too low sensitivity of SCE test to detect any effect at the low exposure levels found in our study that, however, seem to induce oxidative DNA damage in workers exposed to PAH during paving. This effect, evaluated by Fpg comet test, could be the first event of still repairable DNA damage induced by asphalt fume, consisting of PAH mixtures at low doses where naphthalene is one of the main compounds. In particular the study points out for this exposure an early oxidative DNA damage suggesting the induction of reactive oxygen species (ROS) as the first effect of chronic exposure to low doses of PAHs, confirming the hypothesized role of these species in the carcinogenic process induced by PAHs (Kumagai, 1997; Takano, 2002; Zhao, 2004). These results confirm the high sensitivity of the comet test to assess early DNA damage, including in exposure to low doses, and show that the Fpg comet test can furnish a useful contribution to the identification of biomarkers of early effect in workers exposed to complex mixtures of PAHs. Since it is not always possible to perform environmental or biological monitoring to establish the precise dose absorbed, the complete evaluation of exposure performed in our study on paving workers can be applied to other similar occupational conditions of pavers, but it is necessary to take into account that our subjects used protective equipment. In view of the growing numbers of workers exposed to complex PAH mixture at low doses and the rising environmental PAH pollution, particularly in urban areas, further investigations on mechanisms of

218 D. Cavallo et al. action of PAH mixtures and possible indicators of susceptibility and/or effect are needed and have worthwhile public health implications. CONCLUSIONS In conclusion our study shows early oxidative DNA damage and lack of SCE induction by exposure to asphalt fume in paving workers chronically exposed to low doses of PAH mixtures. Moreover our work shows the suitability of urinary 1-OH-pyrene as a biomarker of PAH exposure and of Fpg-modified comet test as a biomarker of early oxidative genotoxic effect. These findings suggest the combined use of both biomarkers, to furnish useful indications in terms of characterization, prevention and management of risk in occupational exposure to mixtures of potentially carcinogenic substances. REFERENCES Bender AP, Parker DL, Johnson RA et al. (1989) Minnesota highway maintenance worker study: cancer mortality. Am J Ind Med; 15: 545 56. Boffetta P, Jourenkova N, Gustavsson P. (1997) Cancer risk from occupational and environmental exposure to polycyclic aromatic hydrocarbons. Cancer Causes Control; 8: 444 72. Boffetta P, Burstyn I, Partanen T et al. (2003) Cancer mortality among European asphalt workers: an international epidemiological study. II. Exposure to bitumen fume and other agents. Am J Ind Med; 43: 28 9. Burgaz S, Erdem O, Karahalil B et al. (1998) Cytogenetic biomonitoring of workers exposed to bitumen fumes. Mutat Res; 419: 123 30. Chiazze L, Watkins DK, Amsel J. (1991) Asphalt and risk of cancer in man. Br J Ind Med; 48: 538 42. Collins AR, Duthie J, Dobson VL. (1999) Direct enzymic detection of endogenous oxidative base damage in human lymphocyte DNA. Carcinogenesis; 14: 1733 5. De Meo M, Genevois C, Brandt H et al. (1996) In vitro studies of the genotoxic effects of bitumen and coal-tar fume condensates: comparison of data obtained by mutagenicity testing and DNA adduct analysis by 32P-postlabelling. Chem Biol Interact; 101: 73 88. Elovaara E, Heikkila P, Pyy L et al. (1995) Significance of dermal and respiratory uptake in creosote workers: exposure to polycyclic aromatic hydrocarbons and urinary excretion of 1-hydroxypyrene. Occup Environ Med; 52: 196 203. Fuchs J, Hengstler JG, Boettler G et al. (1996) Primary DNA damage in peripheral mononuclear blood cells of workers exposed to bitumen-based products. Int Arch Occup Environ Health; 68: 141 6. IARC. (2001) IARC epidemiological study of cancer mortality among European aphalt workers. IARC Internal Report No. 01/003. Lyon: International Agency for Research on Cancer. IARC. (2002) Monograph on Some traditional herbal medicines, some mycotoxins, naphthalene and styrene Vol. 82, 2002. Lyon: International Agency for Research on Cancer. Jarvholm B, Nordstrom G, Hogstedt B et al. (1999) Exposure to polycyclic aromatic hydrocarbons and genotoxic effects on non smoking Swedish road pavement workers. Scand J Work Environ Health; 25: 131 6. King RW, Puzinauskas VP, Holsworth CE. (1984) Asphalt composition and health effects: a critical review. Washington DC: American Petroleum Institute Technical Publication. pp. 1 11. Kuljukka T, Vaaranrinta R, Veidebaum T et al. (1996) Exposure to PAH compounds among cokery workers in the oil shale industry. Environ Health Perspect; 104 (Suppl. 3): 539 41. Kumagai Y, Arimoto T, Shinyashiki M et al. (1997) Generation of reactive oxygen species during interaction of diesel exhaust particle components with NADPH-cytochrome P450 reductase and involvment of the bioactivation in the DNA damage. Free Radic Biol Med; 22: 479 87. Latt SA, Schreck RR. (1980) Sister chromatid exchange analysis. Am J Hum Genet; 32: 297 313. Machado ML, Beatty PW, Fetzer JC et al. (1993) Evaluation of the relationship between PAH content and mutagenic activity of fumes from roofing and paving asphalts and coal tar pitch. Fundam Appl Toxicol; 21: 492 9; Erratum (1994) Fundam Appl Toxicol; 22(2): 317. McClean MD, Rinehart RD, Ngo L et al. (2004a) Urinary 1-Hydroxypyrene and Polycyclic Aromatic Hydrocarbon Exposure Among Asphalt Paving Workers. Ann Occup Hyg; 48: 565 78. McClean MD, Rinehart RD, Ngo L et al. (2004b) Inhalation and dermal exposure among asphalt paving workers. Ann Occup Hyg; 48: 663 71. NIOSH. (1998) Polycyclic aromatic hydrocarbons by HPLC: Method 5506. NIOSH manual of analytical methods, 4th edn. Cincinnati, OH: NIOSH. NIOSH. (2000) Health effects of occupational exposure to asphalt. Cincinnati, OH: US Department of Health and Human Services. Partanen T, Boffetta P. (1994) Cancer risk in asphalt workers and roofers: review and meta-analysis of epidemiologic studies. Am J Ind Med; 26: 721 40. Partanen TJ, Boffetta P, Heikkila PR et al. (1995) Cancer risk for European asphalt workers. Scand J Work Environ Health; 21: 252 8. Perico A, Gottardi M, Boddi V et al. (2001) Assessment of exposure to polycyclic aromatic hydrocarbons in police in Florence, Italy, through personal air sampling and biological monitoring of the urinary metabolite 1-hydroxypyrene. Arch Environ Health; 56: 506 12. Pyy L, Makela M, Hakala E et al. (1997) Ambient and biological monitoring of exposure to polycyclic aromatic hydrocarbons at a coking plant. Sci Total Environ; 199: 151 8. Qian HW, Ong T, Whong WZ. (1996) Induction of micronuclei in cultured mammalian cells by fume condensates of roofing asphalt. Am J Ind Med; 29: 554 9. Reinke G, Swanson M, Paustenbach D et al. (2000) Chemical and mutagenic properties of asphalt fume condensates generated under laboratory and field conditions. Mutat Res; 469: 41 50. Sivak A, Niemeier R, Lynch D et al. (1997) Skin carcinogenicity of condensed asphalt roofing fumes and their fractions following dermal application to mice. Cancer Lett; 117: 113 23. Takano H, Yanagisawa R, Ichinose T et al. (2002) Lung expression of cytochrome P450 1A1 as a possible biomarker of exposure to diesel exhaust particles. Arch toxicol; 76: 146 51. Toraason M, Hayden C, Marlow D et al. (2001) DNA strand breaks, oxidative damage, and 1-OH pyrene in roofers with coal-tar pitch dust and/or asphalt fume exposure. Int Arch Occup Environ Health; 74: 396 404. Väänänen V, Hämeilä M, Kontsas H et al. (2003) Air concentrations and urinary metabolites of polycyclic aromatic hydrocarbons among paving and remixing workers. J Environ Monit; 5: 739 46. Zhao HW, Yin XJ, Frazer D et al. (2004) Effects of paving asphalt fume exposure on genotoxic and mutagenic activities in the rat lung. Mutat Res; 557: 137 49.