Tobacco Documents Online

~ram (1997) rained with relationship ttation Res. :cogitations me aberra- ~ Research, romosomal ical Report es, Science • MN from uman lym- od, Radiat. Effect of ,n analysis 4- 1-47. -'thotrexate 57 (1975) I~UTAT RES-GEN TOX EH 97 (C)ELSEVlER SCIEHOE BV HE ELSEVIER Mutation Research 391 (1997) 99-1 I0 Genetic Tox~cotogy and Environmental Mulagenesis The use of a urine mutagenicity assay in the monitoring of environmental exposure to genotoxins v Milena Cerna a,*, Anna Pastorkovfi a, Steven R. Myers b, Pavel R~Sssner Blanka Binkov~ c ~ National Institute of Public Health, Division of Environmental Health, ~robrrot'a 48, 100 42 Prague, Czech Republic b Universi~. of Louisville, School of Medicine, Department of Pharmacology and Toxicology, Louisville, KY 40292, USA hzstitute of ~rperimental Medicine, Academy of Sciences of the Czech Repttblic and Regional Institute of Hygiene of Central Bohemia, Wdehsk6 1083, 142 20 Prague, Czech Republic Received 3 December 1996; revised 12 February 1997; accepted 12 February 1997 ) 47-53. Abstract Urinary bacterial mutagenicity was used as a biomarker of exposure to ambient air pollution in a group of women working outdoors in the city of Teplice (TP; Northern Bohemia) with higher levels of air pollution than a similar group of women in the city of Prachatice (PT; Southern Bohemia). The Sabnonella ~.'phhnurium plate incorporation assay with the TA98 and YGI041 strains and microsuspension assay with the YGI041 strain were used for testing the urinary mutagenicity. PAH and their metabolites were analyzed by HPLC and GC/MS methods. The significantly higher values of most PAHs/metabolites detected in a TP group confirmed the differences of PAH exposures between both ~oups. In the plate incorporation assay, the TA98 strain was not able to detect the increase in urinary mutagenicity, but, for the YGI041 strain, the urinary mutagenicity was clearly determined with a significant difference in number of ¥G1041 + $9 revertants between the TP and PT groups. The microsuspension assay increased the mean response by about 10-fold over the standard plate test; however, no statistical difference between TP and PT groups was found due to high interindividual variability and small sample size. Comparing the urinary PAH/metabolites to urinary mutagenicity, significant correlations were observed between the plate incorporation mutagenicity results with the YGI04I revertants in the presence of metabolic activation and several of the urinary PAH/metabolites. On the contrary, in the microsuspension assay, several urinary PAH/metabolites correlated significantly with the YGI041 revertants only in the absence of metabolic activation. This may indicate the influence of different treatment conditions of assays on the urinary mutagenicity results. The results suggest the insufficient sensitivity of the TA98 tester strain to determinate low urinary level of mutagens. On the contrary, the use of the YGI041 tester strain increases the probability of detecting an effect of environmental exposure and seems to be applicable to biological monitoring. To definitely replace the standard plate incorporation assay with the microsuspension method is not possible without further comparative studies. Keywords: Biomonitoring; Ambient air exposure; Polycyclic aromatic hydrocarbon: Metab0iite: Urinary mutagenicity; Ames test; Kado assay; YG tester strain Corresponding author. Tel.: +42 (2) 6708-2378; Fax: +42 (2) 6708-2378. 1383-5718/97/$17.00 Copyright tD 1997 Elsevier Science B.V. All rights reserved. PII SI383-5718(97)00058-2 THIS ARTICLE 15 FOR IHDZVIDUAL USE OHLY AHD HAY HOT BE FURTHER REPRODUCED OR STORED ELECTRONICALLY ~ZTHOUT ~RZTTEN PER~ISSIOH FRO~ THE COPYRIGHT HOLDER. UHAUTHORZZE9 REPRODUCTIOH HAY RESULT ZN FINANCIAL AND OTHER pEHALTXES.

Ioo M. ~ernd et al. / Mutation Research 391 (1997) 99-/10 1. Introduction Exposure to genotoxic substances in the environ- ment is frequently viewed as a major risk factor to human health. To assess the risk, it is necessary to obtain biologically relevant exposure data. Different biological monitoring techniques have frequently been used for the evaluation of exposure to geno- toxic chemicals or their biologically effective dose Ill, Urine mutagenicity assessment is one of these traditional biomarker tools. The bacterial mutagenic- ity test (Ames test) has been most frequently used as a short-term test for about 20 years for this purpose [2,31. The first reports indicated the presence of muta- genicity in the urine of patients treated with either cytostatics [4] or other therapeutic agents, e.g., metronidazole [5]. Urine mutagenicity tests have since been often applied to monitor occupational exposure to genotoxic chemicals in different work- places [6]. Some of these studies suggested that urinary mutagen levels reflected occupational expo- sure to genotoxic carcinogens, whereas others em- phasized the importance of possible interference with many confounding factors not related to definite occupational exposure, like active or passive smok- ing, diet, drugs, and hobbies, which lowered the detectability of defined occupational exposure. The use of urine mutagenicity testing for monitor- ing human exposure to genotoxins from the general environment depends on the careful consideration of confounding exposures, given that the concentrations of mutagenic compounds in the environment are substantially lower than in occupational exposures. Urinary mutagen levels may also reflect complex exposure related to personal habits and lifestyle in addition to mutagenic pollutants in the environment [7]. Ho~'ever, the sensitivity of the urine mutagenic- ity test can be increased by using a microsuspension assay modification [8] and recently developed YG indicator strains capable of detecting mutagenic metabolites formed from nitroarenes and aromatic amines [9]. The aim of this study was to evaluate exposure to genotoxic contaminants in the environment by the means of both urinary mutagenicity assays (plate incorporation and microsuspension). The urinary mu- tagenicity assay results were completed with the results of the analysis of urinary polycyclic aromatic hydrocarbons (PAHs) and their metabolites per- formed in the same urine samples. The study subjects were from the city of Teplice" (TP) located in an area of Northern Bohemia pol- luted by brown coal combustion (annual average of air pollution data in ~xg/m3: SO2, 62.9; NO.,, 80; TSP, 110). A comparative group of subjects was from the city of Prachatice (PT) with a lower annual average of air pollution (SO2, 21.6; NO.~, 19.7; TSP, 38.5). The study was conducted within the Teplice Program [ 10]. 2. Materials and methods 2.1. Subjects and sampling A group of 30 healthy women with approximately 6 h of daily outdoor work exposure (postal workers, gardeners, nursery school teachers), aged 20-50 years and residents of the district for at least 5 years, was selected in each district. Prior to beginning the study, the informed consent of each subject was obtained. A qfiestionnaire was administered to each subject to determine their individual life style. The results of personal exposure monitoring (the last week in November, 1992 for TP, the first week in December, 1992 for PT) are presented elsewhere [10]. Spot urine samples from each person were col- lected twice: at the end of the working day and the next morning at the end of the personal exposure monitoring. The aliquots for the PAH metabolite analysis were separated. For the urinary mutagenicity analysis, both urine samples were pooled. Urinary creatinine was evaluated by the modification of the Jaffe colorimetric method [11]. All samples were stored frozen in polypropylene bottles until analysis. 2.2. Urinary. PAH metabolite analysis PAH and their metabolites were analyzed by high-performance liquid chromatography (HPLC) and gas chromatography/mass spectrometry (GC/MS). The method is briefly summarized here. Urine aliquots (5-7 ml) were adjusted to pH 5.0, incubated with 13-glucuronidase/arylsulphatase. The metabo cartrid~ of mett and the extractt reverse metabo metabo and ev: analysi The quantit; using ~ Table I Standar~d Sample I 9 10 14 15 16

lic aromatic ~olites per- of Teplice .hernia pol- average of NO.~, 80; bjects was wet annual 19.7; TSP, he Teplice !1 !| !1 !1 M. ~emd et aL/ Mutation Research 391 (1997) 99- I I0 101 metabolites were extracted using the Sep-Pak Ct8 cartridges (Waters, Milford, MA), eluted with 10 ml of methanol and evaporated at 60°C under nitrogen and the residue was dissolved in 200 Ixl methanol for HPLC and GC/MS analyses. The aliquots of the extracted metabolites (2-5 ~zl) were analyzed by reversed-phase HPLC. The scanning fluorescent detector was programmed to detect the specific metabolites quantitated. The individual PAH/ metabolites fractionated by HPLC were collected and evaporated to dryness under nitrogen prior to the analysis by GC/MS selective ion monitoring (SIM). The PAH/metabolite data reported here were quantitated by three separate determinations each using both HPLC and GC/MS/SIM and the mean v,qlues are reported as ng of PAH metabolite per mg creatinine. 2.3. Urinar3' mutagenici~. Fifty-two urine samples obtained from 20 non- smokers and 8 smokers (3-20 cigarettes/day) from Teplice, and from 24 non-smokers from Prachatice, were tested for mutagenicity in the plate incorpora- -tion assay. Two samples from TP and 6 samples from PT were excluded because of the insufficient amount of urine. The urine samples were thawed at room tempera- ture and filtered through a paper filter. One hundred and fifty millilitres of each sample was incubated Table 1 Standard plating test: individual mutagenicity results in the Teplice group "oximately 1 workers, Sample 'd 20-50 i in°'ars, . Smoking status TA98 - $9 TA98 + $9 YGI041 L $9 YGI041 + $9 Rev/ml Rev/mg Rev/ml Rev/mg urine creatinine urine creatinine Rev/ml Rev/mg Rev/ml Rev/mg urine creatinine urine creatinine the 1 2 ~ject was ! - I3 d to each 4 tyle. The 5 (the last I6 week in ~- 7 ;Isewhere il . 8 9 10 ,' and the 12 exposure 14 15 ,etabolite ~__i 16 agenicity 17 Urinary ' 19 ,n of the 20 es were 22 analysis. 23 24 26 27 • zed by 28 (HPLC) 29 ometry ~!ili 30 n 1.4 0.7 0.7 0.4 n 1.2 0.5 1.5 0.6 n ND ND 1.3 1.7 n ND ND ND ND n 0.2 0.1 0.8 0.2 s/15 a 5.0 4.2 2.3 1.9 n ND ND 2.2 ! .9 s / 10 1.6 0.5 0.6 0.2 n ND ND ND ND s/15 1.1 0.7 0.1 0.1 s/20 3.7 2.2 2.8 1.7 n ND ND ND ND n 0.3 0.4 0.2 0.3 s/7 ND ND ND ND n 0.3 0.4 ND ND n 0.6 0.2 1.5 0.6 n 3.0 1.6 3.4 1.8 n ND 0.03 0.02 0.01 n 0.04 0.03 0.5 0.5 n 1.6 1.7 0.04 0.05 s/10 ND ND 1.6 0.8 n 1.5 0.8 0.9 0.5 n 2.0 1.1 2.6 1.5 n 2.1 2.0 0.6 0.5 n ND ND 0.8 0.7 s/10 ND ND 0.4 0.6 s/8 0.3 0.4 t.8 2.5 n 0.5 11.6 0.5 0.6 ND ND 5.7 3.0 5.6 2.1 ND ND - ND ND 5.8 7.6 ND ND 3.3 2.5 ND ND ND ND 2.7 2.3 11.3 9.4 ND ND 5.1 4.5 11.9 4.0 7.9 2.6 ND ND 1. I 0.9 ND ND 11.3 7.1 ND ND 19.6 11.8 1.4 1.3 22.0 20.0 53.8 61.1 32.3 36.7 4.9 6.7 2.8 3.9 ND ND 19.6 25.1 2.3 1.0 4.2 1.7 20.5 11.0 2.6 1.4 ND ND 19.8 13.5 10.1 8.8 9.6 8.4 2.1 2.3 33.4 37.3 ND ND 54.3 28.3 0.9 0.5 15.3 8.3 61.0 35.1 67.2 38.7 ND ND 6.5 6.0 ND ND ND ND 1.0 1.6 8.2 13.0 3.8 5.2 19,4 26.5 ND ND ND ND Bold values indicate statistically significant, " n cigarettes/day. B~-- linear slope values ( p < 0.05). ND, no dose-response effect observed.

102 M. ~ernd et al. / Mutation Research 391 (1997) 99- 110 with 50 U/ml of [3-glucuronidase/arylsulphatase. (Sigma) for 3 h at 37°C. After the incubation, the mutagens were separated and concentrated by pass- ing the urine through 6 ml C18 (octadecyl) resin columns using a vacuum manifold (Baker 12) system [12]. Methanol (8 ml) was used to extract the muta- gens, the volume normalized to 10 ml with methanol and stored frozen at -20°C. 2.4. Bioassay sample preparation For the plate incorporation mutagenicity assay, the extractable mass dissolved in 10 ml of methanol was evaporated under a stream of nitrogen to near dryness and solvent exchanged into dimethylsulph- oxide (DMSO) (Merck) to give 10 IAI DMSO per ml bioassay vial containing 5 I~1 of DMSO each. Methanol was evaporated under nitrogen in a water bath at 35°C to dryness. 2.5. Mutagenicity assays The plate incorporation assay was carried out as described [13] with the Salmonella o'phimurium tester strains TA98, kindly provided by Prof. B.N. Ames (Berkeley, CA, USA) and YGI041, a deriva- tive of the TA98 parent.strain with elevated levels of both nitroreductase and O-acetyltransferase activi- ties, kindly donated by Drs. T. Nohmi and M. Watanabe (National Institute of Hygienic Sciences, Japan) It4]. The microsuspension assay was con- ducted using the strain YG1041 only. Both assays before b mixture The pro TA98, [ 1 ovemigI ampicill added tt the sele In tl were te: control~ crolitre: urine e: of the ~ molten urine. For the microsuspension mutagenicity assay, doses of methanol extract corresponding to 0.25, 0.5, 1.5 and 3 ml of urine volume were pipetted into the Table 2 Standard plating test: individual mutagenicity results in the Prachatice group were performed with and without metabolic activa- plates. tion using a liver $9 fraction for metabolic activation [" I The which was prepared from male rats pretreated 5 days .. YG1041-$9 YG1041÷$9 [il I ~ Rev/ml Rev/mg Rev/ml Rev/mg __ urine creatinine urine creatinine c Summar2. test! TA98 ~ YGIO,! Sample Smoking no. status TA98-$9 TA98 + $9 Rev/ml Rev/mg Rev/ml Rev/mg urine creatinine urine creatinine 31 n 0.1 0.1 0.7 0.4 ND ND ND 32 n 0.3 0.2 3.7 2.5 ND ND ND 34 n ND ND ND ND ND ND ND 35 n 2.9 1.2 0.05 0.02 ND ND 7.6 36 n 0.5 0.7 0.I 0.1 ND ND ND 37 n 1.3 0.7 1~ 0.6 3.0 1.5 0.2 38 n 0.1 0.1 0.1 0.1 ND ND ~.1 39 n 2.6 1.8 1.6 1.0 ND ND ND ~ n 0.1 0.1 1.7 2.4 ND ND ND 41 n ND ND 0.1 0.1 7.5 6.8 43 42 n 0.6 0.4 1.3 0.7 13.0 7.3 1.8 43 n 2.9 3.9 0.9 1.3 0.3 0.3 0.~ 46 n 2.9 3.0 0.1 0.1 &9 9.0 12.4 47 n ND ND 0.2 0.1 3.0 1.5 8.0 48 n 2.9 2.4 0.2 0.2 ND ND ND ~ n 2.2 1.4 0.5 0.3 8.0 4.8 2.4 51 n 0.I 0.5 0.6 2.3 ND ND l.l 52 n 0.5 0.4 03 ~2 48.0 36.6 0.7 53 n ND ND 0.2 0.2 ND ND 6.3 55 n 0.4 0.2 0.3 0.2 ND ND 12.5 56 n 1.9 0.9 ND ND ND ND 0.8 57 n 1.7 1.4 0.7 0.5 I.I 0.9 9.4 58 n 0.3 0.3 0.8 0.7 ND ND 3.9 60 n 1.2 1.0 2.9 2.3 ND ND ND Bold values indicate statistically significant B~ - linear slope values ( p < 0.05). ND. no dose-response effect observed. ND ND ND 3.1 ND 0.1 81.2 ND ND 3.9 1.0 0.04 12.6 4.1 ND 1.5 4.2 0.6 9.9 6.3 0.4 7.5 3.6 ND "-- Micros~ YGI0 a Mean ~ Medi: " I c The r _ t~ :• _a Signi

in a water || ied out as 'himuriurn~l 'rof. B.N. a deriva- levels°f I i ~e activi- and M. " Sciences, vas con- hassays I I c actiVa- ctivation d5days ! ! nine M. ~ern6 et al. / Mutation Research 391 (1997) 99-110 103 before being killed with 500 mg/kg Delor 103 (PCB mixture comparable to Aroclor 1254) in corn oil. The protein concentration [15] was 22.6 mg/ml. For TA98, ampicillin (25 lxg/ml) was added to the overnight culture in Oxoid II nutrient broth. Both ampicillin and kanamycin (25 ixg/ml of each) were added to the cultivation medium for YGI041 to keep the selective pressure on the YG1041 strain. In the plate incorporation assay, urine extracts were tested in duplicate at 4 doses corresponding to 1, 3, 7 and 10 ml of urine. All positive and negative controls were tested in triplicate. One hundred mi- crolitres of the tester strain, 100 p~l of each dose of urine extract and 500/~1 of $9 mix containing 30 I~1 of the $9 fraction/ml $9 mix were added to 2 ml of molten top agar and overlaid onto minimal agar plates. The microsuspension assay was performed with modification [i6]. The YG1041 cells were concen- trated (5 × ) by centrifugation (10000 × g, 4°C, 10 min) and resuspended in an ice-cold phosphate- buffer. Doses of 0.25, 0.5, 1.5 and 3 ml equivalent of urine were used with duplicate determination for urine extracts and triplicate for controls. To the glass tubes containing 5 p~l of DMSO urine extract, 50 ~1 of concentrated strain, and 50 p,l of $9 mix (or 0.015 M. sodium phosphate buffer) were added. After the suspension was preincubated at 37°C for 90 rain, 2 ml of top agar was added and the content of the tube poured onto minimal medium. The mean values of revertants/plate in the nega- tive control (DMSO) for the plate incorporation as- say were as follows: 15.4 (TA98 - $9), 15.7 (TA98 + $9), 119.9 (¥G1041 - $9) and 100 (YG1041 + $9). The positive control without metabolic activa- tion (p-nitro-o-phenylenediamine, 10 tzg/plate) Table 3 Summary ~tatistics (means, medians and ranges) for urinary mutagenicity data (rev/mg creatinine) in plate incorporation and microsuspen- sion tests Strains Teplice smokers Teplice non-smokers Prachatice non-smokers (n = 8) (n = 20) (n = 24). Plate incorporation test TA98 - $9 1.0 5:1.47 a 0.52 ± 0.63 0.86 + 1,02 0.475 (0.0-4.2) b 0.315 (0.0-2.0) 0,45 (0.0-3.9) 3/5 ¢ 5/15 4/20 TA98 + $9 0.98 5:0.94 0.59 + 0.63 0.68 5:0.83 0.72 (0.0-2.5) 0.49 (0.0-1.9) 0.275 (0.0-2.5) 1/7 4/16 2/22 YG1041 - $9 2.48 + 2.60 6.16 + 15.24 2.87 + 7.67 1.92 (0.0-6.7) 0.0 (0.0-61.1) 0.0 (0.0-36.6) 3/5 11/9 15/9 YGI041 + $9 12.85 5:9.67 10,78 5:13,37 5,83 5:16,43 a 10.63 (2.6-28.3) 5.23 (0.0-38.7) 0.79 (0.0-81.2) 0/8 3/17 8/16 Microsuspension test YGI041 - $9 192.03 5:134.32 86.15 5:128.32 115.91 + 126.67 238,2 (3.3-338.9) 36.61 (2.4-419.3) 70.66 (0.5-434.3) o/5 o/14 o/18 YGI041 + $9 109.46 5:61.88 70.86 + 120.03 86.21 + 133.99 117.4 (36.3-187.4) 21.28 (0.0-409.0) 41.73 (0.0-502) 0/5 2/12 2/16 Mean + SD, Median (minimum-maximum). The relation between the number of samples without and with mutagenic responses. Significant difference between Prachatice and Teplice (whole group p = 0.010, non-smokers p = 0.086).

r 104 M. ~ern6 et aL / Mutation Research 391 (1997) 99-/I0 gave 407.4 and 1662 rev/plate, with metabolic acti- vation (2-aminofluorene 5 ~zg/plate) 848.2 and 925.6 rev/plate in TA98 and YG1041, respectively. For the microsuspension assay, the negative con- trol revertants/plate were as follows: YGI041 with- out $9 143.2, YGI04I with $9 143.0. The positive control without metabolic activation (p-nitro-o- phenylenediamine, 1 ~g/plate) were 549 and with metabolic activation (2-aminofluorene, 5 ~g/plate) 535 revertants/plate. The revertafits were counted after 72-h incubation at 37°C using a Biotran II colony counter. 2.6. Statistics The maximum mutagenic potency (rev/ml urine) was determined using the GeneTox manager soft- . ware [17]. To correct for interindividual differences Table 4 Microsuspension in urine volume, the creatinine content of the urine was used to determine the mutagenic activity in revertants per mg creatinine. The linear slope values (Bt) for both rev/ml urine and rev/mg creatinine were calculated from Bernstein model [18]. Statistical analysis was performed on the B1 slope values using the STATGRAPHICS Plus 7.0 package (Magnuistics, Rockville, MD). Non-parametric methods (The Mann-Whitney rank sum U-test and Kruskall-Wallis one-way analysis of variance) were genici chosen for the data that did not follow a normal distribution. Correlations were performed by the Spearman rank correlation test. The statistical significance for differences in num- ber of positive responses (defined as B, GeneTox value with p < 0.05) was assessed by ANOVA Pro- gram. assay: individual mmagenicity results in the Teplice group Sample Smoking YGI041 - $9 YGI041 + $9 Rev/mt urine Rev/mg creatinine Rev/ml urine Rev/mg creatinine I n NT NT NT NT 2 n NT NT NT NT 3 n 51.9 68.3 33.1 43.4 4 n NT NT NT "NT 5 n NT NT biT NT 6 s/15 a NT NT NT NT 7 n 60.3 53.3 44.2 39.2 8 s/10 NT NT NT NT 9 n 21.5 17.9 bit NT 10 s/15 NT NT NT NT I 1 s/20 5.4 3,3 99.2 59.8 12 n 55.0 50.5 22.0 20.0 14 n NT NT NT NT 15 s/7 194.4 270.0 84.5 117.4 16 n 244.6 310.6 NT NT 17 n 30. I 12.4 42.4 17.4 19 n 6.0 3.3 94.9 50,9 20 n 8.7 5.9 10.8 7.3 21 n 479.4 419.3 128.2 ! 12.2 22 n 161.8 179.5 368.6 409.0 23 s/10 650.6 338.9 281.20 146.4 24 n 4.3 24 0.5 0.3 25 n NT NT NT NT 26 n 49. l 45.0 292,0 265,4 27 n I 1.3 9.6 26.5 22.5 28 s/10 150.6 238.2 118.2 187.4 29 s/8 811.4 109.8 26.8 36.3 30 n 27.5 28.3 4.3 4.5 (roum Iand r expre: with iwitho predi,_ consI~ ity e,- preset genic Slight Bold values indicate statistically significant B~ ~ n cigarettes/day. - linear slope values ( p < 0.05). NT. not tested: NDI no dose-response effect observed.

M. ~ernd et al. / Mutation Research 391 (1997) 99-110 105 : activity in slope values 3.1. Plate incorporation assay ~gcreatinine iI [~ 18]. Tables 1 and 2 show the individual urine muta- the Bt slope genicity results expressed as linear slope values 7.0 package , (round to one decimal point) for revertants per ml 1-parametric 1 ![ and revertants per mg creatinine. The two ways ~ U-test and expressing the urinary mutagenicity correlated well =iance) were with each other (r=0.998- 1.000). The results 'v a normal 1 !1 without any dose response, where the maximum ~ed by the predicted potency was not possible to count, were considered to be negative. ces in hum- The summary data concerning urinary mutagenic- ', GeneTox ! II ity expressed as revertants per mg creatinine are '~'OVA Pro-~ presented in Table 3. Only the .results with muta- genic responses were included in the statistics. ii il Slightly higher mean and median numbers of TA98 Microst 32 i! -[ 136353437 40 41 revertants both with and without metabolic activation observed in the group of Teplice smokers were not statistically significant as compared to the Teplice or Prachatice non-smoker groups. A higher average number of induced revertants obtained with the. YG1041 strain compared to the TA98 results (about one order of magnitude difference) indicates an in- creased capability of the YG1041 strain to detect urine mutagenicity in environmentally exposed groups. This fact was supported by the finding that the mean number of YG1041 + $9 revertants in the PT group was significantly lower in comparison to the whole TP (p=0.0t0) as well as to the TP non-smokers (p =0.086). YGI041- $9 revertants correlated with YG1041 + $9 results for the whole group (r = 0.295, p = 0.0350) as well as for the non-smokers (r=0.330, p =0.0304 in the Spear- man rank correlation test. observed. Microsuspension assay: individual mutagenicity results in the Prachatice group Smoking YG 1041 - $9 Rev/ml urine n 18.5 n 0.7 n 69.1 n 72.1 n 312.7 n NT n 13.4 n NT n 25.7 n NT n 61.3 n 161.9 n NT n 217.6 n NT n 181.6 YGI041 + $9 Rev/mg creatinine Rev/ml urine Rev/mg creatinine ,!~'[14243 46 47 [-ll 48 50 n N'T I1 51 n 80.3 t -, 52 n 137.1 53 n 186.1 55 n NT 56 n 2.3 58 n 234.4. ~L~._ 60 n 94.9 I1 Bold values indicate statistically significant | 9.7 75.1 39.3 0.5 11.9 8.2 68.1 80.2 79.1 29.8 61.2 25.2 434.3 30.6 42.5 NT NT bit 19.1 10.6 15.0 bit NT t'Wl" 35.8 NT NT 34.4 39.5 22.3 217.0 374.5 502.0 NT NT NT 111.3 80.9 41.4 NT NT NT 109.3 306.1 184.2 NT NT NT 316.3 ND ND 104.9 77.8 59.5 290.7 228.5 357.0 NT NT NT 1.1 85.3 42.1 18.1 89.4 71.3 212.8 16.5 15.1 73.2 62.0 47.6 linear slope values (p < 0.05). NT, not tested. ND, no dose-response effect observed.

106 M. ~ern6 et al. / Mutation Research 391 (1997) 99-110 3.2. Microsuspension assay Due to insufficient volumes of urine specimens, only a limited number of urine samples were tested using the YG1041 tester strain only (Tables 4 and 5). Most of individual mutagenicity results were signifi- cant according to the B1 GeneTox linear slope value, but the values show large interindividual variances. The increase in mutagenicity started at a lower urine volume/plate than in the standard plate test and at higher urine volumes, toxic effect were often ob- served. The mean numbers of induced revertants are about l0 times higher when compared to the plate incorporation test results (Table 3). Neither quantita- tive nor qualitative differences in the microsuspen- sion assay results between the Teplice and Prachatice groups were found. The Spearman rank correlation between the YG1041 - $9 and YG1041 + $9 results revealed a statistical significance for all the overall study (r = 0.377, p -- 0.024) as well as for all non- smokers (r = 0.339, p = 0.059). Besides this, signif- icant correlation was observed between the YGI041 -$9 revertants in plate incorporation and the YG1041 + $9 revertants in Kado assays for both the overall study (r=0.341, p =0.041) and the non- smokers (r = 0.378, p = 0.035). 3.3. PAH / metabolites The data reported here (Table 6) include four parent.PAHs: chrysene, benzo{a]pyrene, dibenz[a, h]anthracene, pyrene and their hydroxylated metabo- lites, and the total of 29 PAH/metabolites. The comparison of the PAH metabolites and parent PAH in urine between the TP and PT overall groups, as well as between TP and PT non-smokers, showed significant differences for total PAH/metabolites and for some parent PAHs (chrysene, benzo[a]pyrene, pyrene) and their metabolites (l-hy- droxypyrene, l-hydroxychrysene). There was no sig- nificant difference between the smokers and non- smokers in Teplice in excretion of PAHs or PAH metabolites. Table 6 Summary of statistics (means, medians and ranges) for urinary PAH/metabolites (ng/mg'creatinine) Teplice smokers Teplice non-smokers Prachatice non-smokers (n = 8) (n = 20) (n = 24) Total PAH/total metabolites d 236.80 + 103.1 a 230.80 + 71.8 126.30 + 39.2 207.40 b (123.3-418) ~ 241.80 (96-357) 131.30 (60-215) Benzo[a]pyrene 7.16 + 4.36 5.37 + 5.76 3.05 + 1.59 5.38 (2.7-15) 5.12 (1.6-21.7) 2.78 e (0.5-7.0) Chrysene 17.51 5:10.89 13.86 5:7.87 6.46 5:3.33 16.23 (5.3-32.7) 14.07 (1.7-35) 6.18 e (1.2-17.5) Dibenz[ a,h]anthracene 3.03 5:2.38 3.07 5:2.64 1.28 5:0.84 2.82 (0.4-7.8) 2.47 (0.6-12.5) 1.15 ~ (0.2-3.5) Pyrene 5.50 + 5.10 14.27 + 10.25 6.72 + 3.11 12.29 (6.28-28.5) 11.98 (4.3-51.2) 6.07 e (2.4-15.1) 3-Hydroxybenzo[a]pyrene 8.48 + 6,87 5.40 5:3.63 4,87 5:1.99 6.22 ( 1.8 - 19.7) 5.32 ( 1.0-15.4) 4.60 ( 1.3 -9.4) l-Hydroxychrysene 16.51 5: I 1.16 13,96 5:9.31 4.29 + 2,13 13.97 (6.33-33.0) I 1.55 (3.8-47.5) 4.11 ~ (1.2-9.7) 7,14-Hydroxydibenz[a.h]anthracene 10.44 5:5.29 10.90 5:5.54 8.68 :t: 3.51 9.21 (2.4-18.8) 9.91 (3.8-28,4) 8.10 (2,5-15.3) l-Hydroxypyrene 5,51 5:5.10 6.54 ~. 3,42 2.06 + 1.53 3.46 ( 1.4-16,6) 5.88 (2.3-14.4) 181 ~ (0.5-7.5) Mean 5: SD; b Median; ~ Range; d Urinary PAH and metabolites are expressed as ng/mg creatinine; ~ p < 0.001. ![ ,il Table 7 Corre!at "-11 Tot~P? Benzol o Chrysen Dibenz[ Pyrene 3-Hydro -Hydro l 7,1.4-Hv l-Hyd~ Statistic: 3.4. C urinar. The gemcit ~dy: '~In: mutag~ with t the no statisti tween of PAl A low~ when Table 8 Correlm Benzo[. PO _~ Chryse! O Dibenz! t~ Pyrene 0"~ 3-Hydt ¢.,O _~ l-Hyd~ ~ 7,14-H 03 I -Hyd~

and the ~r both the the non- l-! II I! M. ~ernd et aL / Mutation Research 39l (1997) 99-110 Table 7 Correlation between urine mutagenicity results and urine PAH/metabolites (Spearman rank correlation) - overall study Standard plate test (n = 52) Kado test (n ~ 37) 107 I ~ YGI041 + $9 YGI041 - $9 -I Total.PAH/total metabolites 0.371 (0.009) 0.248 (0.142) lude four dibenz[a, ! metabo- ites. The rent PAH Benzo[a]pyrene Chrysene Dibenz[ a.h ]anthracene Pyrene 3-Hydroxybenzo[ a]pyrene l-Hydroxychrysene 7,14-Hydroxydibenz[ a.h ]anthracene 1 -Hydroxypyrene 0.467 (0.001) 0.563 (0.000) 0.270 (0.062) 0.407 (0.005) 0.350 (0.016) 0.273 (0.11 I) 0.352 (0.040) 0.251 (0.144) 0.309 (0.071) 0.373 (0.030) roups, as • showed ztabolites lrysene, es (1-hy- s no sig- md non- or PAH II ii II Statistical significance (p) is given in parentheses. 3.4. Correlations between urinary, m~tagenicit3.' and urinaD" PAH / metabolites The Spearman rank correlation analysis was used to determine the relationship between urinary muta- genicity and PAH/metabolite in urine for the overall study and for all non-smokers. In the plate incorporation assay, the TA98- $9 mutagenicity results appeared to correlate marginally with the urine 3-hydroxybenzo[a]pyrene values for the non-smokers (r=0.377, p--0.016). Overall, a statistically significant correlation was observed be- tween the YG 1041 + $9 revertants and the majority of PAH/metabolites excreted in the urine (Table 7). A lower, but still significant correlation was obtained when the data for smokers were excluded (Table 8). In the microsuspension assay, a significant corre- lation between the YGI041 -$9 induced revertants and chrysene, 3-hydroxybenzo[a]pyrene and 7,14- hydroxydibenz[a,h]anthracene urinary values for the overall study was documented. In non-smokers, a correlation for the same strain was found for 7,14- hydroxydibenz[a,h]anthracene only. No correlation between the PAH/metabolites and YGI041 + $9 mutagenicity was observed. 4. Discussion The study design was based on the assumption that the Teplice population is exposed to higher concentrations of PAH and other mutagenic products Table 8 Correlation between urine mutagenicity results and urine PAH/metabolites (Spearman rank correlation) - overall for non-smokers I-.:1 Standard plate test (n = 44) Kado test (n = 32) YGI041 + $9 YGI041 - $9 Total PAH/total metabolites 0.287 (0.063) Benzo[a]pyrene Chrysene 0.362 (0.020) Dibenz[a.h ]anthracene 0.524 {0.001 ) Pyrene 3-Hydroxybenzo[ a]pyrene l-Hydroxychrysene 0.304 (0.052) 7,14-Hydrox,vdibenz[ a.h ]anthracene 1 -Hydroxypyrene 0.305 (0.051 ) 0.320 (0.086) Statistical significance (p) is given in parentheses.

M. ~ernd et al. / Mutation Research 39l (1997) 99-ll0 of the incomplete combustion of fossil fuels in the ambient air than the Prachatice population and that this exposure is likely to be manifested in the pres- ence of mutagenic activity in urine. Mutagenicity analysis of urine is generally accepted to be a suit- able method for the demonstration of an occupa- tional exposure to mutagenic chemicals that may be excreted directly into urine or their excreted muta- genic metabolites [19]. The power of this method for determining the extent of human exposure to geno- toxins in the general environment with a markedly lower level of pollutants has not been demonstrated as yet. A significantly higher level of urinary PAH/metabolites detected in the Teplice group clearly showed a higher PAH exposure in this group. The significantly higher number of YG1041 + $9 revertants in the overall TP group and TP non- smokers manifested the marked exposure to genotox- ins compared with the PT group. Likewise, the num- ber of positive and negative individual mutagenic responses for ¥GI041 + $9 was found to be higher in the TP group. The YG104I + $9 results in the plate incorporation test correlated well with the ma- jority of urinary PAH metabolites in the Spearman rank correlation test. Smoking ranks among the factors which can con- tribute to individual urine mutagenicity. In fact, the majority of investigators found smokers to show higher urine mutagenicity [20]. Similarly, slightly, but non-significantly, higher numbers of all induced revertants both with and without metabolic activation were observed in the urine of 8 light smokers in Teplice. However, a significant increase the number of YG1041 + $9 revertants was also observed in Teplice non-smokers. Besides this, no significant difference in the DNA adduct level between non- smokers and smokers suggested that cigarette smok- ing did not substantially confound environmental exposure in this study [21]. To increase the probability of detecting an effect of ambient air exposure, the TA98 strain used as a standard frameshift mutagen indicator for urinary mutagens [22,23] was complemented with YG1041 which is supposed to be more sensitive in detecting urinary mutagens [24]. The higher number of the induced YG1041 revertant and the significant differ- ence between the TP and PT groups compared to the TA98 results supported this suggestion. A correlation between the YG1041 results with and without metabolic activation was observed. However, in a few urinary samples the excess induction of TA98 revertants prevailed. This could be attributable to the fact that several environmental mutagens, including PAH representative, benzo[a]pyrene and 7,12-di- methylbenz[a]anthracene were described to express almost the same or even lower mutagenicity to YG1041 when compared to conventional TA98 strain [14]. -. Provided that PAHs requiring metabolic activa- tion for the expression of mutagenicity are the domi- nant exposure factor, the presence of S9-dependent urine mutagenicity was expected. The dose-depen- dent direct mutagenic effect observed in individual urine specimens indicated that substances other than indirect-acting PAHs were also present. In addition to the presence of PAH in the ambient air, nitrated polycyclic aromatic hydrocarbons (nitroPAHs) are also widely distributed in the environment as a result of incomplete combustion processes. This fact is supported by the finding of the direct-acting muta- genicity of air particle exhaust in TP [25]. It is possible to hssume that the nitroderivatives could also be responsible for increased urinary mutagenic- ity. Nitroarenes are known to be potent bacterial mutagens [26]. They require metabolic activation by both nitroreductase and acetyltransferase present in indicator bacteria as well as in mammalian cells for exerting their mutagenicity. The overproduction of nitroreductases and O-acetyltransferases in the YG strains results in their higher ability to transform nitroderivatives into mutagenic metabolites. Re- cently, it was shown [27] that l-nitropyrene metabo- lites exhibited both direct-acting and S9-dependent mutagenicity which is further enhanced by O-acety- lation, l-Acetamidopyren-6-ol, the major urinary metabolite of l-nitropyrene, required both $9 and O-acetyltransferase activities. Indeed, the increase of indirect-acting YG1041 mutagenic metabolites ob- served in the study may indicate the excretion of nitroPAH metabolites. The two main difficulties observed using urinary mutagenicity tests are due to the small volume of urine and to the tiny amount of (pro)mutagens ex- creted. The micromethod has been shown to be about I0 times more sensitive than the standard plate assay ![ [ [ [ [ [ [81 a, extra. dard stand obtai The 1 muta canc, inter: the ~ resul twee betw ii lyze~ ' meta the I mere preii strai| the ~ indir A ~_~geni, tage expl: facto - , proc the ~ of .u~ tran~ The a si~ sure: be e t'-l! PAl seei in a met [