Science of the Total Environment 408 (2010) 1166–1173

Contents lists available at ScienceDirect

Science of the Total Environment

j ourna l homepage: www.e lsev ie r.com/ locate /sc i totenv

1-Hydroxypyrene and 3-hydroxybenzo[a]pyrene as biomarkers of exposure to PAHin various environmental exposure situations

Ariane Leroyer a,b,c,⁎, Fanny Jeandel d, Anne Maitre e, Mike Howsam a,b, Dominique Deplanque a,b,f,Muriel Mazzuca a,b, Catherine Nisse a,b,c

a Univ Lille Nord de France, F-59000 Lille, Franceb UDSL, EA 2690, GIP CERESTE, F-59000 Lille, Francec CHULille, Service de Pathologie Professionnelle et Environnement, F-59037 Lille, Franced Laboratoire de l'INRS, Département Polluants et Santé, F-54519 Vandoeuvre-lès-Nancy, Francee Equipe EPSP, Environnement et Prédiction de la Santé des Populations, Laboratoire TIMC (UMR 5525), CHRU Grenoble, F-38700 La Tronche, Francef CHULille, Centre d'Investigation Clinique, CIC 9301 INSERM, F-59037 Lille, France

⁎ Corresponding author. Département Universitaire dFaculté de médecine, pôle recherche 1, place de VerduTel.: +33 3 20 62 68 73; fax: +33 3 20 88 36 64.

E-mail address: ariane.leroyer@univ-lille2.fr (A. Lero

0048-9697/$ – see front matter © 2009 Elsevier B.V. Adoi:10.1016/j.scitotenv.2009.10.073

a b s t r a c t

a r t i c l e i n f o

Article history:

Received 25 June 2009Received in revised form 20 October 2009Accepted 27 October 2009Available online 17 November 2009

Keywords:Polycyclic aromatic hydrocarbonsEnvironmental exposureBiomarkers1-Hydroxypyrene3-Hydroxybenzo[a]pyrene

Objective: To evaluate the utility of urinary 3-hydroxybenzo[a]pyrene (3-OH-B[a]P) as a biomarker ofexposure to polycyclic aromatic hydrocarbons (PAHs) in various environmental exposure scenariosalongside the more usually studied 1-hydroxypyrene (1-OH-Pyr).Methods: Two groups of 15 and 10 non smoking, healthy men and women, were exposed for approximately6h to ambient air at two outdoor locations close to metallurgical industries, and at one indoor location in anurban setting. Atmospheric measurements of 16 “priority” PAHs were carried out during each exposure.Urinary 1-OH-Pyr and 3-OH-B[a]P were also analysed, samples being taken the morning before exposure, atthe end of exposure, then 4 and 15h after the end of exposure. Urinary 1-OH-Pyr and 3-OH-B[a]P wereanalysed using high performance liquid chromatograph with fluorescence detection. Limits of detection(LOD) were 0.092 nmol/L and 0.28 pmol/L for 1-OH-Pyr and 3-OH-B[a]P respectively.Results: Average ambient air concentrations varied from 0.27 to 2.87 ng/m3 for pyrene, 0.04 to 1.20 ng/m3 for

B[a]P, and from 70.0 to 910.6 ng/m3 for the sum of the 16 PAH (ΣPAHs). Concentrations of 1-OH-Pyr werenot correlated with atmospheric concentrations of PAHs to which subjects were exposed, nor with theconcentrations of 3-OH-B[a]P. Nearly 80% of measurements of 3-OH-B[a]P were lower than the LOD and norelationship between atmospheric concentrations and urinary metabolites was observable. However, thepercentage of post-exposure values of 3-OH-B[a]P greater than the LOD increased significantly with themedian of atmospheric concentrations of Pyr, B[a]P and ΣPAH at the different sites (test of linear trend,pb0.02 in all cases).Conclusion: Although we used very sensitive techniques for the measurements of urinary metabolites,especially for 3-OH-B[a]P, neither 1-OH-Pyr nor 3-OH-B[a]P were an unequivocal biomarker of exposure toatmospheric PAHs in environmental exposure scenarios relevant to the general population. It would beinteresting to investigate other urinary monohydroxy PAH metabolites in this context.

© 2009 Elsevier B.V. All rights reserved.

1. Introduction

Exposure to polycyclic aromatic hydrocarbons (PAHs) may in-crease the risk of cancer in humans (Boström et al., 2002; IARC,International Agency for Research on Cancer, 2006). The evaluation ofthis exposure is often performed using measurements of atmosphericconcentrations to which subjects are exposed, sometimes in conjunc-tion with measurements of the classically-studied urinary biomarker

e Médecine et Santé au Travail,n, 59045 LILLE Cedex, France.

yer).

ll rights reserved.

of PAH exposure 1-hydroxypyrene (1-OH-Pyr). The advantage of theuse of a biomarker to assess exposure is that it provides an integratedmeasurement of several routes of exposure to PAHs (i.e. by inhalation,dermal, or digestive routes), while a concurrent disadvantage ofthis approach is that it does not distinguish between different sourcesof exposure. In the case of PAHs, dermal exposure can be importantin occupational exposure scenarios, while food represents the mostsignificant source of exposure to PAHs for non smoking subjectsnot occupationally exposed (ATSDR, Agency for Toxic Substancesand Disease Registry, 1995; Fiala et al., 2001). The analysis of urinary1-OH-Pyr, the primary metabolite of pyrene, is widely used inassessment of occupational exposure to PAHs; however, this bio-marker only provides a direct measure of exposure to pyrene, one of

1167A. Leroyer et al. / Science of the Total Environment 408 (2010) 1166–1173

the least toxic of the sixteen PAHs designated as “priority” con-taminants by the United States Environmental Protection Agency,subsequently adopted by the European Union (Keith and Telliard,1979). The greatest carcinogenic risk to humans among the 16“priority” PAHs is thought to be posed by compounds such as benzo[a]pyrene (B[a]P) and several other PAHs classified as probable humancarcinogens by International Agency of Research on Cancer (IARC)and by European Union (EU) (IARC, International Agency for Researchon Cancer, 2006).

The past few years have seen increased study of urinary bio-markers of PAH exposure other than 1-OH-Pyr, that may providedirect evidence of exposure to PAHs of greater toxicological relevancethan pyrene. Among them, 3-hydroxybenzo[a]pyrene (3-OH-B[a]P),the main primary metabolite of B[a]P present in urine, has receivedparticular attention (Gündel and Angerer, 2000; Gündel et al., 2000;Simon et al., 2000; Lafontaine et al., 2006). There are, however, severalfactors that make the search for biomarkers of exposure to the moretoxic PAHs via the analysis of their primary, urinary metabolitessignificantly more difficult than the analysis of 1-OH-Pyr. Studies haveshown that exposure to pyrene is generally considerably greaterthan to B[a]P (Park et al., 2002; Ravindra et al., 2006), and that this,together with the possibility of only a single hydroxy metabolite forpyrene compared with two primary hydroxy metabolites of B[a]P,as well as the predominance of the faecal over the urinary route ofexcretion for the metabolites of this compound, contribute to the verylow levels of urinary 3-OH-B[a]P found (ATSDR, Agency for ToxicSubstances and Disease Registry, 1995). Lafontaine et al. (2006)studied urinary 1-OH-Pyr and 3-OH-B[a]P in 27 smokers and 27 non-smokers: concentrations of 1-OH-Pyr were higher than the limit ofdetection (LOD, 5 ng/L) in all cases, while the concentrations of 3-OH-B[a]P were lower than the LOD (50 pg/L) in 20 of the non-smokers(range: bLOD-224 pg/24 h) and 9 of the smokers (range: bLOD-367 pg/24 h). The National Health and Nutrition Examination Survey(NHANES) evaluated the exposure to PAHs of a sample of theAmerican population (2748 participants in 2001–2002) (CDC, 2005,pp.79–134): measurements of more than 20 urinary PAH metaboliteswere carried out by GC/MS, though more than half of the 3-OH-B[a]Pmeasurements were lower than the LOD (10.5 ng/L), the 75thpercentile being ca. 33 ng/L (CDC, 2005, pp. 127–128). During thevalidation of a high performance liquid chromatography fluorometric(HPLC-fluo.) analytical method, Hollender et al., (2000) studied theurines of 10 non-exposed children: while 1-OH-Pyr was quantifiablein the urine of all the children, 3-OH-B[a]P was detected in only oneurine but was not quantifiable (LOD: 0.19 nmol/L, LOQ: 1.12 nmol/L).

The use of urinary 3-OH-B[a]P as a biomarker of exposure to B[a]Pin the general population is in the early stages of development,therefore, and the main aim of this study was to contribute to thisprocess. We chose to evaluate its use in various environmentalexposure scenarios which the general population, with no occupa-tional PAH exposure, may experience. Three exposure scenarios weredefined — industrial and semi-industrial air in outdoor settings, andurban air in an indoor setting — and urine samples were collected forparallel analysis of urinary 1-OH-Pyr and 3-OH-B[a]P from partici-pants exposed in a controlled manner before exposure, and at threeintervals after exposure.

2. Methods

2.1. General organization of the study

The study protocol was first validated by the Lille branch of theregional Nord-Pas de Calais ethical committee. The subjects wererecruited from the volunteers enrolled in the Clinical InvestigationCentre of the Lille Regional University Hospital (CHRU). An initialrecruitment visit was undertaken, during which the subjects werefully informed of the study and inclusion/exclusion criteria were

verified, after which each volunteer signed an informed consent form.Twenty five people were divided into 2 groups (G1 and G2 composed,respectively of 15 and 10 volunteers). Each groupwas exposed for oneday, once a week, over 3 consecutive weeks in October 2006 (G1 onTuesdays, G2 on Thursdays), each week corresponding to a differentexposure scenario. For the 48h prior to each exposure, the subjectswere requested to avoid contact with smokers, open fires, consump-tion of drugs, roasted or smoked foodstuffs, vitamin supplements andthe use of machines with a petrol engine. Active or passive smokingwas checked by measuring the urinary cotinine concentration beforeeach exposure, while other recommendations were evaluated by aquestionnaire.

Transfer to the site of exposure was by bus and each exposurelasted approximately 6h, with food and drink supplied for thevolunteers being identical.

2.2. Exposure scenarios

Three exposure scenarios were defined at three separate locations:exposure to outdoor air at two locations in industrial settings (sites Aand B) and to indoor air in an urban setting (site C). Sites A and Bwereclose to iron and steel industrial complexes in the Nord-Pas de Calaisregion of northern France which have been reported to emit 461 and25tonnes of non-methane volatile organic compounds (NMVOCs) peryear, respectively (DRIRE Nord-Pas de Calais, 2007; data for theproportion of these emissions that were PAHs was not available). Theexposure to indoor air (site C) was carried out in an administrativebuilding (non-smoking area), at the Faculty of Medicine, University ofLille 2, Lille being a large conurbation in the Nord-Pas de Calais regionof northern France.

2.3. Study population

The subjects were healthy adults, male or female, non-smokers,20–50 years old, not occupationally exposed to PAHs. The exclusioncriteria were: a clinically detectable pathology during the clinicalexamination prior to inclusion, abnormal blood tests (complete bloodcount, creatinine, γ-glutamyl transferase, transaminases), subjects onlong term medication (except oral contraceptives), exposure topassive smoking in the domestic environment, occupational ordomestic exposure to known or potential carcinogens (classes I and2A as defined by the IARC), expectant or nursing mothers. Residencein an urban area or in the vicinity of an industrial site did not in itselfconstitute an exclusion factor, but each situation was studied prior toinclusion in an attempt to avoid subjects likely to be chronicallyexposed to elevated levels of PAHs in their domestic settings.

2.4. Sampling and analysis of atmospheric PAHs

To evaluate the exposure of each group of subjects for each ofthese exposure scenarios, three air samples were taken during theapproximately 6h of presence of the volunteers, according to themethods recommended in France (INRS, 2007; AFNOR NF X 43–294,1995). An opaque, polypropylene cassette held a 25 mm diameterTeflon filter (VWR®) and was connected in series to a XAD2 tube(Supelco® Orbo 43), thewhole connected to a pump regulated to a flowrate of 1–1.5 L/min. Thus, PAHs ostensibly present in the particulatephase of the atmosphere were trapped on the filter and those in theoperationally defined gas phase were trapped by the XAD2 cartridge.The consistency of the air flow (and by inference, the total volume of airsampled) during sampling was checked by comparing rates at thebeginning and end of sampling: a variation greater than 8% resulted inthe exclusion of the sample. After sampling, the filter and cartridgeassemblies were transported to laboratory at +4 °C then stored in thedark at −20 °C until analysis.

1168 A. Leroyer et al. / Science of the Total Environment 408 (2010) 1166–1173

Atmospheric PAHs were analysed at the Grenoble Regional Univer-sityHospital (CHRUG). Fifteenmillilitres ofHPLCgradedichloromethanewas added to each filter. After ultrasonic extraction for 30 min, the filterwas removed, and the solvent was evaporated using a rotary evaporatorat 30 °C. The residuewas taken up in 1 mL of HPLC grade acetonitrile foranalysis. For the XAD2 tube, PAHs was extracted by elution with 2 mLacetonitrile. Seven PAHs predominantly present in the gas phase ofthe atmosphere (Naphtalene (Naph), Acenaphthene (Acen), Fluorene(Fluo), Phenanthrene (Phe), Anthracene (Ant), Fluoranthene (Flua),Pyrene (Pyr)), and 9 PAHs more or less associated with particulatematter in the atmosphere (Fluoranthene (Flua), Pyrene (Pyr), Benzo[a]anthracene (B[a]A), Chrysene (Chr), Benzo[b]fluoranthene (B[b]F),Benzo[k]fluoranthene (B[k]F), Benzo[a]pyrene (B[a]P), Dibenzo[a,h]anthracene (D[a,h]A), Benzo[ghi]perylene (B[ghi]P)) were measured byHPLC-fluo. The systemwas composed of aHPLCAlliance2695model anda fluorescence detector 2475 Multi λ from Waters®. Chromatographicseparation was realized by injecting 30 µL on a C18-PAH column(Supelco® 250×2.1 mm ID, particle size: 5 µm) with an acetonitrile/water gradient. The column oven temperature was set to 32 °C. Fordetection of the operationally-defined gaseous PAHs, the wavelengths(excitation/emission) used were: 276/323 nm to detect Naph, Acen andFluo., 249/362 nm for Phe, 250/400 nm for Ant, 285/450 nmfor Flua, and333/390 nm for Pyr. For detection of the operationally-defined partic-ulate PAHs, the wavelengths used were: 285/450 nm for Flua, 333/390 nmfor Pyr, 265/380 nmforB[a]A andChr, 295/425 nmforB[b]F, and296/405 nm for B[k]F, B[a]P, D[a,h]A and B[ghi]P. Limits of detection(LODs) varied from0.05 to 0.30 ng/filter or tube. The sumof gaseous andparticulate compounds (ΣPAH) was also determined, and the PAH“profile” (relative contribution of each 2-, 3-, 4-, 5- and 6-ring homo-logues to ΣPAH) was calculated.

2.5. Sampling and analysis of urine

Urine samples were taken at 4 different times: before exposure(T1), at the end of exposure (T2), 4h after the end of the exposure (T3)and 15h after the end of the exposure (T4— first micturation the nextmorning). Three aliquots were immediately taken from each sample:one aliquot of 2 ml for creatinine (stored and transported at 4 °C), oneof 10 ml for 1-OH-Pyr analysis (stored and transported at−20 °C) andone of 4 ml for 3-OH-B[a]P analysis, which was immediately passedthrough a Sep-Pak C18 Plus cartridge (stored and transportedat +4 °C), following the “Uriprel” protocol described in Simon et al.,(2003). A supplementary aliquotwas taken from each T1 urine samplefor cotinine analysis.

Urinary creatinine was measured by spectrophotometry withJaffe's method and all concentrations of 1-OH-Pyr and 3-OH-B[a]Pwere normalised to urinary creatinine. Urinary cotinine was mea-sured by immunoenzymatic technique, using an ELISA kit (detectionlimit was 10 µg/L).

Urinary 1-OH-Pyr was analysed in the EPSP-TIMC laboratory,CHRUG. Aliquots of 2 mL of each urine sample were adjusted to pH4.6 by adding 2 mL of sodium acetate buffer (0.1 M, pH 4.66, Merck®,Darmstadt, Germany). Sampleswere then incubated for 12 hwith 20 µLof glucuronidase-arylsulphatase (Roche® diagnostics, Basel, Switzer-land) at 37 °C in a water bath, before being injected on a Waters®

(Guyancourt, France) system composed of a 510 pump, a solventselector, a valve with 10 positions and 2 tracks, allowing an on-lineextraction of 1-OH-Pyr. The pre-column usedwas a C18 reversed-phaseSupelguard (20×4.6 mm i.d., 5 µm particle size; Sigma-Aldrich, Saint-Quentin Fallavier, France). After injection of urine samples, the pre-column was rinsed for 7.2 min. with a water-methanol mixture (H2O:MeOH, 50:50 v/v), and 1-OHPwas then eluted withMeOH and injectedon a HPLC-fluo. instrument. The system was composed of an Alliance2695 HPLC model and a 2475 Multi λ fluorescence detector fromWaters®. The analytical column was a C18 reversed-phase Licrospher(250×4 mmi.d., 5 µmparticle size) fromMerck. The gradientwas linear

from 15% to 94% of MeOH in H2O over 10 min, and the flow rate was1 mL/min. The excitation wavelength of the fluorescence detector was333 nmand theemissionwavelengthwas390 nm. The LODwas20 ng/L(0.092 nmol/L). The laboratory participates in the German Society forOccupational and Environmental Medicine external quality controlsystem for 1-OH-Pyr.

Urinary 3-OH-B[a]P was analysed in the Pollutants and HealthLaboratory of the INRS, Vandoeuvre. The 3-OH-B[a]P was eluted atabout 3 mL/min from the “Uriprel” HAP cartridge with 2 mL of MeOHthen with 2 mL of a mixture of MeOH and sodium acetate buffer(0.4 M, pH=4.65) in the ratio 1:99 (v/v). After shaking, an aliquot of1.25 mL was mixed with 2.5 mL of the buffer mixture and 15 µLglucuronidase-arylsulphatase (Helix pomatia from Merck, Darmstedt)was added. After incubation for 1 h, 3 mL of the hydrolysed solutionwere directly introduced in the chromatographic system on aminiature precolumn loop.

The HPLC-fluo. system consisted of two chromatographic pumps(Model LC-10AT, Shimadzu, Kyoto, Japan), four automated switchingvalves (Model 7000, Rheodyne, Berkeley, CA, USA, one being equippedwith a miniature precolumn loop), three purification columns, oneanalytical column and a fluorescence detector equipped with a Xe lamp(Model RF-10AXL, Shimadzu) set to excitation and emission wave-lengths of 425 and 460 nm, respectively. The entire system wascontrolled by a programmable microprocessor unit. The output signalwas recorded with a Chromjet integrator (Thermo Separation Products,Fremont, CA, USA). The precolumn loop was a 2 cm×0.32 cm i.d.stainless-steel guard column cartridge (Upchurch Scientific, OakHarbor,WA, USA) filled with approximately 100 mg of 40 µm Bondesil C1(Varian, Palo Alto, CA, USA). The three purification columns (C1, C2and C3) were 5×0.32 cm i.d., 7.5×0.46 cm i.d. stainless-steel and7.5×0.46 cm i.d. stainless-steel tubes, respectively, and packed with :(C1) 5 µm Uptispher C8 (Interchim, Montluçon, France), (C2) 5 µmUptispher CN (Interchim, Montluçon, France) and (C3) 5 µm UptispherPhenyl (Interchim, Montluçon, France). The analytical column (C4) ofpreferencewas a 10×0.32 cm i.d. stainless-steel tube packed with 5 µmUptispher PAH(Interchim,Montluçon, France), otherwise a20×0.32 cmi.d. stainless-steel tubepackedwith5 µmVydac 201 TPB (The SeparationGroup, Hesperia, CA, USA) could be used. With the exception of theprecolumn loop,whichwasdryfilled, all the columnswerepacked in thelaboratory at 4×107Pa using a mixture of 95% ethanol, 2-propanol andtoluene (1:1:1, v/v/v) as slurry solvent, followed by methanol and thenwater as displacement liquids. A schematic diagram of the switchingsystem, representing one of the successive steps of the analyticalprocedure, is shown in Fig. 1. The whole switching system withindicative switching event time table has been detailed in a previouspaper (Simon et al., 2000). The mobile phase used for the purificationcolumns was a mixture of water and methanol (30:70, v/v) containingcitric acid (0.005 M) and ascorbic acid (5 mg/L). The mobile phase usedfor the analytical column was a mixture of the same solvents in theratio 10:90 (v/v), also containing citric and ascorbic acids (0.005 M and5 mg/L, respectively). All columns were eluted in isocratic mode at0.5 mL/min. Prior to analysis, the mobile phases were degassed withhelium for 30min andkept under a reducedpressure (5×104 Pa)heliumatmosphere during analysis. The LOD were 0.08 ng/L (0.28 pmol/L) andthe limit of quantification (LOQ)was 0.25 ng/L (0.75 pmol/L) for 3-OH-B[a]P.

2.6. Statistical analyses

To assure sample validity, urines with excessive physiologicaldilution or concentration were excluded from the statistical analyses(creatinine values lower than 0.3 g/L (2.65 mmol/L) or higher than3.0 g/L (26.52 mmol/L)) (WHO, 1996). Concentrations of 1-OH-Pyrlower than the LOD were assigned values equal to half the LOD whileconcentrations of 3-OH-B[a]P lower than the LOD were not taken intoaccount for the statistical analyses because only 20% exceeded the

Fig. 1. Schematic diagram of the switching system for the analysis of 3-OH-B[a]P (representation of one of the successive steps) (PC: precolumn loop; P1 and P2: pumps;V1: injection valve; V2, V3, V4: switching valves; C1, C2, C3: purification columns; C4: analytical column; D: detector).

Table 1Average ambient air concentrations (ng/m3) of the combined, operationally definedgaseous and particulate phase PAHs at the 3 sites for the 2 groups (n=3 for eachsituation, except for Site A G1 and G2, and Site C G2).

Site/group

Ab Bb Cc

Ringsa PAH G1 G2 G1 G2 G1 G2

2 Naph 879.7 142.4 53.0 78.0 110.0 84.9Acen 0.27 0.66 5.33 2.55 1.56 1.24Fluo 5.20 4.27 4.19 4.20 3.79 3.51

3 Phe 8.91 10.06 4.73 5.97 6.45 6.90Ant 0.43 0.50 0.06 0.18 0.03 0.04Flua 4.46 2.57 0.90 1.90 1.23 1.10

4 Pyr 2.87 2.04 0.27 0.91 0.59 0.42B[a]A 0.65 0.38 0.10 0.21 0.12 0.09Chr 2.48 1.00 0.89 0.85 0.18 0.09B[b]F 2.18 0.59 0.10 0.54 0.35 0.13B[k]F 0.96 0.33 0.05 0.20 0.11 0.04

5 B[a]P 1.20 0.81 0.05 0.30 0.19 0.04D[a,h]A 0.14 0.09 0.10 0.08 0.08 0.09

6 B[ghi]P 1.19 0.78 0.19 0.41 0.32 0.09ΣPAH 910.6 166.5 70.0 96.3 125.0 98.7

a Number of aromatic rings in PAH molecule.b Exposure to outdoor, industrial air.c Exposure to indoor air.

1169A. Leroyer et al. / Science of the Total Environment 408 (2010) 1166–1173

LOD. To evaluate the relative utility of the two biomarkers of PAHexposure studied here, two statistical methods were used (SASsoftware v9.1): firstly, comparisons between the urinary concentra-tions before exposure (T1) and each of the three post-exposuresamples T2, T3 and T4 (Wilcoxon's test); secondly, analyses of thelinear regressions taking atmospheric PAH measurements (gaseousplus particulate pyrene, B[a]P, and concentrations of the sum of the 16PAHs (ΣPAH)) as independent variables, and the logarithms of theurinary concentrations of 1-OH-Pyr and 3-OH-B[a]P as dependentvariables.

3. Results

3.1. Study population

The 25 volunteers had a median age of 29 years (20–47 years old),there were 18 men and 7 women, and all presented normal resultsin their clinical and biological examinations (haematological, renaland hepatic assessment). None was a smoker or exposed to passivesmoking and none had occupational exposure to PAHs. Nineteenof them resided in urban environments (though were not residentin streets characterised as having heavy traffic), 5 in peri-urbanenvironments, and 1 in a rural setting.

3.2. Atmospheric PAH concentrations

Three sample replicates were excluded because of discrepancies inthe flow rate between the beginning and end of sampling (Site A —

Group 1 and Group 2, and Site C — Group 2), and the averages of thetriplicate (or remaining duplicate) air samples are presented in Table 1.The highest concentrations of all individual PAHs (with the exception ofAce), as well as ΣPAH, were observed at Site A, especially for Group 1which was exposed to particularly high concentrations of naphthalene.Atmospheric concentrations at Site A for Group 1were, 910.6 ng/m3 forΣPAH and ranged from 70.0 to 166.5 ng/m3 in all other exposuresituations; Pyr was 2.87 ng/m3 and ranged from 0.27 to 2.04 ng/m3 inother exposure situations; B[a]P was 1.20 ng/m3 and ranged from 0.04to 0.81 ng/m3 in other exposure situations; Naph concentration was879.7 ng/m3 and ranged from 53.0 to 142.4 ng/m3 in other exposuresituations. Interestingly, Sites B and C exhibited quite similar airconcentrations, and exposure to interior air (Site C) was not always tothe lowest concentrations of PAHs.

In addition, Sites B and C had remarkably similar PAH profilescompared with Site A. Ignoring naphthalene (the very high concen-trations of which in relation to the other PAHs measured hereconfounded comparisons of atmospheric PAH concentrations), Site Aexhibited percentage contributions to ΣPAH by 4-ring PAHs of ca. 20%,

and by 5- or 6-ring PAHs of ca. 4% each; comparing this withcontributions to ΣPAH by 4-ring PAHs of ca. 10% and by 5- or 6-ringPAHs of ca. 1–2% each at both Site B and Site C, it was evident thatthese heavier PAHs contributed about twice as much to the PAHprofile at Site A than at Sites B and C.

3.3. Urinary biomarkers of exposure

During the entire experiment, one urine was unavailable and 15/299 samples (5%) with extreme creatinine concentrations (b0.3 orN3.0 g/L) were excluded to eliminate potentially unreliable results.None of the concentration of cotinine measured in the pre-exposureurine samples (T1) was greater than the limit of detection.

For the 1-OH-Pyr analyses, 4 of the 284 measurements were lowerthan the LOD (1.4%). Fig. 2 summarizes the results for urinaryconcentrations of 1-OH-Pyr. The median 1-OH-Pyr concentrations forthepopulationat thedifferent sample timeswere39.8 (range4.9–215.6,n=72), 41.5 (range 13.5–113.0, n=72), 38.8 (range 7.8–265.4, n=70)and 39.9 nmol/mol creatinine (range 12.9–204.2, n=70) before ex-posure (T1), immediately after exposure (T2), 4h (T3) and 15h (T4)after the end of the exposure, respectively.

Two hundred and twenty four of the 284 (79%) analyses had 3-OH-B[a]P concentrations lower than the LOD. The medians of the

Fig. 2. Box and whisker plot (min, P25, P50, P75, max) of urinary 1-OH-Pyr concentrations (nmol/mol creatinine) by site, group and sampling time (T1 before exposure, T2 end ofexposure, T3 end of exposure +4 h, T4 end of exposure +15 h).

1170 A. Leroyer et al. / Science of the Total Environment 408 (2010) 1166–1173

3-OH-B[a]P values measured at the different times pre- and post-exposure were all lower than the LOD (range: bLOD— 0.095 nmol/molcreatinine, n=284). Quantifiable concentrations of 3-OH-B[a]P aresummarized Fig. 3 for Groups 1 and 2 combined.

3.4. Variations of urinary biomarkers by exposure scenario and group

Fig. 4 summarizes the paired differences between urinary 1-OH-Pyr concentrations observed before and after exposure, for eachexposure scenario and for each group of subjects. Significantdifferences in 1-OH-Pyr concentrations after exposure were observedonly at Site B: amongGroup1 subjects, a significant increase (p=0.01)was observed immediately after exposure (T2); for Group 2 subjects, asignificant decrease was observed in all three post-exposure samples(p=0.04 for T2, pb0.01 for T3 and p=0.01 for T4) relative to the pre-exposure concentrations. It should be noted that pre-exposure urinary1-OH-Pyr concentrations were higher in Group 2 than in Group 1(medians 75.4 and 34.8 nmol/mol creatinine, respectively).

Fig. 3. Box and whisker plot (min, P25, P50, P75, max) of quantifiable urinary 3-OH-B[a]P (exposure, T2 end of the exposure, T3 end of exposure +4 h, T4 end of exposure +15 h).

The insufficient number of samples with quantifiable urinary 3-OH-B[a]P made it impossible to study paired differences between pre-and post-exposure, even when groups were combined. Qualitativeanalysis of the results (Table 2) revealed no significant differences inthe percentage of quantifiable concentrations before and after ex-posure (all sites taken together and post-exposure samples com-bined), or between the total percentage of quantifiable 3-OH-B[a]Presulting from the exposure scenarios (Sites A, B and C consideredindividually). Nevertheless, a significantly higher percentage of post-exposure samples (T2+T3+T4 taken together) had quantifiableurinary 3-OH-B[a]P at Site A compared with Sites B and C combined(p=0.01). This analysis was felt to be justified owing to thesimilarities between these latter two sites compared with Site A,both in their atmospheric PAH concentrations as well as theiratmospheric PAH “profile”.

No statistically significant correlation was observed between 1-OH-Pyr and 3-OH-B[a]P concentrations, either globally or between thevarious post-exposure sampling times.

nmol/mol creatinine) by site (min, P25, P50, P75, max) and sampling time (T1 before

Fig. 4. Box and whisker plot (min, P25, P50, P75, max) of the paired differences between the pre- and post-exposure concentrations of urinary 1-OH-Pyr (nmol/mol creatinine) bysite, group and sampling time (T1 before exposure, T2 end of the exposure, T3 end of exposure +4 h, T4 end of exposure +15 h). Differences significantly different from zero areindicated by an asterisk.

1171A. Leroyer et al. / Science of the Total Environment 408 (2010) 1166–1173

3.5. Variations of urinary biomarkers with atmospheric concentrations

Linear regressions for the relationships between the atmosphericconcentrations of parent PAHs and concentrations of urinary PAHmetabolites (Table 3) exhibited no significant relationship betweenatmospheric (gas+particulate) Pyr, B[a]P or ΣPAH concentrations andpost-exposure concentrations of urinary 1-OH-Pyr or 3-OH-B[a]P (theuse of other indicators of exposure such as T2-T1 or (T2–T1)/T1 did notmodify these results, neither did adjustments for gender and bodymassindex of the subjects). However, the percentage of post-exposure values(T2, T3 or T4) of 3-OH-B[a]P greater than the LOD increased significantlywith mean atmospheric concentrations of Pyr, B[a]P and ΣPAH at thedifferent sites (Table 2, test of linear trend, pb0.02 in all cases).

Table 2Number and percentage (in parentheses) of 3-OH-B[a]P values greater than the limit oconcentrations during exposure (average of Groups 1 and 2).

Before exposure Post exposure

(T1) (T2) (T3) (T4

Site A 8 (32%) 7 (29%) 7 (28%) 7 (Site B 8 (35%) 4 (16%) 6 (27%) 2 (Site C 3 (12%) 1 (4%) 1 (4%) 6 (Sites B+C 11 (23%) 5 (10%) 7 (16%) 8 (

a pb0.01, sites B and C combined, post-exposure combined, versus site A.

Table 3Univariate regression analysis of log transformed atmospheric PAH concentrations (gas+pconcentrations for the three post-exposure sampling times (T2, T3 and T4).

Atmospheric concentrations

Urinary concentrations/sampling time Log [Pyr]

β (±se)a pb

Log [1-OH-Pyr]T2 — End of exposure −0.06 (±0.06) 0.28T3 — End+4 h 0.04 (±0.09) 0.65T4 — End+15 h −0.11 (±0.07) 0.14

Log [3-OH-B[a]P]T2 — End of exposure −0.18 (±0.13) 0.18T3 — End+4 h −0.04 (±0.10) 0.73T4 — End+15 h 0.14 (±0.28) 0.28

a is the simple regression coefficient (se: standard error) estimated by the linear regressb p value for the parameter estimate (testing the difference to zero).

4. Discussion and conclusions

Our aim in this studywas to assess thepotential for the routineuse of 3-OH-B[a]P as a biomarker of environmental PAH exposure, it being moretoxicologically relevant than themore traditionally studied1-OH-Pyrwhichwas alsomeasured inparallel for the three exposure scenarios studiedhere.The choice of Sites A and B was made according to known emissions ofNMVOCs in the Nord-Pas-de-Calais Area. Themajority of NMVOCs emittedto theatmosphereare indeedmonocyclic compounds, but themetallurgicalindustries operating at these sites are also characterised by significantemissions of PAHs. The study of exposure to indoor air (Site C)was deemedworthwhile since our study was concerned with “everyday” exposurescenarios relevant to places where people actually live.

f detection (LOD) at each site and sampling time with selected atmospheric PAH

Combined post-exp. Atmospheric concentrations (ng/m3)

) (T2+T3+T4) Pyr B[a]P ΣPAH

28%) 21 (28%) 2.54 1.04 6139%) 12 (17%) 0.53 0.15 8027%) 8 (12%) 0.52 0.13 11418%) 20 (14%)a 0.52 0.14 97

articulate) on log transformed, creatinine adjusted urinary 1-OH-Pyr and 3 OH-B[a]P

Log [B[a]P] Log [ΣPAH]

β (±se)a pb β (±se)a pb

−0.04 (±0.04) 0.37 −0.03 (±0.06) 0.620.02 (±0.06) 0.72 0.01 (±0.08) 0.91

−0.07 (±0.05) 0.13 −0.09 (±0.07) 0.18

−0.12 (±0.09) 0.19 −0.11 (±0.35) 0.35−0.04 (±0.07) 0.61 0.05 (±0.10) 0.61

0.06 (±0.07) 0.42 −0.02 (±0.14) 0.86

ion of atmospheric (X) on urinary (Y) concentrations: Y=α+βX.

1172 A. Leroyer et al. / Science of the Total Environment 408 (2010) 1166–1173

Itwasnot possible to control thevariability of the prevailingweatherconditions for different groups of subjects, nor was it possible to ensurethe same atmospheric PAH concentrations for each group, making itpotentially difficult to compare groups and exposure scenarios. Never-theless, only subjects without occupational exposure were selected andinstructions relating to the avoidance of certain activities/foods prior totheir exposure were on the whole strictly observed.

The median values observed here for urinary 1-OH-Pyr variedbetween ca. 30 and 50 nmol/mol creatinine. These were comparablewith those in various studies of the general population, or of subjectswithout occupational exposure or known exposure to sources of PAHsin food (Van Rooij et al., 1994; Viau et al., 1995; Saint-Amour et al.,2000; Bouchard et al., 2001; Huang et al., 2004; CDC, 2005; Lafontaineet al., 2006).

Our study failed to demonstrate a relationship between concentra-tions of urinary 1-OH-Pyr and pyrene atmospheric concentrations. Nosignificant increasewas observed post-exposure for either of the groupsat Site A, the site at which atmospheric concentrations of Pyr were thehighest. At Site B, Groups 1 and 2 exhibited different trends in post-exposure concentrations, but both nevertheless exhibited a generaldecrease after exposure. At Site C, no significant difference between pre-and post-exposure was observed, with atmospheric levels of Pyrcomparable with those at Site B. Lastly, the trends in excretion of 1-OH-Pyr over time did not show any statistically significant increases in1-OH-Pyr 4h after the endof exposure,whereas a peak is often observedat this time owing to the elimination half-life of thismetabolite (ATSDR,Agency for Toxic Substances and Disease Registry, 1995). These obser-vations are in agreement with other studies. Cirillo et al. (2006), withatmospheric concentrations of Pyr a little lower than ours, observedno relationship between the atmospheric concentrations of Pyr andthe urinary concentrations of 1-OH-Pyr in 7–9 years old children, theauthors concluding that, for low levels of exposure, it is food whichcontributesmost to PAH exposure. Gündel et al. (2000) andMaître et al.(2003) studied occupational exposure to PAHs at levels close to thosewe observed and also noted the absence of a correlation between theatmospheric exposure and urinary 1-OH-Pyr concentrations.

Regarding urinary 3-OH-B[a]P, 79% of the values in our study waslower than the limit of detection, making quantitative statistical analysisof the results impossible: these results were nevertheless similar topublishedobservations incomparablepopulations (CDC,2005; Lafontaineet al., 2006). Gündel et al. (2000), studying occupational exposure tosimilar atmospheric levels inour study, also reporteda lackof a correlationbetween atmospheric PAH concentrations and urinary 3-OH-B[a]Pconcentrations. Lafontaine et al., (2004), again studying occupationalexposure but at levels of B[a]P approximately 500 times greater than inour study, observed a very significant correlation between the atmo-spheric B[a]P concentrations and urinary 3-OH-B[a]P concentrations. Inour study, the percentage of samples with post-exposure concentrationshigher than the LOD increased with the median atmospheric concentra-tions of Pyr, B[a]P and ΣPAH, andwhen Site A was compared with Sites Band C combined, a significantly higher proportion of samples at Site Aexhibited detectable concentrations of 3-OH-B[a]P.

Different factors may have confounded the relationship betweenatmospheric concentration of pyrene (or ΣPAH) to which subjectswere exposed and their urinary 1-OH-Pyr concentrations: the relativelow levels of the 3 exposure situations, the short duration of exposure,the inter-individual variations (differences in PAH metabolism) andother uncontrolled PAH exposures may all play a confounding role.Thus, exposure of the subjects to different atmospheric PAHconcentrations in the three exposure scenarios studied here was notdirectly observable using measurements of their urinary 1-OH-Pyr.On the other hand, the proportion of the population with detectableurinary 3-OH-B[a]P increased in a linear fashion with exposure tohigher atmospheric concentrations of B[a]P. While far from confirm-ing the utility of urinary 3-OH-B[a]P as a biomarker of exposure, thesedata suggest that it is nevertheless useful to pursue its analysis, in

tandemwith measurements of urinary 1-OH-Pyr, at the very least as aqualitative indicator of exposure to B[a]P.

European Directive No. 2004/107/CE, concerning environmentalexposure to PAHs (OJEU, 2005), states that B[a]P should be used as amarker of carcinogenic risk to humans and imposes a target value of1 ng/m3 for B[a]P (annual average for PM10 fraction of the atmosphere).The highest values of B[a]P observed in this study were 1.2 and 0.8 ng/m3, both at Site A, but the air samples in this study represent a muchlarger fraction of the atmosphere than just the PM10 particulate fractioncovered by the legislation. The levels of exposure to B[a]P observed inour study were thus well within the target limits in Europe, but arenevertheless low compared with some studies in the literature (e.g. 10times lower than in the study of Mucha et al., 2006 and 500 times lowerthan the study by Lafontaine et al., 2004). In considering the exposure ofthe wider, non-occupationally exposed population to PAHs, however,they are likely to be fairly typical, indicating that the analysis of 1-OH-Pyr and 3-OH-B[a]P in urine as biomarkers of exposure to airborne PAHsin the general population will remain equivocal.

In conclusion, the atmospheric PAH concentrations to which thesubjects were exposed in the three scenarios in this study were withinthe range of those reported in the literature for the general population,as well as being below the target value of European legislation, butwere nevertheless insufficient to produce observable variations inurinary 1-OH-Pyr or 3-OH-B[a]P concentrations. Every effort was madeto limit the influence of other PAH sources during the experiment, butthese other PAH sources can nevertheless not be completely discountedfor the period prior to the studied exposures. The post-exposure con-centrations of urinary 1-OH-Pyr showed no relationship with atmo-spheric concentrations to which subjects were exposed. There was norelationship between post-exposure concentrations of the two urinaryPAH metabolites studied here. Despite the high percentage of samples(ca. 80%)with concentrations of urinary 3-OH-B[a]P inferior to the LOD,there was a significant relationship between the proportion of thepopulation with detectable urinary 3-OH-B[a]P and the atmosphericconcentrations of B[a]P to which they were exposed. Hence, while farfrom being a completely satisfactory biomarker of B[a]P exposure in thegeneral population, the presence or absence of detectable urinary 3-OH-B[a]P may serve as a qualitative indicator of exposure. In such a contextof low environmental PAH exposure, it would be interesting to enlargethe spectrum of biomarkers investigated to include other urinarymonohydroxy metabolites (such those of naphthalene, fluorene,phenanthrene, chrysene...), using sensitive analytical methods.

Acknowledgments

This project was supported by the Institut de Recherches enENvironnement Industriel (IRENI) and financed by the EuropeanCommunity (SubsidyFEDERpredictsNo. 6912),Nord-PasdeCalais Region.

The promoter of this study was the CHRU of Lille. We are gratefulto the Laboratory of Toxicology and Genopathy of the Lille RegionalHospital Center for urinary cotinine analysis.

References

AFNOR NF X 43-294. Juin 1995. Air des lieux de travail. Echantillonnage et analyse deshydrocarbures aromatiques polycycliques. Paris La-Défense, AFNOR; 1995. 18 pp.

ATSDR (Agency for Toxic Substances and Disease Registry). Toxicological Profile forPolycyclic Aromatic Hydrocarbons (PAHs). US Department of Health and HumanServices; 1995. 487 pp, http://www.atsdr.cdc.gov/toxprofiles/tp69.pdf, (April 2009).

Boström CE, Gerde P, Hanberg A, Jernstrom B, Johansson C, Kyrklund T, et al. Cancer riskassessment, indicators, and guidelines for polycyclic aromatic hydrocarbons in theambient air. Environ Health Perspect 2002;110(Suppl 3):451–88.

Bouchard M, Pinsonneault L, Tremblay C, Weber JP. Biological monitoring of environ-mental exposure to polycyclic aromatic hydrocarbons in subjects living in the vicinityof a creosote impregnation plant. Int Arch Occup Environ Health 2001;74:505–13.

CDC. Third national report on human exposure to environmental chemicals.Department of Health and Human Services, Centers for Disease Control andPreventionNational Center for Environmental Health, Atlanta, NCEH Pub. 05-0570,July; 2005. 475 pp.

1173A. Leroyer et al. / Science of the Total Environment 408 (2010) 1166–1173

Cirillo T, Montuori P, Mainardi P, Russo I, Triassi M, Amodio-Cocchieri R. Multipathwaypolycyclicaromatichydrocarbonandpyreneexposureamongchildren living inCampania(Italy). J Environ Sci Health A Tox Hazard Subst Environ Eng 2006;41: 2089–107.

DRIRE Nord - Pas de Calais. L'industrie au regard de l'environnement en 2006. Edition;2007. 464 pp.

Fiala Z, Vyskocil A, Krajak V, Viau C, Ettlerova E, Bukac J, et al. Environmental exposureof small children to polycyclic aromatic hydrocarbons. Int Arch Occup EnvironHealth 2001;74:411–20.

Gündel J, Angerer J. High-performance liquid chromatographic method with fluorescencedetection for the determination of 3-hydroxybenzo[a]pyrene and 3-hydroxybenz[a]anthracene in the urine of polycyclic aromatic hydrocarbon-exposed workers.J Chromatogr B Biomed Sci Appl 2000;738:47–55.

Gündel J, Schaller KH, Angerer J. Occupational exposure to polycyclic aromatichydrocarbons in a fireproof stone producing plant: biological monitoring of 1-hydroxypyrene, 1-, 2-, 3- and 4-hydroxyphenanthrene, 3-hydroxybenz(a)anthraceneand 3-hydroxybenzo(a)pyrene. Int Arch Occup Environ Health 2000;73: 270–4.

Hollender J, Koch B, Dott W. Biomonitoring of environmental polycyclic hydrocarbonexposure by simultaneous measurement of urinary phenanthrene, pyrene andbenzo[a]pyrene hydroxides. J Chromatogr B 2000;729:225–9.

Huang W, Grainger J, Patterson DG, Turner WE, Caudill SP, Needham LL, et al.Comparison of 1-hydroxypyrene exposure in the US population with that inoccupational exposure studies. Int Arch Occup Environ Health 2004;77:491–8.

IARC (International Agency for Research on Cancer). Air Pollution, Part 1. Some Non-heterocyclic Polycyclic Aromatic Hydrocarbons and Some Related IndustrialExposuresSummaries and Evaluations, Vol. 92; 2006. 18 pp.

INRS. Fiche Métropol N°011. Hydrocarbures aromatiques polycycliques; 2007. 14 pp,h t tp : / /www. i n r s . f r / i n r s -pub / i n r s01 . n s f / i n r s 01_me t ropo l _ v i ew/DBE11A887B43754BC1256D5C0041CD32/$File/011.pdf (April 2009).

Keith L, Telliard W. Priority pollutants I— a perspective view. Envron Sci Tech 1979;13:416–23.

Lafontaine M, Gendre C, Delsaut P, Simon P. Urinary 3-hydroxybenzo[a]pyrene as abiomarker of exposure to polycyclic aromatic hydrocarbons : biological limit value.Polycycl Aromat Compd 2004;24:441–50.

Lafontaine M, Champmartin C, Simon P, Delsaut P, Funck-Brentano C. 3-hydroxybenzo[a]pyrene in the urine of smokers and non-smokers. Toxicol Lett 2006;162:181–5.

Maître A, Collot-Fertey D, Anzivino L, Marques M, Hours M, Stoklov M. Municipal wasteincinerators: air and biological monitoring of workers for exposure to particles,metals and organic compounds. Occup Environ Med 2003;60:563–9.

Mucha AP, Hryhorczuk D, Serdyuk A, Nakonechny J, Zvinchuk A, Erdal S, et al.Urinary 1-hydroxypyrene as a biomarker of PAH exposure in 3-year-old Ukrainianchildren. Environ Health Perspect 2006;114:603–9.

OJEU (Official journal of the European Union). Directive 2004/107/EC of the EuropeanParliament and of the Council of 15 December 2004 relating to arsenic, cadmium,mercury, nickel and polycyclic aromatic hydrocarbons in ambient air; 2005.http://faolex.fao.org/docs/pdf/eur49733.pdf.

Park KK, Kim YJ, Kang CH. Atmospheric polycyclic aromatic hydrocarbons in Seoul,Korea. Atmos Environ 2002;36:2917–24.

Ravindra K, Bencs L, Wauters E, de Hoog J, Deutsch F, Roekens E, et al. Seasonal and site-specific variation in vapour and aerosol phase PAHs over Flanders (Belgium) andtheir relation with anthropogenic activities. Atmos Environ 2006;40:771–85.

Saint-Amour M, Tremblay C, Jacques L, Weber JP. Biological monitoring of exposureto polycyclic aromatic hydrocarbons among people living nearby an alu-minium smelter in the province of Québec. Rev Epidemiol Santé Publ2000;48: 439–48.

Simon P, Lafontaine M, Delsaut P, Morele Y, Nicot T. Trace determination of urinary3-hydroxybenzo[a]pyrene by automated column-switching high-performanceliquid chromatography. J Chromatogr B Biomed Sci Appl 2000;748:337–48.

Simon P, Andre F, Delsaut P, Kivistö H, Gendre C, Fabre N, et al. Surveillance biologique :vers un nouveau dispositif de recueil urinaire. Documents pour le Médecin du Travail2003;93:35–51.

Van Rooij JG, Veeger MM, Bodelier-Bade MM, Scheepers PT, Jongeneelen FJ. Smokingand dietary intake of polycyclic aromatic hydrocarbons as sources of interindivid-ual variability in the baseline excretion of 1-hydroxypyrene in urine. Int Arch OccupEnviron Health 1994;66:55–65.

Viau C, Vyskocil A, Martel L. Background urinary 1-hydroxypyrene levels in non-occupationally exposed individuals in the Province of Quebec, Canada, and com-parison with its excretion in workers exposed to PAH mixtures. Sci Total Environ1995;163:191–4.

World Health Organization. Biological monitoring of chemical exposure in the workplace.Vol.1. Geneva: WHO; 1996.