Evaluation of DNA Adducts, DNA and RNA Oxidative Lesions, and 3

Jul 2, 2010 - Between 2 and 72 h postdosing, BPDE−DNA adducts were significantly increased in lung, liver, and mononucleated blood cells of BaP-trea...
0 downloads 0 Views 142KB Size
Download PDF
Chem. Res. Toxicol. 2010, 23, 1207–1214

1207

Evaluation of DNA Adducts, DNA and RNA Oxidative Lesions, and 3-Hydroxybenzo(a)pyrene as Biomarkers of DNA Damage in Lung Following Intravenous Injection of the Parent Compound in Rats Caroline Marie-Desvergne,†,‡ Anne Maıˆtre,*,‡ Miche`le Bouchard,† Jean-Luc Ravanat,§ and Claude Viau† De´partement de Sante´ EnVironnementale et Sante´ au TraVail, Chaire d’Analyse et de Gestion des Risques Toxicologiques, Institut de Recherche en Sante´ Publique de l‘UniVersite´ de Montre´al, Faculte´ de Me´decine, UniVersite´ de Montre´al, CP 6128, Station Centre-Ville, Montre´al, Que´bec H3C 3J7, Canada, Equipe EnVironnement et Pre´diction de la Sante´ des Populations, Laboratoire TIMC (UMR 5525), CHU de Grenoble, UniVersite´ Joseph Fourier, Faculte´ de Me´decine, Domaine de la Merci, 38700 La Tronche, France, and DSM/INAC/SCIB UMR-E 3 CEA-UJF/FRE 2600 CNRS, Laboratoire “Le´sions des Acides Nucle´iques”, CEA-Grenoble, 38054 Grenoble Cedex 9, France ReceiVed March 3, 2010

Biomarkers of exposure and effect were assessed in 40 male Sprague-Dawley rats injected intravenously with 40 µmol/kg of benzo(a)pyrene (BaP) to determine which biomarkers are more representative of BaP-induced DNA damage in lung. Lung, liver, blood, and urine were collected at t ) 2, 4, 8, 16, 24, 33, 48, 72, and 360 h postdosing. Specific BaP-diol epoxide (BPDE)-DNA adducts, 8-hydroxy-7,8dihydro-2′-deoxyguanosine (8-OHdGuo), were measured in lung, liver, and mononucleated blood cells by high-performance liquid chromatography coupled to tandem mass spectrometry (HPLC-MS/MS). Urinary 8-OHdGuo and 8-hydroxy-7,8-dihydroguanosine (8-OHGuo) were also determined by HPLCMS/MS, and urinary 3-hydroxybenzo(a)pyrene was measured by HPLC/fluorescence. Between 2 and 72 h postdosing, BPDE-DNA adducts were significantly increased in lung, liver, and mononucleated blood cells of BaP-treated rats as compared to controls, with the highest levels found in lung. 8-OHdGuo levels also increased in lung of BaP-treated rats with values reaching statistical significance at 2, 8, and 16 h postinjection. No influence of BaP treatment was found on 8-OHdGuo and 8-OHGuo urinary excretions. BPDE-DNA adducts in lung were strongly correlated to urinary 3-OHBaP (r ) 0.936 and p < 0.001) and to a lesser extent to blood BPDE-DNA adducts (r ) 0.636 and p < 0.001), the latter of which were correlated to each other (r ) 0.573 and p ) 0.002). Urinary 3-OHBaP and BPDE-DNA adducts in mononucleated blood cells appear as relevant biomarkers of BaP genotoxic exposure and are highly promising for health risk assessment in humans. Introduction Incomplete combustion of organic matter leads to the formation of polycyclic aromatic hydrocarbons (PAH),1 a family of hundreds of compounds among which some are carcinogenic to humans. Such is the case for benzo(a)pyrene (BaP), a PAH with five fused benzene rings, classified in group 1 by the International Agency for Research on Cancer in 2005 (1). Different industrial sources contribute to PAH emissions, such as coke and electrode production, aluminum electrolysis, and steel factories. PAH exposures may also occur in the general environment through cigarette smoking, consumption of vegetables and smoked or grilled food, and inhalation of polluted air. Occupational PAH exposure has been associated with lung, skin, and bladder cancers in humans (1–3), and the formation * To whom correspondence should be addressed. Tel: +33(0)4 76 63 75 01. Fax: +33(0)4 76 63 75 02. E-mail: [email protected]. † Universite´ de Montre´al. ‡ Universite´ Joseph Fourier. § CEA Grenoble. 1 Abbreviations: PAH, polycyclic aromatic hydrocarbons; BaP, benzo(a)pyrene; BPDE, BaP-diol epoxide; 8-OHdGuo, 8-hydroxy-7,8-dihydro2′-deoxyguanosine; 8-OHGuo, 8-hydroxy-7,8-dihydroguanosine; dGuo, 2′deoxyguanosine; ROS, reactive oxygen species; 1-OHP, 1-hydroxypyrene; 3-OHBaP, 3-hydroxybenzo(a)pyrene.

of PAH DNA adducts is known to be a key step in the initiation of these cancers (4). Carcinogenic PAHs, such as BaP, are therefore characterized by their ability to form various panels of DNA adducts after metabolism (5). BaP is mainly metabolized by cytochromes P450 and epoxide hydrolases to its diol epoxide derivative, the 7,8-dihydroxy-9,10-epoxy-7,8,9,10tetrahydrobenzo(a)pyrene (BPDE). This metabolite reacts with the exocyclic amino group of purines and more extensively with 2′-deoxyguanosine (dGuo), which results in the formation of BPDE-dGuo adducts. Moreover, the metabolism pathway of carcinogenic PAH leading to the formation of aldo-keto reductase derived BaP-7,8-dione is known to generate reactive oxygen species (ROS) able to produce oxidative DNA damage, such as 8-hydroxy-7,8-dihydro-2′-deoxyguanosine (8-OHdGuo). Such lesions may potentially be involved in cancer initiation (6). It is now commonly recognized that biomonitoring of PAH exposure is essential for health risk assessment (7, 8). However, in that perspective, biomarkers should be representative of DNA damage occurring in target organs, such as the lung. DNA adducts seem to be adequate biomarkers of effective dose, dependent on individual factors of metabolism and DNA repair (9). Oxidative DNA lesions should also be evaluated as biomarkers of effect, keeping in mind that these biomarkers are

10.1021/tx100081p  2010 American Chemical Society Published on Web 07/02/2010

1208

Chem. Res. Toxicol., Vol. 23, No. 7, 2010

not specific and might be influenced by many other endogenous and exogenous factors (10, 11). For the assessment of both DNA adducts and oxidative DNA lesions in humans, target organs such as the lung are not accessible; hence, surrogate organs must be selected. Mononucleated blood cells seem interesting since BPDE-dGuo adducts measured in this surrogate tissue have been shown to be representative of lung DNA adducts in rats (12). Nevertheless, blood sampling is considered an invasive procedure in humans such that urinary biomarkers have also been investigated as indirect biomarkers of exposure, potentially correlated to DNA adduct formation (13). Urinary 1-hydroxypyrene (1-OHP) is routinely determined in human urine to assess PAH exposure, as pyrene is ubiquitous in PAH complex emissions. However, pyrene is not carcinogenic, and as a consequence, 1-OHP is not representative of exposure to carcinogenic PAH. This led to the study of potentially suitable urinary biomarkers of exposure to carcinogenic PAH, such as the BaP metabolite 3-hydroxybenzo(a)pyrene (3-OHBaP). The latter metabolite is now proposed for biomonitoring of PAH occupational exposure, but its use is still limited by the fact that particularly highly sensitive analytical methods are required for its detection in humans (14–16). DNA oxidative lesions, as assessed by urinary 8-OHdGuo, are also measurable in humans (17), and the equivalent lesion on RNA, 8-hydroxy-7,8-dihydroguanosine (8-OHGuo), is excreted in urine in larger relative amounts than 8-OHdGuo (18). This raises the question whether all of these biomarkers are representative of BaP metabolic pathways leading to the formation of DNA damage in the lung. Only animal studies allow measurements of the relevant biomarkers in the lung itself. Godschalk et al. (13) reported that urinary 3-OHBaP was highly correlated to DNA adducts measured by 32P-postlabeling in rat lung, but only after BaP intratracheal exposure and not after oral and dermal exposures. Tzekova et al. (19) found that urinary 3-OHBaP was highly correlated to BaP hemoglobin adducts and to a lesser extent to BaP lung protein adducts in rats following BaP intraperitoneal injection. The latter authors also showed that DNA or protein adducts measured in white blood cells were correlated to lung adducts, but the measurement of specific adducts was not performed in these studies. As for 8-OHdGuo, Stedeford et al. (20) showed a significant increase in the concentration of this biomarker in rat lung following BaP intraperitoneal treatment. However, following oral exposure to BaP in rats, Briede et al. (21) reported a decrease in 8-OHdGuo concentrations in liver and lung as compared to controls, while Kim et al. (22) found a significant increase in liver and kidney concentrations but not in lung. In the study of Briede et al., the decrease of 8-OHdGuo concentrations in lung of treated rats as compared to control animals was combined with an increase in 8-OHdGuo excretion in urine. In humans, contradictory results were also reported for 8-OHdGuo, and an increase of this biomarker excretion in urine due to PAH exposure was not always found (23–25). To our knowledge, few studies have simultaneously assessed all of these biomarkers following BaP exposure in rats and using specific analytical procedures. The aim of our study was therefore to concomitantly determine the time courses of BPDE-dGuo adducts, 8-OHdGuo, 8-OHGuo, and 3-OHBaP in tissues and urine of rats exposed to a single dose of BaP and to assess correlations between the different biomarkers. This work is an ancillary part of a detailed kinetic time-course study published elsewhere (26), which focused on BaP and 3-OHBaP time courses in blood, tissues, and excreta of intravenously

Marie-DesVergne et al.

injected rats. Lung and mononucleated blood cells were chosen as target and surrogate tissues, respectively, and the liver was investigated for comparison purposes given that it is a major organ of metabolism and detoxication of xenobiotics. The intravenous route of administration was used to bypass routeof-entry absorption issues and assess the internal dynamics of BaP and its metabolites. To obtain specific measurements of DNA adducts and oxidative lesions, previously published highperformance liquid chromatography-tandem mass spectrometry (HPLC-MS/MS) analytical methods were used (27, 28). Indeed, it is to be noted that highly specific analytical measurements using tandem mass spectrometry are required for a relevant use of DNA adducts as biomarkers of exposure (29). Such data should allow the determination of biomarkers that are more representative of DNA damage in lung and that may be used for human exposure and health risk assessment.

Materials and Methods Chemicals. BaP (99% purity) was purchased from Sigma-Aldrich (Oakville, Ontario, Canada). Reference standards of 3-OHBaP and 3-benzo(a)pyrenyl-β-D-glucopyranosiduronic acid (3-OHBaPGlu) (99% purity) were obtained from the National Cancer Institute (NCI) Chemical Reference Standards distributed by Midwest Research Institute (Kansas, MO). β-Glucuronidase/arylsulfatase (100000 Fishman U/mL and 800000 Roy U/mL from Helix pomatia) was obtained from Roche Diagnostics (Laval, Quebec, Canada). Alkamuls EL-620 was kindly provided by Debro Chemicals and Pharmaceuticals (Dorval, Quebec, Canada). HPLC-grade methanol, ethyl acetate, ascorbic and citric acids, and thymol were obtained from Fisher Scientific Co. (Ottawa, Ontario, Canada), and ammonium formate was from Sigma-Aldrich. Solvents used for HPLC were of analytical grade. As for DNA enzymatic hydrolysis, nuclease P1 (penicillium citrium), phosphodiesterase II, and deferoxamine were obtained from Sigma (St. Quentin-Fallavier, France). Alkaline phosphatase was obtained from Roche (Neuilly sur Seine, France), and phosphodiesterase I (Crotalus adamentus venom) was purchased from Amersham (Piscataway, NJ). Animal Treatment. The treatment protocol used in this study was initially established for a kinetic time-course study of BaP and 3-OHBaP in blood, tissues, and excreta and has also been described elsewhere (26). Forty male Sprague-Dawley rats (Charles River Canada Inc., St.-Constant, Quebec, Canada) weighing 260-290 g were used. Lighting was maintained on a 12 h light-dark cycle, and the room temperature was kept at 22 ( 3 °C. Prior to intravenous injection, rats were kept in plastic cages in groups of two; following injection, animals were put in individual metabolic cages. The principles and guidelines of the Canadian Council on Animal Care were followed. Rats were provided with food and tap water ad libitum prior to treatment. Water containing D-glucose (40 g/L) was supplied ad libitum 12 h before intravenous injection and throughout the experiment to induce a polydipsic behavior with associated aqueous diuresis allowing frequent urine collection (30). Food was removed 2 h prior to intravenous injection and then provided 1 h per day throughout the experiment. Rats received a single intravenous dose of 40 µmol/kg of BaP. A relatively high dose was administered to allow the detection of low levels of biomarkers in the different studied tissues. The vehicle used for the injection was a 20% alkamuls:80% isotonic glucose solution. Three milliliters of solution containing BaP was injected per kilogram of body weight. Groups of four rats were euthanized by CO2 inhalation at t ) 2, 4, 8, 16, 24, 33, 48, and 72 h and 15 days postdosing. An additional group of four rats used as controls received an injection of the vehicle and was euthanized 24 h postinjection. Total blood (about 8 mL) was withdrawn by cardiac puncture (except at t ) 2 h), and livers and lungs were quickly removed, rinsed with saline, blotted dry, and weighed. Urine voided between the time of injection and euthanasia was also collected except for the group of rats euthanized

Benzo(a)pyrene DNA Adducts and Urinary Biomarkers 15 days after injection, as they were not placed in metabolic cages. Thymol was added to the urine collection tubes prior to sampling to avoid bacterial growth. After collection, volumes of urine samples were measured. All samples were kept on ice before storage at -20 °C until analysis. Blood and Tissue DNA Extraction and Enzymatic Hydrolysis. Mononucleated blood cells were isolated from total blood within 2 h posteuthanasia using Histopaque 1083 medium from Sigma (St. Louis, MO), and cellular pellets were stored at -80 °C prior to DNA extraction. Aliquots of 200 mg of lung and liver were potter homogenized prior to DNA extraction. After DNA extraction from mononucleated blood cells and tissue homogenates using a chaotropic method to minimize spurious DNA oxidation during workup, DNA was subsequently digested to nucleosides as previously described (31). Measurements of BPDE-dGuo Adducts in Blood and Tissue DNA. Prior to analysis of BPDE-dGuo adducts, hydrolyzed DNA samples were lyophilized overnight and dissolved in MeOH: H2O (50:50 v/v). Measurement of BPDE-dGuo DNA adducts was achieved by HPLC-MS/MS in the positive ionization mode using the multiple reaction detection method as described by Marie et al. (27). Briefly, the analytical system consisted of an Agilent (Massy, France) 1100 HPLC apparatus equipped with a binary pump, a thermostatted autosampler, and an UV detector. The HPLC system was coupled to an API 3000 tandem mass spectrometer (Applied Biosystems) equipped with an electrospray ionization source. Chromatographic separations were performed using a C8 reversed phase Uptisphere ODB (150 ( 2 mm i.d., 3 µm particles size) column from Interchim (Montluc¸on, France). The elution was performed with a linear gradient from 0 to 100% of acetonitrile in 2 mM ammonium formate over 30 min, at a flow rate of 0.2 mL/ min. The transitions used for the detection of BPDE-dGuo adducts were m/z 570.4 [MH]+ f m/z 257.2 [BPDE-H2O - CO+], m/z 454.2 [BPDE-Gua+], m/z 285.2 [BPDE - H2O+], and m/z 303.2 [BPDE+]. DNA adducts were quantified using external calibration. The amounts of normal nucleosides were determined with an external calibration using the UV detector set up at 270 nm. The limit of detection of the assay using a 100 µg DNA sample was 1 adduct/108 normal nucleosides. Measurements of 8-OHdGuo in Blood and Tissue DNA. 8-OHdGuo was measured in DNA extracted from mononucleated blood cells, lung, and liver by HPLC-MS/MS with the system described above. The chromatographic separation was performed with a C18 reversed phase Uptisphere octadecylsilyl silica gel column (150 ( 2.1 mm i.d., 5 µm particles size) from Interchim (Montluc¸on, France). The elution was achieved at a flow rate of 0.2 mL/min and began with a 15 min linear gradient from 0 to 2.5% of MeOH in ammonium formate (2 mM) over 15 min, followed by a 15 min plateau at 10% MeOH reached in 5 min. 8-OHdGuo was detected using the transition m/z 284 [M + H]+ f m/z 168 [M - 116 + H]+ and quantified using an external calibration curve. The amounts of normal nucleosides were determined with an external calibration using the UV detector set up at 285 nm. The limit of detection of the assay using a 100 µg DNA sample was 20 adducts/108 normal nucleosides. Measurements of Urinary BPDE-dGuo Adducts. BPDEdGuo adducts were investigated in urine as a potential indicator of DNA adduct formation and repair. The same HPLC-MS/MS method described for blood and tissue DNA adducts was used for this purpose. Urinary samples (10 mL) were diluted with 20 mL of sodium acetate buffer (pH 4.6) prior to purification and concentration on C18 Sep-Pak Vac RC (500 mg) cartridges from Waters (Guyancourt, France). Cartridges were conditioned with 5 mL of MeOH followed by 10 mL of water. After urine filtration, cartridges were rinsed with H2O:MeOH (60:40 v/v) and eluted with MeOH. Methanolic fractions were evaporated to dryness, and the residue was dissolved in 100 µL of MeOH prior to HPLC-MS/MS analysis. Mean recovery ((SD) of BPDE-dGuo in spiked blank rat urine was 65 ( 7% (n ) 3). The limit of detection of the assay was 10 pmol/L of urine using a 10 mL urinary sample.

Chem. Res. Toxicol., Vol. 23, No. 7, 2010 1209 Measurements of Urinary 8-OHdGuo and 8-OHGuo. Aliquots of 500 µL of each urine samples were mixed with 50 µL of internal standards 8-OHdGuo [M + 5] and 8-OHGuo [M + 4] (0.05 µM in deionized water) and lyophilized overnight. Samples were rehydrated in 50 µL of 2 mM ammonium formate and centrifuged at 5000g for 5 min. Supernatants were transferred into vials for HPLC-MS/MS analysis. Chromatographic separations were performed using a C18 reversed phase Altima (250 ( 2.1 mm i.d., 5 µm particles size) column from Altech (Enghien les Bains, France). A linear gradient from 0 to 20% of acetonitrile in 2 mM ammonium formate over 35 min was used for elution, with a flow rate of 0.2 mL/min. Detection of 8-OHdGuo and 8-OHGuo was performed using the transitions m/z 284 [M + H]+ f m/z 168 [M - 116 + H]+ and m/z 300 [M + H]+ f m/z 168 [M - 132 + H]+, respectively. 15N5-8-OHdGuo prepared as previously described (28) was detected using the transition m/z 289 [M + H]+ f m/z 173 [M - 116 + H]+ and was used as the internal standard for the quantification of 8-OHdGuo. 13C1-15N3-8-OHGuo was prepared according to Frelon et al. (28) from 13C1-15N3-guanosine (32) and was used as the internal standard for the quantification of 8-OHGuo with the transition m/z 304 [M + H]+ f m/z 172 [M - 132 + H]+. The limits of detection obtained for 8-OHdGuo and 8-OHGuo were similar and corresponded to 0.1 nmol/L of urine using a 500 µL sample. Measurements of Urinary 3-OHBaP. Measurements of urinary 3-OHBaP were described in detail by Marie et al. (26). Briefly, 1 mL of each urine void was diluted 1:2 (v/v) in sodium acetate buffer (0.1 M, pH 5.0) and incubated for 16 h with 10 µL of β-glucuronidase-arylsulphatase at 37 °C in a shaking bath. Subsequently, 200 µL of Triton X-100 R (50 g/L in methanol) was added to the samples; they were then vortex-mixed and transferred to HPLC vials prior to analysis. The HPLC system consisted of a model AS-100 automatic injector (Bio-Rad, Richmond, CA) and a 1100 series apparatus equipped with two quaternary pumps, three automated switching valves, and a fluorescence detector from Agilent Technologies (Mississauga, Ontario, Canada). The fluorescence detector was set to excitation and emission wavelengths of 382 and 441 nm, respectively, and the detector signal was recorded with a ChemStation software. The analytical procedure, including an online extraction of 3-OHBaP, was adapted from Simon et al. (16), and the limit of detection of the method was 5 fmol of 3-OHBaP injected onto the column, corresponding to 100 pmol/L of urine using a 1 mL sample. The quantification was performed using an external calibration curve of 3-OHBaPGlu prepared in blank rat urine prior to enzymatic hydrolysis. Statistical Analysis. Nonparametric Mann-Whitney tests were performed to compare biomarker levels between control rats and BaP-treated rats. Pearson correlation coefficients were calculated to study the relationships between the different biomarkers, and urinary concentrations of 3-OHBaP and 8-OHdGuo levels in lung were log10-transformed to normalize their distribution. Linear regressions were performed to determine whether biomarkers could be predictive of DNA damage in lung. Half of the limits of detection were assigned to samples with no detected BPDE-dGuo adducts and 3-OHBaP. Values were considered statistically significant at p e 0.05.

Results BPDE-dGuo Adducts in Blood and Tissues. The time courses of BPDE-dGuo DNA adducts in mononucleated blood cells, lung, and liver are shown in Figure 1 and Table 1. BPDE-dGuo adduct concentrations were 1.5 and 4 times higher in lung than in mononucleated blood cells and liver, respectively. Levels of BPDE-dGuo DNA adducts in lung, blood, and liver were significantly higher in rats injected with BaP than in controls at all time points, except in liver 15 days after the injection.

1210

Chem. Res. Toxicol., Vol. 23, No. 7, 2010

Marie-DesVergne et al.

only. Cumulative urinary excretion time course of 3-OHBaP is presented in Table 1. No 3-OHBaP was detected in urine of control rats. Associations between Biomarkers. Correlations between biomarkers influenced by BaP injection were assessed, that is, BPDE-dGuo DNA adducts in blood, lung, and liver, 8-OHdGuo in lung, and urinary 3-OHBaP, excluding controls (Table 2). The strongest correlation was found between lung DNA adducts and urinary 3-OHBaP, and the linear regression shown in Figure 3 was highly significant (p < 0.001). DNA adducts measured in mononucleated blood cells were found to be significantly correlated with DNA adduct levels in lung and, to a lesser extent, to urinary 3-OHBaP, DNA adducts in liver, and 8-OHdGuo in lung.

Discussion

Figure 1. Time courses of BPDE-dGuo adducts in lung ([), blood mononucleated cells (2), and liver (9) over the 15-day period following intravenous injection of 40 µmol/kg of BaP in male Sprague-Dawley rats. Control rats euthanized 24 h following injection of the vehicle were assigned to the time point t ) 0. Each point represents the mean, and vertical bars are standard deviations (n ) 4).

Although there were quantitative differences in DNA adduct levels between the lung, liver, and mononucleated blood cells, the time course of DNA adduct formation was similar in these matrices. Adduct formation increased rapidly postinjection to reach a maximum by 16 h postinjection, and in the 16-72 h period postdosing, the levels of DNA adducts observed in these matrices remained fairly constant. BPDE-dGuo DNA adduct levels 15 days postinjection were significantly lower than the maximum levels reached in blood and tissues and represented 42, 36, and 10% of the maximum levels in blood, lung, and liver, respectively. 8-OHdGuo in Blood and Tissues. Figure 2 and Table 1 depict the time courses of 8-OHdGuo in mononucleated blood cells, lung, and liver. 8-OHdGuo was detected in all samples including samples from control rats. In blood and liver, no significant difference in 8-OHdGuo levels was observed between BaP-treated rats and controls at all time points of the kinetics. In lung, 8-OHdGuo levels were higher in treated rats than in controls with differences reaching statistical significance at t ) 2, 8, and 16 h postinjection. Urinary Biomarkers. With regard to BPDE-dGuo DNA adducts in urine, no adduct was detected following BaP injection. This indicates that if this adduct was excreted in urine, its concentration was below the limit of detection of 10 pmol/L of urine. Furthermore, no significant difference was observed in 8-OHdGuo and 8-OHGuo total urinary excretions over 24 h between BaP-treated rats and controls (data not shown). Therefore, urinary 8-OHdGuo and 8-OHGuo levels were not significantly influenced by BaP treatment. The curves of 8-OHdGuo and 8-OHGuo cumulative excretions in urine expressed as functions of time were linear (r2 ) 0.95 and r2 ) 0.94, respectively), indicating that the corresponding rates of excretion were constant over time (data not shown). As expected, 3-OHBaP urinary excretion over the 24 h period following BaP injection was significantly higher (p < 0.05) in treated rats than in the group of rats injected with the vehicle

This study allowed the identification of potential biomarkers of exposure and effect representative of DNA damage measured in lung following BaP exposure. This was achieved by concomitantly determining the time courses of several biomarkers in lung, blood, and urine at different sampling times following an intravenous injection of BaP in male Sprague-Dawley rats. In lung, BPDE-dGuo DNA adduct levels were significantly increased (more than 1 order of magnitude) following injection (between 2 h and 15 days postinjection) as compared to controls. Lung 8-OHdGuo levels also increased markedly in lung after treatment. However, this increase was heterogeneous, and 8-OHdGuo values were significantly higher only at 2, 8, and 16 h postinjection as compared to controls. This could be due to the fact that 8-OHdGuo is not specific of BaP metabolism and therefore presents more interindividual variability. Such results may indicate that both mono-oxygenation and oxidative metabolic pathways increase in lung following BaP exposure but with more variability for the oxidative pathway. Quantitatively, in the rat inhalation study of Ramesh et al. (33), the profile of metabolites obtained in lung over the 4 h period following BaP exposure indicated that BaP dione formation, involving ROS production, was less prominent than that of BaP diols and phenols. The same observation was made by Molliere et al. (34) in isolated perfused rat lung. Such results may be explained in part by the fact that 8-OHdGuo is more rapidly repaired than BPDE-dGuo adducts in DNA. Indeed, 8-OHdGuo is known to be repaired in less than 1 h (35) as compared to 24 h for BPDE-dGuo adducts (36). 8-OHdGuo in DNA is repaired by base excision repair (37) that releases the free base and also probably to a lesser extent by nucleotide excision repair (NER), which could explain the presence of 8-OHdGuo in urine (38, 39). BPDE-dGuo adducts may then be more relevant biomarkers of DNA damage in lung following BaP injection than 8-OHdGuo. In our study, BPDE-dGuo DNA adducts represented 1.2 × 10-5% of administered dose, indicating that the BaP metabolic fraction resulting from mono-oxygenation and potentially involved in cancer initiation might be very low. Such calculations could not be performed for the metabolic fraction generated through quinone pathway since 8-OHdGuo is not specific of BaP-induced damage. Similar to lung, BPDE-dGuo adducts in liver were significantly increased following BaP treatment as compared to controls, but the concentrations were approximately four times lower than in lung. This appears to reflect that more BPDE is available for DNA adduct formation in the lung. Concerning 8-OHdGuo, no significant increase was observed in liver after

Benzo(a)pyrene DNA Adducts and Urinary Biomarkers

Chem. Res. Toxicol., Vol. 23, No. 7, 2010 1211

Table 1. Mean Values Obtained for the Different Biomarkers Measured in Blood, Tissues, and Urine at the Various Sampling Times (2, 4, 8, 16, 24, 33, 48, 72, and 360 h Postdosing) Following Intravenous Injection of 40 µmol/kg of BaP in Male Sprague-Dawley Rats mean values (standard deviation, n ) 4) of biomarkers 8

BPDE-dGuo (adducts/10 normal nucleosides) time after injection (h) 0a 2 4 8 16 24 33 48 72 360

blood ND,b 0.5 (0) 6.3 (1.9) 7.8 (0.91) 9.8 (1.7) 11.6 (2.2) 10.5 (2.5) 9.6 (0.92) 8.7 (1.4) 4.2 (2.6)

8-OHdGuo (lesions/106 normal nucleosides)

lung

liver

blood

lung

liver

ND,b 0.5 (0) 2.1 (0.52) 5.5 (2.0) 9.7 (0.57) 12.5 (2.4) 13.1 (1.7) 16.2 (2.7) 13.3 (2.1) 15.0 (3.5) 6.8 (1.5)

ND,b 0.5 (0) 2.1 (0.37) 3.2 (1.9) 3.0 (1.2) 3.3 (1.6) 2.9 (0.74) 4.6 (0.70) 2.5 (1.1) 3.3 (0.44) ND,b 0.5 (0)

2.0 (0.90)

1.8 (0.58) 3.5 (1.1) 3.2 (1.3) 4.2 (1.4) 3.3 (0.27) 10.5 (15.8) 3.2 (0.94) 3.2 (1.97) 2.3 (0.66) 2.7 (1.3)

0.94 (0.25) 0.99 (0.24) 0.94 (0.28) 0.69 (0.06) 0.93 (0.17) 0.97 (0.22) 0.93 (0.24) 0.95 (0.17) 1.1 (0.13) 1.1 (0.17)

2.6 (0.94) 2.8 (1.1) 4.4 (4.9) 2.8 (1.7) 2.3 (0.80) 2.3 (0.76) 2.4 (1.3) 1.9 (0.45)

3-OHBaP (pmol) urine ND,b 5.4 (1.1) 15.8 (8.5) 249 (114) 1311 (568) 6579 (5230) 7219 (3896) 21984 (8759) 16416 (8769) 23785 (10839)

a The time point t ) 0 corresponded to control rats euthanized 24 h following injection of the vehicle. b Not detected. Half of the analytical limit of detection was assigned to these samples, and for urinary 3-OHBaP in control rats, the volume of urine excreted over 24 h was also taken into account.

Figure 2. Time courses of 8-OHdGuo lesions in lung ([), blood mononucleated cells (2), and liver (9) over the 15-day period following intravenous injection of 40 µmol/kg of BaP in male Sprague-Dawley rats. Control rats euthanized 24 h following injection of the vehicle were assigned to the time point t ) 0. Each point represents the mean, and vertical bars are standard deviations (n ) 4).

BaP injection, and basal levels of 8-OHdGuo were 2-fold lower than in lung. The higher BPDE-dGuo adduct and 8-OHdGuo levels in lung as compared to the liver may be explained in part by the fact that lung receives total cardiac output after an intravenous injection. However, it was shown that even after intraperitoneal injection of BaP, lung BPDE-DNA adducts were higher in lung than in liver (40). As an explanation, Molliere et al. (34) indicated that phase I metabolism was as efficient in lung than in liver, but phase II metabolism might be more limited, as suggested by the presence of more free diols and phenols in lung than in liver. If the lung is less efficient in phase II conjugation than the liver, more free BaP diols might be likely to undergo further metabolism to BPDE or BaP-7,8-dione and form more DNA adducts and 8-OHdGuo in lung than in liver. BPDE-dGuo adducts were also measured in mononucleated blood cells, which are mainly composed of lymphocytes. It was shown that DNA adducts in mononucleated blood cells rather than in total white blood cells are more representative of lung DNA adducts (12). This may be explained by the long lifespan of lymphocytes (several years) contrary to that of granulocytes

(about a day). DNA adducts measured in lymphocytes may thus be considered as a bioindicator of past exposure and potential accumulation. Blood BPDE-dGuo DNA adducts were significantly increased in BaP-treated rats and were only 1.5 times lower than DNA adduct levels in lung. Furthermore, there was a good correlation in BPDE-DNA adduct levels between these two matrices. Our results are in agreement with published data indicating that DNA adducts measured by 32P-postlabeling in BaP-exposed rats were predominant in lung in comparison to white blood cells, whatever the route of exposure (intraperitonal, intratracheal, oral, or dermal) (13, 41). Tzekova et al. (42) also reported that BPDE protein adducts were higher in lung in comparison to liver and blood following intraperitoneal injection of BaP in rats. Blood DNA adducts were also significantly correlated with 8-OHdGuo in lung. However, no significant increase of 8-OHdGuo in blood was found following BaP injection. As for the persistence of DNA adducts in the various tissues analyzed, it was observed that lung DNA adducts were the most persistent adducts followed by blood adducts and liver adducts since 42% of maximum DNA adduct level was still detected in lung 15 days postdosing in comparison to 36 and 10% in blood and liver, respectively. Nonetheless, the kinetics of DNA adduct repair appeared relatively slow in all tissues analyzed as compared to the rate of adduct formation (0-16 h), which would explain the rather constant level of DNA adducts between 16 and 72 h in lung, liver, and blood mononucleated cells. These results are in agreement with those previously published by Godschalk et al., indicating that 34 and 74% of the maximal adduct levels were still detected in rat lung 21 days after BaP intratracheal and oral exposure (13). They are also in line with those of Qu and Stacey (41) showing that 50% of the adduct level in lung 1 day after BaP intraperitoneal injection was found 15 days after the treatment and that removal of adducts was faster in liver followed by blood mononucleated cells and lung. In our study, no BPDE-dGuo could be detected in urine of BaP-treated rats. If BPDE-dGuo was indeed excreted in urine, its concentration was below our limit of detection. Improvement of the analytical procedure and more precisely of the preanalytical extraction and purification steps should be considered in the future. This molecule may be excreted in urine following NER, as it possibly also occurs for 8-OHdGuo, but our results do not bring any substantial information to confirm or invalidate this hypothesis. In addition, another described metabolic pathway for BaP might be responsible for BaP adduct excretion in urine. The monoelectronic oxidation by peroxydases and

1212

Chem. Res. Toxicol., Vol. 23, No. 7, 2010

Marie-DesVergne et al.

Table 2. Pearson Correlation Coefficients between BPDE-dGuo DNA Adducts in Blood, Lung, and Liver, 8-OHdGuo in Lung, and Urinary 3-OHBaP

lung BPDE-dGuo

r p n

blood BPDE-dGuo

r p n

liver BPDE-dGuo

r p n

urinary log10(3-OHBaP)

r p n

a

blood BPDE-dGuo

liver BPDE-dGuo

urinary log10(3-OHBaP)

lung log10(8-OHdGuo)

0.636a