Identification and Quantification of 1-Nitropyrene Metabolites in

Chem. Res. Toxicol. 2007, 20, 999-1007

999

Identification and Quantification of 1-Nitropyrene Metabolites in Human Urine as a Proposed Biomarker for Exposure to Diesel Exhaust Akira Toriba,*,† Hitomi Kitaoka,† Russell L. Dills,‡ Satoko Mizukami,† Kaori Tanabe,† Naoki Takeuchi,† Mariko Ueno,† Takayuki Kameda,† Ning Tang,† Kazuichi Hayakawa,† and Christopher D. Simpson‡ Graduate School of Natural Science and Technology, Kanazawa UniVersity, Kakuma-machi, Kanazawa 920-1192, Japan, and Department of EnVironmental and Occupational Health Sciences, School of Public Health and Community Medicine, UniVersity of Washington, Box 357234, Seattle, Washington 98195-7234 ReceiVed January 13, 2007

1-Nitropyrene (1-NP) is one of the most abundant nitrated polycyclic aromatic hydrocarbons (NPAHs) in diesel exhaust particulate matter (DEP) and is a main contributor of direct-acting mutagenicity in DEP. Therefore, the metabolites of 1-NP are expected to be a biomarker for assessment of exposure to DEP. In this study, a highly specific and sensitive analytical method using liquid chromatography with tandem mass spectrometry (LC-MS/MS) was developed to determine urinary 1-NP metabolites. After enzymatic hydrolysis of the conjugated metabolites, the analytes were selectively extracted from the urine matrix with blue rayon. The eluate from the rayon was further purified on an acidic alumina cartridge. Hydroxy-N-acetyl- 1-aminopyrenes (6- and 8-OHNAAP) and hydroxy-1-nitropyrenes (3-, 6-, and 8-OHNP) in human urine were identified by their retention times and MS/MS spectra and quantified by using deuterated internal standards. 1-NP metabolites were quantified in urine from all healthy, nonoccupationally exposed subjects. 6-OHNAAP, 8-OHNAAP, 6-OHNP, and 8-OHNP (means of 117, 109, 203, and 137 pmol/mol creatinine, respectively) were the most abundant isomers in human urine. This report is the first to demonstrate the presence of OHNAAPs and OHNPs in human urine, in agreement with previous in vivo and in vitro studies that predicted that these metabolites should be excreted into human urine. This method for determining urinary 1-NP metabolites should be useful for the surveillance of exposure to NPAHs and DEP and will facilitate the study of cancer risk associated with these exposures. Introduction Lung cancer is the most common cancer and the major cause of cancer death in the world (1, 2). While smoking is an important factor for this high incidence, other suspected factors include automobile exhaust and ambient and indoor air pollutions (3-6). Diesel exhaust, a major source of environmental pollution, is classified as a probable human carcinogen (IARC Group 2A) (7, 8). Nitrated polycyclic aromatic hydrocarbons (NPAHs)1 are formed through the incomplete combustion of fossil fuels. They are widespread environmental contaminants observed in the extracts of diesel exhaust particulate matter (DEP) and airborne particulate matter (9, 10). Many NPAHs are carcinogenic/mutagenic compounds, and among these compounds, 1-nitropyrene (1-NP) and dinitropyrenes have been previously reported as the main contributors of direct-acting mutagenicity of DEP (11-13). 1-NP is one of the most abundant NPAHs in DEP and has been proposed as a chemical marker * To whom correspondence should be addressed. Tel: +81-76-234-4457. Fax: +81-76-234-4456. E-mail: [email protected]. † Kanazawa University. ‡ University of Washington. 1 Abbreviations: 1-NP, 1-nitropyrene; LC-MS/MS, liquid chromatography with tandem mass spectrometry; OHNAAP, hydroxy-N-acetyl-1aminopyrene; OHNP, hydroxy-1-nitropyrene; PAHs, polycyclic aromatic hydrocarbons; NPAHs, nitrated polycyclic aromatic hydrocarbons; DEP, diesel exhaust particulate matter; NAAP, N-acetyl-1-aminopyrene; 1-AP, 1-aminopyrene; ESI, electrospray ionization; SRM, selected reaction monitoring; MRM, multiple reaction monitoring; EPI, enhanced product ion.

for diesel exhaust (14, 15). Carcinogenic NPAHs were also detected in the resected lungs of nonsmoking patients with lung cancer (16). These findings have generated considerable interest in the relationship between the increasing incidence of lung cancer and the exposure to airborne NPAHs. The metabolism of 1-NP has been studied using various tissues and species (17-22). 1-NP is metabolized essentially through two routes: cytochrome P450-mediated C oxidation and nitroreduction. Urinary or fecal metabolites that have typically been observed in in vivo studies are hydroxy-1nitropyrenes (3-, 6-, and 8-OHNP), hydroxy-N-acetyl-1-aminopyrenes (3-, 6-, and 8-OHNAAP), trans-4,5-dihydro-4,5dihydroxy-1-nitropyrene, N-acetyl-1-aminopyrene (NAAP), and 1-aminopyrene (1-AP) (17-22). However, it is unclear whether these animal studies with relatively high doses are applicable to human exposure to DEP or environmental 1-NP. Van Bekkum et al. investigated the bioavailability of 1-NP after the intragastric administration of native DEP (SRM 2975) (23). Approximately 13% of the 1-NP present on DEP was excreted in urine and identified as 1-NP metabolites. 6-OHNAAP appeared to be the most abundant metabolite (∼7% of administered dose of 1-NP), followed by 6-OHNP (∼1.6% of administered dose of 1-NP). These results indicate that the 1-NP that is present in DEP is metabolized in the same way as was previously observed following administration of pure 1-NP. The structures of these metabolites and the probable pathways of their metabolism in humans are shown in Figure 1.

10.1021/tx700015q CCC: $37.00 © 2007 American Chemical Society Published on Web 06/20/2007

1000 Chem. Res. Toxicol., Vol. 20, No. 7, 2007

Toriba et al.

Figure 1. Major metabolites of 1-NP in human urine. P450, cytochrome P450; NAT, N-acetyltransferase; UGT, UDP-glucuronyltransferase; and ST, sulfotransferase.

Recently, the use of biomarkers has developed as a valuable tool for assessing human exposure to environmental contaminants. Biomarkers of exposures can be exogenous chemicals or metabolites quantified in biological samples such as urine or blood (24). The metabolites of 1-NP are expected to be a specific biomarker of exposure to DEP because of 1-NP’s strong association with diesel exhaust (25). Hemoglobin adducts of NPAHs, particularly 1-NP, have been detected in human blood samples (26, 27). Hemoglobin adducts of 1-NP have been investigated as biomarkers of the exposure to diesel exhaust (26, 27); however, the difference between the adduct levels in bus garage workers and rural controls was not statistically significant (27). In humans, only a few studies have reported methods for determining urinary metabolites of 1-NP or other NPAHs. 1-AP in human urine has been measured with an immunochemical assay (28, 29) and GC-MS (30, 31). However, 1-NP metabolites such as OHNPs and OHNAAPs, which are expected to be the major metabolites from in vivo and in vitro studies, have never been identified and quantified in human urine. Thus, only limited information is available on the urinary excretion of the metabolites of 1-NP in human and on the actual uptake of 1-NP from the exposure to diesel exhaust and other sources. In this study, we aimed to develop a sensitive and specific analytical procedure for the determination of 1-NP metabolites in human urine, which would facilitate the evaluation of 1-NP metabolites as biomarkers of exposure to diesel exhaust. We report an effective pretreatment method for urine samples coupled with liquid chromatography with tandem mass spectrometry (LC-MS/MS) analysis to identify and quantify 1-NP metabolites. Using this method, we identified OHNPs and OHNAAPs as major metabolites of 1-NP in human urine and quantified their concentrations in urine samples from healthy, nonoccupationally exposed subjects.

Materials and Methods Materials. All reagents were of analytical grade and used without any further purification. Pyrene, 1-AP, and platinum(IV) oxide were

purchased from Aldrich (Milwaukee, WI). β-Glucuronidase/aryl sulfatase (type H-2, from Helix pomatia: β-glucuronidase activity, 100000 units/mL; and sulfatase activity, 7500 units/mL) and zinc (granule, 20 mesh) were obtained from Sigma (St. Louis, MO). Blue rayon, rayon fiber (noncrystalline cellulose) covalently bound to copper phthalocyanine trisulfonate, was purchased from Funakoshi (Tokyo, Japan). Ammonia solution (28%, ultrapure) was from Kanto Chemical (Tokyo, Japan). Water was obtained from a Milli-Q purification system (Millipore, Bedford, MA). Other chemicals were from Wako (Osaka, Japan). Synthesis of 1-NP Metabolites. 3-, 6-, and 8-OHNPs, 3-, 6-, and 8-OHNAAPs, and 1-NAAP were synthesized according to the previously reported procedure (17, 23, 32-35) with some modifications. Acetoxypyrene was prepared from pyrene by the treatment with lead tetraacetate in benzene/acetic acid (9/1, v/v) and then nitrated using concentrated HNO3 in acetic acid. The obtained mixture of three isomers of acetoxynitropyrenes was treated with CH3ONa in methanol/THF (1/1, v/v) to obtain a mixture of OHNPs. Each isomer of OHNPs was purified by preparative normal phase HPLC (Supelco silica-gel column 25 cm × 21.2 mm i.d. eluted with CH2Cl2 containing 0.5 mM CH3COOH at 10 mL/min). OHNPs were reduced to amino compounds using zinc in methanol/trisHCl buffer (pH6) (3/1, v/v) at 90 °C or by hydrogenation with H2 gas (10 psi) and platinum(IV) oxide in ethyl acetate at ambient temperature. The amino analogues were immediately acetylated with an excess of acetic anhydride in ethyl acetate to prepare 3-, 6-, and 8-acetoxy-N-acetylaminopyrenes. OHNAAPs were obtained by the hydrolysis of acetoxy-N-acetylaminopyrenes with CH3ONa in methanol. 1-NAAP was prepared by acetylation of 1-AP with acetic anhydride in ethyl acetate. To identify the synthesized compounds, their GC-MS and proton NMR spectra were compared with data in the literature (17, 34, 35). Similar procedures were applied to synthesize deuterated OHNPs and OHNAAPs from pyrene-d10 for use as internal standards. The isolated fraction of OHNPs-d8 was obtained as a mixture of three isomers, and then, the mixture was reduced and acetylated. Deuterated 6- and 8-OHNAAPs were observed as major products, whereas deuterated 3-OHNAAP hardly existed, because of the small proportion of 3-OHNP in the starting material (ca. 15%). MS analysis of the deprotonated molecular ions [M - H]- for the resultant mixture indicated that it contained d0d8 isomers of deuterated OHNAAPs (data not shown), presumably due to the exchange of the deuterium labels in order of labile

1-Nitropyrene Metabolites in Human Urine positions on the pyrene ring during the reduction. The deprotonated molecule for nondeuterated compound accounted for 0.25% of the compounds in the mixture. Because the deprotonated molecule for OHNAAP-d6 was the most abundant, the signal produced by OHNAAP-d6 was used for internal standards. The synthesized standards were protected from light and refrigerated prior to HPLC analysis. Analysis of 1-NP Metabolites by LC/MS/MS. The Agilent 1100 series LC system consisted of a G1379A degasser, a G1312A binary pump, a G1367A autosampler, and a G1316A column oven (all from Agilent Technologies, Palo Alto, CA). Chromatographic separation of 1-NP metabolites in urine samples was performed on a Zorbax Extend-C18 column (150 mm × 2.1 mm i.d., 5 µm, Agilent) with a guard column Zorbax Extend-C18 column (12.5 mm × 2.1 mm i.d., 5 µm, Agilent). The column temperature was kept at 30 °C. A gradient elution using 0.01% NH4OH in water (eluent A) and 0.01% NH4OH in methanol (eluent B) was carried out (B, 25-75% linear gradient for 40 min) at a flow rate of 0.2 mL/min. Sample volumes of 5 µL were typically used for each analysis. The mass spectrometric analyses were performed using an API 4000 Q-Trap tandem mass spectrometer (Applied Biosystems, Foster City, CA) equipped with an electrospray ionization (ESI) interface and operated in a negative ion mode. Sensitivity of the selected reaction monitoring (SRM) was optimized by testing with an infusion of 1-NP metabolites in a mixture of methanol and water (1/1, v/v) containing 0.01% NH4OH. The spray voltage was maintained at -4.5 kV. Nitrogen gas was used as the collision gas and curtain gas, whereas zero grade air was used as the nebulizer gas and heater gas, with the optimum values set, respectively, at 7, 20, 30, and 70 (arbitrary values). The source temperature was set at 600 °C. The mass spectrometer was operated under multiple reaction monitoring (MRM) mode, and the monitored precursor (Q1) and product (Q3) ions were m/z 262 f 232 for OHNPs, m/z 274 f 231 for OHNAAPs, m/z 270 f 240 for OHNPs-d8 (internal standard), and m/z 280 f 237 for OHNAAPs-d6 (internal standard) with dwell times of 1000 ms. The collision energy for OHNPs and OHNAAPs was -85 and -90 eV, respectively. The declustering potential for OHNPs and OHNAAPs was set at -30 and -32 V, respectively, whereas the entrance potential was set at -10 V. The unit mass resolution was used for both Q1 and Q3 mass analyzers. The structures of 1-NP metabolites in human urine were elucidated using the enhanced product ion (EPI) scan mode in which the product ions are trapped in Q3 (in trap mode) before mass analysis. The EPI scan rate was 1000 amu/s, and the scan range was 100400 amu. Analyst software (version 1.4, Applied Biosystems) was used to control the LC/MS/MS system and to acquire and process the data. Pretreatment of Human Urine Samples. A 100 mL aliquot of a urine sample was used for the routine analysis. The urine sample was adjusted to pH 5.0 with 1.0 M hydrochloric acid solution, and then, 5 mL of 4 M acetate buffer (pH 5.0) was added to the sample. To hydrolyze the conjugated metabolites, the solution was incubated with β-glucuronidase (8500 units)/aryl sulfatase (50 units) at 37 °C for 4 h. After the deuterated internal standards corresponding to 6.25 pmol of OHNAAPs-d6 and 2.5 pmol of OHNPs-d8 as the sum of three isomers were added, the solution was incubated with 100 mg of blue rayon for 1 h with shaking at room temperature. The blue rayon was filtered through glass wool in a solid-phase extraction (SPE) tube, and the water phase was discarded. The rayon was washed with 5 mL of water and then air-dried on the vacuum manifold for less than 3 min. The 1-NP metabolites were extracted with 20 mL of methanol/ammonia solution (50/1, v/v) by sonication for 30 min. After the eluate was dried under a stream of nitrogen gas, the residue was dissolved in 5 mL of methanol/ethyl acetate (1/1, v/v), and then, the solution was applied to a Sep-Pak Alumina A cartridge (Waters, Milford, MA) that had been preconditioned with 50 mL of methanol/ethyl acetate (1/1, v/v). The sample solution was applied to the cartridge and then collected through the cartridge. The cartridge was further eluted with 10 mL of methanol/ethyl acetate (1/1, v/v), and the eluate was dried under nitrogen. The residue was redissolved in 50 µL of methanol, and an aliquot (5

Chem. Res. Toxicol., Vol. 20, No. 7, 2007 1001 µL) of the solution was injected into the LC/MS/MS system. To obtain complete mass spectra to confirm identification of the metabolites, 2 L of a pooled urine sample from three nonsmoking subjects was treated with the same above procedure on a large scale. Finally, the residue from the 2 L of urine was redissolved in 150 µL of methanol. Precision and Accuracy. Intraday (within day) and interday (between day) precision and accuracy were determined by replicate analysis of a pooled urine sample without enzymatic hydrolysis. The unhydrolyzed pooled urine gave the concentrations of free (unconjugated) OHNPs and OHNAAPs below the quantification limits. The urine samples were spiked with 2 or 5 and 25 pM OHNAAPs and 2 and 25 pM OHNPs. Except for 3-OHNAAP, the sample concentrations were quantified from the peak area ratio of the analytes to the corresponding deuterated internal standards. Quantification of 3-OHNAAP was based on peak area relative to 8-OHNAAP-d6. For all other metabolites, the sample concentrations were quantified from the peak area ratio of the analytes to the corresponding deuterated internal standards. Human Studies. The study subjects consisted of 22 subjects (mean age, 26; age range, 20-56; gender, 17 males and five females) who lived in the urban area of Kanazawa city, Japan. They were students or staff of Kanazawa University. Seventeen were nonsmokers, and five were smokers. Human urine samples were collected in the morning in a polyethylene bottle and stored at -20 °C until analysis. 1-OHPyr in the urine samples was measured as described previously (36). To compensate for fluctuations due to diuresis, the urinary concentrations of 1-NP metabolites and 1-OHPyr were normalized to the urinary creatinine concentration (pmol/mol creatinine or nmol/mol creatinine). The concentration of urinary creatinine was determined with alkaline picrate using a test kit (Wako Pure Chemicals, Osaka, Japan) according to Jaffe’s method (37).

Results Development of LC-MS/MS Method and Urine Sample Preparation. The goal of this study was to develop a sensitive and specific analytical method for measuring 1-NP metabolites in human urine. LC-MS/MS, which does not require a derivatization step such as GC-MS (/MS), was effective in identifying and quantifying the metabolites. To optimize the detection of OHNPs and OHNAAPs, ESI mass spectra of the standard compounds were obtained in negative ion mode. The major peaks in the mass spectra of OHNPs and OHNAAPs corresponded to the deprotonated molecular ions [M - H]-, m/z 262.1 and 274.0, respectively (data not shown). Typical SRM chromatograms for OHNPs and OHNAAPs of a standard solution are shown in Figures 2A and 3A. As shown in Figures 2A and 3A, a good separation of all isomers of OHNPs and OHNAAPs was accomplished by gradient elution. The full-scan MS/MS (EPI) spectra, B and D in Figures 2 and 3, were obtained from the same analysis triggered by the same SRM peaks shown in A and C. As an example, the MS/MS spectra from the SRM peaks corresponding to 6-OHNP and 6-OHNAAP were shown. The precursor ion of OHNAAPs was fragmented to m/z 231.0 [M - COCH3 - H]- (Figure 2B). The precursor ion of OHNPs was fragmented to m/z 232.1 [M - NO - H]and 216.1 [M - NO2 - H]- (Figure 3B). The loss of NO (30 Da) observed in the ESI-MS/MS spectra is characteristic of aromatic nitro compounds, as reported previously (38, 39). Similar fragmentation patterns were observed in the MS/MS spectra for isomeric OHNAAPs and for isomeric OHNPs. As a result, the monitored precursor (Q1) and product (Q3) ions were chosen as described in the Materials and Methods. The instrumental detection limits of OHNPs and OHNAAPs ranged from 0.15 to 0.3 and 0.75 to 3.0 fmol/injection (S/N ) 3), respectively. This method was found to be more sensitive than

1002 Chem. Res. Toxicol., Vol. 20, No. 7, 2007

Toriba et al.

Figure 2. SRM chromatograms and MS/MS (EPI) spectra of [M - H]- ion of standard OHNPs (A and B) and a urine sample (C and D). The full-scan MS/MS (EPI) spectra, C and D, were obtained from the same analysis triggered by the SRM peaks shown in A and B, respectively. (A) Each peak of OHNPs corresponds to 75 fmol.

previous methods with HPLC-fluorescence detection and with GC-MS/MS (23). Our assay was optimized for 100 mL urine samples. This volume represents a compromise between a sample volume that is small enough to be collected and handled efficiently but is nevertheless large enough volume to achieve the desired detection limits. We adopted blue rayon, a rayon fiber covalently bound to copper phthalocyanine trisulfonate, to pretreat the volume of urine. After we incubated a hydrolyzed urine sample with blue rayon for 1 h, the rayon was easily recovered from the water phase by filtration through a rough filter such as glass wool. After the initial purification, many interfering peaks were still observed on the MS/MS (SRM) chromatograms (data not shown). Therefore, normal-phase SPE was used as a secondary step. Various kinds of normal phase SPE cartridges including Alumina A, amino propyl (NH2), diol, and silica (all from Waters) were evaluated by combining them with the blue rayon treatment. The analytes dissolved in methanol/ethyl acetate (1/ 1, v/v) were eluted from all of the cartridges without retention on their bonded phases. Using an Alumina A cartridge, interfering substances such as pigments in the extract of the blue rayon were tightly bonded to the sorbent, and the interfering peaks on the chromatograms were also remarkably diminished (data not shown). Therefore, the Alumina A cartridge was adopted as the secondary purification step. Finally, the urine extract was concentrated for analysis by LC-MS/MS. The total recoveries of the spiked OHNPs and OHNAAPs from the urine were in the ranges of 77-85 and 41-54%, respectively. Identification of 1-NP Metabolites in Human Urine. Typical SRM chromatograms for OHNPs and OHNAAPs in a

sample prepared from 2 L of human urine are shown in Figures 2C and 3C. To obtain complete MS/MS spectra with adequate sensitivity, we extracted the analytes from a large volume (2 L) of urine. The retention times of peaks 1, 2, 3, 4, and 5 were consistent with 6-OHNP, 8-OHNP, 3-OHNP, 6-OHNAAP, and 8-OHNAAP, respectively (Figures 2C and 3C). In addition, MS/ MS spectra of the five peaks from the urine sample are shown in Figures 2D and 3D. The fragment ions, m/z 232.1 [M - NO - H]- and 216.1 [M - NO2 - H]- for OHNPs (Figure 2D) and m/z 231.0 [M - COCH3 - H]- for OHNAAPs (Figure 3D), were also consistent with those for the standards (Figures 2B and 3B). On the basis of these results, we conclude that OHNPs and OHNAAPs were identified in the urine of healthy control subjects, and we postulate that these compounds were generated via metabolism of 1-NP. The peaks corresponding to free (unconjugated) OHNPs and OHNAAPs in the unhydrolyzed urine were observed; however, they were below the quantification limits. In some cases, the free metabolites were found to be less than 10% of the metabolites in the hydrolyzed urine. Therefore, the identified 1-NP metabolites were mainly glucuronide and/or sulfate of OHNPs and OHNAAPs. The same urine sample was also analyzed by the LC-MS/MS system for 1-NAAP and 1-AP. Although the total recoveries of 1-NAAP and 1-AP from the urine were approximately 80 and 30%, respectively, and the sensitivities of them were comparable to those of OHNPs and OHNAAPs, no peaks were consistent with the retention times and MS/MS spectra for the standards (data not shown). Calibration Curve and Validation. For the routine quantitative determination of urinary 1-NP metabolites, 100 mL urine

1-Nitropyrene Metabolites in Human Urine

Chem. Res. Toxicol., Vol. 20, No. 7, 2007 1003

Figure 3. SRM chromatograms and MS/MS (EPI) spectra of [M - H]- ion of standard OHNAAPs (A and B) and a urine sample (C and D). The full-scan MS/MS (EPI) spectra, C and D, were obtained from the same analysis triggered by the SRM peaks shown in A and B, respectively. (A) Each peak of OHNAAPs corresponds to 250 fmol.

samples were extracted as previously described, and deuterated OHNPs and OHNAAPs were used as internal standards to quantify the analytes. The calibration curves for the standard compounds were linear at concentrations of 1-500 fmol/ injection (correlation coefficient, r2 > 0.999) for OHNPs and 3-500 fmol/injection (r2 > 0.999) for OHNAAPs. The slopes of these calibration curves were almost identical to those of the working curves obtained by adding standards, OHNPs and OHNAAPs, into human urine. The slopes of the working curves were 0.2915 ( 0.024 (mean ( SD, n ) 3) for 3-OHNAAP, 0.0162 ( 0.0002 for 6-OHNAAP, 0.0306 ( 0.0029 for 8-OHNAAP, 0.0326 ( 0.002 for 3-OHNP, 0.0391 ( 0.002 for 6-OHNP, and 0.0513 ( 0.0018 for 8-OHNP. The analytical intraday and interday precision and accuracy data of urinary OHNPs and OHNAAPs are shown in Table 1. The relative standard deviations (RSD, %) of the intraday precision study (n ) 3) were in the range 4.0-10.8, and the interday assay (n ) 4) was in the range of 1.2-10.3 for the urine samples spiked at the concentrations of 2 or 5 and 25 pM OHNAAPs and 2 and 25 pM OHNPs. The accuracy values (%) of the intraday study (n ) 5) and the interday assays (n ) 5) were in the range of 96-109 and 98-112%, respectively. These values indicate that the proposed method is satisfactory for determining 1-NP metabolites in human urine. The enzymatic hydrolysis was optimized for the routine analysis, and the peak area of the metabolites reached a plateau at 4 h (data not shown). A hydrolysis time of 4 h at 37 °C was adopted in this study. The repetitive analysis of a nonspiked urine to validate the hydrolysis step resulted in a reproducibility of