Detection and Quantification of Depurinated Benzo [a] pyrene

University of Nebraska Medical Center, Omaha, Nebraska 68198-5300, and U.S. Environmental. Protection Agency, Research Triangle Park, North Carolina ...
0 downloads 0 Views 125KB Size
192

Chem. Res. Toxicol. 2001, 14, 192-201

Detection and Quantification of Depurinated Benzo[a]pyrene-Adducted DNA Bases in the Urine of Cigarette Smokers and Women Exposed to Household Coal Smoke George P. Casale,*,† Maria Singhal,† Sumitra Bhattacharya,† Regulan RamaNathan,‡ Kenneth P. Roberts,§ Damon C. Barbacci,‡ John Zhao,‡ Ryszard Jankowiak,§ Michael L. Gross,‡ Ercole L. Cavalieri,† Gerald J. Small,§ Stephen I. Rennard,| Judy L. Mumford,⊥ and Meilan Shen† Eppley Institute for Research in Cancer, 986805, University of Nebraska Medical Center, Omaha, Nebraska 68198-6805, Department of Chemistry, Washington University, St. Louis, Missouri 63130, Department of Chemistry and Ames Laboratory-U.S. Department of Energy, Iowa State University, Ames, Iowa 50011, Internal Medicine-Pulmonary, University of Nebraska Medical Center, Omaha, Nebraska 68198-5300, and U.S. Environmental Protection Agency, Research Triangle Park, North Carolina 27711 Received January 27, 2000

Polycyclic aromatic hydrocarbons (PAH) are metabolized to electrophiles that can bind to DNA bases and destabilize the N-glycosyl bond, causing rapid depurination of the adducted bases. Recent studies support depurination of DNA as a mechanism central to the genesis of H-ras mutations in PAH-treated mouse skin. Depurinating adducts account for 71% of all DNA adducts formed in mouse skin treated with benzo[a]pyrene (BP). This study analyzed urine of cigarette smokers, coal smoke-exposed women, and nonexposed controls for the presence and quantities of the depurinated BP-adducted DNA bases, 7-(benzo[a]pyren-6-yl)guanine (BP-6N7Gua) and 7-(benzo[a]pyren-6-yl)adenine (BP-6-N7Ade). Since these adducted bases originate from reaction of the BP radical cation with double-stranded DNA and not with RNA or denatured DNA, their presence in urine is indicative of DNA damage. Urine samples were fractionated by a combination of SepPak extraction and reverse-phase HPLC, and then analyzed by tandem mass spectrometry and capillary electrophoresis with laser-induced fluorescence. BP-adducted bases were detected in the urine from three of seven cigarette smokers and three of seven women exposed to coal smoke, but were not detected in urine from the 13 control subjects. Concentrations were estimated to be 60-340 and 0.1-0.6 fmol/mg of creatinine equivalent of urine for coal smoke-exposed women (maximum possible BP intake of ca. 23 000 ng/day) and cigarette smokers (BP intake of ca. 800 ng/day), respectively, exhibiting a sensitive response to BP exposures. BP-6-N7Gua was present at ca. 20-300 times the concentration of BP-6-N7Ade in the urine of coal smoke-exposed women, but was not detected in the urine of cigarette smokers. This difference may be due to the remarkably different BP exposures experienced by the two groups of PAH-exposed individuals. These results justify more extensive studies of depurinated BP-adducted DNA bases as potential biomarkers of PAH-associated cancer risk.

Introduction A major goal of molecular epidemiology is to improve the accuracy of determining individual risk of environmentally induced cancers (1, 2). Risks of cancers associated with exposures to polycyclic aromatic hydrocarbons (PAH)1 have received particular attention. Especially high exposures to PAH may be due to occupation, residence in polluted areas, and lifestyle, e.g., cigarette smoking. The very potent carcinogens benzo[a]pyrene * To whom correspondence should be addressed. † Eppley Institute for Research in Cancer, University of Nebraska Medical Center. ‡ Washington University. § Iowa State University. | Internal Medicine-Pulmonary, University of Nebraska Medical Center. ⊥ U.S. Environmental Protection Agency.

(BP) and dibenzo[a,l]pyrene (DB[a,l]P) (3) are present in coal smoke (4) and in cigarette smoke (5), and exposure to coal smoke (6) and cigarette smoke (7, 8) is associated with high incidences of lung cancer. In a study of 250 000 U.S. veterans, Rogot and Murray (8) found that the ageadjusted rate of lung cancer mortality was 65.2 per 1 Abbreviations: B[a]P, benzo[a]pyrene; BPDE, benzo[a]pyrene 7,8diol 9,10-epoxide; BP-6-N7Gua, 7-(benzo[a]pyren-6-yl)guanine; BP-6N7Ade, 7-(benzo[a]pyren-6-yl)adenine); CC, capillary cryostat; CrEq, creatinine equivalents; DB[a,l]P, dibenzo[a,l]pyrene; CE/FLN, capillary electrophoresis with fluorescence line narrowing; CE/LIF, capillary electrophoresis with laser-induced fluorescence; DMBA, 7,12-dimethylbenz[a]anthracene; Me2SO, dimethyl sulfoxide; ELISA, enzyme-linked immunosorbent assay; FLN, fluorescence line narrowing; LCQ/MS, liquid chromatography/quadrupole ion-trap mass spectrometry; LC/ MS/MS, liquid chromatography/tandem mass spectrometry; MEKC, micellular electrokinetic chromatography; MRM, multiple-reaction monitoring; PAH, polycyclic aromatic hydrocarbons; PBL, peripheral blood leukocytes.

10.1021/tx000012y CCC: $20.00 © 2001 American Chemical Society Published on Web 01/10/2001

BP-Adducted DNA Bases in Urine

100 000 annually among cigarette smokers, compared with 5.8 per 100 000 expected among nonsmokers. Mumford et al. (6) documented a remarkable 904 lung cancer deaths per 100 000 annually among women (ages 5559) living in three high-mortality communes in Xuan Wei, China. The women were exposed to unvented household smoke produced by combustion of smoky coal, for cooking and heating. The age-adjusted lung cancer mortality for all women in the three communes was 125 per 100 000 annually, compared with 3.2 per 100 000 for all women living in China (6). Data of this kind have stimulated strong interest in developing effective methods for early determination of individual risk of these cancers. BP has been used as an index carcinogen for determining risk of PAH-associated cancer (9). Laboratory investigations have identified BP 7,8-diol 9,10-epoxide (BPDE) as the carcinogenic species which forms a stable adduct with DNA by covalently binding to the amino group of guanine or adenine (10). Numerous transitional studies have evaluated stable PAH or aromatic adducts to DNA as biomarkers of individual risk of PAH-associated cancers (11-20). Current approaches to quantifying stable PAH-DNA adducts include the enzyme-linked immunosorbent assay (ELISA), 32P-postlabeling, and fluorescence spectroscopy. These procedures were evaluated in transitional studies aimed at comparing the number of adducts per unit of DNA with some measure of exposure to PAH (16, 2127). Of the three procedures, the ELISA is most suited to large-scale epidemiology studies or clinical evaluations because it is a rapid and low-cost procedure not requiring radiolabeled compounds. A disadvantage of this procedure, as well as 32P-postlabeling, is the lack of structural specificity. Both assays detect diverse PAH-DNA adducts, which include BPDE-DNA adducts, with a high degree of sensitivity. It is possible, consequently, that two samples exhibiting the same quantitative value may have different proportions of BPDE-DNA adducts. Fluorescence analysis of BP tetrols released by acid hydrolysis of BPDE-DNA adducts is structurally specific and circumvents those problems associated with the ELISA and 32P-postlabeling (25-27). There is another class of PAH-DNA adducts that may provide effective biomarkers for determining individual risk of PAH-associated cancers. PAH are metabolized to DNA-damaging diol epoxides and radical cations that bind covalently to C-8 or N-7 of guanine or to N-3 or N-7 of adenine. These linkages destabilize the N-glycosyl bond to the deoxyribose phosphate backbone of the DNA, and generate apurinic sites (28). Depurinating adducts provide a straightforward mechanism for PAH-associated G f T transversions in ras and p53 (29-32). In mouse skin treated with BP, 71% of all BP-DNA adducts are depurinating adducts generated by the BP radical cation and BPDE (33). The depurinating adducts 7-(benzo[a]pyren-6-yl)guanine (BP-6-N7Gua), 8-(benzo[a]pyren-6yl)guanine, and 7-(benzo[a]pyren-6-yl)adenine (BP-6N7Ade), formed by the BP radical cation, account for 10, 34, and 22%, respectively, of all BP-DNA adducts that are formed. Depurinated BP-adducted DNA bases have been detected in the urine of rats that were given a single dose of BP (34). Consequently, these adducted bases are potentially useful for determining cancer risk associated with PAH exposures. Toward this end, we developed a monoclonal antibody for specific, high-affinity capture of both BP-6-N7Gua and BP-6-N7Ade from urine samples

Chem. Res. Toxicol., Vol. 14, No. 2, 2001 193

(35). Recently, gold biosensor chips were developed for detection of depurinating carcinogen-DNA adducts by fluorescence line-narrowing spectroscopy (36). In this article, we present the results of our work aimed at evaluating depurinated BP-adducted DNA bases as biomarkers for determining the risk of cancers associated with PAH exposures. The specific goals of this work are (1) to determine whether the depurinated BP-adducted DNA bases BP-6-N7Ade and BP-6-N7Gua can be detected in the urine of individuals exposed to PAH and (2) to determine the relative abundance of these adducts, if they can be detected in the urine of individuals who experience different PAH exposures.

Experimental Procedures Chemicals. The BP-adducted bases BP-6-N7Ade and BP-6N7Gua were synthesized by iodine oxidation of BP in the presence of 2′-deoxyguanosine, 2′-deoxyadenosine, or adenine (37). These served as standards for protocol development, and for detection and quantification of depurinated BP-adducted DNA bases in human urine. Collection of Urine Samples. Study subjects were asked to provide urine first passed in the morning. Each study subject completed a consent form and a questionnaire that provided the following information: age, sex, history of tobacco use, health status, and dietary and occupational exposure to PAH. Our studies were carried out according to the recommendations of the World Medical Association Declaration of Helsinki. They were approved by (1) the Institutional Review Board for the Protection of Human Subjects, University of Nebraska Medical Center, and (2) the U. S. Environmental Protection Agency for international research projects. Urine samples were collected from eight Chinese women (3755 years old) exposed to coal smoke and seven nonexposed controls (30-55 years old). The exposed subjects, who maintained households in Xuan Wei, China, were exposed to unvented smoke produced by combustion of smoky coal, which served as fuel for both household heating and cooking. Twentyfour hour mean concentrations of household BP were 303-1970 ng/m3 of air. In addition, for as many as 7 h/day, these women were near the coal fire where the average concentration of BP was 14 650 ng/m3 of air (4). Nonexposed controls, comprising seven Chinese-American women, reported only ambient exposure to air-born PAH. Neither the exposed subjects nor the nonexposed controls used tobacco products, consumed (2 weeks before urine collection) foods rich in PAH, or experienced occupational exposure to PAH. Study subjects provided urine first passed in the morning. Urine samples were collected from smokers and nonsmokers at the Pulmonary and Critical Care Medicine Section of the Department of Internal Medicine, University of Nebraska Medical Center. The smokers included six males (24-65 years old) and six females (22-51 years old), who smoked an average of 30 cigarettes/day (range of 20-45 cigarettes/day). The nonsmokers, including five males (41-60 years old) and six females (29-41 years old), indicated only ambient exposure to PAH, and did not have a recent (20 years or more) history of using tobacco products. None of the subjects experienced occupational exposure to PAH or consumed (2 weeks before urine collection) foods containing high concentrations of PAH. The subjects provided urine first passed in the morning. The creatinine concentration of each urine sample was determined by a colorimetric procedure (Stanbio Creatinine Procedure 0400, Sigma-Aldrich, St. Louis, MO), upon receipt at the laboratory. All urine samples were stored at -80 °C. Preparation of Urine Samples for Preparative HPLC. In one set of analyses, urine collected from cigarette smokers and nonsmokers was thawed and then pooled to yield a volume containing 800 mg of creatinine. Each sample was divided into two equal volumes of 400 mg of creatinine each, and then 1.0

194

Chem. Res. Toxicol., Vol. 14, No. 2, 2001

pmol of the BP-6-N7Ade standard was added to 1 volume. The second volume was left untreated. The pH was adjusted to 7.0 with 3 M NaOH, and then CH3CN was added to give a final concentration of 20% (v/v). For the balance of the analyses, urine samples collected from PAH-exposed and nonexposed subjects and containing 6-100 mg of creatinine were adjusted to pH 7.0 and then to 20% CH3CN. BP-6-N7Ade was not added to these samples. A precipitate, formed by addition of CH3CN to the urine samples, was removed by centrifugation. All urine samples, at 20% CH3CN, were fractionated with SepPak tC18 cartridges (Waters, Inc., Milford, MA). No more than 150 mL of sample or 150 mg of creatinine equivalents (CrEq) of urine was applied to each cartridge (1 g). After sample application, each cartridge was washed with 20 mL of a 20:80 CH3CN/H2O mixture and then eluted with 20 mL of a 60:40 CH3CN/H2O mixture, followed by elution with 100% CH3CN. The guanine and adenine adducts were found exclusively in the fraction eluted with the 60:40 CH3CN/H2O mixture. This fraction was extracted with CHCl3, and the extract was concentrated by rotary evaporation to ca. 10 mL. At this point, 200 µL of dimethyl sulfoxide (Me2SO) was added, and the CHCl3 was evaporated, leaving the urinary constituents dissolved in 200 µL of Me2SO. This procedure prevented excessive loss of adducted bases that occurred when the CHCl3 extract was evaporated to dryness and the dried constituents were redissolved in Me2SO. These samples were fractionated by preparative HPLC. Preparative HPLC for Detection and Quantification of BP-6-N7Ade and BP-6-N7Gua in Urine. Urine samples, prepared by SepPak fractionation and CHCl3 extraction, were fractionated by reverse-phase HPLC with a Waters 625 LC system (Waters, Inc.). Each 200 µL sample was applied to a 250 mm × 10 mm YMC-ODS-AQ column (Alltech, Deerfield, IL) fitted with a 30 mm × 10 mm YMC-ODS-AQ guard column (Alltech) and equilibrated with a 50:50 CH3OH/H2O mixture. The column was eluted at a rate of 2 mL/min with the following program: (1) from 0 to 5 min, 50:50 CH3OH/H2O; (2) from 5 to 48 min, a linear gradient to 70:30 CH3OH/H2O; (3) from 48 to 61 min, 70:30 CH3OH/H2O; (4) from 61 to 73 min, a linear gradient to 75:25 CH3OH/H2O; (5) from 73 to 85 min, 75:25 CH3OH/H2O; (6) from 85 to 95 min, a linear gradient to 100% CH3OH; and (7) from 95 to 160 min, 100% CH3OH. The column eluate was monitored at 427 nm with a Waters 474 scanning fluorescence detector (Waters, Inc.) set to excite at 365 nm. Elutions from 65 to 70 min (fraction 1) and from 70 to 81 min (fraction 2) were collected for detection and quantification of the depurinated BP-adducted DNA bases, BP-6-N7Gua (fraction 1) and BP-6-N7Ade (fraction 2). To ensure the absence of carryover from one fractionation to the next, we injected the HPLC column with Me2SO, alone, after each sample fractionation, and collected the same fractions. All fractions were extracted with CHCl3, concentrated to 30 µL in Me2SO, and then coded and shipped to the mass spectrometry and fluorescence spectroscopy groups, who were not provided prior information about PAH exposures of the corresponding subjects. In one set of analyses, HPLC fraction 2 was extracted with CHCl3, concentrated to 200 µL in Me2SO, and then fractionated on the same preparative column equilibrated with a 35:65 CH3CN/H2O mixture. The column was eluted at a rate of 2.5 mL/min with the following program: (1) from 0 to 5 min, 35:65 CH3CN/H2O; (2) from 5 to 40 min, a linear gradient to 65:35 CH3CN/H2O; (3) from 40 to 50 min, a linear gradient to 100% CH3CN; and (4) from 50 to 120 min, 100% CH3CN. The eluate from 33 to 39 min was collected, extracted with CHCl3, concentrated to 30 µL in Me2SO, and then analyzed for BP-6N7Ade by tandem mass spectrometry. Identification and Quantification of Urinary BP-6N7Ade by Liquid Chromatography/Triple-Quadrupole Mass Spectrometry (LC/MS/MS). HPLC fractions of urine were screened for the presence of BP-6-N7Ade by LC/MS/MS. The data were collected with a PE Sciex API 365 (PE Sciex, Foster City, CA) triple-quadrupole mass spectrometer equipped

Casale et al. with a Sciex turbo-ion-spray inlet. The mass spectrometer was operated in the positive-ion mode, and the ion-spray needle, orifice, and ring electrodes were maintained at 5000, 55, and 300 V, respectively. Nitrogen was used separately as the collision, turbo-ion-spray, nebulizer, and curtain gas. The turboion-spray gas was heated to 350 °C to assist in desolvation. Data were obtained in the multiple-reaction monitoring (MRM) mode to achieve the best sensitivity for BP-6-N7Ade. The Q1 was set to transmit the precursor ion of m/z 386, and the Q3 was set to transmit the product ion of m/z 251 and 252. The collision energy was between 35 and 50 eV, and the dwell time for each MRM transition was set at 400 ms. The LC separations were performed by using a 2 mm × 100 mm Keystone Betasil C-18 column and Shimadzu LC-10-AS pumps, an SCL-10A controller, and an SIL-10A auto injector. The mobile phase, which consisted of 25 mM ammonium acetate with the pH adjusted to 4.0 with acetic acid (solvent A) and methanol (solvent B), was maintained at a constant flow rate (0.5 mL/min). Gradient elution of BP-6-N7Ade was achieved with programmed changes (linear) in mobile phase composition, as follows: Start with 1% B and hold for 1 min, increase to 70% B over the course of 2 min, increase to 90% B over the course of 2 min, hold at 90% B for 2 min, increase to 99% B and hold for 3 min, and equilibrate at 1% B for 3 min. Concentrations of BP-6-N7Ade were estimated with a calibration plot that was linear over a range of 0.1-10 pmol. Identification and Quantification of Urinary BP-6N7Gua and BP-6-N7Ade by Liquid Chromatography/ Quadrupole Ion-Trap Mass Spectrometry (LCQ/MS). HPLC fractions of urine were analyzed for BP-6-N7Gua and BP-6N7Ade at different times using slightly different procedures. For BP-6-N7Gua, the sample, in Me2SO, was diluted 50:50 (v:v) with starting buffer and 10 µL was injected via a Rheodyne 7125 injector onto a 0.3 mm × 150 mm Zorbax C18 column (MicroTech, Sunnyvale, CA) fitted with a 0.3 mm × 10 mm guard column of the same packing material. The column was eluted as follows. Solvent A (H2O) was kept constant at 75% for the first 4 min, decreased to 30% over the next 8 min, and then decreased to 10% over the next 13 min. Solvent B (2-propanol) and solvent D (methanol) were kept constant at 8% for the first 4 min, increased to 30% over the next 8 min, and then increased to 40% over the next 13 min. Solvent C (10% CH3COOH, 100 mM NH4OCOCH3, and H2O) was kept constant at 9% for the first 4 min, then increased to 10% over the next 8 min, and then kept constant at 10% for the next 13 min. The final solvent composition was kept constant for the next 5 min to elute all sample components. The column was eluted at a flow rate of 4.0 µL/ min. Solvent flow from the HPLC pump (Waters 600MS) was split with an LC Packings Accurate Splitter fitted with a 0.3 mm column calibrator (LC Packings, San Francisco, CA). All the column eluent was introduced into a Finnigan liquid chromatography/quadrupole (LCQ) ion-trap mass spectrometer (Finnigan, San Jose, CA), by electrospray ionization. Eluent from the HPLC column was sprayed into a heated capillary (200 °C), where the spray voltage was 4.5 kV and the nitrogen sheath gas flow rate was 60 mL/min. The mass spectrometer was set in the “MS/MS” mode where the protonated BP-6-N7Gua (M + H+; m/z 402.2) was isolated in a window that was 1.5 m/z wide. The resonant excitation energy was set at 40% of maximum, and the scan range was from m/z 110 to 410. Spectra were acquired in the “profile” mode, and each spectrum was the average of two scans of 1000 ms each. BP-6-N7Gua was considered present in a sample if its principal fragment ions (m/z 252, 277, 360, 378, and 385) were detected. Signals seen elsewhere in the product-ion mass spectra represent ions of coeluents. Concentrations of BP-6-N7Gua were estimated with a calibration plot of the peak area of the m/z 252 ion versus the quantity of injected standard. The plot was linear over a range of 5-25 fmol. For analysis of BP-6-N7Ade, the sample, in 20 µL of Me2SO, was diluted 50:50 (v:v) with starting buffer and then injected

BP-Adducted DNA Bases in Urine with a Rheodyne 7125 injector, onto a 0.3 mm × 50 mm Zorbax C18 column (MicroTech) fitted with a 0.3 mm × 10 mm guard column of the same packing material. The column was eluted as follows. Solvent A (H2O) was kept constant at 85% for the first 7 min, decreased to 30% over the next 5 min, and then decreased to 10% at a rate of 1.0%/min. Solvent B (50% CH3OH and 50% 2-propanol) was kept constant at 5% for the first 7 min, increased to 60% over the next 5 min, and then increased to 80% at a rate of 1%/min. Solvent C (10% CH3COOH, 100 mM NH4OCOCH3, and H2O) was kept constant at 10% during the run. The column was eluted at a rate of 4.0 µL/min. Solvent flow from the HPLC pump (Waters 600 mass spectrometer with a pulse-damping silk board) was split with an LC Packing Accurate Splitter fitted with a 0.3 mm column calibrator (LC Packing). All of the column eluate was introduced into a Finnigan liquid chromatography/quadrupole (LCQ) ion-trap mass spectrometer (Finnigan), by electrospray ionization. Eluate from the HPLC column was sprayed into a heated capillary (200 °C). The spray voltage was 5.7 kV, and the nitrogen sheath gas flow was 55 mL/min. The mass spectrometer was set in the “MS/MS” mode where the molecular ion of BP-6-N7Ade (M + H+; m/z 386.2) was isolated in a window that was 1.5 m/z wide. Resonant excitation energy was set at 25% of maximum, and the scan range was from m/z 240 to 390. Spectra were taken in the profile mode, and each spectrum was the average of two scans of 1000 ms each. BP-6-N7Ade was considered present in a sample if its principal fragment ions (m/z 369, 342, and 252) were detected. Abundance ratios were variable because the fragments were present so close to their detection limits, and consequently, ioncounting statistics were low. Signals seen elsewhere in the tandem mass spectra represent fragment ions of coeluates and not electronic noise. Concentrations of BP-6-N7Ade were estimated with a calibration plot of the peak area of the m/z 369 ion versus the quantity of injected standard. The plot was linear over a range of 5-25 fmol. Identification and Quantification of Urinary BP-6N7Ade and BP-6-N7Gua by Capillary Electrophoresis with Laser-Induced Fluorescence (CE/LIF). HPLC fractions were analyzed for both BP-6-N7Ade and BP-6-N7Gua by micellar electrokinetic chromatography (MEKC) coupled with LIF (38). The separation buffer (pH 9.0) consisted of 20 mM dioctylsulfosuccinate (micelle-forming surfactant), 4 mM sodium tetraborate, and 15% acetonitrile. Before injection, the separation buffer was filtered through a 0.22 µm filter and then degassed for 5 min. Samples (dissolved in Me2SO) were hydrodynamically injected (20 mBar for 1.8-3 s), resulting in approximately 10 nL injection volumes. Prior to injection, samples were warmed to room temperature, agitated for 1 min, and sonicated for 30 s to reduce adhesion of the adducts to the walls of the plastic sample tubes. UV-transparent separation capillaries (85 cm in length with a 75 cm effective length; 75 µm i.d., 365 µm o.d.) were conditioned with 0.1 M NaOH for 15 min (30 min for new capillaries), water for 5 min, and separation buffer for 15 min. The capillary was rinsed with several volumes of separation buffer between runs and completely reconditioned every 3 h. A DC voltage (22 kV) was applied across the capillary, generating a current of 19 µA. MEKC separation of the urinary components was monitored as a plot of the integrated fluorescence signal (385-500 nm) versus time, producing an electropherogram of the separated sample components. The CE/LIF electropherograms were obtained with a CW excitation source (Model Innova 90C argon ion laser, Coherent, Santa Clara, CA) operating at 351.1 nm with an output power of 100 mW. Fluorescence was collected with a reflecting objective (Ealing, Holliston, MA), passed through a model 218 0.3-m monochromator (McPherson Inc., Acton, MA), and detected by an intensified charge-coupled device (Roper Scientific, Trenton, NJ). The migration times of the adducted bases were established with standards. BP-6-N7Ade moves past the laser beam ca. 3 min later than BP-6-N7Gua and is well separated from the latter.

Chem. Res. Toxicol., Vol. 14, No. 2, 2001 195 Adduct quantification was established via standard curves that were linear over a range of 5-200 amol. CE-fluorescence line-narrowing (CE/FLN) studies were performed in a specially designed capillary cryostat (CC) (39). The CC consisted of a double-walled quartz cell (8 cm in length) with inlet and return lines for liquid helium introduction. The separation capillary was positioned in the central region of the CC. The outer portion of the CC was evacuated. As the molecules of interest migrated into the portion of the separation capillary that was within the 8 cm quartz cell, liquid helium was introduced into the CC. Because of the low thermal capacity of the capillary and the small dimensions of the CC, cooling to 4.2 K was achieved in less than 1 min. The rapid cooling rate provided the disordered matrix needed for FLN spectroscopy in typical CE buffers (38, 40). A precision stage provided translation of the CC along the capillary axis by (4 cm, allowing the separated analyte zones to be sequentially characterized by FLN spectroscopy. FLN spectra were obtained with selective excitation from a Lambda Physik FL-2002 pulsed dye laser, pumped by a Lambda Physik Lextra 100 XeCl excimer laser (Lambda Physik, Ft. Lauderdale, FL). Detection was accomplished with the setup described above. These low-temperature spectra were obtained in both gated and nongated modes of detection with a 200 ns gate width and various delay times. The spectral resolution was 0.05 nm.

Results Detection of 1 pmol of BP-6-N7Ade Added to Urine Collected from Cigarette Smokers and Nonsmokers. The levels of fluorescent constituents were increased substantially in urine collected from cigarette smokers and coal smoke-exposed individuals. This is evident by comparison of the time-integrated fluorescence signal of HPLC chromatograms generated with SepPak fractions of urine samples containing the same amount of total creatinine (Figure 1). LC/MS/MS and CE/LIF analysis of SepPak fractions prepared with urine from cigarette smokers revealed complex profiles that precluded identification of the adducted bases (data not shown). Our goal, consequently, was to develop a combined SepPak fractionation/HPLC purification procedure for urine samples representing up to 25% (400 mg of CrEq) of the daily excretory output, and containing the smallest quantity of added BP-6-N7Ade standard that could be identified and quantified by reverse-phase HPLC with fluorescence detection. We elected to work with BP-6N7Ade, alone, because this would simplify protocol development and because the newly developed conditions should be applicable to both BP-6-N7Ade and BP-6N7Gua. Samples of urine from cigarette smokers and nonsmokers were treated by addition of 1 pmol of BP-6-N7Ade or were left untreated. SepPak fractions of these samples were fractionated by reverse-phase HPLC with a MeOH/ H2O program (cf. Figure 1). Components eluting at 7081 min (fraction 2, BP-6-N7Ade) were collected and reanalyzed by reverse-phase HPLC using a CH3CN/H2O program. Results obtained with urine from a cigarette smoker are representative. Peaks coeluting with BP-6N7Ade were not resolved in the first HPLC fractionation (Figure 2A,B), but were detected after the second HPLC fractionation (Figure 2C,D). Added BP-6-N7Ade standard was resolved as a sharp peak with a high signal-tobackground ratio (Figure 2D). This protocol permitted high resolution and reproducible identification and quantification of 1.0 pmol of BP-

196

Chem. Res. Toxicol., Vol. 14, No. 2, 2001

Figure 1. High-performance liquid chromatography (HPLC) profiles of urine samples collected from a cigarette smoker, a coal smoke-exposed resident of Xuan Wei, China, and corresponding controls. A SepPak fraction of each urine sample was analyzed by reverse-phase HPLC with a CH3OH/H2O program. Eluate excited at 365 nm was monitored for emission at 427 nm. (A) One hundred milligrams of creatinine equivalents (CrEq) of urine from a cigarette smoker (CSm-7 of Table 1, 42year-old male). (B) One hundred milligrams of CrEq of urine from a nonsmoker (41-year-old male). (C) Thirty-five milligrams of CrEq of urine from a coal smoke-exposed woman (CSE-7 of Table 1, 40 years old). (D) Thirty-five milligrams of CrEq of urine from a Chinese-American woman (36 years old) who was not exposed to cigarette or coal smoke. The solid horizontal bars (F1 and F2) bracket the retention times of standards added to human urine samples, and indicate the fractions collected for further analysis. BP-6-N7Gua elutes between 65 and 70 min (F1), and BP-6-N7Ade elutes between 70 and 81 min (F2).

6-N7Ade added to 400 mg of CrEq (ca. 400 mL) of urine collected from both nonsmokers and smokers. Recovery efficiencies were 70-75% as determined by comparing peak areas of the standard added to urine with peak areas of the standard in Me2SO alone. These areas were linear over a range of 0.1-2.0 pmol of BP-6-N7ade. Split-Sample Study of BP-6-N7Ade in Urine Collected from PAH-Exposed and Control Individuals. The protocol was then applied in a split-sample study to detect and quantify BP-6-N7Ade in urine collected from cigarette smokers and nonsmokers, and one coal smokeexposed individual. Two or three urine samples collected from each of five smokers and five nonsmokers were combined such that each pool, from one individual, contained 800 mg of creatinine. Each pool was split into two equal volumes, and 1.0 pmol of BP-6-N7Ade standard was added to only one of the volumes. The split-sample design provided for (1) precise tracking of BP-6-N7Ade in individual urine samples, (2) evaluation of the purification protocol for each urine sample, (3) evaluation of the analysis of BP-6-N7Ade by LC/MS/MS, in the context of constituents present in individual urine samples from PAH-exposed and control individuals, and (4) quantitative estimation of BP-6-N7Ade in urine samples of exposed and control individuals, against the standard detected in the same environment. Because the quantity (6 mg of CrEq) of urine from the Chinese woman exposed to coal smoke was small, and availability of samples from coal smoke-exposed individuals was limited, this urine sample was analyzed without addition of the BP-6-N7Ade standard. Components eluting at 33-39 min, during HPLC fractionation with the CH3CN/H2O program (Figure

Casale et al.

Figure 2. Purification of the BP-6-N7Ade standard added to urine collected from a cigarette smoker. Pooled urine of a cigarette smoker (CSm-A, 65-year-old male) was divided into two equal volumes of 400 mg of creatinine equivalents, each. BP-6-N7Ade (1.0 pmol) was added to one volume, and the remaining volume was left untreated. A SepPak fraction of each sample was fractionated by reverse-phase HPLC with a CH3OH/ H2O program. Fraction 2 (F2, 70-81 min) was analyzed by reverse-phase HPLC with a CH3CN/H2O program. (A) HPLC (CH3OH/H2O) of a SepPak fraction of untreated urine. (B) HPLC (CH3OH/H2O) of a SepPak fraction of urine treated by addition of 1.0 pmol of BP-6-N7Ade standard. (C) HPLC (CH3CN/H2O) of fraction 2 (F2) of untreated urine. (D) HPLC (CH3CN/H2O) of fraction 2 (F2) of urine treated by addition of 1.0 pmol of BP6-N7Ade standard. S indicates the BP-6-N7Ade standard peak. The vertical arrows indicate the retention time of the BP-6N7Ade standard added to human urine.

2C,D), were analyzed by LC/MS/MS. BP-6-N7Ade was evident in untreated samples from three of the five smokers. The adducted base was identified as a resolved peak with an area equal to or greater than 15-20% of the peak area of the standard added to the same urine sample. A representative spectrum demonstrates detectable BP-6-N7Ade (m/z 386 f 251 at 6.0 min) in urine from a cigarette smoker (Figure 3A). This was confirmed by a marked increase in the peak area and signal-tobackground ratio, upon addition of the BP-6-N7Ade standard to the corresponding split sample (Figure 3B). Urine from a second cigarette smoker lacked detectable BP-adducted adenine (Figure 3C). BP-6-N7Ade was evident in the untreated sample from the coal smokeexposed individual (Figure 3D), but was not evident in untreated samples from the five nonsmokers (data not shown). The ratio of peak areas, corresponding to the amount of BP-6-N7Ade in untreated urine samples to the amount of standard (1.0 pmol) added to urine, provided a minimal estimate of the quantity of BP-6-N7Ade in untreated urine samples. Urine from the BP-6-N7Ade positive smokers contained an estimated 0.5-1.0 fmol of the adducted base/mg of creatinine, and urine from the coal smoke-exposed individual contained an estimated 20 fmol/mg of creatinine. Detection and Quantification of BP-6-N7Ade and BP-6-N7Gua in Urine Collected from Cigarette Smokers and from Chinese Women Exposed to Household Coal Smoke by Capillary Electrophoresis/Laser-Induced Fluorescence (CE/LIF) and by Liquid Chromatography-Quadrupole Ion-Trap Mass Spectrometry (LCQ/MS). On the basis of the information acquired in our split-sample study of BP-6-N7Ade,

BP-Adducted DNA Bases in Urine

Chem. Res. Toxicol., Vol. 14, No. 2, 2001 197 Table 1. Detection and Quantification of the Depurinated Benzo[a]pyrene-Adducted DNA Bases, BP-6-N7Ade and BP-6-N7Gua, in the Urine of Cigarette Smokers and Women Exposed to Coal Smokea detectionb of BP-6-N7GUA detectionb of BP-6-N7ADE

Figure 3. Identification of BP-6-N7Ade in human urine by liquid chromatography/triple-quadrupole mass spectrometry (LC/MS/MS). Urine samples of two cigarette smokers were divided into two equal volumes of 400 mg of creatinine equivalents, each. For each subject, 1.0 pmol of BP-6-N7Ade was added to one volume and the second volume was left untreated. An untreated sample of urine (6 mg of creatinine equivalents) collected from a Xuan Wei resident exposed to coal smoke was also analyzed. A SepPak fraction of each urine sample was fractionated by reverse-phase HPLC with a CH3OH/H2O program. Fraction 2 (F2, 70-81 min) was fractionated by reversephase HPLC with a CH3CN/H2O program (cf. Figure 2). Eluate collected from 33 to 39 min (bracketing the 37 min retention time of BP-6-N7Ade) was analyzed by LC/MS/MS. The mass spectrometer was set to select first for the parent ion (m/z 386.2) and then for a major product ion (m/z 251.0). Each panel presents selected-ion monitoring of m/z 386.2 f 251.0, in real time. The retention time of the BP-6-N7Ade standard was 6.0 min. (A) Untreated urine of a cigarette smoker (CSm-A of Figure 2). (B) Urine collected from cigarette smoker CSm-A and treated by addition of 1.0 pmol of the BP-6-N7Ade standard. (C) Untreated urine of a second cigarette smoker (CSm-B, 57-yearold male). BP-6-N7Ade was not detected in this sample. (D) Untreated urine of a Xuan Wei woman exposed to coal smoke (CSE-A, 55 years old).

we analyzed urine samples for the presence and abundance of both BP-6-N7Gua and BP-6-N7Ade. Untreated urine samples from seven smokers (100 mg of CrEq), six nonsmokers (100 mg of CrEq), seven coal smoke-exposed women (10-30 mg of CrEq), and seven nonexposed women (10-30 mg of CrEq) were fractionated with SepPak cartridges followed by HPLC with a MeOH/H2O program. Fractions 1 and 2 (cf. Figure 1) were analyzed by CE/LIF and by LCQ/MS. Depurinated BP-adducted bases were detected in urine from three of the seven coal smoke-exposed women and from three of the seven smokers (Table 1), whereas adducted bases were not detected in urine from the nonexposed controls. To ensure the absence of carryover from one fractionation to the next, we injected the HPLC column with Me2SO, alone, after each sample fractionation. The fractions from these “blank” runs lacked detectable BP-6-N7Gua and BP-6N7Ade. BP-6-N7Gua was quantified by CE/LIF from its timeintegrated fluorescence signal (Figure 4A) and by LCQ/ MS from the time-integrated signal of its m/z 252 product ion (Figure 5A). Proof of structure was obtained by CE/ LIF under fluorescence line-narrowing (FLN) conditions (Figure 4C) and by LCQ/MS (Figure 5). The urine sample of the coal smoke-exposed woman produced a fragmentation pattern (Figure 5A) which is the average of the product-ion spectra acquired from 34.7 to 35.9 min. This

subjectc

CE/LIFd

LCQ/MSe

CE/LIF

LCQ/MS

CSE-1 CSE-2 CSE-3 CSE-4 CSE-5 CSE-6 CSE-7 CSm-1 CSm-2 CSm-3 CSm-4 CSm-5 CSm-6 CSm-7

160 f

290 340 60 -

10 NQ 0.6 NQ NQ -

4.4 NQh NQ 0.6 NQ NQ -

-g 100 130 -

a BP-adducted bases were not detected in urine from any of the 13 control subjects. b When detected, BP-6-N7Gua was in HPLC fraction 1 and BP-6-N7Ade was in HPLC fraction 2 (cf. Figure 1), except for CSE-3 where BP-6-N7Gua (nonquantifiable) trailed into fraction 2, as determined by LCQ/MS. c An individual exposed to coal smoke (CSE-x) or a cigarette smoker (CSm-x). d Capillary electrophoresis/laser-induced fluorescence. e Liquid chromatography/ quadrupole ion-trap mass spectrometry. f Estimated concentration as femtomoles of BP-adducted base per milligram of creatinine equivalents of urine (roughly 1 mL of urine). g None detected. h Detected but not quantifiable.

pattern shows clearly the fragment ions at m/z 252, 277, 360, 378, and 385 that are characteristic of BP-6-N7Gua (Figure 5B). Three other abundant ions of m/z 265, 383, and 384 are not from BP-6-N7Gua, but from an interference that is at least 20 times more abundant and elutes at approximately 32 min. This component trails into the elution window of the BP-6-N7Gua. The precursor ion of the interference is likely to be of m/z 401, which is unavoidably coselected with the m/z 402 ion in the MS/ MS experiment. Consequently, m/z 383 and 384 are viewed as losses of 17 and 18 amu, respectively. Ignoring these interference ions, we see that the remaining spectrum is very similar to that of the reference material except that the ratio of abundances of the m/z 277/278 and 252 ions is greatly increased in the spectrum of the unknown. This suggests that coeluting with BP-6-N7Gua was BP-6-C8Gua, which would account for the high abundance ratio of the m/z 277/278 and 252 ions (41). Nonetheless, the contribution of BP-6-C8Gua to our quantitative estimates of BP-6-N7Gua likely is minimal since these estimates were based on peak areas of the m/z 252 ion that is generated with markedly lower efficiency from BP-6-C8Gua than from BP-6-N7Gua (41). BP-6-N7Ade was quantified by CE/LIF from its timeintegrated fluorescence signal (Figure 6) and by LCQ/ MS from the time-integrated signal of its m/z 369 ion (Figure 7). The presence of BP-6-N7Ade was proven by the combined results from CE/LIF (Figure 6) and LCQ/ MS (Figure 7). Results from detection and quantification of BP-6-N7Gua and BP-6-N7Ade in urine from PAHexposed individuals, by both procedures, were in remarkable agreement (Table 1). Profiles of BP-DNA bases in the urine of coal smokeexposed women and cigarette smokers were qualitatively and quantitatively distinct. Concentrations of the adducted bases were markedly higher in the coal smokeexposed women than in the cigarette smokers (Table 1),

198

Chem. Res. Toxicol., Vol. 14, No. 2, 2001

Figure 4. Capillary electrophoresis/laser-induced fluorescence (CE/LIF) demonstrating the presence and abundance of BP-6N7Gua in urine of a coal smoke-exposed resident of Xuan Wei, China. A SepPak fraction of urine from a coal smoke-exposed woman (CSE-1 of Table 1, 49 years old) was fractionated by reverse-phase HPLC with a CH3OH/H2O program. Fraction 1 (F1, 65-70 min, bracketing the retention time of BP-6-N7Gua) was extracted with CHCl3, concentrated to 30 µL in Me2SO, and then analyzed by CE/LIF. (A) Real-time electropherogram of the integrated fluorescence signal (385-500 nm). (B) On-line broadband (non-line-narrowed) fluorescence spectrum of the CEseparated peak at ∼22.5 min, obtained with excitation at 351.1 nm at 4.2 K. A fluorescence origin band near 405 nm is indicative of BP-DNA bases produced by the radical cation of BP. (C) Fluorescence line-narrowed (FLN) spectra obtained online for the ∼22.5 min peak (s) and off-line for the BP-6-N7Gua standard (- - -) (λex ) 386.5 nm, T ) 4.2 K). The numbers correspond to excited-state vibrational frequencies in inverse centimeters. These spectra clearly show that the peak at ∼22.5 min corresponds to BP-6-N7Gua.

whose urine contained quantities of BP-6-N7Ade close to our limit of detection (ca. 0.1 fmol/mg of CrEq of urine). BP-6-N7Gua was the predominant BP-adducted base in the urine of coal smoke-exposed women, exhibiting concentrations of 60-340 fmol/mg of CrEq of urine, but was not detected in the urine of cigarette smokers (Table 1).

Discussion The depurinated BP-adducted DNA bases, BP-6N7Ade and BP-6-N7Gua, are promising biomarkers of cancer risk associated with PAH exposures. As demonstrated in previous studies (28), these adducted bases are formed by cytochrome P450-dependent one-electron oxidation of BP to its radical cation. BP-6-N7Ade and BP6-N7Gua are formed by reaction of the radical cation of BP with double-stranded DNA, but not with denatured

Casale et al.

Figure 5. Liquid chromatography/quadrupole ion-trap mass spectrometry (LCQ/MS) demonstrating the presence and abundance of BP-6-N7Gua in the urine of a Xuan Wei resident exposed to coal smoke. A SepPak fraction of urine was fractionated by reverse-phase HPLC (CH3OH/H2O). Fraction 1 (F1, 6570 min, bracketing the elution time of BP-6-N7Gua) was extracted with CHCl3, concentrated to 30 µL in Me2SO, and then analyzed by LCQ/MS. The principal fragments were m/z 252.3, 277.3, 360.1, 378.0, and 385.1. m/z 402.3 (parent ion) f 385.1 was selected for real-time monitoring of the column eluate. Each panel presents a fragment scan. (A) A coal smoke-exposed woman (CSE-1 of Table 1, 49 years old). (B) The BP-6-N7Gua standard. The loss of 42 units (not depicted), producing the m/z 360.1 fragment, is probably due to a through-ring fragmentation eliminating H2NCN from the guanine ring. The fragmentation producing the m/z 378.0 fragment is unknown.

DNA or with RNA (42). Consequently, we conclude that BP-adducted bases detected in urine were produced by depurinating adducts in the DNA of the host. Furthermore, the urinary concentrations of BP-6-N7Ade and BP6-N7Gua were higher among individuals experiencing greater levels of exposure to BP. Adducted bases were present at 0.1-0.6 and 60-340 fmol/mg of CrEq of urine from cigarette smokers and coal smoke-exposed individuals, respectively. The respiratory intake of BP among the cigarette smokers is estimated to be ca. 800 ng/day, assuming 30 ng of BP/cigarette (16) and 90% retention of the BP-containing particulates inhaled. The respiratory intake of BP among the coal smoke-exposed women is estimated to be as high as ca. 23 000 ng/day. This estimate is based on inhalation of BP-contaminated air (14 650 ng/m3) vicinal to the coal fire, for 7 h/day (4). The BP intake of the nonexposed control subjects is estimated to be 9-40 ng/day (16). Urine of the three positive smokers contained BP-6-N7Ade at levels close to the limit of quantification (0.5 fmol/mg of CrEq of urine) and no detectable BP-6-N7Gua. Urine of the positive coal smokeexposed women contained 60-340 fmol of BP-6-N7Gua/ mg of CrEq of urine. Neither BP-6-N7Ade nor BP-6N7Gua was detected (detection limit of ca. 0.1 fmol/mg of CrEq of urine) in urine of the nonexposed controls.

BP-Adducted DNA Bases in Urine

Figure 6. Capillary electrophoresis/laser-induced fluorescence (CE/LIF) demonstrating the presence and abundance of BP-6N7Ade in urine collected from a Xuan Wei resident exposed to coal smoke. Urine of a coal smoke-exposed woman (CSE-1 of Table 1, 49 years old) was extracted with a SepPak cartridge and then fractionated by reverse-phase HPLC with a CH3OH/ H2O program. Fraction 2 (F2, 70-81 min, bracketing the elution time of BP-6-N7Ade) was extracted with CHCl3, concentrated to 30 µL in Me2SO, and then analyzed by CE/LIF. Separation was monitored as the integrated fluorescence signal (385-500 nm) vs time. BP-6-N7Ade was identified by comparing an untreated aliquot of urine (top electropherogram) with an aliquot of the same sample treated by addition of the BP-6N7Ade standard (bottom electropherogram). The added standard clearly identified the location of BP-6-N7Ade in the untreated sample. The peak area of the untreated sample provided an estimate of the concentration of BP-6-N7Ade.

Thus, among samples positive for BP-6-N7Ade or BP-6N7Gua, the concentration of depurinated bases exhibited a large range reflecting the range of the level of BP exposure. The profiles of depurinated BP-adducted bases are remarkably different for the two PAH-exposed groups. BP-6-N7Gua dominates in the urine of individuals exposed to coal smoke, exceeding the quantity of BP-6N7Ade by ca. 20-300-fold, whereas BP-6-N7Ade is the only BP-adducted base detected in the urine of cigarette smokers. The basis for these differences is not known, but it is clear that the levels of BP exposure of the two groups are markedly different both qualitativley and quantitatively. Coal smoke-exposed individuals experience ca. 20-30 times the level of exposure to BP compared with cigarette smokers. The latter, in contrast to coal smoke-exposed individuals, experience BP exposures in the context of many additional components unique to cigarette smoke (5). Given the relatively large concentrations of BP-6N7Gua in the urine of three coal smoke-exposed individuals, the absence of detectable adducted bases in urine samples of the remaining four coal smoke-exposed individuals is surprising. It is unlikely that this result is a consequence of incomplete purification of adducted bases,

Chem. Res. Toxicol., Vol. 14, No. 2, 2001 199

Figure 7. Liquid chromatography/quadrupole ion-trap mass spectrometry (LCQ/MS) demonstrating the presence and abundance of BP-6-N7Ade in urine collected from a cigarette smoker and a coal smoke-exposed resident of Xuan Wei, China. Urine was extracted with a SepPak cartridge and then fractionated by reverse-phase HPLC with a CH3OH/H2O program. Fraction 2 (F2, 70-81 min, bracketing the elution time of BP-6-N7Ade) was extracted with CHCl3, concentrated to 30 µL in Me2SO, and then analyzed by LCQ/MS. The principal fragment ions were m/z 252, 342, and 369. m/z 386.2 (parent ion) f 369 was selected for real-time monitoring of the column eluate. Each panel presents a fragment scan. (A) A coal smoke-exposed woman (CSE-1 of Table 1, 49 years old). (B) A cigarette smoker (CSm-2 of Table 1, 33-year-old female). (C) The BP-6-N7Ade standard.

by the combination of SepPak extraction and reversephase HPLC. Neither a fluorescent peak comigrating with the BP-6-N7Ade or BP-6-N7Gua standard nor product ions characteristic of these modified bases were seen above background levels by CE/LIF and LCQ/MS, respectively. It is also unlikely that this result stems from a sampling problem, since coal smoke exposure occurred on a daily basis and urine samples were collected during periods of documented exposure (4). This remarkable dichotomy in the quantity of adducted bases in their urine may be genetically determined, e.g., polymorphisms of cytochrome P450, which activate PAH to carcinogenic metabolites (2). If our estimates are representative of the population of coal smoke-exposed women, BP-6-N7Gua concentrations of approximately 50-100 fmol/mg of CrEq of urine may identify women at high risk of lung cancer. This can be determined in a prospective study. Depurinating BP-DNA adducts have unique characteristics as potential biomarkers of PAH-associated cancer risk. First, the profile of depurinating adducts produced by a series of carcinogenic PAH applied to mouse skin correlates with the profile of mutations in the H-ras oncogene of carcinogen-induced papillomas of the skin (28, 29). Ninety-nine percent of the DNA adducts formed by DB[a,l]P and 7,12-dimethylbenz[a]anthracene (DMBA) were depurinating adducts, and in both cases, the large majority (ca. 80%) of these were adenine adducts. In DB-

200

Chem. Res. Toxicol., Vol. 14, No. 2, 2001

[a,l]P- and DMBA-induced papillomas, the majority (80 and 100%, respectively) of H-ras mutations were A f T transversions as expected on the basis of misreplication of unrepaired apurinic sites (28, 29). Depurinating adducts represented 71% of all the BP-DNA adducts formed in BP-treated mouse skin, and consisted of 46% guanine adducts and 25% adenine adducts (33). In BPinduced mouse papillomas, ca. 75 and 25% of the H-ras mutations were G f T and A f T transversions, respectively. Depurinating BP-DNA adducts provide a straightforward mechanism for the large proportion of G f T transversions in the p53 tumor suppressor gene of human lung tumors (29-32). Second, depurinated BP-adducted DNA bases enter the general circulation and subsequently accumulate in urine. Consequently, these adducted bases are readily available for analysis. Furthermore, the hepatic portal system may act as an effective “trap” for adducted bases of dietary origin. Depurinated BP-adducted bases formed in the gut epithelium or in the liver parenchyma may be excreted preferentially in bile, rather than in urine. In contrast, a greater proportion of depurinated adducted bases released from the large, epithelial surfaces of the lungs may be expected to be excreted in the urine, since they do not pass, first, through the liver. The strength of these considerations can be determined only by direct investigation. It must be noted in this context, however, that in our study depurinated BP-DNA adducts were not detected in the urine of any control subject. Third, the short life of depurinating adducts in the DNA and the rapid excretion of depurinated BP-adducted bases (34) may make these bases particularly useful as rapid-response indicators of the effectiveness of preventive interventions. Fourth, this study indicates that in adduct-positive individuals, depurinated BP-adducted bases exhibit a high degree of sensitivity to BP exposure. Concentrations of adducted bases in urine from individuals exposed to coal smoke reflected a 20-30-fold greater level of exposure to BP than from cigarette smoke. This conclusion is based on (1) quantification of depurinated BP-DNA adducts in urine from three coal smoke-exposed women and one cigarette smoker and (2) positive identification of adducted bases in the urine of two other smokers (Table 1). Concentrations of BP-adducted bases in the latter two samples were between 0.1 fmol/mg of creatinine (detection limit) and 0.5 fmol/mg of creatinine (limit for quantification). Previous studies (11, 14, 16) showed that quantities of stable PAH-DNA adducts in peripheral blood leukocytes (PBL) of adduct-positive, coal smoke-exposed individuals were ca. 2- and 3-fold greater than in PBL of adduct-positive, cigarette smokers and nonexposed control subjects, respectively. The results of the study presented here justify development of improved protocols for purification and quantification of these adducted bases in urine of PAH-exposed individuals. To this end, we are developing an immunoaffinity/HPLC protocol (35) and gold biosensor chips (36) with a monoclonal antibody that has high specificity and affinity for BP-6-N7Ade and BP-6-N7Gua. These protocols should be useful for higher through-put studies aimed at examining (1) the mean and range of BPadducted base concentrations in urine, for larger sample sizes and for additional PAH-exposed populations; (2) intraindividual tracking of BP-adducted base concentrations, over time; (3) dietary contributions to the urinary

Casale et al.

profile of BP-adducted bases; (4) the contributions of genetically determined polymorphisms of cytochrome P450 to urinary profiles of BP-adducted bases; and (5) association of urinary concentrations of BP-adducted bases with lung cancer risk, in prospective studies that stratify individuals on the basis of genetic polymorphisms of PAH-metabolizing enzymes (2), myeloperoxidase (43), DNA repair capacity (44), and the major histocompatibility complex (45, 46).

Acknowledgment. U.S. Public Health Service Grant 2 PO1 CA49210 from the National Cancer Institute supported this research. The National Centers for Research Resources of the NIH (Grant P41RR00954) and the Office of Health and Environmental Research, Office of Energy Research (U.S. Department of Energy), provided partial support. The research described here has been reviewed by the National Environmental and Health Effects Research Laboratory, U.S. Environmental Protection Agency, and approved for publication. Approval does not signify that the contents necessarily reflect the views and policies of the Agency, nor does mention of trade names or commercial products constitute endorsement or recommendation of their use.

References (1) Perera, F. P., and Weinstein, I. B. (1982) Molecular epidemiology and carcinogen-DNA adduct detection: new approaches to studies of human cancer causation. J. Chronic Dis. 35, 581-600. (2) Perera, F. P. (1996) Molecular epidemiology: Insights into cancer susceptibility, risk assessment, and prevention. J. Natl. Cancer Inst. 88, 496-509. (3) Higginbotham, S., RamaKrishna, N. V. S., Johansson, S. L., Rogan, E. G., and Cavalieri, E. C. (1993) Tumor-initiating activity and carcinogenicity of dibenzo[a,l]pyrene versus 7,12-dimethylbenz[a]anthracene and benzo[a]pyrene at low doses in mouse skin. Carcinogenesis 14, 875-878. (4) Mumford, J. L., Li, X., Hu, F., Lu, X. B., and Chuang, J. C. (1995) Human exposure and dosimetry of polycyclic aromatic hydrocarbons in urine from Xuan Wei, China with high lung cancer mortality associated with exposure to unvented coal smoke. Carcinogenesis 16, 3031-3036. (5) International Agency for Research on Cancer (1986) Tobacco smoking. IARC Monographs on the Evaluation of the Carcinogenic Risk of Chemicals to Humans, Vol. 38, pp 100-102, International Agency for Research on Cancer, Lyon, France. (6) Mumford, J. L., He, X. Z., Chapman, R. S., Cao, S. R., Harris, D. B., Li, X. M., Xian, Y. L., Jiang, W. Z., Xu, C. W., Chuang, J. C., Wilson, W. E., and Cooke, M. (1987) Lung cancer and indoor air pollution in Xuan Wei, China. Science 235, 217-220. (7) Doll, R., and Peto, R. (1981) The causes of cancer: Quantitative estimates of avoidable risks of cancer in the United States today. J. Natl. Cancer Inst. 66, 1191-1308. (8) Rogot, E., and Murray, J. L. (1980) Smoking and causes of death among U.S. veterans: 16 years of observation. Public Health Rep. 95, 213-222. (9) Perera, F., Santella, R., and Poirier, M. (1986) Biomonitoring of workers exposed to carcinogens: Immunoassays to benzo[a]pyrene-DNA adducts as a prototype. J. Occup. Med. 28, 11171123. (10) Conney, A. H. (1982) Induction of microsomal enzymes by foreign chemicals and carcinogenesis by polycyclic aromatic hydrocarbons: G.H.A. Clowes Memorial Lecture. Cancer Res. 42, 4875-4917. (11) Perera, F. P., Santella, R. M., Brenner, D., Poirier, M. C., Munshi, A. A., Fischman, H. K., and Van Ryzin, J. (1987) DNA adducts, protein adducts, and sister chromatid exchange in cigarette smokers and nonsmokers. J. Natl. Cancer Inst. 79, 449-456. (12) Phillips, D. H., Schoket, B., Hewer, A., Bailey, E., Kostic, S., and Vincze, I. (1990) Influence of cigarette smoking on the levels of DNA adducts in human bronchial epithelium and white blood cells. Int. J. Cancer 46, 569-575. (13) Assennato, G., Ferri, G. M., Tockman, M. S., Poirier, M. C., Schoket, B., Porro, A., Corrado, V., and Strickland, P. T. (1993) Biomarkers of carcinogen exposure and cancer risk in a coke plant. Environ. Health Perspect. 99, 237-239.

BP-Adducted DNA Bases in Urine (14) Mumford, J. L., Lee, X., Lewtas, J., Young, T. L., and Santella, R. M. (1993) DNA adducts as biomarkers for assessing exposure to polycyclic aromatic hydrocarbons in tissues from Xuan Wei women with high exposure to coal combustion emissions and high lung cancer mortality. Environ. Health Perspect. 99, 83-87. (15) Kang, D. H., Rothman, N., Poirier, M. C., Greenberg, A., Hsu, C. H., Schwartz, B. S., Baser, M. E., Groopman, J. D., Weston, A., and Strickland, P. T. (1995) Interindividual differences in the concentration of 1-hydroxypyrene-glucuronide in urine and polycyclic aromatic hydrocarbon-DNA adducts in peripheral white blood cells after charbroiled beef consumption. Carcinogenesis 16, 1079-1085. (16) Mooney, L. A., Santella, R. M., Covey, L., Jeffrey, A. M., Bigbee, W., Randall, M. C., Cooper, T. B., Ottman, R., Tsai, W., Wazneh, L., Glassman, H., Young, T., and Perera, F. P. (1995) Decline of DNA damage and other biomarkers in peripheral blood following smoking cessation. Cancer Epidemiol., Biomarkers Prev. 4, 627634. (17) Tang, D., Santella, R. M., Blackwood, A. M., Young, T., Mayer, J., Jaretzki, A., Grantham, S., Tsai, W., and Perera, F. P. (1995) A molecular epidemiological case-control study of lung cancer. Cancer Epidemiol., Biomarkers Prev. 4, 341-346. (18) Van Schooten, F. J., Jongeneelen, F. J., Hillebrand, M. J. X., van Leeuwen, F. E., de Looff, A. J. A., Dijkmans, A. P. G., van Rooij, J. G. M., den Engelse, L., and Kriek, E. (1995) Polycyclic aromatic hydrocarbon-DNA adducts in white blood cell DNA and 1-hydroxypyrene in the urine from aluminum workers: Relation with job category and synergistic effect of smoking. Cancer Epidemiol., Biomarkers Prev. 4, 69-77. (19) Dallinga, J. W., Pachen, D. M. F. A., Wijnhoven, W. P., Breedijk, A., van’t Veer, L., Wigbout, G., van Zandwijk, N., Maas, L. M., van Agen, E., Kleinjans, J. C. S., and van Schooten, F. (1998) The use of 4-aminobiphenyl hemoglobin adducts and aromatic DNA adducts in lymphocytes of smokers as biomarkers of exposure. Cancer Epidemiol., Biomarkers Prev. 7, 571-577. (20) Kriek, E., Rojas, M., Alexandrov, K., and Bartsch, H. (1998) Polycyclic aromatic hydrocarbon-DNA adducts in humans: Relevance as biomarkers for exposure and cancer risk. Mutat. Res. 400, 215-231. (21) Santella, R. M., Weston, A., Perera, F. P., Trivers, G. T., Harris, C. C., Young, T. L., Nguyen, D., Lee, B. M., and Poirier, M. C. (1988) Interlaboratory comparison of antisera and immunoassays for benzo[a]pyrene-diol-epoxide-I-modified DNA. Carcinogenesis 9, 1265-1269. (22) Phillips, D. H., Hewer, A., Martin, C. N., Garner, R. C., and King, M. M. (1988) Correlation of DNA adduct levels in human lung with cigarette smoking. Nature 336, 790-792. (23) Van Schooten, F. J., Hillebrand, M. J. X., van Leeuwen, F. E., van Zandwijk, N., Jansen, H. M., den Engelse, L., and Kriek, E. (1992) Polycyclic aromatic hydrocarbon-DNA adducts in white blood cells from lung cancer patients: no correlation with adduct levels in lung. Carcinogenesis 13, 987-993. (24) Van Schooten, F. J., Godschalk, R. W. L., Breedijk, A., Maas, L. M., Kriek, E., Sakai, H., Wigbout, G., Baas, P., van’t Veer, L., and van Zandwijk, N. (1997) 32P-Postlabelling of aromatic DNA adducts in white blood cells and alveolar macrophages of smokers: saturation at high exposures. Mutat. Res. 378, 65-75. (25) Alexandrov, K., Rojas, M., Geneste, O., Castegnaro, M., Camus, A., Petruzzelli, S., Giutini, C., and Bartsch, H. (1992) An improved fluorometric assay for dosimetry of benzo(a)pyrene diol-epoxideDNA adducts in smokers’ lung: Comparisons with total bulky adducts and aryl hydrocarbon hydrolase activity. Cancer Res. 52, 6248-6253. (26) Rojas, M., Alexandrov, K., van Schooten, F. J., Hillebrand, M., Kriek, E., and Bartsch, H. (1994) Validation of a new fluorometric assay for benzo[a]pyrene diolepoxide-DNA adducts in human white blood cells: comparisons with 32P-postlabeling and ELISA. Carcinogenesis 15, 557-560. (27) Rojas, A. M., Alexandrov, K., Auburtin, G., Wastiaux-Denamur, A., Mayer, L., Mahieu, B., Sebastien, P., and Bartsch, H. (1995) Anti-benzo[a]pyrene diolepoxide-DNA adduct levels in peripheral mononuclear cells from coke oven workers and the enhancing effect of smoking. Carcinogenesis 16, 1373-1376. (28) Cavalieri, E., and Rogan, E. (1998) Mechanisms of tumor initiation by polycyclic aromatic hydrocarbons in mammals. In The Handbook of Environmental Chemistry: PAHs and Related Compounds (Neilson, A. H., Ed.) Vol. 3J, pp 81-117, Springer, Heidelberg, Germany. (29) Chakravarti, D., Pelling, J. C., Cavalieri, E. L., and Rogan, E. G. (1995) Relating aromatic hydrocarbon-induced DNA adducts and c-H-ras mutations in mouse skin papillomas: The role of apurinic sites. Proc. Natl. Acad. Sci. U.S.A. 92, 10422-10426.

Chem. Res. Toxicol., Vol. 14, No. 2, 2001 201 (30) Chary, P., Latham, G. J., Robberson, D. L., Kim, S. J., Han, S., Harris, C. M., Harris, T. M., and Lloyd, R. S. (1995) In vivo and in vitro replication consequences of stereoisomeric benzo[a]pyrene7,8-dihydrodiol 9,10-epoxide adducts on adenine N6 at the second position of N-ras codon 61. J. Biol. Chem. 270, 4990-5000. (31) Penning, T. M., Burczynski, M. E., Hung, C., McCoull, K. D., Palackal, N. T., and Tsuruda, L. S. (1999) Dihydrodiol dehydrogenases and polycyclic aromatic hydrocarbon activation: Generation of reactive and redox active o-quinones. Chem. Res. Toxicol. 12, 1-18. (32) Hollstein, M., Sidransky, D., Vogelstein, B., and Harris, C. C. (1991) p53 mutations in human cancers. Science 253, 49-53. (33) Chen, L., Devanesan, P. D., Higginbotham, S., Ariese, F., Jankowiak, R., Small, G. J., Rogan, E. G., and Cavalieri, E. C. (1996) Expanded analysis of benzo[a]pyrene-DNA adducts formed in vitro and in mouse skin: Their significance in tumor initiation. Chem. Res. Toxicol. 9, 897-903. (34) Rogan, E. G., RamaKrishna, N. V. S., Higginbotham, S., Cavalieri, E. C., Jeong, H., Jankowiak, R., and Small, G. J. (1990) Identification and quantitation of 7-(benzo[a]pyren-6-yl)guanine in the urine and feces of rats treated with benzo[a]pyrene. Chem. Res. Toxicol. 3, 441-444. (35) Casale, G. P., Rogan, E. G., Stack, D., Devanesan, P., and Cavalieri, E. C. (1996) Production of a high-affinity monoclonal antibody specific for 7-(benzo[a]pyren-6-yl)guanine and its application in a competitive enzyme-linked immunosorbent assay. Chem. Res. Toxicol. 9, 1037-1043. (36) Duhachek, S., Kenseth, J., Casale, G. P., Small, G. J., Porter, M. D., and Jankowiac, R. (2000) Monoclonal antibody-gold biosensor chips for detection of depurinating carcinogen-DNA adducts by fluorescence line-narrowing spectroscopy. Anal. Chem. 72 (16), 3709-3716. (37) Hanson, A. A., Rogan, E. G., and Cavalieri, E. L. (1998) Synthesis of adducts formed by iodine oxidation of aromatic hydrocarbons in the presence of deoxyribonucleosides and nucleobases. Chem. Res. Toxicol. 11, 1201-1208. (38) Roberts, K. P., Cheng, H. L., Singhal, M., Casale, G. P., Small, G. J., and Jankowiak, R. (2000) On-Line identification of depurinating DNA adducts in human urine by capillary electrophoresis-fluorescence line-narrowing spectroscopy. Electrophoresis 21, 1-8. (39) Jankowiak, R., Zamzow, D., Ding, W., and Small, G. J. (1996) Capillary electrophoresis-fluorescence line-narrowing system for on-line structural characterization of molecular analytes. Anal. Chem. 68, 2549-2553. (40) Zamzow, D., Small, G. J., and Jankowiak, R. (1996) Capillary electrophoresis fluorescence line narrowing (CE-FLN) system for DNA adduct characterization. Mol. Cryst. Liq. Cryst. 291, 155162. (41) Rogan, E. G., Cavalieri, E. C., Tibbels, S. R., Cremonesi, P., Warner, D. L., Nagel, D. L., Tomer, R. L., Cerny, R. L., and Gross, M. L. (1988) Synthesis and identification of benzo[a]pyreneguanine nucleoside adducts formed by electrochemical oxidation and by horseradish peroxidase catalyzed reaction of benzo[a]pyrene with DNA. J. Am. Chem. Soc. 110, 4023-4029. (42) Devanesan, P. D., RamaKrishna, N. V. S., Todorovic, R., Rogan, E. G., Cavalieri, E. C., Jeong, H., Jankowiak, R., and Small, G. J. (1992) Identification and quantitation of benzo[a]pyrene-DNA adducts formed by rat liver microsomes in vitro. Chem. Res. Toxicol. 5, 302-309. (43) Petruska, J. M., Mosebrook, D. R., Jakab, G. J., and Trush, M. A. (1992) Myeloperoxidase-enhanced formation of (+-)-trans-7,8dihydroxy-7,8-dihydrobenzo[a]pyrene-DNA adducts in lung tissue in vitro: A role of pulmonary inflammation in the bioactivation of a procarcinogen. Carcinogenesis 13, 1075-1081. (44) Cheng, L., Spitz, M. R., Hong, W. K., and Wei, Q. (2000) Reduced expression levels of nucleotide excision repair genes in lung cancer: a case-control study. Carcinogenesis 21, 1527-1530. (45) Casale, G. P., Cheng, Z., Liu, J., Cavalieri, E. L., and Singhal, M. (2000) Profiles of cytokine mRNAs in the skin and lymph nodes of SENCAR mice treated epicutaneously with dibenzo[a,l]pyrene or dimethylbenz[a]anthracene reveal a direct correlation between carcinogen-induced contact hypersensitivity and epidermal hyperplasia. Mol. Carcinog. 27, 125-140. (46) Elmets, C. A., Athar, M., Tubesing, K. A., Rothaupt, D., Xu, H., and Mukhtar, H. (1998) Susceptibility to the biological effects of polyaromatic hydrocarbons is influenced by genes of the major histocompatibility complex. Proc. Natl. Acad. Sci. U.S.A. 95, 14915-14919.

TX000012Y