Quantification of Urinary Excretion of 1,N6-Ethenoadenine, a Potential

Joseph H. Banoub, Russell P. Newton, Eddy Esmans, David F. Ewing, and Grahame Mackenzie. Chemical Reviews 2005 105 (5), 1869-1916. Abstract | Full ...
0 downloads 0 Views 151KB Size
Chem. Res. Toxicol. 2004, 17, 963-971

963

Quantification of Urinary Excretion of a Potential Biomarker of Lipid Peroxidation, in Humans by Stable Isotope Dilution Liquid Chromatography-Electrospray Ionization-Tandem Mass Spectrometry: Comparison with Gas Chromatography-Mass Spectrometry 1,N6-Ethenoadenine,

Hauh-Jyun Candy Chen* and Chia-Ming Chang Department of Chemistry and Biochemistry, National Chung Cheng University, 160 San-Hsing, Ming-Hsiung, Chia-Yi 62142, Taiwan Received September 26, 2003

Etheno DNA adducts are promutagenic DNA lesions derived from exogenous as well as endogenous sources. The levels of etheno adducts in tissue DNA are elevated in cancer prone tissues, and the urinary excretion of etheno adducts is associated with oxidative stress. In this report, a new assay based on isotope dilution liquid chromatography-electrospray ionization-tandem mass spectrometry (LC-ESI-MS/MS) is developed for the quantification of 1,N6-ethenoadenine (Ade) in human urine samples without the need for derivatization. Sample purification before analysis by MS only requires a reversed phase solid phase extraction column. Two multiple reaction monitoring transitions with two product ion fragments generated from a common parent ion were used to quantify urinary Ade. The detection limit of Ade using LC-ESI-MS/MS is 2 pg injected standard Ade on-column, and the assay allows accurate quantification of urinary Ade at concentrations higher than 10 pg/mL. The presence of Ade in human urine is confirmed by the collision-induced daughter ion spectrum. Using this assay, the levels of Ade in the 24 h urine samples from 18 healthy individuals are determined, and the results are in very good agreement with those obtained using isotope dilution gas chromatography-negative ion chemical ionization-mass spectrometry. The high specificity and simple sample pretreatment of this LC-ESI-MS/MS method render it a valuable tool in measuring Ade in the complex mixture of human urine as a promising noninvasive biomarker for DNA damage associated with oxidative stress and for cancer chemoprevention studies.

Introduction The modification of DNA appears to play a critical role in the multistage carcinogenesis processes (1, 2). The formation of the promutagenic etheno DNA adducts, Ade,1 Cyt, N2,3-Gua, and 1,N2-Gua, has been shown to derive from exogenous exposure to industrial chemicals, such as vinyl chloride (3), as well as endogenously generated reactive species (4), including epoxides of degraded lipid hydroperoxides (5, 6). Background levels of etheno DNA adducts detected in unexposed rodents and humans might originate from endogenous lipid peroxidation (7, 8). Increased levels of etheno adducts were found in cancer prone tissues, and their levels increase with oxidative stress. The formation of etheno DNA adducts may contribute to genetic instability and cancer progression, and it is thus implicated in Wilson’s * To whom correspondence should be addressed. Tel: 886-5-2428176. Fax: 886-5-272-1040. E-mail: [email protected]. 1 Abbreviations: CID, collision-induced dissociation; , etheno; Ade, 1,N6-ethenoadenine; Cyt, 3,N4-ethenocytosine; 1,N2-Gua, 1,N2-ethenoguanine; N2,3-Gua, N2,3-ethenoguanine; ESI, electrospray ionization; LOD, limit of detection; LOQ, limit of quantification; MS, mass spectrometry; MS/MS, tandem mass spectrometry; NICI, negative ion chemical ionization; SIM, selective ion monitoring; S/N, signal-to-noise ratio; SPE, solid phase extraction; MRM, multiple reaction monitoring; %RSD, percent relative standard deviation.

disease, colon and breast cancers, and cancers related to chronic infections and inflammation (9-14). Therefore, etheno DNA adducts can be used as biomarkers for oxidative DNA damage and cancer chemoprevention studies (15). The most sensitive analysis for unequivocal chemical characterization and quantification is probably to use GC-NICI-MS with SIM (16, 17). Assays of etheno DNA adducts in tissue DNA based on GC-NICI-MS have been developed for Ade (18), Cyt (19), N2,3-Gua (20, 21), and 1,N2-Gua (22). These methods involve acid hydrolysis of tissue DNA to the nucleic acid bases, followed by derivatization with an electrophore before GC-NICI-MS analysis. Nature has evolved repair systems to reverse DNA adduction. The repair of these promutagenic etheno DNA adducts by base excision repair enzymes is known (23). The excised adducted bases are excreted into urine without further metabolism. Using HPLC with fluorescence detection, Ade was detected in high levels upon the exposure of rats with chloroethylene oxide, the metabolite of vinyl chloride, which led to a 50-fold increase of urinary Ade concentration (24). Although levels of Ade in pancreatic DNA or pulmonary DNA did not differ between smokers and nonsmokers (25, 26),

10.1021/tx0341963 CCC: $27.50 © 2004 American Chemical Society Published on Web 06/26/2004

964

Chem. Res. Toxicol., Vol. 17, No. 7, 2004

urinary excretion of 1,N6-etheno-2′-deoxyadenosine was found to correlate with salt excretion and intake of ω6polyunsaturated fatty acid in postmenopausal women (27). In addition, our recent study showed a statistically significant association between urinary excretion and cigarette smoking in males (28). Because of the fast development of instrumentation, LC coupled with MS has become a very useful technique in the measurement of DNA adducts, although the sensitivity is in general lower than GC-NICI-MS, in which the detection limits are in the range of the sub to hundreds of attomoles level (18-22, 28). However, because LC-MS is suitable for analyzing polar compounds, analysis by LC-MS has the advantage of being devoid of sample derivatization and postderivatization cleanup procedures. An endogenous level of Ade in the urine of untreated rats was determined by immunoaffinity extraction combined with LC-ESI single quadrupole MS under the SIM mode. The detection limit of this assay was 270 fmol (29). The 32P-postlabeling technique can only be used for analysis of adducts on DNA as nucleotides, but not nucleic acid bases or nucleosides, because the method is based on the labeling of the 3′-monophosphate with radioactive γ-ATP. The intrinsic specificity of the tandem mass spectrometric analysis increases the sensitivity of the method by lowering the background. The LC-ESI-MS/MS methods have been developed for the deoxyribonucleosides of etheno adducts (30-33), oxidized adducts, including 8-oxo-7,8-dihydro-2′-deoxyguanosine (34), and malondialdehyde-2′-deoxyguanosine adduct (35) in tissue DNA. Typically, tissue DNA is hydrolyzed enzymatically to the 2′-deoxynucleosides and tandem mass spectrometric analysis is performed using MRM of transition from the molecular ion [M + H]+ to the corresponding base ion [M - deoxyribose]+ for each adducted nucleoside. Despite its high sensitivity, this approach cannot be applied for analysis of adducted nucleobases. However, it is necessary to analyze urinary DNA adducts at the levels of both nucleobases and the 2′-deoxynucleosides because they represent different biological significances. In our previous report, an assay based on stable isotope dilution GC-NICI-MS was used to quantify Ade in human urine (28). To reduce the sample pretreatment processes and to avoid the need for derivatization, we now develop a stable isotope dilution LC-ESI-MS/MS method for accurate quantification of urinary Ade under the unique MRM modes. To our knowledge, this new method is the first report of analyzing an etheno DNA base in human biological fluid using stable isotope dilution LCESI-MS/MS.

Materials and Methods Chemicals. Standard Ade was purchased from Sigma Chemical Co. (St. Louis, MO). Bond Elut C18-OH SPE columns (500 mg, 3 mL) were from Varian (Harbor City, CA). The stable isotope [13C1,15N2]Ade was synthesized from the reaction of 2-chloroacetaldehyde with [13C1,15N2]adenine (Cambridge Isotope Laboratories, Andover, MA), followed by purification with a C18-OH SPE column and quantification by HPLC-UV as reported (28). Urine Pretreatment. Urine collected over a 24 h period was stored as 1.0 mL aliquots at the -80 °C freezer. To examine the possibility that the analyte was contained in the precipitate, two procedures were employed after defrosting the samples in an ice water bath. In procedure A, the urine sample (1.0 mL)

Chen and Chang Table 1. Optimal Parameters of EAde, [13C1,15N2]EAde, ECyt, and [13C4,15N3]ECyt in Tandem Mass Spectra compound

cone voltage (V)

collision energy (eV)

Ade [13C1,15N2]Ade

50 55

method A 25 30

Ade [13C1,15N2]Ade

50 55

method B 25 25

MRM m/z 160 f m/z 106 m/z 163 f m/z 108 m/z 160 f m/z 133 m/z 163 f m/z 136

was added to internal standard [13C1,15N2]Ade (1.0 ng) and centrifuged at 23000g for 10 min at 4 °C. Alternatively, the sample was added to [13C1,15N2]Ade (1.0 ng), diluted 1:1 (v/v) with lithium acetate (100 mM, pH 6.5), incubated at 37 °C for 10 min, and followed by centrifugation at 5000g for 10 min at room temperature (procedure B) (36, 37). The precipitate was discarded, and the leftover samples were not reused. The creatinine contents were analyzed by a picric acid method (38). Adduct Enrichment by C18-OH SPE Column. Before the use for samples, each new batch of SPE columns was tested for consistency in their elution pattern with 1.0 µg each of standard Ade. After elution with conditions described below, the fractions were collected every 3 mL. The eluant was dried and quantified based on the molar UV absorbance of Ade on the reversed phase HPLC equipped with a Hitachi L-7000 pump system with a D-7000 interface, a Rheodyne injector, a L-7450A photodiode array detector (Hitachi, Tokyo, Japan), and a Rheodyne injector connecting to a Prodigy ODS (3) 250 mm × 4.6 mm 5 µm column (Phenomenex, Torrance, CA). The chromatographic system was performed with a linear H2O and CH3CN gradient: 0-5 min, 0% CH3CN; 5-20 min, 0-50% CH3CN at a flow rate of 1.0 mL/ min. The stable isotope [13C1,15N2]Ade was quantified by HPLC using the molar UV absorbance of commercially available Ade standard. Sample Enrichment. The supernatant (1.0 mL) of pretreated urine sample was loaded on a C18-OH SPE column preconditioned with 15 mL of methanol, followed by 15 mL of H2O. After the sample was loaded and eluted, the column was washed with 12 mL of H2O, followed by 3 mL of 10% CH3OH in H2O and 3 mL of 15% CH3OH in H2O. The fraction containing Ade was eluted with 3 mL of 25% CH3OH solution in a 4 mL silanized glass vial. The fraction was evaporated under vacuum with a centrifuge concentrator and reconstituted in 25 µL of 0.01% acetic acid before LC-ESI-MS/MS analysis. LC-ESI-MS/MS Analysis of EAde. Twenty microliters of the reconstituted urine samples collected after the C18-OH SPE column were injected into a LC system consisting of a Hitachi L-7000 pump system with a D-7000 interface (Hitachi), a Rheodyne injector, and a reversed phase C18 column [Luna C18 (2), 2.0 mm × 150 mm, 3 µm, Phenomenex]. It was eluted at a flow rate of 0.2 mL/min with a linear gradient of 0.01% acetic acid (pH 3.8) to 30% CH3OH in 0.01% acetic acid from 0 to 15 min, followed by a gradual increase of the CH3OH concentration in 1 min and washing with 100% CH3OH for 5 min before conditioning with 0.01% acetic acid for 20 min. The column was connected to a three way valve to divert the LC effluent to waste before 10 min and after 13 min. The effluent between 10 and 13 min was subjected to analysis by a triple quadrupole mass spectrometer (Quattro Ultima, Micromass, Manchester, United Kingdom) equipped with an ESI interface. After the MS acquisition was completed, the valve was switched to waste during washing and conditioning before the next run. A voltage of 3.0 kV was applied to the electrospray needle. N2 was used as the desolvation gas (500 L/h) to help nebulization and as the nebulization gas (150 L/h) to help desolvation and to stabilize the spray. The source temperature was at 120 °C, and the stainless steel capillary was heated to 350 °C to obtain optimal desolvation. Argon was used as the collision gas in MS/MS experiments. The pressure of the collision cell was 2.5 × 10-3 mBar. In the MRM experiment (dwell time, 0.2 s), the precursor [M + H]+ ion was generated in the ESI source under the positive

LC-ESI-MS/MS Analysis of Ade in Human Urine Scheme 1. Quantification of EAde in Human Urine by Isotope Dilution LC-ESI-MS/MS

ion mode and focused in quadrupole 1 (Q1) and dissociated in a collision cell (quadrupole 2, Q2) yielding the product ion, which was analyzed in quadrupole 3 (Q3). Four precursor-product reactions were monitored in Q1 and Q3: m/z 160 f m/z 106 for Ade and m/z 163 f m/z 108 for [13C1,15N2]Ade in method A and m/z 160 f m/z 133 for Ade and m/z 163 f m/z 136 for [13C1,15N2]Ade in method B. The optimized cone voltage and collision energy for each MRM experiment are listed in Table 1. The low energy CID daughter ion spectra were obtained by scanning Q3 from 20 to 200 amu selecting the [M + H]+ ion at m/z 160 and 163 on Q1 with the same cone voltages and collision energies as those in the MRM experiments. Calibration Curve. The stock solutions of Ade (1.0 mg/mL) in water were stored at -80 °C. The sample solutions for calibration were freshly prepared by diluting the stock solutions in water for each analysis. Samples with various amounts of Ade ranging from 0, 10, 20, 50, 100, 200, and 500 pg each were added to [13C1,15N2]Ade (1.0 ng) as an internal standard. The samples were enriched by C18-OH SPE columns, evaporated, and reconstituted before LC-ESI-MS/MS analysis. The equations of the calibration curves were obtained by linear regression. GC-NICI-MS Analysis of EAde. The experiments were performed using the published procedures (28) with 0.1 mL of urine for each analysis. Statistical Analysis. The correlation coefficient was calculated using Spearman rank correlation by GraphPad InStat version 3.00 for Windows 95, GraphPad Software (San Diego, CA, www.graphpad.com).

Results and Discussion LC-ESI-MS/MS Assay for EAde. The goal of this study is to develop a quantitative and highly specific assay for analyzing Ade in human urine samples without the need for derivatization. Because of the complexity of the urine matrix, the assay was optimized to minimize contamination by potential interfering compounds. The assay involves adduct purification by a reversed phase C18-OH SPE column, followed by analysis using LC-ESI-MS/MS (Scheme 1). The nonend-capped C18-OH SPE column under the optimized elution conditions removes the early eluting polar metabolites as well as the nonpolar urinary components, which are retained on the column. Use of the disposable SPE column eliminates the laborious cleaning process between samples and the possible cross-contamination in studies, which collection by the HPLC column is used to purify DNA adducts from urine (39, 40). This assay does not use immunoaffinity chromatography for adduct enrichment

Chem. Res. Toxicol., Vol. 17, No. 7, 2004 965

because the results could be poorly reproducible due to batch-to-batch preparation of the antibody and the age of the antibody (29). To obtain accurate quantification of Ade in the complex media, a stable isotope standard [13C1,15N2]Ade 3 mass units higher than the analyte is used as an internal standard to monitor the recovery during sample processing. No unlabeled Ade is detected in the isotopomer. The advantage of using a stable isotope standard is that the analyte can be unambiguously identified in the chromatography of different mass transitions because the stable isotope standard is chemically identical to the analyte. In the chromatography, the analyte has virtually the same retention time with the analyte; the isotopomer eluted slightly (e0.1 min) earlier than the analyte. In addition, the subnanogram quantity of isotopomer serves as a carrier for the trace amount of analyte, which can be lost substantially during sample processing. To examine the possibility that the analyte is bound to urinary precipitate, as reported for 8-oxo-dGuo (36, 37), urinary Ade levels were analyzed with two different sample pretreatment procedures before C18-OH SPE enrichment. Urine samples were centrifuged at high speed at 4 °C, followed by the addition of isotope standard (procedure A). Alternatively, the isotope standard was diluted with buffer and incubated at elevated temperature to release the analyte from the precipitate, followed by centrifugation at a lower speed at room temperature (procedure B). No consistent increase was observed for adduct levels obtained using procedure A or B. In three samples analyzed, Ade concentrations from duplicated experiments were determined as 95, 80, and 167 pg/mL using procedure A and 89, 84, and 157 pg/mL using procedure B, respectively. The adduct levels are within the range of the precision, suggesting that no significant amounts of adduct are adsorbed on the precipitate, which might be due to the low concentration of Ade in the urine. Thus, procedure A was used throughout the study. After sample pretreatment and adduct enrichment, the separation of Ade from other components in the sample was achieved by a reversed phase HPLC column with a small diameter to limit the amount of solvent entering the electrospray source and to increase the sensitivity consequently. The LC elution condition was optimized to allow the maximum amount of organic solvent entering the spray and yet to resolve the analyte peaks from the interfering peaks. Under this condition, Ade eluted at approximately 11.5 min. Although the samples had been prepurified by the SPE column before injection into the HPLC column, only the elution between 10 and 13 min was acquired by the mass spectrometer. The early and later elutions were connected to the waste to minimize the amount of sample entering the MS and thus avoid ionization suppression and contamination of the cone and the source. MRM experiments were performed with a triple quadrupole system to achieve high specificity of the assay. Under the positive ion mode, the protonated precursor ion generated in the ESI source of Q1 is dissociated in a collision cell (Q2) and the resulting fragment (product) ion is analyzed in Q3. The greatly reduced background under the MRM leads to an increase in the sensitivity of the measurement. The CID mass spectrum of standard Ade shows the major fragment ions at m/z 106 and 133 from the [M + H]+ ion at m/z 160 (Figure 1a), while those for the

966

Chem. Res. Toxicol., Vol. 17, No. 7, 2004

Chen and Chang

Figure 1. CID mass spectra of (a) Ade and (b) [13C1,15N2]Ade. (c) Proposed fragmentation pathway for Ade. (d) Proposed fragmentation pathway for [13C1,15N2]Ade. Table 2. Precision of the LC-ESI-MS/MS Assay for EAde in Urine Method A Ade levels (pg/mL) (RSD, %) (n ) 6) samples

day 1

day 2

day 3

interday variation RSD (%)

1 2 3

43 ( 2 (4.9%) 75 ( 3 (7.7%) 165 ( 3 (3.3%)

37 ( 4 (8.3%) 77 ( 2 (4.2%) 163 ( 7 (7.5%)

50 ( 6 (12.0%) 79 ( 3 (7.8%) 158 ( 6 (7.0%)

7.5 2.8 2.3

Method B Ade levels (pg/mL) (RSD, %) (n ) 6) samples

day 1

day 2

day 3

interday variation RSD (%)

1 2 3

34 ( 3 (6.9%) 85 ( 3 (7.6%) 111 ( 4 (3.3%)

29 ( 3 (5.7%) 73 ( 3 (8.5%) 116 ( 8 (7.2%)

32 ( 2 (5.9%) 80 ( 2 (5.9%) 103 ( 8 (7.6%)

7.9 7.9 6.0

corresponding [13C1,15N2]Ade are m/z 108, 136, and 163, respectively (Figure 1b). The product ion of Ade at m/z 106 is assigned as losing two units of HCN groups (29), and it is in agreement with the corresponding m/z 108 fragment ion of [13C1,15N2]Ade labeling at N6, C8, and N9 (Figure 1c). On the other hand, the product ion of Ade at m/z 133 corresponds to the loss of one HCN

moiety from the pyrimidine ring, as evident by the m/z 136 ion in the daughter ion spectrum of [13C1,15N2]Ade (Figure 1d). If HCN is lost from the imidazole ring, the corresponding daughter ion for [13C1,15N2]Ade should be m/z 135. In this assay, two MRM transitions from the same parent ion are used. Method A monitors m/z 160 f 106 and m/z 163 f 108 for Ade and [13C1,15N2]Ade,

LC-ESI-MS/MS Analysis of Ade in Human Urine

Chem. Res. Toxicol., Vol. 17, No. 7, 2004 967

Figure 2. LC-ESI-MS/MS analysis of Ade in a smoker’s urine using (a) method A and (b) method B. The concentration of Ade in this sample was calculated as 136 pg/mL using method B. The chromatographic scheme of a nonsmoker’s urine using methods A and B is shown in panels c and d, respectively. The concentration of Ade in this sample was calculated as 78 and 82 pg/mL by methods A and B, respectively. Urine (1.0 mL) was added to [13C1,15N2]Ade, centrifuged, enriched with a C18-OH SPE column, and analyzed by LC-ESI-MS/MS under MRM mode as described in the Materials and Methods.

Figure 3. Daughter ion spectra of the Ade (upper trace) and [13C1,15N2]Ade (lower trace) peaks in a smoker’s urine. The spectra are obtained by selecting the peaks corresponding to [Ade + H]+ and [[13C1,15N2]Ade + H]+ ions at m/z 160 and 163 in the first quadrupole. After collision activation of the selected ions in the collision cell, the daughter ion spectra are recorded by scanning the third quadrupole.

respectively, whereas method B monitors m/z 160 f 133 and m/z 163 f 136 for Ade and [13C1,15N2]Ade, respectively. The reason for employing both methods is to confirm the adduct levels when interferences are observed in some of the samples using method A. The sensitivity of these two transitions is compared as described below. Detection Limit, Calibration, Precision, and Accuracy. The most commonly used MRM transition for

analysis of nucleosides is to monitor the loss of the deoxyribose moiety from the protonated molecular ion in the positive ESI mode. The sensitivity of this type of MRM transition is in the range from 40 amol to 200 fmol injected on-column (31-34). For instance, the detection of the 2′-deoxyribonucleosides of Ade in the enzymatic hydrolysates of DNA monitors the transition of [Ade2′-deoxyribose + H]+ ion to the nucleobase [Ade + H]+ ion (31). A high sensitivity was obtained for Ade-2′-

968

Chem. Res. Toxicol., Vol. 17, No. 7, 2004

Chen and Chang

Figure 4. Parent ion spectra of the Ade (upper trace) and [13C1,15N2]Ade (lower trace) peaks in a smoker’s urine. The spectra are obtained by selecting the peaks corresponding to [Ade + H]+ and [[13C1,15N2]Ade + H]+ ions at (a) m/z 106 and 108 or (b) m/z 133 and 136 in the third quadrupole. After collision activation of the selected ions in the collision cell, the parent ion spectra are recorded by scanning the first quadrupole.

deoxyribose (LOD of 10 fg or 63 amol) since the glycosidic linkages between bases and 2-deoxyribose are relatively labile. On the other hand, the etheno base is much more stable, and thus, a more harsh collision condition is required for fragmentation. As a consequence, the sensitivity of Ade analysis under MRM is much lower than that for the corresponding nucleoside. Weimann et al. recently reported that quantification of 8-oxo-guanine is four times less sensitive than its deoxynucleoside in human urine using isotope dilution LC-ESI-MS/MS (37). The cone voltages and collision energies for collision in the four MRM transitions (m/z 160 f 106, 163 f 108, 160 f 133, and 163 f 136) are optimized individually

for maximum sensitivity, and the results are summarized in Table 1. The fact that the four optimal collision conditions vary is likely due to the isotope labeling, which is evident by the different dissociation patterns in the daughter ion spectra between standard Ade and [13C1,15N2]Ade (Figure 1a,b). In method A, fragmentation of [13C1,15N2]Ade involves cleavage of a bond between C8 and N9, which are both heavy isotopes, and thus requires higher cone voltage and collision energy to achieve the maximum signal intensity. Estimated by a S/N of 3, the lower LODs injected on-column are 2 and 3 pg for Ade using transition methods A and B, respectively. The calibration curves constructed from the two MRM transi-

LC-ESI-MS/MS Analysis of Ade in Human Urine

Chem. Res. Toxicol., Vol. 17, No. 7, 2004 969

Figure 5. Correlation of urinary Ade levels obtained by LC-ESI-MS/MS vs GC-NICI-MS. Results obtained from LC-MS/MS experiments were average from triplicated measurement using MRM method B, while the GC-MS analyses were performed in duplicate for each sample.

tions of unlabeled Ade are both linear ranging from 10 to 500 pg in the presence of fixed amounts (1.0 ng) of [13C1,15N2]Ade (Figure 1S, Supporting Information). No unlabeled Ade could be detected in the control experiment containing only isotopomer standard; thus, intercepts of these two calibration curves were zero (Figure 2S, Supporting Information). The results are plotted as the ratio of the peak area of Ade to that of the corresponding isotopomer standard in relation to the amount of Ade added. Linear regression gave the leastsquares correlation coefficients (r2) of 0.9992 and 0.9994 for Ade using methods A and B, respectively. The difference in collision conditions and fragmentation patterns for each transition might explain why the slopes in these two calibration curves are different. Comparing the signal of [13C1,15N2]Ade in the urine samples to that obtained from 1.0 ng of standard [13C1,15N2]Ade, the recovery of the assay is estimated to be 85%, which could be due to loss in the C18-OH SPE column. The accuracy of the assay was confirmed by adding various amounts (100-400 pg) of standard Ade to a urine sample and analyzing the total concentrations (Figure 3S, Supporting Information). Extrapolation gave a y-intercept of 97 and 96 pg/mL, while the urine without added Ade was determined as 95 and 94 pg/mL using MRM methods A and B, respectively (Figure 3S, Supporting Information). The precision of the method was performed with three urine samples at the low, middle, and high concentrations. Each sample was analyzed with both MRM methods A and B in six replicates per day for three separate days. The results demonstrated that method A is to some extent more accurate and precise than method B. Both the intraday and interday %RSDs were lower using method A (Table 2). The intraday %RSD ranged from 3.3 to 12% and from 3.3 to 8.5%, while the interday %RSDs were between 2.3 and 7.5% and 6.0 and 7.9% using methods A and B, respectively. These accuracy and precision determinations are comparable to those obtained using GC-NICI-MS analysis with SIM (28). Analysis of EAde in Human Urine. The major difficulty in measuring DNA adducts in DNA hydrolysate is the enormous amount of normal bases or nucleosides.

On the other hand, the complex components of various metabolites interfere with the analysis of urinary DNA adducts. Thus, a highly specific assay with accurate quantification is demanded. The 18 urine samples analyzed in this study include 8 smokers and 10 nonsmokers. In the LC-MS/MS chromatography of most, but not all, smokers’ urine samples, there is a peak (retention time, 10.86 min) eluting before [13C1,15N2]Ade using MRM method A (Figure 2a). This peak affects accurate quantification of [13C1,15N2]Ade and consequently the Ade concentration. The interference peak is absent using MRM method B (Figure 2b), and the level of Ade corresponds to 136 pg/mL. The interference peak is not observed in samples from nonsmokers. Figure 2c,d shows the elution of Ade and [13C1,15N2]Ade under methods A and B, respectively, in the urine of a nonsmoker. The levels of Ade are 78 and 82 pg/mL by MRM transitions A and B, respectively. It is thus reasonable to estimate the LOQ of the assay being 20 pg/mL with a S/N of 10 in chromatograms of the urine samples. The LOQ of this LC-ESI-MS/MS assay is four times that of GC-NICI-MS (LOQ ) 5.0 pg/mL) (28). Because 1.0 mL of the urine sample is processed, the concentration detection limit is equivalent to 126 pM for urinary Ade, which is much lower than that for 8-oxoguanine in human urine (2 nM) (37). Although the signal intensity of Ade using method B is slightly lower than method A, no interference has been encountered in all of the samples. Nonetheless, the urinary levels of Ade obtained by these two MRM transitions are in excellent agreement, with a slope of 1.019 and an R2 of 0.9987 by linear regression analysis, for the 12 samples without the interference problem. The levels of Ade in these 18 urine samples range from 0 to 201 pg/mL. The Ade concentrations in smokers (112 ( 57 pg/mL) are significantly higher than nonsmokers (42 ( 40 pg/mL) (p ) 0.0266). To confirm the identity of Ade in urine, collisioninduced daughter ion spectra are obtained by selecting the peaks corresponding to [Ade + H]+ and [[13C1,15N2]Ade + H]+ ions at m/z 160 and 163 in the first quadrupole. After collision activation of the selected ions in the collision cell, the daughter ion spectra are recorded

970

Chem. Res. Toxicol., Vol. 17, No. 7, 2004

by scanning the third quadrupole. Figure 3 shows the fragment ions at m/z 106 and 133 produced from the peak of Ade (upper panel) and the corresponding fragment ions at m/z 108 and 136 from the peak of [13C1,15N2]Ade (lower panel) in the daughter ion spectra. Furthermore, parent ion spectra were obtained by selecting the peaks corresponding to [Ade + H]+ and [[13C1,15N2]Ade + H]+ ions at m/z 106 and 108 (Figure 4a) and at m/z 133 and 136 (Figure 4b) in the third quadrupole. After collision activation of the selected ions in the collision cell, the parent ion spectra are recorded by scanning the first quadrupole. These results provide qualitative evidence for the presence of Ade and the added [13C1,15N2]Ade in these urine samples. Correlation of Urinary EAde Levels Obtained by GC-NICI-MS vs LC-ESI-MS/MS. Urinary Ade levels obtained by this LC-ESI-MS/MS assay are compared to those measured by GC-NICI-MS under the SIM mode using the same isotopomer standard. Statistical analysis of these results by linear regression showed that they are in very good agreement. The y-intercept of 3.7 is very small, and the slope is 0.96 with an R2 value of 0.89 (Figure 5). Because the sensitivity of this LC-ESI-MS/ MS assay is lower than that of GC-NICI-MS, four out of the 18 samples cannot be accurately quantified using LCESI-MS/MS. Levels of these samples are in the range between 0 and 25 pg/mL as determined by GC-NICI-MS. Although this LC-ESI-MS/MS assay is not as sensitive as the GC-NICI-MS method, partly due to the intrinsically lower resolution of LC as compared to GC, it is convenient and highly specific. This LC-ESI-MS/MS assay greatly reduces the number of steps in the GCNICI-MS analysis, which requires C18-OH SPE enrichment, derivatization, and postderivatization cleanup by a Si SPE column. A substantial decrease in time and cost of the reagents as well as the Si SPE columns required for the GC-NICI-MS analysis makes the LC-ESI-MS/MS assay an attractive candidate for measuring urinary Ade in a large number of samples. Urinary Ade from individuals occupationally not exposed to industrial chemicals represents a form of oxidative DNA damage. A statistically significant correlation was previously found between urinary Ade excretion and cigarette smoking, a form of oxidative stress (28). Without the need for the time-consuming and costly derivatization steps, the measurement of urinary Ade by this new assay should provide a promising noninvasive biomarker for cancer risk assessment and in chemoprevention studies (15).

Chen and Chang

(3)

(4)

(5)

(6)

(7)

(8)

(9)

(10)

(11)

(12)

(13)

(14)

(15)

(16)

(17) (18)

(19)

Acknowledgment. This work was supported by grants from the National Science Council of Taiwan (H.J.C.C.). Supporting Information Available: Calibration curves for the LC-ESI-MS/MS analysis of Ade, LC-ESI-MS/MS analysis of a control sample containing [13C1,15N2]Ade using methods A and B, and analysis of various amounts of standard Ade to a urine sample containing [13C1,15N2]Ade. This material is available free of charge via the Internet at http://pubs.acs.org.

(20)

(21)

(22)

References (1) Lawley, P. D. (1989) Mutagens as carcinogens: Development of current concepts. Mutat. Res. 213, 3-25. (2) Poirier, M. C., and Beland, F. A. (1992) DNA adduct measurements and tumor incidence during chronic carcinogen exposure

(23)

in animal models: Implications for DNA adduct-based human cancer risk assessment. Chem. Res. Toxicol. 5, 749-755. Laib, R. J. (1986) The role of cyclic base adducts in vinyl-chlorideinduced carcinogenesis: studies on nucleic acid alkylation in vivo. In The Role of Cyclic Nucleic Acid Adducts in Carcinogenesis and Mutagenesis (Singer, B., and Bartsch, H., Eds.) pp 101-108, IARC Sci. Publ., 70, Oxford University Press, London. Nair, J., Barbin, A., Velic, I., and Bartsch, H. (1999) Etheno DNAbase adducts from endogenous reactive species. Mutat. Res. 424, 59-69. Chen, H.-J. C., and Chung, F.-L. (1996) Epoxidation of trans-4hydroxy-2-nonenal by fatty acid hydroperoxides and hydrogen peroxide. Chem. Res. Toxicol. 9, 306-312. Lee, S. H., Oe, T., and Blair, I. A. (2001) Vitamin C-induced decomposition of lipid hydroperoxides to endogenous genotoxins. Science 292, 2083-2086. Nair, J., Barbin, A., Guichard, Y., and Bartsch, H. (1995) 1,N6Ethenodeoxyadenosine and 3,N4-ethenodeoxycytidine in liver DNA from humans and untreated rodents detected by immunoaffinity/32P-postlabeling. Carcinogenesis 16, 613-617. Chung, F.-L., Chen, H.-J. C., and Nath, R. G. (1996) Lipid peroxidation as a potential endogenous source for the formation of exocyclic DNA adducts: a commentary. Carcinogenesis 17, 2105-2111. Nair, J., Sone, H., Nagao, M., Barbin, A., and Bartsch, H. (1996) Copper-dependent formation of miscoding etheno-DNA adducts in the liver of Long Evans Cinnamon (LEC) rats developing hereditary hepatitis and hepatocellular carcinoma. Cancer Res. 56, 1267-1271. Nair, J., Carmichael, P. L., Fernando, R. C., Phillips, D. H., Strain, A. J., and Bartsch, H. (1998) Lipid peroxidation-induced ethenoDNA adducts in the liver of patients with the genetic metal storage disorders Wilson’s disease and primary hemochromatosis. Cancer Epidemiol. Biomarkers Prev. 7, 435-440. Nair, J., Vaca, C. E., Velic, I., Mutanen, M., Valsta, L. M., and Bartsch, H. (1997) High dietary omega-6 polyunsaturated fatty acids drastically increase the formation of etheno-DNA base adducts in white blood cells of female subjects. Cancer Epidemiol. Biomarkers Prev. 6, 597-601. Schmid, K., Nair, J., Winde, G., Velic, I., and Bartsch, H. (2000) Increased levels of promutagenic etheno-DNA adducts in colonic polyps of FAP patients. Int. J. Cancer 87, 1-4. Nair, J., Gal, A., Tamir, S., Tannenbaum, S. R., Wogan, G. N., and Bartsch, H. (1998) Etheno adducts in spleen DNA of SJL mice stimulated to overproduce nitric oxide. Carcinogenesis 19, 20812084. Ohshima, H., and Bartsch, H. (1994) Chronic infections and inflammatory processes as cancer risk factors: Possible role of nitric oxide in carcinogenesis. Mutat. Res. 305, 253-264. Bartsch, H., and Nair, J. (2000) New DNA-based biomarkers for oxidative stress and cancer chemoprevention studies. Eur. J. Cancer 36, 1229-1234. Dizdaroglu, M., and Gajewski, E. (1990) Selected-ion mass spectrometry: Assays of oxidative DNA damage. Methods Enzymol. 186, 530-544. Giese, R. W. (1997) Detection of DNA adducts by electron capture mass spectrometry. Chem. Res. Toxicol. 10, 255-270. Chen, H.-J. C., Zhang, L., Cox, J., Cunningham, J. A., and Chung, F.-L. (1998) DNA adducts of 2,3-epoxy-4-hydroxynonanal: Detection of 7-(1′,2-dihydroxy-heptyl)-3H-imidazo[2,1,i]-purine and 1,N6-ethenoadenine by gas chromatography/negative ion chemical ionization/mass spectrometry. Chem. Res. Toxicol. 11, 1474-1480. Chen, H.-J. C., Lin, T.-C., Hong, C.-L., and Chiang, L.-C. (2001) Analysis of 3,N4-ethenocytosine in DNA and in human urine by isotope dilution gas chromatography/negative ion chemical ionization/mass spectrometry. Chem. Res. Toxicol. 14, 1612-1619. Fedtke, N., Boucheron, J. A., Walker, V. E., and Swenberg, J. A. (1990) Vinyl chloride-induced DNA adducts. II: Formation and persistence of 7-(2′-oxoethyl)guanine and N2,3-ethenoguanine in rat tissue DNA. Carcinogenesis 11, 1287-1292. Ham, A. J., Ranasinghe, A., Morinello, E. J., Nakamura, J., Upton, P. B., Johnson, F., and Swenberg, J. A. (1999) Immunoaffinity/ gas chromatography/high-resolution mass spectrometry method for the detection of N2,3-ethenoguanine. Chem. Res. Toxicol. 12, 1240-1246. Morinello, E. J., Ham, A.-J. L., Ranasinghe, A., Sangaiah, R., and Swenberg, J. A. (2001) Simultaneous quantitation of N2,3ethenoguanine and 1,N2-ethenoguanine with an immunoaffinity/ gas chromatography/high-resolution mass spectrometry assay. Chem. Res. Toxicol. 14, 327-334. Singer, B., and Hang, B. (1997) What structural features determine repair enzyme specificity and mechanism in chemically modified DNA? Chem. Res. Toxicol. 10, 713-732.

LC-ESI-MS/MS Analysis of Ade in Human Urine (24) Holt, S., Yen, T. Y., Sangaiah, R., and Swenberg, J. A. (1998) Detection of 1,N6-ethenoadenine in rat urine after chloroethylene oxide exposure. Carcinogenesis 19, 1763-1769. (25) Kadlubar, F. F., Anderson, K. E., Haussermann, S., Lang, N. P., Barone, G. W., Thompson, P. A., MacLeod, S. L., Chou, M. W., Mikhailova, M., Plastaras, J., Marnett, L. J., Nair, J., Velic, I., and Bartsch, H. (1998) Comparison of DNA adduct levels associated with oxidative stress in human pancreas. Mutat. Res. 405, 125-133. (26) Godschalk, R., Nair, J., van Schooten, F. J., Risch, A., Drings, P., Kayser, K., Dienemann, H., and Bartsch, H. (2002) Comparison of multiple DNA adduct types in tumor adjacent human lung tissue: effect of cigarette smoking. Carcinogenesis 23, 2081-2086. (27) Hanaoka, T., Nair, J., Takahashi, Y., Sasaki, S., Bartsch, H., and Tsugane, S. (2002) Urinary level of 1,N6-ethenodeoxyadenosine, a marker of oxidative stress, is associated with salt excretion and omega 6-polyunsaturated fatty acid intake in postmenopausal Japanese women. Int. J. Cancer 100, 71-75. (28) Chen, H.-J. C., and Chiu, W.-L. (2003) Detection and quantification of 1,N6-ethenoadenine in human urine by stable isotope dilution capillary gas chromatography/negative ion chemical ionization/mass spectrometry. Chem. Res. Toxicol. 16, 1099-1106. (29) Yen, T.-Y., Holt, S., Sangaiah, R., Gold, A., and Swenberg, J. A. (1998) Quantitation of 1,N6-ethenoadenine in rat urine by immunoaffinity extraction combined with liquid chromatography/ electrospray ionization mass spectrometry. Chem. Res. Toxicol. 11, 810-815. (30) Chen, H.-J. C., Chiang, L.-C., Tseng, M.-C., Zhang, L. L., Ni, J., and Chung, F.-L. (1999) Detection and quantification of 1,N6ethenoadenine in human placental DNA by mass spectrometry. Chem. Res. Toxicol. 12, 1119-1126. (31) Doerge, D. R., Churchwell, M. I., Fang, J.-L., and Beland, F. A. (2000) Quantification of etheno-DNA adducts using liquid chromatography, on-line sample processing, and electrospray tandem mass spectrometry. Chem. Res. Toxicol. 13, 1259-1264. (32) Churchwell, M. I., Beland, F. A., and Doerge, D. R. (2002) Quantification of ultiple DNA adducts formed through oxidative stress using liquid chromatography and electrospray tandem mass spectrometry. Chem. Res. Toxicol. 15, 1295-1301.

Chem. Res. Toxicol., Vol. 17, No. 7, 2004 971 (33) Loureiro, A. P. M., Marques, S. A., Garcia, C. C. M., Di Mascio, P., and Medeiros, M. H. G. (2002) Development of an on-line liquid chromatography-electrospray tandem mass spectrometry assay to quantitatively determine 1,N2-etheno-2′-deoxyguanosine in DNA. Chem. Res. Toxicol. 15, 1302-1308. (34) Frelon, S., Douki, T., Ravanat, J.-L., Pouget, J.-P., Tornabene, C., and Cadet, J. (2000) High-performance liquid chromatographytandem mass spectrometry measurement of radiation-induced base damage to isolated and cellular DNA. Chem. Res. Toxicol. 13, 1002-1010. (35) Rouzer, C. A., Chaudhary, A. K., Nokubo, M., Ferguson, D. M., Reddy, G. R., Blair, I. A., and Marnett, L. J. (1997) Analysis of the malondialdehyde-2′-deoxyguanosine adduct pyrimidopurinone in human leukocyte DNA by gas chromatography/electron capture/ negative chemical ionization/mass spectrometry. Chem. Res. Toxicol. 10, 181-188. (36) Bogdanov, M. B., Beal, M. F., McCabe, D. R., Griffin, R. M., and Matson, W. R. (1999) A carbon column-based liquid chromatography electrochemical approach to routine 8-hydroxy-2′-deoxyguanosine measurements in urine and other biologic matrices: a one-year evaluation of methods. Free Radical Biol. Med. 27, 647-666. (37) Weimann, A., Belling, D., and Poulsen, H. E. (2002) Quantification of 8-oxo-guanine and guanine as the nucleobase, nucleoside and deoxynucleoside forms in human urine by high-performance liquid chromatography-electrospray tandem mass spectrometry Nucleic Acids Res. 30, 7e. (38) Heinegard, D., and Tiderstrom, G. (1973) Determination of serum creatinine by a direct colorimetric method. Clin. Chim. Acta 43, 305-310. (39) Bianchini, F., Donato, F., Faure, H., Ravanat, J. L., Hall, J., and Cadet, J. (1998) Urinary excretion of 5-(hydroxymethyl) uracil in healthy volunteers: Effect of active and passive tobacco smoke. Int. J. Cancer 77, 40-46. (40) Ravanat, J. L., Guicherd, P., Tuce, Z., and Cadet, J. (1999) Simultaneous determination of five oxidative DNA lesions in human urine. Chem. Res. Toxicol. 12, 802-808.

TX0341963