Chem. Res. Toxicol. 2005, 18, 1593-1599
1593
Association between Cigarette Smoking and Urinary Excretion of 1,N2-Ethenoguanine Measured by Isotope Dilution Liquid Chromatography-Electrospray Ionization/Tandem Mass Spectrometry Hauh-Jyun Candy Chen* and Wei-Loong Chiu Department of Chemistry and Biochemistry, National Chung Cheng University, 168 University Road, Ming-Hsiung, Chia-Yi 62142, Taiwan Received June 3, 2005
Levels of the promutagenic 1,N2-ethenoguanine (1,N2-Gua), an etheno DNA adduct derived mainly from lipid peroxidation, in experimental animals are associated with risk of cancer formation. Since 1,N2-Gua can be repaired by human glycosylases, it is possible to use it as a biomarker for cancer risk assessment in humans. In the present study, a highly sensitive and specific stable isotope dilution liquid chromatography-electrospray ionization/tandem mass spectrometry (LC-ESI/MS/MS) was developed for accurate quantification of 1,N2-Gua in human urine. The sample pretreatment involved a consecutive strong cation exchange solid-phase extraction (SPE) and reversed phase SPE chromatography. The pretreated sample was analyzed by LC-ESI/MS/MS under multiple reaction monitoring mode (MRM) using a triple quadrupole mass spectrometer. The detection limit of 1,N2-Gua using this LC-ESI/MS/MS assay was 1.0 pg (5.8 fmol) injected on-column. Levels of urine samples collected from healthy volunteers were found to range from 0 to 199 pg/mL, and levels as low as 5.0 pg/mL (29 pM) could be accurately quantified. After adjusting for creatinine levels and body weight, an statistically significant association was observed between urinary levels of 1,N2-Gua and cigarette smoking (p ) 0.0006). This highly specific and sensitive assay should be valuable in measuring urinary 1,N2-Gua as a potential noninvasive biomarker for oxidative DNA damage.
Introduction DNA modification plays a critical role in the multistage carcinogenesis processes. Previous studies showed that epoxidation of the enal products of lipid peroxidation, such as 4-hydroxy-2-nonenal, could contribute to the endogenous formation of etheno DNA adducts in untreated rodents and humans (1-4). Etheno adducts are also formed in nitric oxide-induced lipid peroxidation and are thus associated with DNA damages by chronic infections and inflammation. Levels of etheno adducts appear to increase with oxidative stress and are implicated in cancer development (3-7). Among etheno adducts, 1,N2-ethenoguanine (1,N2-Gua)1 is a major product formed in DNA exposed to fatty acids or lipid peroxidation-derived aldehydes under peroxidizing conditions in vitro, whereas its isomer N2,3-ethenoguanine (N2,3-Gua) is originated predominately from industrial chemicals such as vinyl chloride (8). 1,N2-Gua plays a minor role relative to N2,3-Gua in vinyl chloride-induced carcinogenesis, but 1,N2-Gua may be formed to a larger * Address reprint request to Hauh-Jyun Candy Chen at the Department of Chemistry and Biochemistry, National Chung Cheng University, 168 University Road, Chia-Yi 62142, Taiwan. Phone: (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; S/N, signal-to-noise ratio; SPE, solid-phase extraction; MRM, multiple reaction monitoring; %RSD, percent relative standard deviation.
extent from endogenous oxidative processes. Moreover, 1,N2-ethenoguanine was not detected in untreated calf thymus DNA nor in hepatocyte DNA from rats exposed to vinyl chloride (8). Evidence showed that 1,N2-Gua is mutagenic in mammalian cells, inducing mainly G f A transition mutation (9). Nonetheless, it can be repaired by human alkylpurine-DNA-N-glycosylase (10). It is therefore not surprising to detect this genotoxic lesion in the biological fluid of human. Indeed, Gonzalez-Reche et al. reported detection of 1,N2-Gua in human urine, together with its isomer N2,3-Gua by both liquid chromatography-electrospray ionization/tandem mass spectrometry (LC-ESI/ MS/MS) and gas chromatography/mass spectrometry (11). However, the sum of urinary 1,N2-Gua in the forms of base, nucleoside, and nucleotide was measured after acid hydrolysis of urine samples. Furthermore, no internal standards were used in this study, and no information on levels of the two smokers compared to the nonsmoking study populations was provided. In this present study, a highly sensitive and specific stable isotope dilution LC-ESI/MS/MS assay was developed for accurate quantification of 1,N2-Gua in human urine as a result of the base excision repair pathway. Cigarette smoking is accompanied by increasing oxidative stress. We previously reported association between urinary levels of 1,N6-ethenoadenine (Ade), 3,N4-ethenocytosine (Cyt), and its nucleoside with cigarette smoking to different degrees (12-14). In this study, the results from analysis of 30 human urine samples suggest a
10.1021/tx050145p CCC: $30.25 © 2005 American Chemical Society Published on Web 08/20/2005
1594
Chem. Res. Toxicol., Vol. 18, No. 10, 2005
statistically significant correlation between levels of 1,N2Gua and cigarette smoking.
Materials and Methods Materials. Standard 2′-deoxyguanosine was purchased from Sigma Chemical Co. (St. Louis, MO). Bond Elut strong cation exchange SCX and reversed phase C18-OH solid-phase extraction (SPE) columns (500 mg, 3 mL) were from Varian (Harbor City, CA). The standard 1,N2-Gua and its stable isotope [13C1,15N2]1,N2-Gua was synthesized as reported (15) from the reaction of 2-chloroacetaldehyde with 2′-deoxyguanosine and [13C1,15N2]2′-deoxyguanosine (Cambridge Isotope Laboratories, Andover, MA), respectively, followed by mild acid hydrolysis, purification with a C18-OH SPE column, and quantification by HPLC-UV. Subjects. The subjects of this study were healthy volunteers recruited from graduate and undergraduate students, faculty members, and firemen from Chia-Yi city and county. All the subjects provided information on gender, age, body weight, height, number of cigarettes smoked per day, and years of smoking. They received a written warrantee stating that the information was for research purposes only and would be kept confidential. Urine Pretreatment. Urine collected over a 24-h period was stored as 1.0 mL aliquots at the -80 °C freezer. The urine sample (1.0 mL) was added to the internal standard [13C1,15N2] 1,N2-Gua (1.0 ng) and centrifuged at 23 000g for 20 min at 4 °C. The precipitate was discarded, and the left-over samples were not reused. The creatinine contents were analyzed by a picric acid method (16). Adduct Enrichment by SCX/C18-OH SPE Columns. Before the use for samples, each new batch of SPE columns was tested for consistency in the elution pattern with 1.0 µg each of standard 1,N2-Gua. 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 1,N2Gua 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 MeOH gradient: 0-30 min, 0-100% MeOH at a flow rate of 1.0 mL/min. The stable isotope [13C1,15N2]1,N2-Gua was quantified by HPLC using the molar UV absorbance of 1,N2Gua. The supernatant (1.0 mL) of pretreated urine sample was loaded on a SCX SPE column preconditioned with 15 mL of methanol, followed by 24 mL of 0.01% acetic acid (pH 3.8). After the supernatant was loaded and the volume eluted, the column was washed with 3 mL of 0.01% acetic acid, followed by 3 mL of methanol, 3 mL of 50 mM ammonium formate (pH 5), and 3 mL of 80 mM ammonium formate (pH 7). The fraction containing 1,N2-Gua was eluted with 3 mL of 100 mM ammonium formate (pH 7) and evaporated under vacuum with a centrifuge concentrator. One milliliter of water was added and evaporated again to assist in complete removal of ammonium formate. The residue was reconstituted in 1.0 mL of water before loading onto the C18-OH SPE column. The C18-OH SPE column was preconditioned with 15 mL of methanol, followed by 24 mL of H2O. After the sample was loaded and eluted, the column was washed with 12 mL of H2O. The fraction containing 1,N2-Gua was eluted with 3 mL of 15% 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 1,N2-Gua. Twenty microliters of the enriched urine samples collected after the SPE columns described above were injected into a LC system consisting of a Hitachi L-7000 pump system with a D-7000 interface (Hitachi, Tokyo, Japan), a Rheodyne injector, and a reversed phase C18
Chen and Chiu column [Luna C18 (2), 2.0 mm × 150 mm, 5 µm, Phenomenex, Torrance, CA]. 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 25% acetonitrile in 0.01% acetic acid from 0 to 20 min, followed by a gradual increase of the acetonitrile concentration in 5 min to 100% acetonitrile and washing with 100% acetonitrile 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 8 min and after 20 min. The effluent between 8 and 20 min was subjected to analysis by a triple quadrupole mass spectrometer (Quattro Ultima, Micromass, Manchester, U.K.) equipped with an electrospray ionization 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 (550 L/h) to help nebulization and as the nebulization gas (160 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 2.35 × 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 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). The MRM method A monitored Q1 and Q3: m/z 176 f m/z 121 for 1,N2-Gua with cone voltage and collision energy of 55 V and 25 eV, respectively, and m/z 179 f m/z 122 for [13C1,15N2]1,N2-Gua with cone voltage and collision energy of 35 V and 25 eV, respectively. Method B monitored Q1 and Q3: m/z 176 f m/z 148 for 1,N2Gua with cone voltage and collision energy of 55 V and 15 eV, respectively, and m/z 179 f m/z 151 for [13C1,15N2]1,N2-Gua with cone voltage and collision energy of 35 V and 15 eV, respectively. The low-energy collision-induced dissociation (CID) daughter ion spectra were obtained by scanning Q3 from 20 to 200 amu selecting the [M + H]+ ion at m/z 176 and 179 on Q1 with the same cone voltages and collision energies as those in the MRM experiments. Calibration Curve. The stock solutions of 1,N2-Gua (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 1,N2-Gua ranging from 0, 5, 10, 20, 50, 100, 200, to 300 pg each were added to [13C1,15N2]1,N2-Gua (1.0 ng) as an internal standard. The samples were enriched by SCX SPE followed by C18-OH SPE columns, evaporated, and reconstituted as described above before LC-ESI/MS/MS analysis. The equations of the calibration curves were obtained by linear regression. Statistical Analysis. GraphPad InStat version 3.00 for Windows 95, GraphPad Software (San Diego, CA, www.graphpad.com) was used for statistical analysis. The parametric Welch’s t-test was used to compare data from the smokers’ and nonsmokers’ groups since they follow Gaussian distribution as tested using the Kolmogorov and Smirnov method, and the data were not assumed to be with the same standard deviation. The correlation between two variables was calculated using Pearson linear correlation.
Results and Discussion Measurement of urinary 1,N2-Gua involves addition of the stable isotope [13C1,15N2]1,N2-Gua as an internal standard, adduct enrichment by two consecutive solidphase extraction (SPE) columns, and liquid chromatography-electrospray ionization/tandem mass spectrometry (LC-ESI/MS/MS) analysis as described in Scheme 1. Adduct Enrichment by SPE Columns. As human urine samples contain complex metabolites, the adduct purification and enrichment procedures are optimized to
LC-ESI/MS/MS Analysis of 1,N2-Gua in Human Urine
Chem. Res. Toxicol., Vol. 18, No. 10, 2005 1595
Scheme 1. Quantification of 1,N2-EGua in Human Urine by Isotope Dilution LC-ESI/MS/MS
Figure 1. Collision-induced dissociation mass spectra of 1,N2Gua (upper panel) and [13C1,15N2]1,N2-Gua (lower panel).
ensure a clean chromatography in the subsequent analysis by mass spectrometry. A strong cation exchange (SCX) SPE column followed by a nonend-capped reversed phase C18-OH SPE column was found to fulfill this requirement. Either a SCX or C18-OH SPE column alone did not give satisfactory LC-ESI/MS/MS results. The recovery of this procedure in samples containing 1.0 ng of [13C1,15N2]1,N2-Gua with various amounts of 1,N2-Gua was 87% in average. One of the advantages of using subnanogram quantity of isotope internal standard is to serve as a carrier for trace amounts of the analyte. When the disposable SPE columns are used, we avoid the laborious collection of adducts from urine by the HPLC column and the possible cross-contamination between samples (17, 18). Although immunoaffinity chromatography is very specific for the adduct, it is very costly and thus might not be practical for use in analyzing large quantities of samples required in epidemiological studies. LC-ESI/MS/MS Assay for Urinary 1,N2-Gua. Under the positive ion mode of the ESI/MS, the collisioninduced dissociation (CID) mass spectrum of standard 1,N2-Gua showed the major fragment ions from the protonated precursor ion ([M + H]+) at m/z 176 were m/z 148 and 121, while those for the corresponding [13C1,15N2]1,N2-Gua at m/z 179 were m/z 151 and 122, respectively (Figure 1). The product ion of 1,N2-Gua at m/z 121 from [M + H]+ at m/z 176 (19) is assigned as losing CO from the pyrimidine ring and HCN from the imidazole ring, and this postulation is in agreement with the corresponding m/z 122 fragment ion of [13C1,15N2]1,N2Gua labeling at N7, C8, and N9 (Figure 2a). On the other hand, the product ion of 1,N2-Gua at m/z 148 corresponds to the loss of one CO from the pyrimidine moiety (20), which is in accord with the m/z 151 ion in the daughter ion spectrum of [13C1,15N2]1,N2-Gua (Figure 2b). The multiple reaction monitoring (MRM) experiments were performed by a triple quadrupole mass spectrometer with high specificity and sensitivity. Under the positive ion mode, the protonated precursor ion generated in the ESI source of quadrupole 1 (Q1) is dissociated in a collision cell (Q2) and the resulting fragment (product) ion is analyzed in quadrupole 3 (Q3). An enormous
Figure 2. Proposed fragmentation pathway for (a) 1,N2-Gua and (b) [13C1,15N2]1,N2-Gua.
enhancement in sensitivity using MRM experiments is due to the greatly reduced background. The intensities of two MRM transitions with individually optimized cone voltage and collision energy for the same parent ion were compared. Method A monitored m/z 176 f 121 and m/z 179 f 122 for 1,N2-Gua and [13C1,15N2]1,N2-Gua, respectively. Method B monitored m/z 176 f 148 and m/z 179 f 151 for 1,N2-Gua and [13C1,15N2]1,N2-Gua, respectively. Comparing the peak areas obtained from various amounts of 1,N2-Gua using both MRM transitions demonstrated that method A was about 5 times as sensitive as method B consistently (data not shown). Separation of 1,N2-Gua from other interfering constituents in the enriched urine sample was achieved by a
1596
Chem. Res. Toxicol., Vol. 18, No. 10, 2005
Chen and Chiu Table 2. Characteristics of the Study Population
sex (male/female) age (years) cigarettes/day smoking years smoking indexb
smokers (n ) 17) mean ( SD (range)
nonsmokers (n ) 13) mean ( SD (range)
17/0 45 ( 15 (24-68)a 15 ( 5 (5-25) 24 ( 15 (3-50) 370 ( 233 (15-750)
9/4 33 ( 9 (20-45) -
a Results are expressed as mean ( standard deviation (SD). The ranges are expressed in parentheses. b Smoking index ) number of cigarettes smoked per day × years of smoking.
Figure 3. Calibration curve of for the LC-ESI/MS/MS analysis of 1,N2-Gua. To samples containing various amounts (0-300 pg) of 1,N2-Gua was added a fix amount (1.0 ng) of [13C1,15N2]1,N2-Gua and subjected to the assay procedures described in Materials and Methods. The ratio of each analyte to the internal standard was calculated based on the peak areas. The data are combined from at least separate experiments in duplicates.
Figure 5. LC-ESI/MS/MS analysis of 1,N2-Gua in a smoker’s urine. The concentration of 1,N2-Gua in this sample was calculated as 21 pg/mL. Urine (1.0 mL) was added to [13C1,15N2]1,N2Gua, centrifuged, enriched with a SCX and C18-OH SPE column, and analyzed by LC-ESI/MS/MS under MRM mode as described in the Materials and Methods.
Figure 4. Accuracy of the LC-ESI/MS/MS assay was confirmed by addition of various amounts of standard 1,N2-Gua to urine samples containing [13C1,15N2]1,N2-Gua. The adduct concentrations were obtained from the y-intercept by the linear regression of the total concentrations in triplicate experiments. The adduct concentration of the urine measured without added 1,N2-Gua was (a) 147 pg/mL and (b) 0 pg/mL. Table 1. Precision of the LC-ESI/MS/MS for Urinary 1,N2-EGua
samples
1,N2-Gua levels (pg/mL) (RSD, %) (n ) 6) day 1 day 2 day 3
interday variation RSD (%)
1 2 3
20 ( 1 (6.7%) 20 ( 1 (6.6%) 23 ( 2 (9.1%) 93 ( 2 (1.8%) 95 ( 3 (3.1%) 91 ( 2 (1.6%) 143 ( 4 (2.5%) 138 ( 3 (1.9%) 146 ( 4 (2.9%)
7.4 1.9 3.0
reversed phase C18 HPLC column with an optimized eluting condition with a maximum content of the organic solvent to enhance the sensitivity of the ESI/MS/MS. Measurement of 1,N2-Gua in urine samples by method A was used throughout this study for its higher sensitivity. Sensitivity, Calibration, Precision, and Accuracy of the LC-ESI/MS/MS Assay. The individually opti-
Figure 6. Reverse addition experiment. To a smoker’s urine (1.0 mL) was added 1.0 ng of 1,N2-Gua and processed according to the assay procedures described in the Materials and Methods.
mized cone voltage and collision energy provide high sensitivity to 1,N2-Gua analysis. The limit of detection (LOD) for standard 1,N2-Gua was 1.0 pg (5.8 fmol) injected on-column, with a signal-to-noise (S/N) ratio greater than 3. The calibration curve constructed from 0 to 300 pg of 1,N2-Gua in the presence of 1.0 ng of [13C1,15N2]1,N2-Gua after sample pretreatment and adduct enrichment procedures was linear (r2 ) 0.9997) from 5.0 to 300 pg (Figure 3). In the control experiments containing only the isotope standard, no signal of 1,N2Gua was detected. Simulating the amount of adduct in 1.0 mL of urine, the concentration quantification limit (LOQ) was 5.0 pg/mL or 29 pM. This LOQ was 10 times lower than that reported by Gonzalez-Reche et al. (11) in which 20 mL of urine was used and an equivalent of 8 mL of urine was analyzed. The LOD and LOQ of 1,N2Gua are about 2 times lower than those of Ade (21).
LC-ESI/MS/MS Analysis of 1,N2-Gua in Human Urine
Chem. Res. Toxicol., Vol. 18, No. 10, 2005 1597
Table 3. Levels of 1,N2-EGua in Human Urine Samples and Statistical Data
1,N2-Gua
(pg/mL) 1,N2-Gua (nM) 1,N2-Gua/creatinine (pmol/mmol) 1,N2-Gua/body weight/creatinine (pmol/kg/g)
smokers male (n ) 17)a
nonsmokers (n ) 13)a
p-value b
nonsmokers male (n ) 9)a
p-value c
95 ( 41 (22-169) 0.54 ( 0.23 (0.13-0.97) 64 ( 28 (18-111)
50 ( 56 (0-199) 0.29 ( 0.32 (0-1.14) 26 ( 16 (0-53)
0.0247 0.0247
