Liquid Chromatography Electrospray-Mass Spectrometry of Urinary

Jun 27, 2001 - A liquid chromatography electrospray tandem mass spectrometry (LC-ESI-MS/MS) method for the measurement of aflatoxin biomarkers in urin...
4 downloads 8 Views 130KB Size
Download PDF
Chem. Res. Toxicol. 2001, 14, 919-926

919

Liquid Chromatography Electrospray-Mass Spectrometry of Urinary Aflatoxin Biomarkers: Characterization and Application to Dosimetry and Chemoprevention in Rats Michael Walton,† Patricia Egner,† Peter F. Scholl,‡ Jewell Walker,† Thomas W. Kensler,† and John D. Groopman*,† Department of Environmental Health Sciences, Johns Hopkins School of Hygiene and Public Health, 615 North Wolfe Street, Baltimore, Maryland 21205, and Applied Physics Laboratory, Johns Hopkins University, 11100 Johns Hopkins Road, Laurel, Maryland 20733 Received March 23, 2001

A liquid chromatography electrospray tandem mass spectrometry (LC-ESI-MS/MS) method for the measurement of aflatoxin biomarkers in urine has been developed and validated. The two major aflatoxin-DNA adducts formed in rat tissues, aflatoxin N7-guanine and its imidazole ring opened derivative, 8,9-dihydro-8-(2,6-diamino-4-oxo-3,4-dihydropyrimid-5-yl formamido)9-hydroxy-aflatoxin B1, were detected and quantified in urine by the LC-ESI-MS/MS technique. Other metabolites derived from the conjugation and/or oxidation of aflatoxin B1 measured in the urine of dosed rats included aflatoxin P1, aflatoxin P1-glucuronide, aflatoxin Q1, aflatoxin M1, 8,9-dihydro-8,9-dihydroxy aflatoxin B1, aflatoxin B1-mercapturic acid, the aflatoxin-cysteine glycine adduct derived from the aflatoxin-glutathione conjugate, aflatoxin M1P1 and the aflatoxin B1-dialcohol. For in vivo studies to determine the dosimetry of certain aflatoxin metabolites, aflatoxin B2 was used as an internal standard for recovery since this compound is not naturally produced in rats. In the final method using the internal standard, the coefficient of variation of six replicate analyses of in vivo rat urine samples for aflatoxin N7-guanine, aflatoxin B1-mercapturic acid, and aflatoxin M1 was 12.5, 12.8, and 5.8%, respectively. Further, the LC-ESI-MS/MS method to detect aflatoxin N7-guanine in in vivo rat urine samples was at least 20-fold more sensitive than prior techniques. Using the LC-ESI-MS/MS technique, the dosimetry, on a weekly basis, of major urinary aflatoxin metabolites was assessed in animals chronically dosed over a 5-week period. Of particular importance was the application of this method to determine the modulation of levels of urinary aflatoxin metabolites by treatment with oltipraz, a chemopreventive agent that can completely ablate aflatoxin hepatocarcinogenesis in the rat. After 1 week, oltipraz administration diminished urinary aflatoxin N7guanine, aflatoxin B1-mercapturic acid and aflatoxin M1 levels by 83, 92, and 82%, respectively. The magnitude of this reduction was persistent at the day 14, 21, 28, and 35-day time points with the average decrease of aflatoxin N7-guanine, aflatoxin B1-mercapturic acid and aflatoxin M1 being 73, 92, and 90%, respectively. Importantly, even under circumstances where the oltipraz intervention was most efficient in reducing aflatoxin metabolite levels, the LC-ESIMS/MS method was still sensitive enough to detect the reduced biomarker content. This outcome has important translational implications for the application and analysis of the efficacy of primary and secondary prevention interventions in human populations where ambient exposure levels are low, but the toxicologic hazards of these exposures remain high.

Introduction 1

Aflatoxin B1 (AFB1) is a naturally occurring mycotoxin that has been demonstrated to be one of the most potent * To whom correspondence should be addressed. E-mail: jgroopma@ jhsph.edu. † Department of Environmental Health Sciences. ‡ Applied Physics Laboratory. 1 Abbreviations: AFB , Aflatoxin B ; AFB -GS, AFB -glutathione 1 1 1 1 conjugate; AFB1-NAC, AFB1-mercapturic acid; AFB-CysGly, aflatoxincysteine glycine conjugate; AFM1, aflatoxin M1; AFQ1, aflatoxin Q1; AFP1, aflatoxin P1; AFB-N7-Gua, aflatoxin N7-guanine; AFB-FAPyr, 8,9-dihydro-8-(2,6-diamino-4-oxo-3,4-dihydropyrimid-5-yl formamido)9-hydroxy-aflatoxin B1; AFB2, aflatoxin B2; AFM1P1, aflatoxin M1P1; AFB-diol, 8,9-dihydro-8,9-dihydroxy aflatoxin B1; oltipraz, 4-methyl5-(pyrazin-2-yl)-1,2-dithiole-3-thione; LC-ESI-MS, liquid chromatography/electrospray ionization mass spectrometry; LC-ESI-MS/MS, liquid chromatography/electrospray ionization tandem mass spectrometry; PBS, 0.1 M phosphate-buffered saline, pH 7.4.

liver carcinogens tested in experimental animal models (1). Further, human epidemiologic studies demonstrate that exposure to AFB1 is a significant risk factor in the etiology of hepatocellular carcinoma, especially in combination with hepatitis B virus exposure (2, 3). Prevention of the adverse health effects from the unavoidable, chronic, and often high AFB1 exposure of Asian and subSaharan African populations is an important public health goal. Experimental models have provided insights into the molecular mechanisms of action of this carcinogen and these have been used to design chemopreventive protocols (4-6). The development of successful intervention strategies should be hastened by the use of aflatoxin biomarkers that efficiently assess their efficacy in individuals.

10.1021/tx010063a CCC: $20.00 © 2001 American Chemical Society Published on Web 06/27/2001

920

Chem. Res. Toxicol., Vol. 14, No. 7, 2001

Chemical specific biomarkers have been developed in response to the difficulties of traditional metrics of analysis to assess the exposure status and disease risk of individuals. In this context, evolving methods for the sensitive and specific detection of aflatoxin biomarkers are providing important tools for identifying individuals at high risk for disease from dietary exposure to aflatoxins. Substantial work has focused upon developing methods for analyzing the DNA and protein adducts derived from the ultimate carcinogenic metabolite of AFB1, aflatoxin-8,9-epoxide (7-11). Other biomarkers reflecting the overall fate of the aflatoxin-epoxide detoxication pathways are also useful, since a major factor for assessing the overall toxicological hazard following exposure of a person to AFB1 is the integrated balance between the activation and detoxication processes (1, 12). In this regard, the identification of the major aflatoxinmercapturic acid in rat urine has been of interest since it is the ultimate product of glutathione S-transferase mediated shunting of the epoxide to polar, excretable metabolites (13). Other oxidative metabolites of AFB1 formed by the same enzymes responsible for epoxide production have also been identified in animal and human urine, including AFM1, and have been applied in dosimetry studies (14, 15). Indeed, the urinary excretion of AFM1 has been shown to not only reflect human exposure (16-18); levels of this biomarker also presage chemopreventive effectiveness and liver cancer risk (2, 14, 15). While a number of studies have used urinary aflatoxin biomarkers as study endpoints, to date these investigations have only surveyed a fraction of the total spectrum of aflatoxin metabolites excreted in the urine of experimental animals or humans. For example, AFP1, the demethylated metabolite, is also found in urine, but the utility of this marker has yet to be evaluated in chemoprevention trials (17, 19). Further, significant levels of AFP1-glucuronide have been detected in urine and experimentally it is known to be increased by cytochrome P450 inducers such as phenobarbital (20, 21). Another oxidative metabolite, AFQ1, has also been detected in human urine (22) and its glucuronide occurs in biologic samples (23). Thus, a comprehensive approach using highly specific and sensitive analytical methods to evaluate the predictive value of these metabolites as exposure and/or risk markers needs to be developed (24). Traditional methods for the identification and measurement of urinary aflatoxin biomarkers have lacked the requisite sensitivity to make facile measurements in small sample volumes. For example, the recent identification and characterization of the aflatoxin-mercapturic acid in rat urine required large volumes of urine obtained from animals exposed to high doses of AFB1 (13). Fortunately, in the past five years significant advances in mass spectrometry instrumentation have occurred, which now makes practical the characterization of these biomarkers in manageable samples sizes. The work reported herein describes a quantitative method and validation strategy for measuring aflatoxin biomarkers in urine by LC-ESIMS/MS.

Materials and Methods Chemicals. Caution: AFB1 is a human carcinogen. Great care should be exercised to avoid personal exposure and proper decontamination procedures using 1% sodium hypochlorite

Walton et al. (bleach) should be used. AFB1, AFM1, AFQ1, AFB2 and AFP1 were purchased from Sigma Chemical Co. (St. Louis, MO). AFBN7-Gua and AFB-FAPyr were purified by HPLC following the in vitro reaction of AFB1-epoxide and calf thymus DNA and subsequent hydrolysis by 0.1 N HCl. The AFB-NAC and AFBCysGly had previously been synthesized and characterized (13). All metabolites were identified and quantified by spectrophotometric, HPLC, and MS/MS techniques following comparison to authentic standards generously contributed by Dr. Gerald Wogan, MIT. All chromatographic solvents utilized were 99.9% + HPLC, MS grade purity (Burdick and Jackson, Inc., Muskegon, MI). All other chemicals were of the highest grade commercially available. Animals, Diets, and Treatments. Male F344 rats (100-150 g; Harlan, Indianapolis, IN) were housed under controlled conditions of temperature, humidity, and lighting. Food and water were available ad libitum. Purified diet of the AIN-76A formulation (Teklad, Madison, WI) without the recommended addition of 0.02% ethoxyquin was used, and fresh diet was provided to animals at least every other day. Rats were acclimated to the AIN-76A diet for 1 week before beginning the experiments. In total, three separate experiments were performed. In the first study, six rats were dosed by gavage for 10 consecutive days with 25 µg of AFB1 in 100 µL of dimethyl sulfoxide (DMSO). These animals were housed in glass metabolic cages and urine was collected daily on dry ice. In the second experiment, three rats per group were dosed with four levels of AFB1 (0.2, 1.0, 5.0, and 25 µg) in 100 µL of DMSO by gavage. These animals were then placed in metabolic cages and the urine was collected on dry ice for the following 24 h. For the third experiment, eight rats starting at 7 weeks of age, were fed either control diet or diet supplemented with 0.05% oltipraz for 5 consecutive weeks. At eight weeks of age, all rats received 20 µg of AFB1 in 100 µL of DMSO by gavage daily for 4 weeks. Twenty-four hour urine samples were then collected on dry ice during this dosing period (25). Identification and Characterization of Aflatoxins Metabolites in Urine By ESI-LC/MS/MS. Rat urine samples were first adjusted to an acidic pH using 1 M ammonium formate, pH 4.5, and then the total volume of the samples was increased to 5 mL with water. The sample was spiked with 2 ng of aflatoxin B2, as an internal standard, and loaded onto a Waters Oasis HLB 3 mL column (Waters Corp. Milford, MA) previously conditioned with 5 mL of MeOH followed by 5 mL of water. The preparative column was sequentially washed with 5 mL of water, and the aflatoxins were eluted from the column into a 5 mL ReactiVial (Pierce, Rockford, IL) with 3 mL of total (3 × 1 mL) of 100% MeOH; 150 µL of 1% glacial acetic acid: water was then added to the sample. The MeOH was reduced using an ultra-high-purity argon stream to a final volume of approximately 200 µL. This fraction was then brought up to 2 mL with water and loaded onto an aflatoxin-specific preparative monoclonal antibody immunoaffinity column at flow rate of 0.3 mL/min, as described previously (8, 15, 26). The affinity column was washed twice with 5 mL of PBS and once with 10 mL of water to remove nonspecifically bound materials. Aflatoxin derivatives were then eluted from the immunoaffinity column with 4 vol of 70% DMSO:water followed by another 2 vol water wash. The DMSO and water eluates were combined, diluted with 20 mL of water and then applied to another preprimed Waters Oasis HLB 3 mL column. The Oasis column was washed with 10 mL of water to remove DMSO. Aflatoxin derivatives were subsequently eluted with 3 mL of total (3 × mL) of 100% MeOH; 150 µL of 1% glacial acetic acid:water was then added to the sample. The MeOH fraction was then concentrated to a final volume of 30 µL under an argon stream for analysis by LC-ESI-MS. The monoclonal antibodies used in preparing the immunoaffinity resins originated from ascites raised against AFB1lysine, AFM1, and the general AFB moiety (26). Column capacities exceeded 1 µg of AFB-N7-gua, AFB-NAC, and AFM1. Using rat urine samples spiked with a 2 ng of AFB2 internal

LC-ESI-MS of Urinary Aflatoxin Biomarkers Table 1: Identification of in Vivo Rat Urinary Aflatoxin Metabolites by ESI-LCMS and ESI-LC MS/MS metabolite

MH+

MS/MS ions

nucleic acid adducts AFB-N7-Gua 480.1 329.1 (100),a 152.4 (94) AFB-FAPyr 498.1 480.1 (100), 452.0 (62) oxidation metabolites AFM1 329.1 301.1 (100), 273.1 (85) 329.1 311.1 (100), 301.1 (35) AFQ1 AFP1 299.1 271.1 (100) AFB-Diol 347.0 329.1 (100), 283.0 (32) 315.1 269.1 (100), 297.1 (20) AFM1P1 AFB-dialcohol 351.1 333.1 (100) conjugate metabolites AFB-NAC 492.1 329.1 (100) AFP-glucuronide 475.1 299.2 (100) AFB-CysGly 507.1 471.1 (95), 329.1 (100), 425.1 (25) a

Values in parentheses indicate relative abundance of ions.

standard, a 85-90% recovery for this procedure was routinely obtained. A Thermoquest Finnigan LCQ LC/MS system was used to conduct the LC-ESI-MS and LC-ESI-MS/MS analyses of the immunoaffinity processed urine to identify and quantify the aflatoxin derivatives. A Thermal Systems Products HPLC was used to provide a constant flow of 100 µL/min to a 1 × 250 mm J’Sphere ODS-M80 microbore column (Waters Assoc., Milford, MA). The HPLC column temperature was maintained at 55 °C. For these analyses, gradients using combinations of acetonitrile and methanol were used to effect separation of the aflatoxins. The initial conditions for the HPLC separation was 2% acetonitrile: 4% MeOH and a linear gradient over 15 min was used that led to final conditions of 17% acetonitrile:34% MeOH. The aqueous buffer used in all the separations was 1% glacial acetic acid: water. The LC-ESI-MS/MS was conducted with the capillary temperature set at 200 °C. The sheath gas was set to maximum of 70 arbitrary units and the source voltage was 4 kV. The instrument was tuned daily using a solution of AFB-N7-Gua at a concentration of 10 µg/mL in 50% MeOH:1% glacial acetic acid: water. This tuning solution was also used to optimize the collision energy needed for the MS/MS analyses. Throughout the experiments a collision energy of 26% was found to be optimal for the dissociation of the aflatoxin-DNA adducts. Collision energy settings were also optimized for AFM1 and AFP1 at 35 and 40%, respectively. Linear standard calibration curves were obtained for AFB-N7-Gua, AFM1, and AFB-NAC with limits of detection of 1, 2, and 10 pg, respectively. Urine samples were normalized using a spectrophotometric creatinine kit (Sigma, St. Louis, MO). Statistical Analysis. All analytical data are expressed as mean ( SE (standard error) and levels of biomarkers were compared across the study and statistically analyzed by Student t-test.

Results Identification of Aflatoxin-Nucleic Acid, Oxidation, and Conjugate Metabolites in Rat Urine by LC-ESI-MS/MS. Experiments were performed to identify and characterize by liquid chromatography and mass spectrometry the spectrum of aflatoxin metabolites present in rat urine following acute and chronic administration of AFB1. By using the retention time data obtained with authentic standards, where available, and the mass spectrometry information, a number of metabolites were characterized as summarized in Table 1. The data in Table 1 are separated into three broad classification areas: nucleic acid adducts, oxidation metabolites, and conjugates.

Chem. Res. Toxicol., Vol. 14, No. 7, 2001 921

The two major aflatoxin-DNA adducts formed in rat liver are the guanine derived adducts, AFB-N7-Gua and its imidazole-ring opened derivative, AFB-FAPyr (2729). AFB-N7-Gua has been detected previously in rat urine using analytical chemical methods (26, 30), but not using LC-ESI-MS/MS, while the excretion of the AFBFAPyr adduct has not been reported previously. In Figures 1 and 2 are shown the LC-MS and LC-MS/MS chromatographic and spectral profiles of the two major nucleic acid adducts isolated from rat urine. Figures 1A and 2A illustrate the retention time of each of these DNA adducts by LC-ESI-MS monitoring of the major MH+ ions. In Figure 1B, the MS/MS signal for AFB-N7-Gua using the 329.1 and 152.4 MH+ fragment ions is shown. The signal-to-noise ratio of the MS peak in Figure 1A is 7 and this signal-to-noise ratio was increased to 523 by using the selectivity of MS/MS shown in Figure 1B. In Figure 1C and 2B, the MS spectra showing the major MH+ ions for the AFB-N7-Gua and AFB-FAPyr at 480.1 and 498.1, respectively, are illustrated. The MS/MS spectra obtained for the urinary AFB-N7-Gua (Figure 1D) reveals two major ions of approximately equal proportions at MH+ 329.1 and 152.4, reflecting the cleavage of this nucleic acid adduct into 9-hydroxy-aflatoxin and guanine positive ions. The MS/MS spectrum (Figure 2C) derived from the AFB-FAPyr adduct reveals a predominate fragment ion at 480.1, resulting from the loss of water (m/z 18), and a positive ion with a relative abundance of 62 at 452.0 that reflects the loss of a carbonyl (m/z 28); most likely from the carbonyl group derived from the C-8 position of the original guanine moiety. The major fragment ions were then used to develop a MS/MS quantitative method to detect AFBN7-Gua and AFB-FAPyr. The limit for detection of the nucleic acid adducts was 1 pg (about 2 fmol). Using this LC/MS method, the dose response characteristics for the excretion of AFB-N7-Gua in 24 h urine samples were examined in rats dosed with 0.2-25 µg AFB1. These data are shown in Figure 3 and indicate a linear excretion of the AFB-N7-Gua adduct over this exposure range, with a correlation coefficient of 0.96. In Groopman et al. (26), we reported the dose-response characteristics of this adduct in rat urine using a monoclonal antibody immunoaffinity column cleanup followed by HPLC with UV diode-array detection. Those earlier studies also indicated a linear excretion of the nucleic acid adduct (r2 ) 0.99); however, the lowest dose of AFB1 from which adducts could be detected was 4 µg. Since the same monoclonal antibody immunity affinity purification method was used in these current studies, we combined the data reported in Figure 3 with our earlier results. This analysis revealed that our current data extended the linear dose response excretion curve and that the overall combined data had a correlation coefficient of 0.962. Thus, the LC-ESI-MS/MS detection method decreased the limit of detection by at least 20fold compared to the previously reported analytic method (26). These current findings lend some indirect, but useful confirmation of the previously reported nucleic acid adduct excretion data (8, 17, 30). Two of the major oxidation metabolites of AFB1 found in rat urine were AFM1 and AFQ1. Both of these hydroxylated metabolites have a parent MH+ ion of 329.1; however, their MS/MS spectra were quite distinct. AFM1 was previously identified in human urine using a similar LC-ESI-MS/MS procedure (15). The characteristic frag-

922

Chem. Res. Toxicol., Vol. 14, No. 7, 2001

Walton et al.

Figure 1. Liquid chromatography-electrospray mass spectroscopy of AFB-N7-Gua isolated from rat urine of animals multiply dosed dosed by gavage for 10 consecutive days with 25 µg of AFB1. Panel A is a chromatogram of the HPLC separation of aflatoxin metabolites following isolation from rat urine by immunoaffinity chromatography. The profile illustrates the elution of 480.1 MH+ molecular ions including the parent compound, AFB-N7-Gua, at 15.8 min. Panel B depicts the MS/MS scan of AFB-N7-Gua by monitoring the 329.1 and 152.1 fragment ions. The signal-to-noise ratio in panel A is 7 and in panel B is 523. Panel C depicts the full-scan MS for the major peak found at a retention time of 15.8 min in panel A. Panel D depicts the MS/MS profile resulting from colliding the major peak found in panel A.

ment ions for AFM1 were found at 301.1 and 273.1, probably reflecting sequential losses of carbonyl groups (m/z 28) from the parent ion molecule. In contrast, AFQ1 has an MS/MS profile whose predominant fragment ion was at 311.1 with a relative intensity of 100. This fragment ion was the consequence of the loss of water (m/z 18) from the secondary hydroxy group of AFQ1 at the 3-position of the aflatoxin structure. In contrast, the tertiary nature of the hydroxy group at the 10-position in AFM1 renders greater stability to this functional group. The limit of detection for the MS/MS method for AFM1 and AFQ1 was 2 pg (6 fmol). AFP1, a demethylated metabolite of AFB1, was identified in urine by its parent ion at 299.1 and its MS/MS major fragment ion at 271.1 reflecting the loss of a carbonyl residue (19). AFP1 was previously found to be a major aflatoxin metabolite in human urine samples (2, 7). AFB-diol which is derived from the addition of water to the aflatoxin-8,9-epoxide (31) was identified by its parent ion at 347.0 with the major fragment ion occurring at 329.1, reflecting the loss of water (m/z 18). Previous data in the literature has described the occurrence in bile of a dihydroxy aflatoxin metabolite, AFM1P1, generated by the demethylation of AFB1 and a subsequent oxidation at the 10-position (32). The AFM1P1 metabolite was detected in rat urine from its parent ion at 315.1 and from its MS/MS spectra. The predominant fragment ion of AFM1P1 was at 269.1, reflecting a m/z loss of 46 that can be attributed to a simultaneous loss of water and a

carbonyl moiety from the parent ion structure. Finally, the AFB-dialcohol that has been described by Judah et al. (33) was tentatively identified in the rat urine from its MH+ ion at 351.1 and a predominant fragment ion at 333.1, that reflects the loss of water (m/z 18). Three major conjugate metabolites of AFB1 were detected in rat urine, including the mercapturic acid, AFB-NAC, that had previously been shown in both rat and human urine using mass spectrometry (13, 15). Another thiol metabolite from the glutathione cascade, AFB-CysGly, was also detected. The AFB-CysGly parent ion was found at 507.1 and the major fragment ions were at 471.1 and 329.1 reflecting the loss of two water molecules and the resulting 9-hydroxy aflatoxin moiety, respectively. The AFB-CysGly conjugate had previously been detected by traditional analytical chemical methods in rat and hamster samples (34). To date, three aflatoxin glucuronides, formed from AFP1, AFQ1, and aflatoxicol, have been reported in the literature (19, 35). The AFP-glucuronide (19) was identified in the rat urine samples by the parent ion at 475.1 and the MS/MS fragment revealing the 299.1 ion for AFP1. Thus, the neutral loss of the glucuronic acid leads to this positive ion fragment. The glucuronides of aflatoxicol or AFQ1 were not detected in any of the rat urine samples. Optimization of the Immunoaffinity Chromatography/LC-ESI-MS/MS Method. The immunoaffinity chromatography/LC-ESI-MS/MS preparation and analy-

LC-ESI-MS of Urinary Aflatoxin Biomarkers

Chem. Res. Toxicol., Vol. 14, No. 7, 2001 923

Figure 3. Dose response for the excretion of AFB-N7-Gua into rat urine over 24 h as determined by the LC-ESI/MS/MS technique.

Figure 2. Liquid chromatography-electrospray mass spectroscopy of AFB-FAPyr isolated from rat urine of animals multiply dosed dosed by gavage for 10 consecutive days with 25 µg of AFB1. Panel A is a chromatogram of the HPLC separation of aflatoxin metabolites following isolation from rat urine by immunoaffinity chromatography. The profile illustrates the elution of 498.1 MH+ molecular ions including the parent compound, AFB-FAPyr, at 18.2 min. Panel B depicts the fullscan MS for the major peak found at a retention time of 18.2 min in panel A. Panel C depicts the MS/MS profile resulting from colliding the major peak found in panel A.

sis method for the urinary aflatoxin biomarkers was optimized for the major products derived from oxidation by CYP1A2 and CYP3A4, AFB-N7-Gua, AFB-NAC, and AFM1. To affect the MS and MS/MS optimization, 10 µg/ mL solutions of AFB-N7-Gua, AFB-NAC, and AFM1 were

prepared. These solutions then were used to independently tune the mass spectrometer by infusion into a T-fitting at a flow rate of 3 µL/min. HPLC mobile phase was introduced through the other part of the T-fitting and initially aqueous mobile phases containing 1.0 and 0.1% acetic acid and 1.0 and 0.1% formic acid were tested. For each of the aflatoxin metabolites, the greatest ion signal was obtained using 1% acetic acid. Thus, all dosimetry studies used 1% acetic acid as aqueous the mobile phase for HPLC. Following the determination that 1% acetic acid produced the highest signal for the aflatoxins, the capillary temperature was examined over a range of 150 °C to 250 °C. In these studies, 200 °C was found to give the highest ion level. These aflatoxin tuning solutions were used to adjust the mass spectrometer parameters on a daily basis. Since ESI-MS/MS analysis was used to specifically determine the levels of each of these aflatoxin metabolites, the collision energy to produce the maximal collision ions were also optimized experimentally using these tuning solutions. In the final immunoaffinity clean-up mass spectrometry method, AFB2 was used as an internal standard for recovery since this compound is not produced in vivo in rats. In this final method the coefficient of variation of six replicate analyses of in vivo rat urine samples for AFB-N7-Gua, AFBNAC, and AFM1 was 12.5, 12.8, and 5.8%, respectively. Modulation of Urinary Aflatoxin Biomarkers During Chemoprevention Intervention by Oltipraz in the Rat. Levels of measured urinary aflatoxin biomarkers during an oltipraz intervention in aflatoxintreated rats are shown in Figure 4 and Table 2. Aflatoxin was administered daily to the animals beginning on day 1 for the 7, 14, 21, 28, and 35-day time points where the urine metabolite levels were determined. At day 7, concurrent oltipraz administration diminished urinary AFB-N7-Gua, AFB-NAC, and AFM1 levels by 83, 92, and 82%, respectively. This magnitude of reduction persisted throughout the 14, 21, 28, and 35-day time points with the average decreases of AFB-N7-Gua, AFB-NAC and AFM1 being 73, 92, and 90%, respectively. The AFB-N7Gua, AFB-NAC, and AFM1 biomarkers were also measured in the control and oltipraz treatment arms at day 39, which was 4 days after the cessation of aflatoxin administration. In accord with their short biological halflives, absolute levels of these biomarkers were substantially reduced compared to levels during the active dosing periods. Moreover, levels of these biomarkers were

924

Chem. Res. Toxicol., Vol. 14, No. 7, 2001

Walton et al.

Figure 4. Dosimetry for the urinary excretion of AFB-N7-Gua, AFB-NAC, and AFM1 in control and oltipraz-treated rats. Aflatoxin (20 µg, gavage) was administered daily for 35 consecutive days. Aflatoxin biomarkers were determined on 24 h urine samples collected at the end of each week. Table 2 control

oltipraz

day

AFM1

AFB-N7-Gua

AFB1-NAC

day

AFM1

AFB-N7-Gua

AFB1-NAC

7 14 21 28 35 39

8.5 ( 3.4a 6.3 ( 2.0 9.0 ( 3.4 15.7 ( 8.5 13.0 ( 8.1 2.0 ( 1.6

3.4 ( 1.1 2.0 ( 0.3 3.6 ( 1.1 3.9 ( 1.3 3.8 ( 1.3 0.9 ( 0.06

5.0 ( 2.8 3.5 ( 1.2 10.0 ( 4.0 6.9 ( 2.1 8.2 ( 3.1 1.0 ( 0.3

7 14 21 28 35 39

1.5 ( 0.3 1.3 ( 0.3 1.0 ( 0.2 0.8 ( 0.1 0.05 ( 0.1 0.05 ( 0.02

0.6 ( 0.1 0.7 ( 0.1 1.0 ( 0.2 0.9 ( 0.3 0.9 ( 0.1 0.3 ( 0.09

0.4 ( 0.05 0.3 ( 0.03 0.4 ( 0.06 0.5 ( 0.2 0.8 ( 0.3 0.2 ( 0.04

a

(Nanograms of AF per milligram of creatinine) ( SD.

reduced by 68, 77, and 98%, respectively, by oltipraz. For each of the three biomarkers, the reductions produced by oltipraz were statistically significant (P < 0.005). Further, even under circumstances where oltipraz chemoprevention was most efficient in reducing aflatoxin metabolite levels, the LC-ESI-MS/MS method was still sensitive enough to detect the reduced levels of the biomarkers. This outcome has important translational implications for the application and analysis of the efficacy of interventions in human populations where ambient exposure levels are low, but the toxicologic hazards of these exposures remain high.

Discussion The application of biomarkers to investigate the dosimetry of environmental carcinogens has been a major focus for molecular epidemiology during the past 15 years (36). Given the relatively low levels of ambient chemical exposures, the analytical methods needed to measure these carcinogen biomarkers must be both specific and highly sensitive. This need is even more critical for the assessment of the efficacy of an intervention that seeks to reduce the amount of the biomarker expected to be present in a sample. A number of different analytical strategies have been developed for the quantitation of these biomarkers in biological samples, including chromatography, especially high-performance liquid chromatography, immunological assays with specific antibodies

or antisera, such as enzyme-linked immunosorbent assays (ELISA) and radioimmunoassays (RIA), and immunohistochemistry visualization in tissues, and instrumentation-based methods (synchronous fluorescence and mass spectroscopy). Each analytic methodology has unique specificity and sensitivity and, depending upon the application, the user can choose which is most appropriate. Thus, to measure a single aflatoxin metabolite present at relatively high levels, a chromatographic method can resolve the mixtures of aflatoxins into individual compounds, and then both ultraviolet (UV) and fluorescence detectors can be used to quantify the level of metabolite or parent compound. UV detection at the absorbance maxima for most aflatoxin metabolites results in a limit of detection of between 0.5 and 1 ng (17, 27). This limit of detection was decreased to about 20 pg when fluorescence detection was used following a postcolumn derivatization procedure (37). Unfortunately, postcolumn derivatization of the aflatoxins is only applicable for those species that contain an unsaturated double bond in the furo-furan ring and many of the metabolites of interest to toxicologists involve enzymatic alterations to this part of the molecule. The operational limit of detection for most of the aflatoxin metabolites, by direct fluorescence was in the picogram range. For example, the direct fluorescence properties of AFM1 permits its measurement at the 5-10 pg level (15). Synchronous fluorescence spectroscopy methods have

LC-ESI-MS of Urinary Aflatoxin Biomarkers

also been applied to aflatoxin metabolites and the limit of detection for a spectrum of aflatoxin metabolites was found to be between 30 and 100 pg (38). An advantage of the synchronous fluorescence method was that some of the aflatoxin metabolites exhibited unique spectra and this permitted some selectivity for the ultimate analysis. Over the past 20 years the combination of separation using HPLC with a variety of UV and fluorescence detectors have provided the major means for measuring a number of significant metabolites of the aflatoxins. HPLC using standard reversed-phase columns with either UV or fluorescence detection have probably reached their ultimate sensitivity limit. In the future, the application of microbore and nanobore HPLC columns may decrease the practical limit of detection of the aflatoxins by perhaps 10-fold, but these columns have yet to be evaluated in many molecular dosimetry applications. Mass spectroscopy is becoming the standard for metabolite identification and quantification in biological samples. In the present study, mass spectroscopy has been used to identify the excretion products of the two major aflatoxin-DNA adducts formed in rat tissue, AFBN7-Gua and AFB-FAPyr (29, 39). Other metabolites derived from the conjugation and/or oxidation of AFB1 that have been identified in rat samples by mass spectroscopy include AFB-NAC, AFP1, AFP1-glucuronide, AFQ1, AFB-DIOL, AFM1, AFB-CysGly, AFM1P1, and the AFB-dialcohol. The sensitivity of the LC-ESI-MS instrumentation enabled us several years ago to develop an efficient analysis of aflatoxin mercapturic acid conjugates in both immunoaffinity processed urine samples and crude reaction mixtures (13). In that work, the mass spectrum of the AFB-NAC formed in vivo exhibited the expected parent and sodium adduct ions at m/z 492 and 514. To develop the LC-ESI-MS/MS method reported here, we have taken advantage of the strengths of antibodies to isolate and purify trace levels of aflatoxin metabolites in biological samples and then use MS/MS to quantify content. An earlier iteration of this method was used to verify the measurement of aflatoxin metabolites from a clinical trial of oltipraz that were measured using HPLC with fluorescence UV for detection. Although significant changes in the critical levels of AFM1 and AFB-NAC were observed, a substantial number of undetectable values mandated a larger than desirable size for the clinical trial to preserve statistical power (15). Since oltipraz was found to be both an inhibitor of cytochrome P450 1A2 (14) and an inducer of glutathione-S-transferase (15) mediated pathways for aflatoxin metabolites, we needed to develop sensitive and specific methods to examine the modulation of the major aflatoxin biomarkers formed from these pathways. The studies described in this report on the approach of LC-ESI-MS/MS to the dosimetry of aflatoxin in rats indicates that the method brings heretofore unavailable specificity to the analysis of urinary aflatoxin biomarkers. While the method appears to be quantitatively robust for measurements of biomarker levels in experimental models, challenges remain for enhancing its sensitivity efficiently for routine application in human biomonitoring and intervention studies. Clearly, the 20-fold increase in sensitivity of LC-ESI-MS/MS for the detection of AFBN7-Gua relative to previous HPLC-UV methods is an encouraging step in the right direction. While the use of AFB2 as an internal standard promotes general estimates

Chem. Res. Toxicol., Vol. 14, No. 7, 2001 925

for the efficiency of recovery of analytes, the development of isotopically enriched standards for all metabolites will ultimately be needed. Nevertheless, we anticipate that the current methodology will greatly facilitate the analysis of several ongoing intervention studies (exposure reduction and chemoprevention) in high-risk populations.

Acknowledgment. Financial support for this work was provided in part by grants from NIH including P01 ES06052 and P30 ES03819. The authors also wish to acknowledge the technical assistance provided by Xia He.

References (1) Eaton, D. L., and Gallagher, E. P. (1994) Mechanisms of aflatoxin carcinogenesis. Annu. Rev. Pharmacol. Toxicol. 34, 135-172. (2) Qian, G.-S., Ross, R. K., Yu, M. C., Yuan, J.-M., Gao, Y.-T., Henderson, B. E., Wogan, G. N., and Groopman, J. D. (1994) A follow-up study of urinary markers of aflatoxin exposure and liver cancer risk in Shanghai, People’s Republic of China. Cancer Epidemiol. Biomarkers Prev. 3, 3-10. (3) Jackson, P. E., and Groopman, J. D. (1999) Aflatoxin and liver cancer. Bailliere’s Clin. Gastroenterol. 13, 545-555. (4) Roebuck, B. D., Liu, Y.-L., Rogers, A. E., Groopman, J. D., and Kensler, T. W. (1991) Protection against aflatoxin B1 induced hepatocarcinogenesis in F344 rats by 5-(2-pyrazinyl)-4-methyl1,2-dithiole-3-thione (oltipraz): Predictive role for short-term molecular dosimetry. Cancer Res. 51, 5501-5506. (5) Kensler, T. W., Egner, P. A., Davidson, N. E., Roebuck, B. D., Pikul, A., and Groopman, J. D. (1986) Modulation of aflatoxin metabolism, aflatoxin-N7-guanine formation, and hepatic tumorigenesis in rats fed ethoxyquin: Role of induction of glutathione S-transferases. Cancer Res. 46, 3924-3931. (6) Kensler, T. W., Groopman, J. D., Eaton, D. L., Curphey, T. J., and Roebuck, B. D. (1992) Potent inhibition of aflatoxin-induced hepatic tumorigenesis by the monofunctional enzyme inducer 1,2dithiole-3-thione. Carcinogenesis 13, 95-100. (7) Ross, R. K., Yuan, J.-M., Yu, M. C., Wogan, G. N., Qian, G.-S., Tu, J.-T., Groopman, J. D., Gao, Y.-T., and Henderson, B. E. (1992) Urinary aflatoxin biomarkers and risk of hepatocellular carcinoma. Lancet 339, 943-946. (8) Groopman, J. D., Jiaqi, Z., Donahue, P. R., Pikul, A., Lisheng, Z., Jun-shi, C., and Wogan, G. N. (1992) Molecular dosimetry of urinary aflatoxin-DNA adducts in people living in Guangxi Autonomous Region, People’s Republic of China. Cancer Res. 52, 45-52. (9) Gan, L.-S., Skipper, P. L., Peng, X., Groopman, J. D., Chen, J., Wogan, G. N., and Tannenbaum, S. R. (1988) Serum albumin adducts in the molecular epidemiology of aflatoxin carcinogenesis: Correlation with aflatoxin B1 intake and urinary excretion of aflatoxin M1. Carcinogenesis 9, 1323-1325. (10) Wild, C. P., Hudson, G. J., Sabbioni, G., Chapot, B., Hall, A. J., Wogan, G. N., Whittle, H., Montesano, R., and Groopman, J. D. (1992) Dietary intake of aflatoxins and the level of albumin-bound aflatoxin in peripheral blood in The Gambia, West Africa. Cancer Epidemiol. Biomarkers Prev. 1, 229-234. (11) Bolton, M. G., Mun˜oz, A., Jacobson, L. P., Groopman, J. D., Maxuitenko, Y. Y., Roebuck, B. D., and Kensler, T. W. (1993) Transient intervention with olitpraz protects against aflatoxininduced hepatic tumorigenesis. Cancer Res. 53, 3499-3504. (12) Ramsdell, H. S., and Eaton, D. L. (1990) Species susceptibility to aflatoxin B1 carcinogenesis: Comparative kinetics of microsomal biotransformation. Cancer Res. 50, 615-620. (13) Scholl, P. F., Musser, S. M., and Groopman, J. D. (1997) Synthesis and characterization of aflatoxin B1 mercapturic acids and their identification in rat urine. Chem. Res. Toxicol. 10, 1144-1151. (14) Scholl, P., Musser, S. M., Kensler, T. W., and Groopman, J. D. (1996) Inhibition of aflatoxin M1 excretion in rat urine during dietary intervention with oltipraz. Carcinogenesis, 17, 1385-1388. (15) Wang, J.-S., Shen, X., He, X., Zhu, Y., Zhang, B.-C., Wang, J.-B., Qian, G.-S., Kuang, S.-Y., Zarba, A., Egner, P. A., Jacobson, L. P., Munoz, A., Helzlsouer, K. J., Groopman, J. D., and Kensler, T. W. (1999) Protective Alterations in Phase 1 and 2 Metabolism of Aflatoxin B1by Oltipraz in Residents of Qidong, People’s Republic of China. J. Natl. Cancer Inst. 91, 347-354. (16) Campbell, T. C., Caedo, J. P., Jr., Bulatao-Jayme, J., Salamat, L., and Engel, R. W. (1970) Aflatoxin M1 in human urine. Nature 227, 403-404. (17) Groopman, J. D., Donahue, P. R., Zhu, J. Q., Chen, J. S., and Wogan, G. N. (1985) Aflatoxin metabolism in humans: detection

926

(18)

(19) (20) (21)

(22)

(23) (24)

(25)

(26)

(27)

(28)

Chem. Res. Toxicol., Vol. 14, No. 7, 2001 of metabolites and nucleic acid adducts in urine by affinity chromatography. Proc. Natl. Acad. Sci. U.S.A. 82, 6492-6496. Zhu, J. Q., Zhang, L. S., Hu, X., Xiao, Y., Chen, J. S., Xu, Y. C., Fremy, J., and Chu, F. S. (1987) Correlation of dietary aflatoxin B1 levels with excretion of aflatoxin M1 in human urine. Cancer Res. 47, 1848-1852. Dalezios, J., Wogan, G. N., and Weinreb, S. M. (1971) Aflatoxin P-1: a new aflatoxin metabolite in monkeys. Science 171, 584585. Ehrich, M., Huckle, W. R., and Larsen, C. (1984) Increase in glucuronide conjugation of aflatoxin P1 after pretreatment with microsomal enzyme inducers. Toxicology 32, 145-152. Holeski, C. J., Eaton, D. L., Monroe, D. H., and Bellamy, G. M. (1987) Effects of phenobarbital on the biliary excretion of aflatoxin P1-glucuronide and aflatoxin B1-S-glutathione in the rat. Xenobiotica 17, 139-153. Kussak, A., Andersson, B., and Andersson, K. (1994) Determination of aflatoxin Q1 in urine by automated immunoaffinity column cleanup and liquid chromatography. J. Chromatogr., B: Biomed. Sci. Appl. 656, 329-334. Rohrig, T. P., and Yourtee, D. M. (1983) In vitro metabolism of aflatoxin Q1 by rat liver post-mitochondrial homogenates. Res. Commun. Chem. Pathol. Pharmacol. 40, 457-464. Kensler, T. W., Davidson, N. E., Groopman, J. D. and Munoz, A. (2001) Biomarkers and surrogacy: Relevance to chemoprevention. In Biomarkers in Cancer Chemoprevention (Miller, A. B., Bartsch, H., Boffetta, P., Dragsted, L., and Vainio, H., Eds.) No. 154, pp 27-47, IARC Scientific Publications. Kensler, T. W., Gange, S. J., Egner, P. A., Dolan, P. M., Munoz, A., Groopman, J. D., Rogers, A. E., and Roebuck, B. D. (1997) Predictive value of molecular dosimetry: Individual versus group effects of oltipraz on aflatoxin-albumin adducts and risk of liver cancer. Cancer Epidemiol. Biomarkers Prev. 6, 603-610. Groopman, J. D., Hasler, J. A., Trudel, L. J., Pikul, A., Donahue, P. R., and Wogan, G. N. (1992) Molecular dosimetry in rat urine of aflatoxin-N7-guanine and other aflatoxin metabolites by multiple monoclonal antibody affinity chromatography and immunoaffinity/high performance liquid chromatography. Cancer Res. 52, 267-274,. Croy, R. G., Essigmann, J. M., Reinhold, V. N., and Wogan, G. N. (1978) Identification of the principal aflatoxin B1-DNA adduct formed in vivo in rat liver. Proc. Natl. Acad. Sci. U.S.A. 75, 17451749. Croy, R. G., and Wogan, G. N. (1981) Quantitative comparison of

Walton et al.

(29)

(30) (31) (32)

(33)

(34) (35)

(36) (37) (38)

(39)

covalent aflatoxin-DNA adducts formed in rat and mouse livers and kidneys. J. Natl. Cancer Inst. 66, 761-767. Hertzog, P. J., Smith, J. R. L., and Garner, R. C. (1982) Short Communication. Characterisation of the imidazole ring-opened forms of trans-8,9-dihydro-8-(7-guanyl)9-hydroxy aflatoxin B1. Carcinogenesis 3, 723-725. Bennett, R. A., Essigmann, J. M., and Wogan, G. N. (1981) Excretion of an aflatoxin-guanine adduct in the urine of aflatoxin B1-treated rats. Cancer Res. 41, 650-654. Guengerich, F. P., Johnson, W. W., Shimada, T., Ueng, Y.-F., Yamazaki, H., and Langouet, S. (1998) Activation and dexotication of aflatoxin B1. Mutat. Res. 402, 121-128. Eaton, D. L., Monroe, D. H., Bellamy, G., and Kalman, D. A. (1988) Identification of a novel dihydroxy metabolite of aflatoxin B1 produced in vitro and in vivo in rats and mice. Chem. Res. Toxicol. 1, 108-114. Judah, D. J., Hayes, J. D., Yang, J.-C., Lian, L.-Y., Roberts, G. C. K., Farmer, P. B., Lamb, J. H., and Neal, G. E. (1993) A novel aldehyde reductase with activity towards a metabolite of aflatoxin B1 is expressed in rat liver during carcinogenesis and following the administration of an anti-oxidant. Biochem. J. 292, 13-17. Raj, H. G., and Lotlikar, P. D. (1984) Urinary excretion of thiol conjugates of aflatoxin B1 in rats and hamsters. Cancer Lett. 22, 125-133. Loveland, P. M., Nixon, J. E., and Bailey, G. S. (1984) Glucuronides in bile of rainbow trout (Salmo gairdneri) injected with [3H]aflatoxin B1 and the effects of dietary beta-naphthoflavone. Comp Biochem. Physiol C. 78, 13-19. Groopman, J. D., and Kensler, T. W. (1999) The light at the end of the tunnel for chemical-specific biomarkers: daylight or headlight? Carcinogenesis 20, 1-11. Shepherd, M. J., and Gilbert, J. (1984) An investigation of HPLC postcolumn iodination conditions for the enhancement of aflatoxin B1 fluorescence. Food Addit. Contam. 1, 325-335. Harris, C. C., LaVeck, G., Groopman, J., Wilson, V. L., and Mann, D. (1986) Measurement of aflatoxin B1, its metabolites, and DNA adducts by synchronous fluorescence spectrophotometry. Cancer Res. 46, 3249-3253. Essigmann, J. M., Croy, R. G., Nadzan, A. M., Busby, W. F., Jr., Reinhold: V. N., Bu¨chi, G., and Wogan, G. N. (1977) Structural identification of the major DNA adduct formed by aflatoxin B1 in vitro. Proc. Natl. Acad. Sci. U.S.A. 74, 1870-1874.

TX010063A