![]() |
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Department of Chemistry, Hong Kong Baptist University, Kowloon, Hong Kong SAR, China
(Received November 19, 2006; Accepted March 6, 2007)
| Abstract |
|---|
|
|
|---|
|
Lactam formation was the major in vivo and in vitro metabolic pathway of AAs after nitroreduction (Krumbiegel et al., 1987
; Chan et al., 2006
). With the support of spectroscopic data from in vitro incubations, Pfau et al. (1990b
) postulated that AAs were metabolized to N-hydroxyaristolactam, which might be further reduced to aristolactams or rearranged to 7-hydroxyaristolactams via the Bamberger rearrangement. The 7-hydroxyaristolactams could also be produced from the 7-hydroxylation of the aristolactams. This article reports results of the investigation of in vivo metabolic pathways by analyzing a series of positional isomers of hydroxyaristolactams. In addition to the LC-MS and MS/MS analyses, theoretical computations on the hydroxylated metabolites of aristolactams were performed to augment the isomer assignment. In addition, several new phase II metabolites of AAs were identified in the rat urine, for the first time.
| Materials and Methods |
|---|
|
|
|---|
(Waters). Animal Experiment and Sample Preparation. Male Sprague-Dawley rats (n = 4) weighing 200 to 220 g were obtained from the Laboratory Animal Service Centre, Chinese University of Hong Kong. The animals were kept in a controlled room with constant temperature (23°C ± 1) and artificial dark/light cycles. The rats were fasted overnight, but water was given ad libitum before a single oral dose of 6 mg of AAs in 0.5% NaHCO3 solution. The dosed rats were then housed in a metabolic cage, which allowed the separated collection of urine and feces. The rat urine from 0 to 24 h after the oral administration was collected and kept at -20°C.
Urine sample after thawing was filtered through a cellulose filter disc (25 mm, 0.2 µm) before the SPE. Five milliliters of urine was loaded onto a preconditioned SPE cartridge and washed sequentially with 3 ml of water, 2 ml of water/methanol (4:1), and eluted with 3 ml of methanol. The methanolic eluent was collected and evaporated to dryness under a stream of nitrogen. The residual was dissolved in 50 µl of methanol and then centrifuged at 13,000 rpm for 3 min before the LC-MS analysis.
Calibration plots for the reductive metabolites of AAs were obtained from LC-MS determination of aristolactam I and aristolactam II at concentrations of 0.2, 1, 2.5, 5, and 10 µg/ml in blank urine matrix. Linear regression analysis gave calibration curves that were used to calculate the amount of aristolactam I and aristolactam II excreted in rat urine. To determine the recovery of the SPE process for aristolactams, aristolactam I and aristolactam II were spiked to blank urine (2 ml) at a final concentration of 0.5 µg/ml (n = 5), processed, and measured in the same way as the samples from the AA-dosed rats.
LC-MS Analysis. HPLC experiments were performed on an HP1100 HPLC system equipped with an autosampler and a micropump (Agilent Technologies, Palo Alto, CA). A reverse-phase column (Lunar C18, 150 mm x 2.0 mm, 5 µm; Phenomenex, Torrance, CA) was used to separate AAs and their metabolites. The compartment of the autosampler was set at 4°C. The sample injection volume was 8 µl. The mobile phase system consisted of two components, component I being 0.2% acetic acid and 5 mM ammonium acetate (A), and component II being 0.2% acetic acid in methanol (B). The solvent gradient started from 20% B and was held for 5 min, then programmed to 80% B in 5 min, and held for another 15 min before reconditioning, at a flow rate of 200 µl/min. The effluent of the first 5 min from the LC was diverted to waste.
Electrospray ionization mass spectrometry (ESI-MS) and MS/MS analyses were conducted on a Qq-TOF tandem mass spectrometer (API Q-STAR Pulsar i; Applied Biosystems, Foster City, CA). TurboIonspray parameters for positive ion mode ESI-MS were optimized as follows: ionspray voltage 5300 V, declustering potential I 20 V, declustering potential II 15 V, focusing potential 70 V. The mass range was from m/z 200 to 800. Depending on the compound stability, the collision energy for product ion scans of AAs and their metabolites varied from 15 eV to 45 eV for the MS/MS experiments. The ion source gas I, gas II, curtain gas, and collisionally activated gas were set at 30, 15, 30 and 3 psi, respectively. The temperature of ion source gas II was set at 350°C.
|
| Results |
|---|
|
|
|---|
0.999) was obtained for both analytes. The recoveries of aristolactam I and aristolactam II at the level of 0.5 µg/ml (n = 5) were 93.5 ± 4.0% and 93.2 ± 4.1% (mean ± S.D.), respectively. In the rat urine samples collected after the oral administration of AAs (n = 4), 0.04 ± 0.01 µg/ml aristolactam I and 0.10 ± 0.03 µg/ml aristolactam II were detected. The lower concentration of aristolactam I than of aristolactam II might be due to the lower availability of AAI. It has been reported that AAI could be easily converted to aristolochic acid Ia via demethylation (Fig. 1) (Schmeiser et al., 1986Identification of AA Metabolites. Decarboxylated metabolites of AAs. New metabolites from decarboxylation of AAs (Fig. 1) were detected in the rat urine samples, for the first time. Figure 2 shows the extracted ion chromatogram of the decarboxylated metabolite of AAI (M1) (Fig. 2A) and its MS/MS spectrum (Fig. 2B). The MS/MS analyses of the decarboxylated metabolite of AAI resulted in the dissociate loss of -CH3 moiety, producing the fragment ion at m/z 300 as the base peak. The characteristic loss of -NO2 and -OCH3 moieties were also observed from the fragment ions at m/z 268 and m/z 283, respectively.
Phase II metabolites of AAIa. Three new phase II metabolites of aristolochic acid Ia (AAIa, M2) formed by acetylation (M3), glucuronidation (M4), and sulfation (M5) (Fig. 1) were detected in rat urine. The measured HR-MS data of the [M + NH4]+ ion of these metabolites (m/z 387.0802, 425.0259, and 521.1089) matched the theoretical mass of the corresponding elemental composition (m/z 387.0828, 425.0291, and 521.1044), with a mass difference of 6.8 ppm, 7.5 ppm, and 8.7 ppm, respectively. Figure 3 shows the extracted ion chromatograms of the detected phase II metabolites. The MS/MS analysis revealed the characteristic fragmentation at the conjugation linkage. The dissociate loss of 42 Da (Fig. 4A), 176 Da (Fig. 4B), and 80 Da (Fig. 4C) from the intact molecules signified the fragmentation loss of acetyl, glucuronide, and sulfate groups, respectively.
|
|
Hydroxylated metabolites of aristolactams. Three hydroxylated metabolites of aristolactam II (M6, M7, and M8) and two hydroxylated metabolites of aristolactam I (M9 and M10) were detected in the rat urine (Fig. 5). The identification was based on the observation of [M + H]+ ion and [2M + H]+ dimer ion. The MS/MS spectra of the hydroxylated metabolites of aristolactam II (M6, M7, and M8) were acquired from the protonated molecular ion at m/z 280 under the identical collision energy (40 eV). Similar fragmentation patterns were obtained (Fig. 6), except that fewer fragment ions were observed for M7. The common fragmentation included the water loss and the loss of CH2O2,CH2O2 + CO moieties. The common ion peak at m/z 178 represents the phenanthrene moiety from the aristolactam isomers. A fragment ion at m/z 222 was also observed for M6 and M8, resulting from the dissociative loss of both -OCH2 and -CO moieties (Fig. 6, A and C).
|
|
Investigation of Isobaric Hydroxylated Metabolites of Aristolactams by Theoretical Calculations. The molecular structure, relative energy, and dipole moment of the three hydroxylated metabolites of aristolactam II (M6, M7, and M8) in aqueous and gas phases are summarized in Table 1. Among the three hydroxylated metabolites of aristolactam II postulated in previous mechanism studies of AAs (Pfau et al., 1990b
; Arlt et al., 2002
), aristolactam Ia (M6) has the largest dipole moment (7.5 debye), whereas 7-hydroxyaristolactam had the smallest (4.6 debye). All of them showed similar orientation of the electric dipole moment. N-Hydroxyaristolactam II had the highest relative energy (38 kJ/mol), whereas aristolactam Ia had the smallest (0 kJ/mol). The absence of imaginary frequencies in our theoretical calculations verified that all the structures were at their true minima at the specified computational level.
|
| Discussion |
|---|
|
|
|---|
In this study, two major metabolites from the nitroreduction of AAs (aristolactam I and aristolactam II) were quantified in rat urine. The quantitative analysis of aristolactams is of great interest not only because the lactam formation was the major in vivo and in vitro metabolic pathway of AAs after nitroreduction (Krumbiegel et al., 1987
; Chan et al., 2006
), but also because the intermediates (aristolactam-nitriumion) of the reduction process play a key role on biotransformation to the ultimate carcinogen of AAs (Schmeiser et al., 1986
; Pfau et al., 1990b
, 1991
). The preliminary quantitative analysis indicated that a submicrogram per milliliter level of aristolactam I and aristolactam II existed in the rat urine samples collected after the oral administration of AAs. The concentration of aristolactam II was approximately 2 times higher than that of aristolactam I, probably because the parent compound of aristolactam I (i.e., AAI) could be easily demethylated to aristolochic acid Ia (Fig. 1). Further study would focus on the determination of mass balance of AAs and the lactam metabolites after the in vivo experiments. The quantitative analysis might provide important information on carcinogenesis of AAs because the aristolactams are associated with the formation of DNA adducts (Arlt et al., 2002
).
Several in vivo metabolites were identified in the rat urine samples for the first time. In addition to the observation of a decarboxylative metabolite of AAs, three new conjugated metabolites formed by acetylation, glucuronidation, and sulfation of AAIa were identified. A series of positional isomers was detected as hydroxyaristolactam I and hydroxyaristolactam II. All metabolites were characterized from the HR-MS and MS/MS analyses.
Being peri-substituted nitrophenanthrene carboxylic acids, AAs are weak mutagens in Salmonella typhimurium strain TA100 compared with other nitroaromatic compounds (Purohit and Basu, 2000
). It has been reported that steric repulsion between the carboxyl and nitro groups caused the nitro group to deviate from coplanarity with reference to the aromatic rings (Pfau et al., 1990a
). Nitroaromatics with perpendicularly orientated nitro group have been proven to be poor substrates for nitroreductases in mutagenicity assays (Fu et al., 1985
), which might be the reason for the observed low mutagenicity of AAs. However, the decarboxylation of AAs (M1) alleviates the steric repulsion experienced by the nitro group, making it less deviated from coplanarity with reference to the aromatic rings, which might enhance their mutagenicity.
Conjugation is one of the most important routes for the elimination of xenobiotics from biological systems. The conventional methodology for the analysis of conjugated metabolites comprises a chemical or enzymatic hydrolysis step. Combining the high separation efficiency of HPLC and the soft ionization technology of ESI-MS, LC-MS provides a sensitive and efficient way for direct analysis of phase II metabolites in the complex sample matrix. In our previous study, three glucuronide conjugates of aristolactams were detected and characterized (Chan et al., 2006
).
An extensive O-demethylation has been observed for AAI in both in vitro and in vivo studies, producing AAIa as the metabolite (Schmeiser et al., 1986
; Krumbiegel et al., 1987
; Chan et al., 2006
). However, the metabolic fate of AAIa is not well understood apart from the lactam formation, which produces aristolactam Ia. Herein, we report the first identification of three conjugated metabolites formed by acetylation (M3), glucuronidation (M4), and sulfation (M5) of AAIa in rat urine after the oral administration of AAs. The phase II metabolites showed characteristic fragmentation at the conjugation linkage in MS/MS experiments. The loss of 42 Da (Fig. 4A), 176 Da (Fig. 4B), and 80 Da (Fig. 4C) from the intact molecules signified the neutral loss of acetyl, glucuronide, and sulfate groups, respectively. The assignments made for the conjugates were further supported by the HR-MS data. The water and carbon dioxide loss after the cleavage of the conjugated moieties in LC-MS/MS analysis was consistent with the data obtained when AAs and AAIa were analyzed (Chan et al., 2006
).
The mechanism underlying N-hydroxylation of arylamines and the nitroreduction of nitropolycyclic hydrocarbons to aryl hydroxyamines has been considered common. Similar to other nitroaromatics, nitroreduction is the crucial step in metabolic activation of AAs to their ultimate carcinogen. Being the first example of intramolecular acetylation upon metabolic activation, the mechanism for the in vitro reductive conversion of AAI to aristolactam I was postulated (Pfau et al., 1990b
). It was proposed that AAI was first metabolized to its N-hydroxyaristolactam I, which was either reduced to aristolactam I or underwent the Bamberger rearrangement to give 7-hydroxyaristolactam I. However, the intermediate N-hydroxyaristolactam I had not been identified, probably because of the lack of analytical sensitivity.
In this study, three hydroxylated metabolites of aristolactam II (M6-M8) (Fig. 5A) and two hydroxylated metabolites of aristolactam I (M9, M10) (Fig. 5B) were detected in rat urine. Based on the previous metabolism study of AAs (Pfau et al., 1990b
; Arlt et al., 2002
), the hydroxylated metabolites from aristolactam II (mol. wt. 279) were identified as aristolactam Ia, N-hydroxyaristolactam II, and 7-hydroxyaristolactam II, whereas those from aristolactam I (mol. wt. 309) were identified as N-hydroxyaristolactam I and 7-hydroxyaristolactam I. However, distinguishing the positional isomers was a challenge because the corresponding authentic standards were not available. Theoretical computation was therefore performed to help the assignment of the hydroxyaristolactams in the LC-MS chromatograms.
The reverse-phase liquid chromatographic retention time generally depends on the polarity and structural parameters (e.g., size/shape) of the eluted compounds (Kaliszan et al., 1986
). Of a congeneric series of analytes, the more polar compounds generally have less retention under the identical chromatographic conditions. The chemical polarity can be expressed in terms of electronic properties such as dipole moment and polar surface area. Compounds with larger dipole moment tend to have shorter retention times (Kaliszan et al., 1986
; Niessen, 1999
). For the three hydroxylated metabolites of aristolactam II (M6-M8), the calculated dipole moments of aristolactam Ia, N-hydroxyaristolactam II, and 7-hydroxyaristolactam II in the aqueous phase were 7.5, 6.1, and 4.6 debye, respectively, with a similar orientation (Table 1), indicating that aristolactam Ia might have the highest polarity. Accordingly, the reversed-phase liquid chromatographic peaks at retention time of 21.15 min, 22.77 min, and 24.02 min were assigned to be aristolactam Ia (M6), N-hydroxyaristolactam II (M7), and 7-hydroxyaristolactam II (M8), respectively.
The isomer identification of the three hydroxylated metabolites of aristolactam II (M6-M8) was further confirmed by the comparison of the MS/MS results (Fig. 6) together with the relative energy values obtained from the computation (Table 1). The MS/MS data under the identical collision energy were in agreement with the molecular stability represented by the calculated relative energy. The most stable, aristolactam Ia, with the lowest relative energy had the most intensive parent ion peak at m/z 280 (Fig. 6A), whereas the least stable, N-hydroxyaristolactam II, showed intensive fragmentation (Fig. 6B). Moreover, the MS/MS fragmentations of aristolactam Ia (M6) (Fig. 6A) and 7-hydroxyaristolactam II (M8) (Fig. 6C) were similar but far more intensive than that of N-hydroxyaristolactam II (M7) (Fig. 6B), which agreed well with the proposal that M6 and M8 were the same type of (phenolic) compounds, whereas M7 was different (N-hydroxy).
|
In summary, the in vivo metabolic pathway underlying the nitroreduction of AAs to the corresponding aristolactams was proposed. The animal study revealed extensive phase II metabolism of AAIa, which was the demethylated metabolite of AAI. In addition to the identification of the metabolites from decarboxylation of AAs, a series of positional isomers was detected as hydroxyaristolactams. Interpretation of LC-MS and MS/MS results combined with theoretical calculations was used for the identification of the hydroxyaristolactam metabolites. The results from this work may represent one step forward in understanding the metabolism and bioactivation of AAs.
| Footnotes |
|---|
We thank the Faculty Research Grant of the Hong Kong Baptist University and the Research Grant Council, University Grants Committee of Hong Kong (HKBU2154/04M) for their financial support of this study.
ABBREVIATIONS: AA, aristolochic acid; AAIa, aristolochic acid Ia; SPE, solid-phase extraction; HPLC, high performance liquid chromatography; ESI-MS, electrospray ionization-mass spectrometry; LC-MS, liquid chromatography-mass spectrometry; MS/MS, tandem mass spectrometry; HR-MS, high-resolution mass spectrometry.
Address correspondence to: Dr. Zongwei Cai, Department of Chemistry, Hong Kong Baptist University, Kowloon, Hong Kong. E-mail: zwcai{at}hkbu.edu.hk
| References |
|---|
|
|
|---|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||