Glutathione-Derived Pyrroles as Potential Ex Vivo

Dec 9, 2014 - ABSTRACT: Many furan-containing compounds have been reported to be cytotoxic and/or carcinogenic agents. The toxic furans exert their ...
0 downloads 0 Views 2MB Size
Download PDF
Article pubs.acs.org/crt

N‑Acetyl Lysine/Glutathione-Derived Pyrroles as Potential Ex Vivo Biomarkers of Bioactivated Furan-Containing Compounds Chunyan Li,† Dongju Lin,‡ Huiyuan Gao,†,§ Huiming Hua,†,§ Ying Peng,*,‡ and Jiang Zheng*,§,∥ †

School of Traditional Chinese Medicine, ‡School of Pharmacy, §Key Laboratory of Structure-Based Drug Design & Discovery of Ministry of Education, Shenyang Pharmaceutical University, Shenyang, Liaoning 110016, P. R. China ∥ Center for Developmental Therapeutics, Seattle Children’s Research Institute, Division of Gastroenterology and Hepatology, Department of Pediatrics, University of Washington School of Medicine, Seattle, Washington 98101, United States ABSTRACT: Many furan-containing compounds have been reported to be cytotoxic and/or carcinogenic agents. The toxic furans exert their adverse effects possibly through metabolic activation of the furans to corresponding epoxides or/and cisenediones. Detection of the reactive metabolites is a challenge since the electrophiles often have short lives in vivo, and they are too reactive to be isolated for characterization. Seven compounds, including 2,5-dimethylfuran, R-(+)-menthofuran, R-(+)-pulegone, caesalmin C, furanodiene, diosbulbin B, and limonin, were selected for a biomarker search study. Glutathione (GSH) conjugates derived from the test compounds were detected in bile of rats, and the types of the biliary GSH conjugates observed differed from each other and were unpredictable. However, upon mixing of the bile with a solution of N-acetyl lysine (NAL), pyrroles derived from NAL and GSH were exclusively detected in all the bile samples without any exception. The formation of the pyrrole-NAL/GSH conjugates was verified by microsomal incubations and chemical synthesis. The findings facilitate the development of in vivo biomarkers of metabolic activation of furanoids.

■

INTRODUCTION Furan-containing compounds are plentiful in fruits, food, and medicinal herbals, and many synthetic medicines contain a furan function group.1,2 A number of furan compounds have been reported to be cytotoxic and/or carcinogenic agents.3−6 Furans have also been documented as mechanism-based inactivators of cytochromes P450.7−10 Covalent binding of proteins is suggested to be associated with the toxicities and enzyme inactivations of harmful furans. The proposed mechanism for their adverse effects includes the metabolic activation of the toxic furans to the corresponding epoxides or/and cis-enediones, which both readily react with tissue nucleophiles.2,11 Detection of the epoxides and enediones generated in situ is a challenge, due to the high reactivities of the electrophilic species. Chemical derivatization is often employed to detect these reactive metabolites. Currently, epoxides are trapped with thiols, such as N-acetyl cysteine (NAC) and glutathione (GSH), while the detection of cis-enediones is achieved by the reactions with N-acetyl lysine (NAL) and GSH or NAC to form chemically stable pyrrole derivatives in vitro.12,13 The established in vitro system is often composed of microsomal incubation in the presence of NAL and NAC or GSH. Unfortunately, there is not yet an in vivo system recognized for the detection of reactive metabolites of furans, i.e., cis-enediones and epoxides. Earlier, we detected an epoxide/enedial-derived GSH conjugate in the bile of rats given 4-ipomeanol.14 A biliary furan-derived GSH conjugate was reported in animals treated with R-(+)-menthofuran.15,16 Hamberger and co-workers © XXXX American Chemical Society

reported several pyrroles derived from GSH and N-acetyl cysteine in bile of animals administered furan.6 Various pyrroles derived from amino acids were detected in urine of rats after exposure to furan.17 There is a lack of conclusive metabolites in common that can represent the biomarkers in vivo or ex vivo of metabolic activation of furans. The objective of the present study was to develop a general GSH conjugate from reactive metabolites of furans in vivo. As an initial step, we proposed five types (Scheme 1) of possible GSH conjugates derived from reactive metabolites of furans, including furan-GSH-OH, furan-GSH, pyrrole-GSH (cyclic), pyrrole2GSH, and pyrrolinone-GSH. In the study, we tested seven compounds and carefully examined the proposed GSH conjugates in bile of rats treated with the individual test compounds.

■

MATERIALS AND METHODS

Chemicals and Materials. 2,5-Dimethylfuran was obtained from Shijiazhuang Lida Chemical Co., Ltd.. Furanodiene was purchased from Hainan Bikai Pharmaceutical Co., Ltd. Caesalmin C, diosbulbin B, and limonin were isolated and purified in our laboratory, and their structures were confirmed by mass spectrometry and NMR.18−20 R-(+)-Menthofuran, R-(+)-pulegone, glutathione, N-acetyl lysine, and reduced nicotinamide adenine dinucleotide phosphate (NADPH) were purchased from Sigma-Aldrich Co. (St. Louis, MO). The purity of the test compounds Special Issue: Chemical Toxicology in China Received: August 18, 2014

A

dx.doi.org/10.1021/tx500334m | Chem. Res. Toxicol. XXXX, XXX, XXX−XXX

Chemical Research in Toxicology

Article

Scheme 1. Potential GSH Conjugates Derived from Reactive Metabolites of Furans to Be Monitored

Chemical Synthesis. Each furan-containing compound (10 mg) was dissolved in acetone (200 μL) and vortexed until completely dissolved. Saturated sodium bicarbonate solution (40 μL) and Oxone (6.0 mg) were added successively to the resulting solution. The mixture was stirred for 15 min at room temperature, followed by the addition of GSH (60 mg) dissolved in 500 μL of saturated sodium bicarbonate solution. The resulting mixture was stirred at room temperature for 30 min. After centrifugation, the supernatants were harvested and evaporated to dryness under a stream of nitrogen gas at 40 °C. The resultant samples were reconstituted with 200 μL of PBS buffer containing NAL (6.0 mg). Following further stirring for 30 min at 70 °C, the reaction was analyzed by LC-MS/MS. LC-MS/MS Method. All samples were analyzed on a 4000 Q-Trap LC-MS/MS system and a hybrid triple quadrupole/LIT (linear ion trap) mass spectrometer (Applied Biosystems, Foster City, CA) equipped with a TurboIonSpray ion source. The analytes were analyzed by multiple reaction monitoring−information-dependent acquisition− enhanced product ion mode (MRM-IDA-EPI) scan mode. The data were processed using Analyst software (versions 1.6 and 1.6.1). Chromatographic separations were performed on a Hypersil BDS-C18 column (4.6 × 75 mm, 3.5 μm, Thermo, Pittsburgh, PA). The mobile phase consisted of acetonitrile with 0.1% formic acid (A) and 0.1% formic acid in water (B) with a gradient elution of 20% A at 0−2 min, 20−100% A at 2−10 min, 100−100% A at 10−12 min, and 100−20% A at 12−15 min. The flow rate was 0.8 mL/min, and the column temperature was maintained at 25 °C. LC-MS/MS analyses were performed on a 5 μL aliquot sample. An MRM scan of the GSH conjugates was run in positive/negative ion mode. The monitored ion pairs for detection of the GSH conjugates are listed in Table 1. The operation conditions were as follows: ion spray voltage (IS) = 5500/−4500 V; curtain gas (CUR) = 20 psi; collision gas (CAD), medium; turbo spray temperature (TEM) = 650 °C; nebulizer gas 1 (GS1) = 50 psi; heater gas 2 (GS2) = 50 psi; entrance potential (EP), 10/−10 V; and collision cell exit potential (CXP), 5/−5 V. Nitrogen was used as the nebulizer and auxiliary gas. IDA was used to trigger the acquisition of EPI spectra for ions exceeding 5000 cps with the exclusion of former target ions after three occurrences for 10 s. The EPI scan range of furan-containing compounds is described in Table 1. The EPI scanning conditions were as follows: scan mode = profile; step size = 0.08 Da; and scan rate = 1000 Da/s, 5.0 ms pause between mass ranges.

was >98% determined by high-performance liquid chromatography (HPLC) with a diode array detector (DAD). All organic solvents were from Fisher Scientific (Springfield, NJ). All reagents and solvents were of either analytical or HPLC grade. Experimental Animals. Male Sprague−Dawley (200−250 g) rats were obtained from Animal Center of Shenyang Pharmaceutical University, Shenyang, China. The animals were maintained on standard rat chow and tap water ad libitum in a 25 °C room with a 12 h dark/light cycle. Rat liver microsomes (RLMs) were prepared as described by our laboratory.21 Animal Studies and Sample Collection. Before administration, the experimental animals were deprived of food for 12 h. The animals were anesthetized with 10% chloral hydrate (3 mL/kg). After laparotomy, PE-10 tubing was inserted into the bile duct and fixed by placing a ligature around it to prevent dislocation during the bile sampling period. The test compounds dissolved in corn oil were individually administered intraperitoneally in rats (n = 3, for each compounds tested) at a designed dosage as listed in Table 1, and bile was collected for 2 h following dosing. Drug-blank bile samples from these rats had also been collected before the treatment. Sample Preparation. The collected bile samples were divided into two parts. One was directly subjected to LC-MS/MS analysis. The other (100 μL) was mixed with 200 μL of 500 mM N-acetyl lysine (NAL) in potassium phosphate buffer (pH 7.4). Both bile samples with and without chemical derivatization were centrifuged at 4 °C (16,000 rpm, 10 min). The supernatants were harvested and mixed with 300 μL of acetonitrile. After centrifugation, the supernatants were evaporated to dryness under a stream of nitrogen gas at 40 °C. The resulting concentrates were reconstituted with 100 μL of 50% acetonitrile in water and then centrifuged at 16,000 rpm for 10 min. The resultant supernatants (5 μL) were injected onto LC-MS/MS for analysis. Microsomal Incubations. The test compounds (200 μM) were individually incubated with rat liver microsomes (1.0 mg protein/mL) fortified with NADPH (1.0 mM) and GSH (1.0 mM) in the absence or the presence of NAL at a final concentration of 1.0 mM. The total incubation volume was 500 μL. After 90 min of incubation at 37 °C, the reactions were quenched by adding equal volumes of ice-cold acetonitrile, followed by vortex mixing. The reaction mixture was centrifuged to remove precipitated protein at 16,000 rpm for 10 min before being subjected to LC-MS/MS analysis. B

dx.doi.org/10.1021/tx500334m | Chem. Res. Toxicol. XXXX, XXX, XXX−XXX

Chemical Research in Toxicology

Article

Table 1. Dose and Mass Spectrometric Profile Information of Test Compounds

The synthetic samples were also analyzed on a hybrid quadrupoletime-of-flight (Q-TOF) mass spectrometer (Bruker microQ-TOF, Germany) with an ESI source equipped with an Agilent 1200 Series Rapid Resolution LC system. The parameters of ESI-MS were set as follows: capillary voltage (−4500 V), nebulizer gas pressure (1.2 bar), dry gas flow rate (8.0 L/min), and temperature (180 °C). LC conditions similar to those for the Q-Trap MS system described above were applied. The data were analyzed by Bruker Daltonics Data Analysis 3.4 software.

2,5-dimethylfuran, R-(+)-menthofuran, R-(+)-pulegone, caesalmin C, furanodiene, diosbulbin B, and limonin. Bile samples were collected before and after the treatment, followed by LC-MS/MS analysis. To reach the highest detection sensitivity, MRM was applied to detect the GSH conjugates excreted in bile. The selection of ion pairs acquired was based on predicted m/z values to product ions derived from neutral loss (NL) of 129 Da as a survey scan in positive ion mode. Ion m/z 272 is a characteristic fragment ion resulting from the cleavage of a GSH molecule in negative mode. With the same rationale, daughter ion m/z 272 and the product ions as ion pairs were selected for MRM scanning to detect the biliary GSH conjugates. The GSH conjugates

■

RESULTS Detection of GSH Conjugates in Bile. Rats were individually treated (i.p.) with seven test compounds, including C

dx.doi.org/10.1021/tx500334m | Chem. Res. Toxicol. XXXX, XXX, XXX−XXX

Chemical Research in Toxicology

Article

responsible for 2,5-dimethylfuran-derived furan-GSH-OH (420 → 291/418 → 272), furan-GSH (402 → 273/400 → 272), pyrrole-GSH (384 → 255/382 → 272), pyrrole-2GSH (691 → 562/689 → 272), and pyrrolinone-GSH (402 → 145/ 400 → 143, 402 → 162/400 → 160) were acquired in both positive and negative modes. Apparently, no GSH conjugates derived from 2,5-dimethylfuran were detected in bile (Table 2). R-(+)-Menthofuran (2), a furanomonoterpene, is a major component of mint plants. The furan compound has been reported to produce hepatotoxicity in rats.23 R-(+)-Menthofuran was also found to be a potent mechanism-based inactivator of cytochrome P450 2A6.8 It appears that two types of GSH conjugates, including furan-GSH-OH and furan-GSH (Scheme 2), were detected, but neither pyrrole-derived GSH conjugates nor pyrrolinone-GSH conjugates were observed in bile (Table 2). R-(+)-Pulegone (3), a monoterpene ketone, is abundant in pennyroyal. The terpene ketone is a nonfuran compound and is reportedly metabolized to R-(+)-menthofuran that executes the hepatotoxicity.24 As expected, the same GSH conjugates as those found in the bile of rats treated with R-(+)-menthofuran (Table 2) were detected. Caesalmin C (4) was isolated and purified from the seeds of Caesalpinia minax. It is a disubstituted furanoditerpene and has shown anticancer activity.25 Three types of GSH conjugates were detected in bile, and they were furan-GSH-OH, furan-GSH, and pyrrole-GSH (cyclic) (Scheme 2 and Table 2). Furanodiene (5), a trisubstituted furan, is an active ingredient of Rhizoma Curcumae, used for treatment of tumors in China.26 An exclusive GSH conjugate responsible for furan-GSH (Scheme 2) was detected in the bile of rats given furanodiene, and no other types of GSH conjugates were observed (Table 2). Diosbulbin B (6), a bioactive component of Dioscorea bulbifera L., was found to have antitumor properties,27 but it also produced liver injury.28 In contrast to the observation in the furanodiene study, no GSH conjugates corresponding to furan-GSH were detected in bile of animals treated with diosbulbin B. Instead, an exclusive GSH conjugate responsible for pyrrole-GSH (cyclic) (Scheme 2) was observed in the bile (Table 2). Limonin (7), found in the fruit of Tetradium ruticarpum (WuZhu-Yu in Chinese) widely used in China as herbal medicine,29 has been reported to be a mechanism-based inhibitor of P450 3A4.7 Three types of GSH conjugates, including furan-GSH-OH, furan-GSH, and pyrrole-GSH (cyclic) (Scheme 2), were detected in bile (Table 2). Detection of Pyrrole-NAL/GSH Conjugates in Bile after Treatment with N-Acetyl Lysine. The other portions of the bile samples collected were mixed with N-acetyl lysine (NAL) and vortexed, followed by LC-MS/MS analysis. The resulting pyrrole-NAL/GSH conjugates (Scheme 3) derived from the corresponding reactive intermediates were monitored by MRMIDA-EPI. The combination of MRM with EPI provides accurate quantitation of targeted analytes at low concentrations, along with highly sensitive information about fragment ions from selected precursor ions to confirm the targeted analytes for reliable qualitation. The transitions used for the detection of the pyrrole-NAL/ GSH conjugates (1′, Scheme 3) responsible for 2,5-dimethyfuran (1) were [M + H]+ m/z 572.0 → m/z 443.0 in positive ion mode and [M − H]− m/z 570.0 → m/z 272.0 in negative mode. A chromatographic peak at a retention time of 6.5 min was observed in both positive (Figure 1b and Table 2) and negative (data not shown) modes, and no such peak was observed in the control samples. The MS/MS of the analyte displayed the

Scheme 2. Proposed Structures of GSH Conjugates Detected in Bile of Rats Treated with the Test Compounds

corresponding to furan-GSH-OH, furan-GSH, pyrrole-GSH (cyclic), pyrrole-2GSH, and pyrrolinone-GSH (Scheme 1) were monitored in the two modes. 2,5-Dimethylfuran (1), a reported genotoxicant,22 is the smallest molecule among the test compounds tested. Ion pairs D

dx.doi.org/10.1021/tx500334m | Chem. Res. Toxicol. XXXX, XXX, XXX−XXX

Chemical Research in Toxicology

Article

Table 2. Summary of Pyrrole-NAL/GSH Conjugates and Other Proposed GSH Conjugates Derived from Test Compounds GSH-conjugates (biliary) test compounds 2,5-dimethylfuran (1) R-(+)-menthofuran (2) R-(+)-pulegone (3) caesalmin C (4) furanodiene (5) diosbulbin B (6) limonin (7)

pyrrole-NAL/GSH conjugates

furan-GSH-OHa furan-GSHb pyrrole-GSH (cyclic)c pyrrole-2GSHd × √ √ √ × × √

× √ √ √ √ × √

× × × √ × √ √

× × × × × × ×

pyrrolinoneGSHe

NAL-treated bile

synthetic

microsomal incubation

× × × × × × ×

√ √ √ √ √ √ √

√ √ √ √ √ √ √

√ √ √ √ √ √ √

a

GSH conjugates derived from furanoepoxide intermediates. bDehydration product of furan-GSH-OH conjugates. cMono-GSH conjugates resulting from intramolecular cyclization of GSH with cis-enedione intermediates. dDi-GSH conjugates derived from cis-enedione intermediates by reaction of two molecules of GSH. ePyrrolinone-GSH conjugates resulting from the condensation of cis-enediones with GSH.

Scheme 3. Proposed Structures of Pyrrole-NAL/GSH Conjugates Detected in Bile of Rats Given the Test Compounds after the Bile Samples Were Treated with NAL

(negative mode). A peak responsible for the ion pairs with a retention time of 8.3 min was observed in both positive (Figure 2b and Table 2) and negative (data not shown) modes in the bile samples obtained from animals treated with either R-(+)-pulegone or R-(+)-menthofuran. The MS/MS spectrum of 2′ obtained by MRM-EPI scanning (ion transition m/z 626.1 → 497.1) showed the indicative characteristic fragment ions associated with the

characteristic neutral loss of 129 Da from the cleavage of GSH (Figure 1e). R-(+)-Pulegone (3) is the metabolic precursor of R-(+)-menthofuran (2). The transitions employed for analysis of pyrrole-NAL/GSH conjugate 2′ (Scheme 3) derived from R-(+)-menthofuran/R-(+)-pulegone (2/3, Scheme 2) were m/z 626.1 → m/z 497.1 (positive mode) and m/z 624.1 → m/z 272.0 E

dx.doi.org/10.1021/tx500334m | Chem. Res. Toxicol. XXXX, XXX, XXX−XXX

Chemical Research in Toxicology

Article

Figure 1. Extracted ion (m/z 572.0 → 443.0) chromatograms obtained from LC-Q-Trap MS analysis of the bile of rats before (a) and after (b) treatment with 2,5-dimethylfuran. c: Extracted ion (m/z 572.0 → 443.0) chromatogram obtained from LC-Q-Trap MS analysis of rats liver microsomal incubations containing 2,5-dimethylfuran, GSH, and NAL in the presence of NADPH. d: Extracted ion (m/z 572.0 → 443.0) chromatogram obtained from LC-Q-Trap MS analysis of synthetic 1′. e: MS/MS spectrum of 1′ detected in bile of rats given 2,5-dimethylfuran.

Figure 2. Extracted ion (m/z 626.1 → 497.1) chromatograms obtained from LC-Q-Trap MS analysis of bile of rats before (a) and after (b) administration with R-(+)-menthofuran or R-(+)-pulegone. c: Extracted ion (m/z 626.1 → 497.1) chromatograms obtained from LC-Q-Trap MS analysis of rats liver microsomal incubations containing R-(+)-menthofuran or R-(+)-pulegone, GSH, and NAL in the presence of NADPH. d: Extracted ion (m/z 626.1 → 497.1) chromatograms obtained from LC-Q-Trap MS analysis of synthetic 2′. e: MS/MS spectrum of 2′ detected in bile of rats treated with R-(+)-menthofuran or R-(+)-pulegone.

cleavage of the GSH moiety (Figure 2e), such as fragment m/z 497.1 derived from the loss of the γ-glutamyl portion (−129 Da) from m/z 626.1. Pyrrole-NAL/GSH conjugate 4′ (Scheme 3) derived from caesalmin C (4) was monitored by acquiring ion pairs of m/z 950.2 → m/z 821.2 (positive mode) and m/z 948.2 → m/z 272.0 (negative mode). A peak (retention time = 8.3 min) responding

to the conjugate was detected in positive mode (Figure 3b and Table 2) but not in negative mode. The MS/MS spectrum of the analyte obtained through MRM-EPI scanning showed the indicative characteristic fragment ions resulting from the cleavage of GSH (Figure 3e). Furanodiene-derived pyrrole-NAL/GSH conjugates (5′, Scheme 3) were monitored by scanning of ion pairs of m/z F

dx.doi.org/10.1021/tx500334m | Chem. Res. Toxicol. XXXX, XXX, XXX−XXX

Chemical Research in Toxicology

Article

Figure 3. Extracted ion (m/z 950.2 → 821.2) chromatograms obtained from LC-Q-Trap MS analysis of the bile of rats before (a) and after (b) treatment with caesalmin C. c: Extracted ion (m/z 950.2 → 821.2) chromatograms obtained from LC-Q-Trap MS analysis of rat liver microsomal incubations containing caesalmin C, GSH, and NAL in the presence of NADPH. d: Extracted ion (m/z 950.2 → 821.2) chromatograms obtained from LC-Q-Trap MS analysis of synthetic 4′. e: MS/MS spectrum of 4′ detected in bile of rats treated with caesalmin C.

Figure 4. Extracted ion (m/z 692.1 → 563.1) chromatograms obtained from LC-Q-Trap MS analysis of the bile of rats before (a) and after (b) treatment with furanodiene. c: Extracted ion (m/z 692.1 → 563.1) chromatograms obtained from LC-Q-Trap MS analysis of rat liver microsomal incubations containing furanodiene, GSH, and NAL in the presence of NADPH. d: Extracted ion (m/z 692.1 → 563.1) chromatograms obtained from LC-Q-Trap MS analysis of synthetic 5′. e: MS/MS spectrum of 5′ detected in bile of rats given furanodiene.

692.1 → m/z 563.1 (positive mode) and m/z 690.1 → m/z 272.0 (negative mode). A peak (retention time = 7.2 min) responsible for conjugate 5′ was found in positive mode (Figure 4b and Table 2) but not in negative mode. The MS/MS spectrum of the conjugate showed the major fragments from the cleavage of the GSH portion (Figure 4e). Ion pairs of m/z 820.1 → m/z 691.1 in positive ion mode and m/z 818.1 → m/z 272.0 in negative mode were acquired to monitor the pyrrole-NAL/GSH conjugates derived from diosbulbin B (6). A peak (retention time = 6.7 min) related to

the pyrrole-NAL/GSH conjugate (6′) was detected in positive mode (Figure 5b and Table 2) but not in negative mode. The MS/MS spectrum of the analyte elicited the same fragments as we reported recently (Figure 5e).30 Pyrrole-NAL/GSH conjugates 7′ (Scheme 3) derived from limonin (7) were monitored by scanning of ion pairs of m/z 946.0 → m/z 817.0 (positive mode) and m/z 944.0 → m/z 272.0 (negative mode). A peak with a retention time at 6.1 min responsible for conjugate 7′ was found in both positive (Figure 6b and Table 2) and negative (data not shown) modes. As expected, the G

dx.doi.org/10.1021/tx500334m | Chem. Res. Toxicol. XXXX, XXX, XXX−XXX

Chemical Research in Toxicology

Article

Figure 5. Extracted ion (m/z 820.1 → 691.1) chromatograms obtained from LC-Q-Trap MS analysis of bile of rats before (a) and after (b) treatment with diosbulbin B. c: Extracted ion (m/z 820.1 → 691.1) chromatograms obtained from LC-Q-Trap MS analysis of rat liver microsomal incubations containing diosbulbin B, GSH, and NAL in the presence of NADPH. d: Extracted ion (m/z 820.1 → 691.1) chromatograms obtained from LC-Q-Trap MS analysis of synthetic 6′. e: MS/MS spectrum of 6′ detected in bile of rats administered with diosbulbin B.

Figure 6. Extracted ion (m/z 946.0 → 817.0) chromatograms obtained from LC-Q-Trap MS analysis of the bile of rats before (a) and after (b) treatment with limonin. c: Extracted ion (m/z 946.0 → 817.0) chromatograms obtained from LC-Q-Trap MS analysis of rat liver microsomal incubations containing limonin, GSH, and NAL in the presence of NADPH. d: Extracted ion (m/z 946.0 → 817.0) chromatograms obtained from LC-Q-Trap MS analysis of synthetic 7′. e: MS/MS spectrum of 7′ detected in bile of rats administered with limonin.

indicative characteristic fragments resulting from the cleavage of the GSH moiety were observed in the MS/MS spectrum (Figure 6e). Detection of Pyrrole-NAL/GSH Conjugates in Microsomal Incubations Supplemented with GSH and NAL. The test compounds above were individually incubated in rat liver microsomes supplemented with GSH and NAL in the presence of NADPH, followed by LC-MS/MS analysis. The pyrrole-NAL/GSH conjugates detected in NAL-treated bile

of animals given the test compounds were all found in the corresponding microsomal incubations (Figures 1c−6c and Table 2), based on their retention time and mass spectrometric behaviors. Detection of Pyrrole-NAL/GSH Conjugates Generated from Chemical Oxidation of Furans Trapped with GSH and NAL. The furan-containing compounds above were oxidized by Oxone in acetone, and the resulting products were H

dx.doi.org/10.1021/tx500334m | Chem. Res. Toxicol. XXXX, XXX, XXX−XXX

Chemical Research in Toxicology

Article

Table 3. Summary of High Resolution Mass Spectrum Profiles of Synthetic Pyrrole-NAL/GSH Conjugates [M + H]+

[M + H]+

error

compd

formula

calculated

detected

ppm

mDa

sigma

1′ 2′ 4′ 5′ 6′ 7′

C24H37O9N5S C28H43O9N5S C44H63O16N5S C33H49O9N5S C37H49O14N5S C44H59O16N5S

572.2385 626.2854 950.4063 692.3324 820.3069 946.3750

572.2384 626.2846 950.4014 692.3357 820.3038 946.3756

0.21 −0.07 −0.88 −4.87 3.60 −0.60

0.12 −0.05 −0.85 −3.37 4.50 −0.57

0.015 0.018 0.025 0.021 0.030 0.025

precursor of furan−GSH type conjugates, were detected in bile of rats administered furanodiene. Given together, most furans tested were metabolized to GSH conjugates, but the type(s) of GSH conjugates formed and excreted in bile is unfortunately not predictable. Primary amines are the common nucleophiles employed to trap cis-enediones with the assistance of GSH, and the trapping reactions produce chemically stable pyrrole-GSH conjugates (Scheme 3). To seek predictable metabolites as ex vivo biomarkers of metabolic activation of furans, we simply mixed the bile samples collected with NAL. Interestingly, pyrrole-NAL/ GSH conjugates derived from all seven test compounds were detected by LC-MS/MS. To verify the formation of the pyrroleNAL/GSH conjugates, we biosynthesized the conjugates by incubating the furans in rat liver microsomes fortified with GSH and NAL. As expected, pyrrole-NAL/GSH conjugates derived from the seven test compounds were detected by LC-MS/MS. They showed identical chromatographic and mass spectrometric behaviors as those of the conjugates detected in bile after treatment with NAL. Furthermore, we chemically synthesized the conjugates by oxidation of the furans with Oxone in acetone, followed by reaction with GSH and NAL. Again, the retention times and MS/MS spectra of the resulting pyrrole-NAL/GSH conjugates were the same as that of the conjugates generated ex vivo and in the microsomal reactions. The observed consistencies led us to propose the pyrrole-NAL/GSH conjugates as the candidates of ex vivo biomarkers of bioactivation of furans. Among the five types of GSH conjugates proposed (Scheme 1), only the furan-GSH-OH conjugate can react with NAL to form pyrroles. The remaining four are condensation products and are stable aromatic compounds that are unlikely reactive to NAL at room temperature. Additional possible intermediates that can react with NAL to produce pyrrole derivatives are cis-enediones and their GSH conjugates (Scheme 1). We failed to detect the cisenedione intermediates resulting from the seven test compounds (data not shown) in the bile samples. The exact identities of the biliary intermediates reactive to NAL to produce the pyrroleNAL/GSH conjugates remain unknown. In conclusion, furans are metabolized to electrophilic intermediates that sequentially react with GSH and are excreted in bile. The identities of the GSH conjugates in bile vary and are unpredictable. The treatment of bile samples containing reactive intermediates of furans with NAL exclusively produced pyrroleNAL/GSH conjugates, which may be considered as ex vivo biomarkers of metabolic activation of furan-containing compounds.

reacted with GSH and NAL. As expected, the chemical reactions produced the same pyrrole-NAL/GSH conjugates derived from the corresponding furans as those detected in the bile samples and microsomal incubation mixtures (Figures 1d−6d and Table 2). The synthetic pyrrole-NAL/GSH conjugates were all analyzed by the Q-TOF MS system. As expected, the measured mass in the MS spectra agreed with the theoretical mass within 5 ppm, based on the predicted formula (Table 3). Additionally, the sigma values given by the workstation that characterize the degree of the isotope matching were found to be