DMD Celsis microsomes equal better data

Home Help [Feedback] [For Subscribers] [Archive] [Search] [Contents]
QUICK SEARCH:   [advanced]


     

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Chen, L.-J.
Right arrow Articles by Burka, L. T.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Chen, L.-J.
Right arrow Articles by Burka, L. T.

Vol. 29, Issue 12, 1567-1577, December 2001


Metabolism of (R)-(+)-Pulegone in F344 Rats

Ling-Jen Chen, Edward H. Lebetkin, and Leo T. Burka

Laboratory of Pharmacology and Chemistry, National Institute of Environmental Health Sciences, Research Triangle Park, North Carolina


    Abstract
Top
Abstract
Introduction
Materials and Methods
Results
Discussion
References

(R)-(+)-Pulegone, a monoterpene ketone, is a major component of pennyroyal oil. Ingestion of high doses of pennyroyal oil has caused severe toxicity and occasionally death. Studies have shown that metabolites of pulegone were responsible for the toxicity. Previous metabolism studies have used high, near lethal doses and isolation and analysis techniques that may cause degradation of some metabolites. To clarify these issues and further explore the metabolic pathways, a study of 14C-labeled pulegone in F344 rats at doses from 0.8 to 80 mg/kg has been conducted. High-pressure liquid chromatography (HPLC) analysis of the collected urine showed the metabolism of pulegone to be extensive and complex. Fourteen metabolites were isolated by HPLC and characterized by NMR, UV, and mass spectroscopy. The results demonstrated that pulegone was metabolized by three major pathways: 1) hydroxylation to give monohydroxylated pulegones, followed by glucuronidation or further metabolism; 2) reduction of the carbon-carbon double bond to give diastereomeric menthone/isomenthone, followed by hydroxylation and glucuronidation; and 3) Michael addition of glutathione to pulegone, followed by further metabolism to give diastereomeric 8-(N-acetylcystein-S-yl)menthone/isomenthone. This 1,4-addition not only took place in vivo but also in vitro under catalysis of glutathione S-transferase or mild base. Several hydroxylated products of the two mercapturic acids were also observed. Contrary to the previous study, all but one of the major metabolites characterized in the present study are phase II metabolites, and most of the metabolites in free forms are structurally different from those previously identified phase I metabolites.


    Introduction
Top
Abstract
Introduction
Materials and Methods
Results
Discussion
References

Pulegone is a monoterpene ketone present in essential oils from many mint species (Grundschober, 1979). Two mints, Hedeoma pulegoides and Mentha pulegium, both commonly called pennyroyal, contain essential oils, which are chiefly pulegone (Budavari, 1996). Pennyroyal oil has been used as a flavoring agent in foods and beverages, as well as a component in fragrance products and flea repellents (Hall and Oser, 1965; Tyler, 1993). Pennyroyal herb has also been used for the purpose of inducing menstruation and abortion (Tyler, 1993). However, high doses of pennyroyal oil have sometimes been taken in attempted abortion and have resulted in central nervous system toxicity, gastritis, hepatic and renal failure, pulmonary toxicity, and death (Anderson et al., 1996). Pulegone was found to constitute greater than 80% of the terpenes in pennyroyal oils that were obtained from health food stores and was found to be both hepatotoxic and pneumotoxic in mice (Gordon et al., 1982).

Gordon et al. (1987) have shown that metabolites of pulegone were responsible for its toxicity and have implicated menthofuran as a proximate toxin. Metabolism studies in rats treated with relatively high doses of pulegone have demonstrated complex metabolic pathways. About 14 phase I pulegone metabolites were fully characterized after acid treatment and ether extraction of urine samples from rats dosed orally with four daily 250- or 400-mg/kg doses (Moorthy et al., 1989; Madyastha and Raj, 1993). No quantitative data were provided in these studies. Ten phase II biliary metabolites of pulegone were partially characterized by tandem mass spectrometry after the rats were treated with a single i.p. dose of 250 mg/kg (Thomassen et al., 1991). These metabolites accounted for only 3% of total radioactivity excreted in bile, and their structures could not be established solely from mass spectral analysis. The dose (250 mg/kg) used in these metabolism studies has been shown to result in centrilobular hepatic necrosis and widespread alkylation of tissue proteins (McClanahan et al., 1989).

Pulegone has been nominated to the National Toxicology Program (NTP) for toxicity and carcinogenicity studies based on the potential for human exposure and the absence of carcinogenicity data. We were concerned that the high doses used in previous studies might alter the metabolite profile, and we wanted to characterize the expected phase II metabolites excreted in urine. To explore the metabolic pathway in detail and to quantitate the metabolites, we performed metabolism studies at lower doses and minimized chemical treatment to harvest the metabolites. A metabolism study of 14C-labeled pulegone in F344 rats at single oral (0.8, 8, and 80 mg/kg) or i.v. (0.8 mg/kg) doses or four daily oral doses (80 mg/kg/day) has been conducted. Fourteen major urinary metabolites have been characterized.



    Materials and Methods
Top
Abstract
Introduction
Materials and Methods
Results
Discussion
References

Chemicals. Methylethylidene-14C-(R)-(+)-pulegone (specific activity, 60.6 mCi/mmol; radiochemical purity, 97.4%) was obtained from Wizard Laboratories, Inc. (West Sacramento, CA). Unlabeled pulegone (98% pure), trifluoroacetic acid (TFA1), and tetrabutylammonium bromide (99% pure) were purchased from Aldrich Chemical Co. (Milwaukee, WI). Menthofuran (95% pure) was obtained from Acro Organics (Pittsburgh, PA). Piperitone (92% pure) was obtained from Pfaltz & Bauer, Inc. (Waterbury, CT). Reduced glutathione (GSH) (97% pure) was purchased from Fluka BioChemika (Milwaukee, WI). Glutathione S-transferase (GST) (75% pure) from rat liver, glucose 6-phosphate, glucose-6-phosphate dehydrogenase, and NADP+ were purchased from Sigma Chemical Co. (St. Louis, MO). 2'-Hydroxy-4'-methylacetophenone (approximately 75% pure) was purchased from Indofine Chemical Co., Inc. (Somerville, NJ).

Spectra. 1H NMR spectra were acquired on a Nicolet NT-360 NB (Thermo Nicolet, Madison, WI) or a Varian 300 MHz NMR spectrometer (Varian, Palo Alto, CA). The chemical shifts are reported in parts per million relative to solvents. Electrospray ionization (ESI) mass spectra were obtained on a Finnigan/ThermoQuest LCQ DUO ion trap mass spectrometer (Thermo Finnigan, San Jose, CA). Tandem mass spectra (MS/MS) were produced by collision-induced dissociation of the selected parent ions with the He gas present in the mass analyzer. Most samples were dissolved in MeOH-H2O (1:1) for direct infusion analysis (2.5 µl/min) unless otherwise indicated. The heated capillary was maintained at 200°C and the source voltage at 4.5 kV. The GC/MS instrument used for analyzing 10-hydroxypulegone was a Finnigan/ThermoQuest TraceMS, equipped with a Trace 2000 GC. Injections were made using a J&W Scientific (Folsom, CA) cold on-column injector. Samples were injected with an SGE 10-µl syringe (SGE, Inc., Austin, TX) fitted with a fused silica needle. GC conditions were: carrier gas, He; oven temperature program, held for 10 min at 100°C, then increased at 10°C/min to 300°C, and held at 300°C for 15 min. The retention time of 10-hydroxypulegone was 14.4 min.

HPLC. HPLC analyses were carried out with one of two systems. System A consisted of two Waters model 510 pumps (Milford, MA), an automated gradient controller, and a model 481 UV detector. System B consisted of a Beckman System Gold model 126 solvent module pump and a model 168 photodiode array detector, controlled by Nouveau software (Beckman Coulter, Inc., Fullerton, CA). System B was connected to an IN\US (Tampa, FL) beta -Ram flow detector equipped with a liquid cell (500 µl) for radiochemical detection. Liquid scintillation fluid Ultima-Flo M (Packard Instrument Company, Meriden, CT) was delivered in 3:1 scintillation/elute ratio. A Metachem (Torrance, CA) Inertsil C18 5-µm column (4.6 × 250 mm) was used for all studies unless otherwise indicated.

Several HPLC methods were used. All flow rates were 1.5 ml/min. The solvents used were as follows: solvent A: 0.1% TFA in H2O; solvent B: 100% CH3CN; solvent C: 20 mM HCO2NH4 in H2O; solvent D: 20 mM potassium phosphate buffer, pH 7.0, in H2O; solvent E: 20 mM NH4OAc in H2O; solvent F: H2O.

Method 1 used a linear gradient system from 100% A to 50% A and 50% B over 28 min, then back to 100% A over 2 min. Method 2 used a linear gradient system from 100% A to 0% A and 100% B over 28 min, then back to 100% A over 2 min. Method 3 used a linear gradient system from 85% A and 15% B to 25% A and 75% B over 40 min, then back to 85% A and 15% B over 5 min. Method 4 used an isocratic system 85% A and 15% B. Method 5 used an isocratic system 80% A and 20% B. Method 6 used an isocratic system 75% A and 25% B. Method 7 used an isocratic system 70% A and 30% B. Method 8 used an isocratic system 65% A and 35% B. Method 9 used an isocratic system 88% C and 12% B. Method 10 used an isocratic system 82% C and 18% B. Method 11 used an isocratic system 80% C and 20% B. Method 12 used an isocratic system 90% D and 10% B. Method 13 used an isocratic system 85% D and 15% B. Method 14 used a linear gradient system from 100% E to 0% E and 100% B over 43 min, then back to 100% E over 2 min. Method 15 used an isocratic system 85% E and 15% B. Method 16 used an isocratic system 75% F and 25% B.

Animal Dosing and Sample Collection. Male and female F344 rats were obtained from Taconic Farms, Inc. (Germantown, NY). Female rats were 12 to 13 weeks old and weighed 160 to 195 g. Male rats were 11 weeks old and weighed 225 to 333 g.

Single and multiple doses of pulegone were administered by gavage or i.v. to rats (n = 4-8/treatment group), as described in Table 1. Oral doses were administered at 40 µCi/kg in a dose volume of 4 ml/kg in corn oil. The i.v. dose (0.8 mg/kg) was administered at 40 µCi/kg in a dose volume of 1 ml/kg in water (80%), Emulphor (10%), and ethanol (10%). Animals were housed individually in plastic metabolism cages and provided with food (NIH no. 31) and distilled water for ad libitum consumption. Urine was collected at room temperature 4, 8, 12, 24, 48, and 72 h after dosing. Urine samples were stored at -20°C and centrifuged at low gravity before analysis by HPLC. The Institutional Animal Care and Use Committee approved all animal procedures.

                              
View this table:
[in this window]
[in a new window]
 

TABLE 1
Pulegone-derived radiolabeled peaks in cumulative 24-h rat urine (mean ± S.D. of four to eight rats)

All doses were given singly and by mouth unless otherwise noted.

Syntheses.

Pulegol Pulegol was synthesized according to a known method (Moorthy et al., 1989). Its NMR data are consistent with the literature report.

Microsomal incubation of pulegone to prepare 10-hydroxypulegone. F344 rat liver microsomal fractions were prepared by the method of Guengerich (1982). A large scale (50 ml) incubation of [14C]pulegone (1 mM; specific activity, 0.02 µCi/µmol) with rat liver microsomes (2 mg of protein/ml) was conducted in 0.1 M potassium phosphate buffer, pH 7.4, in the presence of 25 mM glucose 6-phosphate, glucose-6-phosphate dehydrogenase (2 units/ml), 4 mM NADP+, 3 mM MgCl2, and 1 mM EDTA. Pulegone was added as a dimethyl sulfoxide solution (0.5 ml). After a 30-min incubation at 37°C in capped vials, reactions were terminated by addition of 0.3 N Ba(OH)2 (5 ml) and 0.3 N ZnSO4 (5 ml). Following centrifugation, the supernatant was filtered through an Acrodisc (Gelman, Ann Arbor, MI; 0.45 µm, 13 mm), and the filtrate was extracted with ether (2 × 50 ml). The combined ether layers were evaporated, and the residue was redissolved in H2O (5 ml) for HPLC separation (method 16, system A, 250 nm). Spectral properties of the major radiolabeled metabolite (10.9 min): 1H NMR (300 MHz, D2O): delta  4.17 (AB quartet, J = 12.6 Hz, 2H, 10-CH2), 2.82 (dt, J = 15.1, 4.5 Hz, 1H, 3-CHeq), 2.56 (ddd, J = 14.8, 4.4, 1.9 Hz, 1H, 6-CHeq), 2.29 (br. t, J = 12.6 Hz, 1H, 3-CHax), 2.22 (dd, J = 14.6, 10.7 Hz, 1H, 6-CHax), 2.13 to 2.01 (m, 1H, 5-CHax), 1.97 to 1.89 (m, 1H, 4-CHeq), 1.86 (d, J = 1.1 Hz, 3H, 9-CH3), 1.40 (qd, J = 11.7, 4.4 Hz, 1H, 4-CHax), 1.00 (d, J = 6.3 Hz, 3H, 5-CH3). Positive ion ESI-MS/MS [in MeOH-H2O (1:1) + 2% acetic acid]: m/z 169 (M + H+), 151 (M + H+ - H2O). GC/EI-MS (in acetone): m/z 168 (M·+), 150 (M - H2O), 139 (M - HCO), 108 (M - H2O - C3H6); UV: lambda max 248 nm. The GC/MS data of the major product are consistent with those of 10-hydroxypulegone, as reported in the literature (McClanahan et al., 1988).

7a-Hydroxy-3,6-dimethyl-5,6,7,7a-tetrahydro-2(4H)-benzofuranone. This compound was prepared by a known method (Thomassen et al., 1992). 1H NMR (CDCl3, 360 MHz): delta  2.96 (s, 1H, -OH), 2.68 (dt, J = 13.8, 2.1 Hz, 1H, 4-CHeq), 2.41 to 2.32 (m, 2H), 2.06 to 1.93 (m, 2H), 1.81 (s, 3H, 3-CH3), 1.27 (t, J = 12.7 Hz, 1H, 6-CHax), 1.02 (qd, J = 13.2, 4.1 Hz, 1H, 5-CHax; overlapping with 6-CH3), 0.98 (d, J = 6.2 Hz, 3H, 6-CH3); UV: lambda max 220 nm.

2-(2-Hydroxy-4-methylphenyl)propionic acid. The synthesis was carried out following modification of a literature method for the desmethyl analog (Rewcastle et al., 1991). 2'-Hydroxy-4'-methylacetophenone (1.98 g, 0.013 mol) was stirred with benzyl chloride (2.3 ml), aqueous NaOH (0.67 g in 10 ml of H2O), and tetrabutylammonium bromide (0.45 g, 0.001 mol) in CH2Cl2 (10 ml) at room temperature. After 24 h, the organic layer was separated, washed three times with H2O, and dried with anhydrous CaCl2. After the solvent was removed, the product was purified by flash column chromatography (ether/hexane, 1:4). Thin layer chromatography (silica gel 60 plate, ether/hexane, 1: 4); Rf = 0.22; yield, 2.07 g (81%). 1H NMR (CDCl3, 360 MHz): delta  7.69 (d, J = 8.1 Hz, 1 H, 6-H), 7.46 to 7.38 (m, 5 H, benzyl-Ph), 6.84 (s, 1H, 3-H), 6.83 (d, J = 7.7 Hz, 1 H, 5-H), 5.15 (s, 2H, benzyl-CH2), 2.57 (s, 3H, 4-CH3), 2.38 (s, 3H, COCH3).

The benzylated product (2.07 g, 0.009 mol) in EtOH (25 ml) was added to NaBH4 (0.186 g, 0.005 mol). After the mixture was refluxed for 1 h, acetone (10 ml) was added to neutralize excess NaBH4, and the solvents were removed. Water was added to the residue, and the mixture was extracted three times with EtOAc. The EtOAc layer was dried with anhydrous Na2SO4, and the solvent was removed to give the desired product (yield, 1.72 g; 83%). 1H NMR (CDCl3, 360 MHz): delta  7.44 to 7.35 (m, 5H, benzyl-Ph), 7.25 (d, J = 8.8 Hz, 1H, 6-H), 6.80 (d, J = 8.8 Hz, 1H, 5-H), 6.79 (s, 1H, 3-H), 5.16 (q, J = 7.0 Hz, 1H, CHOH), 5.12 (s, 2H, benzyl-CH2), 2.58 (br. s, 1H, OH), 2.34 (s, 3H, 4-CH3), 1.51 (d, J = 7.0 Hz, 3H, CHCH3).

The alcohol (1.72 g, 0.007 mol) in dioxane (10 ml) was added to a solution of anhydrous CaCl2 (1.69 g, 0.015 mol) in concentrated HCl (3.4 ml). The viscous mixture was stirred at room temperature for 1 h and then diluted with a mixture of EtOAc and ice. The organic layer was washed with H2O and 2 N NaOH (20 ml), dried with anhydrous Na2SO4, and the solvent was removed (yield, 1.45 g; 80%). 1H NMR (CDCl3, 360 MHz): delta  7.47 to 7.32 (m, 5H, benzyl-Ph), 7.35 (d, J = 8.8 Hz, 1H, 6-H), 6.82 (d, J = 8.1 Hz, 1H, 5-H), 6.77 (s, 1H, 3-H), 5.64 (q, J = 6.6 Hz, 1H, CHCl), 5.11 (ABq, J = 11.7 Hz, 2H, benzyl-CH2), 2.34 (s, 3H, 4-CH3), 1.82 (d, J = 7.0 Hz, 3H, CHCH3).

The chloride from the above reaction (1.45 g, 0.006 mol) was dissolved in dimethyl sulfoxide (10 ml). NaCN (0.49 g, 0.009 mol) was added, and the mixture was refluxed for 1 h. After cooling and dilution with ice, the mixture was extracted twice with EtOAc, dried with anhydrous Na2SO4, and the solvent was removed (yield, 2.28 g; 150%). 1H NMR (CDCl3, 360 MHz): delta  7.32 to 7.23 (m, 5H, benzyl-Ph), 7.24 (d, J = 10.0 Hz, 1H, 6-H), 6.78 (d, J = 9.2 Hz, 1H, 5-H), 6.77 (s, 1H, 3-H), 5.18 to 5.06 (m, 3H, benzyl-CH2 + CHCN), 2.33 (s, 3 H, 4-CH3), 1.49 (d, J = 6.6 Hz, 3H, CHCH3).

The crude nitrile from the above reaction (2.28 g, 0.009 mol) was dissolved in a solution of EtOH (15 ml) and 2 N NaOH (15 ml) and refluxed for 24 h. After EtOH was removed, the pH of the remaining mixture was adjusted to 9 by addition of 2 N HCl. The mixture was extracted with EtOAc, and the aqueous layer was acidified to pH 2 with 2 N HCl, then extracted again with EtOAc. The second EtOAc extraction layer was dried with anhydrous Na2SO4, and the solvent was removed. 1H NMR of the residue showed a mixture of two products, which were separated by HPLC (method 7, system A, 225 nm). 1H NMR of the product with retention time 10.3 min (CDCl3, 300 MHz): delta  7.05 (d, J = 7.8 Hz, 1H, 6-H), 6.74 (d, J = 8.4 Hz, 1H, 5-H), 6.71 (s, 1H, 3-H), 3.93 (q, J = 7.5 Hz, 1H, CHCOOH), 2.29 (s, 3H, 4-CH3), 1.56 (d, J = 7.5 Hz, 3H, CHCH3); UV: lambda max 204, 217, 275 nm; negative ion ESI-MS/MS: m/z 179 (M - H+), 135 (M - COOH+). The spectral results are consistent with formation of 2-(2'-hydroxy-4'-methylphenyl)propionic acid.

2-(N-Acetylcystein-S-yl)menthofuran. The synthesis was carried out following modification of a literature method for 2-(glutathion-S-yl)menthofuran (Thomassen et al., 1991). To CH3CN/H2O (3:1) (10 ml) was added alpha ,alpha '-dimethoxydihydromenthofuran (100 mg, 0.47 mmol), which was synthesized by a method described by McClanahan et al. (1989), and N-acetylcysteine (776 mg, 4.75 mmol). The mixture was stirred at room temperature for 1 h. 2-(N-Acetylcystein-S-yl)menthofuran was separated by HPLC using a Rainin (Varian) Microsorb C18 column (5 µm, 4.6 × 250 mm) (method 3, system A, 250 nm, 24.5 min). 1H NMR (CD2Cl2, 300 MHz): delta  6.61 (br. s, 1H, CONH), 4.55 (q, J = 5.8 Hz, 1H, Cys alpha -CH), 3.14 to 3.02 (m, 2H, Cys beta -CH2), 2.63 (dd, J = 16.2, 5.2 Hz, 1H), 2.35 to 2.25 (m, 2H), 2.17 to 1.79 (m, 3H), 2.03 (s, 3H, COCH3), 1.94 (s, 3H, 3-CH3), 1.38 to 1.24 (m, 1H), 1.05 (d, J = 6.6 Hz, 3H, 6-CH3). 1H NMR (D2O, 300 MHz): delta  4.27 (dd, J = 8.4, 2.9 Hz, 1H, Cys alpha -CH), 3.24 (dd, J = 14.0, 3.3 Hz, 1H, Cys beta -CHa), 2.95 (dd, J = 14.0, 8.8 Hz, 1H, Cys beta -CHb), 2.66 (dd, J = 16.5, 4.7 Hz, 1H), 2.40 to 2.29 (m, 2H), 2.21 to 1.81 (m, 3H), 2.07 (s, 3H, COCH3), 1.96 (s, 3H, 3-CH3), 1.40 to 1.27 (m, 1H), 1.06 (d, J = 6.6 Hz, 6-CH3); negative ion ESI-MS/MS: m/z 310 (M - H+), 181 (2-thiomenthofuran anion); UV: lambda max 202, 221, 256 nm.

Reactions of Pulegone with GSH.

In basic medium The reaction was carried out to produce authentic standards for comparison with the GST incubation products. GSH (200 mg, 0.65 mmol) and NaHCO3 (167 mg, 2 mmol) were dissolved in H2O (1 ml). 14C-Labeled pulegone (107 µl, 100 mg, 0.65 mmol, 0.0108 µCi/µmol) was added. The mixture was stirred as tetrahydrofuran was added dropwise until 1 ml had been added. The head-space was flushed with N2, capped, and then stirred at room temperature for 72 h. The unreacted pulegone was extracted with ether. The two radiolabeled products were isolated from aqueous phase by two successive HPLC systems (retention times 12.5 and 15.2 min, method 15, system A, 220 nm; 14.2 and 18.5 min, method 5, system A, 220 nm). The peak at 12.5 min (method 15): 1H NMR (D2O, 360 MHz): delta  4.54 (dd, J = 8.8, 5.5 Hz, 1H, Cys alpha -H), 3.96 (s, 2H, Gly alpha -CH2), 3.81 (t, J = 5.7 Hz, 1H, Glu alpha -CH), 3.10 (dd, J = 13.2, 5.5 Hz, 1H, Cys beta -CHa), 2.89 (dd, J = 13.2, 8.8 Hz, 1H, Cys beta -CHb), 2.76 to 2.67 (m, 2H), 2.58 to 2.47 (m, 2H, Glu gamma -CH2), 2.43 to 2.38 (m, 1H), 2.24 to 2.20 (m, 1H), 2.16 (q, J = 7.1 Hz, 2H, Glu beta -CH2), 2.08 (dd, J = 12.5, 2.2 Hz, 1H), 1.95 to 1.87 (m, 2H), 1.68 to 1.64 (m, 1H), 1.43 (s, 3H, 9-CH3), 1.38 (s, 3H, 10-CH3), 0.90 (d, J = 7.0 Hz, 3H, 5-CH3); negative ion ESI-MS/MS: m/z 458 (M - H+), 306 (glutathione anion). This molecule was assigned as 8-(glutathion-S-yl)isomenthone. The peak at 15.2 min (method 15): 1H NMR (D2O, 300 MHz): delta  4.57 (dd, J = 8.3, 5.5 Hz, 1H, Cys alpha -CH), 3.99 (s, 2H, Gly alpha -CH2), 3.85 (t, J = 6.6 Hz, 1H, Glu alpha -CH), 3.11 (dd, J = 13.8, 5.4 Hz, 1H, Cys beta -CHa), 2.91 (dd, J = 13.5, 8.7 Hz, 1H, Cys beta -CHb), 2.76 (dd, J = 12.9, 5.4 Hz, 1H), 2.54 (td, J = 7.5, 2.9 Hz, 2H, Glu gamma -CH2), 2.45 to 2.36 (m, 1H), 2.29 to 2.22 (m, 2H), 2.18 (q, J = 6.9 Hz, 2H, Glu beta -CH2), 1.97 to 1.88 (m, 2H), 1.61 (q, J = 12.6 Hz, 1H), 1.51 to 1.36 (m, 1H), 1.45 (s, 3H, 9-CH3), 1.38 (s, 3H, 10-CH3), 1.02 (d, J = 6.0 Hz, 3H, 5-CH3); negative ion ESI-MS/MS: m/z 458 (M - H+), 306 (glutathione anion). This molecule was assigned as 8-(glutathion-S-yl)menthone.

Under catalysis of GST. A similar method as in the literature was used (Thomassen et al., 1990). Initial conditions were 1.0 mM GSH, 1.0 mM 14C-labeled pulegone (1 µCi/µmol), 0.2 mg of GST in 0.1 M potassium phosphate buffer, pH 7.7. Pulegone was added as a CH3CN solution (5 µl). Controls lacking GST or GSH were included. The final volume was 0.5 ml. Three replicates of each set were carried out. The components were combined in an ice-cold 3-ml vial that was subsequently capped, vortexed, and incubated at 37°C for 15 min. The reactions were stopped by placing these vials on ice. The products were analyzed by HPLC (method 14, system B) for comparison with the standards prepared above.

Enzyme Hydrolysis of 24 h Urine and Metabolites D2, E2, and 9-Hydroxypulegone Glucuronide. Samples of male and female rat urine (single doses, 80 mg/kg, 24 h) were incubated with glucuronidase (5150 units) or sulfatase (33 units) in a total volume of 0.4 ml of sodium acetate buffer (0.05 M, pH 5.0), as described previously (Burka et al., 1996). All reaction mixtures, including enzyme-free controls, were incubated at 37°C for 17 h. Samples of each incubated mixture were analyzed by HPLC (method 1 and 2, system B) to determine whether any of the peaks were affected by enzymatic hydrolysis.

Individual metabolites D2, E2, and 9-hydroxypulegone glucuronide were incubated with glucuronidase, as described above. An enzyme-free control of 9-hydroxypulegone glucuronide was also included. Samples of each incubated mixture were analyzed by HPLC (method 1 or 2, system B) before any treatment. The reaction mixture of 9-hydroxypulegone glucuronide was then mixed with an equal amount of CH3CN and kept at -20°C to separate the CH3CN and H2O layers. Both layers were analyzed by HPLC (method 2, system B).

Standards of 10-hydroxypulegone, 7a-Hydroxy-3,6-dimethyl-5,6,7,7a-tetrahydro-2(4H)-benzofuranone, pulegone, menthofuran, 2-(N-acetylcystein-S-yl)menthofuran, piperitone, and pulegol were subjected to the same HPLC analyses for comparison.



    Results
Top
Abstract
Introduction
Materials and Methods
Results
Discussion
References

Identification and Quantitation of Urinary Metabolites. At 24 h postdosing, 44 to 71% of pulegone-derived radioactivity was present in urine (Table 1). An additional 14 and 5% of radioactivity was excreted in the urine of male and female rats dosed with 8 mg/kg pulegone from 24 to 72 h, respectively. Urine samples (24 h) from rats treated with various doses were analyzed by HPLC (method 1, system B) to reveal several major radiolabeled peaks (A-M) (Fig. 1). The percentage of each major peak (A-M) in urine with each dosing method was calculated based on the 14C count in each peak compared with the total amount of 14C excreted in each urine sample (Table 1). HPLC peaks A through M accounted for approximately 60% of the radioactivity that was excreted in urine.


View larger version (15K):
[in this window]
[in a new window]
  		Fig. 1.   Representative HPLC radiochromatogram (method 1, system B) of pulegone-derived radioactivity excreted in cumulative 24-h urine of a male rat receiving a single oral dose (80 mg/kg, 40 µCi/kg) in corn oil.

This study focused on 13 radioactive peaks (A-M) in the urinary metabolite profile.

Urinary metabolites were isolated from urine samples (0-4, 4-8, and 8-12 h) from rats receiving 80-mg/kg single and multiple doses. Fractions containing peaks A through M were isolated by HPLC (method 1, system A) with UV detection at 250 nm. Each fraction was concentrated using a Speed-Vac (Savant, Brooklyn, NY) and analyzed by NMR to reveal the major constituents in each peak.

The individual HPLC fractions were subjected to further HPLC separation using various HPLC methods, as described below. The retention times reported below are based on HPLC system A unless otherwise indicated. During HPLC separation, a small portion of the isolated peaks were added to liquid scintillation fluid Ecolume (ICN Research Products Division, Costa Mesa, CA) and counted for 14C in a liquid scintillation counter (Beckman LS 6500) to ensure that radioactive metabolites were being collected. The purified metabolites (Fig. 2) were concentrated using a Speed-Vac before analysis by NMR, MS, and HPLC (method 1, system B).


View larger version (16K):
[in this window]
[in a new window]
  		Fig. 2.   Structures of characterized urinary metabolites of pulegone.

Peak A (RT = 14.5 min, method 1) appeared only in urine of male rat, not in urine of female rat treated with pulegone. It was further purified twice by HPLC (RT = 7.0 min, method 4, 250 nm; RT = 2.9 min, method 9, 250 nm) to give a metabolite with the following spectral properties: 1H NMR (D2O, 360 MHz): delta  4.43 (br. s, 1H), 3.04 (s, 3H), 2.74 to 2.67 (m, 1H), 2.05 (t, J = 15.4 Hz, 4H), 1.76 to 1.66 (m, 2H), 1.51 (s, 3H), 1.45 (s, 3H); UV: lambda max 219 nm. However, all attempts to obtain a molecular weight of this metabolite failed. It is also not clear whether all the NMR peaks belong to the metabolite. We cannot deduce the structure of metabolite A at this point.

Peak B (RT = 16.7 min, method 1) was further separated by HPLC (method 5, 225 nm) to give two main metabolites [RT = 6.6 min (B1) and 7.0 min (B2)]. The more abundant metabolite (B1) had the following spectral properties: 1H NMR (D2O, 360 MHz): delta  4.53 (dd, J = 7.7, 4.8 Hz, 1H, Cys alpha -CH), 3.11 (dd, J = 12.8, 4.8 Hz, 1H, Cys beta -CHa), 2.95 (dd, J = 12.8, 7.7 Hz, 1H, Cys beta -CHb), 2.77 (dd, J = 12.8, 4.4 Hz, 1H, 2-CH), 2.71 (d, J = 12.8 Hz, 1H, 6-CHa), 2.37 to 2.31 (m, 1H), 2.28 (dd, J = 13.6, 1.1 Hz, 1H, 6-CHb), 2.05 (s, 3H, COCH3), 2.00 to 1.84 (m, 2H), 1.46 (s, 3H, 9-CH3), 1.37 (s, 3H, 10-CH3), 1.33 (s, 3H, 5-CH3); negative ion ESI-MS/MS: m/z 330 (M - H+), 162 (N-Ac-Cys anion); UV: lambda max 202 nm. MS suggested that this metabolite was a hydroxylated product of metabolite K or L (8-(N-acetylcystein-S-yl)menthone/isomenthone). The NMR spectrum showed the presence of 5-CH3 group at 1.33 ppm as a singlet, indicating the ---OH substitution to be at the C-5 position. The analogous CH3 group in pulegone is a doublet at 1.0 ppm. B1 was more abundant than its diastereomer C3, and its 5-CH3 group is 0.15 ppm more downfield (see below); therefore, it was identified as 8-(N-acetylcystein-S-yl)-5-hydroxymenthone for the same assignment for K and L. Nevertheless, the assignment of the stereochemistry at the C-2 position was not definitive.

The less abundant metabolite (B2) had the following spectral properties: 1H NMR (D2O, 360 MHz): delta  2.05 (s, 3H, COCH3), 1.91 (s, 3H, 9-CH3), 1.90 (s, 3H, 10-CH3), 1.04 (d, J = 7.7 Hz, 3H, 5-CH3); the other signals were not well resolved; negative ion ESI-MS/MS: m/z 328 (M - H+), 162 (N-Ac-Cys anion); UV: lambda max 195, 248 nm. MS analysis showed that the mercapturic acid metabolite had a molecular weight 2 Da less than B1, indicating the possibility of one carbon-carbon double bond in the structure. The metabolite had a UV maximum at 248 nm, and its NMR spectrum showed the presence of two allylic CH3 groups at 1.91 and 1.90 ppm. The spectral results indicated that the cyclic-isopropylidene ketone structure of the parent pulegone (lambda max 255 nm) remained intact in the metabolite. The 5-CH3 group remained a doublet at 1.04 ppm, so B2 appeared to be a pulegone with one hydroxyl and one N-acetylcystein-S-yl substitutions at two of the C-3, C-4 and C-6 positions.

Peak C (RT = 17.3 min, method 1) was further separated by HPLC (method 11, 225 nm) to give the major metabolite [RT = 3.3 min (C1)] and several minor metabolites (RT = 3.5-3.9 min). Metabolite C1 was further purified by HPLC (RT = 7.5 min, method 5, 250 nm). Its spectral properties were as follows: 1H NMR (D2O, 360 MHz): delta  4.56 (d, J = 8.1 Hz, 1H, Gluc 1'-CH), 4.20 (br. s, 1H, 4-CH), 3.77 to 3.71 (m, 1H, Gluc 5'-CH), 3.56 to 3.48 (m, 2H, Gluc 2'-, 4'-CHs), 3.33 to 3.28 (m, 1H, Gluc 3'-CH), 2.94 (dd, J = 14.7, 3.3 Hz, 1H, 3-CHeq), 2.58 (dd, J = 15.7, 1.5 Hz, 1H), 2.51 to 2.37 (m, 2H), 2.37 to 2.30 (m, 1H), 1.95 (s, 3H, 9-CH3), 1.85 (s, 3H, 10-CH3), 1.04 (d, J = 6.2 Hz, 3H, 5-CH3); negative ion ESI-MS/MS: m/z 343 (M - H+), 325 (M - H3O+), 193 (glucuronide ion), 175 (glucuronide ion - H2O), 157 (glucuronide ion - 2 H2O); UV: lambda max 257 nm. MS of C1 displayed a molecular ion peak at m/z 343, consistent with the anionic form (M - H+) of a hydroxypulegone glucuronide. C1 had a UV maximum at 257 nm, and its NMR spectrum showed the presence of two allylic CH3 groups at 1.95 and 1.85 ppm, indicating that the cyclic-isopropylidene ketone structure of pulegone remained intact in the metabolite. NMR spectrum demonstrated that there was one proton geminal to the ---OGluc substituent present at 4.20 ppm, so the hydroxylation/glucuronidation must have taken place in one of the methylene protons; that is, C-3, C-4, or C-6 positions. The metabolite was not 3-hydroxypulegone glucuronide, which was characterized (E1, 3-CH is at 5.29 ppm). The ---OGluc substitution was more likely to be at C-4 position because the proton at 4.20 ppm was upfield compared with what would be expected for the one at the C-6 position. In addition, the 3-CHeq was a doublet of doublets in C1 compared with a doublet of triplets (J = 15.5, 4.0 Hz) in the parent pulegone, an indication of a substitution at the C-4 position. Metabolite C1 was identified as 4-hydroxypulegone glucuronide with unknown stereochemistry at the C-4 position. Small amounts of the products from Michael addition of water to C1 and other hydroxypulegone glucuronides were generated in acidic conditions but not in neutral conditions, so it was necessary to remove TFA right after the HPLC isolation.

The minor metabolites were separated by HPLC [RT = 10.7 (C2) and 11.3 min (C3 and C4), method 9, 225 nm]. Metabolite C2 was further purified by HPLC (RT = 4.6 min, method 6, 225 nm). Its spectral properties were as follows: 1H NMR (D2O, 360 MHz): delta  7.19 (br. s, 1H, 3-CH), 4.35 (d, J = 7.3 Hz, 1H, Gluc 1'-CH), 1.56 (s, 3H, 9-CH3), 1.46 (s, 3H, 10-CH3), 1.00 (d, J = 4.0 Hz, 3H, 5-CH3); the other signals were not well resolved; negative ion ESI-MS: m/z 343 (M - H+), 193 (glucuronide ion). Positive ion ESI-MS: m/z 345 (M + H+), 169 (5-methyl-2-(1'-hydroxy-1'-methylethyl)-2-cyclohexene-1-one + H+), 151 (M - glucuronide ion); UV: lambda max 233 nm. MS showed that C2 was a glucuronide with the same molecular weight as hydroxypulegone glucuronide metabolites, but its NMR and UV spectra indicated that the carbon-carbon double bond in the parent pulegone has rearranged. NMR of C2 showed a rather downfield olefinic proton at 7.19 nm and two CH3 groups at 1.56 and 1.46 ppm. The NMR signals are comparable with those of 5-methyl-2-(1'-hydroxy-1'-methylethyl)-2-cyclohexene-1-one, which has an olefinic H at 6.88 ppm and two CH3 groups at 1.4 ppm (Madyastha and Thulasiram, 1999). C2 had a UV maximum at 233 nm, in agreement with the presence of 2-cyclohexene-1-one as part of the structure as in piperitone (UV maximum, 235 nm). Metabolite C2 was tentatively identified as 5-methyl-2-(1'-hydroxy-1'-methylethyl)-2-cyclohexene-1-one glucuronide, an allylic alcohol isomer of metabolite E1. The presence of a bulky glucuronide substituent probably moves the NMR signals more downfield in C2 compared with those of its aglycone. Positive ion ESI-MS/MS analysis gave a fragment with m/z = 151, consistent with formation of a stable allylic cation through removal of a glucuronic acid.

Metabolites C3 and C4 were a mixture of mercapturic acids. Based on the ratio of these two metabolites, some of their spectral properties were resolved as followed: C3: 1H NMR (D2O, 360 MHz): delta  2.05 (s, 3H, COCH3), 1.45 (s, 3H, 9-CH3), 1.36 (s, 3H, 10-CH3), 1.18 (s, 3H, 5-CH3); the other signals were not well resolved; negative ion ESI-MS/MS: m/z 330 (M - H+), 162 (N-Ac-Cys anion). These two mercapturic acids could be separated by HPLC using method 1, system B (C3, RT = 17.1 min; C4, RT = 16.8 min) to obtain individual UV spectra. UV for C3 was the following: lambda max 202 nm. MS analysis suggested that C3 was a hydroxylated 8-(N-acetylcystein-S-yl)menthone/isomenthone and the NMR results showed the 5-CH3 group was a singlet at 1.18 ppm, indicating the ---OH substitution was at the C-5 position. C3 was a diastereomer of B1; their 5-CH3 groups showed significant difference in chemical shifts as the corresponding CH3 groups in diastereomeric metabolites K and L. Metabolite C3 was tentatively identified as 8-(N-acetylcystein-S-yl)-5-hydroxyisomenthone.

C4: 1H NMR (D2O, 360 MHz): delta  2.01 (s, 3H, COCH3), 1.54 (s, 3H, 9-CH3), 1.52 (s, 3H, 10-CH3), 1.10 (d, J = 5.9 Hz, 3H, 5-CH3); the other signals were not well resolved; negative ion ESI-MS/MS: m/z 328 (M - H+), 162 (N-Ac-Cys anion); UV: lambda max 235 nm. C4 had the same molecular weight as B2, but the NMR chemical shifts of the CH3 groups and the UV maximum suggested that C4 was likely to be 5-methyl-2-(1'-(N-acetylcystein-S-yl)-1'-methylethyl)-2-cyclohexene-1-one with an ---OH substitution at an unknown position. C4 is a mercapturic acid counterpart of metabolite C2 plus an ---OH substitution.

Peak D (RT = 17.9 min, method 1) was further separated by HPLC (method 11, 250 nm) to give two major metabolites [RT = 3.8 (D1) and 4.0 (D2) min]. The less abundant metabolite (D1) had the following spectral properties: 1H NMR (D2O, 360 MHz): delta  4.65 (d, J = 7.7 Hz, 1H, Gluc 1'-CH), 3.73 to 3.68 (m, 1H, Gluc 5'-CH), 3.52 to 3.47 (m, 2H, Gluc 2'-, 4'-CHs), 3.25 to 3.20 (m, 1H, Gluc 3'-CH), 2.77 to 2.66 (m, 1H), 2.73 (d, J = 15.8 Hz, 1H, 6-CHa), 2.61 (d, J = 16.1 Hz, 1H, 6-CHb), 2.53 (dt, J = 15.0, ~5 Hz, 1H), 2.11 to 2.02 (m, 1H), 1.98 to 1.84 (m, 1H), 1.95 (s, 3H, 9-CH3), 1.84 (s, 3H, 10-CH3), 1.34 (s, 3H, 5-CH3); negative ion ESI-MS/MS: m/z 343 (M - H+), 193 (glucuronide ion), 175 (glucuronide ion - H2O); UV: lambda max 259 nm. The spectral results indicated that metabolite D1 was a hydroxypulegone glucuronide. The NMR spectrum showed the presence of the 5-CH3 at 1.34 ppm as a singlet, which suggested that the ---OGluc substitution be at the C-5 position. This metabolite was identified as 5-hydroxypulegone glucuronide (D1).

The more abundant metabolite (D2) had the following spectral properties: 1H NMR (D2O, 300 MHz): delta  4.40 (ABq, J = 12.0 Hz, 2H, 10-CH2OGluc), 4.39 (d, J = 6.9 Hz, 1H, Gluc 1'-H), 3.71 to 3.65 (m, 1H, Gluc 5'-H), 3.55 to 3.46 (m, 2H, Gluc 2'-, 4'-CHs), 3.36 to 3.31 (m, 1H, Gluc 3'-CH), 2.86 (dt, J = 14.8, 4.5 Hz, 1H, 3-CHeq), 2.58 (dd, J = 14.3, 3.2 Hz, 1H, 6-CHeq), 2.32 (br. t, J = 13.7 Hz, 1H, 3-CHax), 2.24 (dd, J = 14.3, 10.4 Hz, 1H, 6-CHax), 2.17 to 2.05 (m, 1H, 5-CHax), 1.96 to 1.82 (m, 1H, 4-CHeq), 1.86 (s, 3H, 9-CH3), 1.42 (qd, J = 12.0, 4.6 Hz, 1H, 4-CHax), 1.01 (d, J = 6.3 Hz, 3H, 5-CH3); negative ion ESI-MS/MS: m/z 343 (M - H+), 325 (M - H3O+), 193 (glucuronide ion), 175 (glucuronide ion - H2O), 167 (10-hydroxypulegone anion), 157 (glucuronide ion - 2 H2O), 149 (10-hydroxypulegone anion - H2O); UV: lambda max 248. The spectral results indicated that metabolite D2 was a hydroxypulegone glucuronide. NMR showed only one allylic CH3 group and two protons at 4.38 ppm as an AB quartet due to the ---OGluc substitution at either the C-9 or the C-10 position. Metabolite D2 was hydrolyzed by glucuronidase to give a product that comigrated (20.7 min, method 1, system B) with 10-hydroxypulegone prepared from microsomal incubation of pulegone. Metabolite D2 was identified as 10-hydroxypulegone glucuronide.

Metabolite D2 partially isomerized to its geometric isomer, 9-hydroxypulegone glucuronide, during isolation in acidic conditions, possibly via acid-catalyzed addition of water followed by elimination. 9-Hydroxypulegone glucuronide (RT = 5.9 min, method 11, 250 nm) had the following spectral properties: 1H NMR (D2O): delta  4.44 (d, J = 12.4 Hz, 1H, 9-CHa), 4.31 (d, J = 7.7 Hz, 1H, Gluc 1'-CH), 4.30 (d, J = 13.0 Hz, 1H, 9-CHb), 3.61 to 3.56 (m, 1H, Gluc 5'-CH), 3.50 to 3.41 (m, 2H, Gluc 2', 4'-CHs), 3.31 to 3.26 (m, 1H, Gluc 3'-CH), 2.78 (dt, J = 15.6, 4.4 Hz, 1H, 3-CHeq), 2.53 (dd, J = 14.7, 2.4 Hz, 1H, 6-CHeq), 2.29 (br. t, J = 14 Hz, 1H, 3-CHax), 2.13 (t, J = 11.0 Hz, 1H, 6-CHax), 2.06 to 1.99 (m, 1H, 5-CHax), 1.95 to 1.87 (m, 1H, 4-CHeq), 1.83 (s, 3H, 10-CH3), 1.46 to 1.35 (m, 1H, 4-CHax), 0.98 (d, J = 6.3 Hz, 3H, 5-CH3); negative ion ESI-MS/MS: m/z 343 (M - H+), 325 (M - H3O+), 193 (glucuronide ion), 175 (glucuronide ion - H2O), 167 (9-hydroxypulegone anion), 157 (glucuronide ion - 2 H2O), 149 (9-hydroxypulegone anion - H2O); UV: lambda max 248 nm. 9-Hydroxypulegone glucuronide was analyzed by HPLC (RT = 18.7 min, method 1, system B), which showed that it was either not a metabolite or a very minor metabolite of pulegone. It has been speculated that 9-hydroxypulegone, if formed, would cyclize and dehydrate to give menthofuran (Gordon et al., 1987). Therefore, it was of interest to investigate the products from glucuronidase treatment of 9-hydroxypulegone glucuronide. HPLC analysis (method 2, system B) of the reaction mixture showed total disappearance of the glucuronide peak, but no other peaks were formed. CH3CN extraction of the reaction mixture did not give the anticipated menthofuran upon HPLC analysis, either.

Peak E (RT = 18.5 min, method 1) was further separated by HPLC (method 11, 250 nm) to three major metabolites [RT = 3.9 (E1), 4.3 (E2), and 4.8 (E3) min]. Metabolite E1 was further purified by two successive HPLC systems (RT = 7.3 min, method 13, 250 nm; RT = 5.0 min, method 6, 250 nm). Its spectral properties were as follows: 1H NMR (D2O, 360 MHz): delta  5.29 (dd, J = 7.0, 2.2 Hz, 1H, 3-CH), 4.17 (d, J = 7.7 Hz, 1H, Gluc 1'-CH), 3.62 (d, J = 9.5 Hz, 1H, Gluc 5'-CH), 3.50 (t, J = 9.2 Hz, 1H, Gluc 2'-CH), 3.41 (t, J = 9.2 Hz, Gluc 4'-CH), 3.24 (t, J = 8.8 Hz, Gluc 3'-CH), 2.47 (dd, J = 17.6, 4.8 Hz, 1H, 6-CHeq), 2.41 to 2.35 (m, 1H), 2.29 (dd, J = 16.8, 11.7 Hz, 1H, 6-CHax), 2.06 to 1.95 (m, 1H), 2.01 (s, 3H, 9-CH3), 1.93 (s, 3H, 10-CH3), 1.63 to 1.54 (m, 1H), 1.03 (d, J = 7.0 Hz, 3H, 5-CH3); negative ion ESI-MS: m/z 343 (M - H+), 193 (glucuronide ion); UV: lambda max 245 nm. The spectral data indicated that the metabolite was a hydroxypulegone glucuronide. There was one proton present at 5.29 ppm as a doublet of doublets, more than 1 ppm downfield than its counterpart in metabolite C1, which suggested the ---OGluc substitution be at the allylic C-3 position. This metabolite was identified as 3-hydroxypulegone glucuronide (E1), although the stereochemistry at the C-3 position was not clear.

Metabolite E2 was further purified by HPLC (RT = 5.5 min, method 6, 210 nm) to give the following spectral properties: 1H NMR (D2O, 360 MHz): delta  4.45 (d, J = 7.7 Hz, 1H, Gluc 1'-CH), 4.01 (dd, J = 10.3, 4.0 Hz, 1H, 2-CHa), 3.81 (d, J = 8.8 Hz, 1H, 2-CHb), 3.56 to 3.48 (m, 3H, Gluc 2', 4', and 5'-CHs), 3.32 (t, J = 8.1 Hz, 1H, Gluc 3'-CH), 2.43 to 2.33 (m, 2H), 2.27 to 2.20 (m, 2H), 2.13 to 2.03 (m, 1H), 2.02 to 1.94 (m, 1H), 1.82 to 1.72 (m, 2H), 1.60 to 1.53 (m, 1H), 0.97 (d, J = 6.6 Hz, 3H, 3-CH3), 0.90 (d, J = 6.6 Hz, 3H, 6-CH3); negative ion ESI-MS/MS: m/z 345 (M - H+), 327 (M - H3O+), 193 (glucuronide ion), 175 (glucuronide ion - H2O), 157 (glucuronide ion - 2 H2O); UV: lambda max <190 nm. The molecular ion peak at m/z 345 (M - H+) suggested that metabolite E2 was a glucuronic acid conjugate of monohydroxylated, reduced pulegone. This metabolite had almost no UV absorption; therefore, the reduced pulegone was more likely to be menthone/isomenthone, which had almost no UV absorption, and less likely to be pulegol, which showed a UV maximum at 201 nm. NMR of E2 showed only two CH3 groups at 0.97 and 0.90 ppm as doublets, which confirmed the reduction of the carbon-carbon double bond in the metabolite and indicated the ---OH substitution on one of the CH3 groups. The ---OH substitution could be at the C-9 position of menthone, which would cyclize to give a five-membered ring, or at the C-7 position of menthone, which could not cyclize. A very similar metabolite G1 showed two CH3 groups at 1.02 and 0.84 ppm as doublets and two protons on the ---OCH2--- at 3.79 and 3.57 ppm. Because the CH2 in cyclopentane (1.50 ppm) is more downfield than in n-pentane (1.25 ppm) (Pouchert and Campbell, 1974), we assigned E2, whose ---OCH2--- appeared at 4.01 and 3.81 ppm as 7a-hydroxy-3,6-dimethyloctahydrobenzofuran glucuronide, and G1 as 7-hydroxymenthone glucuronide. There was a minor metabolite isolated along with E2, which was probably the isomenthone diastereomer of E2.

Metabolite E3 was further purified by HPLC (RT = 5.6 min, method 6, 225 nm) to give a molecule with the following spectral properties: 1H NMR (D2O, 360 MHz): delta  4.41 (d, J = 7.7 Hz, 1H, Gluc 1'-CH), 3.61 (d, J = 9.2 Hz, 1H, Gluc 5'-CH), 3.49 (t, J = 8.4 Hz, 1H, Gluc 2'-CH), 3.44 (t, J = 8.8 Hz, 1H, Gluc 4'-CH), 3.34 (t, J = 8.8 Hz, 1H, Gluc 3'-CH), 2.83 (br. d, J = 13.6 Hz, 1H), 2.45 (br. d, J = 9.4 Hz, 1H), 2.34 (td, J = 12.1, 5.1 Hz, 1H), 2.01 to 1.91 (m, 2H), 1.83 (s, 3H, 3-CH3), 1.32 (t, J = 13.0 Hz, 1H), 1.04 to 1.01 (m, 1H), 0.94 (d, J = 6.6 Hz, 3H, 6-CH3); negative ion ESI-MS/MS: m/z 357 (M - H+), 193 (glucuronide ion), 175 (glucuronide ion - H2O); UV: lambda max 222 nm. NMR spectrum of E3 showed only one allylic CH3 group at 1.83 ppm, which indicated that the other allylic CH3 group was modified. MS analysis gave a molecular weight 14 Da more than hydroxypulegone glucuronide metabolites, consistent with oxidation of the other allylic CH3 group to a carboxylic acid. E3 showed a UV maximum at 222 nm, indicating the carboxylic acid group was syn to the ketone group, which cyclized to give 7a-hydroxy-3,6-dimethyl-5,6,7,7a-tetrahydro-2(4H)-benzofuranone (a hydroxylated alpha ,beta -unsaturated-gamma -lactone). The ---OH group was conjugated with a glucuronic acid. This metabolite was partially hydrolyzed by glucuronidase to give a new product, which had a similar HPLC retention time (23.0 min, method 1, system B) and UV maximum as the synthetic 7a-hydroxy-3,6-dimethyl-5,6,7,7a-tetrahydro-2(4H)-benzofuranone (UV: lambda max 220 nm). Metabolite E3 was identified as 7a-hydroxy-3,6-dimethyl-5,6,7,7a-tetrahydro-2(4H)-benzofuranone glucuronide.

Peak F (RT = 19.3 min, method 1) was further separated by HPLC (method 13, 210 nm) to give two major metabolites (RT = 8.6 min, F1; RT = 9.7 min, F2). After further purification by HPLC (RT = 6.5 min, method 6, 210 nm), metabolite F1 had the following spectral characteristics: 1H NMR (D2O, 300 MHz): delta  4.64 (d, J = 8.0 Hz, 1H, Gluc 1'-H), 3.74 to 3.71 (m, 1H, Gluc 5'-H), 3.54 to 3.48 (m, 2H, Gluc 2'-, 4'-CHs), 3.28 to 3.22 (m, 1H, Gluc 3'-CH), 2.62 (ABq, J = 13.5 Hz, 2H, 6-CH2), 2.30 to 2.23 (m, 1H), 2.19 to 1.81 (m, 5H), 1.36 (s, 3H, 5-CH3), 0.89 (d, J = 6.6 Hz, 3H, 9-CH3), 0.88 (d, J = 6.9 Hz, 3H, 10-CH3); negative ion ESI-MS/MS: m/z 345 (M - H+), 193 (glucuronide ion), 175 (glucuronide ion - H2O); UV: lambda max <190 nm. The spectral data indicated that the metabolite was a hydroxymenthone/isomenthone glucuronide. NMR showed the presence of two CH3 groups at 0.89 and 0.88 ppm as doublets and one CH3 at 1.36 ppm as a singlet, due to the ---OGluc substitution at the C-5 position. Because metabolite F1 was slightly less abundant than metabolite F2, we assigned this metabolite as an isomenthone. Metabolite F1 was identified as 5-hydroxyisomenthone glucuronide, although the assignment of stereochemistry was not certain.

After purification by HPLC (6.3 min, method 6, 210 nm), metabolite F2 had the following spectral characteristics: 1H NMR (D2O, 300 MHz): delta  4.66 (d, J = 7.4 Hz, 1H, Gluc 1'-H), 3.72 to 3.67 (m, 1H, Gluc 5'-H), 3.55 to 3.46 (m, 2H, Gluc 2'-, 4'-CHs), 3.27 to 3.22 (m, 1H, Gluc 3'-CH), 2.63 (s, 2H, 6-CH2), 2.18 to 2.01 (m, 3H), 1.97 (t, J = 5.7 Hz, 2H), 1.74 to 1.64 (m, 1H), 1.30 (s, 3H, 5-CH3), 0.92 (d, J = 6.6 Hz, 3H, 9-CH3), 0.83 (d, J = 6.3 Hz, 3H, 10-CH3); negative ion ESI-MS/MS: m/z 345 (M - H+), 193 (glucuronide ion), 175 (glucuronide ion - H2O); UV: lambda max <190 nm. This metabolite appeared to be the diastereomer of metabolite F1. Metabolite F2 was identified as 5-hydroxymenthone glucuronide.

There was a minor metabolite (7.0 min, method 13, 210 nm) isolated from peak F. It was further purified by HPLC (6.3 min, method 6, 210 nm) and was shown to have a molecular weight of 331. The metabolite was likely to be a hydroxylated 8-(N-acetylcystein-S-yl)menthone/isomenthone, although the position of the ---OH substitution was unknown.

Peak G (RT = 19.8 min, method 1) was separated by HPLC (method 10, 210 nm) to three poorly resolved peaks containing radioactivity (6.8, 7.3, 7.8 min). The peak at 6.8 min contained several minor metabolites, which were not isolated. The peak at 7.3 min was further purified (RT = 7.3 min, method 5, 210 nm) to give the major metabolite G1. 1H NMR (D2O, 300 MHz): delta  4.42 (d, J = 7.7 Hz, 1H, Gluc 1'-CH), 3.79 (t, J = 9.1 Hz, 1H, 7-CHa), 3.71 to 3.68 (m, 1H, Gluc 5'-CH), 3.57 (dd, J = 9.9, 4.2 Hz, 1H, 7-CHb), 3.52 to 3.49 (m, 2H, Gluc 2'-, 4'-CHs), 3.34 to 3.26 (m, 1H, Gluc 3'-CH), 2.71 to 2.63 (m, 1H), 2.36 (dd, J = 12.6, 4.6 Hz, 2H), 2.18 (t, J = 12.1 Hz, 1H), 2.11 to 2.04 (m, 1H), 1.98 to 1.83 (m, 2H), 1.56 to 1.37 (m, 2H), 1.02 (d, J = 6.3 Hz, 3H, 9-CH3), 0.84 (d, J = 6.9 Hz, 3H, 10-CH3); negative ion ESI-MS/MS: m/z 345 (M - H+), 327 (M - H3O+), 193 (glucuronide ion); UV: lambda max <190 nm. Metabolite G1 had very similar spectral data as metabolite E2, and subsequently G1 was identified as 7-hydroxymenthone glucuronide as the result of comparison of NMR spectra.

The peak at 7.8 min was further purified (RT = 7.6 min, method 6, 210 nm) to give a mixture of two glucuronides (G2 and G3). Metabolite G2 had the following spectral properties: 1H NMR (D2O, 300 MHz) showed 2 CH3 groups at delta  1.15 (d, J = 7.2 Hz, 3H), 0.99 (d, J = 6.3 Hz, 3H), and the other signals were not well resolved; negative ion ESI-MS/MS: m/z 359 (M - H+), 193 (glucuronide ion), 183 (3,6-dimethyl-3,4,5,6,7,7a-hexahydro-2-benzofuranone anion), 175 (glucuronide ion - H2O); UV: lambda max <190 nm. Metabolite G2 had a molecular weight 2 Da higher than that of E3, as a result of reduction of either the carbon-carbon double bond or the carbonyl group of E3. NMR indicated that the reduction occurred at the carbon-carbon double bond, as evidenced by the absence of allylic CH3 groups and presence of two CH3 groups at 1.15 and 0.99 ppm as doublets. Metabolite G2 was tentatively identified as 7a-hydroxy-3,6-dimethyl-3,4,5,6,7,7a-hexahydro-2-benzofuranone glucuronide.

Metabolite G3 had the following spectral properties: 1H NMR (D2O, 300 MHz) showed two CH3 groups at delta  0.88 (t, J = 7.2 Hz, 6H), and the other signals were not well resolved; negative ion ESI-MS/MS: m/z 361 (M - H+), 193 (glucuronide ion), 185 (7a-dihydroxy-3,6-dimethyl-octahydrobenzofuran anion), 175 (glucuronide ion - H2O); UV: lambda max <190 nm. Metabolite G3 had a molecular weight 4 Da higher than that of E3, as a result of reduction of both the carbon-carbon double bond and the carbonyl group of E3. Metabolite G3 was tentatively identified as 2,7a-dihydroxy-3,6-dimethyl-octahydrobenzofuran glucuronide.

Peak H (isolated in fraction 20-21 min, method 1) and peak I (isolated in fraction 21-22 min, method 1) contained minor metabolites, which could not be identified. MS analysis showed formation of a hydroxylated 8-(N-acetylcystein-S-yl)menthone/isomenthone (M = 331) contained in peak H, although the position of the ---OH group was unknown.

Peak J (isolated in fraction 22-24 min, method 1) contained a major metabolite J, which was further purified by HPLC (RT = 10.5 min, method 7, 225 nm). Metabolite J had the following spectral characteristics: 1H NMR (CDCl3, 300 MHz): delta  7.05 (d, J = 7.8 Hz, 1H, 6'-H), 6.74 (d, J = 7.5 Hz, 1H, 5'-H), 6.72 (s, 1H, 3'-H), 3.93 (q, J = 6.9 Hz, 1H, CHCH3), 2.29 (s, 3H, 4'-CH3), 1.56 (d, J = 7.2 Hz, 3H, CHCH3); negative ion ESI-MS/MS: m/z 179 (M - H+), 135 (M - COOH+); UV: lambda max 200, 221, 275 nm. The NMR spectrum showed two doublets at 7.05 and 6.74 ppm with a coupling constant of 7.5 Hz, consistent with ortho protons on an aromatic structure. The singlet at 6.72 ppm indicated the presence of a proton isolated from the other aromatic protons, and the singlet at 2.29 ppm was consistent with an aromatic CH3 group. MS/MS gave a peak at m/z 135 (M - COOH+), which showed evidence of a carboxylic acid fragmentation in the metabolite. One proton at 3.93 ppm as a quartet and a CH3 group at 1.56 ppm as a doublet agreed with a 2-phenylpropionic acid structure. This metabolite was identified as 2-(2'-hydroxy-4'methylphenyl)propionic acid. 2-(2'-Hydroxy-4'methylphenyl)propionic acid is also a metabolite of thymol in rats; the reported GC-MS data of its methylated derivative are consistent with the MS result of metabolite J (Austgulen et al., 1987). An independent synthesis of this metabolite was carried out starting from 2'-hydroxy-4'-methylacetophenone. The ---OH group was first benzylated, followed by successive conversion of the ketone group to alcohol, chloride, and then cyanide. Subsequent hydrolysis did not only convert the cyanide group to the carboxylic acid but also partially deprotected the ---OH group. NMR and mass spectra, as well as the HPLC chromatograms of the metabolite and the authentic standard were identical.

Peak K (isolated in fraction 24-25 min, method 1) contained one major metabolite. After further purification by HPLC (RT = 12.5 min, method 7, 225 nm), the major metabolite K had the following spectral characteristics: 1H NMR (D2O, 360 MHz): delta  4.34 (dd, J = 8.1, 4.0 Hz, 1H, Cys alpha -CH), 3.09 (dd, J = 13.2, 4.4 Hz, 1H, Cys beta -CHa), 2.88 (dd, J = 13.2, 8.4 Hz, 1H, Cys beta -CHb), 2.72 to 2.64 (m, 1H), 2.70 (dd, J = 12.1, 5.1 Hz, 1H, 2-CH), 2.41 to 2.34 (m, 1H), 2.25 to 2.18 (m, 1H), 2.09 to 2.03 (m, 1H), 2.05 (s, 3H, COCH3), 1.91 to 1.84 (m, 2H), 1.65 to 1.60 (m, 1H), 1.44 (s, 3H, 9-CH3), 1.37 (s, 3H, 10-CH3), 0.91 (d, J = 7.3 Hz, 3H, 5-CH3); negative ion ESI-MS: m/z 314 (M - H+), 162 (N-Ac-Cys anion); UV: lambda max 202 nm. MS of metabolite K was consistent with an N-acetylcysteine conjugate derived from conjugation of glutathione with unmodified pulegone. NMR demonstrated that the N-acetylcysteine substituent was at the 8-position, as evidenced by the chemical shifts of 9- and 10-CH3 groups at 1.44 and 1.37 ppm as singlets. Michael addition of glutathione to the alpha ,beta -unsaturated ketone group of pulegone took place in vivo to give the ultimate diastereomeric metabolites K and L (see below). Because metabolite K was less abundant than metabolite L, we assigned this reduced pulegone as an isomenthone, although the assignment was not unambiguous. Metabolite K was identified as 8-(N-acetylcystein-S-yl)isomenthone.

Peak L and peak M were isolated in one fraction (RT = 25-26.5 min, method 1). When further purified by HPLC (RT = 9.1 min, method 8, 225 nm), the major metabolite L had the following spectral characteristics: 1H NMR (D2O, 360 MHz): delta  4.42 (dd, J = 7.7, 4.4 Hz, 1H, Cys alpha -CH), 3.08 (dd, J = 12.8, 4.4 Hz, 1H, Cys beta -CHa), 2.90 (dd, J = 12.8, 7.7 Hz, 1H, Cys beta -CHb), 2.75 (dd, J = 13.0, 4.9 Hz, 1H, 2-CH), 2.43 to 2.38 (m, 1H), 2.25 to 2.22 (m, 2H), 2.03 (s, 3H, -COCH3), 1.93 to 1.86 (m, 2H), 1.58 (qd, J = 13.2, 2.2 Hz, 1H), 1.47 to 1.38 (m, 1H), 1.43 (s, 3H, 9-CH3), 1.35 (s, 3H, 10-CH3), 1.00 (d, J = 6.2 Hz, 3H, 5-CH3); negative ion ESI-MS/MS: m/z 314 (M - H+), 162 (N-Ac-Cys anion); UV: lambda max 202 nm. This metabolite appeared to be a diastereomer of metabolite K. Its 5-CH3 group is 0.09 ppm more downfield than the corresponding CH3 group of K. Metabolite L was identified as 8-(N-acetylcystein-S-yl)menthone.

The minor metabolite M (RT = 10.9 min, method 8, 225 nm) showed the following characteristics: 1H NMR (D2O): delta  7.29 (d, J = 7.7 Hz, 1H, 6'-H), 7.28 (s, 1H, 3'-H), 7.14 (d, J = 7.7 Hz, 1H, 5'-H), 5.24 (s, 1H, olefinic H), 5.12 (s, 1H, olefinic H), 2.36 (s, 3H, 4'-CH3), 2.12 (s, 3H, 3-CH3); negative ion ESI-MS/MS: m/z 227 (M - H+), 147 (M - SO3H+); UV: lambda max 205, 236 nm. MS/MS showed a peak at m/z 147 (M - SO3H+), indicating the presence of a sulfate group in the metabolite. The NMR spectrum of metabolite M displayed similar aromatic protons and an aromatic methyl group as that of metabolite J. The presence of the sulfate group on 2'-OH shifted 3'-H downfield, a shift to 7.28 ppm from 6.72 ppm in metabolite J. Singlets at 5.24 and 5.12 ppm were olefinic protons, and the singlet at 2.12 ppm was consistent with an allylic methyl group. Metabolite M was identified as 2-(2'-hydroxy-4'-methylphenyl)propene sulfate. Its phenol precursor, 2-(2'-hydroxy-4'methylphenyl)propene is a known compound; its 1H NMR data in CDCl3 are as follows: delta  7.1 to 6.7 (2 doublets, 2H, 5'-H and 6'-H), 6.75 (s, 1H, 3'-H), 5.7 (br. s, 1H, -OH), 5.4 (br. s, 1H, 1-olefinic H), 5.15 (br. s, 1H, 1-olefinic H), 2.3 (s, 3H, 4'-CH3), 2.1 (s, 3H, 3-CH3) (Madyastha and Gaikwad, 1999). M showed similar NMR results as its phenol precursor except the 3'-H shift due to the effect of the sulfate group.

As shown in Table 1 and Fig. 2, peaks C-G are mostly glucuronides, which account for 38 to 47% of the total radioactivity in urine. Mercapturic acids (peaks B, K, and L) constitute 9 to 3% and phenols (peak J and M) make up 3 to 7% of the total radioactivity in urine.

Reactions of Pulegone and Glutathione in Vitro. The in vivo study showed that at least 9 to 13% of metabolites were derived from direct Michael addition of glutathione to pulegone. These results prompted us to reinvestigate the reactivity of pulegone with glutathione in vitro. Two radiolabeled products were isolated from the reaction of glutathione with [14C]pulegone catalyzed by 3 Eq of NaHCO3. They were identified by MS and NMR to be diastereomeric 8-(glutathion-S-yl)menthone/isomenthone. The chemical shifts of their 5-CH3 groups were 1.02 and 0.9 ppm, respectively. We assigned the product with the more downfield 5-CH3 as a menthone in agreement with the assignment for metabolites K and L. These two authentic standards showed similar retention times with the respective two GST incubation products (Fig. 3). These two GST incubation products were formed only when both GST and GSH were present in the incubation mixtures (Fig. 3). The results demonstrated that pulegone underwent Michael addition with glutathione to give the diastereomeric 8-(glutathion-S-yl)menthone/isomenthone under catalysis of GST in vitro.


View larger version (14K):
[in this window]
[in a new window]
  		Fig. 3.   Comparative HPLC radiochromatograms (method 14, system B) of 8-(S-glutathionyl)menthone/isomenthone standards (a, b) and the products from the reaction of pulegone and glutathione under catalysis by GST (c).

These two products were absent in controls lacking GST or GSH (d, e).

Enzyme Hydrolysis of 24-h Urine and Metabolites D2, E2, and 9-Hydroxypulegone Glucuronide. Glucuronidase treatment of 24-h urine (80 mg/kg, single dose) resulted in hydrolysis of many metabolites to give aglycones with longer retention times, whereas sulfatase treatment did not change the metabolite profile. Pulegone (RT = 21.6 min, method 2, system B), menthofuran (RT = 26.4 min, method 2, system B), 2-(N-acetylcystein-S-yl)menthofuran (RT = 19.0 min, method 2, system B), piperitone (RT = 20.0 min, method 2, system B), and pulegol (RT = 20.4 min, method 2, system B) were not observed in either hydrolyzed or untreated urine (data not shown). The results of glucuronidase hydrolysis of metabolites D2, E2, and 9-hydroxypulegone glucuronide were described previously in this work.



    Discussion
Top
Abstract
Introduction
Materials and Methods
Results
Discussion
References

This study focused on 13 radioactive peaks (A-M) in the rat urinary metabolite profile of pulegone (Fig. 1). None of the peaks contained more than 15% of total radioactivity in urine, and several of the peaks were resolved into more than one component (Table 1). Approximately 40% of the eluted radioactivity was diffuse and poorly resolved and not pursued further. Unchanged pulegone was not detected. Pulegone was found to be metabolized via three pathways (Fig. 4): 1) hydroxylation followed by glucuronidation (metabolites C1, D1, D2, and E1) or further metabolism (metabolites E3, J, and M); 2) reduction to give menthone/isomenthone, followed by hydroxylation/glucuronidation (metabolites E2, F1, F2, and G1); and 3) formation of mercapturic acids (metabolites K and L) followed by hydroxylation (metabolite B1).


View larger version (15K):
[in this window]
[in a new window]
  		Fig. 4.   Three metabolic pathways proposed for the biotransformation of pulegone to urinary metabolites in rats.

a, hydroxylation followed by glucuronidation (C1, D1, D2, and E1) or further metabolism (E3, J, and M); b, reduction to give menthone/isomenthone, followed by hydroxylation/glucuronidation (E2, F1, F2, and G1); c, formation of mercapturic acids (K and L) followed by hydroxylation (B1).

Approximately equal amounts of hydroxypulegone glucuronides and hydroxymenthone/isomenthone glucuronides were the most abundant urinary metabolites. This observation was similar to the study on biliary metabolites of pulegone in which glucuronides of hydroxylated pulegone and hydroxylated reduced pulegone were principle metabolites (Thomassen et al., 1991). Four hydroxypulegone glucuronides, 4- (C1), 5- (D1), 10- (D2), and 3- (E1), were characterized in urine. 9-Hydroxypulegone glucuronide produced from acid-catalyzed isomerization of 10-hydroxypulegone glucuronide (D2) was obtained. 9-Hydroxypulegone glucuronide was not a metabolite or, at best, a minor metabolite. Glucuronidase hydrolysis of D2 gave 10-hydroxypulegone as the sole product. By contrast, glucuronidase treatment of 9-hydroxypulegone glucuronide did not give the anticipated menthofuran or any other detectable products. Metabolites E1 and C2 are related as allylic isomers. 3-Hydroxypulegone and 5-methyl-2-(1-hydroxy-1-methylethyl)-2-cyclohexen-1-one, the alcohol precursors to E1 and C2, respectively, were observed in a fungal incubation mixture of pulegone, and the latter was believed to be converted from the former through an allylic alcohol isomerization (Madyastha and Thulasiram, 1999). Alternatively, during the cytochrome P450-catalyzed oxidation of pulegone, an early step is removal of the allylic hydrogen atom on C-3; the allylic radical thus formed can be oxygenated at either end to give the isomeric alcohols. Isomeric allylic alcohol products have been reported for similar substrates (Groves and Subramanian, 1984).

Four hydroxymenthone/isomenthone glucuronides with hydroxylation at the C-5 position or the methyl groups were fully characterized (E2, F1, F2, and G1). Menthone/isomenthone along with menthofuran and pulegone were detected in plasma of rats after i.p. administration of pulegone (Thomassen et al., 1990). Reduction of the carbon-carbon double bond probably took place before hydroxylation. Reduction to menthone/isomenthone is a detoxification process. Menthone/isomenthone were shown to be neither as hepatotoxic nor as extensively bound to tissue proteins as pulegone (McClanahan et al., 1989). Two minor metabolites (hemiacetal, G3, and lactone, G2) corresponding to higher oxidation products of E2 were isolated.

The metabolic pathway proposed for metabolites M and J begins with hydroxylation of pulegone to 5-hydroxypulegone, followed by dehydration to give piperitenone (Fig. 5). Hydroxylation of piperitenone followed by dehydration and a 1,5-H+ transfer led to formation of 2-(2'-hydroxy-4'-methylphenyl)propene. Sulfation of 2-(2'-hydroxy-4'-methylphenyl)propene gave metabolite M. Previous studies have shown 2-(2'-hydroxy-4'-methylphenyl)propene to be a metabolite of piperitenone (Madyastha and Gaikwad, 1999). Hydration and oxidation of 2-(2'-hydroxy-4'-methylphenyl)propene results in the formation of metabolite J, a transformation similar to metabolism of alpha -methylstyrene to phenylpropionic acid (De Costa et al., 2001).