Journal of Chromatography B, 983–984 (2015) 39–46

Contents lists available at ScienceDirect

Journal of Chromatography B journal homepage: www.elsevier.com/locate/chromb

Metabolites identification of glycyrin and glycyrol, bioactive coumarins from licorice Qi Wang, Xue Qiao, Yi Qian, Chun-fang Liu, Yan-fang Yang, Shuai Ji, Jun Li, De-an Guo, Min Ye ∗ State Key Laboratory of Natural and Biomimetic Drugs, School of Pharmaceutical Sciences, Peking University, Beijing 100191, China

a r t i c l e

i n f o

Article history: Received 12 October 2014 Accepted 29 December 2014 Available online 12 January 2015 Keywords: Coumarin Glycyrrhiza uralensis Glycyrin Glycyrol Metabolites identification

a b s t r a c t Coumarins are an important group of bioactive constituents in licorice (Glycyrrhiza uralensis), a worldwide popular herbal medicine. This study aims to elucidate the metabolism of two major licorice coumarins, glycyrin and glycyrol in rats. After oral administration of 40 mg/kg glycyrin, neither the parent compound nor its metabolites could be detected in rats plasma or urine samples, indicating that glycyrin had poor oral bioavailability. Two hydroxylated metabolites, 4 -hydroxyl glycyrin and 5 -hydroxyl glycyrin, were detected in rat liver microsome incubation system. Among them, the major metabolite 4 -hydroxyl glycyrin, which is a new compound, was obtained by microbial transformation of Syncephalastrum racemosum AS 3.264. Its structure was fully identified by 1D and 2D NMR. Meanwhile, glycyrol, together with three metabolites, were detected in rats urine and fecal samples after oral administration (40 mg/kg). Their structures were tentatively characterized by LC/MS. Glycyrol mainly undertakes hydroxylation metabolism, accompanied by hydration and dehydrogenation as minor reactions. This is the first systematic study on metabolism of glycyrin and glycyrol. The results could be valuable to evaluate druggability of these bioactive natural products. © 2015 Elsevier B.V. All rights reserved.

1. Introduction Licorice, or Gan-Cao in Chinese, is a popular herbal medicine in China and many other countries. It is derived from the roots and rhizomes of Glycyrrhiza uralensis. Licorice is mainly used to treat cough, gastric ulcer, hepatitis, as well as pulmonary and skin diseases [1–5]. The major bioactive constituents of licorice include saponins, flavonoids, and coumarins. Glycycoumarin, glycyrin, and glycyrol are the major coumarins of licorice [6]. Significant bioactivities of these coumarins have been revealed in the recent years. For instance, glycyrin and glycyrol could inhibit HCV virus with IC50 values of 18.85 and 12.57 ␮M, respectively [7]. Glycyrin could lower the blood pressure of spontaneously hypertensive rats through PPAR-␥ binding [8], and glycyrol could suppress collageninduced arthritis in mice through decreasing NF-␬B and NFAT transcriptional activities and inhibiting IL-2 expression [9]. Moreover, glycyrol could strongly inhibit neuraminidase (IC50 3.1 ␮M), and down-regulate the inducible nitric oxide synthase at 5–50 ␮M

∗ Corresponding author at: State Key Laboratory of Natural and Biomimetic Drugs, School of Pharmaceutical Sciences, Peking University, 38 Xueyuan Road, Beijing 100191, China. Tel.: +86 10 82801516; fax: +86 10 82802024. E-mail addresses: [email protected], [email protected] (M. Ye). http://dx.doi.org/10.1016/j.jchromb.2014.12.028 1570-0232/© 2015 Elsevier B.V. All rights reserved.

[10,11]. However, little is known on oral bioavailability and in vivo metabolism of these coumarins, so far. Recently, we have reported the metabolic pathway of glycycoumarin in rats [12]. In this study, we report metabolites identification of glycyrin and glycyrol in rats and rat liver microsomes. 2. Materials and methods 2.1. Chemicals and reagents Glycyrin and glycyrol were isolated from G. uralensis Fisch. by the authors [12]. Their structures were fully characterized by NMR spectroscopy and mass spectrometry (Fig. 1). The purities were above 98% as determined by HPLC/UV analysis. HPLC-grade acetonitrile, methanol and formic acid were from Mallinkrodt Baker (Phillipsburg, NJ, USA). Ultra-pure water was prepared with a Milli-Q water purification system (Millipore, Billerica, MA, USA). Other reagents were of analytical grade. Heparin was purchased from Solarbio (Beijing, China). ˇ-Nicotinamide adenine dinucleotide phosphate hydrate (ˇ-NADP), d-glucose 6-phosphate sodium salt (G-6-P), and glucose-6-phosphate dehydrogenase (G-6-P-DE) were purchased from Sigma–Aldrich (St. Louis, MO, USA). Column chromatography was performed on silica gel (200–300 mesh, Qingdao Marine Chemical Corporation, Qingdao, China), ODS (Fuji

40

Q. Wang et al. / J. Chromatogr. B 983–984 (2015) 39–46

Fig. 1. Chemical structures of glycyrin (GLN) and glycyrol (GLL).

Silysia Chemical Ltd., Kasugai, Japan), and Sephadex LH-20 (Pharmacia Biotech AB, Sweden). 2.2. Animals and drug administration Male Sprague-Dawley rats (200 ± 20 g, body weight) were obtained from the Laboratory Animal Center of Peking University Health Science Center. The rats were bred in a cage (465 mm × 300 mm × 200 mm) in a breeding room at 25 ◦ C, 60 ± 5% humidity, and a 12-h dark-light cycle. The rats were given access to tap water and normal chow ad libitum. All the animals were bred under the above conditions for 3-day acclimation, and were then fasted overnight before the experiments. The animal facilities and protocols were approved by the Animal Care and Use Committee of Peking University Health Science Center. All procedures were based on the National Institutes of Health Guide for the Care and Use of Laboratory Animals (Institute of Laboratory Animal Resources, 1996). Animals were divided into six groups (GLN-A, GLN-B, GLLA, GLL-B, Blank-A, Blank-B), and each group had two rats (n = 2). The pure compound glycyrin or glycyrol was suspended in 0.5% carboxymethylcellulose sodium, and was orally given to rats at 40 mg/kg, respectively. For the control group, the rats were administrated with 2 mL normal saline. 2.3. Preparation of plasma, urine and fecal samples Blood (1 mL) was collected into heparinized tubes from the angular vein at 0.5, 2, 6 h for groups GLN-A, GLL-A, Blank-A, and at 1, 4, 8 h for groups GLN-B, GLL-B, and Blank-B, respectively. The blood was centrifuged at 6000 rpm for 20 min to obtain the plasma. Plasma samples of the two rats were mixed, and an aliquot of 3 mL was treated with 4 volumes of methanol to precipitate protein. The mixture was vortexed (2200 rpm) for 5 min, and centrifuged at 9000 rpm for 10 min. The supernatant was separated, dried in vacuum at 37 ◦ C, dissolved in 150 ␮L of methanol, and then filtered through a 0.22-␮m membrane. The rats were held in metabolism cages (DXL-D, Keke Medical Model Co. Ltd., Shanghai, China), and 0–24 h urine and fecal samples were collected. An aliquot of 4 mL of urine was loaded onto a pretreated SPE column (Oasis HLB, 6 mL, Waters, Milford, MA, USA), washed with 5 mL of water, and then successively eluted with 5 mL of 5% methanol and 5 mL of methanol. The methanol eluate was collected and dried in vacuum at 37 ◦ C. The residue was dissolved in 150 ␮L of methanol and filtered through a 0.22-␮m membrane for LC/MS analysis. Feces were dried in the air and then ground into a crude powder. The powder (1.0 g) was extracted with methanol (20 mL) in an ultrasonic bath for 30 min. The resulting solution was dried, and the residue was dissolved in 150 ␮L of methanol and filtered through a 0.22-␮m membrane for analysis. 2.4. Rat liver microsome incubation The rat liver microsome incubation experiments were carried out according to our recently reported procedure [13]. Glycyrin and

glycyrol were dissolved in methanol respectively, and then diluted with PBS. The incubation mixture (300 ␮L) contained rat hepatic microsomes (0.6 mg/mL), potassium phosphate buffer (pH 7.4, 0.1 mM), MgCl2 (5 mM), and nicotinamide adenine dinucleotide phosphate (NADPH)-generating system. The final concentration of glycyrin or glycyrol was 50 ␮M. The total amount of organic solvent was lower than 1% (v/v). PBS containing methanol was used as the blank control. The reaction was initiated by adding the NADPH-generating system, and was terminated by 1200 ␮L of cold acetonitrile at 4 ◦ C after 2 h. For negative control samples, PBS was added instead of NADPH-generating system. The mixture was stored at 4 ◦ C for 30 min, and the precipitated protein was removed by centrifugation (10,000 × g for 10 min at 4 ◦ C). 2.5. Microbial transformation Microbial transformation experiments were carried out according to our recent report [12]. The fungal strains were purchased from China General Microbiological Culture Collection Center. Syncephalastrum racemosum AS 3.264 was incubated at 25 ◦ C on a rotary shaker (150 rpm) in the dark, and the fermentation was carried out in 1000 mL Erlenmeyer flasks containing 400 mL potato culture medium for scaled-up biotransformation. To each flask of 2-day-old cultures, 2 mL of glycyrin (10 mg/mL in methanol, a total of 120 mg) was added. After 6 days of incubation, the cultures were pooled and filtered, and an equal volume of ethyl acetate was used to extract the supernatant. The organic layer was concentrated to dryness, and was separated by silica gel column chromatography. The column was eluted with mixtures of petroleum ether-ethyl acetate (4:1, 3:1, 2:1, 1:1, 1:2, 1:4, 1:8, v/v) to obtain seven fractions. Fr. 5 was purified by semi-preparative HPLC and eluted with acetonitrile-water (30:70, v/v) to yield GLN-M2 (18 mg). Semipreparative HPLC was performed on an Agilent 1200 instrument equipped with a YMC Pack ODS-A column (10 mm × 250 mm, 5 ␮m, YMC Co. Ltd., Japan). 2.6. HPLC/DAD/ESI-MSn analysis The analysis was performed on an Agilent series 1100 HPLC instrument connected to a Finnigan LCQ Advantage ion trap mass spectrometer (Thermo Fisher Scientific, Waltham, MA) through an ESI ion source. Samples were separated on an Agilent ZORBAX SB-C18 column (4.6 mm × 250 mm, 5 ␮m) protected with a ZORBAX Extend-C18 guard column (4.6 mm × 12.5 mm, 5 ␮m). The column temperature was 30 ◦ C. The mobile phase consisted of acetonitrile (A) and water containing 0.1% (v/v) formic acid (B). A linear gradient elution program was used as follows: 0 min, 12% A; 25 min, 100% A; 30 min, 100% A. The flow rate was 1.0 mL/min, and the effluent was introduced into the ESI source of the mass spectrometer at 0.25 mL/min via a T-union splitter. UV spectra were obtained by scanning from 200 to 400 nm. The MS instrument was operated in positive ion mode for glycyrin and negative ion mode for glycyrol, respectively. The optimized parameters of

Q. Wang et al. / J. Chromatogr. B 983–984 (2015) 39–46

41

Fig. 2. Extracted ion chromatograms for glycyrin metabolites in rat liver microsomes and microbial transformation by HPLC/(+)-ESI-MSn analysis.

glycyrin in the positive ion mode were as follows: ion spray voltage, 4.5 kV; sheath gas (nitrogen), 45 arbitrary units; auxiliary gas (nitrogen), 5 arbitrary units; capillary temperature, 340 ◦ C; capillary voltage, 10 V; and tube lens offset voltage, 50 V. The optimized parameters of glycyrol in the negative ion mode were as follows: ion spray voltage, 4.5 kV; sheath gas (nitrogen), 45 arbitrary units; auxiliary gas (nitrogen), 5 arbitrary units; capillary temperature, 320 ◦ C; capillary voltage, −30 V; and tube lens offset voltage, −30 V. Mass spectra were recorded in the range of m/z 150–1000. MSn (n = 2–4) was triggered by a datadependent threshold. The collision energy for CID was adjusted to 38%, and the isolation width for precursor ions was 2.0 mass units. Data were processed by Xcalibur 2.0.7 software (Thermo Fisher).

2.7. HPLC/ESI-IT-TOF-MS analysis High-resolution mass spectra were obtained on an IT-TOF mass spectrometer connected with an LC-20 system (Shimadzu, Tokyo, Japan). The HPLC conditions were the same as described in Section 2.6. Metabolites of glycyrin, and glycyrol were also detected in positive and negative ion modes, respectively. Collision and cooling gas was high-purity argon (Ar), and the nebulizing gas was high-purity nitrogen (N2 , 1.5 L/min). Curved desolvation line temperature, 200 ◦ C; Interface voltage, −3.5 kV (for glycyrol) or 4.5 kV (for glycyrin); detector voltage, 1.70 kV; drying gas (N2 ) pressure, 100 kPa; heat block temperature, 200 ◦ C; MS full scan range, m/z 200–1000; MSn range, m/z 100–800. Data were processed by the LCMS solution software (Shimadzu, Tokyo, Japan).

Table 1 Characterization of glycyrin (GLN) metabolites in rats by HPLC/ESI-IT-TOF-MS. HR-MS [M+H]+

No.

Formula

RT (min)

Measured

Predicted

 (ppm)

M1

C22 H22 O7

14.24

399.1435

399.1438

0.75

M2a

C22 H22 O7

14.52

399.1430

399.1438

2.02

GLNa

C22 H22 O6

20.95

383.1478

383.1489

2.87

a Identified by comparing with reference standards. ++, detected at high abundance; +, detected; –, not detected.

(+)ESI-MSn (m/z)

Metabolic reaction

Plasma

Urine

Rat liver microsomes

Microbial transformation

MS2 [399]: 350, 381 MS3 [381]: 285, 293, 313, 350 MS2 [399]: 350, 381 MS3 [381]: 285, 313, 350 MS2 [383]: 299 MS3 [299]: 243, 284

+OH

–

–

+

+

+OH

–

–

++

++

–

–

–

++

+

42

Q. Wang et al. / J. Chromatogr. B 983–984 (2015) 39–46

Fig. 3.

1

H NMR spectra (400 MHz, DMSO-d6 ) for glycyrin (GLN, A) and 4 -hydroxyl glycyrin (GLN-M2, B).

2.8. UV, IR, NMR, and HRESIMS spectra acquisition

3. Results and discussion

UV spectra were measured on a Cary 300 Bio UV–visible spectrophotometer. IR spectra were recorded on a Thermo Nicolet Nexus 470 FT-IR spectrophotometer. The HR-ESI-MS data were acquired on a Bruker APEX II Fourier transform ion cyclotron resonance (FT-ICR) mass spectrometer. 1 H and 13 C NMR experiments were performed on a Bruker Avance III instrument (400 MHz and 100 MHz, respectively) in DMSO-d6 with TMS as reference. NOE spectra were recorded on an Inova 600 MHz NMR instrument.

3.1. Metabolites identification of glycyrin After oral administration of 40 mg/kg glycyrin, neither glycyrin nor its metabolites could be detected in rats plasma or urine samples by LC/MS in the negative or positive ionization mode. This result indicated that glycyrin could hardly be absorbed after oral administration. This is consistent with our previous study, where glycyrin was not detected upon oral administration of licorice extract [14,15].

Table 2 13 C and 1 H NMR spectroscopic data for glycyrin (GLN) and its metabolite GLN-M2 (400 MHz for 1 H and 100 MHz for 13 C, in DMSO-d6 ). Position

GLN ıC

2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 1 2 3 4 5 5-OCH3 7-OCH3 2 -OH 4 -OH 4 -OH

159.9 C 119.2 C 136.1 CH 154.6 C 121.4 C 160.5 C 95.2 CH 153.3 C 107.0 C 113.2 C 156.0 C 102.7 CH 158.5 C 106.2 CH 131.5 CH 22.2 CH2 122.3 CH 131.1 C 25.4 CH3 17.6 CH3 62.9 CH3 56.4 CH3

GLN-M2 ıH (J in Hz)

7.83 s

6.87 s

6.37 d (2.0) 6.28 dd (2.0, 8.4) 7.12 d (8.4) 3.28 d (7.2) 5.11 t (7.2) 1.63 s 1.73 s 3.78 s 3.88 s 9.44 br s 9.42 br s

ıC 159.9 C 119.0 C 136.1 CH 154.6 C 119.0 C 160.5 C 95.2 CH 153.4 C 107.1 C 113.2 C 156.1 C 102.7 CH 158.5 C 106.3 CH 131.6 CH 21.7 CH2 121.4 CH 135.6 C 66.1 CH2 13.6 CH3 62.9 CH3 56.4 CH3

ıH (J in Hz)

7.85 s

6.89 s

6.37 d (2.0) 6.28 dd (2.0, 8.4) 7.12 d (8.4) 3.32 d (6.8) 5.32 t (6.8) 3.75 d (5.6) 1.70 s 3.78 s 3.88 s 9.46 br s 9.44 br s 4.65 t (5.6)

Q. Wang et al. / J. Chromatogr. B 983–984 (2015) 39–46

Scheme 1. Proposed metabolic pathway for glycyrin in rat liver microsomes. Bold arrows indicate major metabolite.

Fig. 4. Extracted ion chromatograms for rat metabolites of glycyrol by HPLC/(−)-ESI-MSn analysis.

43

44

Q. Wang et al. / J. Chromatogr. B 983–984 (2015) 39–46

Table 3 Characterization of glycyrol (GLL) metabolites in rats by HPLC/ESI-IT-TOF-MS. No.

Formula

RT (min)

HR-MS [M−H]−

Measured

Predicted

 (ppm)

M1

C21 H18 O7

14.09

381.0998

381.0980

−4.72

M2

C21 H20 O7

14.46

383.1153

383.1136

−4.44

M3

C21 H20 O7

15.07

383.1144

383.1136

−2.09

M4

C21 H18 O7

15.12

381.1000

381.0980

−5.25

C21 H16 O7

16.06

379.0842

379.0823

−5.01

C21 H18 O6

19.50

365.1047

365.1031

−4.38

M5 GLL

a

(−)ESI-MSn (m/z)

Metabolic reaction

Plasma

Urine

Feces

Rat liver microsomes

MS2 [381]: 283, 307, 351, 363 MS3 [351]: 295, 307, 337 MS2 [383]: 283, 296, 309 MS3 [283]: 239, 283 MS2 [383]: 306, 339 MS3 [339]: 324 MS2 [381]: 307, 366 MS3 [366]: 296, 307, 336, 348 MS2 [379]: 307, 351 MS2 [365]: 295, 307, 350 MS3 [307]: 175, 206, 279, 251

+OH

–

–

+

++

+OH + 2H

–

–

–

+

+OH + 2H

–

+

+

–

+OH

–

–

–

+

+OH-2H

–

–

+

+

–

–

+

++

++

a Identified by comparing with a reference standard. ++, detected at high abundance; +, detected; –, not detected.

We had discovered that hydroxylation of the isoprenyl methyl group was a major metabolic reaction for glycycoumarin, a close analog of glycyrin. This reaction was catalyzed by P450 enzymes [12]. In this work, we also studied the metabolism of glycyrin in rat liver microsomes. After 2 h incubation, two hydroxylated metabolites (GLN-M1 and M2), together with glycyrin per se, could be detected by LC/MS analysis (Fig. 2). High-resolution mass spectra of GLN-M1 and M2 presented [M+H]+ signals at m/z 399.1435 and 399.1430, respectively (Table 1). Their molecular formulas were established as C22 H22 O7 , one more oxygen atom than glycyrin, indicating that a hydroxyl group was added. The MS/MS spectrum of m/z 399 showed two fragment ions at m/z 381 ([M+H−H2 O]+ ) and m/z 350 ([M+H−H2 O−CH3 O·]+ ) (Fig. S1). The neutral loss of 31 Da at (+)-ESI mode was similar to the neutral loss of 30 Da at (−)-ESI mode for hydroxylated glycycoumarin [12], indicating that the hydroxylation occurred on the isoprenyl methyl group. However, position of the hydroxylation (4 or 5 ) could not be identified by mass spectra. Subsequently, we used microbial transformation system to obtain GLN-M1 and M2, since microbial catalysis could produce metabolites similar to those of mammals [16,17]. Fortunately, after screening a number of fungal strains and analyzing their products by LC/MS, we found that S. racemosum AS 3.264 could produce M1 and M2 (Fig. 2). We then performed a scaled-up biotransformation reaction using 120 mg of glycyrin as the substrate. After 6 days of

incubation, the cultures were pooled and filtered, and the supernatant was separated by silica gel column and semi-preparative HPLC. Finally, 18 mg of M2 was obtained as yellow amorphous powder, showing blue fluorescent spot in ultraviolet light at 365 nm. However, M1 was not obtained due to its low amount. The NMR spectra for M2 were similar to those of glycyrin (Supplemental Spectra). Comparing the 1 H NMR data of M2 with those of glycyrin, the signal for H-4 (ıH 1.63) was missing, and a new methylene signal appeared at ıH 3.75. In addition, a new resonance for 4 -OH appeared at ıH 4.65 (Fig. 3). According to the 13 C NMR data, 5 -CH3 shifted upfield by 4.0 ppm due to the -gauche effect of the hydroxyl group. Furthermore, the carbon signal for 4 -CH3 (ıC 25.4) disappeared, and a new methylene resonance was observed at ıC 66.1 (Table 2). At the same time, DEPT 135 spectra showed an additional methylene signal than glycyrin. The extra hydroxyl group was established by the HMBC correlations of C-2 (ıC 121.4) and H-4 (ıH 3.75); as well as C-3 (ıC 135.6) and H4 (ıH 3.75). HSQC spectra showed that the ıH 3.75 was directly connected to ıC 66.1. The NOE enhancement between H-4 (ıH 3.75) and H-2 (ıH 5.32) confirmed the hydroxyl group on C-4 , and elucidated the trans-relationship of 5 -CH3 and H-2 (Supplemental Spectra). Thus, M2 was identified as 4 -hydroxyl glycyrin. It is a new compound. Similar to the metabolism of glycycoumarin, the hydroxylation occurred at the trans-terminal methyl group was

Scheme 2. Proposed metabolic pathways of glycyrol in vivo and in vitro. Bold arrows indicated major metabolites. U, detected in urine; F, detected in feces; RLM, detected in rat liver microsomes.

Q. Wang et al. / J. Chromatogr. B 983–984 (2015) 39–46

45

more than cis product [12,18,19]. Therefore, M1 was identified as 5 -hydroxyl glycyrin. The metabolic pathway of glycyrin in rat liver microsomes was proposed as shown in Scheme 1. 3.2. Metabolites identification of glycyrol Similar to glycyrin, glycyrol could not be detected in rats plasma after oral administration of a 40 mg/kg dose. However, glycyrol could be detected in urine samples, together with a minor metabolite GLL-M3. Glycyrol was detected in the fecal samples in high abundance, together with minor metabolites GLL-M1, M3, and M5 (Fig. 4, Table 3). Structures of these metabolites were characterized by LC/MS analysis. High-resolution mass spectra of GLL-M1 exhibited an [M−H]− signal at m/z 381.0998, and its molecular formula was established as C21 H18 O7 (calcd. 381.0980 for [M−H]− ). Upon collision-induced dissociation, the [M−H]− ion at m/z 381 produced the fragment ion at m/z 351 ([M−H−CH2 O]− ), m/z 307 ([M−H−C3 H7 O−CH3 ·]− ), and m/z 295 ([M−H−C4 H7 O−CH3 ·]− ) (Fig. 5). The m/z 307 and 295 fragments were the same as those of glycyrol, indicating that the isopentenyl group was hydroxylated. Similar to the metabolic routes of glycyrin and glycycoumarin [12], we proposed that M1 was hydroxylated at the trans-terminal methyl group. HRMS spectrum established the molecular formula of GLL-M3 as C21 H20 O7 ([M−H]− ion at m/z 383.1144), which has two more hydrogen atoms than M1, indicating that M3 was a hydrated product of glycyrol. The MS/MS spectrum of M3 showed a fragment ion at m/z 339 ([M−H−C3 H8 ]− ). As shown in Fig. 5, the m/z 339 suggested that the hydrogenation, but not the hydroxylation, occurred on the isopentenyl group. The hydroxyl position could not be assigned due to limited structural information. In HPLC/IT-TOF-MS analysis, GLL-M5 gave an [M−H]− signal at 379.0842 (C21 H16 O7 ), indicating that it was a dehydrogenated product of GLL-M1 (C21 H18 O7 ). In tandem mass spectra, the [M−H]− ion at m/z 379 produced fragment ions m/z 351 ([M−H−CO]− ) and m/z 307 ([M−H−CO−CH3 −C2 H5 ·]− ). The m/z 307 suggested that GLL-M5 had the same backbone as M1. Considering the highly unsaturated structure of coumarin backbones, the dehydrogenation should occur on the side chain, where the hydroxyl group was transformed into carbonyl group (Fig. 5). Finally, we proposed the metabolic pathway of glycyrol in rats after oral administration (Scheme 2). We had also tried to prepare the hydroxylated metabolites of glycyrin by microbial transformation. However, no microbial strains we screened could produce these metabolites. In order to investigate whether the metabolites were transformed through P450 enzymes catalysis, glycyrol was incubated with rat liver microsomes for 2 h, and then analyzed by LC/MS. We found that M1 and M5 could be detected in the extracted ion chromatogram of m/z 381 and m/z 379, respectively (Fig. 4). Meanwhile, we discovered two minor metabolites (GLL-M2 and M4) in the incubation mixture. The HRMS spectrum of GLL-M2 established its molecular formula as C21 H20 O7 ([M−H]− m/z 383.1153), indicating that it was also a hydration product. Upon collision-induced dissociation, m/z 383 produced fragment ions at m/z 309 ([M−H−CH3 ·−C3 H7 O]− ), m/z 296 ([M−H−CH3 ·−C4 H8 O]− ), and m/z 283 ([M−H−CH3 ·−C5 H9 O]− ) (Fig. 5). The m/z 309 fragment was comparable to m/z 307 of glycyrol, indicating that the hydration took place at C-3 , C-4 or C-5 on the isoprenyl group. A proposed structure of M2 is shown in Fig. 5. It was assigned as a hydrogenated product of trans-hydroxyl glycyrol. GLL-M4 was a hydroxylated product of glycyrol. It displayed an [M−H]− ion at m/z 381.1000 (calcd for C21 H18 O7 ), and yield of a major product ion at m/z 366 ([M−H−CH3 ·]− ). In the MS3 spectrum, the fragment ions at m/z 336, m/z 307 and m/z 296 were detected,

Fig. 5. Tandem mass spectra of glycyrol and its metabolites.

which were different from those of M1. Therefore, we deduced that the hydroxylation did not occur at the isoprenyl group. A metabolic pathway of glycyrol in rat liver microsomes was proposed as shown in Scheme 2. 3.3. Metabolic pathway of glycyrin and glycyrol In this work, the metabolism of glycyrin and glycyrol in rats was studied. After oral administration of 40 mg/kg glycyrin,

46

Q. Wang et al. / J. Chromatogr. B 983–984 (2015) 39–46

neither the unchanged form nor its metabolites could be detected in rats plasma and urine. We speculated that glycyrin had very poor absorption. After intravenous administration (10 mg/kg), glycyrin could be detected in rats plasma, heart, liver and kidney samples after 0.25 h (Fig. S2). Although glycyrin could be metabolized into hydroxylated products in rat liver microsome incubation system, no metabolites were detected in the rat plasma or tissue samples, probably due to their fairly low amounts. Glycyrol could be detected in rats urine and fecal samples, together with its hydroxylated (GLL-M1), hydrated (GLL-M3), and dehydrogenated (GLL-M5) metabolites. Glycyrol occurred mainly in its unchanged form, and its metabolites only accounted for a small portion (less than 20%). GLL-M1 and M5, along with two other metabolites, were proved to be catalyzed by rat liver microsomes. Majority of glycyrol and its metabolites were eliminated through feces, and only a low amount went into blood circulation. Accordingly, following intravenous administration (10 mg/kg), glycyrol and its two metabolites, GLL-M1 and M3, were detected in rats plasma at 0.25 h (Fig. S3). They were consistent with the metabolites we detected in urine and feces. By comparing the metabolism of glycyrin and glycyrol to their close analog glycycoumarin [12], we found all the three arylcoumarins showed high metabolic stability. They existed in biosamples (feces, urine, plasma, and rat liver microsomes) mainly in their unchanged form. However, oral bioavailabilities of glycyrin and glycyrol were much lower than that of glycycoumarin, which was detected in plasma and urine at remarkable amounts. The presence of 4,2 -furan ring (glycyrol) and 7-OCH3 (glycyrin) may affect the oral bioavailability of licorice coumarins. Hydroxylation of the isoprenyl methyl group was the major metabolic reaction for glycyrin, glycyrol, and glycycoumarin. The trans-methyl group was more readily to be hydroxylated than the cis-methyl group. Minor metabolic reaction of hydration and hydrogenation were also commonly seen for the three arylcoumarins. 4. Conclusion The metabolism of glycyrin and glycyrol in rats was studied. These two coumarins showed poor oral bioavailability and high metabolic stability. Neither the parent compounds nor their metabolites were detected in rats plasma after oral administration (40 mg/kg). Rat liver microsome incubation indicated that glycyrin could be hydroxylated at the isoprenyl methyl group by P450 enzymes. The major metabolite, 4 -hydroxyl glycyrin, was isolated as a new compound from the microbial transformation of S. racemosum AS 3.264. A small portion of glycyrol could be metabolized into hydroxylated, hydrated or dehydrogenated products in vivo, which were catalyzed by P450 enzymes. The results obtained in this study could be useful to evaluate druggability of these bioactive natural products. Acknowledgements This work was supported by National Natural Science Foundation of China (No. 81173644 and No. 81222054), and the Program for New Century Excellent Talents in University from Chinese Ministry of Education (No. NCET-11-0019).

Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.jchromb. 2014.12.028. References [1] A. Agarwal, D. Gupta, G. Yadav, P. Goyal, P.K. Singh, U. Singh, An evaluation of the efficacy of licorice gargle for attenuating postoperative sore throat: a prospective, randomized, single-blind study, Anesth. Analg. 109 (2009) 77–81. [2] C.H. Ma, Z.Q. Ma, X.L. Liao, J.P. Liu, Q. Fu, S.P. Ma, Immunoregulatory effects of glycyrrhizic acid exerts anti-asthmatic effects via modulation of Th1/Th2 cytokines and enhancement of CD4+ CD25+ Foxp3+ regulatory T cells in ovalbumin-sensitized mice, J. Ethnopharmacol. 148 (2013) 755–762. [3] M.N. Asl, H. Hosseinzadeh, Review of pharmacological effects of Glycyrrhiza sp. and its bioactive compounds, Phytother. Res. 22 (2008) 709–724. [4] C. Fiore, M. Eisenhut, E. Ragazzi, G. Zanchin, D. Armanini, A history of the therapeutic use of liquorice in Europe, J. Ethnopharmacol. 99 (2005) 317–324. [5] X.Y. Wang, H. Zhang, L.L. Chen, L.H. Shan, G.W. Fan, X.M. Gao, Liquorice, a unique guide drug of traditional Chinese medicine: a review of its role in drug interactions, J. Ethnopharmacol. 150 (2013) 781–790. [6] X. Qiao, C.F. Liu, S. Ji, X.H. Lin, D.A. Guo, M. Ye, Simultaneous determination of five minor coumarins and flavonoids in Glycyrrhiza uralensis by solid phase extraction and high-performance liquid chromatography/electrospray ionization tandem mass spectrometry, Planta Med. 80 (2014) 237–242. [7] M. Adianti, C. Aoki, M. Komoto, L. Deng, I. Shoji, T.S. Wahyuni, M.I. Lusida, Soetjipto, H. Fuchino, N. Kawahara, H. Hotta, Anti-hepatitis C virus compounds obtained from Glycyrrhiza uralensis and other Glycyrrhiza species, Microbiol. Immunol. 58 (2014) 180–187. [8] T. Mae, H. Kishida, T. Nishiyama, M. Tsukagawa, E. Konishi, M. Kuroda, Y. Mimaki, Y. Sashida, K. Takahashi, T. Kawada, K. Nakagawa, M. Kitahara, A licorice ethanolic extract with peroxisome proliferator-activated receptorgamma ligand-binding activity affects diabetes in KK-Ay mice, abdominal obesity in diet-induced obese C57BL mice and hypertension in spontaneously hypertensive rats, J. Nutr. 133 (2003) 3369–3377. [9] Y.X. Fu, H.L. Zhou, S.Y. Wang, Q. Wei, Glycyrol suppresses collagen-induced arthritis by regulating autoimmune and inflammatory responses, PLOS ONE 9 (2014) e98137. [10] Y.B. Ryu, J.H. Kim, S.J. Park, J.S. Chang, M.C. Rho, K.H. Bae, K.H. Park, W.S. Lee, Inhibition of neuraminidase activity by polyphenol compounds isolated from the roots of Glycyrrhiza uralensis, Bioorg. Med. Chem. Lett. 20 (2010) 971–974. [11] E.M. Shin, H.Y. Zhou, L.Y. Guo, J.A. Kim, S.H. Lee, I. Merfort, S.S. Kang, H.S. Kim, S. Kim, Y.S. Kim, Anti-inflammatory effects of glycyrol isolated from Glycyrrhiza uralensis in LPS-stimulated RAW264.7 macrophages, Int. Immunopharmacol. 8 (2008) 1524–1532. [12] Q. Wang, X. Qiao, C.F. Liu, S. Ji, L.M. Feng, Y. Qian, D.A. Guo, M. Ye, Metabolites identification of glycycoumarin, a major bioactive coumarin from licorice in rats, J. Pharm. Biomed. Anal. 98 (2014) 287–295. [13] X. Qiao, S. Ji, S.W. Yu, X.H. Lin, H.W. Jin, Y.K. Duan, L.R. Zhang, D.A. Guo, M. Ye, Identification of key licorice constituents which interact with cytochrome P450: evaluation by LC/MS/MS cocktail assay and metabolic profiling, AAPS J. 16 (2014) 101–113. [14] C. Xiang, X. Qiao, Q. Wang, R. Li, W.J. Miao, D.A. Guo, M. Ye, From single compounds to herbal extract: a strategy to systematically characterize the metabolites of licorice in rats, Drug Metab. Dispos. 39 (2011) 1597–1608. [15] X. Qiao, M. Ye, C. Xiang, Q. Wang, C.F. Liu, W.J. Miao, D.A. Guo, Analytical strategy to reveal the in vivo process of multi-component herbal medicine: a pharmacokinetic study of licorice using liquid chromatography coupled with triple quadrupole mass spectrometry, J. Chromatogr. A 1258 (2012) 84–93. [16] G.P. Rao, P.J. Davis, Microbial model of mammalian metabolism: biotransformations of HP 749 (besipirdine) using Cunninghamella elegans, Drug Metab. Dispos. 25 (1997) 709–715. [17] A. Schmid, J.S. Dordick, B. Hauer, A. Kiener, M. Wubbolts, B. Witholt, Industrial biocatalysis today and tomorrow, Nature 409 (2001) 258–268. [18] J. Guo, D. Liu, D. Nikolic, D. Zhu, J.M. Pezzuto, R.B. van Breemen, In vitro metabolism of isoliquiritigenin by human liver microsomes, Drug Metab. Dispos. 36 (2008) 461–468. [19] D. Nikolic, Y.M. Li, L. Chadwick, G. Pauli, R. Breemen, Metabolism of xanthohumol and isoxanthohumol, prenylated flavonoids from hops (Humulus lupulus L.), by human liver microsomes, J. Mass Spectrom. 40 (2005) 289–299.