Characterization of the metabolites of irisflorentin by using ultra-high performance liquid chromatography combined with quadrupole/orbitrap tandem mass spectrometry

Xiao Zhanga, Gao-Xing Qiaoa, Gao-Feng Zhaoc,∗∗, Song-Feng Zhaoa,b,∗

Abstract

Irisflorentin is one of the bioactive constituents from the root of Belamcanda chinensis (L.) DC, which displayed anti-inflammatory and anti-tumor activities. In this work, the in vitro metabolism of irisflorentin was investigated using liver microsomes and hepatocytes. The metabolites were identified by ultra-high performance liquid chromatography combined with quadrupole/orbitrap tandem mass spectrometry. Under the current conditions, a total of 11 metabolites were detected and structurally identified according to accurate masses, fragment ions and retention times. Metabolite M10, identified as 6,7-Metabolite characterization Cytochrome P450 nuclear magnetic resonance spectroscopy. The metabolic pathways of irisflorentin included oxidation, demethylation and glucuronidation. M10 was the most abundant metabolite in all tested species. Further phenotyping studies revealed that -naphthoflavone and ketoconazole displayed significant inhibitory effect on the formation of M10. Cytochrome P450 (CYP) 1A2 and 3A4 were the major enzymes responsible for the formation of M10 by using individual recombinant human CYP450 enzymes. For the first time the current study provides an overview of the in vitro metabolic fates of irisflorentin, which is helpful for us to predict the human metabolism and the potential drug-drug interactions caused by irisflorentin.

Keywords:
dihydroxy-5,3-tetramethoxy isoflavone, was biosynthesized and unambiguously characterized by

1. Introduction

Belamcanda chinensis (L.) DC, also known as Shegan, was widely used as a traditional Chinese medicine for hundreds of years in China [1,2]. It was prescribed for the treatment of inflammation, asthma and throat diseases in clinic [1–3]. In the past few decades, more than 100 bioactive constituents have been isolated from the the root of B. chinensis. Isoflavonoids represent the major constituents in this herb medicine [4,5], which have a broad spectrum of biological and pharmacological activities, such as antioxidant and antibacterial [2,6,7]. Irisflorentin is one of the most abundant isoflavonoids presented in this herb, which is officially used as a marker of quality control for B. chinensis [1]. Irisflorentin has been demonstrated to be an immunotherapeutic adjuvant through modulating the property of mouse bone marrow-derived dendritic cells and reducing the allergic contact hypersensitivity response [8]. In addition, irisflorentin showed therapeutic effect on Parkinson’s disease through improving -synuclein accumulation and attenuating 6-hydroxydopamine-induced dopaminergic neuron degeneration[9].
To the best of our knowledge, the study regarding the metabolism of irisflorentin is very limited. Drug metabolism raises the necessity to study the metabolic property to support preclinical pharmacokinetic and toxicity studies [10]. Jia and his co-workers have investigated the metabolism of irisflorentin in rat liver microsomes and five phase I metabolites were identified [11], which did provide some valuable information on the metabolism of irisflorentin. However, this study cannot answer the following questions: does irisflorentin show species-dependent metabolism? Which animal has the comparable metabolic profiles to human? Does irisflorentin undergo bioactivation? Which enzyme is responsible for irisflorentin metabolism?
Hence, the current study was aimed 1) to identify and profile the metabolites of irisflorentin in rat, dog and human liver microsomes and hepatocytes by using ultra-high performance liquid chromatography combined with quadrupole/orbitrap mass spectrometry (UPLC-Q/Orbitrap-HRMS); 2) to propose the metabolic pathway and 3) to disclose the human enzyme(s) responsible for the metabolism of irisflorentin.

2. Experimental

2.1. Chemicals and reagents

Irisflorentin (purity > 98 %) was supplied by Chengdu Alfa Biotechnology Co., Ltd. -Naphthoflavone, sulfaphenazole, ticlopidine, quinidine, ketoconazole, glutathione (GSH) and MgCl2 were purchased from Sigma-Aldrich. -Nicotinamide adenine dinucleotide phosphate tetrasodium salt (NADPH) was obtained from Acros Organics. Pooled rat liver microsomes (n = 50), dog liver microsomes (n = 10), human liver microsomes (n = 150) and recombinant human CYP450 1A2, 2B6, 2C8, 2C9, 2C19, 2D6 and 3A4 were obtained from BD Gentest. Pooled rat hepatocytes (n = 10), dog hepatocytes (n = 5), and human hepatocytes (n = 5) were purchased from XenoTech. Acetonitrile and methanol were of HPLC grade and purchased from Fisher Scientific. All other chemicals and reagents (analytical grade) were commercial available.

2.2. Incubation with liver microsomes

Irisflorentin was initially dissolved in methanol at a concentration of 100 mM. The incubation mixtures contained irisflorentin (20 M), microsomes (1 mg protein/mL), NADPH (1 mM), MgCl2 (3 mM), and 100 mM phosphate buffer (pH 7.4). The total incubation volume was 200 L and the organic solvent was less than 0.5 % (v/v). Incubations without NADPH served as negative controls. The reactions were performed at 37 ◦C in a water bath and acetonitrile (1 mL) was added to terminate the reactions after 60 min. The samples were vortexed and centrifuged at 14 000 g (at 4 ◦C) for 5 min, and the supernatants were then dried under nitrogen gas and the residues were re-dissolved with 200 L of acetonitrile-water solution (v/v, 1:9). After centrifugation at 14000 g (at 4 ◦C) for 5 min, 2 L of the sample was injected into UPLC-Q/Orbitrap-HRMS system. To trap the reactive metabolites, GSH (3 mM) was used as a trapping agent and was included in the microsomal incubation. The other procedures were similar as described above.

2.3. Incubation with hepatocytes

All of the incubations were carried out at 37 ◦C in a humidified CO2 incubator for 2 h in Williams’ E medium containing irisflorentin (50 M) and hepatocytes (1 million cells/mL). The organic solvent in the incubation was 0.5 % (v/v). The total incubation volume was 200 L. Incubations without irisflorentin served as blank controls. Reactions were terminated with 1 mL of acetonitrile. The samples were then centrifuged at 14000 g (at 4 ◦C) for 5 min and the supernatants were dried under nitrogen gas. The residues were reconstituted with 200 L of acetonitrile-water solution (v/v, 1:9) and 2 L of the sample was injected into UPLC-Q/Orbitrap-HRMS system.

2.4. Biosynthesis of metabolite M10

To further characterize the structure of the metabolite M10, the reference standard of M10 was biosynthesized by using rat liver microsomes. The total volume of incubation was 300 mL, containing irisflorentin (100 M), NADPH (1 mM), rat liver microsomes (0.5 mg protein/mL), MgCl2 (3 mM) and 0.1 M phosphate buffer (pH 7.4). The reaction was carried out 37 ◦C in water bath. After incubation for 2 h, 300 mL of methanol was added to terminate the reaction and the mixture was centrifuged at 5000 g (at 4 ◦C) for 10 min. The resulting supernatant was evaporated to dryness in vacuum. And the residue was re-dissolved with 15 mL methanolwater solution (v/v; 1:1). Afterwards, M10 was purified using a semi-preparative HPLC system. Chromatographic separation was performed on an Agilent ZORBAX SB C18 column (250 × 10 mm, i.d. 5 m) maintained at 40 ◦C with methanol and water (35: 65, v/v) as mobile phase, at a flow rate of 3 mL/min. The ultraviolet detector was set at 254 nm. Finally, 5.69 mg of reference standard of M10 was obtained. The standard was dissolved in CDCl for 13C-NMR analysis, which was recorded on a Bruker AV 600 NMR spectrometer (Bruker, Germany).

2.5. Microsomal incubations in the presence of specific chemical inhibitors

To disclose the human CYP450 isozyme(s) responsible for the formation of M10, a variety of specific chemical inhibitors were individually involved in the human liver microsomes, including -naphthoflavone (2 M for CYP1A2), ketoconazole (10 M for CYP3A), quinidine (10 M for CYP2D6), sulfaphenazole (30 M for CYP2C9), ticlopidine (100 M for CYP2B6 and CYP2C19) and montelukast (10 M for CYP2C8). The microsomal incubation (100 L) contained human liver microsomes (0.5 mg protein/mL), irisflorentin (10 M), NADPH (1 mM), MgCl2 (3 mM) and individual CYP450 inhibitors. Each incubation was prepared in duplicate. Control samples without inhibitors were prepared in the same manner. After incubation for 30 min, the reactions were stopped by adding 200 L of acetonitrile containing 100 ng/mL of internal standard (kaempferol). The samples were then centrifuged at 14,000 g (at 4 ◦C) for 10 min and the supernatants (2 L) were submitted to UPLC-Q/Orbitrap-HRMS for analysis.

2.6. Formation of M10 by human recombinant CYP450s

To further clarify the human CYP450 isozyme(s) participating in the formation of M10, recombinant human CYP450 enzymes including CYP 1A2, 2B6, 2C8, 2C9, 2C19, 2D6 and 3A4 were used. The total incubation volume was 50 L. The incubation mixtures contained individual CYP450 isozymes (20 pmol/mL), irisflorentin (2 M), NADPH (1 mM) and MgCl2 (3 mM). Each incubation was prepared in duplicate. After a 30-min incubation at 37 ◦C, the reactions were terminated by addition of 100 L of acetonitrile containing 100 ng/mL of internaal standard. After centrifugation at 14,000 g (at 4 ◦C) for 10 min, 2 L of the supernatant was analyzed by UPLC-Q/Orbitrap-HRMS for the monitoring the production of M10. The relative contribution of each CYP to the formation of M10 was evaluated by the following equation: where the expression abundance in native human liver microsomes were 45, 39, 64, 96, 19, 10 and 108 pmol CYP/mg human liver microsome for CYP 1A2, 2B6, 2C8, 2C9, 2C19, 2D6 and 3A4, respectively [12].

2.7. Analytical conditions

The bioanalysis was conducted on Dionex U3000 ultra-high performance liquid chromatography (UPLC) system equipped with an ACQUITY BEH C18 column (100 × 2.1 mm, i.d. 1.7 m, Waters Corp.) thermostated at a temperature 40 ◦C. The mobile phase consisted of 0.1 % formic acid in water (A) and acetonitrile (B), with gradient elution as following: 5 % B at 0−1 min, 5–35 % B at 1−5 min, 35–45 % B at 5−10 min, 45–85 % B at 10−14 min, and 10 % B at 14−16 min. The flow rate was 0.3 mL/min.
Mass detection was completed on a Q-Exactive-Orbitrap high resolution mass spectrometer (Thermo Fisher Scientific) connected to UPLC system via a positive electrospray ionization (ESI) interface. The source conditions were optimized as following: spray voltage 3.0 kV; capillary temperature 250 ◦C; sheath gas flow rate 40 arb; aux gas flow rate 10 arb; S-lens voltage 50 V. Data were recorded from m/z 100–1000 Da in centroid mode. Collision energy was set at 25−30 V. Instrumental control and data acquisition were achieved by using Xcalibur software (version 2.3.1, Thermo Fisher Scientific, USA).

3. Results and discussion

3.1. UPLC–MS/MS analysis of the metabolites

Liver microsomes and hepatocytes represent the most important tools for drug metabolism study because they retain the abundant drug-metabolizing enzymes. In the present work, irisflorentin was incubated with liver microsomes and hepatocytes (rat, dog and human) and the samples were analyzed by UPLCQ/Orbitrap-HRMS and the post-data processing was achieved with Compound Discoverer software based on mass defect filter function. The combined extracted ion chromatograms (cEICs) of the metabolites from different matrices were shown in Figs. 1–3, which provided overall metabolite profiles from different species. With the HRMS, the accurate masses, fragment ions and the elemental compositions of the metabolites were obtained and summarized in Table 1. According to the accurate masses and the elemental compositions, these metabolites mainly derived from oxidation, demethylation and glucuronidation. No GSH conjugates were detected, suggesting that irisflorentin was devoid of bioactivation although alert structure was present in the molecule of irisflorentin. It can be seen from Fig. 1–3, a total of 11 metabolites, including 9 new metabolites (M1-M6, M8, M9 and M11), were detected and M10 (6,7-dihydroxy-5,3 -tetramethoxy isoflavone) was the most abundant metabolite in all matrices. In rat and human hepatocytes, M4 (glucuronidation metabolite) was also the major metabolite. No human-specific metabolites were found. The structural elucidation was described below.

3.2. Structural elucidation

3.2.1. Fragmentation of the parent
To characterize the structure of the metabolite, the fragmentations of irisflorentin was initially investigated. Irisflorentin showed protonated ion [M+H]+ at m/z 387.1077 (Calcd m/z 387.1074),
with the elemental composition of C20H18O8. The MS/MS spectrum along with the proposed fragmentations were shown in Fig. 4, which afforded three main fragment ions at m/z 372.0839 (loss of methyl radical), 357.0607 (loss of two methyl radicals) and 341.0658 (loss of oxygen atom from m/z 357.0607). The fragment ion at m/z 220.0368 was formed by the loss of trimethoxybenzene moiety. The characteristic fragment ion at m/z 195.0290 was derived from the Retro-Diels-Alder fragmentation of the C ring. These fragmentations provided some structural information, which facilitated the characterization of the metabolites.

3.2.2. Metabolites M1, M2, M3, M6 and M9

M1, M2, M3, M6 and M9 were identified as glucuronidation associated products, which were detected at 4.61, 5.00, 5.35, 6.32 and 6.96 min, respectively. They displayed the same protonated ion [M+H]+ at m/z 537.12, with the elemental composition atC24H24O14, suggesting that these metabolites were from oxidation (ring opening of 1,3-benzodioxole) followed by demethylation and glucuronidation. The MS/MS spectrum (Fig. S1) afforded a predominant fragment ion at m/z 361.0919, which was generated by the loss of glucuronyl moiety (−176 Da), a characteristic neutral loss of glucuronide associated metabolite. However, site for demethylation could not be accurately located in the present study.

3.2.3. Metabolites M4 and M8

M4 and M8 were detected at 5.77 and 6.64 min, respectively, with the identical protonated ion [M+H]+ at m/z 551.14. The elemental composition was proposed as C23H22O14, suggesting that they were glucuronide conjugates of M10. MS/MS spectrum (Fig. S1) afforded a characteristic fragment ion at m/z 375.1075, which was generated through the cleavage of glucuronyl moiety (−176 Da), which further suggested that these metabolites derived from glucuronidation of M10.

3.2.4. Metabolites M5 and M7

M5 and M7 were eluted at the retention times of 6.14 and 6.42 min, respectively. They had an identical protonated ion [M+H]+ at m/z 361.09. The elemental composition of C18H16O8suggested that M5 and M7 were demethylated products of M10. The MS/MS spectrum of M5 (Fig. S2) provided two characteristic fragment ions at m/z 346.0685 and 331.0449, which were formed through the cleavage of -CH and -2CH from the [M+H]+ ion, respectively. An indicative fragment ion at m/z 169.0133, 14 Da lower than that of M10, was formed through the Retro-Diels-Alder fragmentation of C ring, which demonstrated that the demethylation occurred at C-5 position (A ring). The MS/MS spectrum of M7 (Fig. S2) provided two characteristic fragment ions at m/z 346.0685 and 331.0449, which were formed through the cleavage of -CH3 and -2CH3, respectively. A characteristic fragment ion at m/z 208.0368 demonstrated that demethylation occurred at B ring. The fragment ion at m/z 183.0188 resulted from Retro-Diels-Alder fragmentation of C ring and this ion further produced the fragment ion at m/z 153.0185 through -OCH3.

3.2.5. Metabolite M10

M10 was the most abundant metabolite in all tested species and it was detected at 7.44 min. It had a protonated ion [M+H]+at m/z 375.11 (elemental composition C19H18O8), 12 Da lower than that of the parent, suggesting the ring opening of 1,3-benzodioxole. MS/MS spectrum (Fig. S3) displayed three characteristic fragment ions at m/z 360.0840 (-CH3), 345.0607(-2CH3) and 317.0656 (2CH3-CO). The fragment ions at m/z 183.0288 was formed through the Retro-Diels-Alder fragmentation of C ring. To further characterize this metabolite, the reference standard was bio-synthesized and purified. The structure was further characterized by 13C-NMR data[153.1 (C-2), 124.4 (C-3), 174.1 (C-4), 145.7 (C-5), 137.5 (C-6), 152.6(C-7), 99.3 (C-8), 152.1 (C-9), 112.1 (C-10), and C-6 (C-5-CH3), -CH and The NMR data revealed the disappearance of methylene acetal group at 103.5. Hence, M10 was identified as 6,7-dihydroxy-5,3 -tetramethoxy isoflavone.

3.2.6. Metabolite M11

M11 was detected at 8.56 min. It had a protonated ion [M+H]+ at m/z 373.09 (elemental composition C19H16O8), 14 Da lower than that of the parent, suggesting the occurrence of demethylation. MS/MS spectrum (Fig. S3) displayed two characteristic fragment ions at m/z 358.0683 (-CH3) and 343.0450 (-2CH3). The fragment ions at m/z 181.0133 was formed through the Retro-DielsAlder fragmentation of C ring, which suggested that demethylation occurred at C-5 position (A ring).

3.3. Human CYP450 enzymes responsible for the formation of M10

Sulfaphenazole, ticlopidine, montelukast and quinidine showed no inhibition on the formation of M10, as shown in Fig. 5a. While, -naphthoflavone (a specific inhibitor of CYP1A2) and ketoconazole (a specific inhibitor of CYP3A4) displayed significant inhibitory effect on the formation of M10. The remaining reaction activity for the formation of M10 caused by -naphthoflavone and ketoconazole were 55 % and 47 %, respectively. These data suggested that CYP1A2 and 3A4 may be involved in the formation of M10. To further study the enzyme(s) responsible for the formation of M10, recombinant human CYP450 enzymes were used and the results demonstrated that CYP1A2 and 3A4 were the major enzymes responsible for the formation of M10. The relative contribution of CYP1A2 and 3A4 to the generation of M10 were 52 % and 43 %, respectively (Fig. 5b).

3.4. Metabolic pathways

Based on the identified metabolites, the metabolic pathways of irisflorentin were accordingly proposed, as shown in Fig. 6. It appeared that the ring opening of 1,3-benzodioxole was the predominant metabolic pathway both in liver microsomes and hepatocytes in all tested species. It has been well documented that on the one hand oxidative ring opening of 1,3-benzodioxole resulted in carbene intermediate, which can bind to CYPs, with the result of time-dependent inhibition of CYPs [13,14]. Paroxetine served as an example in this regard [13,14]. On the other hand, oxidative ring opening of 1,3-benzodioxole produced the catechol derivative. This derivative was readily susceptible to oxidation to form ortho-quinone intermediate, which can react with GSH to form stable adduct [14,15]. In the current work, GSH was used to trap the potential reactive metabolite; however, no conjugate was detected, which suggested that no bioactivation occurred. In addition to oxidative ring opening of 1,3-benzodioxole, demethylation was also found as the metabolic pathway of irisflorentin. Glucuronidation was also found as the major phase II metabolic pathway. Although glucuronidation raises limited safety concerns, it is important to disclose the enzyme(s) responsible for the glucuronidation of irisflorentin, which aids in predicting the potential of drug-drug interactions mediated by glucuronosyltransferases. Therefore, our next work will focus on the phase II metabolism of irisflorentin and the potential drug-drug interactions caused by irisflorentin.

4. Conclusions

Taken together, the in vitro metabolism of irisflorentin was studied using hepatocytes and liver microsomes. The metabolites were profiled and identified using UPLC-Q/Orbitrap-HRMS and a total of 11 metabolites were identified. M10 unambiguously identified as 6,7-dihydroxy-5,3 -tetramethoxy isoflavone by NMR analysis was the most abundant metabolite in all tested species. No GSH adduct was detected. Furthermore, CYP1A2 and 3A4 were the major enzymes participating in the formation of M10. For the first time the current study provides an overview of the metabolism of irisflorentin, which is helpful for us to understand the actions caused by irisflorentin.

References

[1] Chinese Pharmacopoeia Committee, Pharmacopoeia of the People’s Republic of China, Vol. 1, China Medical Science Press, Beijing, 2020.
[2] L. Zhang, K.H. Wei, J.P. Xu, D.W. Yang, C.H. Zhang, Z.P. Wang, M.H. Li, Belamcanda chinensis (L.) DC-An ethnopharmacological, phytochemical and pharmacological review, J. Ethnopharmacol. 186 (2016) 1–13.
[3] H. Ito, S. Onouse, Y. Miyake, T. Yoshida, Iridal-type triterpenoids with ichthyotoxic activity from Belamcanda chinensis, J. Nat. Prod. 62 (1999) 89–93.
[4] M. Moriyasu, Y. Igi, M. Ichimaru, K. Iwasa, J. Kobayakawa, F. Sato-Nishimori, Y. Matsukawa, C. Nagase, New isoflavones from Belamcandae Rhizoma, J. Nat. Med. 61 (2007) 329–333.
[5] Y.Y. Zhang, Q. Wang, L.W. Qi, X.Y. Qin, M.J. Qin, Characterization and determination of the major constituents in Belamcandae rhizoma by HPLCDAD-ESI-MSn, J. Pharm. Biomed. Anal. 56 (2011) 304–314.
[6] D. Wozniak, J. Oszmianski, A. Matkowski, Antimutagenic and antioxidant activity of the extract from Belamcanda chinensis (L.) DC, Acta Pol.Pharm. 63 (2006) 213–218.
[7] K.B. Oh, H. Kang, H. Matsuoka, Detection of antifungal activity in Belamcanda chinensis by alpha-Naphthoflavone a single-cell bioassay method and isolation of its active compound, tectorigenin, Biosci. Biotechnol. Biochem. 65 (2001) 939–942.
[8] R.H. Fu, C.W. Tsai, R.T. Tsai, S.P. Liu, T.M. Chan, Y.C. Ho, H.L. Lin, Y.M. Chen, H.S. Hung, S.C. Chiu, C.H. Tsai, Y.S. Wang, W.C. Shyu, S.Z. Lin, Irisflorentin modifies properties of mouse bone marrow-derived dendritic cells and reduces the allergic contact hypersensitivity responses, Cell Transplant. 24 (2015) 573–588.
[9] Y.M. Chen, S.P. Liu, H.L. Lin, M.C. Chan, Y.C. Chen, Y.L. Huang, M.C. Tsai, R.H. Fu, Irisflorentin improves -synuclein accumulation and attenuates 6-OHDA-induced dopaminergic neuron degeneration, implication for Parkinson’s disease therapy, Biomedicine (Taipei) 5 (2015) 24–32.
[10] M. Ufer, P.E. Juif, M.L. Boof, C. Muehlan, J. Dingemanse, Metabolite profiling in early clinical drug development: current status and future prospects, Expert Opin. Drug Metab. Toxicol. 13 (2017) 803–806.
[11] Y.W. Jia, Z.Q. Zeng, H.L. Shi, J. Liang, Y.M. Liu, Y.X. Tang, X. Liao, Characterization of in vitro metabolites of irisflorentin by rat liver microsomes using high performance liquid chromatography coupled with tandem mass spectrometry, Biomed. Chromatogr. 9 (2016) 1363–1370.
[12] A.D. Rodrigues, Integrated P450 reaction phenotyping: Attempting to bridge the gap between cDNA-expressed cytochromes P450 and native human liver microsomes, Biochem. Pharmacol. 57 (1999) 465–480.
[13] L.Leung A.S. Kalgutkar, R.S. Obach, Metabolic activation in drug-induced liver injury, Drug Metab. Rev. 44 (2012) 18–33.
[14] A.S. Kalgutkar, I. Gardner, R.S. Obach, C.L. Shaffer, E. Callegari, K.R. Henne, A.E. Mutlib, D.K. Dalvie, J.S. Lee, Y. Nakai, J.P. O’Donnell, J. Boer, S.P. Harriman, A comprehensive listing of bioactivation pathways of organic functional groups, Curr. Drug Metab. 6 (2005) 161–225.
[15] B. Wen, W.L. Fitch, Analytical strategies for the screening and evaluation of chemically reactive drug metabolites, Expert. Opin. Drug Metab. Toxcol. 5 (2009) 39–55.