Metabolism of Diterpenoids Derived from the Bark of Cinnamomum cassia in Human Liver Microsomes

Cinnamomum cassia L. is used as a spice and flavoring agent as well as a traditional medicine worldwide. Diterpenoids, a class of compounds present in C. cassia, have various pharmacological effects, such as anti-inflammatory, antitumor, and antibacterial activities; however, there are insufficient studies on the metabolism of diterpenoids. In this study, the metabolism of seven diterpenoids, namely, anhydrocinnzeylanol, anhydrocinnzeylanine (AHC), cinncassiol A, cinncassiol B, cinnzeylanol, cinnzeylanone, and cinnzeylanine, obtained from the bark of C. cassia was studied in human liver microsomes (HLMs). All studied diterpenoids, except for AHC, exhibited strong metabolic stability; however, AHC was rapidly metabolized to 3% in HLMs in the presence of β-NADPH. Using a high-resolution quadrupole-orbitrap mass spectrometer, 20 metabolites were identified as dehydrogenated metabolites (M1–M3), dehydrogenated and oxidated metabolites (M4–M10), mono-oxidated metabolites (M11–M13), or dioxidated metabolites (M14–M20). In addition, CYP isoforms involved in AHC metabolism were determined by profiling metabolites produced after incubation in 11 recombinant cDNA-expressed CYP isoforms. Thus, the diterpenoid compound AHC was identified in a metabolic pathway involving CYP3A4 in HLMs.


Introduction
Cinnamomum cassia L., which belongs to the family Lauraceae, is used as a spice and flavoring agent as well as a traditional medicine worldwide [1,2]. In traditional medicine, different parts of C. cassia are used for different therapeutic purposes [3]. For example, the leaves are used to treat headaches, chills, and abdominal pain, and the bark is used to treat tussis, gastrointestinal neurosis, diarrhea, edema, and cardiac palpitations. The constituents of the treatments are well known, along with the various medicinal effects of phenylpropanoids, sesquiterpenoids, lignans, and diterpenoids isolated from Cinnamomum species, as shown in previous phytochemical studies [4,5]. Among these, diterpene compounds have been mainly identified in the leaves of C. cassia and were recently isolated from the bark [6,7]. Anhydrocinnzeylanol, anhydrocinnzeylanine (AHC), cinncassiol A, cinncassiol B, cinnzeylanol, cinnzeylanone, and cinnzeylanine have been extracted from the bark of C. cassia and classified as diterpenoids [2,[8][9][10][11] (Figure 1). The pharmacological effects of diterpenoids isolated from C. cassia bark include antiinflammatory, antitumor, and antibacterial activities; thus, these compounds have attracted attention as a new natural active class [12][13][14].
Although the activities of diterpenes derived from C. cassia have been reported, research results have not been reported on the metabolism of these compounds. In the initial Although the activities of diterpenes derived from C. cassia have been reported, research results have not been reported on the metabolism of these compounds. In the initial stages of novel drug development, the study of the metabolism of drug candidates is closely linked to the success of the process. Therefore, for natural active substances to be considered valuable drugs, it is necessary to elucidate their metabolic processes. Notably, metabolism is a process that increases the water solubility of absorbed xenobiotics and converts them into forms that can be easily excreted from the body [15,16]. However, the process of drug metabolism sometimes remains incomplete, leading to increased toxicity of bioactivation [17].
A liquid chromatography-mass spectrometry (LC-MS) system capable of high-resolution (HR) analysis is used as a common platform for drug metabolism studies. In particular, HR-MS can confirm the exact elemental compositions of product ions with HR (>10,000 at full width at half maximum) and accurate mass (<5 ppm deviation) capabilities  Metabolic stability of the seven diterpenoids (10 µM) incubated with 0.5 mg/mL human liver microsomes in the presence of a β-NADPH-regenerating system at 37 • C for 0 or 60 min. Bars indicate standard error (n = 3). * p < 0.05 and *** p < 0.001.
A liquid chromatography-mass spectrometry (LC-MS) system capable of high-resolution (HR) analysis is used as a common platform for drug metabolism studies. In particular, HR-MS can confirm the exact elemental compositions of product ions with HR (>10,000 at full width at half maximum) and accurate mass (<5 ppm deviation) capabilities [18]. Therefore, HR-MS enables the accurate identification of metabolite structures by acquiring product ion information generated by the MS 2 of target compounds. In addition to its use in the qualitative analysis of metabolites, LC-MS is widely used as an essential instrument for metabolic stability evaluation and pharmacokinetic research in the development of new drug candidates [19].
Therefore, metabolism studies of active substances can provide important data for predicting pharmacokinetics and toxicity evaluation. In this study, to examine the pharmacological properties of diterpenes isolated from C. cassia, we evaluated the metabolic stability of seven diterpenes and identified the metabolic pathway of AHC, which had the highest metabolic rate in human liver microsomes (HLMs). Seven diterpenes isolated from C cassia were selected based on their structural similarity. The structures of the generated metabolites of AHC were identified with high resolution and accuracy using a mass spectrometer, and the human CYP isoforms involved in the metabolism of AHC were identified; these data can be used to predict additional drug interactions.

Metabolic Stability Studies
For metabolism profiling, 10 µM of each of the seven compounds isolated from C. cassia bark was added to 0.5 mg/mL of pooled HLMs and 0.1 M potassium phosphate buffer (pH 7.4). The HLMs were then preincubated at 37 • C for 1 min. After adding NGS, the reaction mixture was incubated at 37 • C for 60 min. After the 60 min incubation period, 50 µL of the reaction mixture was transferred to a new tube, and 50 µL of 100% ACN containing 0.1% formic acid and tolbutamide (internal standard) was added to terminate the reaction. The mixture was then vortexed, followed by centrifugation at 15,700× g (13,000 rpm) for 10 min (Eppendorf AG, Hamburg, Germany). Finally, 90 µL of the supernatant was transferred to a vial and injected into a C18 column for LC-MS/ MS analysis.

Identification of AHC Metabolites in HLMs
To identify AHC metabolites, 10 µM AHC was incubated with 1 mg/mL pooled HLMs in 0.1 M potassium phosphate buffer (pH 7.4) at 37 • C for 60 min. The reactions were initiated by adding NGS to obtain a final incubation volume of 200 µL. All experiments were conducted in triplicate. The reaction was terminated by adding 400 µL of 100% ACN, and the mixture was vortexed and then centrifuged for 10 min at 15,700× g. Next, the supernatant (550 µL) was transferred to a new tube and dried using a speed-vac concentrator with a cold trap (Labconco, Kansas, MO, USA). Subsequently, the dried samples were reconstituted in 100 µL of 20% MeOH containing 0.1% formic acid before centrifugation at 15,700× g for 10 min. Finally, the mixture was injected into a liquid chromatograph with a tandem mass spectrometer for analysis.

Metabolism of AHC by SKF-525A: A Nonselective Inhibitor
To investigate CYP-mediated metabolism, 10 µM AHC was mixed with 1 mg/mL pooled HLMs and 0.1 M potassium phosphate buffer (pH 7.4). Then, SKF-525A, a general inhibitor of CYPs, was added at concentrations of 0.5, 1, and 5 mM. The reactions were initiated with the addition of NGS at 3 • C for 60 min. After 60 min, the reactions were terminated by adding 400 µL of 100% acetonitrile. The mixtures were then centrifuged at 15,700× g for 10 min. Subsequently, 550 µL of supernatant was dried in a vacuum concentrator. The residue was dissolved in 100 µL of 20% MeOH (0.1% formic acid), and 5 µL of this sample was injected into a C18 column for LC-MS/MS analysis.

Determination of Recombinant cDNA-Expressed CYP Isoforms Included in AHC Metabolism
To identify the metabolic enzymes for AHC, 10 µM AHC was incubated with 10 pmol of human recombinant cDNA-expressed CYP isoforms (CYP1A1, 1A2, 2A6, 2B6, 2C8, 2C9, 2C19, 2D6, 2E1, 3A4, and 3A5) in the presence of 0.1 M potassium phosphate buffer (pH 7.4). The reaction was initiated by adding 100 µL of NGS at 37 • C for 60 min. Then, the mixture was suspended by adding 400 µL of ice-cold 100% ACN. Following centrifugation at 15,700× g for 10 min at 4 • C, the supernatant was transferred to a new tube and evaporated using a speed-vacuum concentrator. Next, the dried sample was diluted with 100 µL of 20% MeOH containing 0.1% formic acid, and a 5 µL aliquot was injected into a C18 column for LC-MS/MS analysis.

LC-MS/MS
The metabolism screening of the seven compounds was performed using an HPLC system (Thermo Fisher Scientific, Bremen, Germany) equipped with an HPG-3200SD standard binary pump, a WPS-3000 TRS analytical autosampler, and a TCC-3000 SD column compartment. The HPLC system was coupled with an HR-MS (Q-Exactive Focus Quadrupole-Orbitrap MS; Thermo Fisher Scientific, Waltham, MA, USA) at the Mass Spectrometry Convergence Research Center. The full MS resolution and scan ranges were 70,000 and 60-900 m/z, respectively. The MS/MS resolution was 35,000. The mass spectra were obtained in negative ion electrospray mode. A heated electrospray ionization (HESI-II) probe was used as an ion generator, with nitrogen as the sheath gas at 35 aux units and auxiliary gas at 12 aux units. The mass spectrometer was operated in negative ionization mode. The other parameters were as follows: spray voltage, 3.5 kV; capillary temperature, 320 • C; S-lens radio frequency level, 50; and aux gas heater temperature, 400 • C.

Statistical Analysis
All data were expressed as the mean ± standard deviation. Statistical analysis was performed using IBM SPSS Statistics version 21 to determine significant differences. Values of p < 0.05 were considered statistically significant and are indicated with asterisks in the figures (i.e., * p < 0.05, ** p < 0.01, and *** p < 0.001).

Metabolic Stability of the Seven Diterpenoids in HLMs
To confirm their metabolic stability, seven diterpenoids (AHC, anhydrocinnzeylanol, cinncassiol A, cinncassiol B, cinnzeylanol, cinnzeylanone, and cinnzeylanine) derived from the bark of C. cassia were incubated with HLMs in the presence of the NGS for 60 min ( Figure 1B). The metabolic stability of each compound was determined according to the ratio of the peak area at 60 min to that at 0 min. This analysis showed that AHC showed the most rapid metabolization among the seven compounds. In contrast, the stabilities of anhydrocinnzeylanol and cinneylanone decreased by approximately 20%, whereas the remaining four compounds did not significantly decrease; thus, the six compounds other than AHC were metabolically stable in HLMs. Therefore, we identified the metabolites of AHC produced in HLMs.

Identification of Phase I Metabolites of AHC
To identify the phase I metabolites of AHC in HLMs, 10 µM AHC was incubated in pooled HLMs without or with the NGS for 60 min. The total ion chromatogram showed that when AHC was incubated in HLMs in the presence of the NGS, AHC was reduced. AHC decreased rapidly with the same trend as the metabolic stability results, and peaks presumed to be metabolites were generated ( Figure S1). The extracted ion chromatograms of the generated AHC metabolites are shown in Figure 2A Figure 2. Extract ion chromatograms (EIC) for anhydrocinnzeylanine (AHC) and its metabolites. AHC (10 µM) was incubated with 1 mg/mL pooled human liver microsomes for 60 min in the absence (A) or presence (B) of the β-NADPHregenerating system (NGS). Synthesis of AHC metabolites by recombinant cDNA-expressed CYP3A4 after incubation at 37 • C for 60 min (C).

Interpretation of Metabolite Structure
To elucidate the chemical structure of the metabolites, we performed HR-MS/MS analysis of AHC and its metabolites using a quadrupole-orbitrap MS. The fragment ions of AHC and its metabolites were indicated in their elemental composition within a mass error of <5 ppm (Table S1). To identify the metabolite structure, the fragment pattern of the parent compound, i.e., AHC, was analyzed ( Figure 3A). The MS spectrum of deprotonated AHC was detected at m/z 407.2072 (C 22 H 31 O 7 ) in the negative ion mode; the spectra of AHC showed eight product ions. Of these ions, three major product ions served as markers of fragment ions for metabolite identification. Three key product ions were observed at m/z 135.0802 (C 9 H 11 O), 149.0962 (C 10 H 13 O), and 153.0910 (C 9 H 13 O 2 ). The spectra of m/z 135 and 153 indicated the loss of the isopropyl-methylcylopentane moiety to -H 2 O (−272 Da) and the isopropyl-methylcylopentane (−254 Da) moiety, respectively. The structure of m/z 149 indicated the breakage of bonds of C6 and C7, bonds of C8 and C9, bonds of C10 and C11, and part of C1. The major ions were denoted as symbols: m/z 135 as "a", m/z 149 as "b", and m/z 153 as "c", to simplify the identification of subsequent metabolite structures.
Additionally, although not providing information on metabolite structure determination, fragment ions that were derived by cleavage of the hydroxy groups in the parent existed. The product ions at m/z 347.1861 (C 20 Figure S2B). In M3, the same product ions at m/z 151.0756 and 149.0961 as those in M1 and M2 were observed, indicating dehydrogenation at the isopropyl-methylcylopentane moiety. Although exact structural analysis was not possible to identify, three product ions were observed in M3 that differed from those in M1 and M2. The product ions at m/z 219.1020 (C 13 H 15 O 3 ) and 203.1070 (C 13 H 15 O 2 ) indicated the loss of CO from 247.0972 and the loss of O from m/z 219.1020, respectively. The product ion at m/z 219.1020 was identified with breakage of the bond between C6 and C7, the bond between C5 and C9, and the single bond between oxygen and C11, and with dehydrogenation at the isopropyl-methylcylopentane moiety. Therefore, M1-M3 were dehydrogenated metabolites in the isopropyl-methylcylopentane moiety.
H2O from m/z 361, and the loss of the carbonyl moiety, respectively.

Time-Dependent Formation of AHC Metabolites
To identify the time-dependent metabolism of AHC in HLMs, we incubated AHC with pooled HLMs in the presence of the NGS for 0, 5, 20, 40, 60, and 90 min ( Figure 5). The hydrogenated metabolites (M1-M3) showed a sharp increase in the initial production of the reaction but decreased as the reaction time increased, indicating their further conversion into other metabolites after the hydrogenation of AHC. The hydrogenated and oxidative metabolites (M5, M8, M9, and M10) tended to gradually decrease after increasing continuously up to 20 min, except for M5, which continued to increase depending on the reaction time. By comparing the relative production over time, it was revealed that AHC was first metabolized by hydrogenation, followed by oxidation. Considering the oxidative metabolites of AHC, the mono-oxidative metabolites (M11, M12, and M13) peaked at 20 min and then decreased with the reaction time, whereas the dioxidative metabolites (M15, M17, M18, and M19) increased continuously depending on the reaction time.

Time-Dependent Formation of AHC Metabolites
To identify the time-dependent metabolism of AHC in HLMs, we incubated AHC with pooled HLMs in the presence of the NGS for 0, 5, 20, 40, 60, and 90 min ( Figure 5). The hydrogenated metabolites (M1-M3) showed a sharp increase in the initial production of the reaction but decreased as the reaction time increased, indicating their further conversion into other metabolites after the hydrogenation of AHC. The hydrogenated and oxidative metabolites (M5, M8, M9, and M10) tended to gradually decrease after increasing continuously up to 20 min, except for M5, which continued to increase depending on the reaction time. By comparing the relative production over time, it was revealed that AHC was first metabolized by hydrogenation, followed by oxidation. Considering the oxidative metabolites of AHC, the mono-oxidative metabolites (M11, M12, and M13) peaked at 20 min and then decreased with the reaction time, whereas the dioxidative metabolites (M15, M17, M18, and M19) increased continuously depending on the reaction time.

Characterization of AHC Metabolism in cDNA-Expressed Recombinant CYP Isoforms
To confirm the metabolism of AHC by CYP enzymes, AHC was incubated in HLMs with SKF-525A, i.e., a nonspecific CYP inhibitor ( Figure S7). The metabolism of AHC was inhibited, and the residual amount of AHC after the reaction was significantly increased depending on the SKF-525A treatment concentration; thus, AHC was metabolized in a CYP-dependent manner. Consequently, as the metabolism of AHC was inhibited by SKF-

Characterization of AHC Metabolism in cDNA-Expressed Recombinant CYP Isoforms
To confirm the metabolism of AHC by CYP enzymes, AHC was incubated in HLMs with SKF-525A, i.e., a nonspecific CYP inhibitor ( Figure S7). The metabolism of AHC was inhibited, and the residual amount of AHC after the reaction was significantly increased depending on the SKF-525A treatment concentration; thus, AHC was metabolized in a CYPdependent manner. Consequently, as the metabolism of AHC was inhibited by SKF-525A treatment, the production of all metabolites was reduced in an SKF-525A concentrationdependent manner.
To identify the CYP isoforms involved in AHC metabolism, AHC (10 µM) was incubated with eleven cDNA-expressed recombinant CYP isoforms (CYP1A1, CYP1A2, CYP2A6, CYP2B6, CYP2C8, CYP2C9, CYP2C19, CYP2D6, CYP2E1, CYP3A4, and CYP3A5) in the presence of the NGS. As a result, CYP3A4 produced M1-M20 as the major CYP responsible for the metabolism of AHC in HLMs ( Figure 2C). Additionally, some metabolites such as M1, M2, M8, M11, M12, and M13 were generated at low levels by CYP3A5 ( Figure S8). However, the metabolites of AHC were not observed from reactions with other CYPs. Thus, in the metabolism of AHC in HLMs, CYP3A4 was involved as the main enzyme, and four types of metabolites were generated. Based on these results, the postulated metabolic pathway of AHC in the HLMs is summarized in Figure 6. 525A treatment, the production of all metabolites was reduced in an SKF-525A concentration-dependent manner.
To identify the CYP isoforms involved in AHC metabolism, AHC (10 µM) was incubated with eleven cDNA-expressed recombinant CYP isoforms (CYP1A1, CYP1A2, CYP2A6, CYP2B6, CYP2C8, CYP2C9, CYP2C19, CYP2D6, CYP2E1, CYP3A4, and CYP3A5) in the presence of the NGS. As a result, CYP3A4 produced M1-M20 as the major CYP responsible for the metabolism of AHC in HLMs ( Figure 2C). Additionally, some metabolites such as M1, M2, M8, M11, M12, and M13 were generated at low levels by CYP3A5 ( Figure S8). However, the metabolites of AHC were not observed from reactions with other CYPs. Thus, in the metabolism of AHC in HLMs, CYP3A4 was involved as the main enzyme, and four types of metabolites were generated. Based on these results, the postulated metabolic pathway of AHC in the HLMs is summarized in Figure 6.

Discussion
Diterpene compounds are known to exhibit excellent biological activity as active substances in several medicinal plants, and research is underway to use diterpene compounds as pharmaceuticals or health functional foods [20]. Although phase I and phase II metabolism experiments were conducted to develop diterpene compounds of several skeletal types as new drug candidates, metabolism studies of diterpene compounds with the same backbone as AHC have not been reported. In this study, the metabolism of seven diterpenoids isolated from C. cassia was studied in HLMs. Of the seven compounds, four were metabolically stable; however, anhydrocinnzeylanol and cinnzeylanone were decreased by approximately 20%, whereas AHC showed a unique reduction in metabolic stability in HLMs ( Figure 1B). Although the compounds have a common structure, i.e., the (3aR,3bR,5S,7aS,8R,8aR)-2-isopropyl-3,5,8-trimethyldecahydro-1H-3b,8(epoxy-ethano)-cyclopenta[a]indene-3a,7a,8a-triol structure, the rapid metabolism of AHC can be explained by structural differences relative to the other six compounds [21,22]. Cinncassiol B, cinnzeylanol, cinnzeylanone, and cinnzeylanine have C11-C12 bond, C13 hydroxylation and C11 hydroxylation structures in common. These similarities are suggested to be related to the enhanced metabolic stability, unlike that observed for AHC. For cinncas-

Discussion
Diterpene compounds are known to exhibit excellent biological activity as active substances in several medicinal plants, and research is underway to use diterpene compounds as pharmaceuticals or health functional foods [20]. Although phase I and phase II metabolism experiments were conducted to develop diterpene compounds of several skeletal types as new drug candidates, metabolism studies of diterpene compounds with the same backbone as AHC have not been reported. In this study, the metabolism of seven diterpenoids isolated from C. cassia was studied in HLMs. Of the seven compounds, four were metabolically stable; however, anhydrocinnzeylanol and cinnzeylanone were decreased by approximately 20%, whereas AHC showed a unique reduction in metabolic stability in HLMs ( Figure 1B). Although the compounds have a common structure, i.e., the (3aR,3bR,5S,7aS,8R,8aR)-2-isopropyl-3,5,8-trimethyldecahydro-1H-3b,8(epoxy-ethano)cyclopenta[a]indene-3a,7a,8a-triol structure, the rapid metabolism of AHC can be explained by structural differences relative to the other six compounds [21,22]. Cinncassiol B, cinnzeylanol, cinnzeylanone, and cinnzeylanine have C11-C12 bond, C13 hydroxylation and C11 hydroxylation structures in common. These similarities are suggested to be related to the enhanced metabolic stability, unlike that observed for AHC. For cinncassiol A, C19 hydroxylation enhanced the metabolic stability, whereas anhydrocinnzeylanol, in particular, differs from AHC in that C1 does not have an acetyl group. C1 acetylation plays an important role in the metabolism of AHC. Anhydrocinnzeylanol, which has an AHC-like structure, except for C1-acetyl, is metabolized by 20%, whereas AHC with the C1-acetyl group is metabolized by 95% under the same conditions. In addition, the C1-acetyl moiety was not directly metabolized, i.e., we did not identify the O-deacetyl metabolite of AHC in HLMs. Thus, the C1-acetyl group is predicted to be important for interactions between AHC and CYP enzymes, especially CYP3A4, in HLMs.
In this study, AHC was predominantly metabolized by CYP3A4 following incubation with cDNA-expressed recombinant CYP3A4 in the presence of the NGS. All metabolites (M1-M20) were formed in CYP3A4, similar to that in HLMs ( Figure 2C). In addition, some metabolites, i.e., M1, M2, M8, M11, M12, and M13, were also detected in small amounts in CYP3A5. CYP3A is the most abundant enzyme in the liver, constituting approximately 30% of all CYP proteins in the liver [23]. Considering that CYP3A oxidizes 40%-50% of drugs, it is of great importance in metabolic reactions and drug-drug interactions [24]. CYP3A4 is the dominant CYP, accounting for 80% of CYPs in small intestinal enterocytes, and is an important factor in the bioavailability of CYP3A4 substrate drugs [25]. Although the rate of metabolism of AHC in human intestinal microsomes was not evaluated in this study, the effect of intestinal CYP3A4-mediated biotransformation should be considered when evaluating the in vivo bioavailability of AHC.
Notably, the dehydrogenation of −2 Da in the parent compound showed that large amounts of metabolites were rapidly produced at the start of the reaction ( Figure 5). Thus, the dehydrogenation reaction may involve enzymes other than CYPs. The major oxidative enzymes other than CYPs involved in the metabolism of drugs and other xenobiotics include flavin-containing monooxygenases, molybdenum hydroxylases (aldehyde oxidase and xanthine oxidase), prostaglandin H synthase, lipoxygenases, amine oxidases (monoamine, polyamine, diamine, and semicarbazide-sensitive amine), alcohol dehydrogenases, and aldehyde dehydrogenase [26]. However, the yield of all metabolites, including dehydrogenated metabolites, was significantly decreased following the addition of SKF-525A. The metabolites were synthesized by cDNA-expressed recombinant CYP3A4, suggesting that CYP3A4 is the main metabolizing enzyme for AHC dehydrogenation and oxidation in HLMs. In a previous study, the dehydrogenation of many compounds by CYP and the chemical mechanism of dehydrogenation were reported in detail [27]. Furthermore, it has been confirmed that testosterone is dehydrogenated by CYP3A, resulting in the formation of the metabolite 6-dehydrotestosterone [21,22]. Moreover, CYP3A4 mediates the dehydrogenation of diverse compounds such as raloxifene and ezlopitant [28][29][30]. This study did not determine exactly where dehydrogenation occurred in the isopropylmethylcylopentane moiety of M1-M3, but dehydrogenated metabolites were generated by CYP3A.
Although we focused on identifying the phase I metabolites of seven diterpenes, including AHC, there is a possibility that phase II metabolism may proceed from a structural viewpoint, such as having -OH groups. Therefore, phase II metabolism studies should be conducted to understand the entire metabolic pathway of diterpene compounds.

Conclusions
We verified that AHC is metabolized to three dehydrogenated metabolites (M1-M3), seven dehydrogenated and oxidated metabolites (M4-M10), three mono-oxidated metabolites (M11-M13), and seven dioxidated metabolites (M14-M20) in HLMs, in which CYP3A4 was involved in the synthesis of all metabolites. Additionally, the structures of the metabolites were determined using HR/high-accuracy MS/MS. Finally, the possible metabolic fate of AHC in HLMs was summarized ( Figure 6).