Preclinical metabolism and metabolic drug–drug interaction profile of pedunculoside and rotundic acid

Abstract

Pedunculoside and rotundic acid, the most abundant components in plants of the genus Ilex L. (Aquifoliaceae), exhibit biological and pharmacological significance in the treatment of cardiovascular diseases. However, there have been few studies on their metabolism. This study performed a systematic metabolism study of pedunculoside and rotundic acid and evaluated their potential for herb–drug interaction. Pedunculoside or rotundic acid was incubated with human liver microsomes and recombinant human metabolic enzymes, and analyzed using LC‐Q‐TOF/MS and LC–MS/MS. Pedunculoside was found to be the most stable in human liver microsomes, whereas rotundic acid was easily metabolized. Eight pedunculoside metabolites and six rotundic acid metabolites were detected and tentatively identified through hydroxylation, glucuronidation, acetylation, and glucose conjugation. Hydroxylation of pedunculoside is mainly catalyzed by CYP3A4/5 and partly by CYP2C8. Hydroxylation of rotundic acid is almost exclusively catalyzed by CYP3A4/5, and its glucuronidation reaction is mediated by UGT1A4. Neither pedunculoside nor rotundic acid showed CYP inhibition (IC₅₀ values > 50 μM) with the probe substrates of major CYP isoforms during incubation with human liver microsomes. This study is the first investigation into the in vitro metabolism of pedunculoside and rotundic acid using human liver microsomes. It also aims to assess their potential as perpetrators of drug–drug interactions involving CYP enzymes. The comprehensive metabolism and drug interaction studies of pedunculoside and rotundic acid enable us to evaluate and manage potential risks with their use in pharmacotherapy.

Study Highlights.

• WHAT IS THE CURRENT KNOWLEDGE ON THE TOPIC?

While pedunculoside exhibits biological and pharmacological significance in treating cardiovascular diseases, there have been few studies on its metabolism.

• WHAT QUESTION DID THIS STUDY ADDRESS?

Metabolic profiles, metabolic pathways, and the major enzymes involved in pedunculoside and rotundic acid metabolism in human liver microsomes require further investigation. Moreover, the potential drug–drug interaction (DDI) of pedunculoside and rotundic acid as cytochrome P450 (CYP) perpetrators need to be evaluated.

• WHAT DOES THIS STUDY ADD TO OUR KNOWLEDGE?

Pedunculoside was stable in human liver microsomes, conversely, rotundic acid was quickly metabolized in the liver. Eight pedunculoside metabolites and six rotundic acid metabolites in human liver microsomes were tentatively identified. CYP3A4/5 primarily catalyzes the hydroxylation of pedunculoside and rotundic acid, while UGT1A4 mediates the glucuronidation reaction of rotundic acid. Neither compound showed inhibition potential for major CYP isoforms.

• HOW MIGHT THIS CHANGE CLINICAL PHARMACOLOGY OR TRANSLATIONAL SCIENCE?

The comprehensive studies on the metabolism and drug interactions of pedunculoside and rotundic acid allow us to assess and mitigate potential risks associated with their use in pharmacotherapy.

INTRODUCTION

The genus Ilex L. (Aquifoliaceae) is used to treat inflammation, thrombosis, and cardiovascular/cerebrovascular diseases. More than 600 species of Ilex have been discovered worldwide, mainly as evergreen or deciduous arbors and shrubs distributed in temperate and tropical regions. ¹ Several Chinese herbal products contain Ilex L. (Aquifoliaceae), such as Compound Hairy Holly and Aluminium Clofibrate Tablets (containing Ilex pubescens and used to treat coronary atherosclerosis heart disease), Compound Jiubiying Capsules (containing Ilex rotunda and used to treat diarrhea and gastroenteritis), and Shiliuwei Dongqing Pills (containing Ilex chinensis and used to treat edema, cold cough, and dizziness). ² Triterpenoids and their glycosides, flavonoids, and phenols are the main chemical constituents reported in Ilex species. Among them, the most widely studied components are triterpenoid saponins, which are also the major pharmacologically active components. ² , ³ Pedunculoside, a triterpenoid saponin, is one of the most abundant components existing in the genus Ilex, particularly in Ilex rotunda Thunb, ⁴ , ⁵ as well as in Ilex pupurea Hassk., Ilex asprella Champ., Ilex pubescens Hook. et Arn., and Ilex cornuta Lindl. et Paxt. ⁶ , ⁷ , ⁸ , ⁹ As an active botanical component, pedunculoside has demonstrated its pharmacological functions in attenuating ischemia–reperfusion‐induced myocardial injury, ¹⁰ , ¹¹ ameliorating hyperlipidemia, improving liver injury, ¹² , ¹³ and protecting against arthritis, ulcerative colitis, and colitis‐associated cancer, ¹⁴ , ¹⁵ , ¹⁶ , ¹⁷ , ¹⁸ showed its potential to be developed as a therapeutic drug or drug precursor in the future.

Researchers have investigated the pharmacokinetic behavior of pedunculoside by administering it orally ¹⁹ , ²⁰ , ²¹ , ²² or intravenously ²² , ²³ in rats. In most studies, pedunculoside showed low bioavailability and short elimination half‐life, indicating that the biotransformation of pedunculoside is particularly important to its effectiveness. Our previous study investigated the metabolic pathways of hydroxylation and glucose conjugation of pedunculoside in rat liver microsomes and rat liver S9, and nine metabolites were detected and identified after oral administration of pedunculoside in rats. ²⁴ However, due to differences in species, the metabolic profiles of pedunculoside may vary from rats to humans, and additional preclinical metabolism studies of pedunculoside based on biological matrices sourced from human beings are still needed to provide complementary information for its clinical use. In addition, pedunculoside was proven to undergo deglycosylation to generate its aglycon rotundic acid through rat intestinal bacteria both in vitro ²⁵ and in vivo, ²¹ , ²⁴ indicating that during its absorption, a part of pedunculoside can be first metabolized by human intestinal flora and enter the blood circulation in the form of rotundic acid and then undergo further biotransformation in the liver. Thus, to comprehensively understand and predict the clinical metabolic profile of pedunculoside, including its subsequent metabolism and multi‐step metabolites generated from rotundic acid, metabolic characteristics and metabolic pathways for both pedunculoside and rotundic acid should be studied together.

The use of herbal medicines has rapidly increased worldwide. Pedunculoside is often administrated with co‐existing components in the genus Ilex, and herbal medicines may also be clinically co‐administrated with other medications, such as diuretics, calcium channel antagonists, and β‐receptor blockers, especially in the treatment of cardiovascular diseases. This might cause drug–drug interactions (DDI), leading to a higher risk of adverse reactions ²⁶ , ²⁷ ; therefore, it is critical to evaluate the metabolic DDI potential of pedunculoside and rotundic acid. Although pedunculoside and rotundic acid are widely used, the specific metabolic enzymes involved in pedunculoside and rotundic acid metabolism and the inhibitory DDI potential of pedunculoside and rotundic acid remain unknown. Therefore, further research is needed to ensure their clinical use.

This study is the first investigation into the in vitro metabolism of pedunculoside and rotundic acid using human liver microsomes (HLM). It also aims to assess their potential as perpetrators of DDI involving cytochrome P450 (CYP) enzymes. The objective of this study was to (1) identify metabolic pathways and metabolic profiles of both pedunculoside and rotundic acid in HLM, (2) identify major enzymes [CYP and UDP‐glucuronosyltransferase (UGT) isoforms] involved in the metabolism of pedunculoside and rotundic acid, and (3) evaluate the DDI potential of pedunculoside and rotundic acid as CYP perpetrators. Metabolite searching, structural elucidation, and simultaneous monitoring of the target components were performed using the analytical instruments of liquid chromatography‐quadrupole‐time of flight/mass spectrometry (LC‐Q‐TOF/MS) and liquid chromatography–tandem mass spectrometry (LC–MS/MS). The findings of this study will enhance the understanding of the metabolism and pharmacokinetics of pedunculoside and rotundic acid and facilitate their clinical application.

METHODS

Pedunculoside and rotundic acid metabolism study by liver microsomal incubation

Metabolites identification of pedunculoside and rotundic acid

The typical reaction mixtures (250 μL for each reaction) contained 100 mM potassium phosphate buffer (pH 7.4, consisting of KCl, NaCl, KH₂PO₄ and Na₂HPO₄·12H₂O), 5 mM MgCl₂, 0.5 mg protein/mL pooled HLM, an nicotinamide adenine dinucleotide phosphate hydride (NADPH) regenerating system [5 mM d‐Glucose 6‐phosphate disodium salt hydrate (G‐6‐P), 1 mM β‐nicotinamide adenine dinucleotide phosphate hydrate (NADP⁺), and 1.2 U/mL glucose‐6‐phosphate dehydrogenase (PDH)], 1 mM d‐saccharic acid 1,4‐lactone monohydrate (DSL), 2 mM uridine 5′‐diphosphoglucuronic acid (UDPGA), and 100 μM pedunculoside or rotundic acid. A control assay was conducted without pedunculoside or rotundic acid.

The pooled HLM was first pretreated with alamethicin at 25 μg/mg of protein on ice for 20 min to diminish the potential latency of UGT activity. After pre‐incubation for 5 min at 37°C, the reaction was initiated by the addition of the NADPH regenerating system and UDPGA. All reactions were incubated at 37°C for 0 (negative control), 8, 24, and 48 h, and terminated by cooling in an ice bath and treatment with 1 mL acetonitrile. After shaking and centrifugation at 16810g (13000 rpm) for 10 min, the supernatant was transferred and evaporated to dryness under vacuum using a CentriVap refrigerated centrifugal concentrator (Labconco, Kansas City, MO, USA) at 40°C. The residues were reconstituted in 200 μL of methanol and centrifuged at 13,000 rpm for 10 min twice. A 2 μL aliquot of the supernatant from each biological sample was injected into the LC‐Q‐TOF/MS system and analyzed (n = 2).

Metabolic stability study of pedunculoside and rotundic acid in HLM

The in vitro metabolic stability studies of both pedunculoside and rotundic acid were performed using HLM (0.25 mg protein/mL) in triplicate. The concentration of the incubated pedunculoside and rotundic acid was ~ 1 μM, and aliquots were collected at 0, 5, 10, 20, 30, 45, 60, 90, and 120 min. The blank control assay (control) was conducted without pedunculoside or rotundic acid, and the negative control assay (negative) was conducted without HLM, NADPH regenerating system, and UDPGA. Serial samples for both blank control and negative control assays were only collected at selected time intervals (0, 60, and 120 min). The incubation systems were also conducted in the absence of the NADPH regenerating system (NADPH‐) or UDPGA (UDPGA‐) to investigate whether the metabolic enzyme systems would affect the metabolic rate of pedunculoside and rotundic acid. In our preliminary metabolic enzyme identification, different incubation systems were used and compared, with the incubation time set at 0, 2, and 4 h, to check whether pedunculoside and rotundic acid could undergo the corresponding biotransformation to generate the metabolites identified. All reactions were terminated with 500 μL ice‐cold acetonitrile containing ilexsaponin A1 [20 ng/mL, internal standard (IS)]. The mixtures were vortexed and centrifuged twice at 13,000 rpm for 10 min, and the supernatant was subjected to LC–MS/MS analysis (n = 3).

Determination of the phase I enzymes responsible for the metabolism of pedunculoside and rotundic acid

All microsomal incubations were performed at 37°C in a final volume of 200 μL containing pooled HLM (0.25 mg protein/mL) and an NADPH regenerating system consisting of MgCl₂ (10 mM), G‐6‐P (10 mM), NADP⁺ (1.5 mM) and PDH (1.5 U/mL). The final concentration of the organic solvent (methanol) in the reaction system was less than 1% (v/v). The incubation time was set to 20 min. All incubations were terminated by adding 200 μL ice‐cold acetonitrile containing ilexsaponin A1 (20 ng/mL, IS), and aliquots of the mixtures were then mixed thoroughly and centrifuged at 13,000 rpm for 10 min twice to obtain the supernatants, of which 2 μL was subjected to LC–MS/MS analysis (n = 3).

To determine the appropriate concentrations of pedunculoside and rotundic acid, enzyme kinetic profile studies were first conducted. The concentrations of pedunculoside used in the incubation system were 1–500 μM, and rotundic acid were at 0.5–50 μM, respectively. Incubation time and HLM protein concentration for the incubation system were also investigated based on the determined proper concentrations of the substrates. The incubation times tested were 0, 5, 10, 20, 30, 45, and 60 min. The HLM protein concentrations tested were 0, 0.025, 0.05, 0.1, 0.25, 0.5, 1, and 2 mg/mL.

To identify the CYP isoforms that catalyze the phase I reaction of pedunculoside and rotundic acid, a set of six known CYP‐specific inhibitors were used, including furafylline (CYP1A2 inhibitor), ticlopidine (CYP2B6/CYP2C19 inhibitor), quercetin (CYP2C8 inhibitor), sulfaphenazole (CYP2C9 inhibitor), quinidine (CYP2D6 inhibitor), and ketoconazole (CYP3A4/5 inhibitor). A range of concentrations was selected for each inhibitor in the incubation system: furafylline 0.01–10 μM, ticlopidine 0.01–10 μM, quercetin 0.01–10 μM, sulfaphenazole 0.005–5 μM, quinidine 0.002–2 μM, and ketoconazole 0.005–1 μM. Mechanism‐based inhibitors, such as furafylline and ticlopidine, were pre‐incubated for 30 min at 37°C with the NADPH regeneration system and HLM before the addition of the substrate. ²⁸ , ²⁹ A control assay was conducted without any inhibitor. The proper concentrations of pedunculoside and rotundic acid were determined and used in the incubation.

Determination of the phase II enzymes responsible for the metabolism of rotundic acid

All microsomal incubations were performed at 37°C in a final volume of 200 μL containing pooled HLM (0.25 mg protein/mL), 1 mM DSL, and 2 mM UDPGA. The final concentration of the organic solvent (methanol) in the reaction system was less than 1% (v/v). The pooled HLM was first pretreated with alamethicin at 25 μg/mg of protein on ice for 20 min to diminish the potential latency of UGT activity. After pre‐incubation for 5 min at 37°C, the reaction was initiated by the addition of UDPGA. The incubation time was set to 20 min. All incubations were terminated by adding 200 μL ice‐cold acetonitrile containing ilexsaponin A1 (20 ng/mL, IS), and aliquots of the mixtures were then mixed thoroughly and centrifuged at 13,000 rpm for 10 min twice to obtain the supernatants, of which 2 μL was subjected to LC–MS/MS analysis (n = 3).

An enzyme kinetic profile study was first conducted to determine the appropriate concentration of rotundic acid. The concentration of rotundic acid used in the incubation system was 0.2–50 μM. To identify the UGT isoforms catalyzing the phase II reaction of rotundic acid, a set of five major recombinant human UGTs were used, including UGT1A1, UGT1A4, UGT1A6, UGT1A9, and UGT2B7. The concentration selected for each recombinant human UGT in the incubation system was also 0.25 mg protein/mL. The positive control assay was conducted using pooled HLM, and the negative control assay was conducted without HLM or recombinant enzymes. The determined concentration of rotundic acid was used for the incubation. Different protein concentrations of the recombinant enzyme were tested at 0.1, 0.5, and 1 mg/mL, with rotundic acid concentrations of 10 and 50 μM to further confirm the enzymatic reaction.

Inhibitory DDI evaluation of pedunculoside and rotundic acid

All microsomal incubations were performed for 20 min at 37°C in a final volume of 200 μL containing pooled HLM (0.25 mg protein/mL) and an NADPH regenerating system consisting of MgCl₂ (10 mM), G‐6‐P (10 mM), NADP⁺ (1.5 mM), and PDH (1.5 U/mL). In the cocktail incubation system, a set of 8 probe substrates for seven specific CYPs were used, including phenacetin (substrate of CYP1A2), bupropion (substrate of CYP2B6), paclitaxel (substrate of CYP2C8), tolbutamide (substrate of CYP2C9), dextromethorphan (substrate of CYP2D6), chlorzoxazone (substrate of CYP2E1), midazolam (substrate of CYP3A4/5), and testosterone (substrate of CYP3A4/5). The final substrate concentrations used for inhibition evaluation were 13.95 μM (phenacetin, 2.5 μg/mL), 9.051 μM (bupropion, 2.5 μg/mL), 5.855 μM (paclitaxel, 5 μg/mL), 18.49 μM (tolbutamide, 5 μg/mL), 2.838 μM (dextromethorphan, 1 μg/mL), 58.97 μM (chlorzoxazone, 10 μg/mL), 17.34 μM (testosterone, 5 μg/mL), and 1.535 μM (midazolam, 0.5 μg/mL). ³⁰ Inhibitory DDI evaluation of pedunculoside and rotundic acid was performed separately, and their concentrations tested in the assay were 0.05, 0.1, 0.2, 0.5, 1, 2, 5, 10, 20, and 50 μM. The control group was not treated with either pedunculoside or rotundic acid. The final concentration of the organic solvent (methanol) in the reaction system was less than 1% (v/v). All incubations were terminated by adding 200 μL ice‐cold acetonitrile containing 20 ng/mL diazepam and 50 ng/mL chloramphenicol (both used as IS). Aliquots of the mixtures were then mixed thoroughly and centrifuged twice at 13,000 rpm for 10 min to obtain the supernatants, of which 2 μL was subjected to cocktail LC–MS/MS analysis (n = 3).

RESULTS

Metabolites profile and identification of pedunculoside and rotundic acid in HLM

In our chromatographic condition, pedunculoside (PE M0, C₃₆H₅₈O₁₀) was eluted at ~ 10.05 min, with measured m/z 695.4026, and it was detected in the ionization mode of [M + FA‐H]⁻ (Table 1, Figures S1 and S2). Rotundic acid (RA M0, C₃₀H₄₈O₅) was eluted at ~ 12.16 min, with measured m/z 487.3437, and it was detected in the ionization mode of [M‐H]⁻ (Table 1, Figures S1 and S2). Eight metabolites of pedunculoside (PE M1–PE M8) and six metabolites of rotundic acid (RA M1–RA M6) were found in the HLM incubation systems (Table 1, Figure 1), with their detailed information shown in Figures S1–S3, Table S2. Pedunculoside and rotundic acid may undergo both phase I and phase II metabolism in HLM through one or two steps of biotransformation; all the metabolites identified were generated only after incubation. For pedunculoside, the metabolic pathways involved were hydroxylation, acetylation, glucuronidation, and glucose conjugation. Among them, hydroxylation and glucose conjugation are reported in both rat liver microsomal and rat liver S9 incubation systems; a glucose‐conjugated metabolite was also found after oral administration of pedunculoside in rats. ²⁴ The metabolic pathways of acetylation and glucuronidation were identified for pedunculoside for the first time. Our results also revealed species differences in pedunculoside metabolism in in vitro microsomal incubation systems, and HLM may provide more biotransformation possibilities for pedunculoside. For rotundic acid, the metabolic pathways involved were hydroxylation, glucuronidation, and their combinations. Previous research has also reported hydroxylated and glucuronidated metabolites after rats were orally administered rotundic acid, and the hydroxylated metabolite was also found in rat liver microsomes and S. racemosum biotransformation. ³¹ , ³² A two‐step metabolite (hydroxylated and glucuronidated rotundic acid) was also reported for the first time.

TABLE 1.

Qualitative information of the identified metabolites of pedunculoside (PE) and rotundic acid (RA) in human liver microsomes.

No.	Ion	R.T. (min)	m/z measured	Formula	m/z predicted	ppm	Metabolic pathway	Fragment ions
PE M0	[M + FA‐H]−	10.05	695.4026	C36H58O10	695.4012	2.0	Prototype	649.3976, 487.3412 a , 469.3330, 207.0508
PE M1	[M + FA‐H]−	7.35	711.3970	C36H58O11	711.3961	1.2	+ O	665.3942, 503.3401 a , 485.3289
PE M2	[M + FA‐H]−	8.04	711.3982	C36H58O11	711.3961	2.9	+ O	665.3964, 503.3400 a , 441.3391 a , 409.3120, 369.2793
PE M3	[M + FA‐H]−	8.39	711.3973	C36H58O11	711.3961	1.7	+ O	665.3867, 503.3388 a , 179.0538
PE M4	[M + FA‐H]−	8.89	711.3966	C36H58O11	711.3961	0.7	+ O	665.3954, 503.3399 a , 485.3275
PE M5	[M + FA‐H]−	10.83	737.4147	C38H60O11	737.4118	4.0	+ COCH3	691.4104, 529.3565, 487.3445 a
PE M6	[M‐H]−	9.57	825.4306	C42H66O16	825.4278	3.4	+ GluA	663.3781 a
PE M7	[M‐H]−	10.65	825.4304	C42H66O16	825.4278	3.1	+ GluA	487.3447, 337.0784 a , 319.0678
PE M8	[M + FA‐H]−	9.15	857.4576	C42H68O15	857.4540	4.2	+ Glu	811.4533 a , 487.3437 a , 323.0980, 179.0555
RA M0	[M‐H]−	12.16	487.3437	C30H48O5	487.3429	1.6	Prototype	469.3323 a , 443.3527, 437.3062, 405.3148, 393.3183
RA M1	[M‐H]−	9.43	503.3384	C30H48O6	503.3378	1.2	+ O	485.3263 a , 383.2957, 355.2647, 351.2692
RA M2	[M‐H]−	11.47	503.3368	C30H48O6	503.3378	−2.0	+ O	485.3284 a , 435.2903
RA M3	[M‐H]−	10.15	663.3762	C36H56O11	663.3750	1.8	+ GluA	487.3410 a , 469.3324, 175.0258, 113.0256
RA M4	[M‐H]−	8.09	679.3693	C36H56O12	679.3699	−0.9	+ O + GluA	503.3386, 441.3377 a , 409.3108, 175.0251, 113.0247
RA M5	[M‐H]−	8.51	679.3691	C36H56O12	679.3699	−1.2	+ O + GluA	503.3375 a , 441.3359 a , 193.0340, 175.0257, 113.0252
RA M6	[M‐H]−	9.77	679.3683	C36H56O12	679.3699	−2.4	+ O + GluA	503.3373 a , 175.0225, 113.0244

Abbreviations: Glu, glucose; GluA, glucuronic acid.

^a The most important fragment ions in MS/MS spectra.

FIGURE 1.

Proposed metabolic pathways of pedunculoside and rotundic acid in HLM (only one possible isomer of the metabolites was given when more than one possible structure existed).

Metabolic stability study of pedunculoside and rotundic acid in HLM

The results for the metabolic stability of pedunculoside and rotundic acid using HLM are presented in Figure 2. A series of control assays were conducted to demonstrate the reliability of the regular HLM incubation system. As expected, no obvious interference from the matrices or the instrument was observed in the blank control assay (control), whereas in the negative control assay (negative), pedunculoside and rotundic acid were stable for up to 120 min (Figure 2a,d). When incubated with microsomes for up to 120 min at 1 μM, there was no loss of drug concentration, indicating that pedunculoside was stable in the regular HLM incubation system (normal group) and not readily metabolized (Figure 2b,c). On the contrary, rotundic acid was diminished quickly, and less than 10% of rotundic acid remained after 120 min (normal group, under regular conditions) (Figure 2e,f). In the absence of the NADPH regeneration system (NADPH‐), we observed a change in the metabolic rate of rotundic acid, with approximately 25% of the compound remaining after 120 min (Figure 2g,h). Conversely, when UDPGA was absent (UDPGA‐), only ~ 50% of rotundic acid was metabolized after incubation for 120 min (Figure 2i,j). When in the absence of the NADPH regeneration system (NADPH‐), phase I metabolism could not occur, and the decrease in rotundic acid was attributed to phase II metabolism; conversely, under the condition when UDPGA (a phase II metabolism cofactor, especially for glucuronidation) was lacking, the decrease in rotundic acid was solely influenced by phase I metabolism. These results demonstrate that both phase I and phase II metabolism are likely to occur in the incubation system, with phase II metabolism being the major metabolic pathway of rotundic acid, as indicated by the differences in metabolic rates.

FIGURE 2.

Metabolic stability study of (a–c) pedunculoside and (d–j) rotundic acid in the HLM incubation system (0–120 min, initial concentration of pedunculoside and rotundic acid at ~ 1 μM, n = 3, the mean ± SD).

Determination of the phase I enzymes responsible for the metabolism of pedunculoside and rotundic acid

In the LC–MS/MS analytical conditions of our preliminary metabolic enzyme identification, three hydroxylated metabolites of pedunculoside and two hydroxylated metabolites of rotundic acid were detected. The retention times for hydroxylated pedunculoside (PE + O) metabolites were ~ 5.37 min [PE + O (1)], 5.64 min [PE + O (2)], and 5.94 min [PE + O (3)], and for hydroxylated rotundic acid metabolites (RA + O) were ~ 6.18 min [RA + O (1)] and 7.91 min [RA + O (2)], respectively (Figure S4).

As a result, when furafylline, ticlopidine, sulfaphenazole, and quinidine were added to the incubation system at different concentration ranges, the generation rate of the corresponding hydroxylated metabolites did not change much (Figure S5), so CYP1A2, CYP2B6/CYP2C19, CYP2C9, and CYP2D6 were not responsible for the hydroxylation reactions. When ketoconazole was added to the incubation, the generation rate of the corresponding metabolites decreased with an increase in the concentration of ketoconazole to a different extent (Figure 3a–e). Ketoconazole inhibited the generation of the five hydroxylated metabolites in a dose‐dependent manner, up to 65%–95% when the concentration of ketoconazole was at 1 μM, indicating the corresponding hydroxylation reactions were mainly catalyzed by CYP3A4/5. The generation rate of PE + O (1) also decreased with increasing concentration of quercetin, and the generation rate was approximately 45% when the concentration of quercetin was 10 μM; for PE + O (2), the generation rate decreased to approximately 60% at a quercetin concentration of 10 μM (Figure 3f–j), suggesting that CYP2C8 also contributes to the hydroxylation of pedunculoside, although it showed lower inhibition (was not the major enzyme).

FIGURE 3.

Effects of (a–e) ketoconazole (0.005–1 μM) and (f–j) quercetin (0.01–10 μM) on the 5 CYP‐mediated hydroxylation reactions of pedunculoside and rotundic acid in HLM (n = 3, the mean ± SD, showed in generation rate–concentration columns).

Determination of the phase II enzymes responsible for the metabolism of rotundic acid

The retention time for glucuronidated rotundic acid (RA + GluA) was approximately 6.73 min under our LC–MS/MS analytical conditions (Figure S4). Five major recombinant human UGTs (UGT1A1, UGT1A4, UGT1A6, UGT1A9, and UGT2B7) were incubated with the determined concentration of rotundic acid to test whether glucuronidated rotundic acid could be generated through specific UGT‐mediated phase II reactions. According to the results, except for the positive control assay, which was conducted with pooled HLM, glucuronidated rotundic acid was only detected in the presence of recombinant human UGT1A4 (Figure 4b,c, Figure S6). It was not detected in either the negative control assay or when co‐incubated with other recombinant UGT enzymes (Figure 4a, Figure S6). Different protein concentrations of recombinant human UGT1A4 were tested. It was found that the generation of glucuronidated rotundic acid increased with higher protein concentrations of UGT1A4, regardless of whether the rotundic acid concentration was at 10 or 50 μM (Figure S7). The results demonstrated that the glucuronidation of rotundic acid was catalyzed by UGT1A4, although, other minor UGT isoforms we did not test may also catalyze this metabolic reaction.

FIGURE 4.

Representative extracted ion chromatograms (EICs) for glucuronidated rotundic acid in the incubation system of (a) negative control, PBS, (b) positive control, HLM, (c) recombinant human UGT1A4.

Inhibitory DDI evaluation

The inhibitory potential of both pedunculoside and rotundic acid against the major CYP isoforms was tested based on the inhibitory DDI evaluation system established and verified in our previous study, in which seven CYP enzymes involving eight CYP‐catalyzed reactions of eight probe substrates were included. ³⁰ As shown in Figures 5 and 6, pedunculoside and rotundic acid did not show significant inhibition of any of the CYPs tested in the concentration range of 0.05–50 μM, and the percentages of remaining activity were mostly between 90% and 110%. Only rotundic acid showed slight inhibition of CYP3A4/5 mediated midazolam 1′‐hydroxylation at concentrations of 10, 20, and 50 μM, and the average percentages of remaining activity decreased to 88.17%, 86.21%, and 84.45%, respectively. For each CYP isoform, the IC₅₀ value was far more than 50 μM, although it could not be exactly calculated or estimated from the current data available. Considering the routine dose and the corresponding plasma concentration reported [C _max of pedunculoside was between 100 and 500 ng/mL (equals to 0.154–0.768 μM) after oral administration], ¹⁹ , ²⁰ , ²¹ , ²² higher concentrations were not tested. Our results indicate that pedunculoside and rotundic acid are unlikely to alter the metabolism of co‐administered drugs metabolized by major CYP enzymes at a normal concentration of 0.05–50 μM.

FIGURE 5.

Inhibition effects of pedunculoside on the eight selected CYP‐mediated probe reactions in the cocktail incubation system (0.05–50 μM, n = 3, the mean ± SD).

FIGURE 6.

Inhibition effects of rotundic acid on the eight selected CYP‐mediated probe reactions in the cocktail incubation system (0.05–50 μM, n = 3, the mean ± SD).

DISCUSSION

Rotundic acid is not only the aglycon of pedunculoside but is also reported as the main metabolite of pedunculoside. A comprehensive understanding of the metabolic pathway associated with rotundic acid enhances our understanding of the holistic metabolic fate of pedunculoside. In this study, we conducted a systematic metabolic analysis of both pedunculoside and rotundic acid. Firstly, the metabolic profiles of pedunculoside and rotundic acid were studied in the HLM incubation system. Pedunculoside was found to be most stable in HLM. Eight metabolites of pedunculoside were detected and tentatively identified, and the metabolic pathways involved were hydroxylation, acetylation, glucuronidation, and glucose conjugation. Acetylation and glucuronidation pathways were reported for the first time. Rotundic acid was easily metabolized by HLM, and its phase II metabolism was more dominant than phase I metabolism. Six metabolites of rotundic acid were found and tentatively identified, and the metabolic pathways involved were hydroxylation, glucuronidation, and their combination; a two‐step metabolite (hydroxylated and glucuronidated rotundic acid) was also reported for the first time. Next, the key metabolic enzymes that participate in pedunculoside and rotundic acid metabolism were investigated, followed by metabolite identification. The NADPH regenerating system was proved to be essential in the hydroxylation metabolism of pedunculoside and rotundic acid, and UDPGA was proved to be essential in the glucuronidation metabolism of rotundic acid. Through co‐incubation with known specific inhibitors for major CYP isoforms, we found that the hydroxylation of pedunculoside was catalyzed mainly by CYP3A4/5 and partly by CYP2C8, whereas the hydroxylation of rotundic acid was mainly catalyzed by CYP3A4/5. Through co‐incubation with major recombinant human UGTs, we found that UGT1A4 could mediate the glucuronidation of rotundic acid. Finally, the potential inhibitory effects of pedunculoside and rotundic acid on the major CYP isoforms were investigated using an established cocktail incubation system consisting of seven major CYP enzymes involving eight CYP‐catalyzed reactions. Both pedunculoside and its aglycon rotundic acid showed no obvious inhibition against the selected CYP isoforms at 0.05–50 μM. Based on in vitro data, there is a low potential risk of pedunculoside and rotundic acid causing clinical DDIs by inhibiting co‐administered drugs metabolized by these CYP enzymes. Taken together, these data add to our knowledge of pedunculoside and rotundic acid metabolism and provide valuable information for their clinical application. This study also provides a basis for investigating the preclinical metabolism and metabolic DDI profile of other herbal medicines.

Pedunculoside and rotundic acid are major constituents of the genus Ilex. In China, several Ilex species, including Ilex rotunda, Ilex chinensis, and Ilex pubescens are used as herbal medicine. However, Ilex species are not limited to China; they also inhabit temperate and tropical regions around the world, including North and South America, tropical and temperate Asia, and parts of Europe and Oceania. ² , ³ As the use of herbal medicine becomes more widespread across various countries and regions, Ilex plants are increasingly likely to be used for the treatment or prevention of cardiovascular diseases and gastrointestinal diseases, either as medicinal materials in herbal products or dietary supplements. Therefore, it' is important to understand the metabolic profile of pedunculoside and rotundic acid, the major constituents of the genus Ilex.

Pedunculoside and rotundic acid were found to have no inhibitory effect on common CYP enzymes and are therefore unlikely to cause drug–herbal interactions mediated by these enzymes. However, our study indicated that they are metabolized by CYP3A4/5, CYP2C8, and UGT1A4. When co‐administrated with other drugs or medications, the metabolic behavior of pedunculoside and rotundic could be changed if certain enzyme inducers or inhibitors are present. For example, clarithromycin, itraconazole, and ketoconazole, which are used to treat fungal or bacterial infections, inhibit CYP3A4/5. Gemfibrozil and clopidogrel, used for the treatment of cardiovascular diseases, by regulating blood lipids and antiplatelet, respectively, inhibit CYP2C8. Additionally, diclofenac, a nonsteroidal anti‐inflammatory drug, is reported to inhibit UGT1A4. When pedunculoside and rotundic acid are co‐administered with these drugs, there is a potential risk of accumulation and an increased likelihood of adverse reactions. As more herbal compounds are used, it is crucial to pay closer attention to drug–herbal interactions to ensure their safe and rational clinical use.

AUTHOR CONTRIBUTIONS

L.W. and Z.Z. wrote the manuscript; L.W. and Z.Z. designed the research; L.W., L.D., X.W., Y.L., X.S., P.W., S.X. and Z.F. performed the research; L.W., L.D., X.W. and Y.L. analyzed the data; L.W. contributed new reagents/analytical tools.

FUNDING INFORMATION

L.W. was supported by the Natural Science Foundation for Colleges and Universities of Jiangsu Province of China [20KJB360010] and the China Postdoctoral Science Foundation [2018M630591]. L.D. was supported by the Innovation and Entrepreneurship Training Program for College Students of Jiangsu Province of China [202110315040Y].

CONFLICT OF INTEREST STATEMENT

The authors declared no competing interests for this work.

Supporting information

Data S1:

Data S2:

Wu L, Dong L, Zhou Z, et al. Preclinical metabolism and metabolic drug–drug interaction profile of pedunculoside and rotundic acid. Clin Transl Sci. 2024;17:e70043. doi: 10.1111/cts.70043