Regioselective Hydroxylation of Oleanolic Acid Catalyzed by Human CYP3A4 to Produce Hederagenenin, a Chiral Metabolite

: Oleanolic acid (OA) is a pentacyclic triterpenoid widely found in plants and foods as an aglycone of triterpenoid saponins or as a free acid. OA exhibits beneﬁcial activities for humans, including antitumor, antivirus, and hepatoprotection properties without apparent toxicity. The metabolites produced by the cytochrome P450 (P450) enzymes are critical for the evaluation of the efﬁcacy and safety of drugs. In this study, the potential metabolites of OA were investigated by P450-catalyzed oxidation reactions. Among the various tested human P450s, only human CYP3A4 was active for the hydroxylation of OA. The major metabolite was characterized by a set of analyses using HPLC, LC–MS, and NMR. It was found to be 4-epi-hederagenenin, a chiral product, by regioselective hydroxylation of the methyl group at the C-23 position. These results indicated that CYP3A4 can hydroxylate an OA substrate to make 4-epi-hederagenenin. Possible drug–food interactions are discussed.


Introduction
The use of natural products to treat chronic conditions such as hepatitis, liver disorder, or cancer has attracted the attention of researchers [1]. Native triterpenoids are of interest due to their availability and multiple biological activities, including antibacterial, anti-inflammatory, antitumor, and hepatoprotection properties. However, triterpenoids are highly hydrophobic, which significantly limits their uses as effective therapeutic agents. One of the most common ways to increase the bioavailability of these triterpenoid compounds is a hydroxylation reaction to add a hydrophilic hydroxyl group to carbon [2].
Oleanolic acid (OA) is a pentacylic triterpenoid ( Figure 1). It is part of a group of natural compounds with diverse structures, including steroids, sterols, and triterpenoid saponins [3]. OA has been found in nearly 2000 types of plants [4]. In addition to its ability to exist as a free acid, OA is also used as a precursor to triterpenoid saponins [5]. It exhibits both pro-and anti-inflammatory properties based on the dose and chemical structure. OA has several beneficial effects, including hepatoprotective, antiviral, and antitumor properties [6][7][8][9][10]. Currently, OA is considered as an important ingredient in alternative therapy in the research, management, and treatment of chronic diseases [11]. Hederagenin is also a pentacyclic triterpenoid ( Figure 1). It is found in the Hedera helix and exists as sapogenins or saponins in many plants. Hederacoside C and alphahederin are the most popular glycosides of hederagenin [12]. Hederagenin has several beneficial effects, including antiapoptotic, anti-inflammatory, anticoagulation, antiatherosclerosis, and antidiabetic activities [13].
Cytochrome P450 (CYP or P450) proteins are a large enzyme family found almost everywhere, including bacteria, plants, animals, and humans [14,15]. P450s play important roles in the metabolism of drugs, carcinogens, and steroids [16]. It has been reported that P450 (moxA) from Nonomuraea recticatena, coexpressed with camAB for pseudomonad redox partners in Escherichia coli, hydroxylated OA to produce queretaroic acid, which is a monohydroxylated OA at the C-30 position [17]. Of the human P450 enzymes, CYP3A4 is known as one of the most active in terms of the metabolism of xenobiotics in the human body.
The purpose of this work was to find an enzymatic strategy for an efficient synthesis of hydroxylated OA derivatives with a high potential of biological activity from OA, an inexpensive substrate, using P450. We found that OA is regioselectively hydroxylated by CYP3A4 to produce hederagenin ( Figure 1).

OA Metabolism by Human Liver Microsomes and Identification of the Major Metabolite
The initial investigations to determine the catalytic activity of P450 enzymes in OA hydroxylation were performed using high-performance liquid chromatography (HPLC). First, to verify the ability of human liver microsomes (HLMs) to hydroxylate OA, the reaction was performed with 500 µM of OA and HLMs at 37 °C for 60 min. Figure 2 shows the HPLC results of OA and its metabolites by HLMs. The retention time of OA was 32.3 min, and a major metabolite of HLMs in the presence of β-nicotinamide adenine dinucleotide phosphate (reduced form) (NADPH) showed a retention time of 22.4 min. When the OA metabolites were formed with the CYP3A4 enzyme in the mixture in the presence of NADPH, the same peak of the metabolite was observed. Hederagenin is also a pentacyclic triterpenoid (Figure 1). It is found in the Hedera helix and exists as sapogenins or saponins in many plants. Hederacoside C and alpha-hederin are the most popular glycosides of hederagenin [12]. Hederagenin has several beneficial effects, including antiapoptotic, anti-inflammatory, anticoagulation, antiatherosclerosis, and antidiabetic activities [13].
Cytochrome P450 (CYP or P450) proteins are a large enzyme family found almost everywhere, including bacteria, plants, animals, and humans [14,15]. P450s play important roles in the metabolism of drugs, carcinogens, and steroids [16]. It has been reported that P450 (moxA) from Nonomuraea recticatena, coexpressed with camAB for pseudomonad redox partners in Escherichia coli, hydroxylated OA to produce queretaroic acid, which is a monohydroxylated OA at the C-30 position [17]. Of the human P450 enzymes, CYP3A4 is known as one of the most active in terms of the metabolism of xenobiotics in the human body.
The purpose of this work was to find an enzymatic strategy for an efficient synthesis of hydroxylated OA derivatives with a high potential of biological activity from OA, an inexpensive substrate, using P450. We found that OA is regioselectively hydroxylated by CYP3A4 to produce hederagenin ( Figure 1).

OA Metabolism by Human Liver Microsomes and Identification of the Major Metabolite
The initial investigations to determine the catalytic activity of P450 enzymes in OA hydroxylation were performed using high-performance liquid chromatography (HPLC). First, to verify the ability of human liver microsomes (HLMs) to hydroxylate OA, the reaction was performed with 500 µM of OA and HLMs at 37 • C for 60 min. Figure 2 shows the HPLC results of OA and its metabolites by HLMs. The retention time of OA was 32.3 min, and a major metabolite of HLMs in the presence of β-nicotinamide adenine dinucleotide phosphate (reduced form) (NADPH) showed a retention time of 22.4 min. When the OA metabolites were formed with the CYP3A4 enzyme in the mixture in the presence of NADPH, the same peak of the metabolite was observed.
Next, LC-MS was used to investigate the molecular weight of the major metabolite generated by CYP3A4. A full-scan chromatogram of the mixture reaction at A 210 ( Figure 3A) and the total ion current (TIC) profile ( Figure 3B  Next, LC-MS was used to investigate the molecular weight of the major metabolite generated by CYP3A4. A full-scan chromatogram of the mixture reaction at A210 ( Figure  3A) and the total ion current (TIC) profile ( Figure 3B  NMR experiments were performed with the major metabolite obtained from the reaction to determine the structure based on the chemical changes of the 13 C NMR spectrum. The result of the chemical shifts of 13 C NMR of the metabolite (M1) matched the reported NMR results of 4-epi-hederagenin ( Figure 4, Table S1, Figures S1, and S2) [18,19].  Next, LC-MS was used to investigate the molecular weight of the major metabo generated by CYP3A4. A full-scan chromatogram of the mixture reaction at A210 ( Fig  3A) and the total ion current (TIC) profile ( Figure 3B   NMR experiments were performed with the major metabolite obtained from the reaction to determine the structure based on the chemical changes of the 13 C NMR spectrum. The result of the chemical shifts of 13 C NMR of the metabolite (M1) matched the reported NMR results of 4-epi-hederagenin ( Figure 4, Table S1, Figures S1 and S2) [18,19]. The results showed the assigned 13 C NMR chemical shifts of the metabolite (M1) with atomic positions (Figure 4 and Table S1). The results showed the assigned 13 C NMR chemical shifts of the metab atomic positions (Figure 4 and Table S1).
When the pH-dependence of 4-epi-hederagenin formation by examined, the pH range of 7.4-9 was optimal with the maximal activity a S3 pH 7.4, a neutral pH

Kinetics Parameters and Total Turnover Numbers of OA Hydroxylati and HLMs
We examined several human CYPs, such as CYP1A2, CYP1B1, CYP CYP2D6, CYP2E1, and CYP3A4, to find which CYPs are involved in OA activities. Among them, only CYP3A4 was found to have an apparent activity of 0.42 min −1 at 500 µM of OA ( Figure 5). Other CYPs did not show activity (less than 0.05 min −1 ) ( Figure S4). Both CYP3A4 and HLMs were used to determine the kinetic par When the pH-dependence of 4-epi-hederagenin formation by CYP3A4 was examined, the pH range of 7.4-9 was optimal with the maximal activity at pH 8 ( Figure S3). All of the experiments in this study were done at pH 7.4, a neutral pH.

Kinetics Parameters and Total Turnover Numbers of OA Hydroxylation by CYP3A4 and HLMs
We examined several human CYPs, such as CYP1A2, CYP1B1, CYP2C9, CYP2C19, CYP2D6, CYP2E1, and CYP3A4, to find which CYPs are involved in OA hydroxylation activities. Among them, only CYP3A4 was found to have an apparent hydroxylation activity of 0.42 min −1 at 500 µM of OA ( Figure 5). Other CYPs did not show any apparent activity (less than 0.05 min −1 ) ( Figure S4). The results showed the assigned 13 C NMR chemical shifts of the metabolite (M1) with atomic positions (Figure 4 and Table S1). When the pH-dependence of 4-epi-hederagenin formation by CYP3A4 was examined, the pH range of 7.4-9 was optimal with the maximal activity at pH 8 ( Figure  S3). All of the experiments in this study were done at pH 7.4, a neutral pH.

Kinetics Parameters and Total Turnover Numbers of OA Hydroxylation by CYP3A4 and HLMs
We examined several human CYPs, such as CYP1A2, CYP1B1, CYP2C9, CYP2C19, CYP2D6, CYP2E1, and CYP3A4, to find which CYPs are involved in OA hydroxylation activities. Among them, only CYP3A4 was found to have an apparent hydroxylation activity of 0.42 min −1 at 500 µM of OA ( Figure 5). Other CYPs did not show any apparent activity (less than 0.05 min −1 ) ( Figure S4). The results showed that the catalyzed efficiency of CYP3A4 is 31-fold higher than that of HLMs. These results suggested that CYP3A4 may be an important CYP in the catalytic conversion of OA to 4-epi-hederagenin.  CYP3A4 is 31-fold higher than that of HLMs. These results suggested that CYP3A4 may be an important CYP in the catalytic conversion of OA to 4-epi-hederagenin. When a time profile of the reactions was determined by prolonging the incubation time to 240 min, the formation of 4-epi-hederagenin increased with the reaction time. The total turnover number (TTN, nmol product/nmol enzyme) reached 40 at 210 min into the reaction ( Figure 6). When a time profile of the reactions was determined by prolong time to 240 min, the formation of 4-epi-hederagenin increased with the total turnover number (TTN, nmol product/nmol enzyme) reached 40 a reaction ( Figure 6).

Inhibition of OA Hydroxylation Activity by Antibodies in HLMs
Immunoinhibition experiments with anti-CYP3A4 were performed role of CYP3A4 in OA hydroxylation. The results showed th concentrations of anti-CYP3A4 inhibited OA hydroxylation immunoinhibitory effect of anti-CYP3A4 on OA hydroxylation in HLM at 42 and 51% inhibition when 5 and 10 mg IgG per nmol of P respectively. The results indicated that CYP3A4 plays an impo hydroxylation of OA in HLMs. This also suggested that other CYPs in ad may be involved in OA hydroxylation. tivity 100 Figure 6. Total turnover numbers (TTNs) for OA hydroxylation by CYP3A4. The TTNs by CYP3A4 were measured by HPLC after the reaction at 37 • C during the indicated time of 10-240 min. The reaction mixture included 0.2 µM of CYP3A4, an NADPH regeneration system, and 0.5 mM of substrate in final volume of 0.25 mL in potassium phosphate buffer (100 mM, pH 7.4).

Inhibition of OA Hydroxylation Activity by Antibodies in HLMs
Immunoinhibition experiments with anti-CYP3A4 were performed to investigate the role of CYP3A4 in OA hydroxylation. The results showed that increasing the concentrations of anti-CYP3A4 inhibited OA hydroxylation (Figure 7). The immunoinhibitory effect of anti-CYP3A4 on OA hydroxylation in HLMs was fairly high, at 42 and 51% inhibition when 5 and 10 mg IgG per nmol of P450 were treated, respectively. The results indicated that CYP3A4 plays an important role in the hydroxylation of OA in HLMs. This also suggested that other CYPs in addition to CYP3A4 may be involved in OA hydroxylation.
When 1-10 mg IgG of anti-CYP2C19 per nmol of P450 was added to the reaction mixtures, the anti-CYP2C19 did not show an apparent inhibition of OA hydroxylation up to 5 mg IgG per nmol of P450. When 10 mg IgG was used, only a 9% inhibition was shown; the inhibition by the anti-CYP2C9 seemed to be nonspecific. When the preimmune IgG was treated with up to 10 mg per nmol of P450, no apparent inhibition was found. at 42 and 51% inhibition when 5 and 10 mg IgG per nmol of P450 w respectively. The results indicated that CYP3A4 plays an important hydroxylation of OA in HLMs. This also suggested that other CYPs in additio may be involved in OA hydroxylation. Regarding the OA oxidation reactions, only a few results have been reported. OA could be hydroxylated at C-30 to produce queretaroic acid, a metabolite of OA produced in several plants, via bacterial P450 (moxA) from Nonomuraea recticatena [17]. Escherichia coli whole-cell and in vitro reactions were performed using the coexpression of P450 (moxA) and Pseudomonas redox partners. However, in this study, we found that human CYP3A4 can hydroxylate at C-23 to produce 4-epi-hederagenin, showing a distinct regioselectivity of the CYP3A4 toward OA. The hydroxide positions for CYP3A4 and P450 (moxA) are the farthest within the OA molecule. This result showed the diverse catalytic activities of P450 from different sources.

Antibody (mg IgG
Hederagenin has been shown to be easily digested and absorbed. It also has many pharmacological benefits such as anti-inflammatory, antidegenerative (for nerves), and antidiabetic properties [13]. In vitro and in vivo experiments have shown that hederagenin induces apoptosis effectively in cisplatin-resistant HNC cells by targeting the Nrf2-ARE antioxidant pathway [20]. These results suggested that hederagenin, a metabolite from OA, may show new and different effects than OA. Further studies with 4-epi-hederagenin need to be done to better understand its role in humans, especially in the treatment of conditions such as cancer and inflammation. Furthermore, other metabolites of OA in addition to hederagenin and queretaroic acid should be studied to find their beneficial pharmacological effects using engineered P450s.
Human P450 plays an important role in the metabolism of drugs and other compounds. Previous studies have shown that CYP3A4 plays a role in the metabolism of many anti-inflammatory and anticancer drugs [21]. In this study, we found a catalytic ability of CYP3A4 to metabolize OA to 4-epi-hederagenin. Notably, OA was found to inhibit the CYP3A4-catalyzed midazolam 1-hydroxylation reaction by 60% at 100 µM [22]. Therefore, drug-food interactions should be considered when consuming foods or vegetables containing OA. The treatment of clinical drugs with OA will inhibit CYP3A4 s ability to function in the body or may cause significant side effects. More research is needed to study the effects of OA on other common foods and supplements that are also likely to be metabolized by CYP3A4, as it is the most common drug-metabolizing enzyme in the small intestine and liver.

Materials
The OA hydrate was purchased from Tokyo Chemical Industry (Tokyo, Japan). The betulinic acid (BA), β-nicotinamide adenine dinucleotide phosphate (reduced form) (NADPH), glucose-6-phosphate dehydrogenase, and glucose-6-phosphate were bought from Sigma-Aldrich (St. Louis, MO, USA). The methanol, formic acid, acetonitrile, and other solvents were purchased from Fisher Scientific (Houston, TX, USA). Other chemicals and solvents used in the research were of high-quality analytical grade and were bought from other companies with no further purification.
The HLMs were purchased from Thermo Fisher Scientific (Waltham, MA, USA). The antibodies against CYP3A4 and CYP2C19 have been studied and implemented in previous reports [23]. The CYPs of CYP1A2, CYP1B1, CYP2C9, CYP2C19, CYP2D6, CYP2E1, and CYP3A4 were expressed in Escherichia coli with a pCW vector containing human P450 cDNA and rat NADPH-P450 reductase (CPR). Human P450 cDNA and pCW vector expression was performed as described in previous studies [24].

Hydroxylation of OA by CYP3A4 and HLMs
The reaction mixtures contained 0.20 µM of CYP3A4 or 0.80 µM of HLMs in 0.25 mL of potassium phosphate buffer (100 mM, pH 7.4) with a final concentration of 0.5 mM of OA. An NADPH-generating system was used to start the reaction, with final concentrations of 10 mM of glucose-6-phosphate, 1 IU of yeast glucose-6-phosphate dehydrogenase per mL, and 0.5 mM of NADP + .
The samples were incubated for 30 min for CYP3A4 and 60 min for HLMs with or without NADPH at 37 • C, and the reaction was stopped with 0.60 mL of ice-cold ethyl acetate. The BA (50 µM) was added as an internal standard to this solution, followed by centrifugation (1000× g, 20 min). The metabolites were analyzed by HPLC and quantified via comparison to the internal standard, BA ( Figures S4 and S5).
Mobile phase A was water containing 0.1% of formic acid and methanol (20:80, v/v), and mobile phase B was acetonitrile; mobile phase A/B (95:5, v/v) had a flow rate of 0.7 mL/min through a gradient pump (LC-20AD; Shimadzu, Kyoto, Japan). The UV wavelength was 210 nm [25].
To study the TTN for CYP3A4, 500 µM of OA was used. The reaction was started by the addition of an NADPH-generating system and incubated at 10, 20, 30, 60, 90, 120, 150, 180, 210, and 240 min at 37 • C. The formation rate of the OA was determined by HPLC as described above.
The kinetic parameters of the OA hydroxylation by CYP3A4 were determined by the reactions, including 10-1000 µM of OA in 100 mM of potassium phosphate buffer (pH 7.4) and 0.20 µM of enzymes. The NADPH-generating systems were added to the initial reaction, and the reaction mixtures were incubated for 60 min at 37 • C. The kinetic parameter results were analyzed using GraphPad Prism software (Graph, San Diego, CA, USA).

LC-MS Analysis of OA Metabolite
To detect OA metabolites by CYP3A4, an LC-MS analysis of the metabolites was executed to compare the fragmentation patterns and LC profiles with authentic compounds. The CYP3A4 was incubated with 500 µM of OA at 37 • C for 2 h with an NADPH-generating system, and the injection samples were prepared as described above. An aliquot (7 µL) of this solution was injected into the LC column. The LC-MS analysis was carried out in electrospray ionization (negative) mode on a Shimadzu LC-MS-2010 EV system (Shimadzu Corporation, Japan) with the LC-MS solution software. The separation was performed on a Shim-pack VP-ODS column (2.0 mm × 250 mm, Shimadzu Corporation, Japan). Mobile phase A was water containing 0.1% of phosphoric acid and methanol (20:80, v/v), and mobile phase B was methanol and acetonitrile (80:20, v/v); mobile phase A/B (77:23, v/v) was delivered at a flow rate of 0.16 mL/min. The retention times for the major metabolite and OA were 48.85 and 71.133 min, respectively. The interface and detector voltages were 4.4 kV and 1.7 kV, respectively. The nebulization gas flow was set at 1.5 L/min. The interface, curve desolvation line, and heat block temperatures were 250, 250, and 200 • C, respectively.

NMR Spectroscopy Analysis of OA Metabolite
The NMR experiments were performed at an ambient temperature on a Varian VN-MRS 600 MHz NMR spectrometer equipped with a carbon-enhanced cryogenic probe. Pyridine-d 5 was used as a solvent, and the chemical shifts for proton NMR spectra were measured in parts per million (ppm) relative to tetramethylsilane. All of the NMR experiments were performed with standard pulse sequences in the VNMRJ (v. 3.2) library and processed with the same software.

Immunoinhibition of Antibodies on OA Hydroxylation Activity
The experiment was performed by incubating a mixture of anti-P450 IgG (anti-CYP3A4 and anti-CYP2C19) at different concentrations at room temperature for 30 min, then adding ingredients such as OA and the NADPH required for the reaction. Controlled experiments with different concentrations of preimmune IgG have been performed as described previously [26]. The substrates and products were tested with HPLC; mobile phase A was water containing 0.1% of formic acid and methanol (20:80, v/v), and mobile phase B was acetonitrile; mobile phase A/B (95:5, v/v) had a flow rate of 0.7 mL/min through a gradient pump (LC-20AD; Shimadzu, Kyoto, Japan). The UV wavelength was 210 nm.

Conclusions
In this study, the potential metabolites of OA were investigated via human P450catalyzed oxidation reactions. Among the tested human P450s, only CYP3A4 was active in the hydroxylation of OA. The major metabolite was characterized by a set of analyses using HPLC, LC-MS, and NMR and was found to be 4-epi-hederagenenin, a chiral product of regioselective hydroxylation of the methyl group at the C-23 position. These results indicate that CYP3A4 can hydroxylate an OA substrate to make 4-epi-hederagenenin. Possible drug-food interactions should be considered when consuming food or vegetables containing OA.