Monoester-Diterpene Aconitum Alkaloid Metabolism in Human Liver Microsomes: Predominant Role of CYP3A4 and CYP3A5

Aconitum, widely used to treat rheumatoid arthritis for thousands of years, is a toxic herb that can frequently cause fatal cardiac poisoning. Aconitum toxicity could be decreased by properly hydrolyzing diester-diterpene alkaloids into monoester-diterpene alkaloids. Monoester-diterpene alkaloids, including benzoylaconine (BAC), benzoylmesaconine (BMA), and benzoylhypaconine (BHA), are the primary active and toxic constituents of processed Aconitum. Cytochrome P450 (CYP) enzymes protect the human body by functioning as the defense line that limits the invasion of toxicants. Our purposes were to identify the CYP metabolites of BAC, BMA, and BHA in human liver microsomes and to distinguish which isozymes are responsible for their metabolism through the use of chemical inhibitors, monoclonal antibodies, and cDNA-expressed CYP enzyme. High-resolution mass spectrometry was used to characterize the metabolites. A total of 7, 8, and 9 metabolites were detected for BAC, BMA, and BHA, respectively. The main metabolic pathways were demethylation, dehydrogenation, demethylation-dehydrogenation, hydroxylation and didemethylation, which produced less toxic metabolites by decomposing the group responsible for the toxicity of the parent compound. Taken together, the results of the chemical inhibitors, monoclonal antibodies, and cDNA-expressed CYP enzymes experiments demonstrated that CYP3A4 and CYP3A5 have essential functions in the metabolism of BAC, BMA, and BHA.


Introduction
Aconitum has been used as an essential drug in China and in other East Asian countries for centuries owing to its extremely excellent effect against rheumatosis and rheumatoid arthritis [1,2]. It was reported that among 500 known prescriptions, the frequency of use of the Aconitum is 13.20%, ranked 9 [3]. Nevertheless, Aconitum is a toxic herb that can cause fatal cardiac poisoning [4][5][6] and is reportedly involved in suicidal and homicidal attempts [6,7]. Since the therapeutic dose of Aconitum is very close to its toxic dose, its poisoning incidents are not uncommon. According to the literature data, in 188 cases of patients with Aconitum, 111 patients had varying degrees of toxic reactions [3]. The toxicological mechanism of Aconitum is widely recognized to be associated with voltage-dependent Na + channels [8].
So the death of patients suffering from Aconitum toxicities always is cardiac arrest or respiratory insufficiency. Up to now, there is no effective antidote to Aconitum intoxication. The initial toxic signs are gastrointestinal including nausea, vomiting, and diarrhea. This is followed by a sensation of burning, tingling, and numbness in the mouth and face, and of burning in the abdomen [9,10].
Therefore, prevention of Aconitum toxicity is a key issue for its safe application. Efflux transporters and metabolic enzymes act as the human body's first line of protection by limiting the invasion of toxicants [11]. Investigation on the roles of efflux transporters and metabolic enzymes in Aconitum is meaningful.
MDAs are the primary active and toxic constituents of processed Aconitum. The alkaloid content in processed Aconitum is reportedly as follows: BMA > HA > BAC > MA > AC [13]. Aside from DDAs, MDAs are detectable in human serum and urine after oral intake of Aconitum extract [14]. Moreover, MDAs are evident in the plasma, liver, and kidney after the administration of DDAs, indicating that DDAs are chemically and biologically hydrolyzed into MDAs [2,15]. BMA could reportedly recover the impaired resistance of thermally injured mice infected by HSV type 1 or Candida albicans [16]. The toxicity of MDAs is much less than those of DDAs (the human lethal dose is 1 mg to 4 mg) [17,18]. However, the toxic reaction will still occur in the event of excessive MDA absorption [12]. Frequent cases of poisoning after the intake of processed Aconitum are encountered in clinics.
Previous studies suggested that DDAs could be metabolized into less toxic products by CYP enzymes (the major function of CYP3A) in the human liver microsomes [19][20][21]. The main metabolic pathways of DDAs include demethylation, dehydrogenation, hydroxylation, and didemethylation [19][20][21]. Efflux transporters such as P-glycoprotein, breast cancer resistance protein, and multidrug resistanceassociated protein 2 were involved in the transport of DDAs [11]. Thus, the toxicity of DDAs is reduced by the effect of metabolic enzymes and efflux transporters [11,[19][20][21]. We investigated the role of efflux transporters in the transport of MDAs and found that efflux transporters did not mediate their transport [11]. This result indicates that the invasion of MDAs could not be limited by efflux transporters. Efflux transporters and metabolic enzymes act as the human body's first line of protection by limiting the invasion of toxicants [11]. However, thus far, the effect of metabolic enzymes on MDAs is unknown. Therefore, investigating the metabolic mechanism of MDAs for clinical practice is significant.
We tested phase I and phase II metabolism of MDAs (BAC, BMA, and BHA) by human liver microsomes in preliminary experiments. No phase II metabolites but several oxidative metabolites of BAC, BMA, or BHA were detected in the reaction systems. This study aims to identify the CYP metabolites of BAC, BMA, and BHA in human liver microsomes as well as the CYP isozymes responsible for their metabolism. High-resolution mass spectrometry was used to characterize the structures of their metabolites. Moreover, chemical inhibitors of specific CYP enzymes, CYP-specific inhibitory monoclonal antibodies, and cDNA-expressed CYP enzymes (CYP1A2, 2C8, 2C9, 2C19, 2D6, 2E1, 3A4, and 3A5) were used to verify which isozymes mediate in the metabolism of the CYP metabolites. These results on MDAs metabolism provide important data for the safe clinical use of processed Aconitum.

Phase I Metabolism of BAC, BMA, and BHA in HLMs.
A typical incubation system contained potassium phosphate (50 mM, pH 7.4), NADP (1.55 mM), 6-P-G (3.3 mM), MgCl 2 (3.3 mM), PDH (0.4 U/mL), HLMs (0.4 mg protein/mL), and BAC, BMA, or BHA (10 M) in a total volume of 500 L. After a 5 min preincubation at 37 ∘ C, the reactions were started by adding NADPH [28]. Incubations were carried out for 90 min in a shaking water bath (150 rpm) at 37 ∘ C and were stopped by the addition of 4 mL ice-cold dichloromethane and 100 L testosterone (200 nM, used as an internal standard). The parent drug and its metabolites were extracted by vortexing for 8 min, after which they were centrifuged for 15 min at 1000 rpm. The supernatant was drawn into another tube and evaporated to dryness. The residue was dissolved with 50% methanol in water (125 L) and analyzed by liquid chromatography with tandem mass spectrometry detection (LC-MS/MS). Control incubations were conducted in the absence of a NADPH-regenerating system.
To determine the protein concentration-dependent specificity of the enzyme, HLMs at concentrations of 0.2, 0.4, and 0.8 mg/mL were incubated at 37 ∘ C for 90 min in the above-described reaction system, respectively. To determine the time-dependent specificity, the reactions were processed for 30, 60, 90, and 120 min at 37 ∘ C in the reaction system, respectively. To determine the substrate concentrationdependent specificity, substrates at concentrations of 2.5, 5, 10, 15, and 20 M were incubated in HLMs at 37 ∘ C for 120 min, respectively. Each experiment was performed using three samples.  [22][23][24][25][26][27]. To obtain the maximum possible inhibition for each of the isoforms, the final inhibitor concentrations in the reaction mixtures were much higher than the reported Ki values. The concentration of each inhibitor was 5 M. The inhibitors were preincubated with all incubation mixtures for 5 min at 37 ∘ C before the initiation of the reaction by the addition of BAC, BMA, or BHA [20]. The mixture was then incubated for 120 min at 37 ∘ C. A comparison was made between the incubations of samples with and without the inhibitors. All incubations were carried out in triplicate.

2.4.
Effect of mAbs on the Metabolism of BAC, BMA, and BHA in HLMs. Immunoinhibition studies were conducted by incubating HLMs with various amounts of human selective mAbs (mAb-1A2, mAb-2C8, mAb-2C19, mAb-2D6, mAb-2E1, and mAb-3A4/5). MAbs (10 L) were incubated with HLMs for 15 min on ice before the addition of the parent drug (BAC, BMA, or BHA, 10 M) and the NADPH-regenerating system [21]. The mixture was incubated at 37 ∘ C for 120 min. As a control, comparable incubations were performed with HLMs and Tris buffer (25 mM). All incubations were conducted in triplicate.

Phase I Metabolism of BAC, BMA, and BHA in cDNA-
Expressed CYP Enzymes. The incubation procedures of BAC, BMA, or BHA in cDNA-expressed CYP enzymes were performed as described for HLMs, except that the concentration of the enzyme used was 40 pmol/mL [21,29]. The isozymes used in this study included CYP 1A2, 2C8, 2C9, 2C19, 2D6, 2E1, 3A4, and 3A5. All incubations were performed in triplicate.

Identification and Quantitation of the Metabolites by LC-MS/MS.
A quadrupole time of flight tandem mass spectrometer (Bruker, Daltonics, USA) was used to calculate the molecular weight of the metabolites, which could be deduced through their molecular formulas. The quantitative analysis procedure was as follows. The UPLC conditions were system, Waters Acquity; column, Acquity UPLC BEH C18 column (50 × 2.1 mm, 1.7 m, Waters, Milford, MA, USA); mobile phase A: 0.1% (v/v) formic acid in water; mobile phase B: 100% acetonitrile; gradient, 0 min to 0.5 min at 2 to 2% B, 0.5 min to 2 min at 2% to 20% B, 2 min to 4 min at 20% to 40% B, 4 min to 6.5 min at 40% to 70% B, and 6 min to 7 min at 70% to 2% B; flow rate, 0.35 mL/min; column temperature, 50 ∘ C; and injection volume, 10 L. The MS/MS detector used was a quadrupole tandem mass spectrometer (Waters, USA). Samples were analyzed using electrospray ionization in the positive model. The main working parameters were set as follows: capillary voltage, 3 kV; cone voltage, 50 V; collision voltage, 30 V; source temperature, 120 ∘ C; desolvation temperature, 500 ∘ C; desolvation gas flow, 600 L/Hr; cone gas flow, 50 L/Hr; and collision gas glow, 0.20 mL/min. Data were collected and analyzed by Waters Quantify software (Masslynx 4.1, Waters, USA).

Data
Analysis. Data are expressed as mean ± SD. Oneway analysis of variance (ANOVA) with or without Tukey-Kramer multiple comparisons (post hoc) tests were used to evaluate statistical differences. Differences were considered significant when values were less than 0.05.

Identification of the BAC, BMA, and BHA Metabolites in
HLMs. To determine the best conditions for incubation, we optimized the incubation time, protein concentration, and substrate concentration. According to the largest formation of metabolites, the final protein concentration was selected to be 0.4 mg/mL and the appropriate incubation time was 120 min. Meanwhile, for the substrate concentration-dependent study, the formations of BAC, BMA, and BHA metabolites were all linear at concentrations of 2.5 M to 20 M, indicating that the concentration of 10 M BAC, BMA, or BHA that we chose for this study was suitable for the experiments. Compared with the negative control (without NADPH-regenerating system), a total of 7 BAC metabolites, 8 BMA metabolites and 9 BHA metabolites were found in the HLMs along with the NADPH-regenerating system ( Figure 1). The metabolites were identified by the retention times, the chromatographic behaviors, and characteristic mass spectrometric fragmentation features, which are summarized in Table 1.      dubious peak existed in the same reaction channel, which was eluted at 4.10 min. Its smart molecular formula was calculated to be C 32 H 45 NO 9 , which did not have similar structure with BAC. Meanwhile, this peak could not only be detected with NADPH-regenerating system, but also be detected without NADPH-regenerating system (negative control). Therefore, the peak eluted at 4.10 min was not the CYP metabolite of BAC. We speculated that it was just the endogenous substance of the phase I metabolic system.      was identified as demethyl-dehydrogen-BHA. To speak of, because the peak could be detected with or without NADPHregenerating system (negative control), the peak having a retention time of 4.3 min was not the CYP metabolite of BHA.  (Figure 2(a)), ketoconazole (CYP3A4/5 inhibitor) produced a remarkable effect on the BAC metabolism. The formation of metabolites decreased by at least 70% compared with that of the control. Moreover, amiodarone (CYP2C9 inhibitor) had a less significant function in the formation of metabolites M1, M2, M3, and M6. Fluvoxamine maleate (CYP1A2 inhibitor) also had a minor effect on M2 and M6. However, the other inhibitors did not inhibit the formation of these seven metabolites. Taken together, CYP3A4/5 exhibited a major contribution to BAC metabolism.

BAC
For BMA (Figure 2(b)), compared to the control group, ketoconazole greatly inhibited more than 65% of the formation of all BMA metabolites. All of the BMA metabolites were also significantly inhibited by the CYP 2C9 inhibitor. The production of M1, M2, M3, M4, M7, and M8 was decreased by the inhibitors of CYP2C8 and CYP2C19. Inhibitory effects on M1, M3, and M8 were observed using fluvoxamine maleate (CYP1A2 inhibitor). Quinidine (CYP2D6 inhibitor) had a negligible function in M1, M3, M4, M7, and M8. Diethyldithiocarbamic acid (CYP2E1 inhibitor) did not exhibit obvious inhibition on all eight metabolites. In general, BMA metabolism was predominantly mediated by CYP 3A, followed by CYP2C9 and CYP2C19.
For BHA (Figure 2(c)), ketoconazole strongly inhibited the formation of BHA metabolites by more than 75% compared with the control samples, suggesting that CYP3A4/5 was the major metabolic isozyme for BHA metabolism. The formation of BHA metabolites was significantly inhibited by amiodarone (CYP2C9 inhibitor). Simultaneously, quercetin (CYP2C8 inhibitor) notably reduced the formation of M3, M4, M7, and M8. Omeprazole (CYP2C19 inhibitor)     contributed to the formation of M1, M3, M4, M6, M7, and M9. Moreover, the production of M6, M7, and M9 was influenced by fluvoxamine maleate (CYP1A2 inhibitor). According to the results, although several isozymes were involved in BHA metabolism, CYP3A4/5 was still the primary isozyme.
3.20. The Effect of mAbs on Metabolism of BAC, BMA, and BHA in HLMs. Percentage inhibition due to chemical inhibitors is never 100% and that selectivity is not always significant. Thus, in the next experiment, we used mAbs to further confirm which CYP isozymes mainly contribute to the metabolism of BAC, BMA, and BHA. We added mAbs 15 min prior to NADPH addition to start the reaction and tried to observe the effect of the possible loss of a particular CYP enzyme on the formation of each of the metabolites. As expected, CYP3A4/5 still had the most significant effect and maximally inhibited the metabolite formation compared with the rest of the mAbs (Figure 3). Other mAbs, such as mAb-1A2, mAb-2C8, mAb-2C9, mAb-2C19, mAb-2D6, and mAb-2E1, had no obvious inhibitory effect on the metabolic transformation of BAC, BMA, and BHA. CYP3A4/5 had an extremely important function in the metabolism of BAC, BMA, and BHA.

Phase I Metabolism of BAC, BMA, and BHA in cDNA-
Expressed CYP Enzymes. In the next experiments, rather than inhibiting the CYPs, we tried to incubate the compounds with specific cDNA-expressed CYPs bought from BD Life Sciences to reconfirm our chemical inhibitor and monoclonal antibody data. As anticipated, CYP3A4 still had the most significant effect, and 3A5 showed high formation rates for all the metabolites of BAC, BMA, and BHA ( Figure 4). And the results indicated that the rate of BAC, BMA and BHA metabolism by CYP 3A4 was 32.96%±3.51%, 19.48%±2.02%, and 29.81% ± 3.54%, respectively.

Discussion
The Chinese medicine Aconitum as toxic drug is commonly used to treat rheumatism and cardiovascular diseases. In addition to efficacy, adverse events of Aconitum poisoning often occur in clinic. Prevention Aconitum toxicity is a key issue for its safe application. Efflux transporters and drugmetabolizing enzymes protect the human body by functioning as the important defense line that limits the invasion of toxicants. In the present study, we have demonstrated that monoester-diterpene alkaloids (MDAs) including BAC, BMA, and BHA were metabolized into several metabolites in HLMs. And CYP3A4 and CYP3A5 were the predominant isozymes responsible for their metabolism by the experiments of chemical inhibitors, mAbs and cDNA-expressed CYP enzymes. A total of 7, 8, and 9 metabolites were found, respectively, when BAC, BMA, or BHA was incubated in the HLMs along with NADPH-regenerating system. A majority of metabolite reaction channels had at least two metabolites, but they did not characterize the exact position of transformation on account of several methyl and hydroxyl groups in the structure of MDAs. What is more, the structures of MDAs are so complex and contain so many similar moieties that it would be extremely difficult to unambiguously identify metabolites solely using mass spectrometry. NMR data would be ideal for such compounds, but it may be difficult to generate sufficient amounts of metabolites. For these reasons, the metabolite characterization of the present study is rather preliminary and speculative, but it may be helpful to understand the metabolic profile of MDAs.
BAC, BMA, and BHA are the main, active and toxic constituents of processed Aconitum. The unprocessed Aconitum is prohibited to sell in the market. Even though the toxicity of BAC, BMA, and BHA is much less than diester diterpene alkaloids [12], attention is still need. The toxic reaction will still occur by excessive absorption of BAC, BMA,    and BHA. As is well known, efflux transporters and metabolic enzymes act as the first line to protect the body by limiting the invasion of toxicants. Previous studies indicated that the efflux transporters including P-gp, BCRP, and MRP2 did not mediate the transport of BAC, BMA, and BHA, suggesting that the invasion of BAC, BMA, and BHA could not be limited by efflux transporters [11]. In this case, the role of metabolic enzymes in BAC, BMA, and BHA is extremely important.
We investigated the metabolism of BAC, BMA, and BHA for the first time. A total of 7, 8, 9 metabolites were found for BAC, BMA, and BHA in HLMs with NADPHregenerating system (Figure 1). BAC, BMA, and BHA have very similar chemical structures. A methyl moiety attaches to the nitrogen atom and a hydroxyl group links to position C3 in the structure of MA, whereas the unit attached to the nitrogen is ethyl in the BAC structure. The group linked to the nitrogen atom in the structure of BHA is methyl, but the hydroxyl moiety at position C3 is not observed. Therefore, their metabolic pathways were also similar, including demethylation, dehydrogenation, demethylationdehydrogenation, hydroxylation, and didemethylation (Figure 5). What is more, these metabolic pathways were similar to those of DDAs [19][20][21]. Previous studies indicated that HA could be metabolized into MA by the effect of CYP enzymes. As expected, BHA was also metabolized into BMA in the present study. It was reported that the high toxicity of Aconitum alkaloids was due to the acetyl group at C8, the hydroxyl group at C13, four methoxyl groups at C1, C6, C16, and C18, and the benzoyl ester group at C14 [9], meaning that the loss of these groups will decrease the compound's toxicity. According to our results, all the metabolites (except for the hydroxylation metabolite) were either loss of methyl from methoxyl group or dehydrogenation from hydroxyl group, indicating that their toxicology was less than that of the parent compound.
The results showed that CYP enzymes mediated the metabolism of BAC, BMA, and BHA, reducing the accumulation of toxic parent compound in the body. In the next experiments, we further investigated which CYP isozymes are responsible for their metabolism. The targeted isozymes included CYP1A2, 2C8, 2C9, 2C19, 2D6, 2E1, 3A4, and 3A5, because up to 90% of human drug metabolism may be attributed to these eight enzymes. In the chemical inhibitors experiments, the major hepatic enzyme responsible for the metabolism of BAC, BMA, and BHA was identified as CYP3A4/5, while chemicals selective for CYP2C8, CYP2C9, and CYP2C19 made a minor effect on their formations. It was assumed that a selective chemical inhibitor raised against a human CYP isoform at high concentration, which was more than the Ki value, will cross-react with other CYP isozymes. For example, recent studies suggest that omeprazole is a substrate probe for CYP2C19, however in vitro studies show that omeprazole could be metabolized by CYP3A4 in the same way [30]. As is well known, the chemical inhibitors are never 100% and their selectivity is not always great. Hence, we used mAbs experimental approach to reconfirm our results. We demonstrated that antibody against human CYP3A4/5 was the most potent and inhibited the formation of BAC, BMA, and BHA. Furthermore, the experiment of cDNAexpressed CYP enzymes showed that CYP3A4 and CYP3A5 were the major isozymes for the metabolism of BAC, BMA, and BHA, and the BAC metabolite M5 generated by CYP2D6, and the BMA metabolites M5 and M6 generated by CYP2C8, 2D6, and 3E1 supported the minimal roles of these enzymes. However, the inhibition experiments of BAC, BMA, and BHA did not support that CYP2C8, 2D6, and 3E1 were the subordinate enzymes which mediated the biotransformation of BAC, BMA, and BHA. Taken together, the results of cDNA-expressed CYP enzymes, chemical inhibitors, and mAbs experiments, we come to the conclusion that CYP3A4    and CYP3A5 were the major isozymes responsible for the metabolism of BAC, BMA, and BHA ( Figure 5).
CYP3A4 and CYP3A5 also contributed a great role to the metabolism of DDAs (AC, MA, and HA) [11,20,31]. DDAs could not be only chemically hydrolyzed into MDAs, but also biologically hydrolyzed [2,15]. In other words, after the intake of processed Aconitum, three kinds of behavior will occur: DDAs transformed into less toxic metabolites by CYP3A4 and CYP3A5; MDAs transformed into less toxic metabolites by CYP3A4 and CYP3A5; and DDAs transformed into MDAs by the esterase, which will greatly reduce the toxicity of Aconitum. Therefore, protection or promotion of CYP3A function could prevent excessive Aconitum absorbed into the bloodstream, thereby avoiding toxic reactions.
CYP3A is one of the most important CYP isoform responsible for drug metabolism by humans, because it is highly expressed in critical tissues such as the gastrointestinal tract and liver, and it is involved in the oxidative biotransformation of numerous clinically useful therapeutic agents. However, many factors could regulate CYP3A expression and activity [32]. Accordingly, the potential for drug interactions with these drugs as well as other CYP3A substrates, when given concomitantly, is high. Metabolism involving CYP3A is also likely to be affected by liver disease as well as aging. Therefore, more attention should be paid when we ingest Aconitum, even though the CYP enzymes could act as a defense line to decrease its toxicity.