Biological and Pharmaceutical Bulletin
Online ISSN : 1347-5215
Print ISSN : 0918-6158
ISSN-L : 0918-6158
Regular Article
Evaluation of the Effect of Aldehyde Oxidase Inhibitors on 6-Mercaptopurine Metabolism
Hinata UedaKatsuya Narumi Ayako FurugenKeisuke OkamotoYoshitaka SaitoMasaki Kobayashi
Author information
JOURNAL OPEN ACCESS FULL-TEXT HTML
Supplementary material

2025 Volume 48 Issue 5 Pages 713-720

Details
Abstract

Thiopurines, such as 6-mercaptopurine (6-MP) and azathioprine, are converted to the inactive metabolites 6-thioxanthin (6-TX) and 6-thiouric acid (6-TUA). Molybdenum-containing oxidoreductases, aldehyde oxidase (AOX) and xanthine oxidase (XO), are involved in the oxidation of 6-MP to 6-TX; XO inhibitors affect the therapeutic efficacy of thiopurines and the incidence of adverse effects, such as liver and blood disorders. However, the role of AOX in the pharmacokinetics of 6-MP remains unclear. To clarify the clinical importance of AOX-mediated drug–drug interactions, we evaluated whether drugs that inhibit AOX affect 6-MP metabolism. The metabolism of 6-MP to 6-TX was strongly inhibited by AOX inhibitors (amitriptyline, chlorpromazine, clomipramine, clozapine, hydralazine, quetiapine, and raloxifene) in a reaction mixture containing human liver cytosol. The inhibition of 6-TX production rate by each AOX inhibitor was 60–70% at high concentrations, although the XO inhibitor febuxostat showed an inhibition rate of 10–30%. Furthermore, the combination of febuxostat and each AOX inhibitor showed greater inhibition than when each compound was added alone. The AOX inhibitor did not alter 6-MP oxidation by recombinant XO. These results suggest that AOX inhibition may affect the pharmacokinetics of thiopurines. However, because of the lower activity of AOX in rats than that in humans, the contribution of AOX could not be assessed using in vivo experiments. Further studies are needed to evaluate the contribution of AOX to the therapeutic and adverse effects of thiopurines, both in clinical studies and in animal models of liver humanization.

INTRODUCTION

6-Mercaptopurine (6-MP) and its prodrug azathioprine have antitumor and immunosuppressive activities and are used to treat acute leukemia, inflammatory bowel disease, and rheumatoid arthritis.1) The pharmacokinetic properties of azathioprine and 6-MP are assumed to be similar because azathioprine is rapidly converted to 6-MP by glutathione S-transferase.2) Dose-independent side effects of azathioprine include pancreatitis and gastrointestinal disturbances, such as nausea and vomiting.3) These adverse events are typically allergic reactions caused by the imidazole ring in azathioprine, and treatment can be continued by switching to 6-MP. However, hematologic and hepatic disorders are dose-dependent side effects of thiopurines, and these side effects are the major reasons for treatment discontinuation.

6-MP is known to be converted by different enzymes into 1 of 3 metabolic pathways: (1) phosphorylation via hypoxanthine guanine phosphoribosyl transferase to 6-thioguanine nucleotides (6-TGNs), (2) methylation via thiopurine S-methyltransferase to 6-methyl mercaptopurine (6-MMP), and (3) oxidation to 6-thioxanthine (6-TX) and 6-thiouric acid (6-TUA).46) 6-TGNs are active metabolites that cause excessive immunosuppression and leukopenia when their concentrations in the blood cells exceed the therapeutic range.4) Blood levels of 6-MMP, an inactive metabolite, correlate with the incidence of hepatic disorders.5,6) Genetic polymorphisms in thiopurine S-methyltransferases are associated with therapeutic efficacy and adverse drug reactions.7) Additionally, genetic polymorphisms of nudix hydrolase 15 (NUDT15) (R139C [rs116855232] and R139H [rs147390019]), which are responsible for the dephosphorylation of thioguanine triphosphate, the active form of 6-TGNs, have been reported to be responsible for the serious side effects of thiopurine preparations.8) In particular, cysteine homozygotes (Cys/Cys) and histidine heterozygotes (Cys/His) showed notable leukopenia and alopecia associated with increased thioguanine triphosphate levels. Although NUDT15 gene polymorphism is a characteristic mutation in East Asians, several patients without this polymorphism also exhibit myelosuppression. In addition, this polymorphism alone cannot predict other side effects, such as hepatic injury. Therefore, a comprehensive elucidation of pharmacokinetic mechanisms, including other metabolic enzymes, is needed.

6-TUA and 6-TX, formed by the oxidation of 6-MP, are inactive metabolites in terms of drug efficacy and toxicity. Previously, only xanthine oxidase (XO) was thought to be responsible for 6-TUA production. The XO inhibitor allopurinol significantly increases the bioavailability of thiopurines, increasing the pharmacological activity of 6-MP and the incidence of adverse effects, such as myelosuppression.9) For the same reason, the XO inhibitors febuxostat and topiroxostat are contraindicated in combination with thiopurines. Previous in vitro studies by Choughule et al. have shown that aldehyde oxidase (AOX), which has a sequence similar to XO, is also involved in 6-MP oxidation.10) Furthermore, AOX may also be associated with clinical outcomes following thiopurine treatment, as some case reports have suggested an association between a genetic polymorphism in AOX (N1135S [rs55754655]) and decreased therapeutic efficacy of azathioprine.11,12) Using a real-world adverse event database, we previously reported that several drugs that inhibit AOX increased reports of hematological disorders, with azathioprine as the suspected drug.13) Although AOX inhibitors can alter the pharmacokinetics of 6-MP, such as XO inhibitors, no studies have evaluated 6-MP metabolic changes with the concomitant use of AOX inhibitors. In this study, we examined the effects of AOX inhibitors on the oxidation of 6-MP to 6-TX and 6-TUA in the human and rat liver cytosol (RLC).

MATERIALS AND METHODS

Chemicals

6-MP and amitriptyline were purchased from Sigma-Aldrich Chemical Corp. (St. Louis, MO, U.S.A.); 2-amino-6-mercaptopurine-13C2, 15N, and 6-TX from Toronto Research Chemicals (Toronto, ON, Canada); 6-TUA was obtained from BIOSYNTH (Compton, U.K.); clozapine, febuxostat, hydralazine, quetiapine, and raloxifene were gained from Tokyo Chemical Industry (Tokyo, Japan); and chlorpromazine and clomipramine from FUJIFILM Wako Pure Chemical Corporation (Osaka, Japan). All other chemicals and reagents were of commercial origin and of the highest grade available.

Incubation in Human Liver Cytosol

XTreme200 pools of human liver cytosol (HLC; Xenotech, Kansas City, MO, U.S.A.) were used for the 6-MP to 6-TX oxidation assay. HLC was diluted in 25 mM phosphate buffer (pH 7.4) to a final concentration of 0.5 mg/mL. The reactions were initiated by adding 6-MP to the HLC mixture after preincubation for 3 min. The concentration of dimethyl sulfoxide used as the compound dilution solution in the reaction mixture was less than 0.1%. The volume of the total reaction mixture was 50 μL. The reaction was terminated by adding an equal volume of ice-cold acetonitrile containing 200 ng/mL 2-amino-6-mercaptopurine-13C2 and 15N as internal standards (IS). The concentration of 6-MP in solution was 50–400 μM. The reaction time ranged from 5 to 90 min. The inhibitory effect of febuxostat and AOX inhibitors was evaluated at 6-MP concentrations of 100 μM and reaction times of 15 min. These inhibitors were evaluated at concentrations of 1 nM to 100 μM. The reaction end mixture was centrifuged at 16000 × g for 15 min and 10 μL of the supernatant was used for determination of the amount of 6-TX. Ultra-performance liquid chromatography-MS/MS (UPLC-MS/MS) was performed with reference to the previous report.10) A Waters ACQUITY UPLC H-Class System and Xevo TQ-S mass spectrometer (Waters Corporation, Milford, MA, U.S.A.) were used to analyze the manufactured products. Samples were separated on a Waters ACQUITY UPLC BEH C18 column (1.7 mm, 2.1 × 100 mm) at a column temperature of 40°C and a flow rate of 0.4 mL/min. Mobile phase A (0.3% formic acid, 0.02 M ammonium acetate solution) and B (acetonitrile) were pumped per the following gradient methods: 90% A for 0.1 min, 90–10% A from 0.1 to 1.0 min, 10–90% A from 1.0 to 1.5 min, and 90% A until 2 min. Negative-ion electrospray MS/MS analysis was used to quantify 6-TX and the IS. The m/z, declustering potential (DP), and collision energy (CE) were as follows: m/z, 167.23 > 133.00 for 6-TX (DP: 44; CE: 16) and 169.27 > 135.05 for 2-amino-6-mercaptopurine-13C2, 15N (DP: 46; CE: 14).

Incubation in Recombinant XO

Recombinant XO (Nacalai Tesque, Kyoto, Japan) was diluted to a final concentration of 0.015 U/mL. 6-MP (final 1–100 μM) and inhibitor (final 1 nM–100 μM) were added to the reaction mixture to a final volume of 100 μL. Ten microliters of 10% HCl was added to stop the reaction and the mixture was centrifuged at 16000 × g for 15 min. Thirty microliters of the supernatant was injected into a Waters ACQUITY UPLC H-Class System. 6-TUA was quantified using a previously reported method.14) 6-TUA was separated on a Waters ACQUITY UPLC BEH C18 column (1.7, 2.1 × 50 mm) at 30°C. Potassium phosphate buffer (20 mM) was pumped isocratically at a flow rate of 0.6 mL/min. UV detection was performed at 342 nm to detect 6-TUA (retention time: 0.6 min).

Incubation in RLC

We used 2 kinds of RLC: (1) pooled Wistar male RLC (pooled RLC) obtained from Xenotech and (2) RLC extracted from the liver of 7-week-old male Jcl:Wistar rats (extracted RLC) that were the same types used for the pharmacokinetic study. Extraction of fresh rat liver was performed with slight modifications to previous reports.15) The rat livers were homogenized in 50 mM tris hydrochloride buffer containing 150 mM potassium chloride (pH 7.4). The liver suspensions were disrupted by sonication. Tissue fractions were centrifuged at 9000 × g for 20 min and then at 105000 × g for 60 min at 4°C. The supernatants of the samples were used as extracted RLC for the 6-MP oxidation assay. The protein concentration of the extracted RLC was determined using a Pierce BCA Protein Assay Kit (Thermo Scientific, Rockford, IL, U.S.A.) following the manufacturer’s instructions. Pooled and extracted RLC were diluted with 25 mM phosphate buffer to a final concentration of 0.2 mg/mL. The reaction and quantification methods were the same as those for the 6-MP oxidation assay described above, using HLC and recombinant XO. All animal experiments were approved by the Laboratory Animal Committee of Hokkaido University (Approval No. 22-0054).

Statistical Analysis

Data were expressed as mean with standard error of the mean (S.E.M.) for in vitro experiments. Statistical analysis was performed using unpaired Student’s t-test, with a p-value of <0.05 considered statistically significant. One-way ANOVA, followed by Dunnett’s test or Tukey’s post hoc test, was used for multigroup comparisons. All statistical analyses were conducted using R version 4.4.0 with the “multcomp” package. Data evaluating the inhibitory effects of febuxostat and AOX inhibitors on 6-MP oxidation were fitted to a 4-parameter logistic equation by nonlinear regression analysis using SigmaPlot 14.5 (Systat Software Inc., San Jose, CA, U.S.A.).

RESULTS

Inhibitory Effects of Febuxostat and AOX Inhibitors on 6-MP Oxidation in HLC

6-MP is converted to the inactive metabolite 6-TUA via the reaction involving the intermediate 6-TX by oxidation of 2 parts of the heterocyclic ring. The amount of 6-TX produced from 6-MP in the HLC increased linearly for up to 90 min (Supplementary Fig. S1). The reaction rate also increased with the 6-MP concentration. Previous studies reported that the final product 6-TUA in HLC is formed after 25 min, when 6-TX reaches a steady state.10) The 1st step of 6-MP catabolism, 6-TX formation, is the rate-limiting step and is considered an important reaction pathway for drug–drug interactions. Therefore, we evaluated the inhibitory effects of febuxostat and AOX inhibitors on 6-MP oxidation in the HLC for 15 min. As in the previous study, we defined AOX inhibitors and decided to examine their inhibitory effects on 6-MP metabolism in this study also.13) Seven drugs (amitriptyline, chlorpromazine, clomipramine, clozapine, hydralazine, quetiapine, and raloxifene) were selected based on Cmax/IC50 calculated from the highest blood concentration of each drug at standard doses and IC50 values in in vitro experiments,16,17) with a cutoff value of 0.1 (Supplementary Table S1). The AOX inhibitors inhibited 6-TX production in a concentration-dependent manner (Fig. 1). Febuxostat, an XO inhibitor, also inhibited 6-MP oxidation; however, its effect was smaller than that of AOX inhibitors. Furthermore, simultaneous addition of febuxostat and various AOX inhibitors strongly inhibited 6-TX production (Fig. 2).

Fig. 1. Effect of AOX Inhibitors and Febuxostat on the Oxidation of 6-MP to 6-TX by HLC

Inhibitor concentrations were set at 1 nM–100 μM; metabolism of 100 μM 6-MP was measured at pH 7.4 for 15 min. Data represent the mean ± standard error of the mean (S.E.M.) of 3 independent experiments. 6-MP, 6-mercaptopurine; 6-TX, 6-thioxanthine; AOX, aldehyde oxidase; HLC, human liver cytosol.

Fig. 2. Inhibitory Effect of AOX Inhibitors in Combination with Febuxostat

Inhibitor concentrations were set at 5 μM except for raloxifene (100 nM); metabolism of 100 μM 6-MP was measured using HLC at pH 7.4 for 15 min. ***p < 0.001 compared with febuxostat; Dunnett’s test. Data represent the mean plus or minus S.E.M. of 3 independent experiments.

Effects of AOX Inhibitors on the Metabolic Function of Recombinant XO

To evaluate the effects of AOX inhibitors on XO oxidation, an oxidation assay of 6-MP to 6-TUA was performed using recombinant XO. In recombinant XO, the formation of 6-TUA increased linearly up to 40 min after the start of the reaction, reaching a maximum reaction at the concentration of 20 μM (Supplementary Fig. S2). Therefore, in subsequent studies with various inhibitors, the reaction time was set to 30 min and the substrate concentration to 5 μM. Febuxostat completely inhibited 6-MP metabolism at a concentration of 0.01 μM. Raloxifene inhibited XO when added at high concentrations; however, the other AOX inhibitors showed no inhibitory effects (Fig. 3). These results indicate that each compound selectively inhibited AOX, but not XO.

Fig. 3. Effect of AOX Inhibitors and Febuxostat on the Oxidation of 6-MP to 6-TUA by Recombinant XO

Inhibitor concentrations were set at 1 nM–100 μM; metabolism of 5 μM 6-MP was measured at pH 7.4 for 30 min. Data represent the mean ± S.E.M. of 3 independent experiments.

6-MP Metabolism in Rats

To determine whether AOX inhibition affects thiopurine pharmacokinetics, we evaluated the effect of concomitant use of hydralazine, a potent AOX-selective inhibitor, on rat plasma 6-MP concentration–time profiles. The reaction kinetics are presented in Supplementary Fig. S3, where the 6-MP and 6-TUA concentrations were higher in the hydralazine group for up to 240 min after administration. Therefore, we used RLCs extracted from Wistar rats and purchased pooled RLC to measure the effects of AOX on 6-MP oxidation. In the pooled RLC, unlike in the HLC, the amount of 6-TX rapidly reached a steady state and 6-TUA was produced shortly after the start of the reaction (Supplementary Fig. S4A). The function of AOX in the extracted RLC was confirmed by evaluating the oxidation of the AOX substrate phthalazine to phthalazone (Supplementary Fig. S4B). In both the pooled and extracted RLC, febuxostat, but not hydralazine, inhibited the oxidation of 6-MP to 6-TX (Fig. 4). A high concentration of hydralazine also did not inhibit the metabolism of 6-MP to 6-TUA (Supplementary Fig. S5). These results suggest that the contribution of AOX in RLC is much smaller than that in HLC; therefore, the contribution of XO to 6-MP metabolism is relatively high.

Fig. 4. Inhibitory Effect of Febuxostat and Hydralazine on 6-TX Production Using (A) Extracted RLC and (B) Pooled RLC

Metabolism of 50 μM 6-MP was measured at pH 7.4 for 3 min. Oxidative activity was evaluated at each concentration of 5 μM with or without febuxostat and/or hydralazine. ***p < 0.001 compared with control; Tukey’s post hoc test. Data represent the mean ± S.E.M. of 3 independent experiments.

DISCUSSION

In this study, to evaluate the relevance of AOX in the pharmacokinetics of thiopurine drugs, we performed 6-MP metabolism experiments using liver-derived fractions and animal models. 6-TX production after the addition of 6-MP was decreased by AOX inhibitors. The magnitude and mode of AOX inhibition depend on the substrate compound used.18) The inhibition strength of 6-MP oxidation evaluated in this study correlates with the previously reported effect of inhibition on phthalazine.16) In the reaction mixture containing HLC, the 6-TX production rate with high concentrations of each AOX inhibitor was 30–40% of that without inhibitors, suggesting that AOX contributed to approximately 60–70% of the 6-TX formation pathway under the conditions of this reaction. Febuxostat, used as an XO inhibitor, showed 10–30% inhibition at increasing concentrations.

The package inserts for thiopurine preparations clearly indicate that the XO inhibitors febuxostat, topiroxostat, and allopurinol are contraindicated or should be used with caution. The results of this study suggest that the contribution of AOX to the 6-MP kinetics is greater than that of XO. However, few clinical reports exist on thiopurine-induced adverse events in combination with AOX inhibitors. Differences may be present between in vitro studies and actual metabolic reactions in the human liver.

Previous reports have shown that the rate of 6-TX formation is biphasic (Vmax1: 0.08 nmol/min/mg; Km1: 88.2 μM; Vmax2: 0.2 nmol/min/mg, Km2: 500 μM), indicating that 2 enzymes with different affinities for the substrate function in the oxidation reaction of 6-MP.10) The Km value for 6-MP metabolism by purified human AOX approximates Km2 above. Therefore, AOX has a lower affinity and faster reaction rate for 6-MP oxidation than that of XO. The 6-TX produced was quantified by UPLC-MS/MS and the concentration of 6-MP was set at 100 μM to account for the lower limit of quantitation. Maximum plasma concentrations of 6-MP were reported to be approximately 7.2–13.2 μM at 2 h after administration of 100 mg 35S-azathioprine.19) Owing to the lack of knowledge regarding tissue translocation of 6-MP, the concentration of 6-MP in the liver is unknown; however, the contributions of XO and AOX vary with the dose of 6-MP. Thus, AOX activity under these experimental conditions may be overestimated compared to the activity at clinical 6-MP liver concentrations.

The complete inhibition of 6-TX production by febuxostat in combination with raloxifene or hydralazine also suggests a significant contribution of AOX to the conversion of 6-MP to 6-TX. Allopurinol may be combined when the immunosuppressive effect of thiopurines is inadequate and hepatotoxicity occurs with increasing dosage.20) XO is inhibited not only by drugs but also by phenolic acids and polyphenols in food.21,22) Furthermore, several genetic polymorphisms are responsible for reduced XO activity.23) Under the XO inhibitor combination treatment, AOX may be responsible for 6-MP metabolism in a compensatory manner. Therefore, further studies focusing on patients with inhibited or suppressed XO function are required.

Hall and Krenitsky found that AOX from rabbit liver recognizes 6-MP as a substrate.24) Rashidi et al. reported that menadione and chlorpromazine, AOX inhibitors, inhibited 6-MP oxidation in guinea pig liver cytosol.25) In the livers of guinea pigs, AOX has been speculated to be involved in the oxidation of 6-TX to 6-TUA. Because the oxidation of 6-MP to 6-TX by XO is the rate-limiting step in guinea pigs, the contribution of AOX to thiopurine metabolism was considered low. In this study, we attempted to elucidate the function of AOX in 6-MP blood dynamics using a rat model. However, the results showed that the 6-MP metabolic profile in humans could not be reproduced because of the different behaviors of 6-MP metabolism in HLC and RLC. The AOX isoforms and their expression levels differ significantly between humans and rodents.26,27) Furthermore, the metabolic rates of drugs that are AOX substrates, such as favipiravir and BIBX1382,28,29) are lower in mice and rats. Thus, the contribution of AOX may be lower in rodent models. Auglurant (VU0424238), an mGlu5 receptor antagonist, is oxidized by both AOX and XO.30) The nitrogen-containing heterocyclic ring in the auglurant structure is oxidized via M1 (6-oxopyrimidine) to M2 (2,6-dioxopyrimidine) in 2 steps, similar to 6-MP. During the metabolism of M1 to M2, XO functions in rodents, and AOX in guinea pigs, rhesus, and cynomolgus. These results confirm that the activities and contribution rates of AOX and XO differ notably among animal species. Therefore, chimeric mice, in which the liver is replaced by the human liver, must be used to predict the pharmacokinetics of AOX substrates.29)

The concomitant use of hydralazine increased plasma 6-MP and 6-TUA levels in rats; additionally, AOX did not contribute to 6-MP metabolism in RLC. As hydralazine does not inhibit XO metabolism, this change in the blood profile may be due to an interaction via metabolic enzymes other than the 6-MP oxidative pathway. In addition, transporters such as multidrug resistance-associated protein 4, organic anion transporting polypeptide 1B1, and organic anion transporter 3 have been reported to be associated with the adverse effects and pharmacokinetics of 6-MP.3133) However, knowledge regarding the direct function of these transporters in the accumulation and excretion of 6-MP is limited. These transporters may be responsible for the interactions observed between 6-MP and hydralazine in this study. More detailed studies using these expression systems and inhibitors are required to elucidate the disposition and interactions of 6-MP in the body.

In the present study, 6-MP is assumed to be oxidized to 6-TUA via 6-TX. The primary intermediate of xanthine, a compound structurally similar to 6-MP, is hypoxanthine, which is produced via oxidation at the C-2 position.34) In contrast, AOX oxidizes the C-8 position of thiopurine compounds to produce 8-oxo-6-MP, whereas XO selectively oxidizes the C-2 position.25,35) Although the amount of 8-oxo-6-MP produced could not be examined in this study due to the lack of a means to synthesize 8-oxo-6-MP, it is possible that the combined use of AOX inhibitors may inhibit the oxidation of the C-8 position, thereby increasing the amount of 6-MP in HLC and RLC.

We confirmed that the conversion of 6-MP to 6-TX was inhibited by various AOX inhibitors. An inhibitory effect was also observed in the presence of febuxostat. However, the evaluation of drug–drug interactions in rats was complicated because of species differences. Clinical studies are important to evaluate the importance of AOX in clinical practice. In addition, when metabolic enzymes other than AOX are involved in the metabolic pathway, the AOX function must be evaluated by considering the reduction in activity due to the genetic polymorphisms of each metabolic enzyme and the concomitant use of inhibitors.

Acknowledgments

We thank the Open Facility, Global Facility Center, Creative Research Institution, Hokkaido University, for allowing us to conduct the analysis using Xevo TQ-S and for providing insight and expertise that greatly assisted the research. This study was partially supported by Grants-in-Aid from the Japan Society for the Promotion of Science Research (Grant Number: 23KJ003903 to HU) and KAKENHI (Grant Number 23K0622903 to KN).

Conflict of Interest

The authors declare no conflict of interest.

Supplementary Materials

This article contains supplementary materials.

REFERENCES
 
© 2025 Author(s).
Published by The Pharmaceutical Society of Japan

This article is licensed under a Creative Commons [Attribution-NonCommercial 4.0 International] license.
https://creativecommons.org/licenses/by-nc/4.0/
feedback
Top