Glucuronidation of the Broadspectrum Antiviral Drug Arbidol by UGT Isoforms

bs_bs_banner

Research Paper Journal of Pharmacy And Pharmacology

Glucuronidation of the broad-spectrum antiviral drug arbidol by UGT isoforms Jin-Hui Songa,*, Zhong-Ze Fangb,e,*, Liang-Liang Zhud, Yun-Feng Caoc, Cui-Min Hue, Guang-Bo Ged and De-Wei Zhaoa aOrthopedics Department, Afﬁliated Zhongshan Hospital of Dalian University, Dalian, bLiaoning Medical University, Jinzhou, cKey Laboratory of Contraceptives and Devices Research (NPFPC), Shanghai Engineer and Technology Research Center of Reproductive Health Drug and Devices, Shanghai Institute of Planned Parenthood Research, Shanghai, dLaboratory of Pharmaceutical Resource Discovery, Dalian Institute of Chemical Physics, Chinese Academy of Sciences, Dalian, China and eLaboratory of Metabolism, Center for Cancer Research, National Cancer Institute, Bethesda, MD, USA

Keywords Abstract arbidol; glucuronidation; UDP-glucuronosyltransferases (UGTs) Objectives The aim of this work was to identify the uridine glucuronosyltransferase (UGT) isoforms involved in the metabolism of the broad-spectrum antivi- Correspondence ral drug arbidol. De-Wei Zhao, Orthopedics Department, Methods A human liver microsome (HLM) incubation system was employed to Affiliated Zhongshan Hospital of Dalian catalyse the formation of arbidol glucuronide. The glucuronidation activity of University, Dalian 116 001, China. E-mail: [email protected] commercially recombinant UGT isoforms towards arbidol was screened. A combination of kinetic analysis and chemical inhibition study was used to determine the Received June 21, 2012 UGT isoforms involved in arbidol’s glucuronidation. Accepted November 14, 2012 Key findings The arbidol glucuronide was detected when arbidol was incubated with HLMs in the presence of UDP-glucuronic acid. The Eadie–Hofstee plot doi: 10.1111/jphp.12014 showed that glucuronidation of arbidol was best fit to the Michaelis–Menten Ϯ *These two authors equally contributed to kinetic model, and Km and apparent Vmax were calculated to be 8.0 0.7 mm and Ϯ this work. 2.03 0.05 nmol/min/mg protein, respectively. Assessment of a panel of recombinant UGT isoforms revealed that UGT1A1, UGT1A3 and UGT1A9 could catalyse the glucuronidation of arbidol. Kinetic analysis and chemical inhibition study demonstrated that UGT1A9 was the predominant UGT isoform involved in arbidol glucuronidation in HLMs. Conclusions The major contribution of UGT1A9 towards arbidol glucuronidation was demonstrated in this study.

Introduction Arbidol, ethyl-6-bromo-4-[(dimethylamino)-methyl]-5- dose of arbidol, the time to reach maximal plasma concen- hydroxy-1-methyl-2-[(phenylthio) methyl]-indole-3- trations (Tmax) was determined to be 1.2 h. The half-life 1 carboxylate (Figure 1), is an antiviral drug developed by the (t /2) of the drug in the body is 17–21 h and about 40% of Chemical Drug Center of All Russian Research Institute of drug is excreted unchanged.[7] It is metabolized in the liver Pharmaceutical Chemistry together with the Institute of and arbidol glucuronide is a major metabolite in human Medical Radiology.[1] It has been demonstrated to exhibit urine.[8] inhibitory activity towards various types of viruses, such as The human uridine glucuronosyltransferases (UGTs) are influenza virus, respiratory syncytial virus, adenovirus, rhi- membrane proteins of the endoplasmic reticulum and can novirus, parainfluenza virus, hepatitis B virus, hepatitis C conjugate various endogenous substances and exogenous virus and coxsackie virus.[2–6] Therefore, arbidol is regarded compounds.[9] Glucuronidation reactions catalysed by as a broad-spectrum antiviral compound. Recently, arbidol UGTs account for >35% of all phase II drug metabolism.[10] has been approved for the treatment of influenza in Russia, UGT-mediated metabolism is normally regarded as a China and some European countries. detoxifying process, and glucuronides have greater polarity The pharmacokinetic behaviour of arbidol has been than the parent drugs. Nevertheless, some glucuronides reported in a previous publication.[7] After a single 50-mg have been reported to show pharmacological and

CH3 donors signed informed consents approved by the local ethical committee. Information on the medication history N of the sample donors was not gained. A panel of human O CH3 H3C liver microsomes (HLMs) was prepared from 12 liver O samples obtained from male and female patients by differ- ential ultracentrifugation as described previously.[13,14] HO Microsomal protein concentrations were determined by the Lowry method with bovine serum albumin as standard.[15] A panel of recombinant human UGT supersomes (UGT1A1, UGT1A3, UGT1A4, UGT1A6, UGT1A9, N S Br UGT2B4, UGT2B7, UGT2B15 and UGT2B17) expressed in baculovirus-infected insect cells were purchased from BD CH3 Gentest Corp. (Woburn, MA, USA). Figure 1 The structure of arbidol. Arbidol glucuronidation assay

[11] toxicological effects. The human UGT superfamily is A typical incubation mixture (200 ml total volume) was classified into two families (UGT1 and UGT2) and three composed of 50 mm Tris-HCl buffer (pH = 7.4), 5 mm subfamilies (UGT1A, UGT2A and UGT2B) based on their MgCl2,5mm UDPGA, 25 mg/ml alamethicin, 10 mm [12] sequence identity. The following UGT enzymes are d-saccharic acid 1,4-lactone, 100 mm arbidol and 0.5 mg/ml expressed in the human liver: UGT1A1, UGT1A3, UGT1A4, HLMs. Arbidol was dissolved in methanol and the final UGT1A6, UGT1A9, UGT2B4, UGT2B7, UGT2B10, concentration of methanol in the incubation system was UGT2B11, UGT2B15, UGT2B17 and UGT2B28. 1% (v/v). After pre-incubation at 37°C for 5 min, the reac- Identification of drug-metabolizing enzymes (DMEs) is tion was initiated by addition of UDPGA. The reaction very important to understand the variation in drug effect mixture was incubated at 37°C for 1 h and terminated by and metabolism. Moreover, identification of DMEs could addition of 100 ml ice-cold acetonitrile. The mixture was provide pivotal information for potential drug–drug inter- centrifuged at 20 000g for 10 min and a portion of the actions, and the information provided can be used to guide supernatant fraction was transferred to a 0.3-ml auto- clinical practice and reduce side effects in clinical treatment. injector vial for HPLC analysis. Control incubations Therefore, the objective of this experiment was to investi- (without UDPGA, without substrate or without micro- gate the glucuronidation of arbidol using an in-vitro somes) were performed to ensure that the formation of human liver microsomal system. Human UGT isoforms metabolites was microsome- and UDPGA-dependent. The responsible for the formation of metabolites were identified HPLC system (Shimadzu, Kyoto, Japan) contained an SCL- by screening with recombinant expressed UGT isoforms 10A system controller, two LC-10AT pumps, a SIL-10A and chemical inhibition study. To our knowledge, this is the auto-injector, an SPD-10AVP UV detector and a C18 column first study to investigate the UGT isoforms involved in the (4.6 ¥ 150 mm, 5 mm, Kromasil, Sigma-Aldrich) was used glucuronidation of arbidol. to separate arbidol and its metabolites. The mobile phase

consisted of CH3CN (A) and H2O containing 0.5% (v/v) Materials and Methods formic acid (B). The following gradient condition was used: 0–15 min, 98–10% B; 15–20 min, 10% B; 20–25 min, 5% B; Materials 25–35 min, 98% B. The ﬂow rate of the mobile phase was Arbidol hydrochloride (purity > 98%) was obtained from set at 1 ml/min. The injection volume was 30 ml and the the National Institute for the Control of Pharmaceutical scan wavelength was set at 316 nm. Due to the absence of and Biological Products (Beijing, People’s Republic of authentic standards for arbidol glucuronide, quantiﬁcation

China). Alamethicin, magnesium chloride (MgCl2), of the glucuronide in the incubation mixtures was per- d-saccharic acid 1,4-lactone, UDP-glucuronic acid triso- formed using a standard curve of arbidol. The calculation dium salt (UDPGA), androsterone, mefenamic acid and curve was generated by peak area of arbidol over the con- erlotinib were purchased from Sigma-Aldrich (St Louis, centration range 0.1–100 mm. The curve was linear over this MO, USA). All other reagents were of HPLC grade or of the concentration range (r2 > 0.99). The limits of detection and highest grade commercially available. quantiﬁcation were determined at signal-to-noise ratios of Human liver samples were obtained from Dalian Medical 3 and 10, respectively. The accuracy and precision of the University (Dalian, Liaoning province, China) with the back-calculated values for each concentration were less approval of the local ethics committee at the university. All than 5%.

Identiﬁcation of arbidol glucuronide by for 20 min. For recombinant human UGT1A1, UGT1A3 mass/mass analysis and UGT1A9, 0.025 mg/ml of protein and various concentrations of arbidol were used to determine the kinetic The arbidol glucuronide was separated and collected for parameters. Metabolite formation was linear under assay mass/mass (MS/MS) analysis. MS/MS analysis was carried conditions. The kinetic constants were estimated by nonlin- out on an ABI 4000 Q-Trap hybrid triple quadrupole linear ear regression analysis using the Michaelis–Menten equa- ion-trap mass spectrometer (Applied Biosystems, Foster tion (Equation 1): City, CA, USA). MS conditions used were as follows: ion- spray voltage, 5000; temperature, 550; collision energy, 35; vV=∗(max [ S ]) ( K m + [ S ]) (1) collision energy spread, 15. A collision-induced dissociation

(CID) MS/MS spectrum was obtained to further elucidate where v is the rate of reaction, Vmax is the maximum veloc- the structure of metabolite. Data were processed using ity, Km is the Michaelis constant (substrate concentration at

Analyst 4.1 software (Applied Biosystems). 0.5 Vmax) and [S] is the substrate concentration.

Assays with recombinant UGT isoforms Statistical methods The formation of arbidol glucuronide was measured The data are presented as mean Ϯ standard deviation (SD). in reaction mixtures containing recombinant human The statistical analysis was performed using two-tailed UGT1A1, 1A3, 1A4, 1A6, 1A9, 2B4, 2B7, 2B15 and 2B17. Student’s t-test; P < 0.05 denoted signiﬁcance in all cases. The incubation conditions were performed as described for the HLM study. The arbidol concentration and protein concentration of the UGT isoforms was 8 mm and 0.1 mg/ml, Results respectively. Reaction time was 30 min for all isoforms. Formation of arbidol glucuronide Chemical inhibition study The formation of arbidol glucuronide by HLMs was detected by HPLC-UV and analysed by MS/MS. When Screening of the inhibition by androsterone towards various arbidol was incubated with HLMs in the presence of UGT isoforms has been carried out, and androsterone UDPGA, a new peak was eluted at 12.2 min (M) (Figure 2). showed relatively strong inhibition towards UGT1A9 and This peak was not detected in the control samples in the UGT2B7.[16] Similarly, mefenamic acid has been demon- absence of UDPGA, arbidol or microsomes. The metabolite strated to exert strong inhibitory effect towards UGT1A9 produces doublet m/z 653/655 ions in the positive-ion and UGT2B7.[17] Due to no catalytic activity of UGT2B7 mode because of the presence of Br in the molecule (Br has towards arbidol glucuronidation, androsterone and the distinct 79Br and 81Br isotopic pattern). As shown in mefenamic acid were considered to be speciﬁc inhibitors of Figure 3, the MS2 spectrum of the metabolite (m/z 655) UGT1A9 in this study. Androsterone and mefenamic acid showed major fragment ions at m/z 479, 434 and 390. As were tested for their inhibitory effects on arbidol glucuroni- reported in elsewhere,[8] the fragment ions corresponded to dation in HLMs, UGT1A1, UGT1A3 and UGT1A9. Arbidol the loss of glucuronic acid (-176 Da), glucuronic acid plus (8 mm) was incubated with HLMs (0.05 mg/ml), rUGT1A1 the dimethylamine (-176 Da to -45 Da) and glucuronic (0.025 mg/ml), rUGT1A3 (0.025 mg/ml) and rUGT1A9 acid plus the dimethylamine plus one neutral acetaldehyde (0.025 mg/ml) in the presence and absence of inhibitors. Various concentrations of androsterone (1–250 mm) and mefenamic acid (1–50 mm) were used to determine the half-inhibition concentrations (IC50) as previously 100 000 [18,19] described. All incubations were performed for 20 min. 75 000 Arbidol A selective inhibitor of UGT1A1, erlotinib (100 mm), was 50 000 also used to investigate its inhibition towards arbidol glu- M 25 000 curonidation. The incubation conditions, including sub- +UDPGA –UDPGA strate concentration, protein concentration and incubation 0

time, were similar to those used for androsterone and Absorbance at 316 nm 0.0 5.0 10.0 15.0 20.0 25.0 mefenamic acid inhibition experiments. Time (min)

Kinetic study Figure 2 Representative HPLC proﬁles of arbidol and its metabolite. Arbidol (100 mM) was incubated with human liver chromosomes To evaluate the kinetic parameters, arbidol (1–80 mm) was (0.5 mg/ml) at 37°C for 60 min with or without UDP-glucuronic acid incubated with HLMs (0.05 mg/ml protein concentration) trisodium salt as described in Materials and Methods.

655.1 Table 1 Inhibition of arbidol glucuronidation in HLMs and recom- 100 binant UGTs by mefenamic acid and androsterone 90 IC50 (mM) 80 UGT source Mefenamic acid Androsterone 70 60 UGT1A1 >50 >250 UGT1A3 >50 >250 50 UGT1A9 0.92 Ϯ 0.02 23.6 Ϯ 5.2 40 HLMs 1.56 Ϯ 0.26 49.3 Ϯ 11.7 Rel. Int. (%) 30 479.0 HLMs, human liver chromosomes, UGT, uridine glucuronosyltrans- 20 434.0 ferase. Arbidol (8 mM) was incubated with microsomes (0.05 mg/ml) 10 and UDP-glucuronic acid trisodium salt (5 mM) for 20 min in the 324.1 389.9 500.2 0 absence or presence of mefenamic acid (1–50 mM) or androsterone 50 100 150 200 250 300 350 400 450 500 550 600 650 (1–250 mM). The IC50 values are given as mean Ϯ SD. m/z

Figure 3 Mass/mass spectrum of metabolite. rUGT1A3 and rUGT1A9 were investigated. As shown in Table 1, androsterone and mefenamic acid weakly inhibited 3.5 the glucuronidation of arbidol by rUGT1A1 and rUGT1A3. Androsterone inhibited arbidol glucuronidation in HLMs 3.0 and rUGT1A9 in a concentration-dependent manner 2.5 (Figure 5a) and the IC50 value was calculated to be 49.3 Ϯ 11.7 mm and 23.6 Ϯ 5.2 mm for HLMs and 2.0 rUGT1A9 (Table 1), respectively. Mefenamic acid also inhibited arbidol glucuronidation in a concentration- 1.5 dependent manner (Figure 5b) and IC50 value was calculated to be 1.56 Ϯ 0.26 mm and 0.92 Ϯ 0.02 mm for HLMs 1.0 and rUGT1A9 (Table 1), respectively. The inhibitory effects

v (nmol/min/mg protein) of erlotinib (100 mm) towards arbidol glucuronidation are 0.5 shown in Figure S1 in Supporting Information. At 100 mm of erlotinib, the residual activity of arbidol glucuronidation 0.0 1A1 1A3 1A4 1A6 1A9 2B4 2B7 2B15 2B17 was 24.1%, 61.0%, 80.6% and 73.1% for UGT1A1, UGT1A3, UGT1A9 and HLMs, respectively. Figure 4 Arbidol glucuronidation by recombinant human uridine glucuronosyltransferase isoforms. Arbidol was incubated with recombinant human UGT1A1, UGT1A3, UGT1A4, UGT1A6, UGT1A9, Kinetic study in HLMs and recombinant UGT2B4, UGT2B7, UGT2B15 and UGT2B17 at 37°C for 30 min. Con- UGT isoforms centrations of 8 mM of arbidol and 0.1 mg/ml of protein were used. Each point represents the average of three independent experiments in Pooled HLMs (n = 12) were used for kinetic studies. duplicate. Arbidol glucuronidation followed the Michaelis–Menten kinetics in HLMs, as demonstrated by the Eadie–Hofstee (-44 Da). All these results demonstrate that glucuronida- plot (Figure 6). The kinetic parameters Km and apparent Ϯ Ϯ tion reaction occurred. Vmax were calculated to be 8.0 0.7 mm and 2.03 0.05 nmol/min/mg protein (Table 2), respectively. For Arbidol glucuronidation by recombinant involved recombinant UGT isoforms, the substrate UGT isoforms concentration–glucuronidation velocity curves also obeyed typical Michaelis–Menten kinetics (data not shown). The As shown in Figure 4, the incubation of arbidol with human kinetic parameters were as follows (Table 2): Km 9.6 Ϯ recombinant UGT isoforms revealed that UGT1A1, 1.1 mm and apparent Vmax 0.34 Ϯ 0.01 nmol/min/mg UGT1A3 and UGT1A9 were involved in the glucuronida- protein for UGT1A1; Km 12.4 Ϯ 1.2 mm and apparent Vmax tion of arbidol. 0.81 Ϯ 0.02 nmol/min/mg protein for UGT1A3; Km

2.7 Ϯ 0.2 mm and apparent Vmax 2.30 Ϯ 0.03 nmol/min/mg Chemical inhibition study protein for UGT1A9. The apparent intrinsic clearance was The inhibitory effects of androsterone and mefenamic acid 253.8, 35.4, 65.3 and 851.9 ml/min/mg protein for HLMs, on arbidol glucuronidation in pooled HLMs, rUGT1A1, UGT1A1, UGT1A3 and UGT1A9, respectively.

(a) 120 HLM 2.0 UGT 1A9 100 1.5 80

2.0 60 1.0 1.8 1.6 1.4 40 1.2

v 1.0 0.5 0.8 Activity (% of control) 20 0.6 v (nmol/min/mg protein) 0.4 0.2 0 0.0 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0 100 200 300 v/[S] Androsterone (μM) 02040 60 80 (b) Arbidol (μM) 120 HLM UGT 1A9 Figure 6 Kinetics of arbidol glucuronidation in human liver micro- 100 somes. Arbidol (1–150 mM) was incubated with pooled human liver microsomes (0.05 mg/ml) at 37°C for 20 min in the presence of UDP- 80 glucuronic acid trisodium salt. The Eadie–Hofstee plot is shown as an inset. Each point represents the average of three independent experi- 60 ments in duplicate.

40 Table 2 Kinetic parameters of arbidol glucuronidation in microsomes obtained from humans and in recombinant UGT1A1, UGT1A3 and Activity (% of control) 20 UGT1A9 Metabolite 0 0204060 UGT Source Apparent Vmax Km Apparent Vmax/Km Mefenamic acid (μM) HLM 2.03 Ϯ 0.05 8.0 Ϯ 0.7 253.8 UGT1A1 0.34 Ϯ 0.01 9.6 Ϯ 1.1 35.4 Figure 5 Inhibition of arbidol glucuronidation in human liver micro- UGT1A3 0.81 Ϯ 0.02 12.4 Ϯ 1.2 65.3 somes and UGT1A9 by androsterone and mefenamic acid. Each point UGT1A9 2.30 Ϯ 0.03 2.7 Ϯ 0.2 851.9 represents the average of three independent experiments in duplicate. a. Inhibition of arbidol glucuronidation by androsterone. Arbidol (8 mM) HLM, human liver microsome; UGT, uridine glucuronosyltransferase. Ϯ was incubated with various concentrations of androsterone The Vmax and Km values are given as mean SD. Units are: apparent (1–250 mM). b. Inhibition of arbidol glucuronidation by mefenamic Vmax, nmol/min/mg protein; Km, mM; apparent Vmax/Km, ml/min/mg acid. Arbidol (8 mM) was incubated with various concentrations of protein. androsterone (1–50 mM).

of arbidol glucuronide was predominantly catalysed by Discussion UGT1A9 in HLMs. This conclusion was supported by several observations First, among the involved UGT iso-

The extended use of arbidol in the clinic prompted us to forms, UGT1A9 has a higher afﬁnity (lower Km) and cata- gain a deeper understanding of its metabolic behaviour. lytic activity (higher apparent Vmax) than UGT1A1 and Previous study performed by Wang et al. showed that glu- UGT1A3. It should be noted that rUGT1A9 showed a lower [8] curonidation is the main elimination pathway of arbidol. Km value than HLMs, which might result from large differ- In the present study, we focused on identifying the UGT ences between the recombinant systems and native tissues. isoforms involved in the glucuronidation of arbidol. Screen- Second, comparable inhibitory effects of androsterone and ing of recombinant UGT isoforms showed that UGT1A1, mefenamic acid on human liver and recombinant UGT1A9 UGT1A3 and UGT1A9 could catalyse the glucuronidation can be seen from Table 1. However, androsterone and of arbidol. Given that UGT1A4 has been widely accepted mefenamic acid weakly inhibited the glucuronidation of to be the common UGT isoform involved in N- arbidol catalysed by rUGT1A1 and rUGT1A3. Additionally, glucuronidation and UGT1A4 was not involved in the a selective inhibitor of UGT1A1, erlotinib, strongly metabolism of arbidol, this glucuronide was proposed to be inhibited rUGT1A1-mediated arbidol glucuronidation, the O-glucuronide. The results indicated that the formation with UGT1A3-, UGT1A9- and HLM-catalysed arbidol

1 glucuronidation negligibly influenced, indicating the minor the AUC, Cmax and t /2 of zidovudine through inhibition of contribution of UGT1A1 to arbidol glucuronidation. For its glucuronidation, which is demonstrated by the reduction UGT1A3, the selective inhibitor trihydroxyvitamin D is not of formation of zidovudine glucuronide and the urinary commercially available, and no other good selective inhibi- molar ratio of zidovudine glucuronide to zidovudine.[34] tors of UGT1A3 have been reported. Therefore, the inhibi- Previous reports showed that valproic acid could inhibit the tion by a selective UGT1A3 inhibitor towards arbidol glucuronidation of lorazepam and result in the elevation of glucuronidation was not performed. However, UGT1A3 has lorazepam exposure.[35,36] Arbidol has been demonstrated to been reported to be expressed with low abundance in the be a good substrate of UGT1A9, so the metabolic elimina- liver (1/20 of the expression of UGT1A1),[20,21] so the contri- tion of arbidol through glucuronidation can be significantly bution of UGT1A3 towards arbidol glucuronidation could affected by the co-administration of inducers or inhibitors be limited. UGTs 1A7, 1A8 and 1A10 were not included in of UGT1A9. Additionally, due to the high binding of this study because these UGT isoforms are expressed in arbidol to UGT1A9, arbidol is speculated to exhibit strong extrahepatic tissues only, most notably the intestine and inhibition towards UGT1A9. Therefore, when arbidol is colon.[22,23] co-administered with drugs mainly undergoing UGT1A9- UGT1A9 is one of the most important UGT isoforms mediated metabolism, potential drug–drug interaction expressed in liver. It has been demonstrated to be involved might occur and close clinical monitoring is needed. in the metabolism of endogenous oestrogenic and thyroid hormones, as well as a variety of drugs, such as paracetamol Conclusions (acetaminophen),[24] diclofenac,[25] propofol[26] and mycophenolic acid.[27] Genetic polymorphisms have been identi- Our study identified UGT1A9 as the prominent UGT fied for UGT1A9 in a previous report;[28] this will result in isoform involved in arbidol glucuronidation in human liver individual differences in drug response. For example, the microsomes. These results are of significance for better influence of UGT1A9 polymorphisms on the metabolism of understanding of arbidol’s pharmacokinetic behaviour and mycophenolic acid has been frequently described. UGT1A9 lend guidance to the clinical application of arbidol. T-275A and C-2152T promoter single nucleotide polymorphisms (SNPs) have been associated with increased metabo- Declarations lism of mycophenolic acid. In contrast, SNPs at nucleotide base positions 8 and 98 could reduce the metabolism of Conflict of interest mycophenolic acid and enhance the exposure to mycophe- The Author(s) declare(s) that they have no conflicts of nolic acid.[29–31] interest to disclose. Serious adverse effects caused by drug–drug interactions could lead to termination of the development of Funding promising new therapies, withdrawal of a drug from the market or severe restrictions/limitations on its use.[32] This work was supported by the National Key Technology Great attention should be paid to UGT-associated clinical R&D Program (2012BAI7B02), the National Science & drug–drug interactions.[33] For example, zidovudine Technology Major Project of China (2012ZX09501001, (3′-azido-3′-deoxythimidine) mainly undergoes UGT2B7- 2012ZX09506001 & 2012ZX10002011), and the National mediated metabolism to form 3′-azido-3′-deoxy-5′-b-d- Natural Science Foundation of China (81202586, 81001473 glucopyranuronosylthymidine. Fluconazole could increase & 81273590).

References 3. Brooks MJ et al. Antiviral chemo- arbidol. Biochemistry 2007; 46: 6050– therapeutic agents against respiratory 6059. 1. Leneva IA et al. Sensitivity of various viruses: where are we now and what’s 6. Zhong Q et al. Antiviral activity of influenza virus strains to arbidol. in the pipeline? Curr Opin Pulm Med arbidol against coxsackie virus B5 in Influence of arbidol combination with 2004; 10: 197–203. vitro and in vivo. Arch Virol 2009; 154: different antiviral drugs on reproduc- 4. Chai H et al. Synthesis and in vitro 601–607. tion of influenza virus A. Ter Arkh anti-hepatitis B virus activities of 7. Liu MY et al. Pharmacokinetic prop- 2005; 77: 84–88. some ethyl 6-bromo-5-hydroxy-1H- erties and bioequivalence of two 2. Burtseva EI et al. Rimantadine and indole-3-carboxylates. Bioorg Med fomulations of arbidol: an open-label, arbidol sensitivity of influenza viruses Chem 2006; 14: 911–917. single-dose, randomized-sequence, that caused epidemic morbidity rise in 5. Pecheu EJ et al. Biochemical mecha- two-period crossover study in healthy Russia in the 2004-2005 season. Vopr nism of hepatitis C virus inhibition Chinese male volunteers. Clin Ther Virusol 2007; 52: 24–29. by the broad-spectrum antiviral 2009; 31: 784–792.

8. Wang Y et al. Metabolite identification 19. Fang ZZ et al. Identification of cyto- enzymes. Pharmacogenomics J 2003; 3: of arbidol in human urine by the chrome P450 (CYP) isoforms involved 136–158. study of CID fragmentation pathways in the metabolism of corynoline, and 29. Levesque E et al. The impact of using HPLC coupled with ion trap assessment of its herb-drug interac- UGT1A8, UGT1A9 and UGT2B7 mass spectrometry. J Mass Spectrom tions. Phythother Res 2011; 25: 256– genetic polymorphisms on the phar- 2008; 43: 1099–1109. 263. macokinetic profile of mycophenolic 9. Miners JO, Mackenzie PI. Drug glu- 20. Ohno S, Nakajin S. Determination of acid after a single oral dose in healthy curonidation in human. Pharmacol mRNA expression of human UDP- volunteers. Clin Pharmacol Ther 2007; Ther 1991; 51: 347–369. Glucuronosyltransferases and applica- 81: 392–400. 10. Kiang TK et al. UDP- tion for localization in various human 30. Inoue K et al. Influence of UGT1A7 glucuronosyltransferases and clinical tissues by real-time reverse and UGT1A9 intronic I399 genetic drug-drug interactions. Pharmacol transcriptase-polymerase chain reac- polymorphisms on mycophenolic acid Ther 2005; 106: 97–132. tion. Drug Metab Dispos 2009; 37: pharmacokinetics in Japanese renal 11. Ritter JK. Roles of glucuronidation 32–40. transplant recipients. Ther Drug Monit and UDP-glucuronosyltransferases in 21. Zhu LL et al. Characterization of 2007; 29: 299–304. xenobiotic bioactivation reactions. hepatic and intestinal glucuronidation 31. Kuypers DR et al. The impact of uridine Chem Biol Interact 2000; 129: 171–193. of magnolol: application of the RAF diphosphate-glucuronosyltransferase 12. Tukey RH, Strassburg CP. Human approach to decipher the contribu- 1A9 (UGT1A9) gene promoter region UDP-glucuronosyltransferases: meta- tions of multiple UGT isoforms. Drug single-nucleotide polymorphisms T- bolism, expression, and disease. Annu Metab Dispos 2012; 40: 522–528. 275A and C-2152T on early mycophe- Rev Pharmacol Toxicol 2000; 40: 581– 22. Cheng Z et al. Studies on the substrate nolic acid dose-interval exposure in de 616. specificity of human intestinal UDP- novo renal allograft recipients. Clin 13. Fang ZZ et al. Time-dependent inhibi- glucuronosyltransferases 1A8 and Pharmacol Ther 2005; 78: 351–361. tion of CYP3A4 and CYP2C9 by 1A10. Drug Metab Dispos 1999; 27: 32. Sun DX et al. Inhibitory effects of cur- noscapine potentially explains clinical 1165–1170. cumenol on human liver cytochrome noscapine-warfarin interaction. Br J 23. Mojarrabi B, Mackenzie PI. Charac- P450 enzymes. Phytother Res 2010; 24: Clin Pharmacol 2010; 69: 193–199. terization of two UDP- 1213–1216. 14. Fang ZZ et al. Metabolic map and bio- glucuronosyltransferases that are 33. Huang T et al. Inhibitory potential activation of the anti-tumor drug predominantly expressed in human of chlormadinone acetate noscapine. Br J Pharmacol 2012; 167: colon. Biochem Biophys Res Commun (CMA) on five important UDP- 1271–1286. 1998; 247: 704–709. glucuronosyltransferases in human 15. Lowry OH et al. Protein measurement 24. Court MH et al. Interindividual vari- liver. Pharmazie 2011; 66: 212–215. with the Folin phenol reagent. J Biol ability in acetaminophen glucuronida- 34. Sahai J et al. Effect of fluconazole on Chem 1951; 193: 265–275. tion by human liver microsomes: zidovudine pharmacokinetics in 16. Uchaipichat V et al. Selectivity of sub- identification of relevant acetami- patients infected with human immu- strate (trifluoperazine) and inhibitor nophen UDP-glucuronosyltransferase nodeficiency virus. J Infect Dis 1994; (amitriptyline, androsterone, canre- isoforms. J Pharmacol Exp Ther 2001; 169: 1103–1107. noic acid, hecogenin, phenylbutazone, 299: 998–1006. 35. Anderson GD et al. Lorazepam- quinidine, quinine, and sulfinpyra- 25. King CD et al. Characterization valproate interaction: studies in zone) ‘probes’ for human udp- of rat and human UDP- normal subjects and isolated perfused glucuronosyltransferases. Drug Metab glucuronyltransferases responsible rat liver. Epilepsia 1994; 35: 221–225. Dispos 2006; 34: 449–456. for the in vitro glucuronidation of 36. Samara EE et al. Effect of valproate on 17. Gaganis P et al. Glucuronidation of diclofenac. Toxicol Sci 2001; 61: 49–53. the pharmacokinetics and pharmaco- fenamates: kinetic studies using 26. Soars MG et al. The effect of incuba- dynamics of lorazepam. J Clin Phar- human kidney cortical micro- tion conditions on the enzyme kinet- macol 1994; 37: 442–450. somes and recombinant UDP- ics of UDP-glucuronyltransferases. glucuronosyltransferase (UGT) 1A9 Drug Metab Dispos 2003; 31: 762–767. Supporting Information and 2B7. Biochem Pharmacol 2007; 73: 27. Bernard O, Guillemette C. The main Additional Supporting Information may 1683–1691. role of UGT1A9 in the hepatic be found in the online version of this 18. Huang T et al. Strong inhibitory metabolism of mycophenolic acid and article: effect of medroxyprogesterone the effects of naturally occurring vari- acetate (MPA) on UDP- ants. Drug Metab Dispos 2004; 32: Figure S1 Inhibition potential of erlotinib glucuronosyltransferase (UGT) 2B7 775–778. towards UGT1A1, 1A3, 1A9 and human might induce drug-drug interactions. 28. Guillemette C. Pharmacogenomics of liver microsomes (HLMs)-catalyzed Pharmazie 2010; 65: 919–921. human UDP-glucuronosyltransferase arbidol glucuronidation.