DMD# human UDP-glucuronosyltransferases, species differences, and interaction potential

Transcription

1 DMD This Fast article Forward. has not been Published copyedited and on formatted. April 9, The 2010 final as version doi: /dmd may differ from this version. Regioselective glucuronidation of tanshinone IIa following quinone reduction: identification of human UDP-glucuronosyltransferases, species differences, and interaction potential Qiong Wang, Haiping Hao, Xuanxuan Zhu, Guo Yu, Li Lai, Yitong Liu, Yuxin Wang, Shan Jiang, and Guangji Wang Key Lab of Drug Metabolism and Pharmacokinetics, Key Unit of SATCM for Pharmacokinetic Methodology of TCM Complex Prescription, China Pharmaceutical University, Nanjing , China (Q.W., H.H., G.Y., L.L., Y.L., Y.W., S.J., G. W.) Pharmacological Laboratory of Clinical Research Institute, Jiangsu Provincial Hospital of Traditional Chinese Medicine, Nanjing , China (X.Z) 1 Copyright 2010 by the American Society for Pharmacology and Experimental Therapeutics.

2 Running title: Regioselective and species dependent glucuronidation of TSA Corresponding Author: Prof. Guangji Wang, Key Lab of Drug Metabolism and Pharmacokinetics, China Pharmaceutical University 24 Tong Jia Xiang, Nanjing , China. Phone: Fax: Number of text pages: 34 Number of tables: 1 Number of figures: 8 Number of references: 37 Number of words in the Abstract: 250 Number of words in the Introduction: 756 Number of words in the Discussion: 1479 Abbreviations: UGT, UDP-glucuronosyltransferase; UDPGA, UDP-glucuronic acid; TSA, tanshinone IIa; NQO1, NAD(P)H: quinone oxidoreductase 1; MPA, mycophenolic acid; NADP, β-nicotinamide adenine dinucleotide phosphate; 4-MU, 4-methylumbelliferone; HPLC, High-performance liquid 2

3 chromatography; IC 50, inhibitor concentration that causes 50% inhibition; LC-MS/MS, liquid chromatography/tandem mass spectrometry; N.D., not detected; MLS9, mouse liver S9; RLS9, rat liver S9; DLS9, dog liver S9; HLS9, human liver S9; MIS9, mouse intestinal S9; RIS9, rat intestinal S9; DIS9, dog intestinal S9; HIS9, human intestinal S9; HLM, human liver microsome; CYP, Cytochrome P450; CL int, intrinsic clearance. 3

4 Abstract We have previously identified that the NQO1-mediated quinone reduction and subsequent glucuronidation is the predominant metabolic pathway for tanshinone IIa (TSA) in rat. The present study contributes to the further research on its glucuronidation enzyme kinetics, the identification of human UDP-glucuronosyltransferase (UGT) isoforms, and the interaction potential with typical UGT substrates. A pair of regioisomers (M1 and M2) of reduced TSA glucuronides was found from human, rat, and mouse, whereas only M1 was found in dog liver S9 incubations. The overall glucuronidation clearance of TSA in human liver S9 was 11.8±0.8µl/min/mg protein, 0.7, 0.8, and 3 fold of that in the mouse, rat and dog, respectively. Using CL int M2/M1 as a regioselective index, opposite regioselectivity was found between human (0.7) and mouse (1.3), whereas no significant regioselectitvity was found in rat. In a sequential metabolism system by applying human liver cytosol as a NQO1 donor in combination with a panel of 12 recombinant human UGTs screening, multiple UGTs were found involved in the M1 formation, whereas only UGT1A9 and to a very minor extent of UGT1A1 and 1A3 contributed to the M2 formation. Further enzyme kinetics, correlation, and chemical inhibition studies confirmed that UGT1A9 played the major role for both M1 and M2 formation. In addition, TSA presented potent inhibitory effect on the glucuronidation of typical UGT1A9 substrates propofol and MPA, with an IC50 value at 8.4±1.8µM and 8.9±1.9µM respectively. This study would be helpful for guiding future studies on characterizing the NQO1-mediated reduction and subsequent glucuronidation of other quinones. 4

5 Introduction Quinones are such a large group of compounds that are ubiquitous in nature and characterized with great biological, pharmacological, and/or toxicological significance. Human beings are unavoidably to be exposed to many endogenous quinones such as co-enzyme Q, vitamin K, oxidized products of endogenous phenols, and more or less to some xenobiotic quinones from foods, environmental pollutants, and drugs. The carbonyl group in the quinone structure is often a determining functional factor for their activities(oppermann, 2007). For example, the quinone/hydroquinone interconversion of Coenzyme-Q plays an important role in the electron transport involved in cellular respiration, and in addition, the one-electron and/or two-electron reduction of the quinone carbonyl groups triggered redox cycle explains largely the pharmacological or toxicological activities of many quinones. Interestingly, the quinone carbonyl reduction and subsequent conjugation constitutes the major biotransformation and metabolic elimination pathway of most quinones. Therefore, the biotransformation study of quinones should be essential for understanding their bioactivation, detoxification, and/or inactivation process. Tanshinones are a class of diterpene phenanthrenequinone compounds isolated from the dried root of Salvia miltiorrhiza (Fam. Labiatae) that is a widely used traditional Chinese medicine with well proven cardiovascular and cerebrovascular efficacies (Han et al., 2008). Danshen is also markedly available in the United States, European countries, Japan and worldwide for improving body function (Zhou et al., 2005). The market share of Danshen pharmaceutics in China has now been estimated to exceeding US$ 2 billion and increases by more than 20% each year. Tanshinones are the major lipophilic components contained in Danshen and account for the majority of its 5

6 pharmacological activities. Of more than 30 tanshinones identified, tanshinone IIA (TSA) may be the most important one, because of its high contents and widely confirmed pharmacological activities. The cardiovascular and cerebrovascular efficacies of TSA had been recognized early in 1970s, and its sodium sulphonate derivative was marketed in China as a cardiovascular drug in 1980s. More recently, various other pharmacological activities including anti-inflammation(ren et al.), liver and renal protective effect(zhang et al., 2009a), learning and memory improvement(kim et al., 2009), and especially its anticancer potential (Wang et al., 2005; Lu et al., 2009; Shi et al., 2009; Zhang et al., 2009b) have been extensively studied and attracted much attention from researchers worldwide. Previous reports strongly suggest that TSA is a promising leading compound for the novel anticancer and cardiovascular drug development. We had previously conducted comprehensive studies on the TSA pharmacokinetics and dispositions both in vivo and in vitro, based on developing some established bioanalytical methods (Li et al., 2005; Hao et al., 2006b). TSA bioavailability was found extremely poor because of its poor solubility and membrane permeability (Hao et al., 2006a), and especially the extensive intestinal first pass metabolism (Hao et al., 2007). NQO1catalyzed quinone reduction, producing a highly unstable catechol intermediate, followed by an immediate glucuronidation was identified as the predominant metabolic pathway of TSA in rat (Hao et al., 2007). Although some hydroxylated metabolites had been identified from the rat bile samples and the in vitro liver microsomal incubation media, most hydroxylated metabolites also underwent further quinone reduction and subsequent glucuronidation (Li et al., 2006). Such a metabolic pathway has been lately confirmed with other tanshinones such as cryptotanshinone (Dai et al., 2008) and dihydrotanshinone I (Liu et 6

7 al., 2007). The evidence strongly indicates that the quinone reduction and following glucuronidation plays an important role on the metabolic elimination of tanshinones, and thus may determine its pharmacological efficacies. The present study was thus designed to extend the research to characterize the enzyme kinetics of TSA glucuronidation in human and different animal species, and to identify the major human UGT isoforms involved, and the potential interactions with typical UGTs substrates. Because NQO1 is mainly distributed in cytosol whereas UGTs is mainly located in microsomes, S9 fractions were applied in the present study for characterizing TSA enzyme kinetics. When human recombinant UGTs assay was involved, the human liver cytosol was used as a NQO1 donor to mimic the in vivo conditions. To identify the UGT isoforms responsible for TSA glucuronidation, the recombinant UGTs screening, correlation analysis, and chemical inhibition study were performed according to the currently adopted protocols for UGTs identification (Fisher et al., 2001; Mano et al., 2007; Wen et al., 2007). Furthermore, as TSA was an UGT1A9 substrate with high affinity indicated by this study, it was of concern whether TSA was a competitive inhibitor of UGT1A9 substrates. For this consideration, in vitro inhibitory effects of TSA on the glucuronidation of the typical UGT1A9 substrates propofol and MPA were determined for evaluating the interaction potential. 7

8 Materials and Methods Chemicals and Reagents Tanshinone IIA, bilirubin and morphine were purchased from the National Institute for the Control of Pharmaceutical and Biological Products (Beijing, China). Glucose 6-phosphate, β-nicotinamide adenine dinucleotide phosphate (NADP), glucose 6-phosphate dehydrogenase (PDH), uridine 5 -diphosphate-glucuronic acid (UDPGA), D-saccharic acid 1,4-lactone, β-glucuronidase (Escherichia coli), alamethicin, mycophenolic acid (MPA), 4-methylumbelliferone (4-MU), Zidovudine (AZT) and propofol were all purchased from Sigma Chemical (St. Louis, MO, USA). Pooled human liver S9, pooled human liver cytosol and a panel of recombinant human UGT Supersomes TM (UGT1A1, 1A3, 1A4, 1A6, 1A7, 1A8, 1A9, 1A10, 2B4, 2B7, 2B15 and 2B17) expressed in baculovirus infected insect cells were obtained from BD Biosciences (Bedford, MA, USA). All other chemicals were of HPLC grade or the best grade that was commercially available. Animals Male ICR mice (18-22g) were supplied by Center of Experimental Animals, China Pharmaceutical University. Male Sprague-Dawley rats ( g) were obtained from Academy of Military Medical Sciences, China. Beagle dogs (8-12kg) were offered by Center of Experimental Animals, China Pharmaceutical University. All animals were acclimated for at least one week before experiments and allowed water and standard chow ad libitum. All animal studies were approved by the Animal Ethics Committee of China Pharmaceutical University. Preparation of Subcelluar Fractions 8

9 Liver and small intestines from mouse, rat, and dog were rinsed with 10 mm phosphate buffer (ph 7.4) containing 5 mm EDTA and 0.5% NaCl. Liver tissues were homogenized to make 20% homogenates in the 10 mm phosphate buffer containing 5 mm EDTA. Small intestines were cut longitudinally and the mucous membrane was scraped gently with a slide. The mucous scrapings were then mixed with 5 volumes (w/v) of 10 mm phosphate buffer containing 5 mm EDTA and homogenized to make 20% homogenates. The homogenates were subjected to centrifugation at 9000 g for 20min to isolate post-mitochondrial fractions (S9). The protein concentrations of the subcellular fractions were determined with a commercially available kit (BCA protein assay; Pierce Chemical Co., Rockford, IL) as described by the manufacturer. All subcellular fractions were stored at -80 C before use. Glucuronidation assay of TSA in S9 fractions The pooled subcellular fraction S9 was pretreated with alamethicin at 25µg/mg protein on ice for 15min to diminish the latency of UGT activity. TSA 20 µm was incubated in reaction mixture consisted of 0.2 mg S9 protein, 2mM UDPGA, 1mM saccharic acid-1,4-lactone, 5mM MgCl 2, a NADPH regenerating system containing 0.2mM NADP, 1.9mM glucose-6-phosphate, 1.2U/ml glucose-6-phosphate dehydrogenase and 100mM potassium phosphate buffer (PH 7.4) in a final volume of 200µl. After preincubation for 5min at 37 C, the reaction was started by adding UDPGA. All reactions were incubated at 37 C for 30 min, terminated by cold acetonitrile followed by centrifugation at 20,000g for 10min to obtain the supernatants and then analyzed by HPLC and LC-MS/MS according to our previous reports (Li et al., 2005; Hao et al., 2006b; Hao et al., 2007). 9

10 Glucuronidation assay of TSA in recombinant UGTs Because TSA is firstly reduced by NQO1 to a highly unstable catechol intermediate which is then subjected to the following glucuronidation, a pooled human liver cytosol (0.2mg protein/ml) used as the NQO1 donor in combination with the recombinant UGTs was applied for reaction phenotype screening. A panel of commercially available UGT isoforms including UGT 1A1, 1A3, 1A4, 1A6, 1A7, 1A8, 1A9, 1A10, 2B4, 2B7, 2B15, and 2B17 (protein concentrations were used as manufacturer s recommendation) was screened for TSA glucuronidation at two different concentrations of 2µM and 20µM. The incubation conditions were the same as that described above for S9 incubations. The reaction was terminated by cold acetonitrile and the supernatants were analyzed by HPLC. Quantification of TSA and its glucuronides A HPLC method was developed and validated for the quantification of TSA and its two glucuroides. A SHIMAZU LC-2010C system with a quaternary pump, autosampler, column oven and UV detector was used. The separation was performed on a Diamonsil-C18 column (250mm 4.6mm I.D., 5µm, Dikma Technologies, Dalian, China) with a guard column. The mobile phase consisted of 0.1% (V/V) acetic acid and 3mM ammonium acetate (PH 4.5) in acetonitrile (A) and water (B) at a flow rate of 1ml/min. The separation of TSA with its metabolites was achieved using the following elution gradient: linear gradient from 30% A to 70% A (0-15 min), 70% A to 90% A (15-25min), and then returned to 30% A for another 10min of equilibration. The column temperature was 40 C and UV detection was achieved at 260nm. The retention time of M1, M2, and TSA was 12.1min, 12.6min and 25.7min, respectively. The 10

11 HPLC-UV method was found linear in a concentration range from to 160µM with the correlation coefficients over The lower limit of quantification was of 0.078µM. The extraction recovery of TSA was determined to be over 85% at all tested concentrations. The intra-batch accuracy for TSA ranged from 103.1~114.9% at the tested concentrations with the precision (R.S.D) within 9.1%. The inter-bath accuracy for TSA ranged from 100.5% ~ 114.1% at the tested concentrations with the precision (R.S.D) within 10.5%. Because no authentic standards of TSA glucuronides were currently available, a method of related quantitative coefficients was developed for quantifying TSA glucuronides. For this purpose, incubated mixtures prepared as described above except with D-Saccharic acid 1,4-lactone omitted were centrifuged (9000g for 10min) and an aliquot of 100 µl was transferred to another tube for the further incubation at 37 C for 4h by adding an equal volume of acetic acid/sodium acetate buffer ph4.5 containing 1500 units of β-d-glucuronidase. The incubated mixture was then treated by cold acetonitrile and centrifuged prior to analysis. The control samples were treated in the same way, except without the addition of β-d-glucuronidase. The related quantitative coefficients were then determined by comparing the peak area of the TSA released from the hydrolysis with that of TSA glucuronides. Enzyme Kinetics analysis in S9 fraction and recombinant UGTs Enzyme kinetics for TSA ( µM) glucuronidation were conducted in pooled liver S9 fractions (mouse, rat, dog, and human), intestinal S9 fractions (mouse and rat), and also in the recombinant human UGT1A9 (0.05mg/ml), and UGT2B7 (0.25mg/ml) in combination with a pooled human liver cytosol. The protein concentrations and incubation time were preferentially 11

12 optimized to ensure linear formation of metabolites. The final selected protein concentration for mouse, rat liver S9 was 0.5mg/ml and was 1mg/ml for dog, human liver S9, and mouse intestinal S9, whereas that for the rat intestinal S9 was 2 mg/ml. The incubation time was all set at 30 min, except that 20 min was used for mouse and rat liver S9 incubation. Other incubation conditions were the same as that described previously. Correlation analysis The glucuronidation of TSA was conducted in seven individual human liver microsomes in combination with a pooled human liver cytosol to assess the individual differences of TSA glucuronidation and to correlate with the typical UGTs substrates. Glucuronidation activities of estradiol 3β-glucuronidation (UGT1A1), trifluoperazine glucuronidation (UGT1A4), and propofol glucuronidation (UGT1A9) were obtained from the manufacturer. AZT glucuronidation (UGT2B7) was measured based on a previously reported method (Sim et al., 1991). Correlation analysis was conducted and p <0.05 was considered statistically significant. Chemical inhibition studies Four typical UGTs probe substrates including bilirubin (UGT1A1), 4-methylumbelliferone (UGT1A6), mycophenolic acid (UGT1A9), and morphine (UGT2B7) were investigated for their potential inhibitory effect on TSA glucuronidation. TSA glucuronidation activities (20 µm) in the pooled human liver S9 were determined in the absence or presence of each of the four typical substrates (0-500µM). IC 50 values representing the 50% inhibition concentration were estimated graphically. 12

13 Inhibitory effect of TSA on propofol and MPA glucuronidation To evaluate the potential inhibitory effect of TSA on UGT1A9 activity, TSA (0-40µM) was co-incubated with the typical UGT1A9 substrates propofol (50µM) and MPA (100µM) in the pooled human liver S9 system containing UDPGA and with or without NAD(P)H as the cofactors. The incubation conditions were the same as that described previously. Propofol, MPA and their glucuronides were determined based on the previous reports (Vree et al., 1999; Miles et al., 2005). Data Analysis Metabolic rate was expressed as pmol/min/mg protein. Kinetic parameters were estimated from the fitted curves using suitable model determined from the goodness of fitting. The formation kinetics of TSA glucuronides M1 and M2 in the pooled S9 fractions of all species, and in the recombinant system (UGT1A9 and UGT2B7) were finally all fitted to the typical Michaelis-Menten kinetics, and thus the Michaelis-Menten equation: v = (Vmax [S]) /(Km + [S]), where Km is the Michaelis-Menten constant, Vmax is the maximum velocity, and S is the substrate concentration, was applied to estimate the apparent kinetic parameters using a nonlinear least squares method. The correlation study was conducted by the Spearman correlation analysis and the correlation is considered significant at the p<0.05. Kinetic parameters and IC 50 values are reported as the means ± standard deviation. 13

14 Results Tanshinone IIA glucuronidation in hepatic and intestinal S9 systems The production of TSA catechol glucuronides in human liver S9, and in mouse, rat and dog liver and intestine S9 incubation systems was profiled by HPLC-UV and LC-MS/MS. The representative HPLC chromatogram for TSA metabolites profiling is shown in Fig.1. The predominant metabolites M1 (Rt, 12.1min) and M2 (Rt, 12.6min) were characterized as the two glucuronides of the quinone reduced TSA at C-10 or 11 position by product ions scan, neutral loss scan and glucuronidase hydrolysis based on the method of our previous reports (Li et al., 2005; Hao et al., 2006b; Li et al., 2006; Hao et al., 2007). It was found that TSA underwent quinone reduction and subsequent glucuronidation metabolism in the liver S9 systems of all species studied here, and also in the intestinal S9 of rat and mouse. Apparent species differences and regioselectivity to produce the two metabolites could be observed directly from Fig.1. Enzyme kinetics assay in the S9 incubation systems was performed to further clarify the species differences and regioselectivity of reduced TSA glucuronidation. As shown in Fig.2, the reduced TSA glucuronidation fitted well to the typical Michaelis-Menten kinetics. Kinetic parameters including the apparent Km, Vmax and CL int (Vmax/Km) for the respective M1, M2, and the sum CL int (M1+M2) are summarized in Table 1. The CL int M2/M1 value was also provided for showing the regioselectivity. Significant regioselectivity was found in the mouse (CL int M2/M1, 1.3) and human liver S9 (CL int M2/M1, 0.7), but not in the rat liver S9 (CL int M2/M1, 1.0). It was interesting to find that only M1 was detected from the dog liver S9, suggesting an extreme regioselective glucuronidation of TSA in dog. The CL int M1 values were almost equivalent throughout the liver S9 of all species except dog (about one fold lower than others); however, the CL int M2 values in 14

15 mouse, rat, dog and human liver S9 being 9.2±0.1, 7.7±0.2, N.D., and 4.5±0.1µl/min/mg, respectively, were significantly different among the species. The CL int of total TSA glucuronidation ranked as mouse>rat>human>dog. The M1 and M2 CL int values in the mouse and rat intestine S9 were about 2-fold lower than those in the liver, and only a very minor amount of glucuronides were detected in the dog intestine S9 so that their clearances were not obtained. Reaction penotyping of TSA glucuronidation Human liver cytosol was used as the NQO1 donor to produce the reduced TSA which was then subjected to the following glucuronidation screening with a panel of commercially available recombinant UGT isoforms (UGT 1A1, 1A3, 1A4, 1A6, 1A7, 1A8, 1A9, 1A10, 2B4, 2B7, 2B15, and 2B17). Of all UGTs screened, multiple UGTs including UGT1A1, 1A3, 1A9, 1A10, 2B7 and 2B15 catalyzed the M1 formation, whereas only UGT1A9 and to a very low extent UGT1A1 and 1A3 contributed to the formation of M2 (Fig.3). UGT1A9 was found to be the major isoform that catalyzed the formation of both M1 and M2, characterized with a catalytic activity at 12.4pmol/min/mg and 3.2pmol/min/mg, respectively, at the TSA concentration of 20µM. UGT2B7 with a catalytic activity of 4.5pmol/min/mg for M1 production may be another important isoform contributing to TSA glucuronidation. Enzyme kinetics in recombinant human UGT1A9 and UGT2B7 Because UGT1A9 and 2B7 were characterized as the major UGT isoforms responsible for TSA glucuronidation, the further enzyme kinetics of TSA glucuronidation in recombinant UGT1A9 or UGT2B7 in combination with human liver cytosol incubation system was performed. Typical 15

16 Michaelis-Menten kinetics was also found for the reduced TSA glucuronidation in the recombinant UGT1A9 and UGT2B7 system (Fig.4). For UGT1A9, the apparent kinetic parameters Km, Vmax, and CL int for M1 were 1.1±0.2µM; 10.4±0.3pmol/min/mg and 9.6±1.1µl/min/mg, for M2 were 0.3±0.1µM, 2.7±0.1pmol/min/mg and 7.9±0.3µl/min/mg, respectively. UGT2B7 could only catalyze the formation of M1 with the Km, Vmax and CL int at 1.5±0.1µM, 3.1±0.1pmol/min/mg and 2.1±0.2µl/min/mg respectively. Correlation analysis with UGT probe substrates in individual HLMs. TSA glucuronidation velocities in 7 individual human liver microsomes in combination with pooled human liver cytosol as a NQO1 donor were measured and correlated with the typical UGT1A1, 1A4 and 1A9 substrates. The TSA M1 and M2 formation velocities in the 7 individual HLMs ranged from 6.2 to 13.9, and 1.5 to 4.2 pmol/min/mg protein, respectively. Both TSA M1 and M2 formation were found correlated well (M1: r=0.714, p=0.024; M2: r=0.951, p=0.001) with the glucuronidation activities of propofol, a typical UGT1A9 substrate (Fig. 5). In contrast, no significant correlation was found with the estradiol 3β-glucuronidation (UGT1A1), trifluoperazine glucuronidation (UGT1A4), and AZT glucuronidation (UGT2B7). Inhibition of TSA glucuronidation by UGT probe substrates The inhibitory effects of bilirubin (UGT1A1), 4-MU (UGT1A6), MPA (UGT1A9) and morphine (UGT2B7) on TSA glucuronidation activities in the pooled human liver S9 were determined. As shown in Fig.6, MPA showed potent inhibitory effect on both M1 and M2 formation, characterized with the IC 50 value of 156±12µM and 122±10µM, respectively. Bilirubin, 16

17 4-methylumbelliferone and morphine showed negligible inhibitory effect on M1 and M2 formation. Inhibitory effect of TSA on propofol and MPA glucuronidation Considering that human UGT1A9 was identified as the predominant isoform contributing to TSA glucuronidation, we further evaluate the potential inhibitory effect of TSA on the glucuronidation of typical UGT1A9 substrates propofol and MPA. As a result, TSA displayed a potent and NAD(P)H dependent inhibitory effect on propofol and MPA glucuronidation with an IC 50 value at 8.4±1.8µM and 8.9±1.9µM, respectively (Fig.7). 17

18 Discussion Glucuronidation defined as the glucuronic acid transfer from UDPGA to a large variety of aglycones catalyzed by UGTs, a class of membrane bound enzymes of endoplasmic reticulum, plays an important role on detoxifying or inactivating of many endogenous compounds and xenobiotics (Tukey and Strassburg, 2000; Fisher et al., 2001). In human, nineteen UGTs classified into two families, UGT1 and UGT2, have been identified (Mackenzie et al., 2005). Compared to CYP450s, the UGTs phenotyping is more difficult due to large substrates overlapping. It may be even more challengeable for characterizing the UGTs involved in quinones carbonyl glucuronidation, considering their characteristic of sequential metabolisms, different subcellular location of enzymes involved, and especially of the production of highly unstable intermediates. As an extension of our previous studies, the present study contributes to identifying the human UGTs responsible for TSA glucuronidation, the regioselective enzyme kinetics in human and different animal species, and the interaction potential with typical UGTs substrates, based on developing an in vitro incubation system for evaluating the sequential quinone reduction followed by an immediate glucuronidation. Previously, we had found that the NQO1 catalyzed two electron reduction followed by immediate glucuronidation was the predominant metabolic pathway of TSA in rat. In the present study, we sought to discover whether TSA underwent similar metabolism in human and other animal species in a S9 incubation system. As a result, two glucuronides identified as a pair of positional isomers have been found from the incubation media of HLS9, MLS9, RLS9, RIS9, and MIS9, whereas only one glucuronide (M1, Rt 12.1min) was found from DLS9 and only a very minor amount of 18

19 both glucuronides from DIS9, suggesting a regioselective and species dependent glucuronidation of reduced TSA. Because no detailed study on such enzymes expression and activity in dog intestine was found in the current literature, it was presently unclear whether the low level of M1 and M2 formation activities in the dog intestine was caused by the low activities of NQO1 or specific UGT isoforms. Further kinetic studies showed that the intrinsic clearance of overall hepatic glucuronidatons CL int (M1+M2) in human was 0.7, 0.8, and 3 fold of that in the rat, mouse and dog, respectively. In comparison with liver, the overall intestinal glucuronidation clearance was about 2-fold lower, which was not surprising considering that the glucuronidation activities of most substrates in intestine were relatively lower than those in liver (Shiratani et al., 2008). However, the overall contribution of intestinal glucuronidation of the oral ingested TSA may be much higher than that of liver, because of the longer retention and higher exposure in intestine (Hao et al., 2007). The enzyme kinetics of TSA in HIS9 was not obtained because the human intestinal S9 was commercially unavailable. Using CL int M2/M1 as a regioselective index, it was interesting to find that the M1 and M2 production in rat (1.0) were almost identical, whereas a significant but opposite regioselectivity was observed between human (0.7) and mouse (1.3). It seems that regioselecitivity and species differences are very common for UGTs, since similar phenomena were found for many other UGT substrates (Rouguieg et al.; Kaji and Kume, 2005b; Kaji and Kume, 2005a; Shiratani et al., 2008) but the underlying mechanism is unclear until now. To identify the UGTs that responsible for TSA glucuronidation, a panel of 12 commercially 19

20 available human recombinant UGTs was screened. Since TSA is necessary to be reduced by NQO1 to form a highly unstable catechol intermediate, an incubation system for evaluating the quinone sequential metabolism has to be developed for UGTs screening. To address such a concern, a more recent study constructed heterologously expressed NQO1 in Sf9 cells and used it as the NQO1 donor for testing menadiol glucuronidation activity in a panel of recombinant UGT isoforms(nishiyama et al., 2008). Although this method was practicable for UGTs screening, it may poorly mimic the in vivo conditions for conducting the enzyme kinetics study in recombinant UGTs. For this consideration, we developed a new incubation system by applying pooled human liver cytosol as the NQO1 donor to conduct the UGTs screening and the followed enzyme kinetics assay, which would be generally applicable for the sequential metabolism of quinones. Using such an incubation system, it was found that multiple UGTs including UGT1A1, 1A3, 1A6, 1A10, 2B7, and 2B15 contributed to the M1 formation, whereas predominantly UGT1A9 and to a very minor extent UGT1A1 and 1A3 was involved in the M2 production. Unlike CYPs (Wojcikowski et al., 2003), it is still difficult to calculate the relative contribution of each UGT isoform to the total TSA glucuronidation in the microsomes because the relative abundance of each UGT isoform to the total UGT contents in the microsomes have not been determined (Mano et al., 2007), in spite of some recent reports handling this issue (Rouguieg et al.). Further kinetic studies in the recombinant UGT1A9 and 2B7, the two isoforms with the highest catalytic activities, in combination with human liver cytosol as the NQO1 donor were performed. Like HLM, the rugt1a9 prefers to produce M1 with the CLint M2/ M1 value of 0.8. UGT2B7 shows extreme regioselectivity in favoring M1 formation. Although the regioselectivity of 20

21 glucuronidation for many endogenous and exogenous compounds (Kaji and Kume, 2005a; Bowalgaha et al., 2007; Kaivosaari et al., 2008) have been reported, the underlying mechanism is still difficult to address since the crystal structures of UGTs are largely unknown (Miley et al., 2007). The CL int (M1+M2) of UGT1A9 was about 8-fold higher than that in UGT2B7 system, suggesting the predominant contribution from UGT1A9 in catalyzing the reduced TSA glucuronidation. It is well known that the substrates of human UGTs are diverse, overlapping and usually glucuronidated by multiple UGT isoforms (Tukey and Strassburg, 2000; Fisher et al., 2001; Wen et al., 2007). Thus, further correlation analysis and chemical inhibition studies are necessary and should be taken together to determine the UGT isoforms responsible (Mano et al., 2007). The correlation analysis with seven individual human liver S9 incubations showed that the M1 and M2 formation of TSA correlated well only with the propofol glucuronidation. Chemical inhibition studies also confirmed that only the typical UGT1A9 substrate MPA (Bernard and Guillemette, 2004) showed potent inhibitory effect on both M1 and M2 formation with an IC 50 value of 156±12µM and 122±10µM, respectively. These results taking together strongly suggest that UGT1A9 is the predominant isoform responsible for TSA glucuronidation, and multiple UGTs may play certain role in M1 (UGT1A1, 1A3, 1A6, 1A10, 2B7, and 2B15) and M2 formation (UGT1A1 and 1A3). Because of several recombinant UGT isoforms (UGT1A5, UGT2A1, UGT2A2, UGT2B10, UGT2B11, and UGT2B28) commercially unavailable has not been screened in the present study, we can not exclude that such isoforms may also participate in TSA glucuronidation. Human UGT1A9, expressed in liver, colon and kidney, is an active gene transcript of the UGT1A 21

22 protein and shares high amino acid homologies with UGT1A7, UGT1A8 and UGT1A10 (Tukey and Strassburg, 2000). Obvious species differences of UGT1A9 gene was found in that human UGT1A9 and mouse Ugt1a9 are functional while rat UGT1A9 is a pseudogene (Mackenzie et al., 2005; Shiratani et al., 2008). It is thus questionable which UGT isoform in rat is mainly involved in the TSA glucuronidation. UGT1A1 and 1A7 may be involved in TSA glucuronidation in rat, similar to that found for MPA (Miles et al., 2005). In view of the large species differences in producing M1 and M2, we consider that the specific UGT isoforms responsible for TSA glucuronidation may be different across various species. Such evidence obtained from this study as well as the previous reports (Mano et al., 2008; Shiratani et al., 2008) suggest that the species differences of UGTs catalyzed reactions may be much more complicated and frequent than that for CYPs. Considering that UGT1A9 exhibits a high affinity (Km, 1.1µM) and moderate capacity (Vmax, 10.4 pmol/min/mg) for TSA, it is of interest whether TSA would present competitive inhibitory effect on UGT1A9 substrates. For this consideration, we tested the potential inhibitory effect of TSA on the glucuronidation of propofol and MPA. As a result, TSA showed a strong and NAD(P)H-dependent inhibitory effect on propofol and MPA glucuronidation in vitro, with an IC 50 value at about 8µM. The NAD(P)H-dependent inhibitory effect of TSA on UGT1A9 activity suggested that such an inhibitory effect was resourced from the NQO1 catalyzed intermediate, reduced TSA, but not from TSA itself. It is noteworthy that UGT1A9 has been found involved in the metabolic elimination of many types of endogenous compounds (fatty acids) (Rowland et al., 2008) and xenobiotic compounds (simple phenols, anthraquinones/flavones, and pimary amines) 22

23 (Tukey and Strassburg, 2000). Therefore, the potential inhibitory effect of TSA against UGT1A9 should be paid close attention when such drugs are taken in combination with tanshinones-contained prescriptions. In summary, the NQO1-mediated quinone reduction and subsequent glucuronidation presents as the major metabolic pathway for TSA in human and various experimental animal species, in spite of large species differences. Although multiple UGT isoforms may be involved, UGT1A9 is the predominant isoform responsible for the formation of both regioisomers (Fig.8). TSA exhibits potent inhibitory effect against UGT1A9 and thus the potential drug-drug interactions should be of concern when tanshinones-contained preparations are prescribed in combination with the UGT1A9 substrates. In addition, the presently developed methodology would be widely applicable for assessing the sequential metabolism of quinone reduction and subsequent glucuronidation. 23

24 References Bernard O and Guillemette C (2004) The main role of UGT1A9 in the hepatic metabolism of mycophenolic acid and the effects of naturally occurring variants. Drug Metab Dispos 32: Bowalgaha K, Elliot DJ, Mackenzie PI, Knights KM and Miners JO (2007) The glucuronidation of Delta4-3-Keto C19- and C21-hydroxysteroids by human liver microsomal and recombinant UDP-glucuronosyltransferases (UGTs): 6alpha- and 21-hydroxyprogesterone are selective substrates for UGT2B7. Drug Metab Dispos 35: Dai H, Wang M, Li X, Wang L, Li Y and Xue M (2008) Structural elucidation of in vitro and in vivo metabolites of cryptotanshinone by HPLC-DAD-ESI-MS(n). J Pharm Biomed Anal 48: Fisher MB, Paine MF, Strelevitz TJ and Wrighton SA (2001) The role of hepatic and extrahepatic UDP-glucuronosyltransferases in human drug metabolism. Drug Metab Rev 33: Han JY, Fan JY, Horie Y, Miura S, Cui DH, Ishii H, Hibi T, Tsuneki H and Kimura I (2008) Ameliorating effects of compounds derived from Salvia miltiorrhiza root extract on microcirculatory disturbance and target organ injury by ischemia and reperfusion. Pharmacol Ther 117: Hao H, Wang G, Cui N, Li J, Xie L and Ding Z (2006a) Pharmacokinetics, absorption and tissue distribution of tanshinone IIA solid dispersion. Planta Med 72: Hao H, Wang G, Cui N, Li J, Xie L and Ding Z (2007) Identification of a novel intestinal first pass metabolic pathway: NQO1 mediated quinone reduction and subsequent glucuronidation. Curr Drug Metab 8:

25 Hao H, Wang G, Li P, Li J and Ding Z (2006b) Simultaneous quantification of cryptotanshinone and its active metabolite tanshinone IIA in plasma by liquid chromatography/tandem mass spectrometry (LC-MS/MS). J Pharm Biomed Anal 40: Kaivosaari S, Toivonen P, Aitio O, Sipila J, Koskinen M, Salonen JS and Finel M (2008) Regioand stereospecific N-glucuronidation of medetomidine: the differences between UDP glucuronosyltransferase (UGT) 1A4 and UGT2B10 account for the complex kinetics of human liver microsomes. Drug Metab Dispos 36: Kaji H and Kume T (2005a) Characterization of afloqualone N-glucuronidation: species differences and identification of human UDP-glucuronosyltransferase isoform(s). Drug Metab Dispos 33: Kaji H and Kume T (2005b) Regioselective glucuronidation of denopamine: marked species differences and identification of human udp-glucuronosyltransferase isoform. Drug Metab Dispos 33: Kim DH, Kim S, Jeon SJ, Son KH, Lee S, Yoon BH, Cheong JH, Ko KH and Ryu JH (2009) Tanshinone I enhances learning and memory, and ameliorates memory impairment in mice via the extracellular signal-regulated kinase signalling pathway. Br J Pharmacol 158: Li J, Wang G, Li P and Hao H (2005) Simultaneous determination of tanshinone IIA and cryptotanshinone in rat plasma by liquid chromatography-electrospray ionisation-mass spectrometry. J Chromatogr B Analyt Technol Biomed Life Sci 826: Li P, Wang GJ, Li J, Hao HP and Zheng CN (2006) Characterization of metabolites of tanshinone IIA in rats by liquid chromatography/tandem mass spectrometry. J Mass Spectrom 25

26 41: Liu J, Wu J, Wang X and Cai Z (2007) Study of the phase I and phase II metabolism of a mixture containing multiple tanshinones using liquid chromatography/tandem mass spectrometry. Rapid Commun Mass Spectrom 21: Lu Q, Zhang P, Zhang X and Chen J (2009) Experimental study of the anti-cancer mechanism of tanshinone IIA against human breast cancer. Int J Mol Med 24: Mackenzie PI, Bock KW, Burchell B, Guillemette C, Ikushiro S, Iyanagi T, Miners JO, Owens IS and Nebert DW (2005) Nomenclature update for the mammalian UDP glycosyltransferase (UGT) gene superfamily. Pharmacogenet Genomics 15: Mano Y, Usui T and Kamimura H (2007) Predominant contribution of UDP-glucuronosyltransferase 2B7 in the glucuronidation of racemic flurbiprofen in the human liver. Drug Metab Dispos 35: Mano Y, Usui T and Kamimura H (2008) Species differences in inhibition potential of nonsteroidal anti-inflammatory drugs against estradiol 3beta-glucuronidation between rats, dogs, and humans. J Pharm Sci 97: Miles KK, Stern ST, Smith PC, Kessler FK, Ali S and Ritter JK (2005) An investigation of human and rat liver microsomal mycophenolic acid glucuronidation: evidence for a principal role of UGT1A enzymes and species differences in UGT1A specificity. Drug Metab Dispos 33: Miley MJ, Zielinska AK, Keenan JE, Bratton SM, Radominska-Pandya A and Redinbo MR (2007) Crystal structure of the cofactor-binding domain of the human phase II drug-metabolism enzyme UDP-glucuronosyltransferase 2B7. J Mol Biol 369:

27 Nishiyama T, Ohnuma T, Inoue Y, Kishi T, Ogura K and Hiratsuka A (2008) UDP-glucuronosyltransferases 1A6 and 1A10 catalyze reduced menadione glucuronidation. Biochem Biophys Res Commun 371: Oppermann U (2007) Carbonyl reductases: the complex relationships of mammalian carbonyland quinone-reducing enzymes and their role in physiology. Annu Rev Pharmacol Toxicol 47: Ren ZH, Tong YH, Xu W, Ma J and Chen Y Tanshinone II A attenuates inflammatory responses of rats with myocardial infarction by reducing MCP-1 expression. Phytomedicine 17: Rouguieg K, Picard N, Sauvage FL, Gaulier JM and Marquet P Contribution of the different UDP-glucuronosyltransferase (UGT) isoforms to buprenorphine and norbuprenorphine metabolism and relationship with the main UGT polymorphisms in a bank of human liver microsomes. Drug Metab Dispos 38: Rowland A, Knights KM, Mackenzie PI and Miners JO (2008) The "albumin effect" and drug glucuronidation: bovine serum albumin and fatty acid-free human serum albumin enhance the glucuronidation of UDP-glucuronosyltransferase (UGT) 1A9 substrates but not UGT1A1 and UGT1A6 activities. Drug Metab Dispos 36: Shi SL, Li QF, Liu QR, Xu DH, Tang J, Liang Y, Zhao ZL and Yang LM (2009) Nuclear matrix protein, prohibitin, was down-regulated and translocated from nucleus to cytoplasm during the differentiation of osteosarcoma MG-63 cells induced by ginsenoside Rg1, cinnamic acid, and tanshinone IIA (RCT). J Cell Biochem 108: Shiratani H, Katoh M, Nakajima M and Yokoi T (2008) Species differences in 27

28 UDP-glucuronosyltransferase activities in mice and rats. Drug Metab Dispos 36: Sim SM, Back DJ and Breckenridge AM (1991) The effect of various drugs on the glucuronidation of zidovudine (azidothymidine; AZT) by human liver microsomes. Br J Clin Pharmacol 32: Tukey RH and Strassburg CP (2000) Human UDP-glucuronosyltransferases: metabolism, expression, and disease. Annu Rev Pharmacol Toxicol 40: Vree TB, Lagerwerf AJ, Bleeker CP and de Grood PM (1999) Direct high-performance liquid chromatography determination of propofol and its metabolite quinol with their glucuronide conjugates and preliminary pharmacokinetics in plasma and urine of man. J Chromatogr B Biomed Sci Appl 721: Wang X, Wei Y, Yuan S, Liu G, Lu Y, Zhang J and Wang W (2005) Potential anticancer activity of tanshinone IIA against human breast cancer. Int J Cancer 116: Wen Z, Martin DE, Bullock P, Lee KH and Smith PC (2007) Glucuronidation of anti-hiv drug candidate bevirimat: identification of human UDP-glucuronosyltransferases and species differences. Drug Metab Dispos 35: Wojcikowski J, Pichard-Garcia L, Maurel P and Daniel WA (2003) Contribution of human cytochrome p-450 isoforms to the metabolism of the simplest phenothiazine neuroleptic promazine. Br J Pharmacol 138: Zhang Y, Can R, Mao C, Zhu L, Shao H, Wang L, Wang A and Xu Z (2009a) The effect of tanshinone IIA on renal and liver functions in ovine fetuses in utero. Drug Chem Toxicol 32:

29 Zhang Z, Gao J, Wang Y, Song T, Zhang J, Wu G, Zhang T and Du G (2009b) Tanshinone IIA triggers p53 responses and apoptosis by RNA polymerase II upon DNA minor groove binding. Biochem Pharmacol 78: Zhou L, Zuo Z and Chow MS (2005) Danshen: an overview of its chemistry, pharmacology, pharmacokinetics, and clinical use. J Clin Pharmacol 45:

30 Footnotes Q.W. and H.H. contributed equally to this work. This study was financially supported by Natural Science Foundation of Jiangsu province (BK , BK ); Science Foundation for Youth Scholars of Ministry of Education of China ( ), a Foundation for the Author of National Excellent Doctoral Dissertation of PR China (200979); National Natural Science Foundation of P.R. China (No , ) and National Key New Drug Creation Special Programme (2009ZX , 2009ZX ). 30

31 Legends for figures: Fig.1 Representative HPLC chromatogram for TSA metabolites profiling in the liver and intestinal S9 incubation systems of different species. The retention time of reduced TSA glucuronidation metabolites M1 and M2 was 12.1 and 12.6 min, respectively. Fig.2 Representative Lineweaver-Burk plots and Eadie-Hofstee plots for the formation of tanshinone glucuronides (M1 and M2). TSA ( µM) was incubated in various pooled liver S9 (A:MLS9; B:RLS9; C: DLS9; D: HLS9) and pooled intestinal S9 (E:MIS9; F:RIS9) systems. The optimal incubation time and protein concentrations were A (20min, 0.5mg/ml); B(20min, 0.5mg/ml); C(30min, 1mg/ml); D(30min, 1mg/ml); E(30min, 1mg/ml); F(30min, 2mg/ml), respectively. Fig.3 Formation of tanshinone IIA glucuronides M1 and M2 by recombinant human UGT isoforms. 2µM TSA (A) and 20µM TSA (B) were incubated with recombinant human UGT isoforms UGT 1A1, 1A3, 1A4, 1A6, 1A7, 1A8, 1A9, 1A10, 2B4, 2B7, 2B15, and 2B17 at the manufacturer s recommended protein concentrations and incubation time, using pooled human liver cytosol as a NQO1 donor as described in materials and methods. Open and closed column represents TSA glucuronides M1 and M2, respectively. Each bar is the means±sd of triplicate determinations. Fig.4 Representative Lineweaver-Burk plots and Eadie-Hofstee plots for the formation of tanshinone glucuronides (M1 and M2) by incubated TSA ( µM) with the recombinant 31

32 human UGT1A9 (A) and UGT2B7 (B) in combination with pooled human liver cytosol as a NQO1 donor. Fig.5 Correlation analysis between TSA glucuronidation in 7 individual human liver microsomes in combination with pooled human liver cytosol as a NQO1 donor and the typical UGT1A1 (Estradiol, A and B), 1A4 (Trifluoperazine, C and D), 1A9 (propofol, E and F) and 2B7(AZT, G) substrates. A p value less than 0.05 was considered statistically significant. Fig.6 Inhibitory effects of typical substrates on TSA glucuronidation in pooled human liver S9. Bilirubin (0-500µM), 4-MU (0-500µM), MPA (0-500µM) and morphine (0-500µM) were used as inhibitors. The protein concentration was 1mg/ml protein and the incubation time was 30min. IC 50 values representing the 50% inhibition concentration were calculated graphically and each data point represents the means of duplicate incubations. Fig.7 Inhibitory effects of TSA (0-40µM) on propofol and MPA glucuronidation in pooled human liver S9. The propofol and MPA concentrations were 50µM and 100µM, respectively. Incubations were performed with (MPA, propofol ) or without NAD(P)H (MPA, propofol ) as the co-factor for initiating the NQO1 catalyzed TSA reduction. IC 50 values representing the 50% inhibition concentration were calculated graphically and each data point represents the means of duplicate incubations. Fig.8 NQO1 catalyzed quinone reduction and subsequent UGT1A9 (predominantly) mediated 32

33 glucuronidation of TSA in human liver S9 system. 33

34 Table 1 Kinetic parameters of TSA glucuronides M1 and M2 formation in pooled liver and intestinal S9. Km Vmax CL int CL int CL int M2/M1 (µmol/l) (pmol/min/mg) (µl/min/mg) (M1+M2) M1 M2 M1 M2 M1 M2 (µl/min/mg) ML S9 6.5± ± ± ± ± ±0.1* # 16.2± RL S9 8.6± ± ± ± ± ±0.2 # 15.3± DL S9 5.6±0.2 N.D. 21.2±0.2 N.D. 3.8±0.2 N.D. 3.8±0.2 - HL S9 4.1± ± ± ± ± ±0.1* 11.8± MI S9 9.2± ± ± ± ± ±0.1* 9.2± RI S9 19.6± ± ± ± ± ± ± *p<0.05, significantly difference from the CLint M1. # p<0.05, significantly difference from the HLS9 CLint M2. 34

35

36

37

38

39

40

41

42

43 .aspetjournals.org at ASPET Journals on November 29, 2018