Simultaneous determination of bilirubin and its glucuronides in liver microsomes and recombinant UGT1A1 enzyme incubation systems by HPLC method and its application to bilirubin glucuronidation studies

Journal of Pharmaceutical and Biomedical Analysis 92 (2014) 149–159

Contents lists available at ScienceDirect

Journal of Pharmaceutical and Biomedical Analysis journal homepage: www.elsevier.com/locate/jpba

Simultaneous determination of bilirubin and its glucuronides in liver microsomes and recombinant UGT1A1 enzyme incubation systems by HPLC method and its application to bilirubin glucuronidation studies Guo Ma ∗,1 , Jiayuan Lin 1 , Weimin Cai, Bo Tan, Xiaoqiang Xiang, Ying Zhang, Peng Zhang ∗∗ Department of Clinical Pharmacy, School of Pharmacy, Fudan University, 826 Zhangheng Road, Shanghai 201203, PR China

a r t i c l e

i n f o

Article history: Received 8 August 2013 Received in revised form 17 January 2014 Accepted 18 January 2014 Available online 27 January 2014 Keywords: Bilirubin Glucuronides UGT1A1 Liver microsomes HPLC

a b s t r a c t Bilirubin, an important endogenous substances and liver function index in humans, is primarily eliminated via UGT1A1-catalyzed glucuronidation. Instability of bilirubin and its glucuronides brings substantial technical challenges to conduct in vitro bilirubin glucuronidation assay. In the present study, we developed a simple and robust HPLC method for simultaneous determination of unconjugated bilirubin (UCB) and its multiple glucuronides, i.e. bilirubin monoglucuronides (BMGs, including BMG1 and BMG2 isomers) and diglucuronide (BDG) in rat liver microsomes (RLM), human liver microsomes (HLM) and recombinant human UGT1A1 enzyme (UGT1A1) incubation systems, and applied it to study in vitro bilirubin glucuronidation. UCB, BMG1, BMG2, BDG and their isomers in the incubation mixtures were successfully separated using a C18 column with UV detection at 450 nm and mobile phase consisted of 0.1% formic acid in water and acetonitrile by a linear gradient elution program. Assay linearities of bilirubin were confirmed in the range 0.01–2 ␮M. Precision of UCB, BMG1, BMG2 and BDG (n = 5) at low, medium and high concentration was within the range of RSD 0.4–3.7%, accuracy expressed in the mean assay recoveries of them (n = 5) ranged from 92.8 ± 1.5% to 104.3 ± 2.2% for intra- and inter-day assays and the mean extraction recoveries of them (n = 5) were above 91.5 ± 1.0%. Stability of bilirubin and its glucuronides was satisfactory at 37 ◦ C in the incubation solutions during the reaction (30 min), 25 ◦ C for 24 h and −70 ◦ C for 7 d in the processed incubation samples with methanol. Furthermore, we established stable, reliable in vitro incubation systems and optimized the incubation conditions to characterize the kinetics of bilirubin glucuronidation by RLM, HLM and UGT1A1, respectively. The kinetic parameters of formation of total bilirubin glucuronides (TBG, the sum of BMG1, BMG2 and BDG) were as follows: Km of 0.45 ± 0.016, 0.40 ± 0.022, 0.44 ± 0.018 ␮M, Vmax of 2.65 ± 0.057, 1.86 ± 0.029, 2.95 ± 0.036 nmol/mg/min, CLint of 5.92 ± 0.22, 4.70 ± 0.079, 6.72 ± 0.27 mL/mg/min by RLM, HLM and UGT1A1, respectively. Bilirubin glucuronidation obeyed the Hill equation by RLM and the Michaelis–Menten equation by HLM and UGT1A1 in the range of substrate concentration selected, respectively. In addition, the relative proportions between BDG and BMGs were in connection with enzyme sources (e.g. RLM, HLM and UGT1A1) and bilirubin concentration. © 2014 Elsevier B.V. All rights reserved.

Abbreviations: UCB, unconjugated bilirubin; CB, conjugated bilirubin; BG, bilirubin glucuronides; BMGs, bilirubin monoglucuronides; BDG, bilirubin diglucuronide; TBG, total bilirubin glucuronides; UGT(s), UDP-glucuronosyltransferase(s); UGT1A1, UDP-glucuronosyltransferases1A1; RLM, rat liver microsomes; HLM, human liver microsomes; UDPGA, uridine diphosphoglucuronic acid; DMSO, dimethylsulfoxide; MRP2, multidrug resistance-associated protein 2; HPLC, high performance liquid chromatography; QC, quality control; LOD, limit of detection; LLOQ, lower limit of quantification; RSD, relative standard deviation; Conc., concentration(s); V, reaction velocity; Km , Michaelis–Menten constant; Vmax , maximum reaction velocity; CLint , intrinsic clearances; R2 , residual sum of squares; AIC, Akaike information criterion; CDER, Center for Drug Evaluation and Research. ∗ Corresponding author. Tel.: +86 21 51980025; fax: +86 21 51980001. ∗∗ Corresponding author. Tel.: +86 21 51980024; fax: +86 21 51980001. E-mail addresses: [email protected], [email protected] (G. Ma), [email protected] (P. Zhang). 1 Co-first authors. 0731-7085/$ – see front matter © 2014 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.jpba.2014.01.025

150

G. Ma et al. / Journal of Pharmaceutical and Biomedical Analysis 92 (2014) 149–159

1. Introduction

kernicterus, Crigler–Najjar syndromes (Types I and II), Gilbert’s syndrome, and even death [1,19–21]. As a result, in drug discovery, development and use settings, the in vitro ability of drug to inhibit bilirubin glucuronidation is commonly evaluated. In addition, some xenobiotics (e.g. phenobarbital, dexamethasone, rifampicin and herbal extracts Yin zhi huang) can also induce UGT1A1 gene expression and enhance UGT1A1 activity by a number of multifunctional nuclear receptors such as constitutive androstane receptor (CAR), pregnane X receptor (PXR), glucocorticoid receptor (GR), aryl hydrocarbon receptor (AhR), and hepatocyte nuclear receptor 1␣ (HNF1␣) [22–25]. These factors contribute to promote bilirubin glucuronidation, and reduce serum UCB level. They might have important clinical application in preventing and treating unconjugated hyperbilirubinemia and neonatal jaundice. It is not difficult to see that establishing a simple and robust assay method for accurate measurement of bilirubin and its glucuronides is vitally important for us to study bilirubin glucuronidation and its inhibition or induction, which has important clinical significance in diagnosis, prevention and treatment of bilirubin-related malady or toxic reaction, for example, jaundice, hyperbilirubinemia and kernicterus. However, as a weakly polar, poorly soluble compound, bilirubin is very labile. It is highly photosensitive and readily oxidized, rapidly degraded in both acidic and alkaline solutions, and high-affinity for proteins (e.g. serum albumin), as well as strong adsorption on experimental equipments and materials (e.g. nonspecific binding of bilirubin to walls of the plastic pipes, tips, vials and tubes, as well as the chromatographic channel and column) [19,26,27]. Equally as problematic is the instability of bilirubin glucuronides, especially BMGs. In aqueous media, BMGs was rapidly transformed into BDG and UCB by dipyrrole exchange mechanism [28]. Furthermore, bilirubin itself is composed of three isomers (i.e. bilirubin IX-␣, XIII-␣ and III-␣), and bilirubin glucuronidation involves a sequential reaction that produces multiple glucuronides (i.e. BMG1, BMG2, BDG and their isomers), resulting in difficult quantitation of glucuronidation assay and establishment of initial rate condition, if not given particular attention. All these factors, especially, in vitro instability, bring the substantial technical challenges in bilirubin glucuronidation [28,29]. These challenges have been manifested in significant disparities in estimated kinetic parameters and mechanism for bilirubin glucuronidation. Three groups [26,30,31] reported that bilirubin glucuronidation obeyed Michaelis–Menten kinetics. One group [32] reported it exhibited substrate inhibition kinetics, and the other group [33] reported it obeyed Michaelis–Menten kinetics at low protein concentration

Bilirubin is the principal constituent of mammalian bile pigment and end-product of heme catabolism. Approximately 250–300 mg of bilirubin is produced in a normal adult each day. Bilirubin is an important index of liver function and biomarker of hepatotoxicity, as well as an important clinical basis for determining jaundice. As an essential endogenous substance in humans and animals, bilirubin was long thought to be a non-functional and toxic waste product. Recent studies [1–3] have shown that bilirubin has multiple biological functions in animals and plants, for example, potent antioxidant and cytoprotective effects at physiological and mildly elevated concentrations, as well as activation of heme oxygenase, and can protect against cardiovascular diseases (e.g. atherosclerosis) and tumor development. However, it can cause apoptosis, cytotoxicity and neurotoxicity at markedly elevated plasma and tissue bilirubin levels, and result in severe, irreversible brain and neurological damage (e.g. kernicterus), especially in neonates [1,4–7]. Bilirubin is mainly metabolized by liver. Before it is transported into liver, bilirubin exists mostly in the form of unconjugated bilirubin (UCB) and binds highly to albumin in the blood. After hepatic uptake, UCB is extensively metabolized to bilirubin glucuronides (BG) by UDP-glucuronosyltransferases1A1 (UGT1A1) localized primarily in smooth endoplasmic reticulum of hepatocyte. In this glucuronidation reaction, a glucuronosyl moiety is conjugated to one of the propionic acid side chains, located on the C8 and C12 carbons of the two central pyrrole rings of bilirubin, resulting in producing two bilirubin monoglucuronides (BMGs) isomers (i.e. BMG1 and BMG2). BMGs were further glucuronidated, and formed bilirubin 8,12-diglucuronide (BDG) [8] (Fig. 1). In adult humans, over 80% of the bilirubin conjugates are normally BDG [9], whereas BMGs predominate in newborns [10]. Finally, BG (i.e. BMGs and BDG) formed are secreted into bile by multidrug resistance-associated protein 2 (MRP2), and subsequently eliminated via feces and urine [11]. UGT1A1 is a critical enzyme responsible for metabolism and detoxification of bilurubin [12]. Glucuronidation by UGT1A1 is an essential step for bilirubin elimination [13]. Xenobiotics (e.g. SN-38 [14], atazanavir, indinavir [15,16], erlotinib [17], sorafenib [18]) inhibiting UGT1A1, and genetic variants resulting in partial or complete loss of UGT1A1 activity, can cause disorder of bilirubin metabolism, and lead to accumulation of bilirubin in blood and/or brain, which further result in jaundice, hyperbilirubinemia, H2C H3C

UGT1A1

CH3

OH HO N

N

N

HN

CH2

UGT1A1

H3C

H2C H3C

HOOC

CH3

OH HO N

N

HN

CO2H

O O

N

H2C

CH3

O

CH2

OH

OH

HO

BMGs

H3C

CH3

HOOC

CH3

OH HO

H3C

UCB

N

N

N

HN

HO

N

HN

CH2

CH3

O OH

O O

O O

O

HO

CO2H

OH

OH

BDG

CH2

CH3 COOH

O OH

N

UGT1A1 H3C

HO2C

HO2C

HO

UGT1A1

HO

N

H3C HO

H2C

COOH

CH3

OH HO

H3C

O O

Fig. 1. The molecular structures of bilirubin and its glucuronides. UCB was metabolized to BMGs (including two isomers BMG1 and BMG2), and BMGs was further metabolized to BDG by UGT1A1. UCB, unconjugated bilirubin; BMGs, bilirubin monoglucuronides; BDG, bilirubin diglucuronide; UGT1A1, recombinant human UGT1A1 enzyme.

G. Ma et al. / Journal of Pharmaceutical and Biomedical Analysis 92 (2014) 149–159

of 0.05 mg/mL and Hill equation at high protein concentration of 0.5 mg/mL. Likewise, estimates of Km and Vmax ranged from 0.20–24 ␮M and 0.08–1.08 nmol/min/mg, respectively. These differences may be caused by different incubation condition selected (e.g. source and concentration of enzyme, concentration range of substrate, reaction time), and some influential factors (e.g. light, heat, oxygen, pH, protein binding, physical adsorption and other factors) in assaying of bilirubin and its glucuronides. In order to carefully characterize the kinetics of bilirubin glucuronidation, we develop a specific, sensitive and robust HPLC method for simultaneous determination of bilirubin and its glucuronides in rat liver microsomes (RLM), human liver microsomes (HLM) and recombinant human UGT1A1 enzyme (UGT1A1) incubation systems, and established and optimized the in vitro incubation conditions in the present study, respectively. Especially, compared with the previous methods [28,29,33–39], we simultaneously determined bilirubin, bilirubin glucuronides and their multiple isomers in three incubation matrix, and disclosed the differences of kinetic mechanism of bilirubin glucuronidation by RLM, HLM and UGT1A1. The in vitro study will provide an important reference for in vivo bilirubin metabolism, and has the potential application in diagnosis, prevention and treatment of bilirubin-related malady or toxic reaction. 2. Materials and methods 2.1. Chemicals and reagents Bilirubin (including three mixed isomers, i.e. bilirubin IX-␣ 90.11%, XIII-␣ 3.12% and III-␣ 5.93%), uridine 5 diphosphoglucuronic acid trisodium salt (UDPGA) and alamethicin were purchased from Sigma–Aldrich (China-mainland). Ascorbic acid was provided by Aladdin Chemistry Co., Ltd. (Shanghai, China). Formic acid, MgCl2 ·6H2 O, K2 HPO4 ·3H2 O, KH2 PO4 , NaH2 PO4 ·2H2 O, Na2 HPO4 ·12H2 O, NaCl (all of analytical grade) were purchased from Sinopharm Chemical Reagent Co., Ltd. (Shanghai, China). Acetonitrile and methanol (both of HPLC grade) were provided by Fisher Scientific International, Inc. (Fair Lawn, NJ, USA). Argon was purchased from Shanghai Lvmin Gas Co. Ltd. Purified water was prepared in a water purification system (EMD Millipore Corp., Billerica, MA, USA). All other reagents were of analytical grade at least. Pooled SD rat liver microsomes (RLM) were prepared in our laboratory. Pooled human liver microsomes (HLM) were purchased from CELSIS, Inc. (Chicago, IL, USA). Recombinant Human UGT1A1 enzyme (UGT1A1, BD-SupersomesTM ) was purchased from BD Biosciences-Discovery Labware (Woburn, MA, USA). 2.2. Chromatographic conditions Chromatographic analyses were performed on a Shimadzu LC2010A HT HPLC system (Kyoto, Japan) equipped with a quaternary pump, an automatic sampler, a UV–vis detector, a system controller and a temperature control oven. System control and data analyses were carried out using a Shimadzu LC solution workstation (Shimadzu, Kyoto, Japan). Bilirubin and its glucuronides were separated on a HPLC column (reverse phase DiamonsilTM C18 column, 200 mm × 4.6 mm, i.d., 5 ␮m particle size, Dikma) with guard column (Cartridge Guard Column E, Inertsil ODS-SP, 10 mm × 4 mm, GL Sciences Inc.). The mobile phase consisted of 0.1% formic acid in water (A) and 100% acetonitrile (B) was delivered at a flow rate of 1 mL/min. The linear gradient elution program was as follows: 0–9 min, 40–75% B; 9–18 min, 75–95% B; 18–27 min, 95% B; 27–30 min, 95–40% B. The column temperature was 45 ◦ C. The detection wavelength was 450 nm. The sample injection volume was 100 ␮L.

151

2.3. Preparation of rat liver microsomes Six Sprague-Dawley rats (male and female, 200 ± 20 g) were provided by SIPPR-BK Laboratory Animal Co., Ltd. (Shanghai, China, Animal study protocol number: 2008001626534). All animal studies were performed according to the requirement of the National Act on the Use of Experimental Animal (China) that was approved by the Ethics Committee for Animal Experiment of School of Pharmacy, Fudan University in Shanghai. RLM were prepared as previously reported [40] with slight modifications. The SD rats were fasted for 12 h before the animal test was conducted. They were sacrificed by decapitation, and were rapidly perfused with ice-cold 0.2 mol/L PBS (pH 7.4) via the portal vein to flush the liver. The livers were removed and placed in the ice-cold PBS at once, washed away the redundant blood and blotted the moisture. After that, the tissue was minced into the slices. The slices were weighed and added the PBS (four times of the weight of the tissue), homogenated with GF-1 dispersator (Kylin-Bell, Haimen, Jiangsu, China). The homogenate was centrifuged in a MICROCL 17R centrifuge (Thermo scientific, Boston, MA, USA) at 9000 × g for 20 min at 4 ◦ C and the supernatant was ultracentrifuged employing a CP-WX ultracentrifuge (TECHCOMP Ltd., Shanghai, China) at 100,000 × g for 1 h at 4 ◦ C in order to obtain the microsomal pellet. The obtained pellet was re-suspended in 30% glycerol–0.2 mol/L PBS (pH 7.4) and stored at −70 ◦ C until use. The protein concentration was determined by a BCA kit (Cwbiotech, Shanghai, China). 2.4. Preparation of samples 2.4.1. Bilirubin stock solution Bilirubin stock solution was prepared by dissolving bilirubin in 100% dimethyl sulfoxide (DMSO) to yield concentration of 2 mM, then rapidly aliquoted and stored at −70 ◦ C. 2.4.2. Standard and quality control (QC) samples The standard and QC samples (n = 5) were prepared as described in Section 2.6. As the incubation mixtures, they contained bilirubin (final concentration of 0.05–2 ␮M, dissolved in 100% DMSO), RLM (HLM or UGT1A1, final concentration of 12.5 ␮g of protein/mL), potassium phosphate buffer (50 mM, pH 7.4), MgCl2 ·6H2 O (0.88 mM), alamethicin (22 ␮g/mL), as well as in the absence or presence of UDPGA (3.5 mM). The standard samples were the simulated incubation mixtures in the absence of UDPGA. The QC samples representing the initial low, medium and high concentrations of bilirubin were set at 0.05, 0.2, 1.5 ␮M for assessing the precision, accuracy, recovery and stability of bilirubin standard solution, and 0.2, 0.75, 1.5 ␮M for assessing the stability of bilirubin and its glucurconides in the incubation solutions during the reaction and the processed incubation samples (the latter are deproteinized and extracted by addition of methanol containing ascorbic acid). Because of lack of commercial products of BMG1, BMG2 and BDG, the samples for stability of glucoronides have to be prepared by bilirubin glucuronidation reaction as described in Section 2.6 with slight modifications. Namely, the reaction was terminated by the addition of ice-cold methanol (a quarter volume of the reaction mixture, in the absence of ascorbic acid). The incubation mixtures were immediately freeze-dried using Liquid Nitrogen Vacuum Freeze Dryer (Tofflon, Shanghai, China) which need not to been deproteinized. The freeze-dried powders were re-dissolved with water, then were used to evaluate the stability of bilirubin and its glucurconides at 37 ◦ C. Finally, the samples were processed as described in Section 2.4.3. The processed incubation samples with methanol for stability at 25 ◦ C and −70 ◦ C were prepared as described in Section 2.6. All these samples were prepared in amber glass vials with screw cap, and processed in a dim light room.

152

G. Ma et al. / Journal of Pharmaceutical and Biomedical Analysis 92 (2014) 149–159

2.4.3. Extraction procedure A 600 ␮L ice-cold methanol containing 200 mM ascorbic acid was added to the above standard or QC samples (200 ␮L), respectively. The mixtures were vortexed for 2 min and then centrifuged at 12,000 rpm for 10 min to precipitate and separate protein. The supernatant was injected into the HPLC system for analysis. 2.5. HPLC validation procedures 2.5.1. Selectivity Selectivity of the method was evaluated by analyzing the blank samples (the incubation samples in the absence of bilirubin, i.e. the incubation matrix described in Section 2.4.2), standard samples (the incubation samples in the absence of UDPGA) and bilirubin glucuronidation samples (the incubation samples in the presence of bilirubin and UDPGA) from the RLM, HLM and UGT1A1 incubation systems, respectively. The samples were prepared as described in Section 2.4. 2.5.2. Calibration curves and linearity The standard samples for calibration curves (n = 3) were prepared as described in Section 2.4. The final concentrations of bilirubin in the standard samples were 0.01, 0.05, 0.1, 0.2, 0.5, 1, 1.5 and 2 ␮M, respectively. The combined peak areas of bilirubin (i.e. the sum of peak areas of bilirubin IX␣, XIII␣ and III␣ isomers) were plotted against the standard concentrations to establish the calibration curves. Quantitation of BMG1, BMG2 and BDG was based on the standard curve of bilirubin as the previous reports [32,33]. Likewise, peak areas of these glucuronides for quantitation were also the sum of peak areas of their isomers in the study, respectively. Limit of detection (LOD) was calculated as the final concentration of bilirubin producing a signal-to-noise ratio of 3. The lower limit of quantification (LLOQ) was considered as the lowest concentration of the calibration curve. 2.5.3. Precision, accuracy and recovery Precision and accuracy of the analytical method were evaluated by analyzing QC samples at three initial concentration levels (0.05, 0.2 and 1.5 ␮M bilirubin). Precision was expressed using relative standard deviation (%, RSD), and accuracy was defined as percent of deviation between the true and the measured value, which both required to be measured using five determinations (n = 5) per concentration. To assess intra-day precision and accuracy, the QC samples were measured within one day. For inter-day assays, QC samples were analyzed for three consecutive days. The precision was required within 15% of the RSD, and accuracies were required not to exceed ±15% of the actual value at three concentration levels [41]. In addition, within-run and inter-run precision for BMG1, BMG2 and BDG from the same batch of RLM, HLM and UGT1A1 incubation samples at three initial bilirubin concentration levels of 0.2, 0.75 and 1.5 ␮M was only assessed owing to in the absence of commercially supplied BMG1, BMG2 and BDG as reference substance hereon. Extraction recovery of bilirubin was determined by comparing the chromatographic peak areas of the analytes extracted from five replicate QC samples at three concentration levels (0.05, 0.2, 1.5 ␮M bilirubin) to that of the pure bilirubin solutions without extraction procedure at the same nominal concentrations. The pure bilirubin solutions were prepared by dissolving bilirubin in the mixed solvent DMSO–methanol (1:4, v/v). 2.5.4. Stability The stability was thoroughly evaluated by analyzing bilirubin stock solutions, standard solutions and QC samples exposed to different conditions. The bilirubin stock solution (2 mM) was stored at 25 ◦ C (room temperature) for 0, 2, 4, and 6 h (short-term stability)

and −70 ◦ C (frozen temperature) for 0, 10, 20, and 30 d (longterm stability), respectively. The bilirubin standard solutions (in the absence of UDPGA) at the concentration level of 0.05, 0.2 and 1.5 ␮M were prepared as described in Section 2.5.2 and stored at 25 ◦ C for 0, 6, 12, and 24 h. The stability of UCB, BMG1, BMG2 and BDG was evaluated at 37 ◦ C (incubation temperature, assessing the short-term stability of bilirubin and its glucurconides in the incubation solutions during the reaction) for 0, 5,10, 15, 30, and 60 min, 25 ◦ C (assessing the post-preparative stability of the processed samples with methanol) for 0, 4, 8, 12, 24 h, and −70 ◦ C (assessing the long-term stability of the processed samples with methanol) for 0, 3, 7 d from the RLM, HLM and UGT1A1 incubation samples (n = 3), respectively. The samples for bilirubin glucuronidation during and after the reaction were prepared as described in Section 2.4.2. The final concentrations of bilirubin in these incubation mixtures were 0.2, 0.75 and 1.5 ␮M at the initial reaction time point. All these samples were processed as described in Section 2.4.3. The peak areas of UCB, BMG1, BMG2 and BDG for stability of the test samples at different time points were compared with the peak area of that at 0 min. All the stability determinations use a set of samples prepared from freshly made stock solution of bilirubin. 2.6. Bilirubin glucuronidation The incubation procedure for bilirubin glucuronidation were as follows: (1) bilirubin (final concentration range of 0.25–10 ␮M), potassium phosphate buffer (50 mM, pH 7.4), MgCl2 ·6H2 O (0.88 mM), alamethicin (22 ␮g/mL) and RLM (HLM or UGT1A1, final concentration 12.5–50 ␮g of protein/mL) were mixed in amber glass vials (full of argon) and pre-incubated at 37 ◦ C for 2 min in a shaking water bath; (2) the reaction was initiated by the addition of UDPGA (3.5 mM); (3) the mixture (total volume 200 ␮L) was incubated at 37 ◦ C for 0–60 min; and (4) the reaction was terminated by the addition of 600 ␮L ice-cold methanol containing 200 mM ascorbic acid. Bilirubin and its glucuronides in the incubation mixture were extracted as described in Section 2.4.3. Bilirubin was dissolved in DMSO just before adding into the incubation mixtures. The final DMSO concentration in the incubation mixture was 1%. The samples were stored in amber glass vials, handled and processed under the dim light. All experiments were performed in triplicates. 2.7. Kinetic analysis Kinetic analysis was performed by fitting the Michaelis–Menten equation (Eq. (1)) or the Hill equation (Eq. (2)) to the kinetic data (substrate concentrations and initial rates) with SigmaPlot 12.0 (Systat Software Inc., San Jose, CA). Glucuronidation velocity (V) in Eqs. (1) and (2) was calculated as nanomoles of glucuronide(s) formed per mg protein amount per reaction time (nmol/mg protein/min). The kinetic parameters Vmax and Km (also depicted as S50 in Eq. (2)) are defined as the maximum velocity and the substrate concentration at which velocity equals to half of the Vmax , respectively. n in Eq. (2) is the Hill coefficient, indicative of the degree of curve sigmoidicity and/or cooperativity. The kinetic parameters CLint (intrinsic clearance, =Vmax /Km ) was calculated as the rate of disappearance of the test compound in units of mL/mg of protein/min [42]. Model appropriateness was determined by visual inspection of the Eadie–Hofstee plots, comparison of the residual sum of squares (R2 ) and Akaike information criterion (AIC) values [43,44]. V= V=

Vmax × [S] Km + [S] Vmax × [S]n n + [S]n Km

(1)

(2)

G. Ma et al. / Journal of Pharmaceutical and Biomedical Analysis 92 (2014) 149–159

2.8. Statistical analysis Statistical analyses were performed by one-way ANONA, twoway ANONA and T-test using GraphPad Prism V5 software for Windows (GraphPad Software, San Diego, CA). Differences were considered significant when p values were less than 0.05 (p < 0.05).

6000

uV

RLM

5000 4000

3.1. Analytical methods In the present study, we simultaneously determined bilirubin and its multiple glucuronides in the RLM, HLM and UGT1A1 incubation systems by an HPLC method, and carried out the analytical method validation including selectivity, linearity, sensitivity, accuracy, precision and stability. Although a number of assay methods for bilirubin and/or its glucuronides in some matrix such as serum/plasma, bile, or cell have been established for the past decades [28,29,33–39], different approaches are needed for the determination of bilirubin in different species and matrix. The conventional analytical approaches, for example, diazo assay and direct spectrophotometry were usually employed to determine concentrations of free and conjugated bilirubin in biological fluids and tissues (e.g. brain or blood). The former was based on the reaction of diazotized sulfanilic acid with bilirubin to form azobilirubin, which formed the quantitative basis of bilirubin in biological fluids. The latter was based on measuring the absorbance of bilirubin near 460 nm [45]. Compared with the newly established HPLC method, the two methods exhibited poor selectivity. It was difficult for them to simultaneously separate and determine UCB, BMG1, BMG2, BDG and their multiple isomers. Moreover, radioassay was used to assess bilirubin levels in brain and CSF after intravenous administration of [14 C]-UCB to Gunn rats or guinea pigs [46,47]. ELISA using an anti-bilirubin antibody was also employed to assay bilirubin and its oxidation in CSF of Alzheimer’s disease patients and in the rat intestinal mucosa [48–50]. It is a pity that the last two methods are not generally accessible due to the commercial unavailability of radiolabeled bilirubin or anti-bilirubin antibody, and, more importantly, underestimate UCB concentrations due to incomplete extraction of the analytes from tissues and organs. In previous assay [38], bilirubin glucuronides and total bilirubin in biological fluids and tissues were usually determined by a complex hydrolysis step, i.e. hydrolytic reagents (e.g. NaOH) were added into the samples consisted of bilirubin glucuronides, and then hydrolyzed these glucuronides into UCB. The procedure involved extraction, separation, hydrolysis and neutralization of samples, easily resulting in loss of the analytes and inaccurate quantification. Furthermore, the indirect method cannot separate and determine the specific constituents of bilirubin glucuronides (i.e. BMG1, BMG2, BDG and their isomers). In a word, we simultaneously determined bilirubin and its multiple glucuronides (including their multiple isomers) in three different incubation systems by a simple, reliable and reproducible HPLC method, and first applied it to systematically investigate and disclose the differences of kinetics of bilirubin glucuronidation by RLM, HLM and UGT1A1 in the present study. In addition, compared with the previous assay [28,29,33–39], the processed procedure of samples (e.g. preparation and extraction) was simple, rapid and reproducible. 3.2. Selectivity The representative chromatograms for bilirubin glucuronidation by RLM, HLM and UGT1A1 are similar (Fig. 2). Ten peaks from bilirubin and its glucuronides including their isomers were detected in the incubation samples. Peak assignment and

UCB

BMG2

3000

BMG1

BDG

2000

3. Results and discussion

153

C1 B1 A1

1000 0 0.0 uV

5.0

10.0

15.0

20.0

25.0

min

HLM

6000 5000 4000

BMG2

3000

UCB BMG1

BDG

2000

C2 B2

1000

A2

0

6000

0.0 uV

5.0

10.0

15.0

20.0

25.0

min

UGT1A1

5000 4000

BMG2

3000

BDG

UCB

BMG1

2000

C3 B3

1000

A3

0 0.0

5.0

10.0

15.0

20.0

25.0

min

Fig. 2. Representative chromatograms for bilirubin glucuronidation in RLM, HLM, UGT1A1 incubation systems, respectively. A1–A3 represent blank samples (the incubation samples in the absence of bilirubin); B1–B3, standard samples (the incubation samples in the absence of UDPGA); C1–C3, bilirubin glucuronidation samples (the incubation samples in the presence of bilirubin and UDPGA) in the RLM, HLM, UGT1A1 incubation systems, respectively. RLM, rat liver microsomes; HLM, human liver microsomes; UGT1A1, recombinant human UGT1A1 enzyme. The peaks at 24.714, 25.279 and 25.761 min were assigned as bilirubin III-␣, IX-␣ (major peak) and XIII-␣, respectively. The peaks at 8.564 and 8.940 min were assigned as the BMG1 IX-␣ (major peak) and XIII-␣, respectively. The peaks at 7.571 and 7.950 min were assigned as the BMG2 III-␣ and IX-␣ (major peak), respectively. The peaks at 4.515, 4.914 and 5.346 min were assigned as the BDG III-␣, IX-␣ (major peak) and XIII-␣, respectively. Incubations were conducted with 1 ␮M bilirubin at 12.5 ␮g/mL microsomal (or UGT1A1) protein concentration for 15 min; chromatographic conditions were described in Section 2.2.

identification of UCB, BMG1, BMG2, BDG and their isomers were based on their lipophilicity and polarity, as well as the elution pattern, chromatographic peak position, relative retention time from previous reports [26,28,32,33] and our current study. As shown in Fig. 2, UCB (including UCB III-␣, IX-␣ and XIII-␣) was metabolized to BMG1 (including BMG1 IX-␣ and XIII-␣), BMG2 (including BMG2 III-␣ and IX-␣) and BDG (including BDG III-␣, IX-␣ and XIII-␣). All these analytes including their multiple isomers were efficiently separated on the HPLC column. No interference was observed at the retention times of each analyte in any incubation samples used for analysis. Although the monoglucuronides BMG1 and BMG2 were efficiently separated, it was difficult to use the techniques described here to assign with certainty the propionic acid side-chain (C8 or C12 ) position for the BMGs peaks. In a word, the HPLC method exhibited good selectivity and high resolution.

154

G. Ma et al. / Journal of Pharmaceutical and Biomedical Analysis 92 (2014) 149–159

Table 1 Regression equations, LOD, LLOQ and linear range for bilirubin. Incubation systems

Regression equations

r

LOD (␮M)

LLOQ (␮M)

RLM HLM UGT1A1

Y = 72186X − 589.32 Y = 70403X − 276.16 Y = 70182X + 283.75

0.9996 0.9998 0.9998

0.005 0.005 0.005

0.01 0.01 0.01

3.3. Calibration curve and linearity The regression equations for calibration curves, LOD, LLOQ and linear range for bilirubin were shown in Table 1, respectively. The HPLC method showed good linearities in the range of 0.01–2 ␮M bilirubin with correlation coefficients (r) ≥ 0.9996 for three calibration curves. The linearities became poor when concentration of bilirubin was beyond 2 ␮M. Hereon, it was possibly due to strong adsorption and overload (or saturation) of bilirubin on the chromatographic column. The LLOQ for bilirubin was set at the lowest concentration in the linear standard curve and equal to 0.01 ␮M, and the accuracies for bilirubin were 105.9 ± 5.1%, 93.4 ± 3.8% and 92.9 ± 3.6% while the precisions (RSD, %) were 4.8%, 4.0% and 3.9% in the RLM, HLM and UGT1A1 incubation system, respectively. It must be pointed out that, bilirubin glucuronides (i.e. BMG1, BMG2 and BDG) are extremely unstable and in the absence of commercial products. Because the added glucuronic acid moiety does not absorb light at 450 nm wavelength, the molar extinction coefficient of the parent compound UCB is not affected. UCB, BMG1, BMG2 and BDG have the same molar extinction coefficient, and their calibration curves are extremely similar, therefore the calibration curves for UCB were used to estimate concentration of BMG1, BMG2 and BDG as previous reports [32,33,46,51]. Moreover, quantification of the analytes only based on the UCB standard curve simplified the quantification process. 3.4. Precision, accuracy and recovery The method for determining bilirubin in the RLM, HLM and UGT1A1 incubation samples was validated according to FDA guidelines for the analysis of drugs in biological fluids. The assay results showed that all the intra- and inter-day precision for bilirubin did not exceed 3.7% of RSD at three concentration levels. Data on accuracies about the mean assay recoveries of bilirubin were 92.8 ± 1.5% to 104.3 ± 2.2% (n = 5) for intra- and inter-day assays at low, medium and high concentrations of bilirubin in the RLM, HLM and UGT1A1 incubation samples (Table 2). The mean extraction recoveries of bilirubin (n = 5) from the RLM, HLM and UGT1A1 incubation samples were satisfactory at low, medium and high concentrations, which varied from 91.5 ± 1.0% to 104.3 ± 6.7% (Table 2). High recovery of bilirubin from the three incubation systems suggested that there was negligible loss during the extraction procedure. Moreover, precision for BMG1, BMG2 and BDG from the RLM, HLM and UGT1A1 incubation samples at three initial bilirubin concentration levels did not exceed 1.1% of RSD for within-run and 2.9% of RSD for inter-run, respectively. All the data on precision, accuracy and recovery complied with the requirements of bioanalytical method validation prepared by Center for Drug Evaluation and Research (CDER) FDA [41]. 3.5. Stability The stability experiment indicated that bilirubin stock solution (2 mM, n = 3) was stable at room temperature (25 ◦ C) for 6 h and −70 ◦ C for 30 d. Compared to that of the initial time point, the percentage remaining of bilirubin were ≥98.2 ± 1.8% and 92.2 ± 2.6% (Table S1 in the supplementary data), respectively. The bilirubin standard solutions were stable at room temperature for 24 h, the

Linear range (␮M) 0.01–2 0.01–2 0.01–2

remaining percentages of bilirubin at the concentration of 0.05, 0.2 and 1.5 ␮M were ≥96.9 ± 5.5% (Table S2 in the supplementary data). Stability of bilirubin and its glucuronides (i.e. UCB, BMG1, BMG2 and BDG) was satisfactory at 37 ◦ C during the reaction (0–30 min) in the RLM, HLM and UGT1A1 incubation solutions, 25 ◦ C for 24 h and −70 ◦ C for 7 d in the processed incubation samples with methanol (Table 3 and Tables S3–S11 in the supplementary data). These studies indicated that bilirubin and its glucuronides were stable in the procedure of preparation, incubation, handling, storage and assay of the tested samples under the conditions selected. The satisfactory stability was probably due to a series of measures that we took. In order to protect from photolysis, oxidation, degradation and adsorption of bilirubin and its glucuronides, we kept the preparation, incubation, extraction and storage of the tested samples in amber glass containers filled with argon in the dark room equipped with dim yellow light (without UV), under the near-neutral (pH 7.4) incubation circumstance, addition of the antioxidant and stabilizer (e.g. ascorbic acid) in the incubation mixtures, sample storage at low temperature (−70 ◦ C), usage of low adsorptive and/or lightresistant experimental materials (e.g. low-binding tips, aluminum foil-wrapped amber glass vials, tubes with screw cap and flask with glass stopper), as well as quick manipulations in the experiment. Among them, ascorbic acid (as a reducing agent) can strongly inhibit non-enzymic hydrolysis of bilirubin glucuronides, and completely diminish the conversion of BMGs into BDG and UCB, as well as prevent from oxidation of bilirubin and its glucuronides [28]. Meanwhile, the satisfactory stability of bilirubin and its glucuronides in the incubation solutions during the reaction was probably related with the stabilizing effect of the protein from the RLM (HLM or UGT1A1). The protein can strongly inhibit the conversion of BMGs into BDG and UCB [33]. It must be pointed out that any decreased time for pretreatment and analysis of samples would decrease the possibility of test sample degradation. In addition, the HPLC column and channel were eluted using 100% acetonitrile over 12 h at the end of every experiment so as to decrease the adsorption of bilirubin on them and prolong their life. In a word, all these measures not only increased stability of the samples, but also improved accuracy and precision of the analytical method. 3.6. Bilirubin glucuronidation To guarantee the process of in vitro bilirubin glucuronidation under the initial rate conditions and formation of appropriate amount of bilirubin glucuronides, meanwhile, taking account of the saturation of bilirubin solubility in the incubation solution, as well as the maneuverability of the experiment (e.g. rate of bilirubin glucuronidation is very quickly), we investigated the incubation conditions, e.g. substrate concentration, microsomal or UGT1A1 protein concentration, and incubation time. Finally, biliruin concentration ranges of 0.25–2 ␮M, microsomal or UGT1A1 protein concentration of 12.5 ␮g/mL, and incubation time of 15 min were chosen as the optimized incubation conditions to characterize bilirubin glucuronidation. The results indicated that, under these conditions, bilirubin glucuronidation obeyed the Hill equation by RLM, and the Michaelis–Menten equation by HLM and UGT1A1, respectively. The kinetic profiles and parameters of bilirubin glucuronidation by RLM, HLM and UGT1A1 were shown in Fig. 3 and Table 4, respectively.

G. Ma et al. / Journal of Pharmaceutical and Biomedical Analysis 92 (2014) 149–159

155

Table 2 Intra- and inter-day precision, accuracy and recovery for the determination of bilirubin in the RLM, HLM and UGT1A1 incubation mixtures, respectively (n = 5, X ± SD). Assay recovery

Extraction recovery

Bilirubin conc. added (␮M)

Intra-day

Inter-day

Conc. assayed (␮M)

Conc. assayed (␮M)

98.2 ± 1.4 98.3 ± 1.8 92.8 ± 1.5

1.4 1.8 1.6

0.050 ± 0.001 0.197 ± 0.004 1.404 ± 0.052

99.7 ± 1.3 98.5 ± 1.8 93.6 ± 3.7

1.3 1.8 3.7

91.5 ± 1.0 104.3 ± 6.7 97.6 ± 5.6

1.1 6.4 5.7

0.051 ± 0.0004 0.208 ± 0.001 1.442 ± 0.005

101.5 ± 0.8 104.1 ± 0.7 96.1 ± 0.3

0.8 0.6 0.4

0.051 ± 0.001 0.209 ± 0.004 1.444 ± 0.017

101.6 ± 2.1 104.3 ± 2.2 96.3 ± 1.2

2.1 2.1 1.2

96.8 ± 0.7 94.7 ± 1.3 94.9 ± 0.7

0.8 1.3 0.7

0.052 ± 0.001 0.202 ± 0.003 1.438 ± 0.022

103.3 ± 1.2 101.2 ± 1.4 95.9 ± 1.5

1.2 1.4 1.5

0.051 ± 0.001 0.204 ± 0.003 1.424 ± 0.015

101.3 ± 1.8 102.2 ± 1.6 94.9 ± 1.0

1.8 1.6 1.1

98.3 ± 2.1 97.3 ± 2.0 92.3 ± 0.8

2.1 2.0 0.9

0.049 ± 0.001 0.197 ± 0.004 1.392 ± 0.022

HLM 0.05 0.2 1.5 UGT1A1 0.05 0.2 1.5

Recovery (%)

RSD

RSD (%)

RLM 0.05 0.2 1.5

Recovery (%)

Recovery RSD %

Conc., concentration(s); RSD, relative standard deviation.

Table 3 Stability of bilirubin and its glucuronides in the RLM, HLM and UGT1A1 incubation samples at 37 ◦ C, 25 ◦ C and −70 ◦ C during or after the reaction, respectively (n =3, X ± SD). Storage condition

Incubation system RLM

HLM

a

a

UGT1A1

0.2 (␮M)

a

a

a

a

0.75 (␮M)

1.5 (␮M)

0.2 (␮M)

0.75 (␮M)

1.5 (␮M)

0.2a (␮M)

0.75a (␮M)

1.5a (␮M)

37 C × 15 min UCBb BMG1b BMG2b BDGb

97.6 ± 2.9 97.5 ± 6.5 102.7 ± 3.0 101.2 ± 3.3

94.1 ± 1.7 98.0 ± 2.6 99.8 ± 3.4 99.1 ± 2.2

99.3 ± 1.0 99.5 ± 5.2 96.9 ± 4.6 98.8 ± 0.9

101.5 ± 3.6 97.6 ± 3.4 99.0 ± 3.0 100.1 ± 5.7

95.8 ± 0.8 100.0 ± 3.5 99.7 ± 1.9 99.9 ± 1.8

97.3 ± 1.3 98.1 ± 3.3 95.7 ± 5.0 98.3 ± 2.4

101.7 ± 4.2 99.1 ± 2.3 98.8 ± 0.9 97.9 ± 4.0

98.9 ± 1.9 99.7 ± 7.8 101.0 ± 1.5 96.5 ± 1.0

99.6 ± 3.3 99.3 ± 2.2 97.3 ± 6.3 99.1 ± 1.8

25 ◦ C × 24 h UCBb BMG1b BMG2b BDGb

87.4 ± 3.6 88.9 ± 5.1 91.2 ± 1.4 96.7 ± 1.5

88.8 ± 0.9 90.4 ± 0.8 85.8 ± 1.0 93.9 ± 1.3

86.6 ± 0.8 85.0 ± 1.7 87.8 ± 0.4 94.5 ± 0.8

100.6 ± 0.2 99.0 ± 1.8 97.4 ± 0.9 99.8 ± 0.2

97.3 ± 1.0 97.4 ± 2.6 95.8 ± 1.0 94.6 ± 1.5

96.7 ± 0.9 94.5 ± 0.5 95.3 ± 0.5 96.6 ± 2.7

95.3 ± 1.8 90.1 ± 8.2 90.0 ± 0.8 89.5 ± 0.5

89.5 ± 2.9 90.9 ± 1.3 90.0 ± 0.4 91.0 ± 1.1

87.0 ± 4.0 90.5 ± 1.2 90.3 ± 0.2 94.5 ± 1.1

−70 ◦ C × 7 d UCBb BMG1b BMG2b BDGb

100.1 ± 2.6 99.8 ± 1.1 100.1 ± 1.3 100.0 ± 0.6

99.3 ± 0.7 100.1 ± 1.7 99.9 ± 0.3 102.5 ± 1.1

100.3 ± 1.5 100.1 ± 0.1 99.6 ± 0.2 101.2 ± 2.1

99.4 ± 1.1 100.5 ± 2.0 99.2 ± 2.1 100.9 ± 4.7

91.4 ± 0.1 98.2 ± 3.2 97.3 ± 5.8 97.4 ± 2.2

87.8 ± 0.9 99.3 ± 0.5 99.8 ± 0.3 97.7 ±1.9

96.2 ± 3.5 97.4 ± 1.5 101.9 ± 0.3 100.4 ± 3.9

102.0 ± 1.6 100.2 ± 0.4 101.1 ± 0.6 98.0 ± 2.6

96.9 ± 0.6 100.9 ± 0.3 101.4 ± 0.6 101.2 ± 2.3

◦

SC, storage condition; IS, incubation system. a The concentrations of bilirubin (i.e. UCB) added in the incubation mixtures at the initial reaction time point (0 min). b The values given in the rows are percentage remaining (PR) (%).

As shown in Fig. 3, rank orders of average formation rates of bilirubin glucuronides were: VUGT1A1 > VRLM > VHLM for BMG1, VRLM > VUGT1A1 > VHLM for BMG2, VUGT1A1 > VHLM > VRLM for BDG, and VUGT1A1 ≈ VRLM > VHLM for TBG (p < 0.05), respectively. Under the same incubation conditions (i.e. the same substrate concentration, protein concentration and incubation time), the average formation rates of TBG by HLM were slower than that by UGT1A1 and

RLM, and it showed no significant difference (p > 0.05) between UGT1A1 and RLM. However, as shown in Table 4, Km and Vmax for total bilirubin glucuronidation by RLM, HLM, UGT1A1 ranged from 0.40 ± 0.022 ␮M to 0.45 ± 0.016 ␮M and 1.86 ± 0.029 nmol/mg/min to 2.95 ± 0.036 nmol/mg/min in the present study, respectively. It is interesting to note that, under the same incubation condition, the apparent kinetic parameters Km , Vmax and CLint

Table 4 Apparent enzyme kinetic parameters of bilirubin glucuronidation by RLM, HLM and UGT1A1, respectively (n = 3, X¯ ± SD). Incubation systems

Kinetic parameters

BMG1

RLM

Km (␮M) Vmax (nmol/mg/min) CLint (mL/mg/min) n

0.45 0.81 1.81 2.17

HLM

Km (␮M) Vmax (nmol/mg/min) CLint (mL/mg/min)

0.47 ± 0.026 0.68 ± 0.012 1.44 ± 0.043

0.48 ± 0.024 0.86 ± 0.014 1.80 ± 0.048

0.25 ± 0.026 0.31 ± 0.007 1.22 ± 0.058

0.40 ± 0.022 1.86 ± 0.029 4.70 ± 0.079

Michaelis–Menten equation

UGT1A1

Km (␮M) Vmax (nmol/mg/min) CLint (mL/mg/min)

0.66 ± 0.03 1.16 ± 0.02 1.75 ± 0.12

0.64 ± 0.027 1.27 ± 0.020 2.00 ± 0.084

0.17 ± 0.017 0.60 ± 0.011 3.68 ± 0.34

0.44 ± 0.018 2.95 ± 0.036 6.72 ± 0.27

Michaelis–Menten equation

TBG = BMG1 + BMG2 + BDG; CLint = Vmax /Km .

± ± ± ±

BMG2 0.017 0.019 0.060 0.17

0.48 1.62 3.39 1.86

± ± ± ±

BDG 0.019 0.038 0.14 0.12

0.29 0.19 0.58 1.13

TBG ± ± ± ±

0.039 0.014 0.10 0.23

0.45 2.65 5.92 1.85

Kinetic mechanism ± ± ± ±

0.016 0.057 0.22 0.11

Hill equation

0.6 V

0.3

1.5

0

2

0

1.5 1.0 0.5

0

0.0 0.0

0

0.5

0.5

1

1.0 V/C

1.5

1.5

1

0.4

1.5

1.0

0.2

2 3

1.5

2

1

1

0.5

0

0

1

0.5

0.0 0.0 0.5 1.0 1.5 2.0 V/C

0

0.5

2

3

0.3

1

0

0.5

1

[Bilirubin (μM)

1

V

0.1

0

0.5

1.6

1

1.5

2

0.5 1.0 V/C

1.5

2

0.8 1

0

0

0

0.5

1

2 V/C

1.5

[Bilirubin (μM)

2

B3

1.0

0.4

0.5

0.2

0.0 0

0

0.5

0.6

1 V/C

1

2

1.5

2

4

2

C3

0.4 0.6

0.2

0.3 0.0 0

0

1.2

0

1.5

0.6

2

D2

0.4

1.5

[Bilirubin (μM)

0.2

0.0 0.0

1

1.0 V/C

0.8

0

2

C2

V/C

0

1.5

0.2

0

2

D1

2.5

0.5

0.5

[Bilirubin (μM)

0.4

V

0.5

0.2 V/C

TBG formation rate (nmol/mg/min)

0

0.1

V

TBG formation rate (nmol/mg/min)

BDG formation rate (nmol/mg/min)

0.1 V

BDG formation rate (nmol/mg/min)

0.2

0

0

[Bilirubin (μM)

0.15

0.0 0.0

0.0 0.0

0

2

0.6

0

2

C1

0.05

1.5

B2

[Bilirubin (μM) 0.2

1.5

0.5

V

1.2

0.4

1

V

B1

0.8

0.5

1.0 V/C

0.3

[Bilirubin (μM)

BMG2 formatio rate (nmol/mg/min)

1.6

V

BMG2 formation rate (nmol/mg/min)

[Bilirubin (μM)

0.5

1.0

V

1

0.0 0.0

0.6

0

0.5

2.5

1

1

2 V/C

3

1.5

4

2

D3

2 1.5

3 2

1

V

0.5

1.0

0.2

BMG2 formation rate (nmol/mg/min)

0

0.5 V/C

0.5

BDG formation rate (nmol/mg/min)

0

0.0 0.0

0.4

TBG formation rate (nmol/mg/min)

0.2

A2

UGT1A1 A3

0.9

V

0.9

0.4

BMG1 formation rate (nmol/mg/min)

A1

0.6

HLM 0.6

V

RLM 0.8

BMG1 formation rate (nmol/mg/min)

G. Ma et al. / Journal of Pharmaceutical and Biomedical Analysis 92 (2014) 149–159

BMG1 formation rate (nmol/mg/min)

156

1

0.5 0

0

0

0.5

0

1

2

4 V/C

1.5

6

8

2

[Bilirubin (μM)

Fig. 3. Kinetics profiles of bilirubin glucuronidation by RLM (A1–D1), HLM (A2–D2) and UGT1A1 (A3–D3), respectively. BMG1 (A1–A3) and BMG2 (B1–B3) represented two bilirubin monoglucuronides, and BDG (C1–C3) represented bilirubin diglucuronides. TBG (D1–D3) represented total bilirubin glucuronides, i.e. TBG = BMG1 + BMG2 + BDG. The Hill equation was fit to the data from the RLM (A1–D1) incubation system, and the Michaelis–Menten equation was fit to the data from the HLM (A2–D2) and UGT1A1 (A3–D3) incubation systems, respectively. The embedded figures are Eadie–Hofstee plots for the same data. Rhombuses and smooth lines denote the observed and predicted rates of bilirubin glucuronidation, respectively. Microsomal or UGT1A1 protein concentration was 12.5 ␮g/mL, incubation time was 15 min. Each data point represented the average of three replicates.

values of bilirubin glucuronidation (i.e. formation of BMG1, BMG2, BDG and TBG) exhibited significant differences (p < 0.05) in the three different incubation systems (Fig. 4). For example, average kinetic parameter values of TBG formation showed Vmax,UGT1A1 > Vmax,RLM > Vmax,HLM Km,UGT1A1 ≈ Km,RLM > Km,HLM , and CLint,UGT1A1 > CLint,RLM > CLint,HLM . The results indicated that bilirubin had the same binding affinity to the UGT1A1 from the recombinant human UGT1A1 enzyme and RLM, and showed the strongest affinity to the UGT1A1 from the HLM. Meanwhile, under the same incubation conditions, recombinant human UGT1A1 enzyme demonstrated the strongest capacity and efficiency for bilirubin glucuronidation, HLM is just the reverse. In fact, Km and Vmax values of bilirubin glucuronidation were reported to range from 0.20 ␮M to as high as 24 ␮M and 0.08 to 1.08 nmol/mg/min when these incubation experiments were

conducted using 0.05–200 ␮M bilirubin, 0.05–2.3 mg/mL protein from the microsomes or UGT1A1, and 5–35 min incubation time, respectively [26,31–33]. The discrepancy in the present experiment and the previous reports is most probably due to different incubation conditions selected, e.g. the substrate concentrations, concentrations and sources of UGT1A1 (e.g. different cell lines, microsomes, supersomes, cDNA-expressed enzymes), reaction time and assay methodology. Especially, the differences of apparent kinetic parameters in our experiment were probably due to different UGT1A1 source (i.e. pooled rat and human microsomes, recombinant human UGT1A1 enzyme from BD-SupersomesTM ), enzymatic activity, and content of UGT1A1 in the RLM, HLM and BD-supersomes. Moreover, all the Km values of formation of BMG1 and BMG2 showed no significant difference (p > 0.05) in the same incubation system. This indicated that C8 and C12 carboxyl group

G. Ma et al. / Journal of Pharmaceutical and Biomedical Analysis 92 (2014) 149–159

A

*

0.8

*

* *

RLM

HLM

UGT1A1

157

RLM

100 80

*

*

0.4

% of TBG

Km (μM)

0.6

*

60

BDG BMGs

40 20

0.2 0 0

0.0 BMG1

BMG2

BDG

TBG

Vmax (nmol/mg/min)

2.5

*

2.0

1.0

*

*

1.5

60

BDG BMGs

40 20

* *

0.5

*

0 0

0.5

BMG2

BDG

TBG

7.0

*

CLint (mL/mg/min)

6.0

* *

4.0

*

*

60

BDG BMGs

40 20 0

3.0 2.0

2

80

% of TBG

*

5.0

1.5

UGT1A1

100

bilirubin glucuronides

C

1

[Bilirubin] (μM)

0.0 BMG1

2

80

* % of TBG

*

3.0

1.5

HLM

100

3.5

1

[Bilirubin] (μM)

bilirubin glucuronides

B

0.5

* *

0

1

1.5

2

[Bilirubin] (μM)

1.0 0.0 BMG1

0.5

BMG2

BDG

TBG

Fig. 5. Proportions of BMGs and BDG formed in the RLM, HLM and UGT1A1 incubation systems, respectively. BMGs = BMG1 + BMG2; microsomal (or UGT1A1) protein concentration 12.5 ␮g/mL; incubation time 15 min.

bilirubin glucuronides Fig. 4. Comparison of kinetic parameters of bilirubin glucuronidation by RLM, HLM and UGT1A1, respectively. A, B and C were the corresponding column plots summarizing the values of Km , Vmax and CLint , respectively. “*”, significant difference (p < 0.05). Microsomal (or UGT1A1) protein concentration 12.5 ␮g/mL; incubation time 15 min.

of bilirubin have the same affinity to UGT1A1 when UCB was glucuronidated to BMGs (i.e. BMG1 and BMG2) by RLM (HLM or UGT1A1). The apparent Vmax values of formation of BMG1 and BMG2 showed no significant difference (p > 0.05) in the UGT1A1 incubation system, but Vmax,BMG2 > Vmax,BMG1 (p < 0.05) in the RLM and HLM incubation system. This indicated that recombinant human UGT1A1 enzyme has the same glucuronidation capacity for formation of BMG1 and BMG2, but RLM and HLM have stronger glucuronidation capacity for formation of BMG2 than that of BMG1. CLint,BMG2 > CLint,BMG1 (p < 0.05) in the three incubation systems indicated that efficiency of intrinsic clearance of BMG2 was higher than that of BMG1.

It needs to be pointed out that it looked the glucuronidation reaction of biliruin was not reach Vmax in some conditions (e.g. A2, B2, A3, B3, etc.) in Fig. 3. This is probably related with the substrate (i.e. bilirubin) concentration selected. The maximum bilirubin concentration selected was 2 ␮M in the study, which was based on the insolubility of bilirubin in the incubation solutions and poor linearities of bilirubin standard solution when its concentration was >2 ␮M. Taking account of the saturation of bilirubin solubility in the incubation solutions, it was not appropriate to further increase bilirubin concentration to >2 ␮M. In fact, as shown in Fig. 3, the observed and predicted rates of bilirubin glucuronidation showed excellent goodness of fit. The Eadie–Hofstee plots, the residual sum of squares (R2 ≥ 0.98) and Akaike information criterion (AIC ≤ −112.97) also exhibited excellent fitting for the data and model. It indicated that apparent enzyme kinetic parameters of bilirubin glucuronidation were reliable in the tested substrate concentration range selected. In addition, our results indicated that proportions of BMGs and BDG formed depended on bilirubin concentration and enzyme sources. As shown in Fig. 5, BMGs were the dominant species

158

G. Ma et al. / Journal of Pharmaceutical and Biomedical Analysis 92 (2014) 149–159

formed (>66%) among the bilirubin glucuronides in the three incubation systems. When concentration of bilirubin rose from 0.25 ␮M to 2 ␮M, the proportions of BMGs increased and proportions of BDG decreased, finally both tended to be constant. The proportions of BMGs formed were the highest in the RLM incubation systems and the lowest in the UGT1A1 incubation systems, whereas the proportions of BDG formed were on the contrary, respectively. Accordingly it can be assumed that, under the same incubation conditions, bilirubin was more easily metabolized to BDG by UGT1A1 than RLM and HLM. 4. Conclusion In conclusion, a simple, reproducible and robust HPLC method for simultaneous determination of bilirubin and its multiple glucuronides (including their isomers) in RLM, HLM and UGT1A1 incubation systems, and a stable and reliable in vitro incubation system were successfully established to investigate the kinetics of bilirubin glucuronidation by RLM, HLM and UGT1A1, respectively. Bilirubin glucuronidation obeyed the Hill equation by RLM, and the Michaelis–Menten equation by HLM and UGT1A1 in the concentration range of 0.25–2 ␮M bilirubin, and exhibited kinetic differences probably due to the difference of enzyme sources and UGT1A1 content. Furthermore, the established HPLC method and in vitro incubation system can also be used to research inhibition and induction of bilirubin glucuronidation by xenobiotics (e.g. drugs and toxics) and endogenous substances (e.g. estradiol). Acknowledgments The author would like to thank Dr. Ming Hu of Department of Pharmacological and Pharmaceutical Sciences, College of Pharmacy, University of Houston for his help in experimental design and analytical procedure. The work was supported by the National Natural Science Funds of China (no. 81374051), the Fundamental Research Funds for the Central Universities (no. 20520133531) and the Fudan s Wangdao Research Program (no. JMH6285113/014/002). Appendix A. Supplementary data Supplementary material related to this article can be found, in the online version, at http://dx.doi.org/10.1016/j.jpba.2014.01.025. References [1] L. Vitek, J.D. Ostrow, Bilirubin chemistry and metabolism; harmful and protective aspects, Curr. Pharm. Des. 15 (2009) 2869–2883. [2] J. Fevery, Bilirubin in clinical practice: a review, Liver Int. 28 (2008) 592–605. [3] L. Vítek, H.A. Schwertner, The heme catabolic pathway and its protective effects on oxidative stress-mediated diseases, Adv. Clin. Chem. 43 (2007) 1–57. [4] S. Gazzin, N. Strazielle, C. Tiribelli, J.F. Ghersi-Egea, Transport and metabolism at blood–brain interfaces and in neural cells: relevance to bilirubin-induced encephalopathy, Front. Pharmacol. 3 (2012) 89. [5] J.D. Ostrow, L. Pascolo, D. Brites, C. Tiribelli, Molecular basis of bilirubin-induced neurotoxicity, Trends Mol. Med. 10 (2004) 65–70. [6] S.M. Shapiro, Bilirubin toxicity in the developing nervous system, Pediatr. Neurol. 29 (2003) 410–421. [7] D.K. Stevenson, H.J. Vreman, R.J. Wong, Bilirubin production and the risk of bilirubin neurotoxicity, Semin. Perinatol. 35 (2011) 121–126. [8] J.M. Crawford, B.J. Ransil, J.P. Narciso, J.L. Gollan, Hepatic microsomal bilirubin UDP-glucuronosyltransferase. The kinetics of bilirubin mono- and diglucuronide synthesis, J. Biol. Chem. 267 (1992) 16943–16950. [9] J. Fevery, D. Van, R. Michiels, J. De Groote, K.P. Heirwegh, Bilirubin conjugates in bile of man and rat in the normal state and in liver disease, J. Clin. Invest. 51 (1972) 2482–2492. [10] M. Muraca, F.F. Rubaltelli, N. Blanckaert, J. Fevery, Unconjugated and conjugated bilirubin pigments during perinatal development. II. Studies on serum of healthy newborns and of neonates with erythroblastosis fetalis, Biol. Neonate 57 (1990) 1–9.

[11] K. Köck, K.L. Brouwer, A perspective on efflux transport proteins in the liver, Clin. Pharmacol. Ther. 92 (2012) 599–612. [12] P.J. Bosma, J. Seppen, B. Goldhoorn, C. Bakker, R.P. Oude Elferink, J.R. Chowdhury, N.R. Chowdhury, P.L. Jansen, Bilirubin UDP-glucuronosyltransferase 1 is the only relevant bilirubin glucuronidating isoform in man, J. Biol. Chem. 269 (1994) 17960–17964. [13] A. Kadakol, S.S. Ghosh, B.S. Sappal, G. Sharma, J.R. Chowdhury, N.R. Chowdhury, Genetic lesions of bilirubin uridine-diphosphoglucuronate glucuronosyltransferase (UGT1A1) causing Crigler–Najjar and Gilbert syndromes: correlation of genotype to phenotype, Hum. Mutat. 16 (2000) 297–306. [14] B. Gupta, C. LeVea, A. Litwin, M.G. Fakih, Reversible grade 4 hyperbilirubinemia in a patient with UGT1A1 7/7 genotype treated with irinotecan and cetuximab, Clin. Colorectal Cancer 6 (2007) 447–449. ´ O. Clewe, D. Röshammar, L. Flamholc, A. Sönnerborg, V. Ormaasen, M. [15] D. Rekic, Gisslén, A. Abelö, M. Ashton, Bilirubin – a potential marker of drug exposure in atazanavir-based antiretroviral therapy, AAPS J. 13 (2011) 598–605. [16] M. Rotger, P. Taffe, G. Bleiber, H.F. Gunthard, H. Furrer, P. Vernazza, H. Drechsler, E. Bernasconi, M. Rickenbach, A. Telenti, Swiss HIV Cohort Study, Gilbert syndrome and the development of antiretroviral therapy-associated hyperbilirubinemia, J. Infect. Dis. 192 (2005) 1381–1386. [17] Y. Liu, J. Ramirez, L. House, M.J. Ratain, Comparison of the drug–drug interactions potential of erlotinib and gefitinib via inhibition of UDPglucuronosyltransferases, Drug Metab. Dispos. 38 (2010) 32–39. [18] C.J. Peer, T.M. Sissung, A. Kim, L. Jain, S. Woo, E.R. Gardner, C.T. Kirkland, S.M. Troutman, B.C. English, E.D. Richardson, J. Federspiel, D. Venzon, W. Dahut, E. Kohn, S. Kummar, R. Yarchoan, G. Giaccone, B. Widemann, W.D. Figg, Sorafenib is an inhibitor of UGT1A1 but is metabolized by UGT1A9: implications of genetic variants on pharmacokinetics and hyperbilirubinemia, Clin. Cancer Res. 18 (2012) 2099–2107. [19] J. Zhou, T.S. Tracy, R.P. Remmel, Correlation between bilirubin glucuronidation and estradiol-3-gluronidation in the presence of model UDP-glucuronosyltransferase 1A1 substrates/inhibitors, Drug Metab. Dispos. 39 (2011) 322–329. [20] M.G. Bartlett, G.R. Gourley, Assessment of UGT polymorphisms and neonatal jaundice, Semin. Perinatol. 35 (2011) 127–133. [21] G. Ma, B. Wu, S. Gao, Z. Yang, Y. Ma, M. Hu, Mutual regioselective inhibition of human UGT1A1-mediated glucuronidation of four flavonoids, Mol. Pharm. 10 (2013) 2891–2903. [22] J. Sugatani, K. Mizushima, M. Osabe, K. Yamakawa, S. Kakizaki, H. Takagi, M. Mori, A. Ikari, M. Miwa, Transcriptional regulation of human UGT1A1 gene expression through distal and proximal promoter motifs: implication of defects in the UGT1A1 gene promoter, Naunyn Schmiedebergs Arch. Pharmacol. 77 (2008) 597–605. [23] D. Chawla, V. Parmar, Phenobarbitone for prevention and treatment of unconjugated hyperbilirubinemia in preterm neonates: a systematic review and meta-analysis, Indian Pediatr. 47 (2010) 401–407. [24] E. Ellis, M. Wagner, F. Lammert, A. Nemeth, J. Gumhold, C.P. Strassburg, C. Kylander, D. Katsika, M. Trauner, C. Einarsson, H.U. Marschall, Successful treatment of severe unconjugated hyperbilirubinemia via induction of UGT1A1 by rifampicin, J. Hepatol. 44 (2006) 243–245. [25] T.K. Chang, Activation of pregnane X receptor (PXR) and constitutive androstane receptor (CAR) by herbal medicines, AAPS J. 11 (2009) 590–601. [26] D.L. Zhang, T.J. Chando, D.W. Everett, C.J. Patten, S.S. Dehal, W.G. Humphreys, In vitro inhibition of UDP glucuronosyltransferases by atazanavir and other HIV protease inhibitors and the relationship of this property to in vivo bilirubin glucuronidation, Drug Metab. Dispos. 331 (2005) 1729–1739. [27] J.D. Ostrow, P. Mukerjee, C. Tiribelli, Structure and binding of unconjugated bilirubin: relevance for physiological and pathophysiological function, J. Lipid Res. 35 (1994) 1715–1737. [28] S. Adachi, T. Uesugi, K. Kamisaka, Study of bilirubin metabolism by highperformance liquid chromatography: stability of bilirubin glucuronides, Arch. Biochem. Biophys. 241 (1985) 486–493. [29] B.T. Doumas, B.W. Perry, E.A. Sasse, J.V. Straumfjord Jr., Standardization in bilirubin assays: evaluation of selected methods and stability of bilirubin solutions, Clin. Chem. 19 (1973) 984–993. [30] J. Seppen, P.J. Bosma, B.G. Goldhoorn, C.T. Bakker, J.R. Chowdhury, N.R. Chowdhury, P.L. Jansen, R.P. Oude Elferink, Discrimination between Crigler–Najjar type I and II by expression of mutant bilirubin uridine diphosphate-glucuronosyltransferase, J. Clin. Invest. 94 (1994) 2385–2391. [31] S.B. Senafi, D.J. Clarke, B. Burchell, Investigation of the substrate-specificity of a cloned expressed human bilirubin UDP-glucuronosyltransferase: UDPsugar specificity and involvement in steroid and xenobiotic glucuronidation, Biochem. J. 303 (Pt 1) (1994) 233–240. [32] W. Udomuksorn, D.J. Elliot, B.C. Lewis, P.I. Mackenzie, K. Yoovathaworn, J.O. Miners, Influence of mutations associated with Gilbert and Crigler–Najjar type II syndromes on the glucuronidation kinetics of bilirubin and other UDPglucuronosyltransferase 1A substrates, Pharmacogenet. Genomics 17 (2007) 1017–1029. [33] J. Zhou, T.S. Tracy, R.P. Remmel, Bilirubin glucuronidation revisited: proper assay conditions to estimate enzyme kinetics with recombinant UGT1A1, Drug Metab. Dispos. 38 (2010) 1907–1911. [34] M. Muraca, N. Blanckaert, Liquid-chromatographic assay and identification of mono- and diester conjugates of bilirubin in normal serum, Clin. Chem. 29 (1983) 1767–1771.

G. Ma et al. / Journal of Pharmaceutical and Biomedical Analysis 92 (2014) 149–159 [35] W. Spivak, M.C. Carey, Reverse-phase h.p.l.c. separation, quantification and preparation of bilirubin and its conjugates from native bile. Quantitative analysis of the intact tetrapyrroles based on h.p.l.c. of their ethyl anthranilate azo derivatives, Biochem. J. 225 (1985) 787–805. [36] H. Saxerholt, V. Skar, T. Midtvedt, HPLC separation and quantification of bilirubin and its glucuronide conjugates in faeces and intestinal contents of germ-free rats, Scand. J. Clin. Lab. Invest. 50 (1990) 487–495. [37] M. Mizobe, F. Kondo, K. Kumamoto, T. Terada, H. Nasu, Rapid and quantitative analysis of bilirubin in equines by high-performance liquid chromatography, Microbios 86 (1996) 39–47. [38] J. Zelenka, M. Lenícek, L. Muchová, M. Jirsa, M. Kudla, P. Balaz, M. Zadinová, J.D. Ostrow, R.J. Wong, L. Vítek, Highly sensitive method for quantitative determination of bilirubin in biological fluids and tissues, J. Chromatogr. B: Analyt. Technol. Biomed. Life Sci. 867 (2008) 37–42. [39] P. Cary, J.V. Jodie, E.Q.J. Martin, P.A. Horacio, L. David, The animal pigment bilirubin identified in Strelitzia reginae, the bird of paradise flower, HortScience 45 (2010) 1411–1415. [40] G.B. Messiano, R.A. Santos, S. Lde Ferreira, R.A. Simões, V.A. Jabor, M.J. Kato, N.P. Lopes, M.T. Pupo, A.R. de Oliveira, In vitro metabolism study of the promising anticancer agent the lignan (−)-grandisin, J. Pharm. Biomed. Anal. 72 (2013) 240–244. [41] Guidance for Industry: Bioanalytical Method Validation, Center for Drug Evaluation and Research (CDER), Center for Veterinary Medicine (CVM), United States Food and Drug Administration. http://www.fda.gov/downloads/ Drugs/Guidances/ucm070107.pdf [42] B. Wu, S. Zhang, M. Hu, Evaluation of 3,3 ,4 -trihydroxyflavone and 3,6,4 trihydroxyflavone (4 -O-glucuronidation) as the in vitro functional markers for hepatic UGT1A1, Mol. Pharm. 8 (2011) 2379–2389.

159

[43] A. Christopoulos, M.J. Lew, Beyond eyeballing: fitting models to experimental data, Crit. Rev. Biochem. Mol. Biol. 35 (2000) 359–391. [44] S. Takeda, Y. Kitajima, Y. Ishii, Y. Nishimura, P.I. Mackenzie, K. Oguri, H. Yamada, Inhibition of UDP-glucuronosyltransferase 2B7-catalyzed morphine glucuronidation by ketoconazole: dual mechanisms involving a novel noncompetitive mode, Drug Metab. Dispos. 34 (2006) 1277–1282. [45] S.F. Lo, B.T. Doumas, The status of bilirubin measurements in U.S. laboratories: why is accuracy elusive, Semin. Perinatol. 35 (2011) 141–147. [46] I. Diamond, R. Schmid, Experimental bilirubin encephalopathy. The mode of entry of bilirubin-14C into the central nervous system, J. Clin. Invest. 45 (1966) 678–689. [47] E.A. Rodriguez Garay, O.U. Scremin, Transfer of bilirubin-14C between blood, cerebrospinal fluid, and brain tissue, Am. J. Physiol. 221 (1971) 1264–1270. [48] Y. Okamura, M. Yamazaki, T. Yamaguchi, Y. Komoda, A. Sugimoto, H. Nakajima, Anti-bilirubin monoclonal antibody. III. Preparation and properties of monoclonal antibodies to unconjugated bilirubin-IX alpha, Biochim. Biophys. Acta 1073 (1991) 538–542. [49] T. Kimpara, A. Takeda, T. Yamaguchi, H. Arai, N. Okita, S. Takase, H. Sasaki, Y. Itoyama, Increased bilirubins and their derivatives in cerebrospinal fluid in Alzheimer’s disease, Neurobiol. Aging 21 (2000) 551–554. [50] K. Otani, S. Shimizu, K. Chijiiwa, T. Morisaki, T. Yamaguchi, K. Yamaguchi, S. Kuroki, M. Tanaka, Administration of bacterial lipopolysaccharide to rats induces heme oxygenase-1 and formation of antioxidant bilirubin in the intestinal mucosa, Dig. Dis. Sci. 45 (2000) 2313–2319. [51] W. Spivak, W. Yuey, Application of a rapid and efficient h.p.l.c. method to measure bilirubin and its conjugates from native bile and in model bile systems. Potential use as a tool for kinetic reactions and as an aid in diagnosis of hepatobiliary disease, Biochem. J. 234 (1986) 101–109.