Human Intestinal Raf Kinase Inhibitor Protein (RKIP) Catalyzes Prasugrel as a Bioactivation Hydrolase

ABSTRACT

Prasugrel is a thienopyridine antiplatelet prodrug that undergoes rapid hydrolysis in vivo to a thiolactone metabolite by human carboxylesterase-2 (hCE2) during gastrointestinal absorption. The thiolactone metabolite is further converted to a pharmacologically active metabolite by cytochrome P450 isoforms. The aim of the current study was to elucidate hydrolases other than hCE2 involved in the bioactivation step of prasugrel in human intestine. Using size- exclusion column chromatography of a human small intestinal S9 fraction, another peak besides the hCE2 peak was observed to have prasugrel hydrolyzing activity, and this protein was found to have a molecular weight of about 20 kDa. This prasugrel hydrolyzing protein was successfully purified from a monkey small intestinal cytosolic fraction by successive four-step column chromatography and identified as Raf-1 kinase inhibitor protein (RKIP) by liquid chromatography-tandem mass spectrometry. Second, we evalu- ated the enzymatic kinetic parameters for prasugrel hydrolysis using recombinant human RKIP and hCE2 and estimated the contributions of these two hydrolyzing enzymes to the prasugrel hydrolysis reaction in human intestine, which were approximately 40% for hRKIP and 60% for hCE2. Moreover, prasugrel hydrolysis was inhibited by anti-hRKIP antibody and carboxylesterase-specific chemical inhibitor (bis p-nitrophenyl phosphate) by 30% and 60%, respectively. In conclusion, another protein capable of hydrolyzing prasugrel to its thiolactone metabolite was identified as RKIP, and this protein may play a significant role with hCE2 in prasugrel bioactivation in human intestine. RKIP is known to have diverse functions in many intracellular signaling cascades, but this is the first report describing RKIP as a hydrolase involved in drug metabolism.

Introduction

Prasugrel (marketed as Effient in the United States and as Efient in Europe, Japan, and other countries) is a third-generation thienopyridine antiplatelet prodrug (Angiolillo et al., 2008) indicated for the reduction of thrombotic cardiovascular events in patients with acute coronary syndrome who are being treated by percutaneous coronary intervention. Prasugrel was not detected in the plasma, urine, or feces after oral administration of prasugrel to humans because it is rapidly absorbed and extensively metabolized in human (Farid et al., 2007). Prasugrel undergoes rapid hydrolysis in vivo to a thiolactone metabolite, which is further converted to a pharmacologically active metabolite by cyto- chrome P450 isoforms (Fig. 1). This rapid hydrolysis reaction of prasugrel may contribute to the rapid onset and the low individual variability of pharmacologic effect of prasugrel compared with other thienopyridine antiplatelet agents such as clopidogrel (Tcheng and Mackay, 2012).

We previously reported, based on in vitro experiments, that prasugrel was converted to a thiolactone metabolite by human carboxylesterase 2 (hCE2) and human carboxylesterase 1 (hCE1) (Williams et al., 2008).The hydrolysis of prasugrel was at least 25 times greater with hCE2 than hCE1. Prasugrel hydrolysis by hCE2 exhibited substrate inhibition at high substrate concentration, although this in vitro observation did not translate to in vivo relevance (Williams et al., 2008). A linear relationship has been shown between the prasugrel dose and the plasma exposure to thiolactone metabolite in human (Asai et al., 2006). Therefore, it was suggested that the formation of thiolactone metabolite from prasugrel was catalyzed by not only hCES but other unknown enzymes. The aim of this study was to identify the hydrolysis enzymes other than hCE2 involved in the hydrolysis of prasugrel in the intestine. Another aim was to determine the contribution of hCE2 and the newly identified enzyme to the hydrolysis of prasugrel to the thiolactone metabolite in human intestine by estimating enzyme kinetic parameters and enzyme inhibition.

Materials and Methods

Materials. Prasugrel, thiolactone metabolites (racemate), and R-135766 (internal standard for thiolactone assay) were synthesized by Ube Industries, Ltd. (Ube, Japan) (Supplemental Fig. S1). Protease inhibitor cocktail (cOmplete mini: mixture of serine, cysteine and metaloprotease inhibitors) was purchased from Roche Applied Science (Mannheim, Germany). Bis p-nitrophenyl phosphate (BNPP) was purchased from Sigma-Aldrich Corporation (St. Louis, MO). Laemmli sample buffer, Flamingo fluorescent gel stain, and 5%–20% SDS acrylamide gel were purchased from Bio-Rad Laboratories, Inc. (Richmond,CA). Bovine serum albumin was purchased from Thermo Scientific Pierce (Rockford, IL).

Biologic Samples. Pooled mixed-gender human intestinal S9 was purchased from Xeno Tech, LLC (Lenexa, KS), and four individual human small intestinal S9 (HIS-063-S3: 16.6 mg protein/ml, HIS-067-S3: 12.6 mg protein/ml, HIS- 084-S3: 19.9 mg protein/ml, HIS-111-S3: 5.7 mg protein/ml) were purchased from the nonprofit Human and Animal Bridging Research Organization (Chiba, Japan). Ethical approval was obtained from the ethics committee of Daiichi Sankyo Co., Ltd. (Tokyo, Japan). FreeStyle 293F cells, cell culture media, and 293fectin were purchased from Invitrogen (Carlsbad, CA). FLAG-M2 agarose was purchased from Sigma-Aldrich. The mouse anti-human Raf Kinase Inhibitor Protein (hRKIP) antibody was obtained from Invitrogen. For immunodepletion, rabbit anti-hRKIP antibodies and the control rabbit IgG were produced by Immuno-Biologic Laboratories Co., Ltd (Gunma, Japan).

A rabbit polyclonal antibody to RKIP (RKIP FL-187: sc-28837) was purchased from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA). A rabbit monoclonal antibody to hCE2 was constructed in Immuno-Biologic Laborato- ries Co., Ltd. (Minneapolis, MN) Amersham (Piscataway, NJ) ECL-anti-rabbit IgG, horseradish peroxidase (HRP)–linked species-specific whole antibody [ECL-anti-rabbit IgG-HRP linked] and Amersham ECL Advance Western Blotting Detection Kit were purchased from GE Healthcare UK Limited. Peptide N-glycosidase F (PNGase F, Lot 0360910) was obtained from New England Biolabs
Japan, Inc. (Tokyo, Japan).

Preparation of Monkey Small Intestinal Subcellular Fractions. All experimental procedures were performed in accordance with the in-house guidelines of the Institutional Animal Care and Use Committee of Daiichi Sankyo Co., Ltd. All preparation was conducted at 4°C. The cynomolgus monkey (HAMURI Co., Ltd., Ibaragi, Japan) was euthanized, and the small intestine was removed. The small intestine was homogenized in 9 volumes of homogenization buffer [10 mM HEPES-NaOH (pH 7.0), 0.25 M sucrose, and protease inhibitor cocktail] using a polytron. The homogenate was centrifuged at 9000g for 30 minutes at 4°C, and the supernatant was used as the S9 fraction. Additionally, the S9 fraction was ultracentrifuged at 105,000g for 1 hour at 4°C, and the supernatant and precipitate fractions were used as the cytosolic and the microsome fraction, respectively. The subcellular fractions were stored at 280°C until use.

Gel Filtration of Human Small Intestine S9 and Monkey Small intestine Cytosol. Two hundred microliters of human small intestinal S9 fraction (10 mg/ml) was loaded onto a gel filtration column (Superdex 75; GE Healthcare, Pittsburgh, PA) and eluted with 20 ml of 100 mM HEPES (pH 7.0) at 0.5 ml/min with a fraction size of 0.5 ml. Similarly, 200 ml of monkey small intestinal cytosolic fraction (10 mg/ml) was separated by the same experimental conditions.

Purification of Endogenous Enzyme. The cytosolic fraction equivalent to 5 g of monkey small intestine was dialyzed against 20 mM sodium acetate (pH 6.0). The dialysate was loaded onto a 20-ml HiPrep heparin column (GE Healthcare) and eluted with a 20-ml linear gradient of 0–0.5 M NaCl. Each fraction was tested for prasugrel hydrolysis activity and analyzed by immuno- blotting with the monoclonal antibody to hCE2. The flow-through fractions (no. 5–10) were pooled and applied to a 1-ml mono S 5/50 GL column (GE Healthcare). The bound proteins were eluted with a 30-ml linear gradient of 0–0.5 M NaCl. Each fraction was tested for prasugrel hydrolysis activity, and the active fractions were pooled and dialyzed against 20 mM Tris-HCl (pH 9.0) buffer. The dialyzed sample was applied to a 1-ml mono Q 5/50 GL column (GE Healthcare), and proteins were eluted with 30 ml linear gradient of 0–0.5 M NaCl. Each fraction was tested for the prasugrel hydrolysis activity, and a part of fractions was loaded on SDS-PAGE. The active fractions were dialyzed against 20 mM sodium acetate (pH 6.0) and applied to a 0.24-ml mini S PC 3.2/3 column (GE Healthcare). The column was eluted with a 7.2-ml linear gradient of 0-0.35 M NaCl, and the presence of purified enzyme in eluted fractions was confirmed by prasugrel hydrolysis activity and SDS-PAGE.
Prasugrel Hydrolase Assay for Subcellular Localization, Gel Filtration, and Endogenous Enzyme Purification. The subcellular fraction at a final concentration of 1 mg of protein/ml or 30 ml of fractions from endogenous enzyme purification was mixed with prasugrel dimethylsulfoxide (DMSO) solution at a final concentration of 6 mM in a final volume of 50 ml of 100 mM HEPES buffer (pH 7.0). The mixture was incubated at 37°C for 15 minutes followed by adding 100 ml of methanol to terminate the reaction. The samples were filtrated by an Ultrafree-MC 0.45 mm PVDF membrane filter unit (Millipore, Billerica, MA), and 2 ml of the filtrate was injected into a liquid chromatography (LC)-mass spectrometry (MS) system to determine the concentrations of the thiolactone metabolite.

In case of fractions of gel filtration separation, 30 ml of each fraction was tested as described with slight modification; that is, fractions were tested with or without 1 mM BNPP, a specific inhibitor of human carboxyl esterase (hCE); incubation time was extended to 60 minutes. Electrophoresis. The monoQ and miniS active fractions were resolved by 5%–20% SDS-PAGE, and the gel was stained with Flamingo (Bio-Rad Laboratories, Inc.) and scanned by molecular imager FX (Bio-Rad Laboratories, Inc.) system.

Protein Identification by LC-MS/MS. For protein separation, the active fractions from the mini S PC 3.2/3 column were loaded on a 5%–20% SDS- PAGE gel. After staining of the gel with the Flamingo staining, each gel piece was excised from the gel. The gel piece was subjected to in-gel reduction and alkylation, followed by trypsin digestion (modified trypsin; Promega, Madison, WI). The resulting peptides were extracted and sequenced with LC-MS/MS on a DiNa nano-flow liquid chromatography system (KYA Tech, Tokyo, Japan) coupled to a LTQ-Orbitrap (Thermo Fisher Scientific). The mass spectrometer was operated in the data-dependent mode to automatically switch between Orbitrap-MS and LTQ MS/MS acquisitions. Orbitrap MS full scans were acquired in the Orbitrap analyzer in using lock mass recalibration in real time. Resolution in the Orbitrap MS acquisition was set to r = 15,000. The tandem mass spectra of the six most intense peptide ions with charge states $2 were collected. Data from LC-MS/MS measurements were searched against the Swiss-Plot database using the Mascot algorithm (Matrix Science, Boston, MA) with the following parameters: trypsin specificity, two missed cleavage, cysteine carbamidomethylation (fixed), protein N-term acetylated, methionine oxidation and asparagine, glutamine deamidated (variable), and electrospray ionization-ion trap fragmentation. The maximum allowed mass deviation for MS and MS/MS scans was 5 ppm and 0.8 Da, respectively. The resulting files were summarized and rearranged by an in-house developed software iMAP2.

Expression and Purification of Recombinant Proteins. The expression vector of hCE2 fused with carboxyl terminal FLAG tag was prepared as described (Ishizuka et al., 2013). The expression vector of the N-terminal FLAG- tagged enhanced green fluorescent protein (GFP) was kindly provided by Dr. Keisuke Fukuchi from Daiichi Sankyo Co., Ltd. Human RKIP was cloned from cDNA library of human 293F cells, and the expression vector was constructed with N-terminal FLAG tag under cytomegalovirus promoter as described (Kubota et al., 2015). GFP, RKIP, and hCE2 expression vectors (30 mg) were transfected into 293F cells (3 × 107 cells) using 293fectin according to the manufacturer’s protocol, respectively. The transfected cells were cultured for 72 hours. The cell culture was centrifuged, and the collected cells were suspended for 5 minutes on ice in lysis buffer [20 mM Tris-HCl (pH 7.5), 500 mM NaCl, 0.1% NP-40]. The cell extracts were centrifuged, and the supernatant was used as a cell lysate. Proteins were purified using their FLAG-tag by affinity chromatography. Five microliters of anti-FLAG M2 agarose was added to 300 ml of the lysate. After 2-hour incubation at 4°C, the resin was collected by centrifuge and washed three times with 500 ml of lysis buffer. Proteins bound to the gel were eluted with 100 ml of lysis buffer containing 0.1 mg/ml FLAG peptide. Each purified recombinant protein (i.e., hCE2, RKIP, GFP as control) was subjected to 5%–20% SDS-PAGE and visualized by Flamingo staining, which confirmed more than 90% purity (Supplemental Fig. S2). Additionally, we evaluated prasugrel hydrolysis activity with each recombinant protein.

Determination of the Enzymatic Kinetic Parameters for Thiolactone Formation from Prasugrel. The assay was performed by using recombinant RKIP and hCE2. The incubation mixture contained 1 mg of protein per milliliter of recombinant RKIP or 0.25 mg of protein per milliliter of recombinant hCE2 and 0.625, 1.25, 2.5, 5, 10, 20, 40, 80, 160, 320, and 640 mM prasugrel in a final volume of 200 ml of 0.1N HEPES buffer (pH 7.0). A mixture without prasugrel was preincubated at 37°C for 5 minutes, and the reaction was started by the addition of 2 ml of a solution of prasugrel in DMSO. After incubation at 37°C for 2 minutes, 50 ml of the incubation mixture was collected and added to 100 ml of acetonitrile and 50 ml of a solution of R-135766 as the internal standard (2 mM in acetonitrile) to terminate the reaction. The mixture was centrifuged at 15,000 rpm at 4°C for 3 minutes, and 5 ml of the supernatant was injected into LC-MS/MS system to determine the concentration of the thiolactone metabolite.

Western Blot Analysis. Western blotting was performed to determine the amount of RKIP and hCE2 in four individual human small intestinal S9.
Assay of the Contents of RKIP in Human Small Intestinal S9. Recombinant RKIP (100 mg/ml) was diluted with 0.1N HEPES buffer (pH 7.0) and prepared at the concentration of 2.5, 5, 10, 15, and 20 mg/ml. Each human small intestinal S9 (n = 4) was prepared at the concentration of 0.8 mg of protein per milliliter in the same way. These prepared samples and GFP as control were diluted twice with Laemmli sample buffer containing 5% mercaptoethanol and boiled for 5 minutes at 100°C. Ten microliters of boiled samples was loaded onto commercially available 12.5% SDS-PAGE (Ready Gel J, Bio-Rad Labora- tories, Inc.), respectively. The electrophoresis was performed at 140 V for 80 minutes. The proteins were electroblotted onto a polyvinylidene difluoride (PVDF) membrane (Bio-Rad Laboratories, Inc.) using a blotter (Trans- blot, Bio-Rad Laboratories, Inc.) at 15 V for 40 minutes. The PVDF membrane was blocked with blocking buffer (Tris-buffered saline containing 0.05% Tween 20 (TTBS) and 2% skimmed milk) for 1 hour at room temperature. After washing the PVDF membrane with TTBS, the PVDF membrane was incubated with the antibody to RKIP (dilution 1: 400) for 1 hour at room temperature. After washing the PVDF membrane with TTBS for 30 minutes, the PVDF membrane was incubated with ECL anti-rabbit IgG-HRP-linked (dilution 1: 5000) for 1 hour at room temperature. After washing with TTBS and water, the PVDF membrane was incubated with coloring reagent (Lumigen TMA-6, Western blotting kits) for 5 minutes and analyzed using Lumino-Image analyzer (LAS-4000UVmini; Fujifilm Co., Ltd., Tokyo, Japan).

Assay of the Contents of hCE2 in Human Small Intestinal S9. PNGase F-treated recombinant hCE2 (50 mg/ml) was diluted with Laemmli sample buffer containing 5% mercaptoethanol and prepared at the concentration of 1, 2.5, 5, 7.5, and 10 mg/ml. Each human small intestinal S9 (n = 4) was prepared at the concentration of 1.6 mg protein/ml with 0.1N HEPES buffer (pH 7.0). The diluted human small intestinal S9 was treated with PNGase F. PNGase F-treated human small intestinal S9 samples were diluted twice with Laemmli sample buffer containing 5% mercaptoethanol and boiled for 5 minutes at 100°C. Ten microliters of boiled samples was loaded onto commercially available 7.5%

SDS-PAGE, respectively. The details of the Western blot analysis are described herein.

In this Western blotting, the monoclonal antibody to hCE2 (dilution 1:5000) was used as the primary antibody and an ECL anti-rabbit IgG-HRP-linked antibody (dilution 1: 20000) was used as the secondary antibody.Determination of RKIP and hCE2 Contents using Lumino-Image Analyzer. The expression protein (ng/mg S9) of RKIP and hCE2 in human small intestinal S9 was determined using LAS-4000UVmini. The amount of chemiluminescence (AU) of each enzyme was computed using Multi Gauge version 3.0 software (Fujifilm Corp.). The calibration curve was constructed by linear least squares regression, plotting the amount of chemiluminescence against each recombinant concentration. The calibration curve range was from 12.5 ng/10 ml to 100 ng/10ml for RKIP and from 10.0 ng/10 ml to 100 ng/10ml for hCE2.

Inhibition Study. Human small intestine S9 fractions (20 mg) were incubated with the anti-RKIP antibody or the control rabbit antibody (20 mg) in TNN buffer (20 mM Tris (pH 8.0), 100 mM NaCl, 1% NP-40) overnight at 4°C. The immune complexes were removed by incubation with 10 ml of protein G-Sepharose (GE Healthcare). Depletion of RKIP was monitored by immu- noblotting with anti-RKIP antibodies. The depleted S9 fractions (10 ml each) were mixed with Laemmli sample buffer containing 5% b-mercaptoethanol and subjected to 5%–20% SDS-PAGE and transferred to a PVDF membrane using the iBlot dry blotting system (Invitrogen). The PVDF membrane was treated with TBS-T (20 mM Tris-HCl (pH 7.4), 150 mM NaCl, and 0.1% Tween 20) containing 10% Aqua block (Rockland) for 1 hour at room temperature. The blot was sequentially incubated with anti-RKIP antibody (1:5000 dilution) or anti-b-actin antibody (1:5000 dilution) as primary antibody and anti-mouse IgG antibody conjugated with horseradish peroxidase as secondary antibodies (1:50,000 dilution; GE Healthcare). The membrane was visualized with the ECL plus or advance (GE Healthcare) and developed using a NightOWL imaging system (Berthold Technologies GmbH, Bad Wildbad, Germany).

After that, the incubation mixture contained each depleted S9 fraction (i.e., the rabbit IgG-depleted S9 fraction, RKIP-depleted S9 fraction) (0.2 mg/ml) in a final volume of 100 ml of 50 mM HEPES buffer (pH 7.0) with or without 0.1 mM BNPP. A mixture without prasugrel was preincubated at 37°C for 2 minutes, and a reaction was started by adding 40 mM prasugrel in a DMSO solution. The reaction mixtures were incubated at 37°C for 20 minutes. A 10-ml aliquot was removed and mixed with 90 ml of methanol/H2O (1:1, v/v) to terminate the reaction; 25 ml of this sample was diluted with 75 ml of methanol/ H2O (1:1, v/v) and injected into the LC-MS/MS system to determine the concentration of thiolactone metabolite.

Assay of Thiolactone Metabolites by LC-MS/MS System. Thiolactone metabolites and R-135766 were analyzed using the LC-MS/MS system as described as follows Yeung et al., 2001. Separation by high-performance liquid chromatography was conducted using an Alliance 2695 Separations Module high-performance liquid chromatography system (Waters Corp., Milford, MA) with an Octa Decyl Silyl column (Inertsil ODS-3, 150 × 2.1 mm ID, 5.0 mm; GL Sciences, Inc., Tokyo, Japan) at a flow rate of 0.2 ml/min with a mobile phase consisting of methanol, distilled water, and trifluoroacetic acid [570/430 /0.5 (v/v/v) or 520/480 /0.5 (v/v/v)]. The column temperature was set at 40°C, and the injection volume was 2 or 5 ml. The mass spectra were determined using a Quattro LC-MS/MS system (Micromass UK Ltd., Wilmslow, UK) in positive-ion detection mode at the electrospray ionization interface. The precursor ions of the thiolactone [M + H]+ at m/z 332 and R-135766[M + H]+ at m/z 548 (for the internal standard) were obtained in the first quadrupole. After the collision-induced fragmentation in the second quadrupole, the product ion of the thiolactone metabolite [M + H]+ at m/z 149 and R-135766[M + H]+ at m/z 206 was monitored in the third quadrupole. The peak area ratio of each compound to the internal standard was computed using MassLynx version 4.0 SP4 (Waters Corp.) software.

Initially the obtained data were analyzed by an Eadie-Hofstee plot (x-axis: V/S, y-axis: V). Eadie-Hofstee plots were used to visually detect deviation from linearity. The formation of the thiolactone metabolites from prasugrel by the recombinant hCE2 indicated a non–straight line in the Eadie-Hofstee plot, suggesting involvement of the substrate inhibition kinetic properties. Therefore, the data were fitted to eq. 3 using WinNonlin Professional (version 4.0.1; Pharsight Corporation, St. Louis, MO) to calculate the Km and Vmax values. S (mM) is the substrate concentration, Km (mM) is the Michaelis-Menten constant, Vmax (pmol/min per micro- gram of protein) is the maximal reaction rate and, Ki is the inhibition constant for the substrate: On the other hand, Km and Vmax for the recombinant RKIP were calculated by using WinNonlin professional based on a pharmacodynamic compiled model (model no.101) since the best model was a Michaelis-Menten kinetics model.

Results

Identification of Several Hydrolases in Human and Monkey Small Intestine. To define subcellular localization of the hydrolase involved in the prasugrel hydrolysis reaction, we determined the prasugrel hydrolysis activity using each subcellular fraction intestinal cytosolic fraction was performed using a Superdex 75 column to confirm the characterization of expressing hydrolysis enzymes of Additionally, the remaining hydrolysis activity (%) in the inhibition study was determined by eq. 6 and eq. 7.

Protein Identification by LC-MS/MS. To identify the protein associated with prasugrel hydrolysis activity, the mini S active fractions were subjected to SDS-PAGE, and the 21-kDa band was excised and subjected to in-gel trypsin digestion. The fragmented peptides were analyzed by LC-MS/MS, identified by Mascot. Twenty-one peptides matched the amino acid coverage of 96% for monkey Raf-1 kinase inhibitory protein (RKIP) (Fig. 6).

Estimation of the Enzymatic Kinetic Parameters for Prasugrel Hydrolysis by hRKIP and hCE2. The V and V/S values of prasugrel hydrolysis for recombinant hRKIP and hCE2 were determined accord- ing to eq. 1 and eq. 2; the Eadie-Hofstee plots of these data are shown in Fig. 7. The enzyme kinetic parameters, Km and Vmax for recombinant hCE2, were estimated according to eq. 3 since the formation of thiolactone from prasugrel indicated a substrate inhibition pattern in the Eadie-Hofstee plots. On the other hand, Km and Vmax values for recombinant RKIP were calculated by Michaelis-Menten model since it fit the data best, although we tried other models such as a substrate inhibition model. The CV (%) for the parameter estimates in the Michaelis-Menten model was the smallest of the models tested.

In the formation of thiolactone metabolite from prasugrel using recombinant RKIP, Km and Vmax were 49.9 6 7.96 mM and 14,114 6 647 pmol/min per microgram of protein (Table 1). Similarly, in the case of using recombinant hCE2, Km and Vmax were 49.8 6 2.54 mM and 54,839 6 1510 pmol/min per microgram of protein (Table 1). Thiolactone formation from prasugrel in GFP was almost never detected (data not shown).

Estimation of the Contribution of hRKIP and hCE2 Involved in the Prasugrel Hydrolysis in Human Small Intestinal S9. We determined the contents of hRKIP
and hCE2 in individual human small intestinal S9 by Western blotting method (Fig. 8). As results, RKIP and hCE2 in individual human small intestinal S9 were 7.48 (ng/mg S9)–15.6 (ng/mg S9) and 2.24 (ng/mg S9)–7.91(ng/mg S9), respectively. From these data and enzyme kinetic parameters, it was estimated that the contribution ratio of RKIP and hCE2 involved in the hydrolysis reaction of prasugrel in human small intestinal S9 was 42.9% 6 9.82% (mean 6 S.D.) and 57.1% 6 9.82%, respectively (Table 2).

Inhibition Study of Prasugrel Hydrolysis in Human Small Intestinal S9. RKIP was immunodepleted from human small intestinal S9 fraction using the anti-RKIP antibody (Fig. 9A). This immunode- pletion of RKIP clearly inhibited prasugrel hydrolysis activity by about 34.7%; inhibition by BNPP was about 50.7% (Fig. 9B; Table 3).

Discussion

Prasugrel and clopidogrel are both thienopyridine-type antiplatelet prodrugs and both need to be bioactivated via a thiolactone interme- diate to their pharmacologically active metabolites (Farid et al., 2010). Prasugrel was converted more rapidly and more efficiently to the thiolactone metabolite compared with clopidogrel since for prasugrel, the thiolactone formation is via rapid ester group hydrolysis during gastrointestinal absorption. Prasugrel itself is not detected in the blood circulation, but there is rapid appearance of the thiolactone metabolite. On the other hand, conversion of clopidogrel to its thiolactone metabolite is via hepatic cytochrome P450 oxidation, including CYP2C19, which is known to have large interindividual variability. Furthermore, most of the clopidogrel undergoes hydrolysis of its ester group to form a carboxylic acid metabolite, which cannot be converted to the active metabolite (Tang et al., 2006; Hagihara et al., 2009; Kazui et al., 2010). We previously reported that prasugrel was converted to the thiolactone metabolite primarily by hCE2, with a lesser contribution by hCE1, and that at high prasugrel concentrations in excess of 109 mM, substrate inhibition was observed for hCE2 (Williams et al., 2008). The prasugrel concentrations in the gastroin- testinal tract at the doses of 2.5, 10, and 75 mg are calculated to be 26.8, 107, and 803 mM, respectively, assuming that these doses of prasugrel are dissolved in a standard glass of water (250 ml). If the metabolic enzyme involved in the formation of thiolactone in the small intestine is only hCE2, the human exposure to thiolactone metabolite might saturate at more than 10-mg doses of prasugrel; however, the human plasma exposure to the thiolactone metabolite after oral administration of prasugrel at the doses of 2.5, 10, and 75 mg were dose dependent (Asai et al., 2006), meaning that the observed inhibition in vitro does not translate to in vivo relevance. Thus, we hypothesized another enzyme other than hCE2 might contribute to the prasugrel hydrolysis in the intestine. We also tried to determine the contributions of hCE2 and unknown enzyme to the prasugrel hydrolysis process.

As first step, we found that the prasugrel hydrolysis activity was localized in the cytosolic fraction of monkey small intestine, in addition to microsome fraction, which contains most of CESs, suggesting that another prasugrel hydrolase exists in the cytosol fraction (Fig. 2). Accordingly, we compared prasugrel hydrolysis activity profiles of human intestinal S9 fraction and monkey intestinal cytosol fraction in size-exclusion column chromatography, which resulted in almost the same profiles. In both matrices, first CES activity peak and second unknown activity peak were observed (Fig. 3, A and B). Therefore, further purification of the unknown enzyme was performed using monkey small intestinal cytosolic fraction since human source availability was limited. As a result of successive four-step column chromatography purification, target enzyme protein was successfully purified. SDS-PAGE and Flamingo staining of the mini S active fractions revealed a single band of 21 kDa associated with prasugrel hydrolysis activity (Fig. 5, A and B). The fragmented peptides of the 21-kDa band were analyzed by LC-MS/MS, identified by Mascot. As a result, 21 peptides were found to match with the amino acid coverage of 96% for RKIP (Fig. 6). RKIP is a widely known protein as a member of the phosphatidylethanolamine-binding protein family. It is a small, evolutionarily conserved cytosolic protein that plays a pivotal modula- tory role in several protein kinase signaling cascades (Bazzi et al., 1992; Yeung et al., 2000; Yeung et al., 2001; Corbit et al., 2003; Lorenz et al., 2003). Additionally, has been reported that RKIP exerts a significant impact on controlling the cell cycle and is also associated with centrosomes and kinetochores in cultured mammalian cells (Al-Mulla et al., 2013). Taken together, RKIP play several roles in regulating the process of cell growth. It is already known that a lot of hydrolysis enzymes such as carboxylesterase, paraoxonase, butyrylcholinesterase, acetylcholinesterase, carboxymethylenebutenolidase, and albumin are involved in the bioconversion of ester-based prodrugs (Liedere and Borchardt, 2006; Ishizuka et al., 2013); however, it is not known that RKIP plays a role of the drug-metabolizing enzyme (i.e., hydrolase). This study is, to the best of our knowledge, the first report describing hydrolase activity of RKIP.

We determined the contributions of hRKIP and hCE2 for prasugrel hydrolysis by estimating the enzymatic kinetic parameters of recombi- nant enzymes and by the inhibition study using human small intestinal S9. The experiment for estimating the enzymatic kinetic parameters showed that both RKIP and hCE2 have similar Km values for the hydrolysis of prasugrel, suggesting that they bind prasugrel with similar affinity (Table 1), but the Vmax value for hCE2 appears to be four times higher than that for RKIP (Table 1).

From these enzyme kinetic parameters with the enzyme contents determined by Western method, the contribution ratio of RKIP and hCE2 involved in prasugrel hydrolysis in the human small intestinal S9 was estimated to be 42.9% 6 9.82% and 57.1% 6 9.82%, respectively (Table 2). Additionally, the prasugrel hydrolysis in human small intestinal S9 was inhibited about 30%–40% by anti-RKIP antibody and a further 40%–50% by BNPP (Table 3, Fig. 9B). The data from the inhibition study were consistent with the estimated contribution ratio from the enzyme kinetic parameters obtained using the recombinant enzymes. Therefore, we judged that the contribution of RKIP and hCE2 to prasugrel hydrolysis in human intestine was about 30%–40% and 60%, respectively.

In conclusion, hRKIP was identified as capable of hydrolyzing prasugrel to its thiolactone metabolite and may play a significant role with hCE2 in prasugrel bioactivation in the human intestine. RKIP is known to have diverse functions as a master modulator of many intracellular signaling cascades. This is the first report describing a Lifirafenib new function of RKIP as hydrolase involved in drug metabolism.