BIOPHARMACEUTICS & DRUG DISPOSITION Biopharm. Drug Dispos. 36: 352–363 (2015) Published online 21 April 2015 in Wiley Online Library
(wileyonlinelibrary.com) DOI: 10.1002/bdd.1945
Bioactivation of loxoprofen to a pharmacologically active metabolite and its disposition kinetics in human skin Ryoko Sawamuraa,*,†, Hidetaka Sakuraib,†, Naoya Wadab, Yumi Nishiyaa, Tomoyo Hondaa, Miho Kazuia, Atsushi Kuriharaa, Akira Shinagawab, and Takashi Izumia a
Drug Metabolism & Pharmacokinetics Research Laboratories, Daiichi Sankyo Co., Ltd, Japan Discovery Science and Technology Department, Daiichi Sankyo RD Novare Co., Ltd, Japan
b
ABSTRACT: Loxoprofen (LX) is a prodrug-type non-steroidal anti-inflammatory drug which is used not only as an oral drug but also as a transdermal formulation. As a pharmacologically active metabolite, the trans-alcohol form of LX (trans-OH form) is generated after oral administration to humans. The objectives of this study were to evaluate the generation of the trans-OH form in human in vitro skin and to identify the predominant enzyme for its generation. In the permeation and metabolism study using human in vitro skin, both the permeation of LX and the formation of the trans-OH form increased in a time- and dose-dependent manner after the application of LX gel to the skin. In addition, the characteristics of permeation and metabolism of both LX and the trans-OH form were examined by a mathematical pharmacokinetic model. The Km value was calculated to be 10.3 mM in the human in vitro skin. The predominant enzyme which generates the trans-OH form in human whole skin was identified to be carbonyl reductase 1 (CBR1) by immunodepletion using the antihuman CBR1 antibody. The results of the enzyme kinetic study using the recombinant human CBR1 protein demonstrated that the Km and Vmax values were 7.30 mM and 402 nmol/min/mg protein, respectively. In addition, it was found that no unknown metabolites were generated in the human in vitro skin. This is the first report in which LX is bioactivated to the trans-OH form in human skin by CBR1. Copyright © 2015 John Wiley & Sons, Ltd. Key words: human skin; NSAID; active metabolite; carbonyl reductase 1
Introduction Loxoprofen (LX) is a prodrug of the pharmacologically active trans-alcohol form (trans-OH form), which shows a strong inhibitory effect on prostaglandin synthetase [1]. It was initially marketed as an oral drug, with a strong anti-inflammatory effect against rheumatoid arthritis, backache, tooth pain, etc. [2]. In addition to the oral drug, a
*Correspondence to: Ryoko Sawamura, Drug Metabolism & Pharmacokinetics Research Laboratories, Daiichi Sankyo Co., Ltd, 1-2-58, Hiromachi, Shinagawa-ku, Tokyo 140-8710, Japan. E-mail: [email protected] † These authors contributed equally to this work.
Copyright © 2015 John Wiley & Sons, Ltd.
topical formulation of loxoprofen, LX gel (LX-G) (Loxonin® gel 1%, Daiichi Sankyo Co., Ltd, Tokyo, Japan) was developed [3]. It is used widely in Japan for the treatment of muscle soreness or pain due to osteoarthritis, sprains, contusions, and so forth [4,5]. As a good pharmacokinetic characteristic of dermal application, it is known that the drug concentration achieves a higher level in the tissues subjacent to the treated site after dermal application of some drugs compared with that in the systemic circulation [6]. Therefore, it is important to examine the bioactivation of LX to the trans-OH form in human skin after dermal application of LX, considering the efficacy at the treated site. Received 27 October 2014 Revised 24 February 2015 Accepted 4 March 2015
353
BIOACTIVATION OF LOXOPROFEN IN HUMAN SKIN
It is well known that human skin contains enzymes which are able to catalyse the metabolism of both endogenous chemicals (hormones, steroids inflammatory mediators, etc.) and xenobiotics, including drugs. Regarding the metabolizing enzyme of LX to the trans-OH form, Tanaka et al. reported that LX is reduced mainly to the transOH form by 3α-hydroxysteroid dehydrogenase in rabbit liver cytosol [7]. In human livers, both dihydrodiol dehydrogenase and carbonyl reductase (CBR) are reported to catalyse the generation of the trans-OH form [8]. Among the carbonyl reducing enzymes, the CBR family has been one of the enzyme families most studied in terms of the biotransformation of xenobiotics [9]. Carbonyl reductases belong to the short-chain dehydroge nase/reductase (SDR) superfamily. Regarding the CBR family, three CBRs have been found in humans: CBR1, CBR3 and CBR4 [10]. CBR1 distributes abundantly and widely in various human tissues including the skin and liver [9]. CBR3 is highly homologous with CBR1 in nucleotide and amino acid sequences, while it exhibits considerably less activity than CBR1 [11]. It has been reported that CBR4 localizes in the mitochondrial matrix in human livers and kidneys [12], however, expression of CBR4 in the human skin is still unknown. Since genetic polymorphisms of CBR1 and CBR3 in the human liver have been reported [13–16], the properties of the enzymes could become key determinants for the pharmacokinetics of the pharmacologically active metabolite and, consequently, for the efficacy of the drug. Therefore, an identification and characterization of enzymes responsible for prodrug bioactivation are important. However, an activating enzyme of LX in the human skin has not been identified so far. Considering the above, a permeation and metabolism study of LX was conducted using human in vitro skin to examine the permeability of LX and the formation of the trans-OH form. Furthermore, the results were analysed using the mathematical pharmacokinetic model to evaluate the characteristics of the permeation and metabolism of LX. In addition, we tried to identify the predominant enzyme which generates the trans-OH form in the human skin and to evaluate its enzyme kinetics. The metabolism of LX in the human skin was also investigated using 14C-labeled LX ([14C]-LX) to Copyright © 2015 John Wiley & Sons, Ltd.
determine whether any unknown metabolites are generated or not.
Materials and Methods Materials Loxoprofen was synthesized in-house. The transOH form, its stereoisomers, its stable isotope (d3-trans-OH form) and the cis-alcohol (cis-OH) form were synthesized in Chemtech Labo., Inc. (Tokyo, Japan). Their structures are shown in Figure 1. [14C]-Labeled LX (2.30 MBq/mg, radiochemical purity > 98%) was synthesized in Daiichi Pure Chemicals Co., Ltd (Ibaraki, Japan). Propyl p-hydroxybenzoate, internal standard (IS), was purchased from Wako Pure Chemical Industries, Ltd (Osaka, Japan). LX gel (LX-G) containing 1%, 3% or 5% LX was prepared by Toko Yakuhin Kogyo Co., Ltd (Osaka, Japan). TESTSKIN (3-Dimensional Cultured Human Skin Model, LSE-high) was purchased from Toyobo Co., Ltd (Osaka, Japan). TESTSKIN contains human fibroblasts in a collagen-gel matrix co-cultured with stratified human epidermal keratinocyte including a stratum corneum layer. All the skin cells for TESTSKIN were obtained from healthy neonatal male foreskin. TESTSKIN was stored in a CO2 incubator at 37 °C until use. Entry clones of human CBR1 (product ID: FLJ95780AAAF) and human carbonyl reductase 3 (CBR3, product ID: FLJ81707AAAF) were purchased from the National Institute of Technology and Evaluation (Tokyo, Japan). Recombinant human carbonyl reductase 4 (CBR4) protein fused to glutathione S-transferase protein
Figure 1. Chemical structures of loxoprofen (A), the trans-OH 14 form (B), the cis-OH form (C). * indicates the site of the C label in loxoprofen Biopharm. Drug Dispos. 36: 352–363 (2015) DOI: 10.1002/bdd
354
R. SAWAMURA ET AL.
at N-terminal (Cat. No. H00084869-P01) was purchased from Abnova (Taipei, Taiwan). Human abdominal carver skin blocks (lot number 0101-113155, 0101-113200 and 0101-112777) were collected by National Disease Research Interchange (Philadelphia, PA), and provided by the Human Animal Bridging Research Organization (Tokyo, Japan). They contained stratum corneum, epidermis and dermis. The age and gender of the donors were as follows: 0101-113155: 59 years, female; 0101-113200: 65 years, male; 0101-112777: 57 years, male. None of the subjects had any known skin disease. Ethical approval was obtained from the Research Ethics Committee at Daiichi Sankyo Co., Ltd (Tokyo, Japan) and Daiichi Sankyo RD Novare Co., Ltd (Tokyo, Japan). All reagents and solvents were of reagent grade or HPLC grade and used without further purification.
dLXskin ¼ Per Area LXabs Volgel dt
An assay medium (1.2 ml) was added to each transwell (skin tissue) of an assay plate. Regarding the detection of LX and the trans-OH form, 100 mg of 1%, 3% or 5% LX-G was applied to the transwell plate containing TESTSKIN (human in vitro skin). The human in vitro skin was incubated at 37 °C in the CO2 incubator. Medium samples of 0.4 ml each were taken at 1, 3 and 6 h after the application and the same volume of fresh medium was added. The samples (100 μl) were each mixed with 1 ml of propyl p-hydroxybenzoate (50 μg/ml-mobile phase) and 900 μl of mobile phase using a vortex mixer. The mobile phase consisted of 60% of 0.0333% phosphoric acid and 40% of acetonitrile. The samples were analysed by HPLC using method A (see Analysis). The permeation and metabolism of LX in the human in vitro skin is illustrated by the model shown in Figure 2. This model was developed based on the model of Knaak et al. [17]. The amount of LX in LX-G or in LX solution (LXabs) is described by Equation (1):
=
! (1)
where Per is the permeability constant of LX (cm/h), Area is the area of the skin to which Copyright © 2015 John Wiley & Sons, Ltd.
LX-G was applied (0.785 cm2) and Volgel is the volume of LX-G or solution of LX (0.1 cm3). Per was set to be 0.00224 cm/h based on the preliminary study (data not shown). The amount of LX in the skin (LXskin) is described by Equation (2):
=
Permeation and metabolism of LX in human in vitro skin
dLXabs ¼ Per Area LXabs Vol gel dt
Figure 2. Schematic presentation of the model for permeation and metabolism of loxoprofen in the human in vitro skin
! (2)
kpLX LXskin vmetab where kp_LX is the first-order rate constant of permeation from the skin tissue to medium, and vmetab is the metabolic rate of LX to the trans-OH form in the skin. The vmetab is described by Equation (3): V max LXskin VolTESTSKIN (3) vmetab ¼ Km þ LXskin =VolTESTSKIN where Vmax is the maximum metabolic rate, Km is the Michaelis-Menten constant and VolTESTSKIN is the volume of the skin to which LX was applied (0.0236 ml). The VolTESTSKIN was calculated using the diameter (10 mm) and thickness (0.3 mm) of the skin. The amount of the trans-OH form in the skin (Trans-OHskin) is described by Equation (4): dTrans - OH skin ¼ vmetab kpt Trans - OH skin dt (4) where kp_t is the permeation rate of the trans-OH form from the skin to the medium. After the permeation from the human in vitro skin, the amounts of LX (LXmed) and the trans-OH Biopharm. Drug Dispos. 36: 352–363 (2015) DOI: 10.1002/bdd
355
BIOACTIVATION OF LOXOPROFEN IN HUMAN SKIN
form (Trans-OH med) in the medium are described by Equations (5) and (6): dLXmed ¼ kpLX LXskin dt
(5)
dTrans - OH med ¼ kpt Trans - OH skin dt
(6)
The model analysis was performed by Phoenix WinNonlin (Ver. 6.3, Certara USA, Inc., St Louis, MO). In the model analysis, the data from the permeation and metabolism study using LX-G were used.
Clarification of metabolism of LX in human in vitro skin [14C]-LX aqueous solution containing 1% or 3% LX (100 μl) was applied to the transwell plate of TESTSKIN. The plate was incubated at 37 °C in the CO2 incubator. Medium samples of 0.4 ml each were taken at 8 h. Four transwells were used for each time point. A five-fold volume of methanol was added to the collected medium (100 μl), and sonicated for 10 min. The samples were shaken for 10 min, and centrifuged (1800 × g, 10 min, 4 °C) to separate the supernatant. The collected skin tissue was homogenized with a 4-fold volume of phosphate buffer (pH 7.0) to the wet weight and extracted in the same way as the assay medium. The extracts were evaporated to dryness under reduced pressure at room temperature and then the residue was reconstituted in 200 μl of methanol. The radioactivity in 10 μl of the reconstituted solution was measured after the addition of 10 ml of the scintillator. The residue was solubilized with 2 ml of tissue solubilizer Soluene-350 (PerkinElmer, Inc., Walthem MA). Radioactivity in two aliquots (1 ml) of each sample was measured after the addition of 10 ml of the scintillator Hionic-fluor. The medium extracts were diluted to a 20-fold volume and the skin tissue extract to a 5-fold volume. The metabolite profile was assessed by thin-layer chromatography (TLC) using method B (see Analysis).
Preparation of recombinant proteins The CBR family was selected as a candidate for an LX-reducing enzyme because CBR was reported Copyright © 2015 John Wiley & Sons, Ltd.
to catalyse the metabolism of LX in the human liver [7]. CBR1 was expressed as a fusion protein with an N-terminal FLAG tag, and CBR3 was expressed as fusion proteins of a C-terminal FLAG tag. The construction of the expression vectors was confirmed by DNA sequencing. Each expression vector was transfected into human embryonic kidney FreeStyle 293 cells (293-F cell, 3 × 107 cells, Life Technologies, Carlsbad, CA) using 293 fectin (Life Technologies, Carlsbad, CA) according to the manufacturer’s protocol. Ten micrograms of DNA was transfected for the N-terminal FLAG-tagged enhanced green fluorescent protein (EGFP) vector, or 30 μg of DNA was transfected for the other vectors. The transfected cells were cultured for 72 h in FreeStyle 293 Expression Medium (Life Technologies, Carlsbad, CA). The transfected cells were collected by centrifugation and suspended in 20 mM HEPES pH 7.5, 150 mM NaCl containing 1% 3-[(3-cholamidopropyl)dimethylammonio]-1propanesulfonate (CHAPS). After sonication, the lysates were centrifuged at 100000 × g for 60 min. The supernatants were incubated with 25 μl of anti-FLAG M2 affinity gel in 20 mM HEPES pH 7.5, 150 mM NaCl containing 0.1% CHAPS for 2 h at 4 °C on a 0.45 μm spin top filter. Then, the resins were washed five times with 20 mM HEPES pH 7.5, 150 mM NaCl containing 0.1% CHAPS. The FLAG-tagged recombinant proteins bound to the resins were eluted with 100 μg/ml FLAG peptide. Protein expression of CBR1 and CBR3 was confirmed and quantified by Western blotting. The purified proteins were loaded onto SDS-PAGE gel (4–20% gradient Mini-PROTEAN TGX precast gel, Bio-Rad Laboratories, Inc., Hercules, CA) and then electrotransferred onto a polyvinylidene difluoride membrane by iBlot (Life Technologies Corp., Carlsbad, CA). After blocking with AQUA Block (East Coast Biologics Inc., North Berwick, ME), the membrane was incubated with an antiFLAG M2 antibody conjugated with horseradish peroxidase, followed by detection using an ECL Western blotting detection system (GE Healthcare, Little Chalfont, UK). The band intensity was scanned by a NightOWL imaging system (Berthold Technologies, Bad Wildbad, Germany) and quantified using C-terminal FLAG-bacterial alkaline phosphatase as a standard protein. Biopharm. Drug Dispos. 36: 352–363 (2015) DOI: 10.1002/bdd
356
R. SAWAMURA ET AL.
To examine the LX-reducing activity of recombinant proteins, the tissue homogenate or the recombinant protein was incubated with 100 μM NADPH in 99 μl of 100 mM potassium phosphate, pH 6.0 for 5 min at 37 °C, and then 1 μl of 10 mM LX (final concentration of 100 μM LX) was added. After incubation for 30 min at 37 °C, the reaction was quenched by mixing with 2 volumes of acetonitrile. The quenched samples were analysed by LC-MS using method C (see Analysis).
Kinetic constants for the recombinant CBR1 protein were obtained by fitting experimental data to the Michaelis-Menten model (Eq. (7)) using Phoenix WinNonlin 6.3 (Certara USA, Inc., St Louis, MO). The data points represent the mean of two experiments.
Immunodepletion of CBR1 in human skin
where Activity is the rate of trans-OH form generation, Vmax is the maximum reaction rate, Km is the Michaelis-Menten constant (substrate concentration at half of Vmax) and ConcLX is the concentration of LX.
Human skin blocks were minced and homogenized in 9 volumes of homogenate buffer, 0.25 M sucrose containing 20 mM HEPES, pH 7.0 and the proteinase inhibitor cocktail (Complete, Roche Applied Science, Mannheim, Germany), using a Physcotron homogenizer (Microtec Co., Ltd, Chiba, Japan). The homogenates were centrifuged at 900 × g at 4 °C for 2 min and the supernatants were collected as skin homogenates and stored at 80 °C. Twenty microliters of the human skin homogenates was incubated with an anti-human CBR1 antibody (Abnova Corp., Taipei, Taiwan) or normal rabbit IgG together with Protein G agarose in 20 mM HEPES pH 7.5, 150 mM NaCl containing 0.1% CHAPS for 2 h at 4 °C on a 0.45 μm spin top filter. Subsequently, the mixture was filtered by centrifugation to remove the resin, and the filtrate was subjected to an LX carbonyl reductase assay. The activity was normalized by the activity of each input homogenate without the anti-human CBR1 antibody (+) and normal rabbit IgG (-).
Kinetic analysis of trans-OH form generation in recombinant human CBR1 protein CBR1 activity in the recombinant CBR1 protein (10 μg protein/ml) was measured as described below. Each recombinant protein sample was incubated with LX (final concentration: 0.5, 1, 2.5, 5, 10, 25, 50, 100 mM), β-nicotinamide-adenine dinucleotide phosphate (2.5 mM), glucose-6-phosphate (25 mM), glucose-6-phosphate dehydrogenase (0.5 units/ml) and magnesium chloride (10 mM) in 10 mM potassium phosphate buffer (pH 6.0) at 37 °C for 15 min (n = 2 for each LX concentration). The formation of the trans-OH form was analysed by LC-MS/MS using method D (see Analysis). Copyright © 2015 John Wiley & Sons, Ltd.
Activity ¼
V max ConcLX Km þ ConcLX
(7)
Analysis Method A. This analytical method was developed to determine the medium concentrations of LX and its metabolites in the study using human in vitro skin. Chromatographic separation was obtained using an LC 10Advp system (Shimadzu Corp., Kyoto, Japan) and a COSMOSIL 5C18-AR column (4.6 × 150 mm) (Nacalai Tesque, Inc., Kyoto, Japan). Detection analyses were run at a flow rate of 1 ml/min at 40 °C. The detection wavelength was set at 225 nm. The lower limit of quantification (LLOQ) and upper limit of quantification (ULOQ) of both LX and the trans-OH form were 1 μg/ml and 100 μg/ml, respectively. For the calculation of mean values, concentrations below the LLOQ were treated as zero. Values of SD were not calculated for the samples when concentrations of more than two samples were below the LLOQ. Method B. This analytical method was developed to identify the metabolite profiles of LX in the human in vitro skin. The radioactivity in the extracted samples was measured by the previously reported method [18]. TLC analysis was conducted with the mobile phase for 40–50 min. The plates were exposed in lead shield boxes for 15 h. Distributions of radioactivity on the plate corresponding to unchanged LX and radioactive metabolites were visualized and quantified using a Bio-imaging analyser (FUJIX BAS2000, Fujifilm Corp., Tokyo, Japan). Biopharm. Drug Dispos. 36: 352–363 (2015) DOI: 10.1002/bdd
357
BIOACTIVATION OF LOXOPROFEN IN HUMAN SKIN
Each plate was divided into the area corresponding to the labeled compound (radioactive area), the other area, and the background area (undeveloped area on the TLC plate) on the radioluminogram. Photo stimulated luminescence (PSL) in each area was determined, and the relative amount of labeled compound was calculated according to the equation shown below (Eq. (8)). In this equation, photo stimulated luminescence was used after subtracting the background value (PSL/mm2) corresponding to each area.
length, Agilent Technologies Inc., Santa Clara, CA). The mobile phase consisted of 33% acetonitrile in 0.1% trifluoroacetic acid. The flow rate employed was 0.3 ml/min and the column temperature was set at 50 °C. The analytes were detected by a quadrupole mass spectrometer (Agilent SL, Agilent Technologies Inc., Santa Clara, CA) equipped with an electrospray ionization source operated in the positive mode. Mass spectra were acquired in a selected ion monitoring mode of m/z 247.10 for the trans-OH form, m/z
Relative amount of labeled compound ð% on TLCÞ Photo stimulated luminescence of labeled compound ðPSLÞ ¼ 100 Total photo stimulated luminescence in areas ðPSLÞ
Radioactivity in the supernatant and residue obtained by methanol extraction of the incubation mixture was measured to calculate the ratio of radioactivity recovered in the methanol extract. The recovery of radioactivity through the sample preparation (%) was calculated according to the following equation (Eq. (9)):
245.10 for LX and m/z 247.10 for the cis-OH form. The assay enabled the detection of these analytes with a limit of quantitation of 2.5 pg and analytical time of 5 min per sample. The LLOQ and the ULOQ of LX were 1 nM and 100 μM, respectively. The LLOQ and the ULOQ of the trans-OH form were 10 nM and 100 μM, respectively.
Recovery through sample preparation ð%Þ Total radioactivity in supernatant ðdpmÞ 100 ¼ Total radioactivity in supernatant ðdpmÞ þ Total radioactivity in residue ðdpmÞ
The relative amount of each labeled compound on the TLC plate (% on TLC) was multiplied by the recovery through the sample preparation to calculate the percentage in the sample of each labeled compound. Method C. This analytical method was developed to examine the generation of the metabolite of LX after the incubation of LX with the recombinant proteins. Chromatographic separation was obtained using an Agilent Series 1100 System or an Agilent Series 1200 System (Agilent SL, Agilent Technologies Inc., Santa Clara, CA) and a reversed-phase chromatography column (Zorbax Eclipse XDB-C18 1.8 μm, 2.1 mm I.D. × 50 mm Copyright © 2015 John Wiley & Sons, Ltd.
(8)
(9)
Method D. This analytical method was developed to examine the generation of the trans-OH form after the incubation of LX with the recombinant CBR1 protein. Chromatographic separation was obtained using an Acquity UPLC system (Waters Corp., Milford, MA) and Ascentis Express C18 column (2.7 μm, 2.1 × 50 mm) (Sigma-Aldrich Corp., St Louis, MO). Mobile phase A consisted of acetonitrile /0.1 M ammonium acetate /formic acid (950:50:2, v/v/v), whereas mobile phase B consisted of distilled water /0.1 M ammonium acetate /formic acid (950:50:2, v/v/v). From 0 min to 0.2 min, 5% of mobile phase A was applied (0.3 ml/min) and then increased to 40% and maintained from 0.2 to 2.7 min (0.2 ml/min). Biopharm. Drug Dispos. 36: 352–363 (2015) DOI: 10.1002/bdd
358 The eluent was coupled to a TQ Detector (Waters Corp., Milford, MA). Mass spectra were acquired in a selected ion monitoring mode of m/z 266 > 231 for the trans-OH form and m/z 269 > 188 for d3-trans-OH form. The LLOQ and the ULOQ of both LX the trans-OH form and were 0.5 mM and 500 mM, respectively.
Results Permeation and metabolism of LX in human in vitro skin The permeation and metabolism study of LX was performed using the human in vitro skin to examine the permeation of LX and the generation of trans-OH form in the skin. The permeation and accumulation of LX increased in a time-dependent manner with a lag time (around 1 h) after the application of 1%, 3% and 5% LX-G to the human in vitro skin (Figure 3). The accumulation of LX in the medium increased in a linear manner with respect to the dose. Formation of both the trans-OH form and the cis-OH form were identified after the incubation. The formation, permeation and accumulation of the trans-OH form also increased in a time-dependent and saturable manner while increasing the dose. The accumulation amount of the trans-OH form at 6 h was from 13% to 25% of that for LX, while the accumulation amount of the cis-OH form was about 1% of that for LX at 6 h.
R. SAWAMURA ET AL.
Kinetic analysis for permeation and metabolism of LX in human in vitro skin The permeation rate and metabolic rate of LX after application of LX-G to the human in vitro skin were calculated by the model analysis (Figure 2, Table 1). The appearance rate constant of LX from the skin to the medium (kp_LX) was calculated to be 0.158 h-1. Km and Vmax for the trans-OH form production were calculated to be 10.3 mM and 0.0234 μmol/h, respectively. The appearance rate constant of the trans-OH from the skin to the medium (kp_t) was calculated to be 0.922 h-1, which was around six times larger than that of LX.
Clarification of metabolism of LX in human in vitro skin The metabolic profile of LX in the human in vitro skin was examined to investigate the generation of the unknown metabolite. The recoveries of the radioactivity from the medium and the tissue homogenates were more than 94%. In both the tissue homogenate and the medium incubated for 8 h in the human Table 1. Parameter estimates of LX and the trans-OH form in permeation and metabolism study using the human in vitro skin
kp_LX kp_t Vmax Km
-1
h -1 h μmol/h mM
Estimate
CV(%)
0.158 0.922 2.34E-02 10.3
5 44 12 27
Figure 3. Accumulation of loxoprofen (A) and the trans-OH form (B) in the medium after application of loxoprofen to the human in vitro skin. Applied drugs are as follows. □; 1% LX-G, △; 3% LX-G, ○; 5% LX-G. The lines represent the best-fit lines as predicted by the model for permeation and metabolism of loxoprofen Copyright © 2015 John Wiley & Sons, Ltd.
Biopharm. Drug Dispos. 36: 352–363 (2015) DOI: 10.1002/bdd
BIOACTIVATION OF LOXOPROFEN IN HUMAN SKIN
in vitro skin, LX, the trans-OH form and the cis-OH form were detected in this order (Figure 4, Table 2). No other metabolites except for the trans-OH form and the cis-OH form were observed in either the medium or the tissue homogenates.
Examination of recombinant CBR proteins as candidates of LX reducing enzymes To investigate whether reported CBR proteins have LX reducing activity, recombinant CBR proteins were prepared and examined. CBR1 and CBR3 were expressed as fusion proteins in the
359 mammalian cell line. The purified proteins were confirmed and quantified by a Western blot assay using an anti-FLAG antibody. Regarding human CBR4 protein, recombinant human CBR4 protein fused to GST at the N-terminal was used which was purchased from Abnova (Taipei, Taiwan). The recombinant human CBR1 and CBR3 proteins showed LX-reducing activity, while recombinant CBR4 protein did not show this activity (Figure 5). In particular, CBR1 had 15-fold higher carbonyl reductase activity for LX compared with CBR3.
Effect of CBR1 immunodepletion on generation of trans-OH form in human skin homogenates Immunodepletion experiments were attempted using an anti-human CBR1 antibody to determine the contribution of CBR1 to the generation of the trans-OH form in the human skin homogenates. As shown in Figure 6, the immunodepletion of CBR1 decreased LX-reducing activity by more than 80% in each skin homogenate from three individuals, whereas the human skin homogenates treated with normal rabbit IgG maintained 99% of the control activity. The immunodepletion of CBR1 decreased by more than 90% in recombinant human CBR1 protein. Therefore, it was demonstrated that CBR1 is the predominant LX reducing enzyme in the human skin.
Figure 4. Thin layer chromatograms of LX and its metabolites in the medium (A) and the tissue homogenate (B) after the ap14 plication of [ C]-LX (30 mg/ml as LX) to the human in vitro skin. Sampling time point for both medium and tissue were 8 h for the application Table 2. Composition of LX, the trans-OH form and the cis-OH form in the medium or tissue homogenates after the incubation 14 of [ C]-LX with the human in vitro skin for 8 h % of total radioactivity
Tissue Medium
Conc. (mg/ml)
LX
trans-OH
cis-OH
Total recovery
10 30 10 30
84.9 87.7 82.2 83.8
5.7 3.4 13.3 10.1
1.1 0.9 1.8 1.4
94.0 94.4 97.6 96.6
The average values of two samples are shown.
Copyright © 2015 John Wiley & Sons, Ltd.
Figure 5. Carbonyl reductase activities of the recombinant human CBR proteins for the trans-OH form generation. The recombinant human CBR proteins (CBR1 and CBR3) and EGFP as a negative control protein were expressed in 293-F cells and purified using an anti-FLAG antibody. CBR4 was purchased as a purified protein from Abnova Corp. LX carbonyl reductase activities of these recombinant proteins were examined. The data are shown as the average of duplicates with error bars to the minimum and maximum values (n = 2) Biopharm. Drug Dispos. 36: 352–363 (2015) DOI: 10.1002/bdd
360
R. SAWAMURA ET AL.
values for the recombinant CBR1 protein were calculated to be 7.30 mM and 402 nmol/min/mg protein, respectively.
Discussion
Figure 6. Immunodepletion of CBR1 in the human skin homogenates. The human skin homogenates (Skin 1: 0101-113155, Skin 2: 0101-113200 and Skin 3: 0101-112777) were immunodepleted with an anti-human CBR1 antibody (+) or normal rabbit IgG (-). The treated samples were used to detect CBR1 by Western blotting and to test for CBR1 activity. The activities were normalized by the activity of each input homogenate without the anti-human CBR1 antibody (+) and normal rabbit IgG (-) (Input). The data are shown as the average of duplicates with error bars to the minimum and maximum values (n = 2)
Figure 7. Enzyme kinetics for the trans-OH form generation catalysed by recombinant human CBR1 protein. Data represent the mean values of duplicate determination. Solid lines, best fit to the Michaelis-Menten equation with nonlinear least-squares regression
Enzyme kinetics of trans-OH form generation in recombinant human CBR1 protein. The kinetic study of the metabolism of LX was performed to evaluate the trans-OH form generation by CBR1 quantitatively. The generation of the trans-OH form in the recombinant CBR1 protein exhibited single-enzyme Michaelis-Menten kinetics, as shown in Figure 7. The Km and Vmax Copyright © 2015 John Wiley & Sons, Ltd.
The metabolism of loxoprofen was evaluated using the human in vitro skin and the predominant enzyme which generates the trans-OH form was identified to elucidate the bioactivation of LX in the human skin. In the study using the human in vitro skin, LX was detected in the medium time-dependently. This suggests that LX has a physicochemical property suitable for skin permeation, and LX would permeate through the human skin after the application of LX-G to the human in vivo skin. As there was a lag time between the application of LX-G and the appearance of LX in the medium, one of the factors is thought to be the permeation process of LX through the stratum corneum, which plays a role of the skin’s barrier to protect underlying tissue from infection, dehydration and chemicals [19]. Furthermore, not only LX but the trans-OH form was also detected in both tissue homogenates and medium. From this result, it was shown that the trans-OH form is produced in the human in vitro skin and permeates through it. Efficacy of the trans-OH form in the affected area (e.g. muscle, joint) is anticipated after the dermal application of LX to the human skin. In fact, the trans-OH form was determined in the plasma after the dermal application of LX-G to humans and LX-G showed the pharmacological effect after topical application to patients [3–5]. The data of the cis-OH form were not shown because the pharmacological activity of the cis-OH form was weak [1] and its concentrations were very low compared with LX and the trans-OH form as described in the Results section. TESTSKIN was used as the human in vitro skin. TESTSKIN is a three-dimensional primary cultured human skin model, which is generally used for the estimation of permeation, metabolism and toxicity in the human skin [20–22]. For instance, Shibayama et al. examined not only the permeation of disodium isostearyl 2-O-L-ascorbyl phosphate (VCP-IS-2Na), but the conversion of VCP-IS-2Na to vitamin C using TESTSKIN [22]. Biopharm. Drug Dispos. 36: 352–363 (2015) DOI: 10.1002/bdd
BIOACTIVATION OF LOXOPROFEN IN HUMAN SKIN
The pharmacokinetic model adequately described the amount–time profiles of LX and the trans-OH form in the medium. The analysis demonstrated that the permeation rate constant of the trans-OH form (kp_t) was about six times larger than that of LX (kp_LX). Since it is well known that log P is important for skin permeation [23], the difference in log P (clogP for LX: 1.97, clogP for the trans-OH form: 2.23) might result in a difference in the permeation rate constant. While the accumulation of LX in the medium increased in a linear manner with respect to the dose, the accumulation of the trans-OH form increased in a saturable manner. The reason was assumed to be because the concentration of LX in the skin tissue was higher than the Km value, though the compound concentrations in the skin tissue could not be determined due to the difficulty in the separation from endogenous substances. Km for generation of the trans-OH form in the human in vitro skin (10.3 mM) was similar to that in recombinant human CBR1 protein (7.30 mM). This provides the evidence that CBR1 is a predominant LX-reducing enzyme in human skin. Ohara et al. reported the Km value of LX in the human liver CBR and it was comparable to our results at the order level (38 mM) [8]. The prediction of the human pharmacokinetic (PK) profile of the compound is an essential part of drug development. In recent years, the use of IVIVE-PBPK linked models has increased in the pharmaceutical industries for human PK prediction [24,25]. The human PK can be predicted using the IVIVE-PBPK linked model without concern about the accuracy in the parameter estimation, although the parameter estimation performed well in our model analysis. The human PK of LX-G will be predicted using the IVIVE-PBPK linked models by adding the data (e.g. the CBR1 amount in human skin), and the predictability of the human PK examined in further studies. It has been reported that CBR catalyses the metabolism of LX in the human liver in addition to dihydrodiol dehydrogenase [8]. However, the CBR isoform was not identified in the report by Ohara et al. [8]. From the examination of the CBR1 contribution on the generation of the trans-OH form using the human whole-skin homogenate, it was demonstrated that CBR1 is the predominant LX reducing enzyme in the human Copyright © 2015 John Wiley & Sons, Ltd.
361 skin, as immunodepletion of CBR1 from human whole-skin homogenate reduced more than 80% of the LX reduction activity. Loxoprofen is considered to be bioactivated to the trans-OH form by CBR1 after LX permeates through the stratum corneum, because the trans-OH form concentrations were not detected in the stratum corneum after dermal application of LX-G to healthy male skin (data not shown). It is reported that CBR1 expresses in the outer root sheath of a human hair follicle [26], although the other location of CBR1 in human skin has not been identified yet. Further studies will be required to identify the location of CBR1 in human skin in order to further our understanding of the disposition of LX-G in humans. CBR1 catalyses the carbonyl reduction of various endogenous substrates such as prostaglandins, vitamin K, Coenzyme Q and so forth [9]. In addition to the endogenous substrates, xenobiotics such as haloperidol and anti-cancer anthracyclines (daunorubicin and doxorubicin) [27,28], belong to the substrate of CBR1. Although the metabolism of CBR1 is considered to play an important role in the detoxification of endogenous and xenobiotic ligands in the human body [9], it was found that CBR1 forms the pharmacologically active metabolite (the trans-OH form) with higher potency when compared with the parent drug (LX) [1]. Prodrugs require bioactivation by metabolizing enzymes to demonstrate their therapeutic effect. Genetic polymorphism in the prodrug activating enzymes could cause an interindividual variability in the drug exposure and response. Regarding CBR1, genetic polymorphism has been identified in previous reports. Gonzalez-Covarrubias et al. detected the CBR1 V88I polymorphism in DNA samples from individuals with African ancestry. Kinetic studies revealed that the Vmax of CBR1 I88 was more than 66% of that of CBR1 V88 for daunorubicin (CBR1 V88, 181 ± 13 versus Vmax CBR1 I88, 121 ± 12 nmol/min/mg protein) [14]. Bains et al. reported that the CBR1 wild-type enzyme has within a two-fold higher Vmax for doxorubicin compared with the CBR1 V88I [15]. Interindividual variability of CBR1 activity in human skin has not been reported yet. Although interindividual variability in the efficacy of LX-G has not been reported yet, it is believed that investigations on the interindividual variability Biopharm. Drug Dispos. 36: 352–363 (2015) DOI: 10.1002/bdd
362 of CBR1 in the human skin are important for evaluating therapeutic efficacy. It has been reported that skin metabolism might play a role in not only the formulation of active metabolites and detoxification, but also in the manifestation of local toxicities, such as skin sensitization or skin cancer [28,29]. For instance, cutaneous CYP1A enzymes metabolize polycyclic aromatic hydrocarbons, such as benzo(a) pyrene and dimethylbenz(a)anthracene, to toxic metabolites which are sufficiently reactive to bind to DNA, thereby causing skin cancer. Since a drug is highly exposed at the application site of the skin, it is possible that a larger amount of unknown metabolites tend to be generated after dermal application of a drug. Taken together, risk assessment for skin metabolism should be performed for dermal formulation. In the present study, however, the metabolic profile of LX in human in vitro skin and the medium was revealed to be the same as that in plasma after oral administration of LX to healthy volunteers without any unknown metabolites [30]. This suggests that it would not be necessary to consider toxicity derived from unknown metabolites of LX after dermal application. In addition, it is assumed that both metabolites which are generated in the skin after the dermal application of LX would be metabolized or excreted in the same way as oral administration after they move to the circulating blood.
Conclusion It was found that the permeation of loxoprofen and the generation of the trans-OH form increased dose- and time-dependently in the human in vitro skin. The characteristics of the permeation and metabolism of LX were evaluated by the analysis using the mathematical pharmacokinetic model. The predominant enzyme generating the transOH form in the human skin was identified to be CBR1 by immunodepletion using the anti-human CBR1 antibody. In addition, the Km value in recombinant human CBR1 protein was similar to that in human in vitro skin calculated by the model analysis. It was revealed that no metabolites other than the trans-OH form and the Copyright © 2015 John Wiley & Sons, Ltd.
R. SAWAMURA ET AL.
cis-OH form were formed in the human skin. This is the first report that LX is bioactivated to the trans-OH form in human skin by CBR1. Our preclinical studies supported the development of LX-G.
Acknowledgements We thank the National Disease Research Interchange and the Human Animal Bridging Research Organization for the provision of human skin blocks and the preparation of human skin homogenates. We gratefully acknowledge Hiroshi Katsuno and Mayumi Hayashi for the mass spectrometry experiments. We also appreciate Dr Kazuishi Kubota for his expert advice on writing the manuscript.
Conflict of Interest The authors have declared that there is no conflict of interest.
References 1. Matsuda K, Tanaka Y, Ushiyama S, Ohnishi K, Yamazaki M. Inhibition of prostaglandin synthesis by sodium 2-[4-(2-oxocyclopentylmethyl)phenyl] propionate dihydrate (CS-600), a new antiinflammatory drug, and its active metabolite in vitro and in vivo. Biochem Pharmacol 1984; 33: 2473–2478. 2. Honma M. Investigation of efficacy and safety in Loxonin® by new post-marketing surveillance. J Clin Ther Med 1994; 10: 2107–2120. 3. Daiichi Sankyo Co., Ltd. Investigator’s Brochure: Loxonin® Gel 1%. Tokyo, Japan. 4. Matsubara T, Arai K, Shiro Y, et al. Analgesic effects of NSAID gel containing loxoprofen (Loxonin gel 1%) in chronic neck pain females. Pain Clinic 2011; 32: 715–719. 5. Kajita Y, Yamazaki T, Sawada S, et al. Patient evaluation of the utility and usability of liniment containing loxoprofen sodium hydrate (Loxonin Gel) – study on the amount of single dose by affected site. J New Rem Clin 2012; 61: 2039–2051. 6. Heyneman CA, Lawless-Liday C, Wall GC. Oral versus topical NSAIDs in rheumatic diseases: a comparison. Drugs 2000; 60: 555–574. Biopharm. Drug Dispos. 36: 352–363 (2015) DOI: 10.1002/bdd
BIOACTIVATION OF LOXOPROFEN IN HUMAN SKIN
7. Tanaka Y, Nishikawa Y, Matsuda K, Yamazaki M, Hayashi R. Purification and some properties of ketone reductase forming an active metabolite of sodium 2-[4-(2-oxocyclopentylmethyl)-phenyl] propionate dihydrate (loxoprofen sodium), a new anti-inflammatory agent, in rabbit liver cytosol. Chem Pharm Bull (Tokyo) 1984; 32: 1040–1048. 8. Ohara H, Miyabe Y, Deyashiki Y, Matsuura K, Hara A. Reduction of drug ketones by dihydrodiol dehydrogenases, carbonyl reductase and aldehyde reductase of human liver. Biochem Pharmacol 1995; 50: 221–227. 9. Malátková P, Maser E, Wsól V. Human carbonyl reductases. Curr Drug Metab 2010; 11: 639–658. 10. Matsunaga T, Shintani S, Hara A. Multiplicity of mammalian reductases for xenobiotic carbonyl compounds. Drug Metab Pharmacokinet 2006; 21: 1–18. 11. Miura T, Nishinaka T, Terada T. Different functions between human monomeric carbonyl reductase 3 and carbonyl reductase 1. Mol Cell Biochem 2008; 315: 113–121. 12. Endo S, Matsunaga T, Kitade Y, et al. Human carbonyl reductase 4 is a mitochondrial NADPHdependent quinone reductase. Biochem Biophys Res Commun 2008; 377: 1326–1330. 13. Lakhman SS, Ghosh D, Blanco JG. Functional significance of a natural allelic variant of human carbonyl reductase 3 (CBR3). Drug Metab Dispos 2005; 33: 254–257. 14. Gonzalez-Covarrubias V, Ghosh D, Lakhman SS, Pendyala L, Blanco JG. A functional genetic polymorphism on human carbonyl reductase 1 (CBR1 V88I) impacts on catalytic activity and NADPH binding affinity. Drug Metab Dispos 2007; 35: 973–980. 15. Bains OS, Karkling MJ, Grigliatti TA, Reid RE, Riggs KW. Two nonsynonymous single nucleotide polymorphisms of human carbonyl reductase 1 demonstrate reduced in vitro metabolism of daunorubicin and doxorubicin. Drug Metab Dispos 2009; 37: 1107–1114. 16. Zhang J, Blanco JG. Identification of the promoter of human carbonyl reductase 3 (CBR3) and impact of common promoter polymorphisms on hepatic CBR3 mRNA expression. Pharm Res 2009; 26: 2209–2215. 17. Knaak B, Al-Bayati M, Raabe OG, Blancato JN. In vivo percutaneous absorption studies in the rat: pharmacokinetics and modeling of isofenphos absorption. In Prediction of Percutaneous Penetration, Scott RC (ed.). IBC Technical Services Ltd, London, 1991; 1–18.
Copyright © 2015 John Wiley & Sons, Ltd.
363 18. Sawamura R, Kazui M, Kurihara A, Izumi T. Absorption, distribution, metabolism and excretion of loxoprofen after dermal application of loxoprofen gel to rats. Xenobiotica 2014; 11: 1–13. 19. Trommer H, Neubert RH. Overcoming the stratum corneum: the modulation of skin penetration. A review. Skin Pharmacol Physiol 2006; 19: 106–121. 20. Osborne R, Perkins MA. An approach for development of alternative test methods based on mechanisms of skin irritation. Food Chem Toxicol 1994; 32: 133–142. 21. Andronis V, Mesiha MS, Plakogiannis FM. Design and evaluation of transdermal chlorpheniramine maleate drug delivery system. Pharm Acta Helv 1995; 70: 301–306. 22. Shibayama H, Hisama M, Matsuda S, Ohtsuki M. Permeation and metabolism of a novel ascorbic acid derivative, disodium isostearyl 2-O-L-ascorbyl phosphate, in human living skin equivalent models. Skin Pharmacol Physiol 2008; 21: 235–243. 23. Katritzky AR, Dobchev DA, Fara DC, et al. Skin permeation rate as a function of chemical structure. J Med Chem 2006; 49: 3305–3314. 24. Rostami-Hodjegan A. Physiologically based pharmacokinetics joined with in vitro-in vivo extrapolation of ADME: a marriage under the arch of systems pharmacology. Clin Pharmacol Ther 2012; 92: 50–61. 25. Chen Y, Jin JY, Mukadam S, Malhi V, Kenny JR. Application of IVIVE and PBPK modeling in prospective prediction of clinical pharmacokinetics: strategy and approach during the drug discovery phase with four case studies. Biopharm Drug Dispos 2012; 33: 85–98. 26. Colombe L, Vindrios A, Michelet JF, Bernard BA. Prostaglandin metabolism in human hair follicle. Exp Dermatol 2007; 16: 762–769. 27. De Jong J, Schoofs PR, Snabilié AM, Bast A, van der Vijgh WJ. The role of biotransformation in anthracycline-induced cardiotoxicity in mice. J Pharmacol Exp Ther 1993; 266: 1312–1320. 28. Hotchkiss SAM. Dermal metabolism. In Dermal Absorption and Toxicity Assessment, Roberts MS, Walter KA (eds). Marcel Dekker, New York, (Drug and the Pharmaceutical Sciences Vol. 91). 1998; 43–101. 29. Svensson CK. Biotransformation of drugs in human skin. Drug Metab Dispos 2009; 37: 247–253. 30. Daiichi Sankyo Co., Ltd. Investigator’s Brochure: Loxonin®. Tokyo, Japan. Website: http://www. info.pmda.go.jp/ippan/O200900004/430549000_ 22200APX00095000_V100_1.pdf
Biopharm. Drug Dispos. 36: 352–363 (2015) DOI: 10.1002/bdd
(wileyonlinelibrary.com) DOI: 10.1002/bdd.1945
Bioactivation of loxoprofen to a pharmacologically active metabolite and its disposition kinetics in human skin Ryoko Sawamuraa,*,†, Hidetaka Sakuraib,†, Naoya Wadab, Yumi Nishiyaa, Tomoyo Hondaa, Miho Kazuia, Atsushi Kuriharaa, Akira Shinagawab, and Takashi Izumia a
Drug Metabolism & Pharmacokinetics Research Laboratories, Daiichi Sankyo Co., Ltd, Japan Discovery Science and Technology Department, Daiichi Sankyo RD Novare Co., Ltd, Japan
b
ABSTRACT: Loxoprofen (LX) is a prodrug-type non-steroidal anti-inflammatory drug which is used not only as an oral drug but also as a transdermal formulation. As a pharmacologically active metabolite, the trans-alcohol form of LX (trans-OH form) is generated after oral administration to humans. The objectives of this study were to evaluate the generation of the trans-OH form in human in vitro skin and to identify the predominant enzyme for its generation. In the permeation and metabolism study using human in vitro skin, both the permeation of LX and the formation of the trans-OH form increased in a time- and dose-dependent manner after the application of LX gel to the skin. In addition, the characteristics of permeation and metabolism of both LX and the trans-OH form were examined by a mathematical pharmacokinetic model. The Km value was calculated to be 10.3 mM in the human in vitro skin. The predominant enzyme which generates the trans-OH form in human whole skin was identified to be carbonyl reductase 1 (CBR1) by immunodepletion using the antihuman CBR1 antibody. The results of the enzyme kinetic study using the recombinant human CBR1 protein demonstrated that the Km and Vmax values were 7.30 mM and 402 nmol/min/mg protein, respectively. In addition, it was found that no unknown metabolites were generated in the human in vitro skin. This is the first report in which LX is bioactivated to the trans-OH form in human skin by CBR1. Copyright © 2015 John Wiley & Sons, Ltd. Key words: human skin; NSAID; active metabolite; carbonyl reductase 1
Introduction Loxoprofen (LX) is a prodrug of the pharmacologically active trans-alcohol form (trans-OH form), which shows a strong inhibitory effect on prostaglandin synthetase [1]. It was initially marketed as an oral drug, with a strong anti-inflammatory effect against rheumatoid arthritis, backache, tooth pain, etc. [2]. In addition to the oral drug, a
*Correspondence to: Ryoko Sawamura, Drug Metabolism & Pharmacokinetics Research Laboratories, Daiichi Sankyo Co., Ltd, 1-2-58, Hiromachi, Shinagawa-ku, Tokyo 140-8710, Japan. E-mail: [email protected] † These authors contributed equally to this work.
Copyright © 2015 John Wiley & Sons, Ltd.
topical formulation of loxoprofen, LX gel (LX-G) (Loxonin® gel 1%, Daiichi Sankyo Co., Ltd, Tokyo, Japan) was developed [3]. It is used widely in Japan for the treatment of muscle soreness or pain due to osteoarthritis, sprains, contusions, and so forth [4,5]. As a good pharmacokinetic characteristic of dermal application, it is known that the drug concentration achieves a higher level in the tissues subjacent to the treated site after dermal application of some drugs compared with that in the systemic circulation [6]. Therefore, it is important to examine the bioactivation of LX to the trans-OH form in human skin after dermal application of LX, considering the efficacy at the treated site. Received 27 October 2014 Revised 24 February 2015 Accepted 4 March 2015
353
BIOACTIVATION OF LOXOPROFEN IN HUMAN SKIN
It is well known that human skin contains enzymes which are able to catalyse the metabolism of both endogenous chemicals (hormones, steroids inflammatory mediators, etc.) and xenobiotics, including drugs. Regarding the metabolizing enzyme of LX to the trans-OH form, Tanaka et al. reported that LX is reduced mainly to the transOH form by 3α-hydroxysteroid dehydrogenase in rabbit liver cytosol [7]. In human livers, both dihydrodiol dehydrogenase and carbonyl reductase (CBR) are reported to catalyse the generation of the trans-OH form [8]. Among the carbonyl reducing enzymes, the CBR family has been one of the enzyme families most studied in terms of the biotransformation of xenobiotics [9]. Carbonyl reductases belong to the short-chain dehydroge nase/reductase (SDR) superfamily. Regarding the CBR family, three CBRs have been found in humans: CBR1, CBR3 and CBR4 [10]. CBR1 distributes abundantly and widely in various human tissues including the skin and liver [9]. CBR3 is highly homologous with CBR1 in nucleotide and amino acid sequences, while it exhibits considerably less activity than CBR1 [11]. It has been reported that CBR4 localizes in the mitochondrial matrix in human livers and kidneys [12], however, expression of CBR4 in the human skin is still unknown. Since genetic polymorphisms of CBR1 and CBR3 in the human liver have been reported [13–16], the properties of the enzymes could become key determinants for the pharmacokinetics of the pharmacologically active metabolite and, consequently, for the efficacy of the drug. Therefore, an identification and characterization of enzymes responsible for prodrug bioactivation are important. However, an activating enzyme of LX in the human skin has not been identified so far. Considering the above, a permeation and metabolism study of LX was conducted using human in vitro skin to examine the permeability of LX and the formation of the trans-OH form. Furthermore, the results were analysed using the mathematical pharmacokinetic model to evaluate the characteristics of the permeation and metabolism of LX. In addition, we tried to identify the predominant enzyme which generates the trans-OH form in the human skin and to evaluate its enzyme kinetics. The metabolism of LX in the human skin was also investigated using 14C-labeled LX ([14C]-LX) to Copyright © 2015 John Wiley & Sons, Ltd.
determine whether any unknown metabolites are generated or not.
Materials and Methods Materials Loxoprofen was synthesized in-house. The transOH form, its stereoisomers, its stable isotope (d3-trans-OH form) and the cis-alcohol (cis-OH) form were synthesized in Chemtech Labo., Inc. (Tokyo, Japan). Their structures are shown in Figure 1. [14C]-Labeled LX (2.30 MBq/mg, radiochemical purity > 98%) was synthesized in Daiichi Pure Chemicals Co., Ltd (Ibaraki, Japan). Propyl p-hydroxybenzoate, internal standard (IS), was purchased from Wako Pure Chemical Industries, Ltd (Osaka, Japan). LX gel (LX-G) containing 1%, 3% or 5% LX was prepared by Toko Yakuhin Kogyo Co., Ltd (Osaka, Japan). TESTSKIN (3-Dimensional Cultured Human Skin Model, LSE-high) was purchased from Toyobo Co., Ltd (Osaka, Japan). TESTSKIN contains human fibroblasts in a collagen-gel matrix co-cultured with stratified human epidermal keratinocyte including a stratum corneum layer. All the skin cells for TESTSKIN were obtained from healthy neonatal male foreskin. TESTSKIN was stored in a CO2 incubator at 37 °C until use. Entry clones of human CBR1 (product ID: FLJ95780AAAF) and human carbonyl reductase 3 (CBR3, product ID: FLJ81707AAAF) were purchased from the National Institute of Technology and Evaluation (Tokyo, Japan). Recombinant human carbonyl reductase 4 (CBR4) protein fused to glutathione S-transferase protein
Figure 1. Chemical structures of loxoprofen (A), the trans-OH 14 form (B), the cis-OH form (C). * indicates the site of the C label in loxoprofen Biopharm. Drug Dispos. 36: 352–363 (2015) DOI: 10.1002/bdd
354
R. SAWAMURA ET AL.
at N-terminal (Cat. No. H00084869-P01) was purchased from Abnova (Taipei, Taiwan). Human abdominal carver skin blocks (lot number 0101-113155, 0101-113200 and 0101-112777) were collected by National Disease Research Interchange (Philadelphia, PA), and provided by the Human Animal Bridging Research Organization (Tokyo, Japan). They contained stratum corneum, epidermis and dermis. The age and gender of the donors were as follows: 0101-113155: 59 years, female; 0101-113200: 65 years, male; 0101-112777: 57 years, male. None of the subjects had any known skin disease. Ethical approval was obtained from the Research Ethics Committee at Daiichi Sankyo Co., Ltd (Tokyo, Japan) and Daiichi Sankyo RD Novare Co., Ltd (Tokyo, Japan). All reagents and solvents were of reagent grade or HPLC grade and used without further purification.
dLXskin ¼ Per Area LXabs Volgel dt
An assay medium (1.2 ml) was added to each transwell (skin tissue) of an assay plate. Regarding the detection of LX and the trans-OH form, 100 mg of 1%, 3% or 5% LX-G was applied to the transwell plate containing TESTSKIN (human in vitro skin). The human in vitro skin was incubated at 37 °C in the CO2 incubator. Medium samples of 0.4 ml each were taken at 1, 3 and 6 h after the application and the same volume of fresh medium was added. The samples (100 μl) were each mixed with 1 ml of propyl p-hydroxybenzoate (50 μg/ml-mobile phase) and 900 μl of mobile phase using a vortex mixer. The mobile phase consisted of 60% of 0.0333% phosphoric acid and 40% of acetonitrile. The samples were analysed by HPLC using method A (see Analysis). The permeation and metabolism of LX in the human in vitro skin is illustrated by the model shown in Figure 2. This model was developed based on the model of Knaak et al. [17]. The amount of LX in LX-G or in LX solution (LXabs) is described by Equation (1):
=
! (1)
where Per is the permeability constant of LX (cm/h), Area is the area of the skin to which Copyright © 2015 John Wiley & Sons, Ltd.
LX-G was applied (0.785 cm2) and Volgel is the volume of LX-G or solution of LX (0.1 cm3). Per was set to be 0.00224 cm/h based on the preliminary study (data not shown). The amount of LX in the skin (LXskin) is described by Equation (2):
=
Permeation and metabolism of LX in human in vitro skin
dLXabs ¼ Per Area LXabs Vol gel dt
Figure 2. Schematic presentation of the model for permeation and metabolism of loxoprofen in the human in vitro skin
! (2)
kpLX LXskin vmetab where kp_LX is the first-order rate constant of permeation from the skin tissue to medium, and vmetab is the metabolic rate of LX to the trans-OH form in the skin. The vmetab is described by Equation (3): V max LXskin VolTESTSKIN (3) vmetab ¼ Km þ LXskin =VolTESTSKIN where Vmax is the maximum metabolic rate, Km is the Michaelis-Menten constant and VolTESTSKIN is the volume of the skin to which LX was applied (0.0236 ml). The VolTESTSKIN was calculated using the diameter (10 mm) and thickness (0.3 mm) of the skin. The amount of the trans-OH form in the skin (Trans-OHskin) is described by Equation (4): dTrans - OH skin ¼ vmetab kpt Trans - OH skin dt (4) where kp_t is the permeation rate of the trans-OH form from the skin to the medium. After the permeation from the human in vitro skin, the amounts of LX (LXmed) and the trans-OH Biopharm. Drug Dispos. 36: 352–363 (2015) DOI: 10.1002/bdd
355
BIOACTIVATION OF LOXOPROFEN IN HUMAN SKIN
form (Trans-OH med) in the medium are described by Equations (5) and (6): dLXmed ¼ kpLX LXskin dt
(5)
dTrans - OH med ¼ kpt Trans - OH skin dt
(6)
The model analysis was performed by Phoenix WinNonlin (Ver. 6.3, Certara USA, Inc., St Louis, MO). In the model analysis, the data from the permeation and metabolism study using LX-G were used.
Clarification of metabolism of LX in human in vitro skin [14C]-LX aqueous solution containing 1% or 3% LX (100 μl) was applied to the transwell plate of TESTSKIN. The plate was incubated at 37 °C in the CO2 incubator. Medium samples of 0.4 ml each were taken at 8 h. Four transwells were used for each time point. A five-fold volume of methanol was added to the collected medium (100 μl), and sonicated for 10 min. The samples were shaken for 10 min, and centrifuged (1800 × g, 10 min, 4 °C) to separate the supernatant. The collected skin tissue was homogenized with a 4-fold volume of phosphate buffer (pH 7.0) to the wet weight and extracted in the same way as the assay medium. The extracts were evaporated to dryness under reduced pressure at room temperature and then the residue was reconstituted in 200 μl of methanol. The radioactivity in 10 μl of the reconstituted solution was measured after the addition of 10 ml of the scintillator. The residue was solubilized with 2 ml of tissue solubilizer Soluene-350 (PerkinElmer, Inc., Walthem MA). Radioactivity in two aliquots (1 ml) of each sample was measured after the addition of 10 ml of the scintillator Hionic-fluor. The medium extracts were diluted to a 20-fold volume and the skin tissue extract to a 5-fold volume. The metabolite profile was assessed by thin-layer chromatography (TLC) using method B (see Analysis).
Preparation of recombinant proteins The CBR family was selected as a candidate for an LX-reducing enzyme because CBR was reported Copyright © 2015 John Wiley & Sons, Ltd.
to catalyse the metabolism of LX in the human liver [7]. CBR1 was expressed as a fusion protein with an N-terminal FLAG tag, and CBR3 was expressed as fusion proteins of a C-terminal FLAG tag. The construction of the expression vectors was confirmed by DNA sequencing. Each expression vector was transfected into human embryonic kidney FreeStyle 293 cells (293-F cell, 3 × 107 cells, Life Technologies, Carlsbad, CA) using 293 fectin (Life Technologies, Carlsbad, CA) according to the manufacturer’s protocol. Ten micrograms of DNA was transfected for the N-terminal FLAG-tagged enhanced green fluorescent protein (EGFP) vector, or 30 μg of DNA was transfected for the other vectors. The transfected cells were cultured for 72 h in FreeStyle 293 Expression Medium (Life Technologies, Carlsbad, CA). The transfected cells were collected by centrifugation and suspended in 20 mM HEPES pH 7.5, 150 mM NaCl containing 1% 3-[(3-cholamidopropyl)dimethylammonio]-1propanesulfonate (CHAPS). After sonication, the lysates were centrifuged at 100000 × g for 60 min. The supernatants were incubated with 25 μl of anti-FLAG M2 affinity gel in 20 mM HEPES pH 7.5, 150 mM NaCl containing 0.1% CHAPS for 2 h at 4 °C on a 0.45 μm spin top filter. Then, the resins were washed five times with 20 mM HEPES pH 7.5, 150 mM NaCl containing 0.1% CHAPS. The FLAG-tagged recombinant proteins bound to the resins were eluted with 100 μg/ml FLAG peptide. Protein expression of CBR1 and CBR3 was confirmed and quantified by Western blotting. The purified proteins were loaded onto SDS-PAGE gel (4–20% gradient Mini-PROTEAN TGX precast gel, Bio-Rad Laboratories, Inc., Hercules, CA) and then electrotransferred onto a polyvinylidene difluoride membrane by iBlot (Life Technologies Corp., Carlsbad, CA). After blocking with AQUA Block (East Coast Biologics Inc., North Berwick, ME), the membrane was incubated with an antiFLAG M2 antibody conjugated with horseradish peroxidase, followed by detection using an ECL Western blotting detection system (GE Healthcare, Little Chalfont, UK). The band intensity was scanned by a NightOWL imaging system (Berthold Technologies, Bad Wildbad, Germany) and quantified using C-terminal FLAG-bacterial alkaline phosphatase as a standard protein. Biopharm. Drug Dispos. 36: 352–363 (2015) DOI: 10.1002/bdd
356
R. SAWAMURA ET AL.
To examine the LX-reducing activity of recombinant proteins, the tissue homogenate or the recombinant protein was incubated with 100 μM NADPH in 99 μl of 100 mM potassium phosphate, pH 6.0 for 5 min at 37 °C, and then 1 μl of 10 mM LX (final concentration of 100 μM LX) was added. After incubation for 30 min at 37 °C, the reaction was quenched by mixing with 2 volumes of acetonitrile. The quenched samples were analysed by LC-MS using method C (see Analysis).
Kinetic constants for the recombinant CBR1 protein were obtained by fitting experimental data to the Michaelis-Menten model (Eq. (7)) using Phoenix WinNonlin 6.3 (Certara USA, Inc., St Louis, MO). The data points represent the mean of two experiments.
Immunodepletion of CBR1 in human skin
where Activity is the rate of trans-OH form generation, Vmax is the maximum reaction rate, Km is the Michaelis-Menten constant (substrate concentration at half of Vmax) and ConcLX is the concentration of LX.
Human skin blocks were minced and homogenized in 9 volumes of homogenate buffer, 0.25 M sucrose containing 20 mM HEPES, pH 7.0 and the proteinase inhibitor cocktail (Complete, Roche Applied Science, Mannheim, Germany), using a Physcotron homogenizer (Microtec Co., Ltd, Chiba, Japan). The homogenates were centrifuged at 900 × g at 4 °C for 2 min and the supernatants were collected as skin homogenates and stored at 80 °C. Twenty microliters of the human skin homogenates was incubated with an anti-human CBR1 antibody (Abnova Corp., Taipei, Taiwan) or normal rabbit IgG together with Protein G agarose in 20 mM HEPES pH 7.5, 150 mM NaCl containing 0.1% CHAPS for 2 h at 4 °C on a 0.45 μm spin top filter. Subsequently, the mixture was filtered by centrifugation to remove the resin, and the filtrate was subjected to an LX carbonyl reductase assay. The activity was normalized by the activity of each input homogenate without the anti-human CBR1 antibody (+) and normal rabbit IgG (-).
Kinetic analysis of trans-OH form generation in recombinant human CBR1 protein CBR1 activity in the recombinant CBR1 protein (10 μg protein/ml) was measured as described below. Each recombinant protein sample was incubated with LX (final concentration: 0.5, 1, 2.5, 5, 10, 25, 50, 100 mM), β-nicotinamide-adenine dinucleotide phosphate (2.5 mM), glucose-6-phosphate (25 mM), glucose-6-phosphate dehydrogenase (0.5 units/ml) and magnesium chloride (10 mM) in 10 mM potassium phosphate buffer (pH 6.0) at 37 °C for 15 min (n = 2 for each LX concentration). The formation of the trans-OH form was analysed by LC-MS/MS using method D (see Analysis). Copyright © 2015 John Wiley & Sons, Ltd.
Activity ¼
V max ConcLX Km þ ConcLX
(7)
Analysis Method A. This analytical method was developed to determine the medium concentrations of LX and its metabolites in the study using human in vitro skin. Chromatographic separation was obtained using an LC 10Advp system (Shimadzu Corp., Kyoto, Japan) and a COSMOSIL 5C18-AR column (4.6 × 150 mm) (Nacalai Tesque, Inc., Kyoto, Japan). Detection analyses were run at a flow rate of 1 ml/min at 40 °C. The detection wavelength was set at 225 nm. The lower limit of quantification (LLOQ) and upper limit of quantification (ULOQ) of both LX and the trans-OH form were 1 μg/ml and 100 μg/ml, respectively. For the calculation of mean values, concentrations below the LLOQ were treated as zero. Values of SD were not calculated for the samples when concentrations of more than two samples were below the LLOQ. Method B. This analytical method was developed to identify the metabolite profiles of LX in the human in vitro skin. The radioactivity in the extracted samples was measured by the previously reported method [18]. TLC analysis was conducted with the mobile phase for 40–50 min. The plates were exposed in lead shield boxes for 15 h. Distributions of radioactivity on the plate corresponding to unchanged LX and radioactive metabolites were visualized and quantified using a Bio-imaging analyser (FUJIX BAS2000, Fujifilm Corp., Tokyo, Japan). Biopharm. Drug Dispos. 36: 352–363 (2015) DOI: 10.1002/bdd
357
BIOACTIVATION OF LOXOPROFEN IN HUMAN SKIN
Each plate was divided into the area corresponding to the labeled compound (radioactive area), the other area, and the background area (undeveloped area on the TLC plate) on the radioluminogram. Photo stimulated luminescence (PSL) in each area was determined, and the relative amount of labeled compound was calculated according to the equation shown below (Eq. (8)). In this equation, photo stimulated luminescence was used after subtracting the background value (PSL/mm2) corresponding to each area.
length, Agilent Technologies Inc., Santa Clara, CA). The mobile phase consisted of 33% acetonitrile in 0.1% trifluoroacetic acid. The flow rate employed was 0.3 ml/min and the column temperature was set at 50 °C. The analytes were detected by a quadrupole mass spectrometer (Agilent SL, Agilent Technologies Inc., Santa Clara, CA) equipped with an electrospray ionization source operated in the positive mode. Mass spectra were acquired in a selected ion monitoring mode of m/z 247.10 for the trans-OH form, m/z
Relative amount of labeled compound ð% on TLCÞ Photo stimulated luminescence of labeled compound ðPSLÞ ¼ 100 Total photo stimulated luminescence in areas ðPSLÞ
Radioactivity in the supernatant and residue obtained by methanol extraction of the incubation mixture was measured to calculate the ratio of radioactivity recovered in the methanol extract. The recovery of radioactivity through the sample preparation (%) was calculated according to the following equation (Eq. (9)):
245.10 for LX and m/z 247.10 for the cis-OH form. The assay enabled the detection of these analytes with a limit of quantitation of 2.5 pg and analytical time of 5 min per sample. The LLOQ and the ULOQ of LX were 1 nM and 100 μM, respectively. The LLOQ and the ULOQ of the trans-OH form were 10 nM and 100 μM, respectively.
Recovery through sample preparation ð%Þ Total radioactivity in supernatant ðdpmÞ 100 ¼ Total radioactivity in supernatant ðdpmÞ þ Total radioactivity in residue ðdpmÞ
The relative amount of each labeled compound on the TLC plate (% on TLC) was multiplied by the recovery through the sample preparation to calculate the percentage in the sample of each labeled compound. Method C. This analytical method was developed to examine the generation of the metabolite of LX after the incubation of LX with the recombinant proteins. Chromatographic separation was obtained using an Agilent Series 1100 System or an Agilent Series 1200 System (Agilent SL, Agilent Technologies Inc., Santa Clara, CA) and a reversed-phase chromatography column (Zorbax Eclipse XDB-C18 1.8 μm, 2.1 mm I.D. × 50 mm Copyright © 2015 John Wiley & Sons, Ltd.
(8)
(9)
Method D. This analytical method was developed to examine the generation of the trans-OH form after the incubation of LX with the recombinant CBR1 protein. Chromatographic separation was obtained using an Acquity UPLC system (Waters Corp., Milford, MA) and Ascentis Express C18 column (2.7 μm, 2.1 × 50 mm) (Sigma-Aldrich Corp., St Louis, MO). Mobile phase A consisted of acetonitrile /0.1 M ammonium acetate /formic acid (950:50:2, v/v/v), whereas mobile phase B consisted of distilled water /0.1 M ammonium acetate /formic acid (950:50:2, v/v/v). From 0 min to 0.2 min, 5% of mobile phase A was applied (0.3 ml/min) and then increased to 40% and maintained from 0.2 to 2.7 min (0.2 ml/min). Biopharm. Drug Dispos. 36: 352–363 (2015) DOI: 10.1002/bdd
358 The eluent was coupled to a TQ Detector (Waters Corp., Milford, MA). Mass spectra were acquired in a selected ion monitoring mode of m/z 266 > 231 for the trans-OH form and m/z 269 > 188 for d3-trans-OH form. The LLOQ and the ULOQ of both LX the trans-OH form and were 0.5 mM and 500 mM, respectively.
Results Permeation and metabolism of LX in human in vitro skin The permeation and metabolism study of LX was performed using the human in vitro skin to examine the permeation of LX and the generation of trans-OH form in the skin. The permeation and accumulation of LX increased in a time-dependent manner with a lag time (around 1 h) after the application of 1%, 3% and 5% LX-G to the human in vitro skin (Figure 3). The accumulation of LX in the medium increased in a linear manner with respect to the dose. Formation of both the trans-OH form and the cis-OH form were identified after the incubation. The formation, permeation and accumulation of the trans-OH form also increased in a time-dependent and saturable manner while increasing the dose. The accumulation amount of the trans-OH form at 6 h was from 13% to 25% of that for LX, while the accumulation amount of the cis-OH form was about 1% of that for LX at 6 h.
R. SAWAMURA ET AL.
Kinetic analysis for permeation and metabolism of LX in human in vitro skin The permeation rate and metabolic rate of LX after application of LX-G to the human in vitro skin were calculated by the model analysis (Figure 2, Table 1). The appearance rate constant of LX from the skin to the medium (kp_LX) was calculated to be 0.158 h-1. Km and Vmax for the trans-OH form production were calculated to be 10.3 mM and 0.0234 μmol/h, respectively. The appearance rate constant of the trans-OH from the skin to the medium (kp_t) was calculated to be 0.922 h-1, which was around six times larger than that of LX.
Clarification of metabolism of LX in human in vitro skin The metabolic profile of LX in the human in vitro skin was examined to investigate the generation of the unknown metabolite. The recoveries of the radioactivity from the medium and the tissue homogenates were more than 94%. In both the tissue homogenate and the medium incubated for 8 h in the human Table 1. Parameter estimates of LX and the trans-OH form in permeation and metabolism study using the human in vitro skin
kp_LX kp_t Vmax Km
-1
h -1 h μmol/h mM
Estimate
CV(%)
0.158 0.922 2.34E-02 10.3
5 44 12 27
Figure 3. Accumulation of loxoprofen (A) and the trans-OH form (B) in the medium after application of loxoprofen to the human in vitro skin. Applied drugs are as follows. □; 1% LX-G, △; 3% LX-G, ○; 5% LX-G. The lines represent the best-fit lines as predicted by the model for permeation and metabolism of loxoprofen Copyright © 2015 John Wiley & Sons, Ltd.
Biopharm. Drug Dispos. 36: 352–363 (2015) DOI: 10.1002/bdd
BIOACTIVATION OF LOXOPROFEN IN HUMAN SKIN
in vitro skin, LX, the trans-OH form and the cis-OH form were detected in this order (Figure 4, Table 2). No other metabolites except for the trans-OH form and the cis-OH form were observed in either the medium or the tissue homogenates.
Examination of recombinant CBR proteins as candidates of LX reducing enzymes To investigate whether reported CBR proteins have LX reducing activity, recombinant CBR proteins were prepared and examined. CBR1 and CBR3 were expressed as fusion proteins in the
359 mammalian cell line. The purified proteins were confirmed and quantified by a Western blot assay using an anti-FLAG antibody. Regarding human CBR4 protein, recombinant human CBR4 protein fused to GST at the N-terminal was used which was purchased from Abnova (Taipei, Taiwan). The recombinant human CBR1 and CBR3 proteins showed LX-reducing activity, while recombinant CBR4 protein did not show this activity (Figure 5). In particular, CBR1 had 15-fold higher carbonyl reductase activity for LX compared with CBR3.
Effect of CBR1 immunodepletion on generation of trans-OH form in human skin homogenates Immunodepletion experiments were attempted using an anti-human CBR1 antibody to determine the contribution of CBR1 to the generation of the trans-OH form in the human skin homogenates. As shown in Figure 6, the immunodepletion of CBR1 decreased LX-reducing activity by more than 80% in each skin homogenate from three individuals, whereas the human skin homogenates treated with normal rabbit IgG maintained 99% of the control activity. The immunodepletion of CBR1 decreased by more than 90% in recombinant human CBR1 protein. Therefore, it was demonstrated that CBR1 is the predominant LX reducing enzyme in the human skin.
Figure 4. Thin layer chromatograms of LX and its metabolites in the medium (A) and the tissue homogenate (B) after the ap14 plication of [ C]-LX (30 mg/ml as LX) to the human in vitro skin. Sampling time point for both medium and tissue were 8 h for the application Table 2. Composition of LX, the trans-OH form and the cis-OH form in the medium or tissue homogenates after the incubation 14 of [ C]-LX with the human in vitro skin for 8 h % of total radioactivity
Tissue Medium
Conc. (mg/ml)
LX
trans-OH
cis-OH
Total recovery
10 30 10 30
84.9 87.7 82.2 83.8
5.7 3.4 13.3 10.1
1.1 0.9 1.8 1.4
94.0 94.4 97.6 96.6
The average values of two samples are shown.
Copyright © 2015 John Wiley & Sons, Ltd.
Figure 5. Carbonyl reductase activities of the recombinant human CBR proteins for the trans-OH form generation. The recombinant human CBR proteins (CBR1 and CBR3) and EGFP as a negative control protein were expressed in 293-F cells and purified using an anti-FLAG antibody. CBR4 was purchased as a purified protein from Abnova Corp. LX carbonyl reductase activities of these recombinant proteins were examined. The data are shown as the average of duplicates with error bars to the minimum and maximum values (n = 2) Biopharm. Drug Dispos. 36: 352–363 (2015) DOI: 10.1002/bdd
360
R. SAWAMURA ET AL.
values for the recombinant CBR1 protein were calculated to be 7.30 mM and 402 nmol/min/mg protein, respectively.
Discussion
Figure 6. Immunodepletion of CBR1 in the human skin homogenates. The human skin homogenates (Skin 1: 0101-113155, Skin 2: 0101-113200 and Skin 3: 0101-112777) were immunodepleted with an anti-human CBR1 antibody (+) or normal rabbit IgG (-). The treated samples were used to detect CBR1 by Western blotting and to test for CBR1 activity. The activities were normalized by the activity of each input homogenate without the anti-human CBR1 antibody (+) and normal rabbit IgG (-) (Input). The data are shown as the average of duplicates with error bars to the minimum and maximum values (n = 2)
Figure 7. Enzyme kinetics for the trans-OH form generation catalysed by recombinant human CBR1 protein. Data represent the mean values of duplicate determination. Solid lines, best fit to the Michaelis-Menten equation with nonlinear least-squares regression
Enzyme kinetics of trans-OH form generation in recombinant human CBR1 protein. The kinetic study of the metabolism of LX was performed to evaluate the trans-OH form generation by CBR1 quantitatively. The generation of the trans-OH form in the recombinant CBR1 protein exhibited single-enzyme Michaelis-Menten kinetics, as shown in Figure 7. The Km and Vmax Copyright © 2015 John Wiley & Sons, Ltd.
The metabolism of loxoprofen was evaluated using the human in vitro skin and the predominant enzyme which generates the trans-OH form was identified to elucidate the bioactivation of LX in the human skin. In the study using the human in vitro skin, LX was detected in the medium time-dependently. This suggests that LX has a physicochemical property suitable for skin permeation, and LX would permeate through the human skin after the application of LX-G to the human in vivo skin. As there was a lag time between the application of LX-G and the appearance of LX in the medium, one of the factors is thought to be the permeation process of LX through the stratum corneum, which plays a role of the skin’s barrier to protect underlying tissue from infection, dehydration and chemicals [19]. Furthermore, not only LX but the trans-OH form was also detected in both tissue homogenates and medium. From this result, it was shown that the trans-OH form is produced in the human in vitro skin and permeates through it. Efficacy of the trans-OH form in the affected area (e.g. muscle, joint) is anticipated after the dermal application of LX to the human skin. In fact, the trans-OH form was determined in the plasma after the dermal application of LX-G to humans and LX-G showed the pharmacological effect after topical application to patients [3–5]. The data of the cis-OH form were not shown because the pharmacological activity of the cis-OH form was weak [1] and its concentrations were very low compared with LX and the trans-OH form as described in the Results section. TESTSKIN was used as the human in vitro skin. TESTSKIN is a three-dimensional primary cultured human skin model, which is generally used for the estimation of permeation, metabolism and toxicity in the human skin [20–22]. For instance, Shibayama et al. examined not only the permeation of disodium isostearyl 2-O-L-ascorbyl phosphate (VCP-IS-2Na), but the conversion of VCP-IS-2Na to vitamin C using TESTSKIN [22]. Biopharm. Drug Dispos. 36: 352–363 (2015) DOI: 10.1002/bdd
BIOACTIVATION OF LOXOPROFEN IN HUMAN SKIN
The pharmacokinetic model adequately described the amount–time profiles of LX and the trans-OH form in the medium. The analysis demonstrated that the permeation rate constant of the trans-OH form (kp_t) was about six times larger than that of LX (kp_LX). Since it is well known that log P is important for skin permeation [23], the difference in log P (clogP for LX: 1.97, clogP for the trans-OH form: 2.23) might result in a difference in the permeation rate constant. While the accumulation of LX in the medium increased in a linear manner with respect to the dose, the accumulation of the trans-OH form increased in a saturable manner. The reason was assumed to be because the concentration of LX in the skin tissue was higher than the Km value, though the compound concentrations in the skin tissue could not be determined due to the difficulty in the separation from endogenous substances. Km for generation of the trans-OH form in the human in vitro skin (10.3 mM) was similar to that in recombinant human CBR1 protein (7.30 mM). This provides the evidence that CBR1 is a predominant LX-reducing enzyme in human skin. Ohara et al. reported the Km value of LX in the human liver CBR and it was comparable to our results at the order level (38 mM) [8]. The prediction of the human pharmacokinetic (PK) profile of the compound is an essential part of drug development. In recent years, the use of IVIVE-PBPK linked models has increased in the pharmaceutical industries for human PK prediction [24,25]. The human PK can be predicted using the IVIVE-PBPK linked model without concern about the accuracy in the parameter estimation, although the parameter estimation performed well in our model analysis. The human PK of LX-G will be predicted using the IVIVE-PBPK linked models by adding the data (e.g. the CBR1 amount in human skin), and the predictability of the human PK examined in further studies. It has been reported that CBR catalyses the metabolism of LX in the human liver in addition to dihydrodiol dehydrogenase [8]. However, the CBR isoform was not identified in the report by Ohara et al. [8]. From the examination of the CBR1 contribution on the generation of the trans-OH form using the human whole-skin homogenate, it was demonstrated that CBR1 is the predominant LX reducing enzyme in the human Copyright © 2015 John Wiley & Sons, Ltd.
361 skin, as immunodepletion of CBR1 from human whole-skin homogenate reduced more than 80% of the LX reduction activity. Loxoprofen is considered to be bioactivated to the trans-OH form by CBR1 after LX permeates through the stratum corneum, because the trans-OH form concentrations were not detected in the stratum corneum after dermal application of LX-G to healthy male skin (data not shown). It is reported that CBR1 expresses in the outer root sheath of a human hair follicle [26], although the other location of CBR1 in human skin has not been identified yet. Further studies will be required to identify the location of CBR1 in human skin in order to further our understanding of the disposition of LX-G in humans. CBR1 catalyses the carbonyl reduction of various endogenous substrates such as prostaglandins, vitamin K, Coenzyme Q and so forth [9]. In addition to the endogenous substrates, xenobiotics such as haloperidol and anti-cancer anthracyclines (daunorubicin and doxorubicin) [27,28], belong to the substrate of CBR1. Although the metabolism of CBR1 is considered to play an important role in the detoxification of endogenous and xenobiotic ligands in the human body [9], it was found that CBR1 forms the pharmacologically active metabolite (the trans-OH form) with higher potency when compared with the parent drug (LX) [1]. Prodrugs require bioactivation by metabolizing enzymes to demonstrate their therapeutic effect. Genetic polymorphism in the prodrug activating enzymes could cause an interindividual variability in the drug exposure and response. Regarding CBR1, genetic polymorphism has been identified in previous reports. Gonzalez-Covarrubias et al. detected the CBR1 V88I polymorphism in DNA samples from individuals with African ancestry. Kinetic studies revealed that the Vmax of CBR1 I88 was more than 66% of that of CBR1 V88 for daunorubicin (CBR1 V88, 181 ± 13 versus Vmax CBR1 I88, 121 ± 12 nmol/min/mg protein) [14]. Bains et al. reported that the CBR1 wild-type enzyme has within a two-fold higher Vmax for doxorubicin compared with the CBR1 V88I [15]. Interindividual variability of CBR1 activity in human skin has not been reported yet. Although interindividual variability in the efficacy of LX-G has not been reported yet, it is believed that investigations on the interindividual variability Biopharm. Drug Dispos. 36: 352–363 (2015) DOI: 10.1002/bdd
362 of CBR1 in the human skin are important for evaluating therapeutic efficacy. It has been reported that skin metabolism might play a role in not only the formulation of active metabolites and detoxification, but also in the manifestation of local toxicities, such as skin sensitization or skin cancer [28,29]. For instance, cutaneous CYP1A enzymes metabolize polycyclic aromatic hydrocarbons, such as benzo(a) pyrene and dimethylbenz(a)anthracene, to toxic metabolites which are sufficiently reactive to bind to DNA, thereby causing skin cancer. Since a drug is highly exposed at the application site of the skin, it is possible that a larger amount of unknown metabolites tend to be generated after dermal application of a drug. Taken together, risk assessment for skin metabolism should be performed for dermal formulation. In the present study, however, the metabolic profile of LX in human in vitro skin and the medium was revealed to be the same as that in plasma after oral administration of LX to healthy volunteers without any unknown metabolites [30]. This suggests that it would not be necessary to consider toxicity derived from unknown metabolites of LX after dermal application. In addition, it is assumed that both metabolites which are generated in the skin after the dermal application of LX would be metabolized or excreted in the same way as oral administration after they move to the circulating blood.
Conclusion It was found that the permeation of loxoprofen and the generation of the trans-OH form increased dose- and time-dependently in the human in vitro skin. The characteristics of the permeation and metabolism of LX were evaluated by the analysis using the mathematical pharmacokinetic model. The predominant enzyme generating the transOH form in the human skin was identified to be CBR1 by immunodepletion using the anti-human CBR1 antibody. In addition, the Km value in recombinant human CBR1 protein was similar to that in human in vitro skin calculated by the model analysis. It was revealed that no metabolites other than the trans-OH form and the Copyright © 2015 John Wiley & Sons, Ltd.
R. SAWAMURA ET AL.
cis-OH form were formed in the human skin. This is the first report that LX is bioactivated to the trans-OH form in human skin by CBR1. Our preclinical studies supported the development of LX-G.
Acknowledgements We thank the National Disease Research Interchange and the Human Animal Bridging Research Organization for the provision of human skin blocks and the preparation of human skin homogenates. We gratefully acknowledge Hiroshi Katsuno and Mayumi Hayashi for the mass spectrometry experiments. We also appreciate Dr Kazuishi Kubota for his expert advice on writing the manuscript.
Conflict of Interest The authors have declared that there is no conflict of interest.
References 1. Matsuda K, Tanaka Y, Ushiyama S, Ohnishi K, Yamazaki M. Inhibition of prostaglandin synthesis by sodium 2-[4-(2-oxocyclopentylmethyl)phenyl] propionate dihydrate (CS-600), a new antiinflammatory drug, and its active metabolite in vitro and in vivo. Biochem Pharmacol 1984; 33: 2473–2478. 2. Honma M. Investigation of efficacy and safety in Loxonin® by new post-marketing surveillance. J Clin Ther Med 1994; 10: 2107–2120. 3. Daiichi Sankyo Co., Ltd. Investigator’s Brochure: Loxonin® Gel 1%. Tokyo, Japan. 4. Matsubara T, Arai K, Shiro Y, et al. Analgesic effects of NSAID gel containing loxoprofen (Loxonin gel 1%) in chronic neck pain females. Pain Clinic 2011; 32: 715–719. 5. Kajita Y, Yamazaki T, Sawada S, et al. Patient evaluation of the utility and usability of liniment containing loxoprofen sodium hydrate (Loxonin Gel) – study on the amount of single dose by affected site. J New Rem Clin 2012; 61: 2039–2051. 6. Heyneman CA, Lawless-Liday C, Wall GC. Oral versus topical NSAIDs in rheumatic diseases: a comparison. Drugs 2000; 60: 555–574. Biopharm. Drug Dispos. 36: 352–363 (2015) DOI: 10.1002/bdd
BIOACTIVATION OF LOXOPROFEN IN HUMAN SKIN
7. Tanaka Y, Nishikawa Y, Matsuda K, Yamazaki M, Hayashi R. Purification and some properties of ketone reductase forming an active metabolite of sodium 2-[4-(2-oxocyclopentylmethyl)-phenyl] propionate dihydrate (loxoprofen sodium), a new anti-inflammatory agent, in rabbit liver cytosol. Chem Pharm Bull (Tokyo) 1984; 32: 1040–1048. 8. Ohara H, Miyabe Y, Deyashiki Y, Matsuura K, Hara A. Reduction of drug ketones by dihydrodiol dehydrogenases, carbonyl reductase and aldehyde reductase of human liver. Biochem Pharmacol 1995; 50: 221–227. 9. Malátková P, Maser E, Wsól V. Human carbonyl reductases. Curr Drug Metab 2010; 11: 639–658. 10. Matsunaga T, Shintani S, Hara A. Multiplicity of mammalian reductases for xenobiotic carbonyl compounds. Drug Metab Pharmacokinet 2006; 21: 1–18. 11. Miura T, Nishinaka T, Terada T. Different functions between human monomeric carbonyl reductase 3 and carbonyl reductase 1. Mol Cell Biochem 2008; 315: 113–121. 12. Endo S, Matsunaga T, Kitade Y, et al. Human carbonyl reductase 4 is a mitochondrial NADPHdependent quinone reductase. Biochem Biophys Res Commun 2008; 377: 1326–1330. 13. Lakhman SS, Ghosh D, Blanco JG. Functional significance of a natural allelic variant of human carbonyl reductase 3 (CBR3). Drug Metab Dispos 2005; 33: 254–257. 14. Gonzalez-Covarrubias V, Ghosh D, Lakhman SS, Pendyala L, Blanco JG. A functional genetic polymorphism on human carbonyl reductase 1 (CBR1 V88I) impacts on catalytic activity and NADPH binding affinity. Drug Metab Dispos 2007; 35: 973–980. 15. Bains OS, Karkling MJ, Grigliatti TA, Reid RE, Riggs KW. Two nonsynonymous single nucleotide polymorphisms of human carbonyl reductase 1 demonstrate reduced in vitro metabolism of daunorubicin and doxorubicin. Drug Metab Dispos 2009; 37: 1107–1114. 16. Zhang J, Blanco JG. Identification of the promoter of human carbonyl reductase 3 (CBR3) and impact of common promoter polymorphisms on hepatic CBR3 mRNA expression. Pharm Res 2009; 26: 2209–2215. 17. Knaak B, Al-Bayati M, Raabe OG, Blancato JN. In vivo percutaneous absorption studies in the rat: pharmacokinetics and modeling of isofenphos absorption. In Prediction of Percutaneous Penetration, Scott RC (ed.). IBC Technical Services Ltd, London, 1991; 1–18.
Copyright © 2015 John Wiley & Sons, Ltd.
363 18. Sawamura R, Kazui M, Kurihara A, Izumi T. Absorption, distribution, metabolism and excretion of loxoprofen after dermal application of loxoprofen gel to rats. Xenobiotica 2014; 11: 1–13. 19. Trommer H, Neubert RH. Overcoming the stratum corneum: the modulation of skin penetration. A review. Skin Pharmacol Physiol 2006; 19: 106–121. 20. Osborne R, Perkins MA. An approach for development of alternative test methods based on mechanisms of skin irritation. Food Chem Toxicol 1994; 32: 133–142. 21. Andronis V, Mesiha MS, Plakogiannis FM. Design and evaluation of transdermal chlorpheniramine maleate drug delivery system. Pharm Acta Helv 1995; 70: 301–306. 22. Shibayama H, Hisama M, Matsuda S, Ohtsuki M. Permeation and metabolism of a novel ascorbic acid derivative, disodium isostearyl 2-O-L-ascorbyl phosphate, in human living skin equivalent models. Skin Pharmacol Physiol 2008; 21: 235–243. 23. Katritzky AR, Dobchev DA, Fara DC, et al. Skin permeation rate as a function of chemical structure. J Med Chem 2006; 49: 3305–3314. 24. Rostami-Hodjegan A. Physiologically based pharmacokinetics joined with in vitro-in vivo extrapolation of ADME: a marriage under the arch of systems pharmacology. Clin Pharmacol Ther 2012; 92: 50–61. 25. Chen Y, Jin JY, Mukadam S, Malhi V, Kenny JR. Application of IVIVE and PBPK modeling in prospective prediction of clinical pharmacokinetics: strategy and approach during the drug discovery phase with four case studies. Biopharm Drug Dispos 2012; 33: 85–98. 26. Colombe L, Vindrios A, Michelet JF, Bernard BA. Prostaglandin metabolism in human hair follicle. Exp Dermatol 2007; 16: 762–769. 27. De Jong J, Schoofs PR, Snabilié AM, Bast A, van der Vijgh WJ. The role of biotransformation in anthracycline-induced cardiotoxicity in mice. J Pharmacol Exp Ther 1993; 266: 1312–1320. 28. Hotchkiss SAM. Dermal metabolism. In Dermal Absorption and Toxicity Assessment, Roberts MS, Walter KA (eds). Marcel Dekker, New York, (Drug and the Pharmaceutical Sciences Vol. 91). 1998; 43–101. 29. Svensson CK. Biotransformation of drugs in human skin. Drug Metab Dispos 2009; 37: 247–253. 30. Daiichi Sankyo Co., Ltd. Investigator’s Brochure: Loxonin®. Tokyo, Japan. Website: http://www. info.pmda.go.jp/ippan/O200900004/430549000_ 22200APX00095000_V100_1.pdf
Biopharm. Drug Dispos. 36: 352–363 (2015) DOI: 10.1002/bdd
