Human mass balance, metabolite profile and identification of metabolic enzymes of [14C]ASP015K, a novel oral janus kinase inhibitor

Abstract

1. The human mass balance of 14C-labelled ASP015K ([14C]ASP015K), an orally bioavailable Janus kinase (JAK) inhibitor, was characterized in six healthy male subjects after a single oral dose of [14C]ASP015K (100 mg, 3.7 MBq) in solution. [14C]ASP015K was rapidly absorbed with tmax of 1.6 and 1.8 h for ASP015K and total radioactivity in plasma, respectively. Mean recovery in urine and feces amounted to 36.8% and 56.6% of the administered dose, respectively. The main components of radioactivity in plasma and urine were ASP015K and M2 (5′-O-sulfo ASP015K). In feces, ASP015K and M4 (7-N-methyl ASP015K) were the main components.
2. In vitro study of ASP015K metabolism showed that the major isozyme contributing to the formation of M2 was human sulfotransferase (SULT) 2A1 and of M4 was nicotinamide N-methyltransferase (NNMT).
3. The in vitro intrinsic clearance (CLint_in vitro) of M4 formation from ASP015K in human liver cytosol (HLC) was 11-fold higher than that of M2. The competitive inhibitory effect of nicotinamide on M4 formation in the human liver was considered the reason for high CLint_in vitro of M4 formation, while each metabolic pathway made a near equal contribution to the in vivo elimination of ASP015K. ASP015K was cleared by multiple mechanisms.

Keywords : ASP015K, human mass balance study, liver cytosols, metabolic enzyme

Introduction

4-{[(1R,2s,3S,5s,7s)-5-Hydroxy-2-adamantyl]amino}-1H- pyrrolo[2,3-b]pyridine-5-carboxamide (ASP015K) (Figure 1), synthesized at Astellas Pharma Inc. (Ibaraki, Japan), is an orally bioavailable Janus kinase (JAK) inhibitor with moder- ate selectivity for JAK1 and JAK3, and it inhibits JAK1 and JAK3 more potently than JAK2 by 1.3 and 7.0 times, respectively (Higashi et al., 2012). JAK1 is involved in the signaling pathways of various cytokines, such as interleukin (IL)-6 and interferon (IFN)-g, which are involved in inflam- matory responses in psoriasis and rheumatoid arthritis (RA) (Imada & Leonard, 2000; Leonard & Lin, 2000; Sakuma et al., 2001). JAK3 is involved in the signaling pathways of various cytokines (e.g. IL-2, IL-4, IL-7, IL-9, IL-15, and IL-21) which play a crucial role in T-cell differentiation, proliferation, and survival (Imada & Leonard, 2000; Leonard & Lin, 2000). In a Phase 2 study to evaluate the efficacy of ASP015K versus placebo in RA patients, ASP015K at once- daily doses of 100 and 150 mg exhibited superior efficacy compared to placebo (Takeuchi et al., 2014). ASP015K was also deemed safe and well tolerated in healthy subjects after single and multiple administrations up to 100 mg twice daily in Phase 1 clinical trials of non-radiolabeled ASP015K (Zhu, 2012).

The objectives of the present study were to (1) quantify the levels of total radioactivity in blood and plasma and of ASP015K in plasma, (2) investigate the routes and rates of elimination of ASP015K, and (3) examine the metabolite profiles of ASP015K after oral administration of [14C]ASP015K to healthy male subjects. Further, the in vitro metabolism of ASP015K was examined using recombinant human drug metabolizing enzymes to identify the isozymes responsible for the formation of metabolites. In addition, the in vitro intrinsic clearance (CLint_in vitro) values for the formation of metabolites from ASP015K were calculated using human liver cytosol (HLC) and then compared to the CLint_in vitro values to determine the relative contribution of each metabolic enzyme to the metabolic clearance of ASP015K.

Material and methods

Radiolabeled material and other materials

[14C]ASP015K (Figure 1) was synthesized at PerkinElmer Health Sciences Inc. (Boston, MA), with a certificate of analysis of the radiochemical purity (>99%) and specific activity (1.25 MBq/mg) and was stored at —80 ◦C. Authentic standards of non-labeled ASP015K and its metabolites
AS2628528 (M2), AS2604202 (M4), and AS2645866 (M1) were supplied by the Chemistry Research Laboratories of Astellas Pharma Inc. (Ibaraki, Japan). Deuterated internal standards (IS) d3-ASP015K (Figure 1), d3-M2, and d3-M1 were also supplied by the Chemistry Research Laboratories of Astellas Pharma Inc. AS1940152 (an enantiomer of ASP015K) was used as the IS for M4. Pooled HLC were purchased from Xenotech LLC (Lenexa, KS). Recombinant human sulfotransferase [SULTs] (SULT1A1*1, 1A1*2, 1A2, 1A3, 1B1, 1C2, 1C4, 1E1, and 2A1) were purchased from Cypex Ltd (Dundee, UK). 3′-Phosphoadenosine 5′-phospho- sulfate (PAPS) and S-adenosyl-L-methionine (SAM) were purchased from Sigma-Aldrich (St. Louis, MO). Nicotinamide adenine dinucleotide phosphate (reduced form) (NADPH·4Na) was purchased from Roche Diagnostics, Inc. (Indianapolis, IN). All other reagents were of high-performance liquid chromatography (HPLC) or analytical grade and obtained from commercial sources.

Dose preparation

The total dose of ASP015K was 100 mg/subject, and it contained 3.7 MBq of [14C]ASP015K. The dosing solution was prepared at the clinic site on the day of dosing. Briefly, an equivalent amount of ASP015K (radiolabeled plus non- radiolabeled) was dissolved in water to make the final concentration of 1 mg/mL. Subjects drank 100 mL of solution. The container was rinsed with water, and the subjects drank the rinse water. The 100-mg dose of ASP015K was the target for clinically effective dose. The radioactive dose of 100 mg/ 3.7 MBq of [14C]ASP015K was expected to be sufficient to determine the total recovery of radioactivity (mass balance), the routes and rates of excretion of radioactivity, the pharmacokinetic behavior of total radioactivity in blood components and excreta, and the pharmacokinetics of unchanged drug in plasma and urine.

Study design

This study was an open-label study involving six healthy male white subjects, who were aged 18 to 55 years with a body weight between 61 and 90 kg, and a body mass index between 21 and 28 kg/m2. The clinical phase of this study was conducted at Covance Clinical Research Unit (Covance CRU, Madison, WI) in accordance with good clinical practice (GCP) guidelines and the ethical principles that have their origin in the Declaration of Helsinki, International Conference on Harmonisation (ICH) guidelines, and applic- able laws and regulations. Eligible subjects were in good health as determined by the Investigator’s assessment of medical history, physical examination, clinical laboratory tests and 12-lead safety electrocardiogram (ECG).
Subjects were administered drug in the morning on day 1 after breakfast. On all other clinic days, standardized meals were served at appropriate times. Subjects were to abstain from alcohol, caffeine and blood oranges and grapefruit juice within 24 h prior to study drug administration. Standardized clinic meals (breakfast, lunch, dinner, and snack), non- alcoholic, and decaffeinated beverages were served at appro- priate times relative to dosing and performance of study procedures. Water was allowed with breakfast and then 2 h after study drug administration and lunch was served 4 h after study drug administration. Safety evaluations including 12-lead ECG, vital signs, and laboratory tests (hematology, biochemistry, and urinalysis) were performed throughout the study.

Sample collection

Blood samples (2 × 10 mL and 1 × 3 mL) were collected in Vacutainer® tubes containing K3- ethylenediaminetetraacetate (EDTA) at pre-dose and 0.5, 1, 1.5, 2, 3, 4, 6, 8, 12, 24, 36, 48, 72, 96, 120, 144, and 168 h after administration. Each 3-mL sample of whole blood was refrigerated and used for assessment of blood radioactivity. Each 10-mL sample of blood was centrifuged at 4 ◦C to separate the plasma for assessment of radioactivity, bioanalysis of ASP015K and for metabolic profiling. Plasma samples for assessment of radioactivity were stored refrigerated until analysis. Plasma samples for ASP015K assay and metabolic profiling were stored frozen at —70 ◦C until analysis.Urine samples were collected at 0 h (within 1 h prior to dosing), between 0 and 6 h, 6 and 12 h, and 12 and 24 h after administration, and at subsequent 24-h intervals until 7–9 d (168–216 h) after administration. During the collection inter- val, the urine was stored in a refrigerator. At the end of each interval, the samples were mixed and the total volume was recorded, and three aliquots were subsequently taken for radioactivity counting (10 mL), ASP015K analysis (4 mL), and metabolite profiling (100 mL). The urine for radioactivity counting was transferred to a polypropylene tube, and stored at approximately —20 ◦C. The samples for analysis of ASP015K analysis and metabolite profiling were stored at —70 ◦C.

Fecal samples were collected at 0 h (collected on the day before administration) and over 24-h intervals until 7–9 d (168–216 h) after administration. For each collection interval, the weight of the feces was recorded, and the whole samples were homogenized with water. The volume of water added was recorded to quantify the dilution of the samples. Portions of the homogenized feces were used for radioactivity counting. For metabolite profiling, a 25-g aliquot of the homogenized feces was transferred into a polypropylene tube, and then stored at —70 ◦C.

Analysis of total radioactivity in samples

Aliquots of plasma (approximately 0.2 g) and urine (approxi- mately 0.5 g) samples were mixed with liquid scintillation fluid (Ultima Gold XR; PerkinElmer Life and Analytical Sciences, Waltham, MA). An aliquot of blood (approximately 0.2 g) samples and fecal homogenate samples (approximately 0.5 g) were combusted using a Packard 307 sample oxidizer (PerkinElmer Life and Analytical Sciences). The 14CO2 generated was trapped in a mixture of PermaFluorTM (Thermo Fisher Scientific, Waltham, MA) and an absorbing fluid (CarboSorb; PerkinElmer Life and Analytical Sciences). All of the samples in the scintillation fluid were counted in the scintillation counter (Packard Tri-Carb 2900TR; PerkinElmer Life and Analytical Sciences), and were analyzed in duplicate if sample size allowed. If the result of a sample aliquot (calculated as 14C dpm/g sample) differed by more than 10% of the mean of two replicates and the aliquot had radioactivity greater than 200 dpm, the sample was re- mixed and re-analyzed as long as the amount of sample permitted.

Determination of unchanged ASP015K in plasma and urine

The liquid chromatography with tandem mass spectrometry (LC-MS/MS) system consisted of API4000 (Applied Biosystems, Foster, CA) and Series 10 ADVP series HPLC systems (Shimadzu, Kyoto, Japan) and was used for the measurement of concentrations of unchanged ASP015K in plasma and urine. The assay methods were validated for selectivity, linearity, accuracy and precision, and stability prior to the analysis. The processing of samples consisted of a supported liquid extraction of ASP015K and internal standard d3-ASP015K using an SLE + plate (Isolute SLE + Supported Liquid Extraction Plate, 200 mg [Biotage, Uppsala, Sweden]) and tert-butyl methyl ether. Separation of the analytes from matrix constituents was achieved using a HILIC Thermo Electron Betasil Silica-100 column (50 × 3 mm, 5 mm) with a mobile phase composed of 5:95:0.5:0.05 (v/v/v/v) water:acetonitrile:acetic acid:trifluoroacetic acid (TFA). Tandem mass spectrometric (MS/MS) detection was con- ducted by monitoring the following precursor/fragment ions: 327.14 → 160.20 (ASP015K) and 329.99 → 163.20 (internal standard) coupled to an API4000 mass spectrometer using an electron spray ionization interface in positive ion mode.
Calibration curves ranged from 0.25 to 250 ng/mL for plasma and from 2.5 to 2500 ng/mL for urine. Accuracy and precision of Quality Control samples at the concentration of 0.250, 0.750, 10 and 200 ng/mL were —13.4% to —7.9% and 3.4% to 5.6% RSD (relative standard deviation), respectively, for plasma. Accuracy and precision of quality control samples at the concentration of 2.50, 7.50, 100 and 2000 ng/mL were —5.4% to 3.5% and 3.6% to 5.5% RSD, respectively, for urine in the validation studies.

Pharmacokinetic analysis

Pharmacokinetic parameters were calculated from the indi- vidual subject data by non-compartmental methods using the computer program WinNonlin® Professional Version 5.01 (Pharsight Corp., Mountain View, CA). Pharmacokinetic parameters included area under the drug concentration–time curve extrapolated to infinity [AUCinf], maximum concentra- tion [Cmax], time to reach Cmax [tmax], terminal elimination half-life [t1/2] (0.693/kel), and renal clearance [CLr]. The apparent terminal elimination rate constant (kel) was deter- mined by linear regression of log-transformed concentration data over the terminal elimination phase.

The total amount of radioactivity excreted between time 0 to the last quantifiable time in the urine, and feces, and the amount of ASP015K excreted into the urine were calculated and were expressed as a percentage of the administered radioactive dose.

Sample preparation for metabolite profiling

The frozen urine (0–6, 6–12, and 12–24 h) and fecal homogenate samples (0–24, 24–48, 48–72 and 72–96 h) obtained from each subject were individually analyzed. The samples were thawed and mixed with a 3-fold volume of methanol and centrifuged 1800 × g at 4 ◦C for 10 min to separate the supernatant. The residue was extracted twice with methanol in the same way as the first extraction. All the obtained supernatants were combined and evaporated to dryness under reduced pressure. The residue was reconsti- tuted with Solvent A (100 mM ammonium acetate (pH 8.0)- acetonitrile (95:5, v/v)) or Solvent A containing N-methyl-2- pyrrolidinone (7:3, v/v) for urine samples or fecal samples, respectively. The samples were centrifuged (1800 × g, 4 ◦C, 5 min), and the obtained supernatants were subjected to HPLC analysis. The extraction recoveries of urine and fecal samples were 97.2% to 100.7% and 81.8% to 88.5%, respectively.

Plasma samples at 1, 2 (approximately tmax), 4, 8, and 12 h from the six subjects were pooled for each time point. The samples were thawed and mixed with a 3-fold volume of ethanol containing 1% acetic acid and centrifuged 1800 × g at 4 ◦C for 10 min to separate the supernatant. The residue was extracted twice with the same solvent in the same way as the first extraction. The extracted samples were treated in the same way as the fecal samples described above.

The extraction recoveries were 92.0%, 92.0%, 93.9%, 97.9%, and 72.7% for 1-, 2-, 4-, 8-, and 12-h plasma, respectively.Due to the low radioactivity of plasma samples, the plasma concentrations of metabolites were measured by accelerator mass spectrometry (AMS) (NEC 1.5SDH-1 0.6MV Pelletron AMS system, National Electrostatics Corp., Middleton, WI). For AMS analysis, plasma samples obtained from time-points up to 24 h post-administration were pooled to prepare one sample with a concentration proportional to the pharmacoki- netic AUC24 (area under the drug concentration-time curve [AUC] from time 0 to 24 h) (Hop et al., 1998). Briefly: AUC = Cpool(tm — t0) where Cpool denotes the concentration of the pooled plasma sample, tm denotes the last time-point, and t0 denotes the moment of dosing. The AUC calculated by this equation is theoretically identical to that calculated by the linear trapez- oidal rule. Volumes (v) of the individual aliquots used to create the pooled sample are determined using the following equation: Using this procedure, plasma samples at pre-dose, 0.5, 1, 1.5, 2, 3, 4, 6, 8, 12, and 24 h post-administration were pooled in a ratio of 0.5, 1, 1, 1, 1.5, 2, 3, 4, 6, 16, and 12, respectively, to approximate the AUC24. The pooled plasma was diluted with blank plasma to prepare approximately 10 dpm/mL. The sample was prepared in the same way as the method described previously.

HPLC analysis of metabolites in samples

An aliquot of the reconstituted solution was analyzed under the conditions described below. Relative amounts of metab- olites in the urine, feces, and plasma were determined using an HPLC connected to a radiochemical detector (RAD) or by a liquid scintillation counter (LSC) after collection of HPLC eluates. The metabolites were identified by comparison of retention times between radioactive peaks of samples and UV peaks of authentic standards on the HPLC connected to the UV detector. Identification of the metabolites was further conducted by comparison of retention times between ion peaks of samples and those of authentic standards on an LC-MS system equipped with multiple-stage mass spec- trometry (MSn).

A Develosil C30-UG-5 column (4.6 × 250 mm, 5 mm; Nomura Chemical Co. Ltd., Aichi, Japan) was used as the analytical HPLC column. As the mobile phase, the mixture of 100 mM ammonium acetate (pH 8.0)-acetonitrile-water (10:5:85, v/v/v) (A) and 100 mM ammonium acetate (pH 8.0)-acetonitrile-water (1:8:1, v/v/v) (B) was flowed at 1 mL/min in the following linear gradient mode: starting with 0% B composition, increasing to 20% in 0 to 50 min,increasing to 100% in 50 to 60 min, maintaining 100% in 60 to 70 min. The column was maintained at 40 ◦C. The column eluate was introduced to the UV detector set at a wavelength of 295 nm and, for urine and feces, to the RAD (Radiomatic 525TR or 625TR; PerkinElmer Life and Analytical Sciences). The elution pattern of metabolites in urine and feces was determined using the RAD with 6-s integration. As scintil- lation fluid for urine and feces, Flo-Scint II (PerkinElmer Life Analytical Sciences) was delivered to the HPLC eluate at a 3-fold flow rate of the mobile phase. For the analysis of the plasma of 1-, 2-, 4-, 8-, and 12-h after dosing, the HPLC eluate was fractionated every 18 s and mixed with 4 mL of Hionic-Fluor (PerkinElmer Life and Analytical Sciences) to measure the radioactivity using the LSC. For the measure- ment of metabolite AUC24, the HPLC eluates of pooled plasma prepared as shown in previous section were fractionated every 30 s and measured the radioactivity by AMS. The HPLC eluates collected from the injection until 20 min and from 60 until 70 min were pooled as two separate samples for AMS analysis because there were no significant radioactive peaks observed between these two time ranges. AMS analysis was conducted by the Institute of Accelerator Analysis Ltd (Fukushima, Japan). The AMS instrument and the interface used to generate CO2 from HPLC fraction prepared from plasma samples have been described previ- ously (Kitagawa et al., 1993).

On average, 98% and 97% of the injected radioactivity from urine and fecal extracts was recovered from the HPLC column. HPLC column recovery experiments were not conducted for plasma, because the radioactivity counts in plasma samples were very low. Detection limits of radio- activity for quantification of metabolite peaks in the LSC assays were defined as two times the background values.

The ratio of the radioactivity of [14C]ASP015K or its metabolites to the sum of radioactivity over the run time was determined from the HPLC chromatogram of the sample and multiplied by the recovery of radioactivity through the sample processing to calculate the composition of ASP015K and its metabolites in the sample. The amount of the ASP015K and its metabolites in the urine and feces was expressed as a percentage of the dose administered. When the peak corres- ponding to ASP015K or its metabolites was below the detection limit, the result was expressed as ND (not detected).

Identification of ASP015K, M1, M2, and M4 using LC-MS (MSn)

Analytical samples prepared from plasma, urine and feces were analyzed using an LC-MS (MSn). Metabolites in analytical samples were identified by comparing the retention time, protonated molecule, and fragment ion peaks with the authentic standards achieved by separation on LC with ion trap MS. The conditions were as follows mass spectrometer, LTQ Orbitrap XL (Thermo Fisher Scientific); atmospheric pressure ionization interface, electrospray ionization; acqui- sition, full scans in positive ion modes; spray voltage, 3.7 kV; sheath gas (N2), 53 arb; auxiliary gas (N2), 43 arb; capillary temperature, 200 ◦C. HPLC conditions were described in a previous section. MS2 and MS3 production scans for including the target parent ion were set at m/z 327 and 310 for ASP015K, at m/z 421 and 341 for M1, at m/z 407 and 327 for M2, at m/z 341 and 324 for M4, respectively. Normalized collision energy for each product ion scan was set at 35%.

In vitro sulfation of ASP015K to M2 and M4 to M1

In vitro sulfation of ASP015K and M4 in HLC was measured as previously described (Honma et al., 2002), with slight modifications to HLC conditions (0.05–0.2 mg protein/mL) and incubation times (0–45 min). These modifications resulted in a linear formation of M2 from ASP015K and of M1 from M4, and in an HLC concentration of 0.1 mg/mL and incubation time of 30 min being selected for further sulfation studies. Substrates (ASP015K or M4) were incubated in 50 mM phosphate buffer (pH 7.4) containing pooled HLC of 0.1 mg/mL, 10 mM magnesium chloride, and 10 mM dithio- threitol. After pre-incubation at 37 ◦C, PAPS was added to initiate the enzyme reaction (final concentration = 0.4 mM). Substrate concentrations in the reaction mixture were 2 to 200 mM (ASP015K) or 3 to 300 mM (M4). The reaction was performed for the designated durations and terminated by mixing 2-fold volume of chilled 50% acetonitrile, followed by addition of each IS and centrifugation at 1870 × g for 10 min at 4 ◦C. An aliquot of the supernatant was diluted twice with 10 mM ammonium acetate and injected into an LC-MS/MS apparatus. Concentrations of the metabolites M2 and M1 in the supernatant were determined as in the ‘‘LC-MS/MS Analysis’’ section. Data for the enzymatic reactions were fitted to the Michaelis–Menten equation by a non-linear regression model using the least-squares method. Kinetic parameter (Km and Vmax) values were calculated using GraphPad Prism 5 (GraphPad Software Inc., La Jolla, CA). As non-specific protein binding of ASP015K in the reaction mixture was negligible (data not shown), no correction was performed for the calculation of CLint_in vitro [= Vmax/Km]).

Identification of SULT isozymes for the formation of M2

To identify the SULT isozymes responsible for the formation of M2 from ASP015K, M2 concentration after incubation of ASP015K in recombinant SULTs (SULT1A1*1, 1A1*2, 1A2, 1A3, 1B1, 1C2, 1C4, 1E1, and 2A1) was measured. [14C]ASP015K (1 mM) was incubated for 60 min in 50 mM phosphate buffer (pH 7.4) containing each recombinant SULT (0.2 mg/mL), 0.4 mM PAPS, 10 mM magnesium chloride, and 10 mM dithiothreitol. The reaction was terminated by mixing 2-fold volume of chilled 50% acetonitrile, followed by centrifugation at 8000 × g for 5 min at 4 ◦C. After evaporation of the supernatant, the residue was dissolved with 0.3 mL of mobile phase with sonication. Recovery of radioactivity to the resoluble extract was 92.5% or more from all samples. An aliquot of the supernatant was injected into an HPLC-RAD (Radiomatic 525TR or 625TR; PerkinElmer Japan Co., Ltd.) equipped with a 500-mL liquid cell. Column recovery from HPLC was 94.8% or more from all the samples. HPLC analytical conditions were described in ‘‘HPLC Analysis of Metabolites in Samples’’. A part of column eluate was also introduced to LC-MS/MS to obtain the mass spectra of metabolites.

In vitro methylation of ASP015K to M4 and M2 to M1

In vitro methylation of ASP015K and M2 in HLC was measured in the presence of 100 mM potassium phosphate buffer (pH 7.4) and 0.1 mM EDTA. After pre-incubation at 37 ◦C, SAM was added to initiate the enzyme reaction at a final concentration 1 mM. HLCs at concentrations of 0.05 to
0.2 mg/mL and incubation times of 0 to 30 min resulted in a linear formation of M4 from ASP015K, as well as that of M1 from M2. An HLC concentration of 0.1 mg/mL and incuba- tion time of 20 min were therefore selected for further methylation studies. Substrate concentrations in the reaction mixture ranged from 1 to 100 mM (ASP015K) or 2 to 150 mM (M2). The reaction was terminated by mixing a 2-fold volume of chilled 50% acetonitrile, followed by the addition of each IS and centrifugation at 1870 × g for 10 min at 4 ◦C. An aliquot of the supernatant was diluted twice with 10 mM ammonium acetate and injected into an LC-MS/MS appar- atus. The concentrations of metabolites M4 and M1 were determined as described in the ‘‘LC-MS/MS Analysis’’. Km and Vmax values were obtained as described in the previous section.

Identification of N-methyltransferase (NMT) isozymes for the formation of M4 from ASP015K and M1 from M2

The N-methyltransferase (NMT) isozymes responsible for the formation of M4 from ASP015K and of M1 from M2 were investigated using two recombinant NMT isozymes expressed predominantly in human liver (nicotinamide N-methyltrans- ferase [NNMT] and glycine N-methyltransferase [GNMT]) (Chen et al., 2010; Emanuelli et al., 2010). ASP015K and M2 (1 mM) were incubated in 100 mM potassium phosphate buffer (pH 7.4) containing 5 mg protein/mL of NNMT or GNMT, and 0.1 mM EDTA. SAM was added to initiate the enzyme reaction at a final concentration 1 mM, and mixtures were incubated at 37 ◦C for 30 min. The reaction was terminated by mixing a 2-fold volume of chilled 50% acetonitrile, followed by addition of each IS and centrifuga- tion at 1870 × g for 10 min at 4 ◦C. An aliquot of the supernatant was diluted with 10 mM ammonium acetate and injected into an LC-MS/MS apparatus. Metabolites M4 or M1 concentrations were determined using the method described in ‘‘LC-MS/MS Analysis’’.

Effects of nicotinamide on the formation of M4 from ASP015K

Nicotinamide is a vitamin involved in a physiologically important metabolic pathway supplying pyridine nucleotides to the liver (Pumpo et al., 2001). Given that excess nicotina- mide is metabolized by NNMT to N-methyl nicotinamide in the human liver (Pumpo et al., 2001), nicotinamide in the liver might competitively inhibit NNMT-mediated M4 formation.

In vitro methylation of ASP015K in HLC (0.05 mg protein/ mL) was measured in the presence of 0, 31, 63, 125, and 250 mM of nicotinamide. Enzyme reactions were conducted as described in ‘‘In Vitro Methylation of ASP015K to M4 and M2 to M1’’. M4 concentration was determined using the method as described in ‘‘LC-MS/MS Analysis’’. The observed rates versus substrate concentrations (in the presence and absence of inhibitors) were simultaneously fitted according to the non-linear regression for competitive, non-competitive, and uncompetitive enzyme inhibition (Ring et al., 1995) using Phoenix WinNonlin software (version 6.1, Pharsight Corp.). The initial inhibition constant (Ki value) was calculated from the relationship between concentration of nicotinamide and the slopes of the lines of Lineweaver– Burk plots using GraphPad Prism 5 (Chiba et al., 1994).

LC-MS/MS analysis

The LC-MS/MS system consisted of API4000 and LC-10Avp HPLC systems. Analytical conditions for the determination of ASP015K and M2 were as previously described (Oda et al., 2014). For the determination of M4 and M1 concentrations, the same HPLC condition was also applied for separation of the analytes from ASP015K and M2 as well as matrix constituents. Tandem mass spectrometric detection was conducted by monitoring the following precursor/fragment ions: 341 → 174 (M4), 327 → 91 or 160 (internal standard of M4), 421 → 341 (M1), 424 → 344 (internal standard of M1) coupled to an API4000 mass spectrometer using an electron spray ionization interface in positive ion mode. Ion chro- matograms were integrated and quantified using Analyst software (version 1.4.2, Applied Biosystems, Foster, CA).

Results

Safety assessment

A single-oral dose of 100 mg [14C]ASP015K was well tolerated in the six subjects tested with mild events of arthralgia and dizziness reported in one subject. All adverse events resolved during the course of study. There were no clinically important changes in clinical laboratory values, vital signs, ECG parameters and physical examination data during the study.

Pharmacokinetics of unchanged ASP015K and total radioactivity

Time profiles of the concentrations of radioactivity in the whole blood and plasma as well as unchanged ASP015K in the plasma after a single oral dose of 100 mg of [14C]ASP015K are illustrated in Figure 2. The key pharma- cokinetic parameters are summarized in Table 1. The mean tmax of ASP015K and total radioactivity in the plasma was 1.6 h (median (range): 1.8 h (1.0–2.0 h)) and 1.8 h (median (range): 2.0 h (1.0–3.0 h)), respectively. AUCinf of ASP015K accounted for 33% of total radioactivity in plasma. The whole blood-to-plasma ratios (B/P) of total radioactivity for Cmax and AUCinf were 0.76 and 0.73, respectively.

Excretion and recovery of unchanged ASP015K and total radioactivity

Mean excretion of radioactivity in urine accounted for 36.8% ± 4.4% of the administered dose over 216 h, the majority (25.7% of the administered dose) being excreted within 0 to 6 h post-dose. Mean excretion of radioactivity in feces accounted for 56.6% ± 5.1% of the administered dose over 216 h, the majority (49.7% of the administered dose) being excreted within 72 h after dosing. The cumulative excretion in urine and feces are described in Figure 3, and sum of urinary and fecal excretion was 93.4% of the administered dose by 216 h. The mean total amount of unchanged ASP015K excreted in urine accounted for 36% of the excreted radioactivity.

Metabolic profiles in urine

Representative radiochromatograms of urine are shown in Figure 4. Four peaks were present in these chromatograms. Assignment of the radioactive peaks to ASP015K and its metabolites were done by comparison of retention times and mass spectra including product ion spectra with authentic reference compounds. The peak at 52 min was assigned to unchanged ASP015K because the retention time and the product ion spectra corresponded with those of ASP015K in subsequently conducted identification experiments (Table 2; Figure 5). Similarly, the metabolite peaks at approximately 36, 38, and 44 min were identified as M1 (AS2645866), M2 (AS2628528), and M4 (AS2604202), respectively (Table 2; Figures 6–8). Between 0 and 24 h post-dose, the mean urinary excretion of unchanged ASP015K, M1, M2, and M4 amounted to 14.2%, 2.6%, 13.7%, and 4.1% of the dose,respectively (Table 3). ASP015K and M2 represented the largest components in urine.

Metabolite profiles in feces

In the radiochromatograms of extracts of fecal homogenates, five peaks were present (Figure 9). The peak at 52 min was assigned to unchanged ASP015K based on the retention time and the product ion spectra (Table 2). Similarly, the metab- olite peaks at approximately 38 and 44 min were identified as M2, and M4, respectively (Figure 9). The structure of two metabolites (Peak 1 and 2) could not be identified. Peak 1 found at 39 min generated the protonated molecule at m/z 343.1762 corresponding to C18H23N4O3 (calculated [M+H]+, 343.1765), and its major fragment ions were detected at m/z 326 and 298. Peak 2 was detected in the retention time of 43 min (Figure 9), however, the molecular weight could not be obtained. Between 0 and 96 h post-dose, the mean fecal excretion of unchanged ASP015K, M2, and M4 amounted to 29.8%, 5.8%, and 10.7% of the dose, respectively (Table 3). The mean fecal excretions of Peak 1 and Peak 2 were 1.0% and 0.02%, respectively (Table 3). ASP015K represented the largest component in feces.

Metabolite profiles in plasma

Representative radiochromatograms of plasma extracts are shown in Figure 10. Only two peaks were found in the chromatograms at 1, 2, and 4 h post-dose measured by LSC,and the peak at 38 and 51 min was assigned to M2 and unchanged ASP015K, respectively, based on the retention time and the product ion spectra (Figure 10a; Table 3). The two peaks were below the detection limit at 8 h post-dose thereafter (Figure 10b). A total of six peaks was found in the chromatograms of 0–24 h sample measured by AMS (Figure 10c). The peak at 50 min was assigned to unchanged ASP015K. Similarly, peaks at 34, 37, and 44 min were identified as metabolites M1, M2, and M4, respectively. Two other peaks (Peak 2 and Peak 3) were detected at 42 and 47 min, respectively; their structure could not be identified. The ratios of unchanged ASP015K and M2 to the total profiled radioactivity at 1, 2, and 4 h (% of profiled radioactivity) were 45.1–63.4% and 30.5–45.3%, respectively. The ratios of unchanged ASP015K, M1, M2, and M4 to the total profiled radioactivity in 0–24 h sample (% of profiled radioactivity) were 32.5%, 4.3%, 34.2%, and 7.5%, respect- ively, and those of two un-identified metabolites (Peak 2 and Peak 3) were 1% or less. ASP015K and M2 were the largest components in plasma.

In vitro sulfation of ASP015K to M2 and M4 to M1

For the formation of M2 from ASP015K in recombinant SULT2A1, the Km value was 24.6 mM and Vmax value was 2460 pmol/min/mg protein. For the formation of M1 from M4, the Km value was 216 mM and Vmax value was 259 pmol/ min/mg protein (Table 4).

In vitro methylation of ASP015K to M4 and M2 to M1

The kinetic parameters (Km, Vmax and CLint_in vitro) for the formation of M4 from ASP015K as well as those of M1 formation from M2 in HLC were determined. For M4 formation from ASP015K, the Km value was 13.1 mM and Vmax value was 195 pmol/min/mg protein. For M1 formation from M2, the Km value was 41.3 mM and Vmax value was 258 pmol/min/mg protein (Figure 13 and Table 5). The CLint_in vitro value for the formation of M4 was twice that of M1.

Identification of NMT isozymes for the formation of M4 from ASP015K and M1 from M2

The NMT isozymes responsible for the formation of M4 from ASP015K and of M1 from M2 were investigated using two formation of M2 from ASP015K as well as that of M1 from M4 in HLC were determined (Figure 11). For M2 formation from ASP015K, the Km value was 34.5 mM and Vmax value was 47.4 pmol/min/mg protein. For M1 formation from M4, the Km value was 128 mM and Vmax value was 13.1 pmol/min/ mg protein (Table 4). The CLint_in vitro value (as expressed by Vmax/Km) for the formation of M2 was 14 times higher than that of M1.

Identification of SULT isozymes for the formation of M2

The composition of ASP015K and its metabolites in each recombinant human SULT isozyme was investigated. ASP015K was mainly metabolized by SULT2A1, and the remaining ratio after reaction for 60 min was 13.6% (Figure 12). Almost no ASP015K was metabolized by the other SULT isozymes examined, and the remaining ratios after reaction for 60 min ranged from 99.1% to 100.0%. In the radiochromatogram of SULT2A1, one major metabolite peak was observed, and the percentage peak accounted for 86.4%. The ion at m/z 407 (calcd [M+H]+: 407) corresponding to [M+H]+ ion of M2 was detected at the retention time of M2. As the major fragment ions were the same as those of the reference standard of M2, the radioactive peak was identified as M2.

Effects of nicotinamide on the formation of M4 from ASP015K

The effects on nicotinamide on the formation of M4 from ASP015K in HLC were determined. Of these models, the best-fit model was determined as a competitive inhibition model with a Ki value of 29 mM, suggesting that nicotinamide competitively inhibits M4 formation (Figure 14).

Discussion

[14C]ASP015K was rapidly absorbed with a plasma tmax of 1.6 and 1.8 h for ASP015K and total radioactivity post-dose, respectively, suggesting that orally dosed ASP015K was mainly absorbed in upper gastrointestinal tract. Plasma concentration decreased with a t1/2 of 12 and 3.1 h for ASP015K and radioactivity, respectively (Table 1). Plasma concentrations could be measured until 48 and 12 h post-dose Xenobiotica Downloaded from informahealthcare.com by Nyu Medical Center on 05/24/15 For personal use only.

We also investigated the Km and Vmax values for the formation of M1 from M4 in HLCs and SULT2A1. As shown in Figure 11 and Table 4, M1 was observed when M4 was incubated in both HLCs and SULT2A1. Here again, no significant difference was observed in Km values between SULT2A1 and HLC (Table 4). Although we have not examined M1 formation from M4 in the other recombinant SULTs, these results indicate that SULT2A1 is also the isozyme responsible for the formation of M1 from M4.

SULT2A1 is expressed in human liver as well as in the small intestine (Teubner et al., 2007). Hepatic CLint_in vitro (1.4 mL/min/mg protein) of sulfation to M2 from ASP015K was corrected to 2.7 mL/min/kg, based on the cytosolic protein in liver (80.7 mg protein/g of liver [Cubitt et al., 2011]) and weight of liver in body (24 g of liver/kg of body weight [Hanano et al., 1985]). The intestinal CLint_in vitro of sulfation was corrected to 1.1 mL/min/kg, based on the cytosolic protein in small intestine (18 mg protein/g of small intestine [Cubitt et al., 2011]) and weight of small intestine in body (46 g of small intestine/kg of body weight [Hanano et al., 1985]). The higher CLint_in vitro of liver than that of small intestine suggested that hepatic SULT2A1 significantly contributes to the M2 formation compared to intestinal SULT2A1.

As shown in Figure 13 and Table 5, each of the methylated metabolites were detected when ASP015K or M2 were incubated in HLC and NNMT. It was suggested that the NMT isozyme responsible for the formation of M4 from ASP015K and for the formation of M1 from M2 was NNMT (Figure 13). NNMT is a cytosolic enzyme that catalyzes the N-methylation of the nitrogen atom of nicotinamide (Emanuelli et al., 2010). ASP015K has the same structure as nicotinamide, and the nitrogen atom was methylated to M4. NNMT was therefore considered a reasonable candidate for the catalysis of ASP015K. The postulated metabolic pathways are shown in Figure 15.

Several genetic polymorphisms have been reported in both NNMT (Emanuelli et al., 2010) and SULT2A1 (Thomae et al., 2002). Saito et al. screened DNA from Japanese subjects for single-nucleotide polymorphisms (SNPs) in NNMT genes, and suggested that some SNPs might affect its transcriptional efficacy (Saito et al., 2001). In addition, SULT2A1 activity is reported to vary by more than 5-fold without gender dependence (Aksoy et al., 1993). In African- American subjects, several genetic polymorphisms that alter the amino acid sequence and enzyme activity of SULT2A1 have been reported (Thomae et al., 2002). The influence of the genetic polymorphisms on the pharmacokinetics of ASP015K may therefore be considered.

The CLint_in vitro for the formation of M4 from ASP015K was 11 times higher than that of sulfation (M2 formation from ASP015K), a finding we consider interesting, as the human mass balance study shows that methylation and sulfation are major metabolic pathways of ASP015K. In fact, 20% of the administered dose was excreted as M2 and 15% as M4. Nicotinamide is a vitamin involved in a physiologically important metabolic pathway which supplies pyridine nucleo- tides to the liver (Pumpo et al., 2001). Excess nicotinamide is converted to N-methylnicotinamide by liver NNMT or excreted in a small quantity as free urinary compound (Pumpo et al., 2001). Therefore, hepatic nicotinamide might competitively inhibit M4 formation by NNMT.

As shown in Figure 14, the addition of nicotinamide to HLC resulted in the competitive inhibition of M4 formation with a Ki value of 29 mM, which is the same level of nicotinamide (27–37 mM) in human plasma (Jenks et al., 1987). In general, the magnitude of the effect of an inhibitor on the metabolism of drugs depends on the concentration of the inhibitors at the enzyme site. Although no nicotinamide concentration in human liver is available, the physiological nicotinamide concentration in rat liver was reported as 11 to 500 mM (Hoshino et al., 1984). When such a high concen- tration (500 mM) is anticipated in the human liver, the formation of M4 will be affected by nicotinamide in human liver. We estimate that the competitive inhibitory effect of nicotinamide in the human liver on the formation of M4 from ASP015K was the reason for the 11-fold higher CLint_in vitro (methylation > sulfation) observed in vitro, while each meta- bolic pathway made a near equal contribution to the elimination of ASP015K in vivo. These findings indicate that the inhibitory effect of nicotinamide in the liver should be taken into consideration when estimating the CLint_in vivo of methylation by NNMT from CLint_in vitro.

Of the administered dose of [14C]ASP015K, 36.8% was excreted in urine and 56.6% in feces. In urine, 14% of the administered dose was excreted as unchanged ASP015K, and CLr was calculated to be 13 L/h, indicating that urinary excretion of the unchanged form is one of the elimination pathways of ASP015K.

In conclusion, the present study clarified the absorption and elimination kinetics of ASP015K and the characteristics of metabolites in the excreta and plasma of six healthy male subjects after a single oral administration of 100 mg of [14C]ASP015K in solution. ASP015K was rapidly absorbed after oral administration and circulated in the plasma as unchanged drug, sulfuric acid conjugates, N-methyl conju- gates, and other metabolites. Of the administered dose, 36.8% was excreted in urine, mainly as the unchanged form and sulfuric acid conjugate (M2), and 56.6% was recovered in feces, mainly as the unchanged form and N-methyl conjugate (M4). The isozyme responsible for the formation of M2 was SULT2A1 and of M4 was NNMT. ASP015K was cleared by multiple mechanisms (renal and possibly biliary excretion,Peficitinib and metabolism) with no single dominant clearance pathway.