Drug Metabolism and Disposition Fast Forward
First published on January 7, 2008; DOI: 10.1124/dmd.107.018572
0090-9556/08/3604-641-648$20.00
DMD 36:641-648, 2008
Identification of the in Vitro Metabolites of 3,4-Dihydro-2,2-dimethyl-2H-naphthol[1,2-b]pyran-5,6-dione (ARQ 501; β-Lapachone) in Whole Blood
Xiu-Sheng Miao,
Pengfei Song,
Ronald E. Savage,
Caiyun Zhong,
Rui-Yang Yang,
Darin Kizer,
Hui Wu,
Erika Volckova,
Mark A. Ashwell,
Jeffrey G. Supko,
Xiaoying He, and
Thomas C. K. Chan
Departments of Preclinical Development and Clinical Pharmacology (X.-S.M., P.S., R.E.S., C.Z., T.C.K.C.) and Chemistry (R.-Y.Y., D.K., H.W., E.V., M.A.A.), ArQule, Inc., Woburn, Massachusetts; and Massachusetts General Hospital, Harvard Medical School, Boston, Massachusetts (J.G.S., X.H.)
(Received August 28, 2007;
accepted January 2, 2008)
 |
Abstract
|
|---|
3,4-Dihydro-2,2-dimethyl-2H-naphthol[1,2-b]pyran-5,6-dione (ARQ 501; β-lapachone) showed promising anticancer activity in phase I clinical trials as monotherapy and in combination with cytotoxic drugs. ARQ 501 is currently in multiple phase II clinical trials. In vitro incubation in fresh whole blood at 37°C revealed that ARQ 501 is stable in plasma but disappears rapidly in whole blood. Our data showed that extensive metabolism in red blood cells (RBCs) was mainly responsible for the rapid disappearance of ARQ 501 in whole blood. By comparison, covalent binding of ARQ 501 and/or its metabolites to whole blood components was a minor contributor to the disappearance of this compound. Sequestration of intact ARQ 501 in RBCs was not observed. Cross-species metabolite profiles from incubating [14C]ARQ 501 in freshly drawn blood were characterized using a liquid chromatography-mass spec-trometry-accurate radioactivity counter. The results show that ARQ 501 was metabolized more rapidly in mouse and rat blood than in dog, monkey, and human blood, with qualitatively similar metabolite profiles. Six metabolites were identified in human blood using ultra-high performance liquid chromatography/time-of-flight mass spectrometry, and the postulated structure of five metabolites was confirmed using synthetic standards. We conclude that the primary metabolic pathway of ARQ 501 in human blood involved oxidation of the two adjacent carbonyl groups to produce dicarboxylic and monocarboxylic metabolites, elimination of a carbonyl group to form a ring-contracted metabolite, and lactonization to produce two metabolites with a pyrone ring to form a ring-contracted metabolite. Metabolism by RBCs may play a role in clearance of ARQ 501 from the blood compartment in cancer patients.
3,4-Dihydro-2,2-dimethyl-2H-naphthol[1,2-b]pyran-5,6-dione (β-lapachone) is found in Pau d'arco trees (Tabebuia impetiginosa) and possesses antitumor activity against S-180 leukemic cells, Yoshida sarcoma cells, sarcoma 180 cells, and Walker 256 carcinoma cells in early studies (Docampo et al., 1979
; Schaffner-Sabba et al., 1984; Chau et al., 1998). ARQ 501 is a fully synthetic β-lapachone and has been identified as a potent activator of E2F1. The compound causes tumor cell apoptosis and death in vitro and in vivo (Wang et al., 2005). Promising anticancer activity was observed in phase I and phase II clinical trials.
Blood is generally assumed to be an inert vehicle for the transportation of drugs to target tissues. The primary physiological objective of red blood cells (RBCs) is gas transport and exchange. Beyond this, they perform metabolic functions to produce the necessary cofactors (ATP, NADPH, and NADH) enzymatically for their own persistence (Bossi and Giardina, 1996; Wiback and Palsson, 2002). Though various enzymes in RBCs can metabolize drugs (Carruthers and Melchior, 1988; Hooks, 1994), systematic investigation of whole blood metabolism of therapeutic drugs has been, according to some researchers, traditionally neglected in drug discovery and development (Cossum, 1988; Hinderling, 1997). In particular, few in-depth studies of the metabolism of drugs in whole blood exist.
In vitro incubation in whole blood and in plasma revealed that ARQ 501 was stable in plasma but disappeared rapidly from whole blood. Potentially, sequestration in RBCs, covalent binding to blood components, metabolism by RBCs, or a combination of the above processes could be responsible. The objectives of this study were to explore the underlying mechanism for the rapid disappearance of ARQ 501 from whole blood and to identify its probable in vitro metabolites.
|
Materials and Methods
|
|---|
Materials. [14C]ARQ 501 was synthesized by ABC Laboratories (Columbia, MO) with a radiochemical purity of 99.6%. The specific activity was 5.22 µCi/mg (10 mCi/mmol). Nonlabeled ARQ 501 and deuterated ARQ 501 (D6-ARQ 501) were synthesized by ArQule, Inc. Scintillation cocktail ULTIMA-FLO was purchased from PerkinElmer Life and Analytical Sciences (Waltham, MA). Metabolite standards were synthesized by ArQule, Inc. (Woburn, MA) (Kizer et al., 2007). High purity acetonitrile and water were purchased from EMD Chemicals (Gibbstown, NJ). All other chemicals used were of reagent grade or better.
Sample Collection. Fresh mouse and rat whole blood was collected within ArQule, Inc. Fresh dog and monkey whole blood was supplied by Charles River Laboratories, Inc. (Wilmington, MA). Fresh human whole blood was obtained from healthy volunteers. Blood samples were collected into a K3-EDTA Vacutainer (BD Biosciences, Franklin Lakes, NJ) over ice and were used within 24 h. Plasma was prepared by centrifuging fresh whole blood at 3000g at 25°C for 5 min. Nonlabeled ARQ 501 was used to determine drug stability and partitioning in whole blood, whereas [14C]ARQ 501 was used to investigate covalent binding, metabolite profiling, and structural elucidation. Because ARQ 501 is light-sensitive, photoprotection was achieved by using amber vials or aluminum foil as a seal for sample plates.
Drug stability and partitioning studies were conducted using fresh whole blood and plasma from four species: mouse, rat, dog, and human. ARQ 501 was spiked into whole blood and plasma, respectively, to prepare a 10-µM solution, which was incubated at 37°C in a thermomixer. An aliquot of 150 µl of whole blood was collected at 0, 30, 60, and 120 min after incubation and then immediately centrifuged at 6000g for 1 min to separate plasma from RBCs. The resulting RBCs were mixed with a 2-fold volume of distilled water and incubated at 37°C to generate RBC lysate. The concentrations of ARQ 501 in plasma and RBC lysate were determined with LC-MS-MS after protein precipitation by a 3-fold volume of acetonitrile containing D6-ARQ 501 as internal standard (300 ng/ml).
Covalent binding and metabolism studies were conducted by incubating [14C]ARQ 501 at 10 and 100 µM with human whole blood at 37°C. An aliquot of 150 µl of whole blood was collected at 0, 30, 60, 120, and 180 min after incubation and mixed with a 2-fold volume of distilled water to generate whole blood lysate. Aliquots (20 µl) of the whole blood lysate at each time point were used for total radioactivity counting. To eliminate the color quenching effect of whole blood lysate, four volumes of 30% H2O2 were added into each aliquot for decolorization. The sample was mixed with scintillation cocktail for radioactivity measurement on a Wallac 1450 MicroBeta liquid scintillation counter (PerkinElmer Life and Analytical Sciences). Aliquots (20 µl) of the whole blood lysate at each time point were mixed with three volumes of acetonitrile and centrifuged to separate the protein pellet from the supernatant. The resulting protein pellets were used for covalent binding determination and the supernatant was used for metabolite characterization by LC-MS-ARC. Protein pellets were extensively washed with 80% aqueous ethanol until the radioactivity in the washings was reduced to less than twice background radioactivity (
150 cpm). The protein pellets were then dissolved in 100 µl of 0.1 N NaOH solution and neutralized with an equal volume of 0.1 N HCl solution before the decolorization and radioactivity measurement. Covalent binding (%) was calculated as percentage of the total radioactivity remaining in protein pellets after extensive washing.
LC-MS-MS. An API4000 mass spectrometer (Applied Biosystems, Foster City, CA) equipped with an Agilent 1100 binary pump (Agilent Technologies, Palo Alto, CA) and an HTS PAL injector (LEAP Technologies, Inc., Carrboro, NC) was used for LC-MS-MS analyses in blood stability and partitioning experiments. Chromatographic separations were performed on an Atlantis dC18 HPLC column (2.1 x 50 mm, 5 µm; Waters, Milford, MA) using a binary gradient. The mobile phase solvents [water (A) and acetonitrile (B)] were both modified with 0.1% formic acid. The gradient used was as follows: solvent B started at 40% and held for 0.3 min, then linearly increased to 95% from time 0.3 to 2.75 min, held at 95% for 0.5 min, then decreased to 40% at 3.27 min. Effluent from HPLC was introduced to the mass spectrometer via an electrospray ionization source in positive-ion mode. Multiple reaction monitoring was used to monitor ARQ 501 and D6-ARQ 501 using m/z 243.1 > 159.0 and 249.1 > 159.0, respectively. Good linearity was achieved over the concentration range of 3 to 2000 ng/ml with R2 > 0.99 when a weighting of 1/(concentration)2 was used.
LC-MS-ARC. Radioactivity profiles in whole blood samples were monitored using liquid chromatography coupled with an accurate radioisotope counting system. HPLC was performed on an Agilent 1100 Series Modules system coupled to a Packard Radiomatic 500TR series flow scintillation analyzer (PerkinElmer Life and Analytical Sciences) and an accurate radio-isotope counting XFlow controller (AIM Research Co., Newark, DE). Mass spectra were collected with a Waters Quattro LC mass spectrometer operated in positive-ion electrospray mode. The capillary voltage and cone voltage were set to 3.5 kV and 30 V, respectively. The source and desolvation temperatures were 100 and 350°C, respectively. Collision-induced dissociation (CID) experiments were conducted for the selected target masses with argon as the collision gas. The LC-MS-ARC system was controlled by an ARC data system (version 2.6; AIM Research Co., Newark, DE).
The chromatographic separation was achieved using an Atlantis dC18 column (250 mm x 4.6 mm, 5 µm) or an XTerra MS column (250 x 4.6 mm, 5 µm; Waters) at a flow rate of 1.0 ml/min. The XTerra MS column was applied to run the orthogonal chromatographic conditions at pH 3 and 10. The mobile phase effluent was split to allow 20% to the mass spectrometer via the electrospray ionization source. The remaining 80% of the flow was mixed with scintillation cocktail ULTIMA-FLO and analyzed with ARC using a 500-µl liquid cell. The splitting was conducted through a T-piece with an online check valve that allowed the flow to go only in one direction. The mobile phase solvents [water (A) and acetonitrile (B)] were both modified with 0.1% formic acid (pH 3) or 10 mM ammonium hydroxide (pH 10). The gradient used was as follows: solvent B started at 15% and held for 1 min, then linearly increased to 95% from time 1 to 30 min, held at 95% for 4 min, then decreased to 15% at 35 min.
UPLC–Q-Tof MS. High-resolution MS experiments were conducted on a Waters Q-Tof Premier operated in positive-ion electrospray mode. The instrument was calibrated across the mass range of 50 to 1000 Da using a solution of sodium formate (0.3 mM). Data were acquired using a desolvation temperature of 350°C, source temperature of 100°C, and cone voltage of 30 V. Acquisition and analysis of the data were performed using MassLynx software (version 4.1). Full scan time-of-flight MS spectra were acquired for measurement of accurate masses of ARQ 501 and its metabolites. Data were centroided, and mass was corrected during acquisition using an external reference (LockSpray) solution of 500 ng/ml warfarin infused at 5 µl/min, which provided a reference ion at m/z 309.1127 ([M+H]+). One scan of the Lock-Spray channel was interleaved with 10 scans of the analyte channel. Scans were of 0.2-s duration with a 0.1-s interscan delay.
A Waters ACQUITY UPLC using an HSS T3 column (2.1 x 100 mm, 1.8 µm) was used for the separation of metabolites. The mobile phase consisted of a binary solvent system using water (A) and acetonitrile (B), both acidified with 0.1% formic acid. The gradient program began with 5% eluent B and was maintained for 1 min, then ramped linearly to 95% from 1 to 15 min and held for 2 min. The flow rate was 200 µl/min, and the injection volume was 5 µl.
|
Results
|
|---|
In Vitro Studies of ARQ 501 in Whole Blood. When incubated with fresh human whole blood or the RBC fraction at 37°C, more than 90% of ARQ 501 rapidly disappeared within 30 min. However, ARQ 501 was shown to be stable in plasma for 2 h (Fig. 1). Similar results were observed for other species, including mouse, rat, dog, and monkey. Rational explanations for the rapid disappearance of ARQ 501 in whole blood could include sequestration of ARQ 501 in RBCs, covalent binding of ARQ 501 (and/or its metabolites) to blood components, metabolism by enzymes in blood, or a combination of the above processes. In light of this hypothesis, in vitro studies were conducted. ARQ 501 was distributed between human RBCs and plasma with a partition coefficient of 1 (range: 0.82–1.09), which suggested that the sequestration of ARQ 501 into RBCs was not significant. After a 60-min incubation of [14C]ARQ 501 in human whole blood at 37°C (Fig. 2), covalent binding reached a plateau of approximately 23% of the total radioactivity of [14C]ARQ 501 at 10 µM and 100 µM, which represents 11.5 and 112.3 pmol/mg protein at 10 and 100 µM, respectively. Given that ARQ 501 was stable in plasma, these data clearly suggest that metabolism by an enzyme or enzymes located in RBCs was responsible for the rapid disappearance of the ARQ 501 in whole blood.
View larger version (16K):
[in this window]
[in a new window]
|
FIG. 2. The covalent binding of [14C]ARQ 501 to blood components after incubation with human whole blood at 37°C.
|
|
Identification of Metabolites. Radioactivity detection facilitated metabolite profiling, and accurate mass measurement provided structural identification of the metabolites. This study used [14C]ARQ 501 for metabolite identification as well as covalent binding studies. Figure 3 depicts the results from in vitro time course studies on the metabolism of [14C]ARQ 501 in human whole blood. After 60 min of incubation at 37°C, ARQ 501 was not detectable, whereas four major metabolites (M1, M2, M3, and M5) and two minor metabolite peaks (M4 and M6) were observed. At 120 min, the radioactivity (based on ARC peak area) of M1, M2, M3, M4, M5, and M6 was approximately 13, 35, 14, 1, 20, and 5%, respectively, of the total radioactivity from [14C]ARQ 501. These six major radioactive peaks (corresponding to M1 through M6) were found neither in the control samples (collected at time point 0) nor in the incubated plasma samples. Figure 4 illustrates comparative radiochromatograms derived from LC-ARC analysis of samples from an interspecies metabolic study. The metabolite profiles show qualitative but not quantitative similarities among the different studied species. For example, M5 was not a major metabolite in mouse or rat whole blood but was a major one in dog, monkey, and human blood. In addition, an unidentified peak at 11.75 min (labeled with an asterisk) was only observed in dog blood.
View larger version (15K):
[in this window]
[in a new window]
|
FIG. 4. Comparative radiochromatograms of [14C]ARQ 501 incubated with whole blood of different species for 2 h (1, 2, 3... p = M1, M2, M3... parent).
|
|
Detection and identification of the metabolites was based on radioactive peak monitoring and molecular ion detection. The same protocol was conducted with nonlabeled ARQ 501, and the samples were analyzed using UPLC–time-of-flight MS in order to obtain accurate mass measurement and high-resolution product ion mass spectra for each of the selected metabolite peaks. Table 1 summarizes the accurate masses, elemental compositions, and postulated molecular structures of the metabolites.
Interpretation of product ion mass spectra of metabolites frequently hinges upon the similarity of product ions from the metabolite to those of the parent compound. Figure 5 illustrates the product ion mass spectra of ARQ 501 in positive-ion mode. The characteristic product ion at m/z 187 in the CID mass spectrum of m/z 243 (Fig. 5a) indicated a neutral loss of 56 Da by a cross-ring cleavage at the C(2)-O(1) and C(3)-C(4) bonds on the ether ring C, along with charge retention on the rings A and B moiety (Scheme 1). Further fragmentation of m/z 187 led to prominent loss of a carbonyl group, giving rise to another characteristic ion at m/z 159, which in turn eliminated the second carbonyl group to form m/z 131 (Fig. 5b). The product ion at m/z 105 from m/z 159 indicates whether phenyl ring A is modified through metabolism (Fig. 5c). These characteristic fragmentation patterns were applied toward the structural identification of metabolites. In addition, the synthesized standard of ARQ 501 was characterized as: 1H NMR (400 MHz, CDCl3), 
: 8.06 (dd, J = 7.6 Hz, J'= 1.2 Hz, 1H), 7.81 (d, J = 8.0 Hz, 1H), 7.64 (td, J = 8.0 Hz, J'= 1.2 Hz, 1H), 7.50 (td, J = 8.0 Hz, J'= 1.2 Hz, 1H), 2.57 (t, J = 6.4 Hz, 2H), 1.85 (t, J = 6.8 Hz, 2H), 1.47 (s, 6H).
To identify potential pH-sensitive groups (e.g., –OH and –COOH) within the metabolites, the pH of the mobile phase was run at pH 3 and 10, respectively, and retention times were monitored to detect shift. Significant shift in retention time was observed for M1 to M3 but not for M4 to M6 between pH 3 and 10, indicating that pH-sensitive groups were present in M1 to M3 but not in M4 to M6. The chromatographic behavior of the metabolites was integrated with their mass spectral data for structural characterization.
M1. Accurate mass measurement indicated that M1 was produced by oxidation of ARQ 501 (Table 1). The characteristic ion at m/z 203 suggests an intact ether ring C, and the product ions at m/z 175 and 147 from m/z 203 (produced by the elimination of one and two carbonyl groups, respectively) indicate an intact dicarbonyl ring B (Fig. 6). In addition, the product ion at m/z 121 clearly localized the oxidation to the phenyl ring A. However, none of the synthesized oxidative metabolite standards, 7-, 8-, 9-, and 10-hydroxy-ARQ 501, matched the chromatographic retention time of M1, which eluted much earlier than any of the above standards, indicating that M1 was very polar. Therefore, the structure of M1 was not postulated here.
M2. M2 represents most of the radioactivity among the six metabolites detected in the human whole blood sample after 2 h of incubation. Accurate mass measurement indicated that M2 was formed by the introduction of two oxygen and two hydrogen atoms (34 Da) to the ARQ 501 molecule (Table 1). Figure 7 demonstrates the product ion mass spectra of M2 in positive- and negative-ion modes. In positiveion mode, the relatively unstable parent ion at m/z 277 undergoes facile elimination of H2O to produce a major product ion at m/z 259. Ion at m/z 159 derived from the characteristic ion of m/z 203 by losing a CO2 (44 Da) indicates a carboxylic structure.
The product ion mass spectrum in negative-ion mode clearly revealed carboxylic structural information consistent with that of a carboxylic acid (Fig. 7b), because deprotonation in negative-ion mode initially occurred on the carboxylic group and decarboxylation could readily take place on the negatively charged group upon collisional activation (Miao and Metcalfe, 2007). This is further supported by the fact that sp2 carbons can facilitate the loss of CO2 by stabilizing negative charges via the increased s-character of the carbon (Bandu et al., 2004), which indicates a carboxylic group is located in the phenyl ring. The ions at m/z 231 and 187 were generated by subsequent neutral losses of CO2 from the precursor ion m/z 275. Therefore, interpretation of mass spectra led to a structure corresponding to oxidative opening of the carbonyl ring B and the forming of a dicarboxylic acid structure (Table 1). Ultimately, the structure of M2 was confirmed by comparing its retention time and product ion spectra to the synthesized standard (Table 1). The standard compound was characterized as: m.p., 185–187°C, 1H NMR (400 MHz, DMSO-d6) : 12.00 (br, 2H), 7.79 (d, J = 7.2 Hz, 1H), 7.49 (t, J = 7.6 Hz, 1H), 7.40 (t, J = 7.6 Hz, 1H), 7.19 (d, J = 8.0 Hz, 1H), 2.33 (t, J = 6.4 Hz, 2H), 1.69 (t, J = 6.4 Hz, 2H), 1.26 (s, 6H). The International Union of Pure and Applied Chemistry (IUPAC) name of M2 is 6-(2-carboxyphenyl)-2,2-dimethyl-3,4-dihydro-2H-pyran-5-carboxylic acid.
M3. Accurate mass measurement indicated that M3 was formed by the addition of one O and two H atoms (18 Da) to ARQ 501 (Table 1). Figure 8 shows the product ion mass spectra of M3 in positive- and negative-ion modes. Compared with positive-ion mode (Fig. 8a), more structural information was revealed in negative-ion mode (Fig. 8b). One of the major product ions in negative-ion mode was observed at m/z 215 and appeared to arise from a neutral loss of 44 Da, which suggested that a carboxylic group was within the molecule. Interpretation of the product ion mass spectra suggested that the cleavage between C(5)-C(6) of the two carbonyl groups was initiated by an oxidative process that generated a carboxylic as well as an aldehyde group. Further investigation with 1H NMR on synthesized standards confirmed the proposed structure and provided the positional information of the carboxylic group, which was located at position 6 (Scheme 1; Table 1). The standard compound was characterized as: m.p., 146–147°C, 1H NMR (400 MHz, CDCl3) : 9.20 (s, 1H), 8.10 (dd, J = 8.0, 1.2 Hz, 1H), 7.61 (td, J = 7.6, 1.2 Hz, 1H), 7.56 (td, J = 7.6, 1.6 Hz, 1H), 7.40 (dd, J = 7.2, 1.2 Hz, 1H), 2.43 (t, J = 6.8 Hz, 2H), 1.79 (t, J = 6.4 Hz, 2H), 1.37 (s, 6H). The IUPAC name of M3 is 2-(5-formyl-2,2-dimethyl-3,4-dihydro-2H-pyran-6-yl)benzoic acid.
M4. Accurate mass measurement confirmed the elemental composition of M4 as C14H15O3, only one carbon atom different from that of ARQ 501 (Table 1). Figure 9a shows the product ion mass spectrum of M4 and the major fragmentation pathway. The characteristic neutral loss of 56 Da occurred in M4 and generated one major ion, m/z 175, through cross-ring cleavage of the C(2)-O(1) and C(3)-C(4) bonds. The proposed structure of M4 was confirmed by a synthesized standard characterized as: m.p., 115–117°C, 1H NMR (400 MHz, DMSO-d6) : 7.74 (d, J = 7.2 Hz, 1 H), 7.61 (t, J = 8.0 Hz, 1 H), 7.39–7.33 (m, 2 H), 2.45 (t, J = 6.4 Hz, 2 H), 1.87 (t, J = 6.4 Hz, 2 H), 1.41 (s, 6 H). The IUPAC name of M4 is 2,2-dimethyl-3,4-dihydro-2H,5H-pyrano[3,2-c]chromen-5-one.
View larger version (14K):
[in this window]
[in a new window]
|
FIG. 9. Product ion mass spectra of M4 (a), M5 (b), and M6 (c) in positive-ion mode. The insets show the CID mass spectra of m/z 175 (a) and m/z 159 (b), respectively.
|
|
M5. M5 represents the largest peak observed in human blood after a 60-min incubation (Fig. 3). Accurate mass measurement provided the elemental composition of C14H15O2. The difference of 28 Da between M5 and ARQ 501 implies the loss of CO group. Figure 9b illustrates the product ion mass spectrum of M5 and the major fragmentation pathways. The product ion spectrum of [M+H]+ (m/z 215) was dominated by an intense signal at m/z 159, which was produced from the characteristic cross-ring cleavage of the C(2)-O(1) and C(3)-C(4) bonds. This ion (m/z 159) underwent further fragmentation to yield the product ion m/z 131 by a neutral loss of CO. In addition, the product ion at m/z 105 suggests no metabolic modification to phenyl ring A. Hence, the structure of M5 was postulated to occur by the elimination of a carbonyl group through inductive cleavage followed by subsequent formation of a five-membered ring. The structure of M5 was confirmed with a synthesized standard characterized as: m.p., 53–55°C, 1H NMR (400 MHz, DMSO-d6) : 7.41–7.29 (m, 3H), 7.12 (d, J = 6.8 Hz, 1H), 2.20 (t, J = 6.4 Hz, 2H), 1.78 (t, J = 6.4 Hz, 2H), 1.41 (s, 6H). The IUPAC name of M5 is 2,2-dimethyl-3,4-dihydroindeno[1,2-b]pyran-5(2H)-one.
M6. Accurate mass measurement shows that this metabolite has the same elemental composition as M4, C14H15O3 (Table 1). Figure 9c illustrates the product ion mass spectrum of M6 and the major fragmentation pathways. Apparently, the only major ion at m/z 175 was generated by the characteristic cross-ring cleavage with a neutral loss of 56 Da. The ion at m/z 105 confirms that phenyl ring A remained intact and was not the site of metabolism. Upon locating the biotrans-formation that occurred on ring B, an ester structure was assigned to M6 and was confirmed by a synthetic standard. The standard compound was characterized as: 1H NMR (400 MHz, CDCl3) : 8.24 (d, J = 7.8 Hz, 1H), 7.74–7.72 (m, 2H), 7.49–7.47 (m, 1H), 2.65 (t, J = 6.6 Hz, 2H), 1.90 (t, J = 6.6 Hz, 2H), 1.39 (s, 6H). The IUPAC name of M6 is 2,2-dimethyl-3,4-dihydropyrano[3,2-c]isochromen-6(2H)-one.
|
Discussion
|
|---|
Potential reasons for the rapid disappearance of ARQ 501 in blood were investigated by incubating ARQ 501 in freshly drawn blood at 37°C. Metabolic profiling demonstrated that ARQ 501 was extensively metabolized in whole blood under in vitro conditions and that the enzymatic activity was located in the RBC fraction. These results further showed that metabolism by RBCs was a major contributor, whereas covalent protein binding of ARQ 501 and/or its metabolites was a minor contributor (23%) to the loss of ARQ 501 in whole blood.
Interspecies metabolic profiling implied that the enzymes responsible for ARQ 501 metabolism exist in the RBCs of the studied species, but, for each species, they exhibit notably different enzymatic activity. ARQ 501 showed concentration-dependant metabolism in human whole blood; the metabolic rate of ARQ 501 decreased with increasing concentrations over a range of 10 to 500 µM. In addition, ARQ 501 was completely metabolized after a 60-min incubation in human whole blood (Fig. 3), which coincides with the plateau at 60 min in the time course covalent binding study (Fig. 1).
RBCs contain moderate cytochrome P450-like activity due to the existence of various enzymes, such as aldehyde dehydrogenase, esterase, hemoglobin, methyl transferases, N-acetyl transferases, glutathione transferases, etc. (Owen and Natatsu, 1983; Awasthi and Singh, 1984; Mieyal, 1985; Helander and Tottmar, 1987; Ericsson et al., 1999). Atul et al. (2002) reported that the same metabolite of centpropazine was formed in rat liver S9, intestine, and RBCs. It provided support that the RBCs may share many metabolic enzymes with liver and intestine, and can act as an extrahepatic system for drug clearance. In the present study, the metabolites of ARQ 501 observed in blood were not detected in in vitro incubations of ARQ 501 with liver microsomes or hepatocytes in any species. In liver microsomes, the major metabolites were formed through monohydroxylation on the phenyl ring A or the ether ring C. In hepatocytes, most of the metabolites were produced through phase II metabolic pathways, such as glucuronidation, sulfation, and glucosidation (Miao et al., 2007). This suggested a unique biotransformation pathway in blood compared with the hepatic system. Consistent with the previous studies that few phase II metabolites are generated by RBCs (Cossum, 1988), no phase II metabolites were detected in the current study.
Work is under way to identify the specific enzymes responsible for the biotransformation of ARQ 501 in whole blood and to evaluate the metabolic pathways for other ARQ 501 related drug candidates. It is conceivable that formation of M2 occurs via initial oxidative biotrans-formation of quinone; insertion of molecular oxygen by monooxygenase, followed by hydrolysis, would lead to dicarboxylic acid M2. Similar biotransformations were observed for the metabolism of acenaphthene (Vila et al., 2001). Reduction is an important reaction in the metabolism and activation of quinonic drugs and xenobiotics (Testa, 1995). Alternatively, initial reduction of the quinone ring by carbonyl reductase or quinone reductase through a two-electron reduction would lead to unstable hydroxyquinone (Gaikwad et al., 2007). Oxidative cleavage of the hydroxyquinone would be expected to provide the dicarboxylic acid M2, as reported for the metabolism of pyrene (Vila et al., 2001). We have observed that the reduction of ARQ 501 is involved in its metabolism in hepatocytes (Miao et al., 2007). The conjugation metabolites (e.g., glucuronides, sulfates, etc.) of ARQ 501 arise from a two-step process, which requires reduction of the quinone ring B to a hydroquinone, followed by conjugation at the reduced site(s). After the formation of M2, the carboxylic group on the pyran ring C of M2 can be reduced by carboxylic acid reductase, which yields the aldehyde metabolite M3. The selective generation of M3 indicates regioselectivity of the carboxylic acid reductase since the regioisomer of M3 with the carboxylic group at position 5 was not observed. As observed for M2 and M3 of ARQ 501, etoposide, an anticancer agent, was also metabolized to an open ring metabolite when incubated with intact human RBCs (Loo et al., 1987).
The two conjugated lactone regioisomers, M4 and M6, might be generated from ARQ 501 through similar multiple metabolic steps. The formation of M5 could be initiated from M2, followed by spontaneous enzymatic dehydration and ring cyclization. Funabiki et al. (1986) demonstrated several oxygenase model reactions using 3,5-di-tert-butylcatechol as a probe compound. The studied compound was oxygenated with insertion of molecular oxygen to give intra- and extradiol oxygenation intermediates, which further produced pyrone products through eliminating a carbonyl group. Likewise, ARQ 501 can form unstable intermediates containing a seven-membered ring formed through insertion of oxygen extra-diketone, and the intermediates yield M4 and M6, based on the position of the inserted oxygen, through eliminating a carbonyl group. Likewise, M5 can be produced directly from ARQ 501 through the carbonyl elimination pathway. To our knowledge, this novel ring-contraction biotransformation has not been reported previously.
In conclusion, five in vitro metabolites of ARQ 501 were identified in whole blood. The rapid disappearance of ARQ 501 in blood clearly emphasized the importance of separating plasma from RBCs immediately upon collection of clinical blood samples. Further studies have shown that the identified metabolites here were also detected in clinical samples from cancer patients treated with ARQ 501. The next logical step is to ask how important is the blood metabolism for the systemic clearance of ARQ 501. Investigations are ongoing to search for potential phase II metabolites derived from the blood metabolites and to evaluate the contribution of blood metabolism to the in vivo clearance in patients.
|
Footnotes
|
|---|
X.-S.M. and P.S. contributed equally to this work.
Article, publication date, and citation information can be found at http://dmd.aspetjournals.org.
doi:10.1124/dmd.107.018572.
ABBREVIATIONS: ARQ 501, β-lapachone, 3,4-dihydro-2,2-dimethyl-2H-naphthol[1,2-b]pyran-5,6-dione; ARC, accurate radioisotope counting; RBCs, red blood cells; CID, collision-induced dissociation; HPLC, high-performance liquid chromatography; MRM, multiple reaction monitoring; MS, mass spectrometry; NMR, nuclear magnetic resonance; UPLC, ultra-performance liquid chromatography; LC-MS-MS, liquid chromatography/tandem mass spectrometry; LC-MS-ARC, liquid chromatography/mass spectrometry/accurate radioisotope counting.
Address correspondence to: Dr. Thomas Chan, ArQule, Inc., 19 Presidential Way, Woburn, MA 01801. E-mail: tchan{at}arqule.com
|
References
|
|---|
Atul BV, Singh SK, Madhusudanan KP, Paliwal JK, and Gupta RC (2002) Isolation and characterization of metabolites of Centpropazine in rat liver, intestine, and red blood cell homogenates. J Pharm Sci 91: 2067–2075.[CrossRef][Medline]
Awasthi YC and Singh SV (1984) Purification and characterization of a new form of glutathione S-transferase from human erythrocytes. Biochem Biophys Res Commun 125: 1053–1060.[CrossRef][Medline]
Bandu ML, Watkins KR, Bretthauer ML, Moore CA, and Desaire H (2004) Prediction of MS/MS data. 1. A focus on pharmaceuticals containing carboxylic acids. Anal Chem 76: 1746–1753.[Medline]
Bossi D and Giardina B (1996) Red cell physiology. Mol Aspects Med 17: 117–128.[CrossRef][Medline]
Carruthers A and Melchior DL (1988) Effects of lipid environment on membrane transport: the human erythrocyte sugar transport protein lipid/bilayer system. Annu Rev Physiol 50: 257–271.[CrossRef][Medline]
Chau YP, Shiah SG, Don MJ, and Kuo ML (1998) Involvement of hydrogen peroxide in topoisomerase inhibitor β-lapachone-induced apoptosis and differentiation in human leukemia cells. Free Radic Biol Med 24: 660–670.[CrossRef][Medline]
Cossum PA (1988) Role of the red blood cell in drug metabolism. Biopharm Drug Dispos 9: 321–336.[CrossRef][Medline]
Docampo R, Cruz FS, Boveris A, Muniz RP, and Esquivel DM (1979) β-Lapachone enhancement of lipid peroxidation and superoxide anion and hydrogen peroxide formation by sarcoma 180 ascites tumor cells. Biochem Pharmacol 28: 723–728.[CrossRef][Medline]
Ericsson H, Tholander B, and Regårdh CG (1999) In vitro hydrolysis rate and protein binding of clevidipine, a new ultrashort-acting calcium antagonist metabolized by esterases, in different animal species and human. Eur J Pharm Sci 8: 29–37.[Medline]
Funabiki T, Mizoguchi A, Sugimoto T, Tada S, Tsuji M, Sakamoto H, and Yoshida S (1986) Oxygenase model reactions. 1. Intra- and extradiol oxygenation of 3,5-di-tert-butylcatechol catalyzed by (bipyridine) (pyridine) iron(III) complex. J Am Chem Soc 108: 2921–2932.[CrossRef]
Gaikwad NW, Rogan EG, and Cavalieri E (2007) Evidence from ESI-MS for NQO1-catalyzed reduction of estrogen ortho-quinones. Free Radic Biol Med 43: 1289–1298.[CrossRef][Medline]
Helander A and Tottmar O (1987) Metabolism of biogenic aldehydes in isolated human blood cells, platelets and in plasma. Biochem Pharmacol 36: 1077–1082.[CrossRef][Medline]
Hinderling PH (1997) Red blood cells: a neglected compartment in pharmacokinetics and pharmacodynamics. Pharmacol Rev 49: 279–295.[Abstract/Free Full Text]
Hooks MA (1994) Tacrolimus, a new immunosuppressant—a review of the literature. Ann Pharmacother 28: 501–511.[Abstract]
Kizer D, Miao X, Yang R-Y, Wu H, Volckova E, Nguyen K, Ali S, Tandon M, Savage R, Ashwell MA, et al. Synthesis and characterization of the metabolites of ARQ 501 in whole blood. The 10th Annual Land O'Lakes Conference on Drug Metabolism/Applied Pharmacokinetics; 2007 Sep 17-21; Merrimac, WI.
Loo TL, Lu K, and Savaraj N (1987) Erythrocytes as an extrahepatic drug metabolizing tissue. Pharmacol Res 2: 1645–1650.
Miao X-S and Metcalfe CD (2007) Analysis of neutral and acidic pharmaceuticals by liquid chromatography mass spectrometry, in Comprehensive Analytical Chemistry (Petrovic M and Barceló D eds) vol 50, pp 135–156, Elsevier Science B.V., The Netherlands.
Miao X-S, Zhong C, Savage RE, Yang R-Y, Kizer D, Volckova E, Ashwell MA, and Chan TCK. In vitro metabolism of ARQ 501 (β-lapachone) in mammalian hepatocytes. The 10th Annual Land O'Lakes Conference on Drug Metabolism/Applied Pharmacokinetics; 2007 Sep 17-21; Merrimac, WI.
Mieyal JJ (1985) Monooxygenase activity of hemoglobin and myoglobin. Rev Biochem Toxicol 7: 1–66.
Owen JA and Natatsu K (1983) Diacetylmorphine (heroin) hydrolases in human blood. Can J Physiol Pharmacol 61: 870–875.[Medline]
Schaffner-Sabba K, Schmidt-Ruppin KH, Wehrli W, Schuerch AR, and Wasley JW (1984) β-Lapachone: synthesis of derivatives and activities in tumor models. J Med Chem 27: 990–994.[CrossRef][Medline]
Testa B (1995) Biochemistry of redox reactions, in The Metabolism of Drugs and Other Xenobiotics (Testa B and Caldwell J eds) pp 402–405, Academic Press Inc., San Diego.
Vila J, López Z, Sabate J, Minguillón C, Solanas AM, and Grifoll M (2001) Identification of a novel metabolite in the degradation of pyrene by Mycobacterium sp. strain AP1: actions of the isolate on two- and three-ring polycyclic aromatic hydrocarbon. Appl Environ Microbiol 67: 5497–5505.[Abstract/Free Full Text]
Wang A, Li CJ, Reddy PV, and Pardee AB (2005) Cancer chemotherapy by deoxynucleotide depletion and E2F-1 elevation. Cancer Res 65: 7809–7814.[Abstract/Free Full Text]
Wiback SJ and Palsson BD (2002) Extreme pathway analysis of human red blood cell metabolism. Biophys J 83: 808–818.[Medline]
![[Feedback]](/icons/banner/feedbackACT.gif)
![[For Subscribers]](/icons/banner/subscriberACT.gif)
![[Archive]](/icons/banner/archiveACT.gif)
![[Search]](/icons/banner/searchACT.gif)
![[Contents]](/icons/banner/tocACT.gif)
Top
Abstract
Materials and Methods
) and whole blood (
for plasma and 
for RBCs) at 37°C.
