type II topoisomerase.

http://informahealthcare.com/xen ISSN: 0049-8254 (print), 1366-5928 (electronic) Xenobiotica, 2015; 45(2): 158–170 ! 2015 Informa UK Ltd. DOI: 10.3109/00498254.2014.952799

RESEARCH ARTICLE

In vitro and in vivo metabolism of 14C-AZ11, a novel inhibitor of bacterial DNA gyrase/type II topoisomerase Jian Guo1, Camil Joubran2, Ricardo A. Luzietti Jr.1, Fei Zhou2, Gregory S. Basarab2, and Karthick Vishwanathan1 Department of DMPK and 2Department of Chemistry, Infection Innovative Medicine, AstraZeneca Pharmaceuticals, Waltham, MA, USA

Xenobiotica Downloaded from informahealthcare.com by Yale Dermatologic Surgery on 01/04/15 For personal use only.

1

Abstract

Keywords

1. (2R,4S,4aS)-11-Fluoro-2,4-dimethyl-8-((S)-4-methyl-2-oxooxazolidin-3-yl)-2,4,4a,6-tetrahydro1H,1’H-spiro [isoxazolo[4,5-g][1,4]oxazino[4,3-a]quinoline-5,50 -pyrimidine]-20 ,40 ,60 (30 H)-trione (AZ11) is a novel mode-of-inhibition bacterial topoisomerase inhibitor that entered preclinical development for the treatment of Gram-positive bacteria infection. 2. The in vitro biotransformation studies of AZ11 using mouse, rat, dog and human hepatocytes showed low-intrinsic clearance in all species attributed to microsomal metabolism. 3. After a single intravenous administration of [14C]AZ11 in bile duct cannulated rats, the mean percentage of dose recovered in rat urine, bile and feces was approximately 18, 36 and 42%, respectively. Unchanged AZ11 recovered in rat urine and bile was less than 9% of the dose, indicating that AZ11 underwent extensive metabolism in rats. 4. The most abundant in vivo metabolite detected in urine and bile was M1 formed via ring opening on the piperidine and morpholine rings accounting for 20% of the administered dose. The major fecal metabolite was M5, which accounted for approximately 32% of administered dose. M5 was not formed when AZ11 incubated with rat intestinal microsomes and cytosol but was formed when incubated with fresh rat feces, suggesting that unchanged AZ11 was directly excreted into gut lumen where M5 formed as an intestinal microflora-mediated product. This process could have significant impact on bioavailability or exposure of AZ11 in rat.

AZ11, excretion, mass spectrometry, metabolism

Introduction Type II deoxyibonucleic acid (DNA) topoisomerases are essential cellular enzymes responsible for modulating the topology of DNA, a function vital to DNA replication and repair and protein synthesis. Unlike eukaryotic enzymes that are active as homodimers, the prokaryotic type II topoisomerases, DNA gyrase (Gyr) and topoisomerase IV consist of two subunits, GyrA/GyrB and ParC/ParE, respectively, playing essential roles in DNA replication (Berger et al., 1996; Sanyal & Doig, 2012). Given these distinct features, bacterial DNA topoisomerases have become well-established drug targets for antibiotics (Collin et al., 2011; Sanyal & Doig, 2012). So far, fluoroquinolones (FQs) are one of the most successful antibacterial agents targeting bacterial DNA gyrase and topoisomerase IV in the clinic (Bryskier, 1993). Specifically, FQs form complexes of these enzymes with DNA, consequently, inhibit DNA supercoiling introduced by DNA gyrase and decatenation by topoisomerase IV.

Address for correspondence: Jian Guo, DMPK of Infection Innovative Medicine, AstraZeneca Pharmaceuticals, 35 Gatehouse Drive, Waltham, MA, USA. Tel: +1-781-4725985. E-mail: [email protected]; [email protected]

History Received 15 July 2014 Revised 4 August 2014 Accepted 5 August 2014 Published online 21 August 2014

This stabilizes the DNA cleavable complex for both enzymes preventing bacterial DNA replication. However, the emergence and development of bacterial resistance to existing FQs has become increasingly common (Dalhoff, 2012). The major mechanisms of FQ-resistance involve mutations in the quinolone resistance-determining regions of the FQ target genes, i.e. DNA gyrase and topoisomerase IV, and a decrease in drug accumulation via an efflux-mediated mechanism (Hooper, 2001; Maruri et al., 2012; Shigemura et al., 2012). Therefore, it is extremely important to find new antibacterial agents with novel scaffolds that retain the activities against these resistant bacterial strains. PNU-286607, a spirocyclic pyrimidinetrione moiety connected with tetrahydroquinoline core fused with a morpholine portion, is the first member of a novel class of antibacterial agents that target bacterial DNA gyrase and topoisomerase IV, and exhibit activity against both gram-positive and gramnegative organisms, including methicillin- and quinoloneresistant strains. In addition, no measurable inhibitory effect against a human type II topoisomerase was observed, suggesting that the compound selectively inhibits bacterial enzymes relative to that of human (Miller et al., 2008). Therefore, PNU-286607 represents the first of a promising new class of antibacterial agents (Huband et al., 2007).


DOI: 10.3109/00498254.2014.952799

(2R,4S,4aS)-11-Fluoro-2,4-dimethyl-8-((S)-4-methyl-2oxooxazolidin-3-yl)-2,4,4a,6-tetrahydro-1H,10 H-spiro [isoxazolo[4,5-g][1,4]oxazino[4,3-a]quinoline-5,50 -pyrimidine]20 ,40 ,60 (30 H)-trione (AZ11) incorporates a novel benzisoxazole scaffold with the spirocyclic pyrimidinetrione pharmacophore (Basarab et al., 2013). Biochemical and biological characterizations showed that AZ11 inhibited bacterial DNA replication by targeting the GyrA and ParC subunits of type II topoisomerases and that was not cross-resistant to FQ-resistant strains. The stability studies in liver microsomes showed that AZ11 was stable in the incubations with liver microsomes from preclinical species and human (data not published). Similarly, in pharmacokinetic studies of AZ11 after a single intravenous infusion in mice and rats, AZ11 exhibited low clearance in mouse and rat at 11 and 4 ml/min/ kg, respectively. However, pharmacokinetic study using bile duct cannulated rats showed extremely low-renal clearance and biliary clearance of AZ11, suggesting that AZ11 is not eliminated unchanged via biliary or renal clearance. Therefore, the objectives of the present study were to investigate the in vitro metabolism of [14C]-AZ11 in hepatocytes from mouse, rat, dog and human, and determine excretion routes of [14C]-AZ11 in bile duct-cannulated (BDC) rats after intravenous infusion. The mechanisms for the metabolites formation in vivo were also explored.

Metabolism and excretion of AZ11

159

Synthesis of metabolite standard Preparation of 2-((2 R,4S,4aS)-10-fluoro-9-hydroxy-2,4dimethyl-20 ,40 ,60 -trioxo-2,20 ,30 ,4,4a,40 ,6,60 -octahydro-1H,10 Hspiro[[1,4]oxazino[4,3-a]quinoline-5,50 -pyrimidin]-8-yl)-2oxo-N-(2,2,2-trifluoroethyl)acetamide (M5) A mixture of (2R,4S,4aS)-11-fluoro-2,4-dimethyl-20 ,40 ,60 trioxo-N-(2,2,2-trifluoroethyl)-2,20 ,30 ,4,4a,40 ,6,60 -octahydro1H,10 H-spiro[isoxazolo[4,5-g][1,4]oxazino[4,3-a]quinoline5,50 -pyrimidine]-8-carboxamide (300 mg, 0.58 mmol) and Raneyl Nickel (1.37 g, 11.7 mmol) in Methanol (MeOH) (10 ml) was stirred at room temperature overnight. The mixture was filtered and rinsed with MeOH (10 ml). The filtrate was mixed in hydrochloric acid aqueous solution (aq. 2 M, 1 ml) and stirred at room temperature for another 30 min, then concentrated to give a yellow solid. This crude sample was purified on a reverse phase C-18 column to give the desired product as a yellowish solid. 1H NMR (300 MHz, DMSO-d6) ppm 0.90 (d, J ¼ 6.4 Hz, 3H), 1.13 (d, J ¼ 6.2 Hz 3H), 2.83 (d, J ¼ 14 Hz, 1H), 3.06 (m, 1H), 3.49 (m, 1H), 3.67 (dd, J ¼ 8.9, 6.4 Hz, 1H), 3.75 (m, 1H), 3.94 (d, J ¼ 8.7 Hz, 1H), 4.04 (m, 2H), 4.16 (d, J ¼ 11.9 Hz, 1H); 7.23 (s, 1H), 9.51 (br.s, 1H), 11.49 (br.s, 1H), 11.67 (br.s, 1H), 11.80 (br.s, 1H). High-resolution mass spectrometry measured for C21H19F4N4O7, [MH]– found at m/z 515.1190 (calculated m/z 515.1184).

Materials and methods Chemicals [14C]-AZ11 (>99% radiochemical purity) and authentic standard of M5 [2-((2R,4S,4aS)-10-fluoro-9-hydroxy-2,4dimethyl-20 ,40 ,60 -trioxo-2,20 ,30 ,4,4a,40 ,6,60 -octahydro-1H,10 Hspiro[[1,4]oxazino[4,3-a]quinoline-5,50 -pyrimidine]-8-yl)-2oxo-N-(2,2,2-trifluoroethyl)acetamide] were synthesized at AstraZeneca Pharmaceuticals. HPLC-grade solvents were purchased from Thermo Fisher Scientific (Waltham, MA). All other chemicals were purchased from Sigma-Aldrich (St. Louis, MO). Male CD1 mouse cryopreserved hepatocytes and male Wistar Han rat cryopreserved hepatocytes were purchased from Gibco (Grand Island, NY). Male Beagle dog cryopreserved hepatocytes and human cryopreserved hepatocytes were purchased from CellzDirect (Durham, NC). All other chemicals were of analytical grade or equivalent. Instrumentation The metabolite isolated from rat urine samples for NMR analysis was fractionated based on the indication from on-line mass spectrometer and trapped on a Bruker Metabolic Profiler system equipped with Agilent HP1200 HPLC system (Agilent Technologies, Santa Clara, CA), on-line Prospekt 2 SPE system (Spark Holland, Emmen, The Netherland), Gilson 215 liquid handler (Gilson Inc, Middleton, WI), and Bruker microTOFQ II mass spectrometer (Bruker BioSpin Corporation, Billerica, MA). Nuclear magnetic resonance (NMR) analysis was performed on Bruker Avance III 600 MHz NMR spectrometer (Bruker Daltonics Inc, Billerica, MA) operating at a frequency of 599.80 MHz fitted with a 5-mm TCI cryoprobe.

Collection and isolation of standard metabolite from rat urine samples A single dose of AZ11 formulated as a solution in an aqueous 10% sulfobutylether b-cyclodextrin was administered to three rats at the dose level of 100 mg/kg through intravenous infusion. Urine samples were collected at 0–24 h post-dose. The pooled urine samples were extracted using Thermo Hypersep C18 solid phase extraction column (10 g, 75 ml) (Thermo Scientific, Bellefonte, PA). The column was washed with deionized H2O, and analytes were separated using gradient elution with mobile phase containing water–methanol at different ratio. The portions eluted by 30–50% methanol were mixed and dried for further purification. Individual metabolites were separated using HPLC on a Waters XBridge C18 column (3.5 mm, 4.6 100 mm) with UV detection at 254 nm. The mobile phase consisted of 0.1% formic cid in water (A) and 0.1% formic acid in acetonitrile (B). The metabolites were eluted with a gradient of 20–50% B over 13 min, then to 90% B and held that to 15.5 min then equilibrated at 20% B till 20 min. Automated SPE collection was achieved on HySphere Resin GP-10 SPE cartridges (2 10 mm) (Spark Holland, Emmen, The Netherland) that were preconditioned with acetonitrile followed by water. The target metabolites were monitored using extracted ion chromatography and were collected on SPE cartridges when the signals were above the threshold levels. Metabolite peaks were collected using eight separate 20 ml injections on two SPE cartridges (one for each collected target). One set of cartridges was eluted with deuterated d3-ACN into 1.7 mm NMR tubes using a modified Gilson liquid handler. The second set of loaded cartridges was eluted with deuterated DMSO-d6. H/D exchange experiment was also

160

J. Guo et al.

performed by adding 1 ml of D2O to DMSO-d6-eluted NMR samples. Samples were kept on ice in the dark during sample collection and preparation.

Xenobiotica, 2015; 45(2): 158–170

conducted under aseptic condition. Samples were kept in the dark during sample incubation and preparation.


In vivo studies In vitro studies

Dosing and sample collection

Hepatocyte incubations

A group of bile duct-cannulated (three males) Wistar Hanover rats was housed individually in stainless steel metabolism cages. A single dose of [14C]-AZ11 formulated as a solution in an aqueous 10% sulfobutylether b-cyclodextrin was administered to rats (50 mg/kg) through intravenous infusion. Urine and bile samples were collected at 0–6, 6–24 and 24–48 h post-dose. Feces was collected at 0–24 and 24–48 h. In the circulation study, a single dose of [14C]-AZ11 was administered (three males) at a dose level of 40 mg/kg through i.v. infusion. Blood was collected and processed for plasma at each of the following times: 1, 4 and 8-h post-dose, from single animals. All samples were stored at 20 C before and after analysis. Plasma samples were prepared by centrifugation at 4 C for 10 min at 1500 g within 30 min of blood sampling. Plasma was stored at 70 C before analysis. Samples were kept on ice in the dark during sample collection and preparation.

Mouse, rat, dog or human cryopreserved hepatocytes were thawed and prepared in the Williams E buffer (pH 7.4) containing 2 mM glutamine, insulin transferrin selenium mixtures and 25 mM HEPES. The initial viability of the hepatocytes from mouse, rat, dog and human was more than 80%. Approximately 2 million viable cells in a 1-ml suspension were incubated with [14C]-AZ11 (specific activity approximately 55.8 mCi/mmol) at 1 and 10 mM for 4 h at 37 C. The incubation was terminated by the addition of three volumes of acetonitrile. The resulting mixtures were centrifuged to remove cellular debris. The solvent in supernatant was evaporated to dryness. The residues were reconstituted into 0.3 ml of initial mobile phase used in the liquid chromatographic (LC) analysis. The hepatocyte incubations that were quenched after spiking the test compound in the mediums were used as negative controls. Unlabeled 7-ethoxycoumarin (25 mM) was used as positive control for hepatocytes to assess the activity of enzymes. Samples were kept on ice in the dark during sample collection and preparation. Rat intestinal microsomes and cytosol incubations Incubation solutions contained pooled rat intestinal microsome (2 mg protein/ml) or rat intestinal cytosols (2 mg protein/ml) and 10 mM of AZ11 in 50 mM phosphate buffer (pH 7.4). After pre-incubation at 37 C for 3 min, reactions containing rat liver microsomes were initiated by adding NADPH or NAD (1 mM final concentration) in a total volume of 200 ml and performed at 37 C for 30 min. The incubations were terminated by the addition of 0.6 ml of ice-cold acetonitrile. After centrifugation to remove the precipitated proteins, the supernatant was evaporated to dryness under dry nitrogen stream. Each residue was reconstituted in 200 ml of HPLC mobile phase immediately before analysis using LC-MS/MS and HPLC/UV. Incubation with fresh rat feces and fresh feces from germ-free rats Approximately 1 g of fresh rat feces or feces from germ-free Sprague Dawley rats (Taconic Farms Inc, Hudson, NY) were homogenized with 1 ml of sterile saline and pre-incubated at 37 C for 20 h. The fresh rat feces homogenized with 1 ml of 50% ethanol was used as negative control. Then [14C]-AZ11 (10 mM, 2 mL) was added and incubated for another 20 h at 37 C. The fecal residues were extracted three times with 6 ml methanol:water (80:20). The extracts were combined, and the solvent was reduced to approximately 200 ml under vacuum at 30 C. Then, acetonitrile:water (1:9) was added to the concentrated samples to restore the original sample volume of approximately 1 ml. After centrifugation, 50 ml supernatant of each individual sample was injected onto the LC column for metabolite profile analysis. The experiments were

Radioactivity measurements Radioactivity in urine and bile samples was analyzed directly by liquid scintillation counting on an IrgaSafe liquid scintillation counter (LSC) [PerkinElmer LAS (UK) Limited] by mixing Hionic-Fluor scintillation fluid [PerkinElmer LAS (UK) Limited] to a known amount of the samples. Feces samples were homogenized in an appropriate volume of deionized water. Fecal homogenates and blood were submitted to solubilisation. A suitable volume of solubilising agent (SolueneÕ -350) was added to samples. After an appropriate period of incubation at ca. 50 C, Hionic-Fluor liquid scintillant was added prior to LSC. All radioassays were performed in at least duplicate. Preparation of urine, bile, feces and plasma samples for metabolite profile analysis Urine from 0 to 24 h was pooled proportionally to the original sample volume collected at each time interval to obtain a representative pooled sample. Pooled rat urine samples were centrifuged at 10 000g for 10 min to remove debris and then a 50-ml aliquot of supernatant was injected onto the LC column for metabolite profile analysis. Bile from 0 to 24 h was pooled proportionally to the original sample volume collected at each time interval to obtain a representative pooled sample. A 1-ml aliquot of pooled rat bile sample was mixed with 3 volumes of methanol/acetonitrile (v:v, 1:1) for protein precipitation and desalination. The pellet was washed with methanol/acetonitrile (v:v, 1:1). The solvent in the supernatant was evaporated to dryness under vacuum at 30 C for 4 h. The dry residues were reconstituted in 3 ml water:acetonitrile (90:10) containing 0.05% formic acid. A 50-ml aliquot of reconstituted bile sample was injected onto the LC column for metabolite profile analysis. Fecal homogenates from 0 to 24 and 24 to 48 h pose-dose were pooled separately. Aliquots (3 ml) of pooled samples



DOI: 10.3109/00498254.2014.952799

were centrifuged, and the supernatants were transferred to new vials. The fecal residues were extracted three times with 15 ml methanol:water (80:20). The extracts were combined, and the solvent was reduced to approximately 200 ml under vacuum at 30 C. Then, acetonitrile:water (1:9) was added to the concentrated samples to restore the original sample volume of approximately 3 ml. After centrifugation, 50 ml supernatant of each individual sample was injected onto the LC column for metabolite profile analysis. A 1-ml aliquot of pooled plasma sample was mixed with 3 volumes of acetonitrile for protein precipitation. The solvent in the supernatant was evaporated to dryness under vacuum at 30 C for 4 h using a GeneVac EZ-2 centrifugal vacuum evaporator (GeneVac Inc, Gardiner, NY). The dry residues were reconstituted in 1 ml water:acetonitrile (90:10). A 50-ml aliquot of reconstituted plasma sample was injected onto the LC column for metabolite profile analysis. LC-radioactivity analysis Radiochromatographic analysis of samples from hepatocyte studies and rat mass balance studies was performed on an AcquityTM UPLC system (Waters, Milford, MA). The separation was carried out on an Aquasil C18 column (4.6 150 mm, 5 mm, Thermo Scientific, Waltham, MA) at a flow rate of 1.0 ml/min. The mobile phase consisted of water with 20 mM ammonium formate (pH 4.0) (A) and acetonitrile (B). The analytes were eluted using the following gradient: 10% solvent B for 2 min, a linear increase to 50% solvent B over 33 min followed by a linear increase to 90% B over 8 min, held that composition for 6 min, then the mobile phase composition was returned to the starting solvent mixture over 6 min. The LC elutant was split post-column at approximately a 1:9 ratio between the LTQ-Orbitrap mass spectrometer and a b-RAM Model 5 radio flow detector. A 500-ml liquid scintillation cell was used in the radio flow detector. The flow rate of IN-FLOWTM 2:1 liquid scintillation fluid was 2 ml/min.

161

using an HPLC system described above. Data-dependent analysis was carried out using an LTQ Orbitarp with full scan data acquired at a resolving power of 15 000 in parallel with data-dependent CID MSn scan mode. CID product ion spectra of the three most intense ions were acquired within one acquisition cycle. Helium was used as the collision gas for CID. Normalized collision energies of 16% was set for CID MS2 and MS3. All scan events were detected in the FT detector to give accurate mass data for precursor ions and MSn ions used for metabolite identification.

Results Biotransformation profiles in hepatocytes The representative radiochromatographic profiles from extracts of mouse, rat, dog and human hepatocytes incubations are shown in Figure 1. Unchanged AZ11 accounted for 87, 94, 84 and 92% of the total radioactivity, respectively, after incubation of 10 mM of AZ11 with mouse, rat, dog and human hepatocytes. A total of four metabolites of AZ11 were detected in hepatocytes from mouse and human hepatocytes (Table 1). The metabolites detected in the mouse hepatocyte incubates consisted of a metabolite (M4) where the morpholine ring was excised, a reductive/hydrolytic metabolite M5, and two oxidative metabolites (M6 and M7). The relative amounts of all these metabolites in mouse hepatocyte incubate were less than 4% of the integrated region of interest (ROI) relative to the entire radiochromatogram. M6 was also detected in human hepatocyte incubates (2.1% of integrated

LC-MS and LC-MS/MS analysis Analysis of hepatocytes samples for metabolite identification. Chromatographic separations were carried

out using an Acquity UPLC system described above. Mass spectrometric measurement was performed on a Thermo LTQ-Orbitrap XL mass spectrometer with electrospray ionization in negative ion mode. Data-dependent analysis was preformed with full scan data acquired at a resolving power of 15 000 in parallel with data-dependent collision-induced dissociation (CID) MSn scanning triggered by specific isotopic recognition pattern that was consistent with ratio of unlabeled to [14C]-labeled parent drug used as substrates. The isotopic peak intensity ratio for triggering the dependent scan was set up at M:(M+2) ¼ 1:0.90 (±0.14). CID product ion spectra of the three most intense ions matching the isotopic pattern were acquired within one acquisition cycle. Helium was used as the collision gas for CID. Normalized collision energy of 16% was set for CID MS2 and MS3. Analysis of urine, bile, fecal and plasma samples for metabolite identification. Chromatographic separations were carried out

Figure 1. Representative radioactivity profiles of [14C]-AZ11 metabolites in hepatocyte incubates. (A) mouse, (B) rat, (C) dog and (D) human.

Table 1. Quantitative estimates (% total peak area radioactivity) of AZ11 and metabolites following 3 h incubation in hepatocytes from mouse, rat, dog and human. Percentage of total radioactivity (%)

AZ11 M4 M5 M6 M7 Total (%)

Mouse

Rat

Dog

Human

83.7 2.1 1.2 3.5 3.1 93.6

94.3

92.7

87.2 1.8

94.3

92.7

89.3

162

J. Guo et al.

Xenobiotica, 2015; 45(2): 158–170

ROI). The structures of metabolites found in hepatocytes were interpreted along with metabolites detected in rats. Excretion profiles


The mean recoveries of radioactivity in bile duct-cannulated (BDC) rat urine and feces through 2 days after intravenous administration of [14C]-AZ11 are summarized in Table 2. The mean overall recoveries of total radioactivity were 97% of the administered dose. The mean percentage of dose recovered in rat urine, bile and feces was approximately 18, 36 and 42%, respectively. Radioactivity concentrations in blood and plasma were characterized by peak levels at each sampling time. The ratio of blood to plasma radioactivity were in range of 0.6 to 0.7 at all sampling times, suggesting that there was little or no blood cell binding during the blood sampling regime. Metabolite profiles in rat The urinary metabolite profile from BDC rats after intravenous administration of [14C]-AZ11 is shown in Figure 2(A). The most abundant metabolite of [14C]-AZ11 was identified as metabolite M1, accounting for 5.9% of total dosed radioactivity. M1 is thought to be brought about by a hydration/reduction sequence (see below). Low levels of many other metabolites were also detected, including Phase I metabolites generated by oxidation (M2, M6 and M7), reduction and hydrolysis (M5) and hydration and Table 2. Mean percentage dose recovered in urine, bile and feces from rat after i.v. administration of AZ11. Percentage of radioactive dose (%)

Male (n ¼ 3)

Urine

Bile

Feces

Cage Wash

Total

35.6 (±3.3)

18.3 (±2.2)

42.1 (±11.9)

1.3 (±0.6)

97.3 (±11.3)

Figure 2. Representative radioactivity profiles of [14C]-AZ11 metabolites in (A) urine (0–24 h), (B) bile (0–24 h), (C) feces (0–48 h) from rats after i.v. administration of a single 50 mg/kg dose of [14C]-AZ11, (D) plasma.

O-dealkylation (M3 and M4) that together accounted for 4.6% of the total drug-derived radioactivity in rat (Table 3). The biliary metabolite profile of [14C]-AZ11 in BDC rats is given in Figure 2(B). Only a low level of unchanged AZ11 was detected, and the metabolite profile was similar to that in rat urine in which M1 was the most abundant metabolite. However, there was a high-abundant glucuronide conjugate (M8) that accounted for 8.3% of the total dosed radioactivity (Table 3). The representative radiochromatogram of rat feces after administration of [14C]-AZ11 is shown in Figure 2(C). The most abundant metabolite observed was identified as the reductive/hydrolytic metabolite (M5) accounting for approximately 31% of total dosed radioactivity. Two metabolites M9 and M10 that result from further metabolism of M5 accounted for 7.8 and 2.2% of administered dose, respectively, and were only detected in rat feces. Minor metabolites M1 and M5 were also observed in pooled rat plasma (Figure 2D).

Table 3. Percentage of dose excreted in rat samples identified as AZ11 and its metabolites. Percentage of dose (%) Metabolites AZ11 M1 M2 M3 M4 M5 M6 M7 M8 M9 M10 Total

Urine (0–24 h)

Bile (0–24 h)

7.0 5.9 1.3 0.3 1.1 0.7 0.8 0.4

1.6 14.3 1.7 0.7 2.7

17.5

Feces (0–48 h)

31.1 0.8 0.9 8.3 31.0

7.8 2.2 41.1

Total 8.8 20.2 3.0 1.0 3.8 31.8 1.6 1.3 8.3 7.8 2.2 89.8


DOI: 10.3109/00498254.2014.952799

163

Table 4. Mass spectra data and proposed structures of AZ11 metabolites.


Metabolites

[M–H]a

Dm/z (ppm)

AZ11 M1 M2

512.1190 532.1478 474.1038

1.7 0.3 0.7

M3 M4

472.0874 416.0621

0.1 0.6

M5 M6

515.1194 546.1245

0.4 1.5

M7 M8 M9 M10

528.1144 691.1515 517.1340 434.1003

0.7 0.2 1.9 0.5

MS2 and MS3 product ionsb

Description of metabolites

Samples

Hydration and reduction O-dealkylation and hydrolysis

RU, RB, RP RU, RB, RP RU, RB

O-dealkylation and oxidation Hydration and N-dealkylation

RU, RB MH, RU, RB

Reduction and hydrolysis Hydration and oxidation

MH, RU, RF, RP MH, HH, RU, RB

Oxidation Reduction, hydrolysis and glucuronidation Reduction, hydrolysis and reduction Reduction, imine and amide hydrolyses

MH, RU, RB RB RF RF

2

MS : 426, 387, 301, 247, 205 MS2: 407, 364, 347, 321, 267 MS2: 349, 348, 306, 289, 209 MS3(474!349): 306, 261, 208, 164 MS2: 429, 347, 346, 304, 284, 206 MS2: 373, 291, 290, 248, 205 MS3(416!291): 248, 203, 150 MS2: 472, 429, 375 MS2: 474, 421, 420, 378, 335, 293, 281 MS3(546!281): 191 MS2: 485, 471, 456, 403, 360, 345, 314, 263 MS2: 515, 375 MS2: 517, 377 MS2: 390, 294

MH mouse hepatocytes; HH human hepatocytes; RU rat urine; RB rat bile; RF rat feces; RP rat plasma. a Measured accurate mass of deprotonated molecule (Da). b Major or significant product ions. Figure 3. CID product ion mass spectra of (A) AZ11 ([M–H] ¼ 512). (B) M1 ([M–H] ¼ 532).

Identification of metabolites Based on the radiochromatographic profiles of the metabolites in hepatocytes from four species and in vivo metabolites from rat, a total of 10 metabolites of [14C]-AZ11 were detected, and the structures of prominent metabolites were elucidated by high-resolution accurate mass spectrometry using isotopic pattern recognition, parent mass list or intensity triggered data-dependent product ion scans on a LTQOrbitrap mass spectrometer. The fragment ions of AZ11 and its metabolites are shown in Table 4.

512.1199. The base peak in the product ion spectrum was observed at m/z 247, associated with the loss of amide side chain and the cleavage of pyrimidinetrione in a retro-[2+4] fashion (Figure 3A). The low-abundance ion with m/z ¼ 387 correlated with a characteristic fragmentation generated due to the loss of the amide side chain. Ions with lower abundance were also detected at m/z 301 and 426, corresponding to the cleavage of the pyrimidinetrione and the loss of amide side chain, and the cleavage of pyrimidnetrione, respectively. Hydrogen-deuterium (H/D) exchange analysis indicated that there were three exchangeable protons on AZ11.

AZ11 The deprotonated molecule of Compound I in a full scan highresolution MS (ESI, MH) showed a parent m/z ¼ 512.1190 (Table 4) correlating well with the calculated value of

Metabolite M1 Metabolite M1 was one of the most abundant metabolites detected in rat urine and bile. LC-MS scans of M1 showed a


164

J. Guo et al.

deprotonated molecule at m/z 532, with 20 Da higher than that of parent drug. Accurate mass of deprotonated M1 (observed at m/z 532.1459 within 0.3 ppm mass deviation of the calculated value) supported its empirical formula as C21H22O7N5F4 (equivalent to AZ11 + 4H + O). Additionally, M1 showed six exchangeable protons upon H/D exchange, indicated by a mass increase of +6 in the mass spectrum. The base peak m/z 407 resulted from the characteristic loss of the amide side chain. The fragment ion at m/z 364 resulted from the further dissociation of isocyanate from the pyrimidinetrione (Figure 3B). A minor fragment ion at m/z 267 generated due to the loss of pyrimidinetrione and amide side chain in accord with the assignment for the metabolite with ring-opening of tetrahydroquinoline and morpholine rings. The structure of M1 was further characterized using 1 H-NMR, 1H–1H correlation spectroscopy (COSY) NMR and heteronuclear multiple quantum coherence (HMQC) NMR techniques after preparative chromatographic purification from urine. Two H1 protons were adjacent to

Xenobiotica, 2015; 45(2): 158–170

the aniline-type nitrogen atom as evidenced by the carbon chemical shift of 52 ppm from the HMQC spectrum (Figure 4). The two H1 protons were connected to the same carbon, indicated by the co-linearity of both peaks in the HMQC spectra. The 1H-1H COSY spectrum of M1 (Figure 5) showed the connectivity from the two H1 protons (at ¼ 3.3 and 3.6 ppm) to H2 proton (at ¼ 3.7 ppm). Proton H2 also splits the protons of methyl group (H20 ) into a doublet with coupling constant 2J at 6.5 Hz. Proton H3 (at ¼ 3.9 ppm) splits the second methyl group (H30 , at ¼ 1.2 ppm) into a doublet with coupling constant 2J at 6.5 Hz in the COSY spectrum. The two protons of H4 attach to an oxygen atom supported by the carbon chemical shift of 65 ppm from the HMQC spectrum (Figure 4). The spectrum was void of the OH resonance due to solvent exchange. The two H5 protons were typically observed in the 1H-NMR spectrum of parent compound as an AB pattern. However, these two protons (H5) of M1 appeared as one sharp resonance (Figure 4) due to the presence of a free rotatable benzyl bond. All this information indicated the ring-opening occurred on both morpholine ring and tetrahydroquinoline ring. Therefore, M1 was identified as 7-fluoro-5-((6-hydroxy-2,4-dioxo-1,2,3,4tetrahydropyrimidin-5-yl)methyl)-6-(2-(1-hydroxypropan-2yloxy)propylamino)-N-(2,2,2-trifluoroethyl)benzo[d]isoxazole-3-carboxamide. Metabolite M2

Figure 4. 2D 1H-13C HMQC NMR spectrum of M1 in d3-acetonitrile.

Figure 5. 2D 1H-1H correlation spectroscopy (COSY) NMR spectrum of M1.

LC-MS scans of M2 showed a deprotonated molecule at m/z 474, with 38 Da less than that of parent compound (Figure 6A). Accurate mass of deprotonated M2 (observed at m/z 474.1039 within 0.7 ppm mass deviation of the calculated value) supported its empirical formula as C18H16O6N5F4 (equivalent to AZ11 – C3H6 + 4H) (Table 4). H/D exchange showed that M2 has six exchangeable protons. The diagnostic fragment ion at m/z 349 resulted from the loss of amide side chain and at m/z 263 with subsequent followed by the cleavage of the pyrimidinetrione correlating with M2 resulting from cleavage of the three


DOI: 10.3109/00498254.2014.952799

165


Figure 6. CID product ion mass spectra of M2 ([M–H] ¼ 474). (A) MS2 product ion spectrum (B) MS3 product ion spectrum.

carbon fragment of the morpholine ring. The fragmentation of product ion at m/z 349 generated the MS3 product ion at m/z 208 and 164 that produced by the loss of pyrimidinetrione and further loss of the isopropanol side chain, revealing that two hydrogen atoms were introduced into piperidine ring leading to the ring opening (Figure 6B). Therefore, M2 was tentatively identified as 7-fluoro-6-(2-hydroxypropylamino)N-(2,2,2-trifluoroethyl)-5-((2,4,6-trioxohexahydropyrimidin5-yl)methyl)benzo[d]isoxazole-3-carboxamide. Metabolite M3 LC-MS scans of M3 showed a deprotonated molecule at m/z 472, with 40 Da less than that of the parent drug. Accurate mass of deprotonated M3 (observed at m/z 472.0886 within 0.1 ppm mass deviation of the calculated value) supported its empirical formula as C18H14O6N5F4 (equivalent to AZ11 – C3H6 + 2H), also indicating the cleavage of morpholine ring (Table 4). The high abundant fragment ions at m/z 347 and 261 generated by the loss of the amide side chain and the subsequent cleavage of pyrimidinetrione, respectively, also consistent with the loss of C3H6 group from the morpholine ring for the metabolite. H/D exchange also confirmed that M3 has five exchangeable protons. A diagnostic fragment at m/z 206 generated due to the loss of amide side chain and the pyrimidinetrione ring was in accord with the assignment for the metabolite with ring-opening of tetrahydroquinoline and morpholine rings. Metabolite M4 Metabolite M4 was detected in feces from bile duct-intact rats and in bile from BDC rats. LC-MS scans of M4 showed a deprotonated molecule at m/z 416, with 98 Da less than that of parent drug (Figure 7A). Accurate mass analysis of

deprotonated M4 (observed at m/z 416.0626 within 0.6 ppm mass deviation of the calculated value) supported its empirical formula as C15H10O5N5F4 (equivalent to AZ11–C6H8O). The base peak in the product ion spectrum was observed at m/z 291, associated with the loss of amide side change, again consistent with the loss of C6H8O moiety for M4 (Figure 7A). A characteristic fragment at m/z 151, resulting from the loss of pyrimidinetrione was observed in MS3 spectrum (416 ! 291 !), indicating that the desalkylation occurred on the morpholine ring. Therefore, the data are consistent with the assignment of M4 as 6-amino-7-fluoro-N-(2,2,2trifluoroethyl)-5-((2,4,6-trioxohexahydropyrimidin-5-yl)methyl) benzo[d]isoxazole-3-carboxamide). Metabolite M5 LC-MS scans of M5 showed a deprotonated molecule at m/z 515, with 3 Da higher than that of parent drug. Accurate mass of deprotonated M5 (observed at m/z 515.1188 within 1.4 ppm mass deviation of the calculated value) supported its empirical formula as C21H19O7N4F4 (equivalent to AZ11–N + OH) (Table 4). The base peak in CID product ion spectrum of M5 was at m/z 375 generated by the cleavage of amide bond of the side chain and the loss of CHON (isocyanate) moiety from the pyrimidinetrione. Another diagnostic fragment ion at m/z 429 generated by the loss of two CHON (isocyanate) moieties from the pyrimidinetrione. This is consistent with the M5 structural assignment wherein the benzisoxazole was reduced by cleaving the N–O bond. The resulting imine was hydrolyzed to the ketone. Thereby, the nitrogen atom that was replaced by oxygen was the one from the benzisoxazole ring. Based on these data, M5 was proposed as 2-((2R,4S)-10fluoro-9-hydroxy-2,4-dimethyl-20 ,40 ,60 -trioxo-2,20 ,30 ,4,4a,40 ,6, 60 -octahydro-1H,10 H-spiro[[1,4]oxazino[4,3-a]quinoline-5,50 -

166

J. Guo et al.

Xenobiotica, 2015; 45(2): 158–170


Figure 7. CID product ion mass spectra of (A) M4 ([M–H] ¼ 416). (B) M6 ([M–H] ¼ 546).

pyrimidine]-8-yl)-2-oxo-N-(2,2,2-trifluoroethyl)acetamid. This conclusion was confirmed by comparing retention time and product ion spectrum with that of an authentic synthetic standard. Metabolite M6 LC-MS scans of M6 showed a deprotonated molecule at m/z 546, with 34 Da higher than that of parent drug. The accurate mass analysis of deprotonated M6 (observed at m/z 546.1251 within 0.5 ppm mass deviation of the calculated value) gave its empirical formula as C21H20O8N5F4 (equivalent to AZ11 + 2O + 2H) associated with a ring opening hydrolysis and an oxidation (Figure 7B). The characteristic fragment ion at m/z 421 was generated by the loss of carboxamide substituent and the fragment at m/z 281 generated by the further loss of methylene pyrimidinetrione consistent with the oxidation occurring on the benzisoxazole-tetrahydroquinoline-morpholine ring system and the ring-opening of tetrahydroquinoline. Furthermore, fragmentation of the ion at m/z 281 generated MS3 product ion at m/z 191 due to the loss of C3H6O3 moiety, indicating that the oxidative reactions occurred on the morpholine ring with two oxygen atoms introduced (Figure 7B). H/D exchange also indicating that M6 has six exchangeable protons, very likely that oxidative reactions on morphaline ring resulted in the ring. Metabolite M7 Metabolite M7 was detected in urine and bile from bile-duct cannulated rats and showed a deprotonated molecule at m/z 528, with 16 Da higher than that of the parent drug. Accurate mass analysis of deprotonated M7 (observed at m/z 528.1144 within 0.7 ppm mass deviation of the calculated value) was consistent with the empirical formula as C21H18O7N5F4 (equivalent to AZ11 + O), suggesting that M7 was a monooxidative metabolite of AZ11 (Table 4). The accurate

mass analysis of a diagnostic product ion at m/z 456.0928 gave an empirical elemental composition as C18H14O5N5F4 (equivalent to M7 – C3H6O2 + 2H). The base peak in product ion spectrum at m/z 263 was generated due to the loss of carboxamide and the cleavage of pyrimidinetrione in a retro[2+4] fashion. These typical fragmentation patterns were also observed in AZ11, suggesting that the oxidation occurred on the morpholine ring. The exact location of oxidation could not be determined from mass spectra. Metabolite M8 Metabolite M8 was detected in rat bile. M8 showed a deprotonated molecule at m/z 691, with 179 Da higher than that of the parent drug. The accurate mass of deprotonated M8 (observed at m/z 691.1515 within 0.2 ppm mass deviation of the calculated value) suggested its empirical formula as C27H27O13N4F4 (equivalent to AZ11 + C6H9O7 – N). The fragment ion at m/z 515 was 176 Da less than corresponding precursor ion of M8, indicating that M8 was a glucuronide conjugate (Table 4). The product ions at m/z 375 and 515 were identical to the MS2 spectrum of M5. Therefore, M8 was identified as a glucuronide conjugate of M5 (Table 1). Metabolite M9 Metabolite M9 was only detected in feces from BDC rats. M9 displayed a deprotonated molecule at m/z 517, with 5 Da higher than that of the parent drug. Accurate mass measurement of M9 gave a molecular ion at m/z 517.1342 and an empirical formula of C21H21O7N4F4 (equivalent to AZ11 + 3H + O – N) (D ¼ 1.9 ppm), revealing that M9 was formed through the same sort of metabolic reaction of M5 that involved the benzisoxazole N–O reduction, hydrolysis of the imine to the carbonyl. The fragment ion at m/z 377 was generated by the retro [2+4] cleavage of the pyrimidinetrione (Table 4).



DOI: 10.3109/00498254.2014.952799

167

Metabolite M10

Discussion

Metabolite M10 was detected in rat fecal samples. M10 showed a deprotonated molecule at m/z 434, with 78 Da less than that of the parent compound (Table 4). Accurate mass analysis of M10 gave a deprotonated molecular ion at m/z 434.1003 and an empirical formula of C19H17O8N3F (equivalent to AZ11 – C2HN2F3 + 2O) (D ¼ 0.5 ppm), indicating hydrolysis of the amide bond and the loss of trifluoroethanamine portion of the side chain relative to metabolite M5. This assumption was supported by a minor fragment ion at m/z 390.1 generated by decarboxylation. In addition, the base peak in CID product ion spectrum of M10 at m/z 294 generated by the retro [2+4] loss of pyrimidinetrione also supported the assigned structure (Table 4). Therefore, M10 was tentatively identified as 2-((2R,4S)-10-fluoro-9-hydroxy-2,4-dimethyl20 ,40 ,60 -trioxo-2,20 ,30 ,4,4a,40 ,6,60 -octahydro-1H,10 H-spiro[[1,4] oxazino[4,3-a]quinoline-5,50 -pyrimidine]-8-yl)-2-oxoacetic acid.

To understand the selection of animal species for toxicology studies and address potential safety concerns, the metabolism of AZ11 was examined in mouse, rat, dog and human hepatocytes in the current study. A total of four minor metabolites were observed in mouse and human hepatocytes. The metabolite detected in human hepatocytes was identified as an oxidative metabolite (M6) that accounted for less than 3% of total integrated radioacitivity. Other metabolites, such as M4, M5 and M7, were detected in mouse hepatocyte incubates. No major metabolite was detected in rat or dog hepatocyte incubates. These observations suggested that liver P450 enzymes do not predominantly contribute to the metabolism of AZ11 in preclinical species and human. After a single intravenous administration of [14C]AZ11 to rats, most of the radioactivity was recovered in urine, bile and feces by 24 h post-dose. The total recovery of radioactivity was approximately 97% of the administered dose. An unexpected observation was that 42.1% of administered dose was recovered in rat feces other than the radioactivity recovered in rat urine and bile, suggesting that feces is a major excretion route of AZ11. Given that the rats were dosed through intravenous infusion, and bile samples were collected via cannulation, the plausible pathway through which the AZ11 related radioactivity eliminated into gastrointestinal tract is that (1) AZ11 and its metabolites crossed the gut membrane and excreted into gut lumen, (2) AZ11 was metabolized by intestinal metabolic enzymes, and the metabolites were then eliminated into gut lumen or (3) unchanged AZ11 crossed the gut membrane and was metabolized by the gut microflora. Comparison of the metabolite profile of AZ11 in different matrices showed that the metabolites M9 and M10 detected in rat feces were not observed in rat urine and bile samples. Only a trace of metabolite M5, the predominate fecal metabolite, was seen in the urine. The structure elucidation of the three fecal metabolites also revealed that the metabolites were derived from the reduction and subsequent hydrolysis of the benzisoxazole ring that was mostly not observed in metabolites detected in urine and most metabolites in bile samples. Meanwhile, no metabolites were detected after incubation in rat intestinal microsomes and cytosol, suggesting that formation of the metabolites found in fecal samples was not catalyzed by intestinal metabolic enzymes. These results suggested that the metabolites observed in rat feces appeared to be formed via a route in which unchanged AZ11 crossed the gut membrane into gut lumen where it was metabolized by the gut bacteria. This hypothesis was confirmed in an assay in which the same metabolites of AZ11 were observed after incubation in fresh rat feces, but not in the incubation in feces from germ-free rats. The negative results obtained in the incubation in homogenate containing fresh rat feces and 50% ethanol that had the same chemical environment as that in fresh rat feces, also ruled out the possibility that fecal metabolites were formed due to chemical degradation. AZ11 was incubated with E. coli, a very common bacterium in gut, under anaerobic and aerobic condition. No metabolites were detected (data not shown), suggesting that at least E. coli did not contributed to the metabolite formation in rat

Rat intestinal microsomes and cytosol incubations No metabolites of AZ11 were detected after incubation in pooled rat intestinal microsomes (2 mg protein/ml) or rat intestine cytosol (2 mg protein/ml) (data not shown). Incubation with fresh rat feces and the feces from germ-free rats The representative radiochromatographic profiles from extracts of rat fecal incubations are shown in Figure 8. Except AZ11, three metabolites M5, M9 and M10 were detected in incubate with fresh rat feces. Among them, M5 was the major metabolite that account for approximately 75% to the total detected radioactivity. After incubation in feces from germ-free rats, no major metabolite was observed. Meanwhile, no metabolite was detected in incubate that was homogenized with 50% ethanol undergoing similar incubation.

Figure 8. Representative radioactivity profiles of AZ11 metabolites in the incubation with (A) fresh rat feces, (B) germ-free rat feces and (C) fresh rat feces with 50% ethanol for 24 h.


168

J. Guo et al.

fecal samples. In addition, as the population of bacteria in upper gastrointestinal compartment in rat are much more than that live in human intestine (Sousa et al., 2008), it is very likely that the bioavailability of AZ11 in human is higher than that in rats after oral administration, leading to the higher exposure of AZ11 in human. This factor should be considered in the prediction of human dose in clinical trials. The identification of metabolites was achieved by accurate mass measurement and by analysis of their characteristic fragment ions. In addition to the parent drug, a total of 10 metabolites were observed by radiochemical detection and tentatively identified. The proposed scheme for the in vivo biotransformation of AZ11 is shown in Scheme 1. The most prominent metabolite M1 found in rat urine and bile samples was not observed in rat or human hepatocytes. The structure elucidation was based on accurate mass spectrometry and NMR analysis. It is proposed that M1 formed via acidmediated retro-Mannich reaction opening of the tetrahydroquinoline ring to produce an iminium cation that underwent hydrolysis by water to unravel the morpholine ring. The resulting aldehyde would then be reduced to the primary alcohol. A plausible chemical mechanism for the formation of M1 is illustrated in Scheme 2. The initial retro-Mannich step in which spirocyclic pyrimidinetrione reverted to iminium ion is the reversion reaction of cyclization reaction used from the preparation of the spirocycle scaffold and does occur at higher

Xenobiotica, 2015; 45(2): 158–170

temperatures (Ruble et al., 2009). Thus, it is very possible that the intermediate steps to M1 are not associated with metabolic enzymes. The final step to M1 would involve a reduction reaction that might be the reason why M1 was not seen in hepatocyte incubations. Together with the major metabolite M5, other two metabolites M9 and M10 were observed in rat fecal samples. The formation of these analogues involved reductive ring cleavage of the benzisoxazole ring. The resultant imine would then be hydrolyzed to the carbonyl. The results from in vitro experiments using fresh rat feces and feces from germ-free rats demonstrated that these biotransformation reactions are very likely catalyzed by the intestinal bacteria in rats. Since trace M5 was seen in urine and glucuronide M8 was seen in bile, it can be speculated that some M5 generated in the gut was absorbed systemically and either excreted in the urine or conjugated as the glucuronide. Further, the biotransformation reactions involving gut bacteria catalyzed isoxazole-ring opening have been observed previously as well (Kitamura et al., 1997; Mannens et al., 1993). In summary, no major metabolite of AZ11 was observed in hepatocytes from mouse, rat, dog and human due to slow and low metabolism. After a single intravenous administration of [14C]AZ11 to BDC rats, the large portion of drug-related radioactivity was eliminated via feces along with urine and bile. The high proportion of radioactivity in the feces suggests

Scheme 1. The proposed metabolic pathways of AZ11 in rat. The asterisk (*) indicates the position of

14

C on AZ11.


DOI: 10.3109/00498254.2014.952799

169


Scheme 2. The proposed mechanism for the formation of M1.

that the fecal route plays an important role in the overall elimination of metabolites of AZ11. The metabolite profiles of AZ11 in bile and urine showed that the major biotransformation pathways of AZ11 in rats involved ring-opening on the hydrolytic unraveling of the tetrahydroisoquinoline and morpholine rings followed by an enzyme mediated reduction, indicating that P450s is not the primarily responsible for formation of the most abundant metabolites in rats. These findings also showed slow and low-hepatic clearance of AZ11 in vitro and low-renal and biliary clearance in vivo. Gut efflux mediated elimination of AZ11 and metabolism by gut microflora more significantly contributed to the elimination of AZ11 in rats.

Acknowledgements Authors thank Ying Liu and Vladimir Capka for separation and purification of the renal metabolites for NMR analysis; Ryan Bragg and Nick Bushby for providing 14C-labeled standard; Linda G. Otterson and Michael D. Huband for validating the assay using feces from germ-free rats and their scientific input in microbiology.

Declaration of interest The authors report no conflicts of interest. The authors alone are responsible for the content and writing of this article.

References Basarab GS, Barvian K, Beaudoin M-E, et al. (2013). Novel DNA gyrase inhibiting antibacterial agents: fusion of a benzisoxazole with a spirocyclic octahydropyrido-oxazine pyrimidinetrione. ICAAC F1–1129. Berger JM, Gamblin SJ, Harrison SC, Wang JC. (1996). Structure and mechanism of DNA topoisomerase II. Nature 379:225–32. Bryskier A. (1993). Fluoroquinolones: mechanisms of action and resistance. Int J Antimicrob Agents 2:151–83. Collin F, Karkare S, Maxwell A. (2011). Exploiting bacterial DNA gyrase as a drug target: current state and perspectives. Appl Microbiol Biotechnol 92:479–97. Dalhoff A. (2012). Global fluoroquinolone resistance epidemiology and implictions for clinical use. Interdiscip Perspect Infect Dis 2012: 976273 (1–37). Hooper DC. (2001). Emerging mechanisms of fluoroquinolone resistance. Emerg Infect Dis 7:337–41. Huband MD, Cohen MA, Zurack M, et al. (2007). In vitro and in vivo activities of PD 0305970 and PD 0326448, new bacterial gyrase/topoisomerase inhibitors with potent antibacterial activities versus multidrug-resistant gram-positive and fastidious organism groups. Antimicrob Agents Chemother 51: 1191–201. Kitamura S, Sugihara K, Kuwasako M, Tatsumi K. (1997). The role of mammalian intestinal bacteria in the reductive metabolism of zonisamide. J Pharm Pharmacol 49:253–6. Mannens G, Huang ML, Meuldermans W, et al. (1993). Absorption, metabolism, and excretion of risperidone in humans. Drug Metab Dispos 21:1134–41. Maruri F, Sterling TR, Kaiga AW, et al. (2012). A systematic review of gyrase mutations associated with fluoroquinolone-resistant

170

J. Guo et al.


Mycobacterium tuberculosis and a proposed gyrase numbering system. J Antimicrob Chemother 67:819–31. Miller AA, Bundy GL, Mott JE, et al. (2008). Discovery and characterization of QPT-1, the progenitor of a new class of bacterial topoisomerase inhibitors. Antimicrob Agents Chemother 52:2806–12. Ruble CJ, Hurd AR, Johnson TA, et al. (2009). Synthesis of (-)-PNU286607 by asymmetric cyclization of alkylidene barbiturates. J Am Chem Soc 131:3991–7.

Xenobiotica, 2015; 45(2): 158–170

Sanyal G, Doig P. (2012). Bacterial DNA replication enzymes as targets for antibacterial drug discovery. Expert Opin Drug Discov 7:327–39. Shigemura K, Tanaka K, Yamamichi F, et al. (2012). Does mutation in gyrA and/or parC or efflux pump expression play the main role in fluoroquinolone resistance in Escherichia coli urinary tract infections? A statistical analysis study. Int J Antimicrob Agents 40:516–20. Sousa T, Paterson R, Moore V, et al. (2008). The gastrointestinal microbiota as a site for the biotransformation of drugs. Int J Pharm 363:1–20.

Topoisomerase II and leukemia.

Topoisomerase II mediates meiotic crossover interference.

Developmental regulation of Drosophila DNA topoisomerase II.

Catalytic function of DNA topoisomerase II.

MicroRNA-139 suppresses proliferation in luminal type breast cancer cells by targeting Topoisomerase II alpha.

African swine fever virus ORF P1192R codes for a functional type II DNA topoisomerase.

Topoisomerase II cleavage of herpes simplex virus type 1 DNA in vivo is replication dependent.

Structure and partial amino acid sequence of calf thymus DNA topoisomerase II: comparison with other type II enzymes.

Nitric oxide inhibits topoisomerase II activity and induces resistance to topoisomerase II-poisons in human tumor cells.

Synthesis and evaluation of (18)F-labeled ATP competitive inhibitors of topoisomerase II as probes for imaging topoisomerase II expression.

Tacrine derivatives as dual topoisomerase I and II catalytic inhibitors.

Topoisomerase II inhibition suppresses the proliferation of telomerase-negative cancers.

Mechanisms behind Topoisomerase II SUMOylation in chromosome segregation.

Topoisomerase II inhibitors from the roots of Stellera chamaejasme L.

Ciprofloxacin-induced inhibition of topoisomerase II in human lymphoblastoid cells.

DNA break mapping reveals topoisomerase II activity genome-wide.

Eukaryotic topoisomerase II cleavage of parallel stranded DNA tetraplexes.

8-methoxycaffeine inhibition of Drosophila DNA topoisomerase II.

Regulation of topoisomerase II by murine mastocytoma cells.

The role of topoisomerase II in drug resistance.

Drosophila topoisomerase II-DNA interactions are affected by DNA structure.

Human cytomegalovirus induces expression of cellular topoisomerase II.

Differential Targeting of Human Topoisomerase II Isoforms with Small Molecules.

Chromosome assembly in vitro: topoisomerase II is required for condensation.