In Vitro Identification of Potential Metabolites of Plinabulin (NPI 2358) in Hepatic Preparations Using Liquid Chromatography-Ion Trap Mass Spectrometry.

Plinabulin (1, NPI2358), a vascular disrupting agent (VDA) molecule, is a synthetic analogue of the natural product phenylahistin (2, NPI 2350), which is isolated from Aspergillus ustus. Evaluation of the in vitro metabolic profile of VDA plinabulin using human liver microsomes (HLMs) and HepaRG Cells Cryopreserved is described. HLMs and HepaRG Cells Cryopreserved were prepared in-house and incubated with plinabulin according to published methodologies. The incubated mixtures were analyzed by liquid chromatography-ion trap mass spectrometry to identify possible metabolic products. The incubated plinabulin (1) revealed the presence of several peaks representing 19 tentative metabolites in HLMs and HepaRG Cells Cryopreserved in the presence of NADPH (nicotinamide adenine dinucleotide phosphate) and in the absence of NADPH-generating system, respectively. However, in NADPH absence, no metabolites and microsomes were generated for 1 in incubated HLMs, indicating a likely involvement of CYP450 enzymes in the metabolism. The metabolite structures, obtained from HLMs and HepaRG Cells Cryopreserved incubations, were elucidated by LC-MS/MS fragmentation study. Seventeen phase-I metabolites were proposed to be the results of isomerization, hydroxylation, hydration, and oxygenation of 1 in HLMs and two isomeric phase-II sulfate conjugate metabolites of 1 in HepaRG Cells Cryopreserved incubation.

Introduction

Plinabulin (1, NPI2358) has been found as a potent vascular disrupting agent (VDA) against tubulin depolymerization and is currently being used in phase III clinical trial for non-small cell lung cancer (NSCLC) therapy.[1,2] Plinabulin as a new drug application in combination with granulocyte colony-stimulating factor for chemotherapy-induced neutropenia prevention has been approved by FDA.[3] Plinabulin enhanced the docetaxel antitumor activity positively with a favorable safety profile, as indicated by results with NSCLC patients,[4] and provided significant clinical benefit to patients with solid tumors as well. According to World Health Organization reports, lung cancer is the most common malignant disease worldwide and is the main cause of death from cancer (estimated to be 901,746 new cases each year), particularly among men.[5] From the data estimated in 2008, among the 16,632 cases diagnosed with lung cancer in the Arab league world, 13,826 cases (79.7%) were males and 2806 (20.3%) were females, and this study predicts that there will be 29,576 new cases in 2020.[6] After reviewing the previous literature reviews that were related to plinabulin, we did not find any study on plinabulin metabolism. During the drug discovery process, various issues with respect to pharmacodynamics, pharmacokinetics, and toxicity are commonly encountered. Drug metabolism is a major criterion in the high-throughput screening of prospective drugs. Hence, the nature of the metabolites produced from the drug must be thoroughly studied. Studying in vitro drug metabolism can be useful in the design of clinical studies, particularly those that examine drug–drug interactions. Previously reported studies on plinabulin were focused only on its synthesis, activity, and clinical trial development, and no research was carried out to identify its metabolites. In vitro metabolism should provide data that closely match that acquired from whole animal investigations because the compounds will be in the early phases of development; therefore, the present study was aimed to carry out the in vitro identification and characterization of plinabulin potential metabolites in hepatic preparations using the liquid chromatography–ion trap mass spectrometry (LC-ITMS/MS) technique. Employing human liver microsomes (HLMs) and HepaRG Cells Cryopreserved, the in vitro metabolic profile of 1 was studied. Thus, 19 proposed metabolites resulted from five main reactions including isomerization, oxidation, reduction, hydration, and conjugation. Among them, 17 were produced by HLMs, while the last 2 were by HepaRG Cells Cryopreserved. The presence of two geometric centers (E–Z) in the structure of 1 led to the formation of several isomer metabolites, which is considered a remarkable phenomenon and worthy of study. The E to Z geometrical isomers are considered the main metabolic pathway of 1 metabolites. Four proposed geometrical isomers (EE, EZ, ZE, and ZZ) were detected after incubation of the parent drug with HLMs and HepaRG Cells Cryopreserved. These had different retention times with the same m/z and the same MS/MS spectra of the parent compound.

Materials and Methods

Chemicals and Reagents

Water was obtained from Milli-Q connected to an Elix Millipore water purification system (Millipore, USA). Acetonitrile HPLC grade, ammonium formate, and magnesium chloride were obtained from BDH laboratory supplies (Poole-UK). Nicotinamide adenine dinucleotide phosphate (NADPH) reduced form were obtained from ACROS, USA. Plinabulin (NPI2358) was obtained from Selleck Chemicals, USA. Pooled male human liver microsomes (M0567) were purchased from Sigma-Aldrich company (West Chester, PA, USA). The protein content of HLMs was 20 mg/mL protein in 250 mM sucrose. HepaRG Cells Cryopreserved (HPRGC10) were purchased from ThermoFisher company (Waltham, MA USA).

In Vitro Studies

HLM Incubations

At pH = 7.4, plinabulin (30 μM) in dimethyl sulfoxide was added to the phosphate buffer (0.08 M) with HLMs and stirred for 5 min at 37 °C for equilibration in water bath. The NADPH solution (1 mM) and microsomal protein (1 mg/mL) were added to start the incubation process (total volume of incubation mixture was 1 mL). NAPDH was replaced with water or microsomal protein to prepare the controls. After incubation (90 min), ice-cold acetonitrile (2 mL) was added to precipitate the proteins. At 14,000 rpm, the sample was centrifuged (10 min), and the supernatant was transferred to a fresh container and then evaporated under a stream of nitrogen. The reconstituted residue was transferred into a HPLC vial for analysis.

Incubation of Plinabulin with HepaRG Cells Cryopreserved

The cell suspensions were preincubated at 37 °C for 10 min, and then 2 mL of this suspension was added to the appropriate container at room temperature (RT) except for the control container. The incubation was initiated by adding 2 μl of a 5mM stock solution of 1. After incubation of the samples (2 h), ice-cold acetonitrile (2 mL) was added to precipitate the proteins. At 14,000 rpm, the sample was centrifuged (10 min), and the supernatant was collected in a fresh container and then evaporated under a stream of nitrogen. The reconstituted residue was transferred into a HPLC vial for analysis. All experiments were duplicated.

Chromatographic Conditions

Employing an Agilent 1200 series system containing of G1311A quaternary pump, a G1322A degasser, a G1367B HIP-ALS autosampler, a G1316A thermostatted column compartment, and a G1315B DAD detector, the separation was performed. An eclipse plus C18 (4.6 × 150 mm 3.5 μm) column and a mobile phase composed of 5 mM ammonium formate in water solvent (A) and acetonitrile solvent (B) were used. At a flow rate of 0.4 mL/min, gradient chromatography (a run of 45 min) was achieved in the water/acetonitrile mixture. 95% of solvent A was used for starting the program, and then within 30 min, solvent B was increased from 5 to 60% and held constant for the next 10 min. The injected sample into the HPLC system was 5 μL.

Mass Spectrometric Conditions

LC–MS/MS measurements were performed using an Agilent Ion Trap system model 6320 (Agilent Technologies, USA) equipped with an electrospray ionization (ESI) source. ESI was done at RT in the negative mode, the capillary voltage was maintained at 4.5 kV, and the capillary temperature was 325 °C. The scan range was set between 100 and 700 Da. The flow rate of the drying gas (N2) was 10 L/min. The nebulizer pressure was at 50 psi. Flow injection analysis was used to measure mass spectrometric parameter optimization to attain the highest ion intensity.

Results and Discussion

LC–MS and LC–MS/MS Analysis of Plinabulin (1)

For determination of the possible molecular weights of metabolites of 1, two possible plans were utilized.[7] In the first one, the possible metabolite structures were hypothesized in light of the metabolism rules of the compounds with similar structures. Then, in the full-scan mass spectrum, the corresponding peaks of potential metabolites were identified by comparing the chromatographic peaks of the blank samples with those of the drug-treated samples. In the second strategy, MS fragmentation behaviors of the parent compound and its potential metabolites were investigated. Another LC–MS/MS screening was performed by constant neutral loss scan monitoring for metabolites which lost neutral adducts. Screening of total ion chromatograms of 1 in the HLM incubated mixture showed a prominent peak eluting at ∼34.7 min (Figure ), with a molecular ion peak at m/z 335 corresponding to compound 1.

Figure 1

(a) TIC of 1; (b) MS spectra of 1; (c) chemical structure of 1.

The obtained MS2 spectra of 1 with LC-ITMS showed that the parent peak [M – H]− with m/z 335 gave many fragments as explained in Figure . In the negative scan mode, plinabulin formed a deprotonated molecule [M – H]− at m/z 335. As illustrated by Figure A, MS2 scan of 1 gave four major fragments which were predominantly formed by cleavage of the amide group from the piperazine ring and cleavage of the CH2N2 side chain from the imidazole ring (Figure ).

Figure 2

(A) Proposed chemical structrure of 1 and its MS/MS spectrum. (B) MS/MS spectrum ions of 1

Identification of Proposed Phase-I Metabolites (M1–M17)

Identification of Stereoisomeric Metabolites M1 and M2

In addition to the peak that appeared with m/z 335 [M – H]− at the retention time of 34.7 min. which indicated parent 1, there are two more peaks (M1 and M2) with m/z 335 [M – H]− at the retention times of 31.8 and 34.2 min, respectively (Figure A). These two peaks demonstrated similar deprotonated molecular ion peaks at m/z 335 [M – H]− and showed same MS/MS fragment ions indicating their isomerization for 1. The MS/MS spectra of M1 and M2 metabolites gave the most abundant ions at m/z 307, 292, and 265 which were similar to the fragment ions of parent 1 (Figure B,C). The presence of the fragment ion at m/z 307 with 28 Da less than the parent drug suggested the loss of the carbonyl group from the piperazine ring. Fragment ion at m/z 292 with 43 Da less than the precursor ion indicated the loss of amide group from the piperazine ring. However, the fragment ion at m/z 265 that lost 70 Da from the parent molecule pointed to the loss of carbonyl group from the piperazine ring and CH2N2 side chain from the imidazole ring. Based on these data and a comparison of the retention times on HPLC and mass spectra, we concluded that M1 and M2 metabolites had known metabolic transformation (stereoisomer) with each other.[8]

Figure 3

(A) Proposed chemical structure of M1/M2 and their MS/MS fragments; (B) MS/MS spectrum of M1; (C) MS/MS spectrum of M2.

Identification of Carbonyl Groups Reduced Metabolites (M3–M5)

M3, M4, and M5 were detected at m/z 337 in a full scan mode with retention times of 39.2, 40.6, and 41.5 min, respectively (Figure A). The mass spectrum analysis indicated that the addition of two hydrogens corresponding to the reduced parent 1 most probably occurred.[9]

Figure 4

(A) Proposed chemical structure of M3, M4, and M5 and their most abundant MS/MS fragments; (B) MS/MS spectrum of M3; (C) MS/MS spectrum of M4; D) MS/MS spectrum of M5.

Upon fragmentation of M3, M4, and M5 at [M – H]− at m/z 337, the MS/MS fragment of m/z 337 gave the most abundant fragment ions with m/z 319 and 301 (Figure B–D). The fragment ion (m/z 319) with 18 Da less than the precursor ion indicated the loss of H2O molecule from the piperazine ring, whereas the fragment ion (m/z 301) with 36 Da less than the precursor ion suggested that two molecules of H2O are lost. In light of these data, metabolites M3–M5 with m/z 337 were tentatively proposed to be isomers as shown in Figure .

Identification of Hydroxy Metabolites in Aromatic Ring (M6 and M7)

As shown in Figure A, M6 and M7 were found at the retention times of 28.7 and 29.3 min, respectively, with the deprotonated molecular ion peaks at m/z 351 [M – H]− in the negative ion mode. Furthermore, M6 and M7 gave 16 Da more than 1 suggesting they were monohydroxylated metabolites for 1.[10]

Figure 5

(A) Proposed chemical structure of M6 and M7 and their most abundant MS/MS fragments; (B) MS/MS spectrum of M6; (C) MS/MS spectrum of M7.

Fragmentation of both M6 and M7 provided the same most abundant fragment ions with m/z 333 and m/z 307 (Figure B,C). The product ion at m/z 333 was thought to be formed by the loss of the H2O molecule (18 Da less than the precursor ion). Fragment ion at m/z 307 with a loss of 44 Da from the precursor ion suggests the loss of hydroxyl group from the benzyl ring and carbonyl group from the piperazine ring. Accordingly, compounds M6 and M7 are proposed to be the hydroxylated metabolites of 1.

Identification of Hydroxy Together with Reduced Metabolites (M8 and M9)

M8 and M9 were detected at m/z 367 [M – H]− in full scan mode with retention times of 28.2 and 30.4 min, respectively (Figure A). The presence of [M–H]− ions at m/z 367 suggested that both M8 and M9 mass values (32 Da) were higher than that of 1, indicating that these two proposed metabolites are double-hydroxylated metabolites of 1.

Figure 6

(A) Proposed chemical structure of M8 and/or M9 and their most abundant MS/MS fragments; (B) MS/MS spectrum M8; (C) MS/MS spectrum of M9.

In the MS2 spectra of M8, the presence of the characteristic ions at m/z 324 and 310, which were as the same as those of M9, suggested that both metabolites are thought to be isomers (Figure B,C). The fragment ion at m/z 324 with 43 Da lower than the precursor ion indicated the loss of amide group from the piperazine ring, while the fragment ion at m/z 310 with 57 Da lower than the precursor ion suggested the loss of amide group from the piperazine ring and hydroxyl group either from the aromatic ring or from the aliphatic chain.

Identification of Hydrated Metabolites M10–M13

The metabolites M10, M11, M12, and M13 eluted at the retention times of 25, 26, 26.8, and 28.6 min, respectively (Figure A), showed the same deprotonated molecular ion peaks at m/z 353 [M – H]− with18 Da higher than that of m/z 335, suggesting that the molecule of H2O has attached on the parent drug.[11]

Figure 7

(A) Proposed chemical structure of M10, M11, M12, and M13 and their most abundant fragments; (B) MS/MS spectrum of M10; (C) MS/MS spectrum of M11; (D) MS/MS spectrum of M12; (E) MS/MS spectrum of M13.

The most abundant fragment ion showed by M10-M13 appeared at m/z 335 with18 Da lower than the precursor ion (Figure B–E), suggesting that this fragment ion might be parent 1. Depending on the obtained data, the four metabolites, namely, M10, M11, M12, and M13 are proposed to be isomers of hydrated 1.

Identification of Oxidized Metabolites (M14–M17)

The metabolites M14–M17 that eluted at 26.2, 27, 28, and 28.9 min, respectively, displayed deprotonated molecular ion peaks at [M – H]− at m/z 365, with 30 Da higher than 1, suggested to be the carboxylic acid metabolites of 1(12) (Figure A).

Figure 8

(A) Proposed chemical structure of M14, M15, M16, and M17 and their most abundant MS/MS fragments; (B) MS/MS spectrum of M14; (C) MS/MS spectrum of M15; (D) MS/MS spectrum of M16; (E) MS/MS spectrum of M17.

The fragmentation of M14–M17metabolites provided the same MS/MS fragment ions with m/z 347 and 335 (Figure B–E). The product with ion (m/z 347) could be formed by losing of the H2O molecule [M-H2O]- with 18 Da lower than the precursor ion, while the fragment ion at m/z 335 with 30 Da lower than the precursor ion suggested the losing of −OOH atoms from the carboxylic acid group. In view of the above data, the proposed metabolites (M14–M17) might be isomers.

Identification of Proposed Phase-II Metabolites

Identification of Sulfate Conjugates (M18 and M19)

The metabolites M18 and M19 eluted at the retention times of 17.8 and 18.3 min, respectively (Figure A) showed the same deprotonated molecular ion peaks at m/z 447 [M – H]− with 80 Da higher than that of the precursor ion with m/z 367, suggesting that these products proposed to be sulfate conjugation metabolites of M8 and M9.[13]

Figure 9

(A) Proposed chemical structure of M18 and M19 and their most abundant fragments; (B) MS/MS spectrum of M18; (C) constant neutral loss scan of M18; (D) MS/MS spectrum of M19; (E) constant neutral loss scan of M19.

The MS2 spectra of the proposed metabolites (M18 and M19) showed the same most abundant fragments at m/z 367 and 401 (Figure B,D). The fragment ion at m/z 401 suggested the loss of aliphatic CH2OH group and H2O molecule. Another LC–MS/MS screening for the sulfate adduct was performed by constant neutral loss scan monitoring of ions that lost 80 Da (Figure C,E).[14] As shown in Figure C,E, only two ions at retention times of 17.8 and 18.3 min with m/z 447 were detected, suggesting their isomerization character (Table ).

Table 1

Proposed Metabolites of Plinabulin (1)

entry	mass m/z [M – H]−	most abundant fragment ions (m/z)	retention time (min)	proposed biotransformation
M1	335.15	307, 292, 265	31.8	isomerization
M2	335.15	307, 292, 265	34.2	isomerization
M3	337.7	319, 301, 257	39.2	reduction
M4	337.7	319, 301, 257	40.6	reduction
M5	337.7	319, 301, 257	41.5	reduction
M6	351.15	307, 333	28.7	monhydroxylation
M7	351.15	307, 333	29.3	monohydroxylation
M8	367.14	324, 310	28.2	dihydroxylation
M9	367.14	324, 310	30.4	dihydroxylation
M10	353.16	335	25.0	hydration
M11	353.16	335	26.0	hydration
M12	353.16	335	26.8	hydration
M13	353.16	335	28.6	hydration
M14	365.13	347, 335	26.2	oxidation
M15	365.13	347, 335	27.0	oxidation
M16	365.13	347, 335	28.0	oxidation
M17	365.13	347, 335	28.9	oxidation
M18	447.10	384, 367	17.8	dihydroxylation and sulfate conjugation
M19	447.10	384, 367	18.3	dihydroxylation and sulfate conjugation

Conclusions

In this work, 17 phase-I metabolites of 1 were determined and characterized by the LC–MS/MS method. Upon incubation of plinabulin with HLMs via in vitro metabolic reactions including isomerization, hydroxylation, hydration, and oxygenation in the presence of a NADPH-generating system, the metabolites of 1 were formed. Upon incubation of 1 with HepaRG Cells Cryopreserved (phase-II), two sulfate conjugates of 1 were detected and characterized by the LC–MS/MS method and constant neutral loss scan. Structures of those metabolites/conjugates were elucidated by comparing the fragmentation behavior study of those metabolites with the parent molecules. In conclusion, this study is an addition to the drug development of potent VDA drug plinabulin.