Development and validation of a simple and sensitive high resolution LC/MS method for determination of PF-04620110 in dog plasma: application to the pharmacokinetic study

Fan He*, Wei Zheng, Yongzhuang Chen, Chengke Mo, Yilu Chen

Guangzhou Women and Children’s Medical Center, No. 9 Jinsui Road, Guangzhou 510623, China

Correspondence

Fan He, Ph. D.

Guangzhou Women and Children’s Medical Center, No. 9 Jinsui Road, Guangzhou 510623, China
Tel: +86-020-38076262; Fax: +86-020-38076262

Email: [email protected] (F. He)

ABSTRACT
In this study, a more sensitive and reliable quantitative method based on ultra-high performance liquid chromatography coupled with Q-Exactive-Orbitrap-MS in full-mass scan was developed and validated for the determination of PF-04620110 in dog plasma. After protein precipitation with acetonitrile, the samples separations were carried out on an ACQUITY BEH C18 column with 1 mM ammonium acetate in water and acetonitrile containing 0.1% acetic acid as mobile phase, at a flow rate of 0.4 mL/min. The assay showed excellent linearity over the concentration range of 1-2000 ng/mL with correlation coefficient more than 0.9980 (r > 0.9980). The LLOQ was 1 ng/mL. The inter- and intra-day precision (RSD, %) was within 9.69% while the accuracy (RE, %) was in the range of -8.59-11.24%. The extraction recovery was > 85.37% and the assay was free of matrix effects. PF-04620110 was demonstrated to be stable under various processing and handing conditions. The validated method was successfully applied to the pharmacokinetic study of PF-04620110 in dogs and the results revealed that PF-04620110 was slowly eliminated from plasma with clearance of 60.81 ± 7.11 mL/h/kg for intravenous administration and 81.44 ± 25.79 mL/h/kg for oral administration. The oral bioavailability was determined to be 77.89% in dogs.
Keywords: PF-04620110, pharmacokinetics, dog, LC/MS

1. Introduction
Acyl-CoA/diacylglycerol acyltransferases (DGAT) play a key role in the biosynthesis of triglycerides from diacylglycerol and fatty acyl-CoA (Cao et al., 2011; King et al., 2009). DGAT1 was highly expressed in the small intestine and contributed to the most of the intestinal DAGT activity (Cases et al., 2001). It has been demonstrated that inhibition of DGAT1 could be a new therapeutic approach for hypertriglyceridemia, obesity and diabetes (Subauste and Burant, 2003). PF-04620110 (trans-4-[4-(4-Amino-7,8-dihydro-5-oxopyrimido[5,4-f][1,4]oxazepin-6(5H)-yl)phenyl]-cycl ohexaneacetic acid) developed by Pfizer Global Research was a novel, selective and potent DGAT1 inhibitor with IC50 of 19 nM (Dow et al., 2011; Dow et al., 2013). Enayetallah et al. reported that PF-04620110 significantly decreased the level of triglycerides in rat plasma after oral administration (Enayetallah et al., 2011). Additionally, PF-04620110 could decrease the blood glucose levels by increasing the amount of insulin (Zhang and Ren, 2011).
Pharmacokinetic studies are of great importance in drug discovery and development, andin some cases, unwanted pharmacokinetic behaviors can lead to drug development failure or drug withdrawal from market (Yengi et al., 2007; He and Wan, 2018). Although PF-04620110 was demonstrated to be pharmacologically active in hypertriglyceridemia, obesity and diabetes, the pharmacokinetic profiles of this drug were less explored. Lee et al., reported the pharmacokinetic characteristics of PF-04620110 in rats, which demonstrated that PF-04620110 had a high in vivo exposure, low clearance and high oral bioavailability (Lee et al., 2012). These informations would be valuable for us to understand the in vivo disposition of this drug. However, pharmacokinetics would be species-dependent in some cases, making the prediction of human pharmacokinetics difficult (Martignoni et al., 2006; He and Wan 2018). To the best of knowledge, no information was available for the pharmacokinetics of PF-04620110 in dogs. Hence, it would be meaningful to obtain the pharmacokinetic information of this drug in dogs for better understanding its in vivo disposition and for predicting human pharmacokinetics. As the concentration of drug in plasma is very low, a sensitive quantitative method is necessary for pharmacokinetic study. In recent years, liquid chromatography-mass spectrometry (LC/MS) is frequently used for the determination of drugn biological matrices due to its sensitivity and selectivity. Lee et al. developed and validated an LC-MS/MS method for the determination of PF-04620110 in rat plasma (Lee et al., 2012). However, the sensitivity (LLOQ 50 ng/mL) of this method was poor, which limited its application. Hence, it was necessary to develop a more sensitive LC/MS method for the determination of PF-04620110 in dog plasma.

The purpose of this study was to develop an ultra-high performance liquid chromatography couple with Q-Exactive-Orbitrap-MS (UHPLC-Q-Exactive-Orbitrap-MS) for the determination of PF-04620110 in dog plasma. The validated method was further successfully applied to the pharmacokinetic study of PF-04620110 in dogs after intravenous and oral administration. As far as we know, this was the first report on the pharmacokinetic profiles of PF-04620110 in dogs.

2. Materials and methods

2.1. Chemicals and reagents
PF-04620110 with purity more than 98% was purchased from Multi Science (Lianke) Biotechnology Co. Ltd (Hangzhou, China). GW9508 (internal standard, IS) with purity > 98% was obtained from Selleck Chemicals (Shanghai, China). Acetonitrile and acetic acid were of HPLC- grade and purchased from Fisher Scientific (NJ, USA). Ultra-purified water was generated by Elga purification system (Purelab Elga, Britain). All other chemicals and reagents were of analytical grade and from commercial sources.

2.2. Instruments and analytical conditions
The LC system was Thermo Dionex Ultimate 3000 UHPLC system (Thermo Scientific, USA) equipped with a solvent degasser, a quaternary pump, an auto-sampler and a column compartment. Chromatographic separation was performed on Waters ACQUITY BEH C18 column (1.7 μm, 2.1 × 50 mm) which was kept at a temperature of 37 oC. Mobile phase was composed of 1 mM ammonium acetate in water (A) and acetonitrile containing 0.1% acetic acid (B), with a flow rate of 0.4 mL/min. The gradient programs were optimized as follows: 10% B at 0-1 min, 10-45% B at 1-5 min, 45-90% B at 5-8 min, and 10% B at 8-10 min. The injection volume was 2 μL.Mass detection was carried out on a Thermo Q-Exactive Orbitrap mass spectrometer (Thermo Scientific, USA) coupled with an electrospray ionization (ESI) source operated in positive ion mode. The data were acquired simultaneously in full scan with high resolution acquisition and dd-MS2 modes. The full scan data were recorded from m/z 100 to 800 Da with a resolution of 70000 FWHM. The dd-MS2 data were obtained with a resolution of 17500 FWHM and the collision energy was set at 35 eV. The source conditions were optimized as follows: spray voltage 3.5 kV, sheath gas 35 arb, auxiliary gas 10 arb, sweep gas 5 arb, capillary temperature 300 oC, and auxiliary gas temperature 200 oC. The parameters for integration were set as follows: smoothing points 11, baseline window 30, peak noise factor 5, tailing factor 0.95 and mass tolerance 5 ppm. All operations were controlled by Xcalibur 2.2 software (Thermo Scientific, USA).

2.3. Plasma sample preparation
The plasma samples were processed with acetonitrile. Briefly, an aliquot of 50 μL of plasma sample was mixed with 10 μL of IS working solution and then 300 μL of acetonitrile was added into the mixture. After vortexing for 1 min, the sample was centrifuged at 15000 g for 20 min, and 250 μL of the supernatant was taken out and dried under nitrogen gas at room temperature. Finally, the residue was dissolved with 150 μL of acetonitrile-water (v/v, 10:90) and then centrifuged again at 15000 g for 20 min, and 2 μL of the supernatant was injected into LC/MS system for analysis.
2.4. Preparation of stock solutions, calibrators and quality control samples The stock solution containing PF-04620110 at a concentration of 0.5 mg/mL was prepared by dissolving accurately weighed reference standard in acetonitrile. Subsequently, the stock solution was serially diluted with acetonitrile-water (v/v, 50:50) solution to 25, 125, 250, 1250, 2500, 12500, 25000, and 50000 ng/mL as working solutions. The IS working solution (1 μg/mL) was prepared by dissolving GW9805 in acetonitrile and then diluted with acetonitrile-water (v/v, 50:50). The calibration curve was prepared by spiking 2 μL of each working solution into 50 μL blank dog plasma, resulting in the nominal concentration of 1, 5, 10, 50, 100, 500, 1000, and 2000 ng/mL. The quality control (QC) samples were separately prepared as the same procedure of the calibrators at 3, 160 and 1600 ng/mL. All the solutionswere stored at 4 oC before use.

2.5. Bioanalytical method validation
The newly developed analytical method was validated according to Food and Drug Administration guidance (Food and Drug Administration, 2018). Each validation batch consisted of one set of calibration curve and six replicates of each level of QC samples.

2.5.1. Selectivity and carryover
The developed assay was able to differentiate PF-04620110 and IS from endogenous substances in the plasma. The selectivity was evaluated using six individual sources of blank plasma. The samples, i. e., blank dog plasma samples, blank plasma samples spiked with PF-04620110 at lower limit of quantification (LLOQ) and the plasma samples from the dogs after oral administration of PF-04620110 were processed and analyzed as described above. There should be no obvious interfering peaks at the retention times of PF-04620110 and IS. The interference should not exceed 20% of the LLOQ for PF-04620110 and 5% for IS. Carry-over was evaluated by injecting blank plasma samples after highest calibration standard. The carry-over did not exceed 20% of the LLOQ for PF-04620110 and 5% of the IS.

2.5.2. Linearity
The calibration curve was generated by plotting the peak area ratios of analyte to IS against the nominal concentrations of PF-04620110 in dog plasma samples using a weighed (1/x2) linear regression. The linearity of the calibration curve was assessed by correlation coefficient(r) which should be more than 0.995. The back-calculated concentration of calibrators should not exceed ± 15% of the nominal concentrations.

2.5.3. Sensitivity
The sensitivity of the method was expressed as LLOQ, which was defined as the lowest concentration of PF-04620110 in the sample which can be quantified reliably. At LLOQ, the ratio of signal-to-noise was more than 10 (S/N > 10), with acceptable accuracy (within ± 15%) and precision (< 15%).

2.5.4. Precision and accuracy
The precision and accuracy were assessed in inter- and intra-day which were determined at three concentration levels (3, 160 and 1600 ng/mL) on three successive days. The precision and accuracy were expressed as relative standard deviation (RSD) and relative error (RE), respectively. The RSD did not exceed 15%, while RE was within ±15%.

2.5.5. Extraction recovery and matrix effects
Extraction recovery and matrix effects were assessed at six replicates at each concentration level (3, 160 and 1600 ng/mL). The extraction recovery was determined by comparing the peak area of the pre-extracted samples with those of processed blank plasma spiked standard solutions at corresponding concentrations (post-extraction). For matrix effects assessment, six lots of blank dog plasma from individual donors were involved. The matrix effects were evaluated by comparing the peak area of the analyte that spiked into the post-extraction matrix with those of the standards prepared in initial mobile phase at the same concentration. If the ratio < 85% or >115%, matrix effects were implied.

2.5.6. Stability
The stability of the analyte in dog plasma was evaluated at six replicates at each concentration level (3, 160 and 1600 ng/mL). The short-term stability was tested by evaluating the QC samples after storing the QC samples at room temperature for 12 h. The long-term stability was investigated by analyzing the QC samples after storing samples at-80°C for 60 days. The freeze-thaw stability was evaluated by analyzing the QC samples after three freeze (-80 oC)-thaw (25 oC) cycles. The post-preparative stability was evaluated by analyzing the QC samples after storing processed samples at auto-sampler (10 oC) for 6 h.

2.6. Pharmacokinetic application

2.6.1. Animals
Twelve male dogs with body weight of 10-13 kg were provided by Animal Experiment Center of Guangzhou Women and Children’s Medical Center (Guangzhou, China). The dogs were kept in an environmentally controlled breeding room (temperature 25± 2 oC, humidity 55-65%, dark/light 12h/12h) for a week with free access to food and water. Beforeexperiments, all animals were fasted overnight but free access to water. The animal experiment was approved by the Animal Care and Usage of Guangzhou Women and Children’s Medical Center (Guangzhou, China).

2.6.2. Dosing and sample collection
PF-04620110 was formulated in 0.5% CMC-Na solution for dosing. The dogs were divided into two groups. One group was orally administered with PF-04620110 at a single dose of 1 mg/kg, while other group was intravenously administered with PF-04620110 through tail vein at a single dose of 0.5 mg/kg. Approximately 120 μL of blood samples were collected into heparinized tubes at pre-dose, 0.25, 0.5, 1, 2, 4, 6, 8, 12, 24 and 48 h post-dose. The collected blood samples were immediately centrifuged at 4000 rpm for 5 min and the resulting plasma samples were harvested and then stored at -80 oC until analysis.

2.6.3. Data analysis
The pharmacokinetic parameters including area under the curve (AUC), maximum plasma concentration (Cmax), time to reach Cmax (Tmax), mean residence time (MRT), clearance (CL), volume of distribution (Vd), and half-life (T1/2) were calculated by non-compartmental analysis using DAS 2.0 software. The oral bioavailability (F, %) was calculated in terms of
the following equation: F% = AUC 0−∞, ig × Dose iv × 100%.

3. Results and discussions

AUC 0−∞, ivDose ig

3.1. Method development
Although triple quadrupole mass spectrometers are frequently used for quantitative analysis of drugs in biological samples, high resolution mass spectrometers such as Q-Exactive-Orbitrap mass spectrometer are demonstrated to be powerful tools for the quantification because of its high resolution, selectivity and sensitivity. Most importantly, it can provide some structural information of the target analyte for confirmation. The standard solutions of PF-04620110 and IS were individually infused into Q-Exactive-Orbitrap mass spectrometer operated both in positive and negative ion mode. Compared with that in negative ion mode, PF-04620110 and IS showed much higher (~10 folds) mass responses in positive ion mode. PF-04620110 and IS showed accurate protonated molecular ions [M+H]+at m/z 394.1872 (Calculated m/z 397.1870) and 348.1585 (Calculated m/z 348.1594), respectively. Therefore, the exact mass of each protonated species (m/z 394.1872 for PF-04620110 and 348.1585 for IS) was extracted for quantification, with a 5 ppm extraction window. The MS2 spectra of PF-04620110 and IS along with the proposed fragmentations were shown in Figure 1. The daughter ions of PF-04620110 were observed at m/z 379.1753, 260.1649, 213.0655, 156.0411, 138.0299 and 120.0810. The daughter ions of IS were m/z 330.1488, 288.1379 and 183.0811. These daughter ions offered some structure information of the analyte and were used for the confirmation of the target analyte.

To improve the peak shape of the analyte, different columns including Waters ACQUITY BEH C18 column (1.7 μm, 2.1 × 50 mm), Agilent Poroshell 120 SB-C8 (2.7 μm,
2.1 × 50 mm), Waters XBridge BEH300 column (3.5 μm, 2.1 × 50 mm), Welch Ultimate UHPLC XB-C18 column (1.8 μm, 2.1 × 100 mm) and Waters ACQUITY CSH-C18 column (1.7 μm, 2.1 × 100 mm) in combination with various mobile phase compositions such as 1mM ammonium acetate-acetonitrile, 0.1% acetic acid-acetonitrile and 1mM ammonium acetate-0.1% acetic acid in acetonitrile were tested. Finally, Waters ACQUITY BEH C18 column (1.7 μm, 2.1 × 50 mm) combined with 1mM ammonium acetate-0.1% acetic acid in acetonitrile was found to yield better peak shape, lower background, higher mass response and shorter analysis time.

Compared with the reported method (Lee et al., 2012), the assay described here was more sensitive, with LLOQ being 1 ng/mL. The method could provide the exact mass and some structure information of the analyte. Therefore it would be more selective than triple quadrupole mass spectrometer and this method would be further suitable for metabolite identification.

3.2. Method Validation

3.2.1. Selectivity and carryover
The representative extracted ion chromatograms of the blank dog plasma, blank dog plasma supplemented with PF-04620110 at LLOQ and IS, and plasma samples collected at 2 h post dose, were displayed in Figure 2. The results indicated that PF-04620110 and IS were free of interference. No carryover was observed in the current study.

3.2.2. Linearity
The assay showed excellent linearity over the concentration range of 1-2000 ng/mL, with correlation coefficient > 0.998 (r > 0.998). The regression equation was y = (0.00135 ± 0.0001) x + (0.0027 ± 0.0002), where y means the peak area ratios of analyte to IS and x means nominal concentration of the PF-04620110 in dog plasma. The back-calculated concentrations of calibrators were in the range of 89.25-106.54% of the nominal concentrations.

3.2.3. Sensitivity
The LLOQ of this assay was 1 ng/mL, at which the ratio of signal-to-noise was 12.3. The accuracy (RE, %) at LLOQ was in the range of -12.57-6.84%, while precision (RSD, %) was 8.71%, suggesting that at LLOQ, PF-04620110 could be accurately and precisely quantified.

3.2.4. Accuracy and precision
The results of inter- and intra-day accuracy and precision were summarized in Table 1. Accuracy was in the range of -8.59-11.24%, and precision (RSD, %) did not exceed 9.69%. The results showed that the assay was satisfactory assurance for the determination of PF-04620110 in dog plasma.

3.2.5. Extraction recovery and matrix effects
The extraction recovery of PF-04620110 ranged from 85.37% to 91.31% with RSD% below 15% (as shown in Table 2). The matrix effects of PF-04620110 ranged from 94.39% to 105.09% and the RSD% did not exceed 8.87%. The extraction recovery and matrix effects of IS were 87.56% and 93.92%, respectively. These data suggested that the developed assay had satisfactory extraction efficiency and no endogenous substances impacted the ionization of the analytes.

3.2.6. Stability
The stability results demonstrated that PF-04620110 was stable under the tested storage conditions. The RE% ranged from -7.67% to 9.06%, with RSD% less than 8.79% (Table 3).

3.3. Pharmacokinetics study
The established UHPLC-Q-Exactive-Orbitrap-MS method was further successfully applied tothe pharmacokinetic study of PF-04620110. Figure 3 depicted the plasma concentration versus time profiles of PF-04620110 in dog plasma after single intravenous and oral administration of PF-04620110. The pharmacokinetic parameters were summarized in Table

4. After intravenous administration, PF-04620110 was slowly eliminated from the plasma with clearance (CL) of 60.81 ± 7.11 mL/h/kg, which was far below the blood flow (~1854 mL/h/kg). Its half-life (T1/2) was 6.72 ± 0.95 h. The AUC0-t was 8.30 ± 0.94 μg·h/mL. After oral administration, PF-04620110 showed fast absorption and reached the maximum concentration (1.25 ± 0.15 μg/mL) at 3.50 ± 1.00 h post-dose. It was also slowly eliminated from the plasma with CL of 81.44 ± 25.79 mL/h/kg and T1/2 of 6.53 ± 1.38 h. The oral bioavailability was determined to be 77.89% in dogs, suggesting that PF-04620110 would have an excellent gastrointestinal absorption and low first-pass metabolism.

4. Conclusions
A quantitative method based on UHPLC-Q-Exactive-Orbitrap-MS in full scan mode was developed and validated for the determination of PF-04620110 in dog plasma. The method was demonstrated to be selective, linear, precise and accurate. The LLOQ of the method was 1 ng/mL. The validated method has been successfully applied to the pharmacokinetic study of PF-04620110 in dogs and the results revealed that PF-04620110 showed slow elimination and high in vivo exposure. Its oral bioavailability was calculated to be 77.89%. As far as we know, this was the first report on the pharmacokinetics of PF-04620110 in dog plasma.

Acknowledgements
This work was financially supported by National Natural Science Foundation of China (Project No. 81302870) and Guangdong Science and Technology Planning Project (Project No. 2013B021800043).

Conflict of interests
The authors declared no conflict of interests.

References
Cao J, Zhou Y, Peng H, Huang X, Stahler S, Suri V, Qadri A, Gareski T, Jones J and Hahm S. Targeting Acyl-CoA: diacylglycerol acyltransferase 1 (DGAT1) with PF-04620110 small moleculeinhibitors for the treatment of metabolic diseases, Journal of Biological Chemistry, 2011, 286, 41838–41851Cases S, Stone SJ, Zhou P, Yen E, Tow B, Lardizabal KD, Voelker T and Farese RV. Cloning of DGAT2, a second mammalian diacylglycerol acyltransferase, and related family members, Journal of Biological Chemistry, 2001, 276, 38870–38876Dow RL, Li JC, Pence MP, Gibbs EM, LaPerle JL, Litchfield J, Piotrowski DW, Munchhof MJ, Manion TB and Zavadoski WJ. Discovery of PF-04620110, a potent, selective, and orally bioavailable inhibitor of DGAT-1, ACS Medicinal Chemistry Letters, 2011, 2, 407–412Dow RL, Andrews MP, Li JC, Gibbs EM, Guzman-Perez A, Laperle JL, Li QF, Mather D, Munchhof MJ, Niosi M, Patel L, Perreault C, Tapley S, Zavadoski WJ, Defining the key pharmacophore elements of PF-04620110: Discovery of a potent, orally-active, neutral DGAT-1 inhibitor, Bioorganic & Medicinal Chemistry, 2013, 21, 5081-5097

Enayetallah AE, Ziemek D, Leininger MT, Randhawa R, Yang J, Manion TB, Mather DE, Zavadoski WJ, Kuhn M and Treadway JL. Modeling the mechanism of action of a DGAT1 inhibitor using a causal reasoning platform, PLoS One, 2011, 6, e27009.
Food and Drug Administration, Guidance for industry: bioanalytical method validation, 2018. Available at
https://www.fda.gov/downloads/drugs/guidancecomplianceregulatoryinformation/guida nces/ucm368107.pdf.
He CY, Wan H, Drug metabolism and metabolite safety assessment in drug discovery and development, Expert Opinion on Drug Metabolism and Toxicology, 2018, 14, 1071-1085 King AJ, Segreti JA, Larson KJ, Souers AJ, Kym PR, Reilly RM, Zhao G, Mittelstadt SW and Cox BF. Diacylglycerol acyltransferase 1 inhibition lowers serum triglycerides in
the Zucker fatty rat and the hyperlipidemic hamster, Journal of Pharmacology and Experimental Therapeutics, 2009, 330, 526–531
Lee KR, Choi SH, Song JS, Seo H, Chae YJ, Cho HE, Ahn JH, Ahn SH, Bae MA, Determination of PF-04620110, a novel inhibitor of diacylglycerol acyltransferase-1, in rat plasma using liquid chromatography–tandem mass spectrometry and its application in pharmacokinetic studies, Biomedical Chromatography, 2013, 27, 846-852

Martignoni M, Groothuis GM, de Kanter R. Species differences between mouse, rat, dog, monkey and human CYP-mediated drug metabolism, inhibition and induction, Expert Opinion on Drug Metabolism and Toxicology, 2006, 2, 875–894
Subauste A, Burant CF, DGAT: novel therapeutic target for obesity and type 2 diabetes mellitus, Current Drug Targets Immune Endocrine and Metabolic Disorders, 2003, 3, 263–270
Yengi LG, Leung L, Kao J, The evolving role of drug metabolism in discovery and development. Pharmaceutical Research, 2007, 24, 842–858
Zhang Y and Ren J. Role of cardiac steatosis and lipotoxicity in obesity cardiomyopathy,
Hypertension, 2011, 57, 148–150