Rapid quantification of a chemically synthesized peptide GAP162 in rat plasma by liquid chromatography/triple quadrupole tandem mass spectrometry and application to a pharmacokinetic study

Abstract

Liquid chromatography/tandem mass spectrometry (LC-MS/MS) is a promising analytical platform for the quantification of therapeutic peptide in biological fluids for pharmacokinetics (PK) studies. Herein, an absolute quantification method based on liquid chromatography-electrospray ionization-tandem mass spectrometry (LC-ESI-MS/MS) technique was developed to quantify GAP162, a new synthetic peptide derived from RasGAP_301–326, which is a promising candidate as antitumor drug. A synthetic peptide P119 was used as internal standard. Solid phase extraction (SPE) of the mixed-mode of ion exchange and reversed-phase was employed for sample preparation. Chromatographic separation was performed on a reversed phase C4 column (30 mm × 2.1 mm, 5 μm) with a mobile phase consisting of acetonitrile–water containing 0.1% formic acid with gradient elution at a flow rate of 0.8 mL min⁻¹ for 2.0 min. Multiple reaction-monitoring (MRM) mode was performed with ion pairs of m/z: 748.2 → 830.2, and 526.7 → 585.5 for GAP162 and internal standard of P119, respectively. Calibration curve was linear over a concentration range of 5–500 ng mL⁻¹ with a correlation coefficient >0.99. The lower limit of detection was at 5 ng mL⁻¹ in rat plasma for GAP162. The results of the intra- and inter-day precision and accuracy studies were well within the acceptable limits. The validated method was successfully applied to investigate the pharmacokinetics study of GAP162 after single intravenous administration to male Sprague-Dawley rats at 5 mg kg⁻¹.

---

1. Introduction

The Ras-GTPase activating protein Src homology 3 (SH3) domain binding protein (G3BP), which can recognize the SH3 domain of Ras-GTPase activating protein (RasGAP), is overexpressed in many human tumors and participates in several signaling pathways.¹ G3BPs interact with RasGAP only in growing cells and depend on Ras activation, and participate in Ras signal pathway.² Therefore, the blockage and down-regulation of G3BPs may be a novel strategy for cancer therapy.³ It has been discovered that the caspase-3-generated RasGAP N-terminal fragment (RasGAP_158–455), called N₂, was able to selectively sensitize cancer cells, but not healthy cells, to genotoxin-induced apoptosis.⁴ RasGAP amino acids 301 to 326 within fragment N₂ were found to carry this sensitizing activity.⁵ Two trans-activator protein (TAT)-conjugated RasGAP N₂ fragment-derived peptides, TAT-RasGAP_317–326 and RI.TAT-RasGAP_317–326, were reported to be effective for improving the efficacy of chemotherapy in tumor models.^6,7 Another two TAT-conjugated RasGAP_301–326 mimic peptides, 38GAP and GAP161, could increase the inhibition by cis-diamminedichloroplatinum (CDDP) in vivo through the termination of CDDP-induced G2/M arrest and inhibitions of phospho-Akt, phospho-ERK and NF-κB in colon carcinoma HCT116 cells.^8,9 GAP162 is a novel RasGAP_301–326 mimic peptide, which has the similar effective amino acids sequence (MFIVHNELRRGWMWAEGGRRR) with GAP161 in term of chemical structure. This peptide can markedly suppress HCT116 cell growth and HCT116 xenografts growth in mice with 78% TGI (tumor growth inhibition, treated by the combination of GAP162 and CDDP at 40 and 1 mg kg⁻¹) and 42% TGI (treated byGAP162 only at 40 mg kg⁻¹) in vivo (unpublished data). The anti-tumor effect of GAP162 is much better than GAP161 (32% TGI at 60 mg kg⁻¹).⁹ Therefore, GAP162 has being a novel drug candidate targeted to multiple tumors and under research in our laboratory.

To explore and understand the mechanism underlying its anti-tumor effects, the pharmacokinetic (PK) study should be conducted, along with in vitro and in vivo pharmacological evaluation, to disclose the systemic exposure to animal and PK property of GAP162. In general, the major challenge for PK study is to develop and then validate a rapid, sensitive, and robust analytical method to quantify the concentration of compound of interest in various matrices, such as plasma, urine, bile, and tissue homogenate. Enzyme-linked immuno sorbent assay (ELISA) is the most commonly employed approach for the quantification of biotherapeutics in biological fluids. The limitation of this methodology is that the quantitative accuracy and specificity are often compromised by the interferences from endogenous proteins and protein fragments. In addition, the development of specific antibodies is always time- and labor-consuming, and various species-specific antibodies have to be developed when pharmacokinetics of test compound need to be tested in multiple animal species.

Liquid chromatography-mass spectrometry (LC-MS/MS) has become a mature methodology used in the pharmacokinetic evaluation of chemical small molecule. The application of this technique has also extended to the quantification of protein/polypeptide in various biological matrices and served as a promising alternative to ELISA due to its intrinsic high specificity and sensitivity, and multiplexing capability. More importantly, the LC-MS/MS based method is often more robust in terms of species specificity. Two strategies have been widely adopted when LC-MS/MS technique is used for the quantification of protein/peptides in biological matrix:^10–12 (1) direct detection by MS: intact small proteins/peptides with molecule weight <10 kDa and good ionization are treated as small molecule and detected directly by targeted MS without digestion by trypsin or other enzymes and efforts should be focused on biological sample preparation with high recovery; (2) indirect detection by MS: before injection into LC-MS/MS, proteins/peptides are hydrolyzed into peptides by enzyme digestion and MS-favorable signature peptide(s) will be selected and monitored in MRM mode by LC-MS/MS for quantification. In this way, LC-MS/MS can usually achieve better (or comparable) performance to immunoassays.^10–15 However, in many cases, the pickup of surrogate peptide(s) with both sequence uniqueness and MS-favorable characteristics remains quite difficult and challenging. LC-MS/MS technique is playing a more and more important role in the research and development of biotherapeutics, especially in discovery phase.

In this study, a sensitive and robust LC-MS/MS method for the quantification of GAP162 in rat plasma was established for the first time. In this method, a semi-automated 96-well ion-exchange solid phase extraction (SPE) was used for sample cleanup, a synthetic peptide of P119, an analog to GAP162, was used as internal standard (IS) to monitor both the sample preparation and MS detection process, and the short chromatographic running time by LC-MS/MS (2.0 min) allowed the high throughput of sample analysis and expedite the analytical section of pharmacokinetic study. After a fit for purpose validation, the method was applied to the PK study of GAP162 after single intravenous (IV) administration to Sprague-Dawley rats at 5 mg kg⁻¹ and the PK properties were obtained accordingly.

2. Experimental

2.1 Chemicals and reagents

GAP162 was synthesized at Wuhan KatyGen Pharmaceuticals Inc. (Hubei, China) using guanidine transformation method technology, purified by HPLC, and determined by mass spectrometry. The molecular weight (MW) of GAP162 was 5977.96 Da. Internal standard (P119, MW: 3680.38 Da) was also provided by Wuhan KatyGen Pharmaceuticals Inc.

SPEμElution 96-well plates (MAX, MCX, WAX, and WCX) were purchased from Waters (Milford, MA, USA). All other reagents of the highest available purity were HPLC-grade. Distilled water was produced from Milli-pore water purified system.

2.2 Preparation of standard and quality control solutions

The reference standard of GAP162 was accurately weighed and dissolved in methanol (MeOH) : water (H₂O) (50 : 50, v/v) at 1 mg mL⁻¹, and then diluted to appropriate concentrations using acetonitrile : MeOH : H₂O (5 : 4 : 1, v/v/v, containing 2% formic acid) for the construction of calibration curves in rat plasma. The concentration of stock solution of GAP162 was 1000 μg mL⁻¹. Working solutions of GAP162 were prepared at serial concentrations of 50.0, 100, 200, 500, 1000, 2000, 4000, 5000 ng mL⁻¹. IS (P119) working solution (5 μg mL⁻¹) was prepared by diluting with MeOH : H₂O (50 : 50, v/v). All the stock and working standard solutions were stored at 4 °C prior to further use.

2.3 Sample preparation

Oasis MAX μElution 96-well plate was preconditioned with 200 μL of MeOH followed by 200 μL of H₂O. The thawed plasma samples (100 μL) spiked with 10 μL of IS working solution (5 μg mL⁻¹) were diluted with 4% H₃PO₄ (100 μL) before loading. After sample loaded, the plate was washed with 200 μL of 5% NH₄OH, followed by 200 μL of methanol, and eluted with 100 μL of acetonitrile : MeOH : H₂O (5 : 4 : 1, v/v/v, containing 2% formic acid). The elution was diluted with 100 μL of deionized water (1 : 1, v/v). The diluted elution (100 μL) was transferred to 96-well plate for LC-MS/MS analysis. The injection volume was 5 μL.

2.4 Instrumentation and analytical conditions

The liquid chromatography/quadrupole-linear ion trap mass spectrometry (LC-ESI-(QqLIT)MS/MS) system consisted of a Shimadzu LC (SLC-30A) system (Shimadzu Corporation, Kyoto, Japan) coupled to Q-Trap 4500 (ABSCIEX, Concord, ON, CA) equipped with TurboIonSpray source. The LC-MS/MS system was controlled by Analyst 1.5.2 software, which also acquired and processed the data.

LC separation was performed on a 5 μm ACE C4 column (30 mm × 2.1 mm, inside diameter) at room temperature. The LC linear gradient was increased from 5% B to 90% B in 2 min (A), water with 0.1% formic acid; (B), acetonitrile with 0.1% formic acid), at a flow rate of 0.80 mL min⁻¹. The injection volume was 5 μL and the temperature of autosampler was set at 4 °C to keep the samples cool during the analysis.

The operating parameters of the ion source were optimized using 10 μg mL⁻¹ tuning solution of GAP162 and IS to obtain the best performance from the mass spectrometer for the analysis of GAP162. Before introduced into the MS detector, the tuning solution, delivered at 5 μL min⁻¹ by a syringe pump, was combined through a peek tee-connector with mobile phase at 0.8 mL min⁻¹ of mobile phase by HPLC pump. Typical sets of parameters used in this study were as follows: the source parameters, including curtain gas, gas 1, gas 2, collision gas (CAD), capillary temperature (TEM), and ion spray voltage, were set at 35 psi, 50 psi, 50 psi, medium, 550 °C, and 5.5 kV for positive mode; the compound parameters, including declustering potential (DP), collision energy (CE), and collision cell exit potential (CXP), were 136 V, 31 V, and 10 V for GAP162, and 111 V, 21 V, and 12 V for IS. The MS detector was operated in MRM) mode at unit mass resolution for quantification of GAP162 in rat plasma with the dwell time at 100 ms. The precursor-to-product ion transition for GAP162 was m/z: 748.2 → 830.2 and transition of m/z: 526.7 → 585.5 for IS.

2.5 Calibration curves

Calibration curves were prepared by spiking 10 μL aliquots of appropriate working solutions into 90 μL of blank plasma (heparin sodium as anticoagulant) to produce the standard samples at the final concentration of 5, 10, 20, 50, 100, 200, 400, and 500 ng mL⁻¹. Each sample also contained 500 ng mL⁻¹ of IS. Zero plasma samples used in each run were prepared containing 500 ng mL⁻¹ of IS only. The plasma samples for calibration were prepared under the same conditions as the test samples. In each run, a plasma blank, a zero standard (with IS only), and a set of calibration standards were analyzed where the lower limit of quantification (LLOQ) and upper limit of quantification (ULOQ) were used in duplicate.

2.6 Method validation

According to the international guidelines,^19,20 the method was validated in terms of the selectivity, linearity, lower limit of quantification (LLOQ), accuracy, precision, recovery and stability. The selectivity was evaluated by analyzing six individual samples of blank rat plasma by comparison with the plasma-spiked analyte for endogenous interferences. Each blank sample was tested by LC-MS/MS method in order to investigate potential interferences. Heparin sodium was used as anticoagulant for blank rat plasma in the validation.

The linearity of the method was determined by analyzing a series of standard plasma samples at concentrations of 5.0, 10, 20, 50, 100, 200, 400, and 500 ng mL⁻¹ for GAP162 by least squares linear regression of the peak area ratios of GAP162 to IS obtained against the corresponding concentration (X) with a weighting factor of 1/X². Since the analytical method was developed to support the preclinical PK study in discovery phase, the LLOQ was defined as the lowest concentration on the calibration curve with acceptable precision and accuracy (<25%). The criteria for the calibration included a correlation coefficient (r) of 0.99 or better and for non-LLOQ levels, the precision should be <20% and accuracy within 80–120%. The recovery of GAP162 was determined by comparing the peak area of the extracted plasma samples at three QC concentrations (10, 200 and 400 ng mL⁻¹) in 3 replicates to the peak area of neat standard prepared in acetonitrile : MeOH : H₂O (5 : 4 : 1, v/v/v, containing 2% formic acid). Recoveries were determined at 10, 200, and 400 ng mL⁻¹.

To determine the accuracy and precision, the method developed here was validated by analyzing quality control (QC) samples 3 times, which were prepared by spiking known amounts of GAP162 in blank plasma at low (10 ng mL⁻¹, LQC), medium (200 ng mL⁻¹, MQC), and high concentration (400 ng mL⁻¹, HQC). The measured value was calculated from the calibration curve obtained in the same run, expressed as the mean of the 3 values. The measured value DQC was multiplied by dilute factor. The accuracy was measured as the difference between the nominal value and measured value expressed as a percentage of the nominal value. The precision was expressed as the coefficient of variation (CV), i.e., the standard deviation divided by the mean value multiplied by 100.

The stability of GAP162 was evaluated by comparing the peak areas of stability samples in triplicates (10, 200, and 400 ng mL⁻¹) under conditions likely to be encountered during the sample storage, preparation, and the analytical process to those of freshly prepared QC samples at the same 3 concentrations including stock solution storage at 4 °C for 3 months, storage at 25 °C in rat plasma for 2 h, storage in autosampler for 8 h after extraction, and short-term storage at −80 °C with three cycles of freeze/thaw (−80 ↔ 25 °C) for 7 days. Samples were considered to be stable if assay values were within the acceptable limits of accuracy (80–120% deviation (DEV)) and precision (±20% CV).

2.7 Application to a pharmacokinetic study

Male Sprague-Dawley rats with body weight of ∼220 g were purchased from Beijing Vital River Laboratories Co., Ltd (China). All experimental procedures were approved by the Institutional Experimental Animal Care and Use Committee of BioDuro (Shanghai, China). The rats were maintained in an air-conditioned animal quarter with alternating 12 h light/dark cycles at a room temperature of 22 ± 2 °C and a relative humidity of 50 ± 10%. The rodents were given a commercial rat chow and water ad libitum.

Three days before the experiment, polyethylene cannulas were implanted into the jugular vein of rats after anesthetized with ketamine combined with xylazine. To prevent blood clotting, the cannulas were externalized at the neck back and filled with heparinized saline/glycerol (20 units per mL). Three rats were used on study for single IV administration of GAP162 at 5 mg kg⁻¹. The blood samples (∼200 μL) were taken via cannula implanted in jugular vein from non-restrained non-sedated animals into heparin sodium coated tubes pre-dose and subsequently at 0.083, 0.25, 0.5, 1, 1.5, 2, 3, 4, 6, 8, 10 and 24 h following IV administration. The blood samples were centrifuged at 4000g for 10 min at 4 °C to give 100 μL plasma samples. These samples were stored at −80 °C immediately until LC-MS/MS analysis.

The plasma concentration-time data for GAP162 in rats were analyzed by the WinNonlin 6.3 software from Pharsight (Mountain View, CA, US) employing non-compartmental model. The parameters reflecting the exposure, distribution, and elimination of GAP162 in plasma included C₀, AUC_0–t, CL, V_d, MRT, and T_half following single IV administration. The C₀ is the peak plasma concentration calculated; the AUC_0–t is the area under the concentration-time curve from 0 to t since the drug injection, which was calculated using a linear trapezoidal method. T_half is the half-life calculated according to 0.693/λ_z, where λ_z is the terminal elimination rate obtained using concentration data during 0 to 6 h. CL is the total clearance. V_d is the apparent distribution volume. MRT is the mean residence time calculated as AUMC/AUC_0–t, where AUMC is the integration of C_0–t versus time from 0 to t. All results were expressed as arithmetic mean ± standard deviation (SD).

3. Results and discussion

3.1 Sample preparation

Peptides and proteins often have greater nonspecific binding to vessel walls (e.g., sample vials and 96-well plates) than small molecules, which can compromise the linearity and accuracy of the assay.¹⁶ The interactions of peptide with solvents and vessel surfaces depend largely on the specific side chains of the amino acids of peptide.¹⁷ Positively charged peptides can easily generate electrostatic interaction with for example glass surfaces, carrying a negative potential.¹⁴ As expected, GAP162 showed severe binding to glass tube due to its arginine/lysine rich region. When polypropylene vials were used instead, no nonspecific binding was observed and the loss of GAP162 by this binding was avoided successfully. The adsorption of GAP162 was evaluated by transferring neat peptide solutions at 50 and 200 ng mL⁻¹ sequentially from a polypropylene vial to a glass vial, and then back to polypropylene vial and analyzing a small portion after each transfer step to assess losses. The result showed that the losses from polypropylene vial to glass were 62.8% and 36.9% at 50 and 200 ng mL⁻¹, respectively. There is no obvious loss from the glass vial to polypropylene vial. Thus polypropylene material vessels were used throughout this study. Other than nonspecific binding, the arginine/lysine rich region poises unique challenges for sample cleanup. When the protein precipitation with various organic solvent (e.g., MeOH and acetonitrile) was tried, GAP162 could not be efficiently separated from endogenous proteins/interferences in rat plasma and co-precipitate out resulting in the unacceptable low recovery (data not shown). Then different solid phase extraction columns were tested trying to get more efficient cleanup and higher recovery. When rat plasma sample spiked with test compound was loaded on general reversed-phase SPE (Oasis HLB μElute), most of GAP162 was recovered in loading solution, indicating the poor retention of the peptide on the column and inevitable low recovery in the following elution step. Several ion exchange SPE sorbents including MAX, MCX, WAX, and WCX were evaluated using the same strategy and the mixed-mode ion exchange SPE (MAXμElute) was found finally to have the highest recovery for GAP162 as well as the cleanest eluent for LC-MS/MS injection among those columns with different sorbents. Thus, Oasis MAX μElution 96-well plate was employed for rat plasma sample cleanup throughout this study and the SPE procedure was performed on an automated μElute96-well extraction system, which can increase the throughput for sample preparation significantly.

3.2 Mass spectrometry and chromatography

Peptides and small proteins can produce multiple charged ions via electrospray ionization and thereby the mass-to-charge ratio can be lowered to the range of m/z from 400 to 2000, which is readily accessed by quadrupole, ion trap, orbitrap, and ion cyclotron resonance (ICR) mass analyzers.¹⁸ The multiple charged ions can also improves signal-to-noise ratio, mass resolution, and mass accuracy.¹⁹ However, intact protein analysis by MS usually encounters the challenges of low charging efficiency and overbroad charge-state distribution, which can compromise the assay sensitivity and dynamic range in quantification analysis. The charge state distribution is affected by the factors including analyte gas-phase basicity,^20,21 solution pH, solvent composition droplet size,²² instrument parameters, and protein conformation.²³ It has been proved that supercharging reagent can increase ESI multiple charging of proteins, polypeptides, and protein complex.²⁴ The previous study has shown that a charge distribution frame of GAP161 was displaced to lower m/z region and the ionization efficiency was markedly increased by a variety of supercharging reagents including dimethylacetamide (DMA), acetone, m-nitrobenzyl alcohol (m-NBA), ethyl acetate (EtOAc), tetrahydrofuran (THF), N-methylpyrrolidone (NMP), DMSO, benzyl alcohol (BnOH), and thiodiglycol (TDG) at different concentrations and DMSO achieved the most pronounced enhancement.²⁵ To achieve higher sensitivity, the effect of supercharging reagent to GAP162 was also investigated and 0.5% formic acid (FA), which is compatible to be LC-MS/MS mobile phase additive, was found to have significant impact on the alteration of charge distribution and response intensity of GAP162. GAP162 exhibits a fairly wide charge distribution using conventional mobile phase. The primary charge states from [M + 6H]⁶⁺ to [M + 10H]¹⁰⁺ with the most intense peak at [M + 7H]⁷⁺. When supercharging reagent added, 0.5% formic acid could shift the base peak ion of GAP162, served as the candidate precursor for MRM monitoring, from [M + 7H]⁷⁺ (m/z: 854.6) towards higher charge state [M + 8H]⁸⁺ (m/z: 748.2), and the intensity of the precursor ion intensity was increased at least by four folds (as shown in Fig. 1), which is quite favorable for sensitivity enhancement of the LC-MS/MS quantification method. Based on this observation, [M + 8H]⁸⁺ at m/z: 748.2 was finally utilized as the parent ion for MRM monitoring.

Fig. 1 Mass spectra of GAP162 from Q1 scan. Representative Q1 full-scan spectra of GAP162 from conventional solution (top) and solution containing 0.5% FA (bottom).

Upon collision induced dissociation (CID), the predominant fragment ion was observed at m/z of 830.2 and was thereby selected as the product ion of MRM transition. The MRM transition for IS used for MS detection is of m/z: 526.7 → 585.5. The full-scan product ion mass spectra of GAP162 and IS are shown in Fig. 2. Worth noting that the higher m/z of product ion than parent ion in MRM transition can usually cause pretty low level of noise, which can be used to enhance the sensitivity of analytical method in other hand. Other conditions such as ion spray voltage, curtain gas pressure, nebulizer gas pressure, heater gas pressure, source temperature and collision energy were further optimized to improve the response intensity and stability of GAP162.

Fig. 2 Chemical structures and MS/MS spectra of GAP162 and P119 (IS). Top: the structure and MS/MS spectra of GAP162; bottom: the structure and MS/MS spectra of P119 (IS).

To resolve GAP162 from the endogenous interference in rat plasma, various types of chromatographic column have been tested and reverse phase (RP) columns were found to be able to retain GAP162 well and obtain good resolution. Among the RP columns, large-poreC4 column outperformed conventional C18/C8 RP columns not only because of its superior separation efficiency and peak symmetry, but also due to its low carryover (<1%), which can ensure the measured accuracy of rat plasma sample in low concentration. With the optimized gradient program for elution, the chromatographic retention time of GAP162 and IS, P119, were 1.01 and 0.99 min with good peak shape and symmetry and the total running time for one sample was 2.0 min, which allows the method high-throughput and suitable for routine plasma analysis in pharmacokinetic study.

3.3 Validation of analytical method

The established method was validated for the purpose of application in a pre-clinical pharmacokinetic study in discovery stage. The calibration curves ranging from 5 to 500 ng mL⁻¹ were linear for the quantification of GAP162 in rat plasma. The slopes, intercepts, and correlation coefficients of the regression equations were determined by least squares linear regression using a weight-factor of 1/X². Typical equations for the standard curves were Y = 0.00491X + 0.161 (r: 0.993). Deviations were within ±20% (except for LLOQ: within ±25%) for all regression equations. The lower limit of quantification (LLOQ) was 5 ng mL⁻¹ for GAP162 (peak area was 7.04 × 10²) in rat plasma with signal to noise ratio >5.

The precision of this analytical method was investigated by calculating the relative standard deviation (RSD) of the concentrations on the same day (n = 3) and on three consecutive days (n = 9) for QC samples, which were blank plasma samples spiked with different amounts GAP162 (10, 200, and 400 ng mL⁻¹). As shown in Table 1, the precision was between 2.0–19.3% for intra-day and 3.1–12.8% for inter-day at three concentration levels of 10, 200, and 400 ng mL⁻¹, indicating good assay precision. Meanwhile, the intra-run and inter-run accuracy ranged from 82.9 to 104%.

Table 1 Intra- and inter-day accuracy and precision data (mean ± SD) for GAP162 in rat plasma determined by LC-MS/MS (n = 3)

Batch	Item	Low (10 ng mL^−1)	Medium (200 ng mL^−1)	High (400 ng mL^−1)
Day 1	Mean ± SD	8.74 ± 0.56	173 ± 3.5	399 ± 39
	RSD (%)	6.4	2.0	9.7
	Accuracy (%)	87.4	86.5	99.7
Day 2	Mean ± SD	9.74 ± 1.76	195 ± 38	393 ± 18
	RSD (%)	18.1	19.3	4.5
	Accuracy (%)	97.4	97.7	98.3
Day 3	Mean ± SD	8.29 ± 0.38	183 ± 10	417 ± 11
	RSD (%)	4.6	5.7	2.6
	Accuracy (%)	82.9	91.5	104
Inter-day	Mean ± SD	8.92 ± 1.14	184 ± 22	403 ± 25
	RSD (%)	12.8	11.9	6.09
	Accuracy (%)	89.2	91.9	101

---

The specificity of the method was investigated by analyzing rat plasma of different source. Typical chromatograms of blank sample, blank plasma sample spiked with GAP162 (200 ng mL⁻¹) and IS (500 ng mL⁻¹), and a rat plasma sample collected at 1 h after IV administration are showed in Fig. 3. The retention time for GAP162 and IS were 1.01 and 0.99 min under the described LC-MS/MS method and there were no peaks corresponding to endogenous substances observed at these retention times to interfere with both the analyte of GAP162 and IS of P119 detection, which proved the good specificity of method.

Fig. 3 Typical LC-MS/MS chromatograms of GAP162 and IS in various samples. (A) blank rat plasma; (B) blank rat plasma spiked with GAP162 (200 ng mL⁻¹) and IS (500 ng mL⁻¹); (C) an unknown rat plasma sample collected at 1 h after single IV administration of GAP162 (5 mg kg⁻¹).

The recovery of GAP162 processed by SPE procedure using MAX μElute plate was determined by comparing the peak area of plasma samples at three QC concentrations (10, 200 and 400 ng mL⁻¹) in 3 replicates to the peak area of neat standard at the same 3 concentrations. Recovery efficiencies of GAP162 at three concentrations were ranged from 42.9–56.0%, indicating that the recoveries of GAP162 from rat plasma were consistent and concentration-independent in the range of 10–400 ng mL⁻¹ (Table 2). The recovery of the structure analog IS, P119, was at 59.0% and close to GAP162, suggesting the IS was suitable to monitor both the sample preparation process and MS detection.

Table 2 Recovery of GAP162 and IS by SPE procedure (n = 3)

Nominal conc. (ng mL^−1)	Peak area (mean ± SD (CV%))		Recovery^a
	QC sample	Neat solution
a Recovery efficiency (EE) was calculated as: recovery = (mean peak area)QC Sample/(mean peak area)Neat Solution × 100%.
10	2.07 × 10^2 ± 5.86 × 10^0 (3)	3.69 × 10^2 ± 8.00 × 10^0 (2)	56.0
200	2.38 × 10^3 ± 3.88 × 10^2 (16)	5.41 × 10^3 ± 2.25 × 10^2 (4)	42.9
400	4.30 × 10^3 ± 4.05 × 10^2 (9)	8.96 × 10^3 ± 7.85 × 10^2 (9)	48.0
IS (500 ng mL^−1)	3.08 × 10^3 ± 2.89 × 10^2 (9)	5.22 × 10^3 ± 5.34 × 10^2 (10)	59.0

---

Usually due to the proteases and other enzymes present in biological matrix, protease inhibitors including phenylmethanesulfonyl fluoride (PMSF), dithiothreitol (DTT), and glutathione were added to stabilize peptides/proteins²⁶ during quantitative analysis, which has the risk to introduce the interference to MS detection. The other way is to store bio-samples below −60 °C and handle analytical procedure on ice to decrease the residual protease activity and slow down the rate of any peptides/protein degradation.¹⁴ Thus, we decided to test the storage and bench-top stability in the mimicking conditions that could be encountered in sample analysis without proteases inhibitors added to avoid the possible interference. The stability results were presented in Table 3. The stock solution of GAP162 in 50% MeOH at 1 mg mL⁻¹ was stable for up to 3 months when stored at 4 °C. GAP162 was stable for at least 2 h in rat plasma at room temperature (25 °C), at least8 h in autosampler at 4 °C, and at least 7 days in rat plasma at −80 °C. GAP162 in rat plasma could stand up for 3 cycles of freeze–thaw process. In our previous study, GAP161 was reported to be unstable when stored at room temperature and as a result, GAP162 was designed with one of the purposes to improve the stability property in addition to improving anti-tumor effect and lowering the toxicity. Our data shows that GAP162 in rat plasma does have better stability at room temperature and would not be degraded or chemically converted in analytical process. If the sample preparations including blood collection, plasma separation, plasma thaw, and other handling are conducted on ice, no proteases inhibitors were added necessarily.

Table 3 Stability of GAP162 during sample storage and analysis (n = 3)

Conc. (ng mL^−1)	Condition	Peak area (mean ± SD (CV%))		Stability^a (%)
		Stability sample	Fresh QC sample
a Stability was calculated as: stability = (mean peak area)Stability Sample/(mean peak area)Fresh QC Sample × 100%.b Stability sample of stock and freshly prepared stock solution were diluted to 100 ng mL^−1 before injection into LC-MS/MS for stability evaluation.
10	2 h in rat plasma at 25 °C	2.72 × 10^2 ± 8.74 × 10^0 (3)	2.50 × 10^2 ± 2.60 × 10^1 (10)	109
	8 h in autosampler at 4 °C	2.80 × 10^2 ± 1.59 × 10^1 (6)		112
	7 day in rat plasma at −80 °C and 3 freeze–thaw cycles	2.73 × 10^2 ± 2.59 × 10^1 (9)		109
200	2 h in rat plasma at 25 °C	3.06 × 10^3 ± 1.33 × 10^2 (4)	2.81 × 10^3 ± 4.91 × 10^2 (17)	109
	8 h in autosampler at 4 °C	2.82 × 10^3 ± 1.23 × 10^2 (4)		100
	7 day in rat plasma at −80 °C and 3 freeze–thaw cycles	3.13 × 10^3 ± 7.37 × 10^1 (2)		111
400	2 h in rat plasma at 25 °C	5.45 × 10^3 ± 6.54 × 10^2 (12)	5.99 × 10^3 ± 4.31 × 10^2 (7)	91
	8 h in autosampler at 4 °C	5.35 × 10^3 ± 1.10 × 10^2 (2)		89
	7 day in rat plasma at −80 °C and 3 freeze–thaw cycles	5.30 × 10^3 ± 6.00 × 10^1 (1)		89
Stock^b	3 month at 4 °C	3.96 × 10^3 ± 4.88 × 10^2 (12)	4.12 × 10^3 ± 2.19 × 10^2 (5)	96

---

In conclusion, the present method was demonstrated to have satisfactory performance in agreement with international guideline for the separation and determination of GAP162 in rat plasma. The LLOQ of 5 ng mL⁻¹ by this method was much lower than radiopharmaceutical method with LLOQ of 66.7 ng mL⁻¹ developed in our lab (unpublished).

3.4 Application of the method and pharmacokinetics of GAP162

With this method developed and validated, the pharmacokinetic study of GAP162 after single IV administration to male Sprague-Dawley rats was conducted. To get the comprehensive pharmacokinetic profile of GAP162 and minimize the possible toxic effect to rat, the dosage was set at 5 mg kg⁻¹ without any adverse effect observed. The concentration versus time profile of GAP162 following single IV administration to Sprague-Dawley rats (n = 3) is shown in Fig. 4. The pharmacokinetics parameters have been calculated by WinNolin using the non-compartmental analysis and presented in Table 4. The GAP162 concentration was detectable in rat plasma samples for up to 6 h post-dose. The systemic exposure of GAP162 to rats (AUC_{0–6 h}) was 1283 ± 153 h ng mL⁻¹ after IV dosing at 5 mg kg⁻¹. The linear logarithm has been found in the elimination phase of the pharmacokinetics curve and the T_1/2 of this compound was short at 1.43 ± 0.34 h, which is comparable to GAP161 (∼1.84 h), but much longer than the original polypeptide RasGAP_301–326 (∼20 min). MRT of GAP162 was significantly increased from 0.26 ± 0.05 h for GAP161 to 0.65 ± 0.08 h. Due to the anti-cancer activity released from the in vitro and in vivo pharmacological study and the improved PK property (prolonged T_1/2 and increased exposure), GAP162 has been screened out as a valuable candidate for further investigation or structure modification, if needed. The LC-MS/MS based quantification method for GAP162 in rat plasma has the potential to be the basis for the future PK study in other species and expedite the research of this test peptide.

Fig. 4 Mean plasma concentration–time profiles of GAP162 after single IV administration of GAP162 at 5 mg kg⁻¹. Each point represents mean ± SD (n = 3).

Table 4 Main pharmacokinetic parameters of GAP162 in male SD rats after single IV administration of GAP162 (n = 3, mean ± SD)

Parameters	Unit	IV (5 mg kg^−1)
C0	ng mL^−1	16 886 ± 5363
Thalf	h	1.43 ± 0.34
AUC0–10 h	h ng mL^−1	1283 ± 153
AUCall	h ng mL^−1	1309 ± 160
CL	mL h^−1 kg^−1	3857 ± 445
Vz	mL kg^−1	7872 ± 1062
AUMC0–10 h	h h ng mL^−1	840 ± 210
MRT	h	0.65 ± 0.08

---

4. Conclusion

The LC-MS/MS-based method described in the present study enables determination of GAP162 in rat plasma with P119 as internal standard and SPE as the preparation method. The simple sample cleanup procedure and short chromatographic running time increased the throughput of sample analysis. The method was further applied to study the pharmacokinetics of GAP162 after a single IV administration in rats successfully, which suggests that the method is able and sufficient to be used in pre-clinical pharmacokinetic evaluation of GAP162 in future.

References

1. E. Guitard, F. Parker, R. Millon, J. Abecassis and B. Tocque, Cancer Lett., 2001, 162, 213–221.
2. A. Annibaldi, A. Dousse, S. Martin, J. Tazi and C. Widmann, PLoS One, 2011, 6, e29024.
3. W. Cui, Z. Wei, Q. Chen, Y. Cheng, L. Geng, J. Zhang, J. Chen, T. Hou and M. Ji, J. Chem. Inf. Model., 2010, 50, 380–387.
4. J. Y. Yang, J. Walicki, D. Michod, G. Dubuis and C. Widmann, Mol. Biol. Cell, 2005, 16, 3511–3520.
5. D. Michod, J. Y. Yang, J. Chen, C. Bonny and C. Widmann, Oncogene, 2004, 23, 8971–8978.
6. D. Michod and C. Widmann, Mol. Cancer Res., 2007, 5, 497–507.
7. D. Michod, A. Annibaldi, S. Schaefer, C. Dapples, B. Rochat and C. Widmann, J. Natl. Cancer Inst., 2009, 101, 828–832.
8. H. Zhang, S. Zhang, H. He, W. Zhao, K. Ren, J. Chen and R. G. Shao, Cancer Lett., 2011, 308, 62–70.
9. H. Zhang, S. Zhang, H. He, W. Zhao, J. Chen and R. G. Shao, Cancer Sci., 2012, 103, 1848–1856.
10. M. Dubois, F. Fenaille, G. Clement, M. Lechmann, J. C. Tabet, E. Ezan and F. Becher, Anal. Chem., 2008, 80, 1737–1745.
11. O. Heudi, S. Barteau, D. Zimmer, J. Schmidt, K. Bill, N. Lehmann, C. Bauer and O. Kretz, Anal. Chem., 2008, 80, 4200–4207.
12. C. Ji, N. Sadagopan, Y. Zhang and C. Lepsy, Anal. Chem., 2009, 81, 9321–9328.
13. A. Lesur, E. Varesio and G. Hopfgartner, Anal. Chem., 2010, 82, 5227–5237.
14. M. Ewles and L. Goodwin, Bioanalysis, 2011, 3, 1379–1397.
15. K. Xu, L. Liu, M. Maia, J. Li, J. Lowe, A. Song and S. Kaur, Bioanalysis, 2014, 6, 1781–1794.
16. D. L. Chappell, M. E. Lassman, T. McAvoy, M. Lin, D. S. Spellman and O. F. Laterza, Bioanalysis, 2014, 6, 1843–1857.
17. K. J. Bronsema, R. Bischoff and N. C. van de Merbel, Anal. Chem., 2013, 85, 9528–9535.
18. D. S. Wunschel, L. P. Tolic, B. Feng and R. D. Smith, J. Am. Soc. Mass Spectrom., 2000, 11, 333–337.
19. H. Hahne, F. Pachl, B. Ruprecht, S. K. Maier, S. Klaeger, D. Helm, G. Medard, M. Wilm, S. Lemeer and B. Kuster, Nat. Methods, 2013, 10, 989–991.
20. A. T. Iavarone, J. C. Jurchen and E. R. Williams, J. Am. Soc. Mass Spectrom., 2000, 11, 976–985.
21. E. R. Williams, J. Mass Spectrom., 1996, 31, 831–842.
22. K. Benkestock, G. Sundqvist, P. O. Edlund and J. Roeraade, J. Mass Spectrom., 2004, 39, 1059–1067.
23. M. Samalikova, I. Matecko, N. Muller and R. Grandori, Anal. Bioanal. Chem., 2004, 378, 1112–1123.
24. S. G. Valeja, J. D. Tipton, M. R. Emmett and A. G. Marshall, Anal. Chem., 2010, 82, 7515–7519.
25. X. Zhao, J. Sun, L. Sun, J. Chen, Y. Cheng and X. Duan, Bioanalysis, 2015, 8, 981–990.
26. R. Wolf, F. Rosche, T. Hoffmann and H. U. Demuth, J. Chromatogr. A, 2001, 926, 21–27.

---

Footnote

† L. Kang contributed equally to this study.

---