The determination of 10 compounds of GuiZhi Decoction in rat plasma after oral administration by HPLC-MS/MS and its application to a pharmacokinetic study

Background: Guizhi Decoction (GZD), a traditional Chinese medical formula, has been commonly used to treat fever, sweating, and cold in China. Methods: The high performance liquid chromatography-tandem mass spectrometry (HPLC-MS/MS) method was established for the determination of 10 compounds, including cinnamic acid, paeoniflorin, albiflorin, liquiritin, isoliquiritin, liquiritigenin, isoliquiritigenin, glycyrrhizic acid, glycyrrhetinic acid, and 6-gingerol. And the specificity, linearity, lower limit of quantification (LLOQ), lower limit of detection (LLOD), precision and accuracy, recovery, matrix effect, and stability were used to verify the HPLC-MS/MS method. This validated method was successfully applied for pharmacokinetic study of the 10 compounds in rat plasma after oral administration of GZD in three doses (40 g crude drug·kg − 1 , 20 g crude drug·kg − 1 , 10 g crude drug·kg − 1 ) and intravenous injection of GZD extraction at a dose of 2.0 g crude drug·kg − 1 .The measurements of pharmacokinetic parameters including AUC 0– ∞ , T 1/2 , T max , C max , Vz_F, Cl_F, and MRT, were performed using a non-compartmental model with Winnonlin 8.1 software. Results: The results showed that 10 compounds were detected in plasma after oral administration of GZD. the compounds (except for glycyrrhetinic acid) reached the maximum blood concentration quickly, whose Tmax was about 0.1-0.2 min. And a total of 9 compounds were detected after intravenous injection of GZD. The plasma concentration-time curve of these compounds declines rapidly at the beginning, and then decreased slowly, indicating that the plasma concentration-time curves were double exponential function curves. Conclusions: In this study, the developed method was suitable for pharmacokinetic analysis of the main compounds of GZD in rat plasma, and may reveal the pharmacodynamic material basis of GZD and provide a reference for the rational use of GZD in the clinic.

Animals 30 clean male Sprague-Dawley rats (300 ± 20 g) were purchased from SPF (Beijing) Biotechnology Co., Ltd. (Beijing, China quality certification number: SCXK (Jing) 2016-0002). The temperature and humidity of the controlled environmental conditions were 23 ± 3℃ and 50% ± 10℃, respectively. The principles of laboratory animal care and all protocols were in accordance with the relevant national legislation and local guidelines and were approved by Animal Care and

Preparation of GZD for oral adminstration
The volatile oil of Ramulus Cinnamomiand and Rhizoma Zingiberis Recens was extracted by steam distillation with four folds water for 2 h. The residues of the two herbal slices and other three herbal slices were boiled twice with eight folds water for 30 min each time. All those water decoctions were mixed together and concentrated to 1.0 g crude drugs per milliliter. After the volatile oil were added and mixed together, the extraction was stored at 4℃.

Preparation of GZD extraction for injection
GZD solution (1.0 g·mL −1 ) was centrifuged (15,000 rpm, 15 min, 4℃) and the supernatant was filtered with 0.2 μm filter membranes. Then, 100 mL filtrate was vacuum freeze-dried into powder, which was dissolved with 500 mL sodium chloride injection (0.9%). The solution was filtered with 0.2 μm filter membrane in the super clean bench to remove the bacteria and then packed separately and stored at 4℃. This extraction contained 0.2 g crude drugs per milliliter.

HPLC-MS/MS conditions
Agilent 1260 with the Rapid Resolution Liquid Chromatography (RRLC) system (Agilent Technologies, Santa Clara, CA, USA) was used. The separation of compounds was carried out on the Agilent SB C18 column (2.1 × 50 mm, 1.8 μm) at 35℃.
Agilent 6410 mass spectrometry was equipped with Electron Spray Ionization (ESI), using Multiple Reaction Monitoring (MRM) with negative ion mode detection. The ion source temperature was 350℃; the drying gas was nitrogen (N 2 , purity of 99.9%); the flow rate was 10 L·min −1 ; the nebulizing gas pressure was 40 psi; and the capillary voltage was 4.0 kv. As the analytes and IS had different ionization activities, two optimized mobile phases were used separately to ensure the sensitivity. The MS conditions, including the ion pair, fragmentor voltage, and collision energy were all optimized with standard solutions.

Sample preparation
In this study, the methods between liquid-liquid extraction (LLE) and protein precipitation (PPT) were compared to extract and purify the compounds to be tested in rat plasma. For LLE, ethyl acetate with or without hydrochloric acid were used as the extract liquor. For PPT, methanol and acetonitrile with or without hydrochloric acid were used to precipitate protein. The mixed reference solution was added into the blank plasma samples and prepared with above methods separately and injected into HPLC-MS/MS. The ratio of the peak area of each compound in the plasma sample to that of in the reference solution with same concentration was taken as the extraction rate. The method with a high extraction rate was chosen to prepare the samples.

HPLC-MS/MS method validation Specificity
The specificity of the method was assessed by comparing chromatograms of blank plasma samples from six individual rats, blank plasma spiked with the 10 analytes and 2 ISs. The plasma samples were obtained at 30 min or 4 h from the rats after oral administration of the GZD or GZD extraction.
Plasma calibration curves were constructed using the peak area ratios of the 10 analytes to the ISs, and applying separate weighted (1/χ 2 ) least squares linear regression. The lower limit of quantification (LLOQ) and the lower limit of detection (LLOD) were defined by the signal-to-noise ratio method. LLOQ should be ten times the noise level (S/N ≥ 10), and LLOD should be three times the noise level (S/N ≥ 3).

Precision and accuracy
The accuracy and precision of the method were evaluated by intra-and inter-day variations. Standard solutions were added to the blank plasma and prepared at three different concentrations (low, medium, and high). Intra-day precision was evaluated with six replicates at one day, and interday precision was evaluated at three days. The precision of each compound was evaluated by relative standard deviation (RSD) value, which should not exceed 15.0%, and the accuracy was estimated with the relative error (RE), which should be within ±15.0%.

Extraction recovery and matrix Effect
The recoveries of the analytes from plasma samples were determined by comparing the peak areas of the analytes in plasma samples after extraction to those of the same concentration of the analytes spiked into the solution extracted from plasma samples. The matrix effects were measured by comparing the peak areas obtained from samples with the analytes spiked after extraction, at three concentration levels (low, middle, and high), to those obtained from standard solutions at the same concentrations.

Stability
Stability of the analytes from the plasma samples were investigated by determining three different concentrations (low, medium and high samples) in five replicates under different storage conditions. The stability in plasma was assessed by analyzing (i) samples kept at room temperature (25°C) for 24 h, (ii) samples after three freeze-thaw cycles, (iii) samples after stored at −80℃ for 15 days.

Pharmacokinetic study
30 clean male Sprague-Dawley rats (300 ± 20 g) were divided into oral administration (ig) groups (high, medium, and low dose), intravenous injection (iv) group and blank group. There were 6 rats in each ig group and 5 rats in iv group. All animals were fasted for 12 h but with access to water before experiment.
Rats in ig group were given GZD by gavage according to their body weight (4.0 mL·100 g −1 ). The high dose group was given GZD in a concentration of 1.0 g crude drug·mL −1 , which was about 10 times as much as the clinical dosage (40.0 g crude drug·kg −1 ). The medium dose group was in a concentration of 0.5 g crude drug·mL −1 (20.0 g crude drug·kg −1 ), and the low dose group was in a concentration of 0.25 g crude drug·mL −1 (10.0 g crude drug·kg −1 ). 300 μL blood samples were collected in Eppendorf tubes from the postorbital venous plexus veins of each rat by capillary tubes before dose (0 h) and after doses at 5 min, 15 min, 30 min, 1 h, 2 h, 4 h, 8 h, 12 h, and 24 h.
Rats in iv group were injected with GZD extraction via the caudal vein according to their body weight (1.0 mL·100 g −1 ), which was equivalent to 0.5 times as much as the clinical dosage (2.0g crude drug·kg −1 ). Blood samples were collected at 0 min, 2 min, 15 min, 30 min, 1 h, 2 h, 4 h, 8 h, 12 h, and 24 h after injecting GZD extraction.
Rats in the blank group were used to collect blank blood samples from the abdominal aorta after anesthetized by intraperitoneal injection of 3% sodium pentobarbital in the dosage of 0.5mL·100 g −1 .
The blood samples were centrifuged at 12,000 rpm for 10 min at 4℃, and stored at −80℃ until analysis.

Data analysis
The validated HPLC-MS/MS methods were applied to analyze the concentrations of 10 compounds in rat plasma after oral or parenteral administration GZD at different times. The pharmacokinetic parameters were obtained by the non-compartmental analysis of plasma concentration versus time data using a non-compartmental model of the Phoenix Winnonlin 8.1 software (Certara, USA).

Optimization of chromatographic and Mass conditions
In order to obtain high detection sensitivity and good peak symmetry of the analytes, different ratios of formic acid, acetic acid, or ammonium acetate were chosen as the mobile phase annexing additive. Finally, we found that the separation and detection of the analytes were determined with these two phase conditions. The mobile phase used acetonitrile and 0.1% formic acid was suitable for paeoniflorin, albiflorin, liquiritin, isoliquiritin, liquiritigenin, isoliquiritigenin, glycyrrhizic acid, glycyrrhetinic acid, and icariin (IS1), while the mobile phase condition of acetonitrile and 0.2% acetic acid containing 2 mmoL·L −1 ammonium acetate was suitable for cinnamic acid, 6-gingerol, and aesculetin (IS2). The ionization conditions of those compounds, such as capillary voltage, nebulizer pressure, fragmentor voltage, and collision energy, were optimized by using the standard solutions. The optimized ionization conditions for the 10 compounds and 2 ISs were summarized in Table 1.

Sample pretreatment
The absolute recoveries of 10 analytes and 2 ISs after different sample pretreatments could be found in Table 2. The final sample treatment method was as follows: 20 µL IS solution (500 ng·mL −1 of icariin; 600 ng·mL −1 of aesculetin) and 10 µL hydrochloric acid (20 mmol·L −1 ) were added into each 100 µL plasma sample. The sample was mixed for 1 min and 200 µL methanol was added. Then the mixture was vortex for 3 min. The sample was centrifuged at 12,000 rpm for 15 min at 4℃. 300 µL of the supernatant was collected and evaporated to dryness by vacuum freeze-drying. Finally, the residue was dissolved by 100 µL mobile phase (ACN: H2O = 1:1) and vortexed for 3 min. After centrifuged at 12,000 rpm for 15 min at 4℃, 2 µL aliquot was injected into HPLC-MS/MS system.

Method validation specificity
Under the developed chromatographic and mass conditions, the sample chromatograms of blank plasma, the standard sample solution, the drug-containing plasma samples after oral administration of GZD, and the solution of GZD were presented in Fig. 2. The method had high specificity and could be used for the qualitative determination of these compounds in plasma.

Linearity, LLOQ, and LLOD
The results showed that the 10 compounds had good linearity in the corresponding concentration range. The regression equation, linear range and correlation coefficient (r), LLOQ, and lower limit of detection (LLOD) of each compound were listed in Table 3.

Precision and accuracy
The results of precision and accuracy at the three different concentration levels were presented in Table 4. The intra-and inter-day RSD values were below 6.9% and 9.6% respectively, while the corresponding RE values ranged from -10.8% to 5.3%. All the assay values were within the acceptable criteria.

Matrix effect and extraction recovery
Average recoveries of investigated targets ranged from 85.4% to 113.9%. The RSD values of all analytes were lower than 15.3%, which indicated that there was no significant loss of the compounds among the process of the protein precipitation in the plasma samples. The matrix effect of each compound ranged from 85.4% to 116.8%, indicating that there was no obvious matrix interference. The results showed that the method was accurate and acceptable. The data of extraction recovery and matrix effect were listed in Table 5.

Stability
After storage at -80℃ for 15 days and three freeze-thaw cycles, the stability RSD values of 10 compounds in plasma were all less than 9.9% ( Table 6). The results showed that the detected analytes were all satisfied with the criteria under all conditions, so the samples were stable during the test process.

Pharmacokinetic results
The validated HPLC-MS/MS methods were applied to the pharmacokinetic study of the 10 compounds in rat plasma after oral administration and intravenous injection of GZD. The pharmacokinetic parameters were calculated by Phoenix Winnonlin 8.1 software (Certara, USA). Mean blood concentration-time curves (C-T) of the 10 compounds taken orally were displayed in Fig. 3, and the main pharmacokinetic parameters were listed in Table 7. Simultaneously, the pharmacokinetic results of the compounds administered intravenously were mainly shown in Fig. 4 and Table 8.
After oral administration of GZD, 10 compounds were detected in plasma. Except for glycyrrhetinic acid (metabolite), the other compounds reached the maximum blood concentration quickly, whose T max was about 0.1-0.2 min. This pharmacokinetics phenomenon coincided with the characteristic of GZD as the Jiebiao formula (TCM term), which took effect quickly. All the AUC and C max values of those 10 compounds at different doses were positively correlated with the dose, and T 1/2 was independent with the dose, indicating that the compounds were in line with the firstorder kinetic process in vivo.
A total of 9 compounds were detected after intravenous injection of GZD. The plasma concentration-time curve of these compounds declines rapidly at the beginning, and then decreased slowly, indicating that the plasma concentration-time curves were double exponential function curves. Initially, these compounds did not reach the dynamic balance (the effects of distribution and elimination combined together). After a period of time, only the elimination process existed in vivo. The metabolites, glycyrrhetinic acid, was detected in plasma. However, it could not be fitted with winnonlin software due to its concentration was too low.The V_F values of paenoiflorin, liquiritigenin, and isoliquiritigenin were higher than those of other compounds, which indicated they had broad distribution in tissues.Glycyrrhizic acid had the longest T 1/2 and MRT. The compounds eliminated slowly in vivo.