Pharmacokinetic Study of Thirteen Ingredients after the Oral Administration of Flos Chrysanthemi Extract in Rats by UPLC-MS/MS

A rapid and reliable UPLC-MS/MS method was developed and validated for the simultaneous quantification of thirteen bioactive compounds (luteolin, cynaroside, luteolin 7-O-glucuronide, isochlorogenic acid C, chlorogenic acid, cryptochlorogenic acid, apigenin, apigenin 7-glucoside, acacetin, hyperoside, isoquercitrin, tilianin, and hesperidin) in rat plasma. The compounds were separated on an ACQUITY UPLC BEH C18 column (2.1 × 100 mm, 1.7 μm) with a gradient mobile phase system of acetonitrile and 0.1% (v/v) formic acid aqueous solution at a flow rate of 0.3 mL/min. All compounds were quantitated using Agilent Jet Stream electrospray ionization (AJS ESI) in a negative ion mode. The lower limit of quantification (LLOQ) for all compounds was below 5 ng/mL. The intra- and interday accuracy ranged from -13.0% to 14.0%, and precisions were less than 12.2%. The extraction recoveries of the compounds were in the range of 56.9% to 95.0%, and the matrix effect ranged between 71.6% and 109.3%. Stability studies proved that the thirteen compounds were stable under tested conditions, with a relative standard deviation (RSD) of less than 11.4%. This developed method was successfully applied to the pharmacokinetic study of the 13 bioactive compounds after oral administration of Flos Chrysanthemi extract in rat by UPLC-MS/MS. Pharmacokinetic parameters of 8 out of the 13 compounds investigated are presented in this paper.

Pharmacokinetics plays an important role in drug development by quantitatively describing various dynamic processes in the body. At present, there are a few pharmacokinetic studies on luteolin, apigenin, diosmetin, and chrysoeriol in oral Flos Chrysanthemi extract [14][15][16]. The pharmacokinetics of Flos Chrysanthemi has not been comprehensively evaluated in the existing studies because its extract contains a wide variety of chemical compounds. Hence, it is of relevance to further research the pharmacokinetics of various ingredients in rat plasma after the oral intake of Flos Chrysanthemi extract.
This research develops a sensitive and reliable UPLC-MS/MS method for the determination of thirteen compounds in rat plasma after oral administration of Flos Chrysanthemi extract. In addition, this is the first pharmacokinetic study on the chemical components luteolin 7-O-glucuronide and apigenin 7-glucoside. This study would provide reference for further pharmacological studies on Flos Chrysanthemi extract.
2.4. Stock and Working Solution. Stock solutions of luteolin, cynaroside, luteolin 7-O-glucuronide, isochlorogenic acid C, chlorogenic acid, cryptochlorogenic acid, apigenin, apigenin 7-glucoside, acacetin, hyperoside, isoquercitrin, tilianin, hesperidin, and icariin (internal standard solution) were prepared individually and diluted to 1 mg/mL with methanol. Appropriate amounts of the 13 different stock solutions were added together in methanol for the mixed standard solution. The calibration solutions were prepared by adding 20 μL IS and appropriate volumes of mixed standard solution into 100 μL blank rat plasma. Low, medium, and high concentrations of quality control (QC) samples consisting of appropriate mixed standard solutions and blank plasma sample, adjusted to desired concentrations, were selected as calibration solutions. These solutions were kept at 4°C.

Sample
Preparation. 20 μL methanol (volume corresponding to that of QC samples and calibration curve) and 20 μL IS (1 μg/mL) were added into 100 μL plasma sample, which was then vortex-mixed with 400 μL acetonitrile for 3 min. The mixed solution was then centrifuged for 10 min at 14,000 rpm. After collecting the supernatant in a clean Eppendorf tube, the supernatant was evaporated to dryness under a gentle stream of nitrogen gas. The residue was dissolved in 100 μL methanol, vortex-mixed for 3 min, and then centrifuged for 10 min at 14,000 rpm. Finally, 5 μL supernatant was used for analysis by the UPLC-MS/MS system. 2.6. Method Validation 2.6.1. Specificity. Plasma samples acquired at 0.08 h after oral administration of Flos Chrysanthemi extract were compared to spiked plasma samples (containing working solutions and IS) and blank plasma samples from six different rats to evaluate method specificity and identify endogenous interfering substances.
2.6.2. The Calibration Curves and LLOQ. Blank rat plasma individually spiked with different concentrations of mixed standard solution and IS was quantitatively measured for three consecutive days, in replica, to validate the linearity. Calibration curves were drawn with peak-area ratios (y) of analyte relative to IS against its minimal concentration (x). The weighting factor was1/x 2 . The lowest limit of quantification (LLOQ) was calculated based on a signal-to-noise ratio of approximately 10 (S/N ≥ 10).
2.6.3. Precision and Accuracy. Intra-and interday precision and accuracy were estimated by analysing six duplicated QC samples at different concentrations as follows: 1, 10,

LC-MS/MS Optimization.
Various types of mobile phases were investigated for optimal separation of the 13 compounds, such as the use of acetonitrile or methanol as mobile (B) and 0.05% or 0.1% formic acid in water as mobile (A). Experimental results showed that acetonitrile (B) and 0.1% (v/v) formic acid in water (A) provided better peak shapes and reduced separation timings. Both positive and negative ion modes of the AJS ESI source were experimented for optimal mass spectrometry results. A negative ion mode showed greater signal intensity; thus, the thirteen compounds were quantitated with AJS ESI in the negative ion mode.

Sample Preparation.
In the study, we identified two simple and efficient methods for processing plasma samples: protein precipitation and liquid-liquid extraction. The ethyl acetate liquid-liquid extraction method provided better recovery of flavonoids, but the recoveries of chlorogenic acid, cryptochlorogenic acid, and isochlorogenic acid C were poor. The effects of methanol and acetonitrile on protein precipitation were then compared, and the results demonstrated that acetonitrile precipitation produced better recovery rates.

Method Validation of Bioanalysis
3.3.1. Specificity. The respective chromatograms were compared to evaluate method specificity based on analyte retention times and the presence of interference peaks. Figure 2 shows chromatograms of (a) blank plasma samples, (b) blank plasma samples spiked with analytes and IS, and (c) plasma sample collected after oral administration of Flos Chrysanthemi extract. Results showed consistent retention times and no interference peaks for all analytes across samples.

Precision and Accuracy.
The intraday and interday precision and accuracy were measured with six replicates of QC samples, with analytes set at low, medium, and high concentrations as described above. The results of the analysis are shown in Table 3. Accuracy (RE) ranged from -13.0% to 14.0%, and precision (RSD) ranged from 0.4% to 12.2%, indicating that the developed method was reliable. Table 4, the extraction recovery of all 13 analytes, at three different concentrations, ranged between 56.9% and 95.0%. The matrix effects on all analytes ranged between 71.6% and 109.3%. These results indicated that both the matrix effect and extraction recovery are satisfactory.

Recovery and Matrix Effect. As shown in
3.3.5. Stability. The stability of all 13 analytes during sample collection and processing was evaluated with the various storage conditions being tested on spiked plasma samples at three QC concentrations. As shown in Table 5, the RSD of all tested samples were below 11.4%, suggesting that all 13 analytes were stable in the above four test conditions.

Pharmacokinetic Study.
Plasma samples from rats orally administered with Flos Chrysanthemi extract (10.0 g/kg) were analysed with UPLC-MS/MS. Plasma concentration-time curves are shown in Figure 3, and the primary pharmacokinetic parameters of each analyte are summarized in Table 6.   Five analytes, namely, acacetin, hyperoside, isoquercitrin, tilianin, and hesperidin, were detected only at the first few blood sampling points following oral administration of the Flos Chrysanthemi extract, which made it difficult to plot a complete pharmacokinetic curve. Hence, these 5 analytes were excluded in the following results.
As shown in Table 6, C max of cynaroside and luteolin were 2547:84 ± 1121:18 ng/mL and 2079:55 ± 307:09 ng/mL,  respectively, ranking as the highest two amongst the 8 remaining analytes. In addition, AUC ð0−tnÞ of cynaroside and luteolin were larger than the other analytes, indicating a higher level of plasma exposure.
T 1/2 of chlorogenic acid and isochlorogenic acid C are 0.20 h and 0.24 h, respectively, indicating that these two compounds are eliminated shortly after oral administration. T 1/2 of cynaroside, luteolin 7-O-glucuronide, cryptochlorogenic acid, apigenin 7-glucoside, apigenin, and luteolin range from 5.01 h to 13.87 h, suggesting that these compounds have a relatively longer therapeutic time, especially apigenin 7-glucoside.

Conclusion
A reliable and sensitive UPLC-MS/MS method was developed to measure 13 ingredients (luteolin, cynaroside, luteolin 7-Oglucuronide, isochlorogenic acid C, chlorogenic acid, cryptochlorogenic acid, apigenin 7-glucoside, apigenin, acacetin, hyperoside, isoquercitrin, tilianin, and hesperidin) after the oral administration of Flos Chrysanthemi extract in rat plasma. This method offered adequate specificity, precision, recovery, and stability. In addition, the results showed that the blood concentrations of cynaroside and luteolin were higher than the other 11 analytes following oral administration of the Flos Chrysanthemi extract. Meanwhile, the absorption and elimination of chlorogenic acid and isochlorogenic acid C were rapid compared to other compounds. These pharmacokinetic parameters facilitate further development and clinical application for Flos Chrysanthemi.

Data Availability
The data used to support the findings of this study are available from the corresponding authors upon request.

Conflicts of Interest
The authors declare no conflict of interests.

Authors' Contributions
Qi Jia and Xuhua Huang contributed equally to this work.