Identification of absorbed constituents and evaluation of the pharmacokinetics of main compounds after oral administration of yindanxinnaotong by UPLC-Q-TOF-MS and UPLC-QqQ-MS

Yindanxinnaotong capsule (YDXNT), a traditional Chinese formula, has been used to treat cardio-cerebrovascular diseases for several decades. Previous research has focused on evaluating the pharmacological properties and main compounds of YDXNT in vitro and in vivo. However, the multiple bioactive compounds in vivo remain poorly understood. In the present research, an integrative strategy using UPLC-Q-TOF-MS combined with UPLC-QqQ-MS was employed to detect the absorbed constituents and investigate the pharmacokinetics of main compounds in the plasma after oral administration of YDXNT. UPLC-Q-TOF-MS was developed to detect the absorbed constituents and their metabolites in the plasma after oral administration in rats. A total of 52 constituents, including 44 prototype compounds and 8 metabolites, were identified or tentatively characterized. Then, nine main compounds (quercetin, isorhamnetin, kaempferol, ginkgolide A, ginkgolide B, ginkgolide C, bilobalide, tanshinone IIA, and salvianolic acid B) were chosen to further investigate the pharmacokinetic behavior of YDXNT using UPLC-QqQ-MS. The concentration of nine main constituents were in the range of 27.85–76.54 ng mL−1. This research provides a systematic approach for rapid qualitative analysis of absorbed constituents and for evaluating the pharmacokinetics of the main ingredients of YDXNT following its oral administration. More importantly, this work provides key information on the identification of bioactive compounds and the clarification of their action mechanisms, as well as on the pharmacological actions of YDXNT.


Introduction
Traditional Chinese medicine (TCM), one of the oldest phytomedicine systems in health care, has been used in Asia, such as in China, Korea, and Japan, for thousands of years. [1][2][3][4] The herbs used in TCM usually have complex formulas and function mainly via their multiple constituents, targets and modes of action (3M) in a network-based and holistic manner. 5 Thus, the therapeutic effects of TCM formulae rely on the joint effect of multiple ingredients. Most TCM formulae are taken orally, and only ingredients absorbed into the blood can exert their bioactivities. 6 Therefore, it is necessary to trace the compounds in a TCM prescription in vivo and evaluate their pharmacokinetics to provide more in-depth insight into the main active components and therapeutic mechanisms of TCM formulae. In recent years, an integrative strategy using ultra-highperformance liquid chromatography coupled with quadrupole time-of-ight mass spectrometry (UPLC-Q-TOF-MS) combined with ultra-high-performance liquid chromatography with triple quadrupole mass spectrometry (UPLC-QqQ-MS) has been widely used to identify the absorbed constituents and determine main compounds in complex matrices. [7][8][9][10][11] Hence, they are valuable analytical techniques for identifying key compounds and evaluating the pharmacokinetics of TCM formulae in vivo.
Our previous research focused on the chemical compounds, quality control, and in vitro and vivo pharmacological properties of YDXNT. In studying the chemical compounds of YDXNT, 70 chemical constituents, including 23 avonoids, 4 ginkgolides, 15 phenolic compounds, 18 diterpenoid tanshinones, and 10 ginsenosides, were tentatively identied using various detection approaches. Furthermore, ngerprint analysis was established to evaluate the uniform quality of YDXNT so capsules. 16 In in vivo experiments, taking YDXNT and swimming may prevent atherosclerosis through a synergistic effect between YDXNT and swimming leading to improved blood circulation, hemorheological parameters, blood lipid levels and vascular endothelium in rats. 17 In an isolated heart experiment, YDXNT and its main compounds elevated heart function, coronary ow and SOD levels and decreased levels of MDA and inammatory factors. 18 Ultimately, our research demonstrated that YDXNT and its major constituents relieve atherosclerosis by regulating lipids, reducing lipid particle deposition in the endothelial layer of arteries, enhancing antioxidant power, and repressing inammatory activity by inhibiting the nuclear factor-kappa B signaling pathway both in vivo and vitro. 19 Previous work focus on the pharmacokinetics of main constituents in YXY, DS1, SQ, SZ, and DZXX. [20][21][22][23][24] However, until now, little has been known about how many constituents are absorbed into the blood aer oral administration of YDXNT or about the nal fate of the formula in vivo. This presents a huge obstacle to understanding the pharmacological mechanisms of YDXNT. Thus, the present research was designed to determine the absorbed compounds and metabolites in biological samples from rats and investigate the pharmacokinetics of the main abundant active components.
In this research, UPLC-Q-TOF-MS was used to analyze and identify the absorbed ingredients and possible metabolites in rat plasma aer oral administration of YDXNT due to its high chromatographic resolution, high sensitivity, short analysis times, good separation, and highly accurate mass values. [25][26][27] In addition, UPLC-QqQ-MS was used for simultaneous quantitation of the main components quercetin (QCT), kaempferol (KMF), isorhamnetin (ISR), ginkgolide A (GA), ginkgolide B (GB), ginkgolide C (GC), bilobalide (BB), tanshinone IIA (TSIIA), and salvianolic acid B (SAB) in rats, with baicalein used as the internal standard (IS). 28,29 These compounds exert antiatherosclerotic and endothelium-independent vasodilator effects, prevent inammatory damage and have a protective effect on nitric oxide, antioxidant action, and anti-vascular inammation. [26][27][28][29] The pharmacokinetics of QCT, KMF, ISR, GA, GB, GC, BB, TSIIA, and SAB were investigated in rats aer oral administration of YDXNT, and the pharmacokinetics properties of this formula were further speculated. This research represents the rst detailed investigation into the absorbed constituents, metabolites, and pharmacokinetics of YDXNT. It may provide valuable information for better understanding the pharmacological mechanisms and clinical applications of YDXNT.

Chemicals and reagents
YDXNT so capsules were provided by Guizhou Bailing Group Pharmaceutical Co., Ltd. Standards of rutin and cryptotanshinone were purchased from Chengdu Chroma-Biotechnology Co., Ltd. (Sichuan, China). Kaempferol, quercetin, tanshinone IIA, salvianolic acid B, ginkgolide A, ginkgolide B, bilobalide, and baicalein were purchased from Chengdu Herb Purify Co., Ltd. (Sichuan, China). Isorhamnetin, ginkgolide C, ginsenoside Rg1, ginsenoside Re, ginsenoside Rd, and notoginsenoside R1 were purchased from the National Institutes for Food and Drug Control (Beijing, China), Ltd., and HPLC-grade acetonitrile and methanol were obtained from Thermo Fisher Scientic Inc. (Shanghai, China). Formic acid was purchased from Sinopharm Chemical Reagent Co., Ltd. (Shanghai, China), and the water used in the experiments was puried with a Milli-Q system (Sartorius Arium® Pro, Germany).

Preparation of YDXNT sample
The samples were rst processed by completely removing the outer capsule layer. Then, a 48.0 g sample was precisely weighed and soaked with 100 mL de-ionized water. The sample was ultrasonically dissolved for 30 min and then stored at À4 C until it was orally administered to rats.

Animal experiments
All animal procedures were performed in accordance with the Guidelines for Care and Use of Laboratory Animals of Center for Laboratory Animal Care, China Academy of Chinese Medical Science and approved by the Animal Ethics Committee of Institute of Chinese Material Medica, China Academy of Chinese Medical Science.
Twenty-four male Sprague-Dawley rats (250 AE 30 g) were obtained from the Laboratory Animal Center of the Academy of Military Medical Sciences (Beijing, China) and divided into four groups (six rats each). The rats were housed and fed at a temperature of 25 AE 2 C and a relative humidity of 50% for three days in the breeding room. Standard chow and deionized water were provided prior to the oral treatment. All rats were fasted for 12 h with free access to water before the experiment. YDXNT was administered orally to rats at a dose of 4.8 g kg À1 d À1 , and blood was collected from the abdominal aorta aer 20% urethane anesthesia and then centrifuged at 3000 rpm for 10 min at 4 C to obtain the plasma. All samples were stored at À20 C until analysis. Blank plasma samples were prepared in the same manner.

Serum sample preparation
Six milliliters of 75% methanol was added to two milliliters of prepared plasma and then vortexed for 60 s. The mixture was centrifuged at 4000 rpm for 10 min at 4 C. The supernatant was transferred to another EP tube and dried using a pressure blowing concentrator. The residue was redissolved with 200 mL of 75% methanol, vortexed for 60 s, and then centrifuged at 13 000 rpm for 10 min. To an aliquot of 200 mL plasma, 120 mL of 0.2 mol L À1 acetic acid-sodium acetate buffer solution (pH ¼ 4.6), 100 mL of the IS baicalein and 80 mL of 980 U mL À1 bglucuronidase and sulfatase was added and then vortexed for 1 min. The mixture was hydrolyzed in a water bath at 37 C for 30 min, and 600 mL acetone was added to stop the hydrolysis. The sample was cooled, followed by vortexing for 2 min and then centrifugation at 13 000 rpm for 10 min. The supernatant uid was evaporated to dryness under a ow of nitrogen gas.

Preparation of standard, calibration, and quality control solutions
Stock solutions of QCT, ISR, KMF, GA, GB, GC, BB, TSIIA, SAB and IS were separately prepared by dissolving the compounds in methanol at a concentration of 200 ng mL À1 ; all solutions were stored at 4 C. Working solutions for calibration and QC samples were prepared by adding the diluted stock solutions to blank rat plasma. Calibration standards of plasma-derived working solutions with QCT, ISR, KMF, GA, GB, GC, BB, and TSIIA were all prepared with nal concentrations ranging from 1-100 ng mL À1 . The SAB solution was prepared with nal concentrations ranging from 5-500 ng mL À1 . The IS solution was prepared at a nal concentration of 100 ng mL À1 in methanol. For method validation, QC plasma samples of QCT, ISR, KMF, GA, GB, GC, BB, and TSIIA were all prepared separately at three concentrations (1, 10 and 100 ng mL À1 ). The QC plasma sample of SAB was prepared at three concentrations (5, 50, and 500 ng mL À1 ).

UPLC-Q-TOF-MS for qualitative analysis
UPLC data were produced using the Waters Synapt G2-S UPLC system (Waters, Milford, USA) equipped with a quaternary pump system, online degasser, autosampler, thermostatically controlled column compartment, and diode array detector. The system was controlled with Masslynx V4.1 soware. The chromatographic column ACQUITY BEH C18 (100 Â 2.1 mm, 1.7 mm) equipped with VanGuard™ BEHC18 1.7 mm was used. The column was eluted with a gradient of A (0.1% formic acid in deionized water) and B (CAN) at a ow rate of 0.3 mL min À1 and a temperature of 30 C: 0-2.5 min, 10-16% B; 2.5-6.5 min, 16-16% B; 6.5-9.      For mass detection, mass data were recorded using the Waters Synapt G2-S Q-TOF Mass Spectrometer system equipped with an atmospheric pressure chemical ionization (APCI) electrospray ionization (ESI) interface (Waters, Milford, USA). The conditions of the TOF/MS were set as follows: capillary, 2.5 kV; source temperature, 120 C; cone gas ow, 50 L h À1 ; nebulizer, 6 bar; desolvation temperature, 450 C; sheath gas ow, 12 L min À1 ; nozzle voltage, 10-1000 V; The mass spectrometer was operated from 100 to 1700 m/z. Accurate masses and compositions of the precursor and fragment ions were calculated using MassLynx 4.1 soware. The UPLC and Q-TOF-MS methods have been optimised (ESI 1 †).

UPLC-QqQ-MS for quantitative analysis
Chromatographic analysis was performed using an Agilent 1290 Series UPLC from Agilent Technologies (Palo Alto, CA, USA), equipped with a quaternary pump an online degasser, autosampler and column oven. Chromatographic separation was performed on the Agilent ZORBAX SB-Aq C18 column (100 mm Â 2.1 mm, 1.8 mm, Agilent, USA). The analytical column temperature was maintained at 30 C and eluted with a mobile phase comprising 0.1% formic acid in water (A) and acetonitrile (B) using the following gradient program: 25-90% B from 0 to 1 min, 90% B from 1 to 3.5 min, and 90-25% B from 3.5 to 5 min at a ow rate of 0.3 mL min À1 . The equilibration time was 5 min, and the injection volume of sample and reference constituents was 5 mL.
Mass spectrometry was performed on the Agilent 6490 triplequadrupole mass spectrometer (Palo Alto, CA, USA) equipped with an atmospheric pressure chemical ionization (APCI) interface. The mass spectrometer was set in negative ionization mode, with the capillary voltage set at 3500 V. The other parameters in the source were set as follows: desolvation gas ow, 15 L min À1 ; source temperature, 290 C; N 2 sheath gas temperature, 350 C; nebulizer gas (N 2 ) pressure, 50 psi; ow, 12 L min À1 . The mass spectrometer scanned in multiple reaction monitoring (MRM) mode. The UPLC and QqQ-MS methods have been optimised (ESI 2 †). Fig. 1 shows the MS/MS product ion spectra of the analytes and IS. The optimized extraction ion pairs and parameter values are shown in Table 1.

Method validation
Method validation was performed according to the guidelines for bioanalytical method validation published by the US FDA. 30 The analytical method used in this research was validated to demonstrate its specicity, selectivity, linearity, lower limit of quantication (LLOQ), precision, accuracy, stability, extraction recovery rates and matrix effects of the samples (ESI 3 †).

Pharmacokinetics study
Six male rats were used in the pharmacokinetics study. All the rats were orally administered 2.4 g kg À1 YXCNT suspended in an aqueous solution of 0.5% carboxymethyl cellulose sodium (QCT: 4.118 mg; KMF: 5.067 mg; ISR: 0.965 mg; GA: 1.416 mg; Table 3 (1) Chromatographic and MS data of metabolites analysed by UPLC-Q-TOF-MS in negative mode.

Identication of prototype constituents and metabolites in drug-containing plasma
The optimized LC-MS protocol was used for the separation and identication of compounds in drug-containing plasma. Exogenous constituents of YDXNT in the plasma were divided into prototype components and metabolites. Ultimately, a total of 52 compounds, including 44 prototype constituents and 8 Fig. 3 The mass spectra of protocatechuic aldehyde glucuronide in negative mode and the proposed fragmentation pathway. Fig. 4 The mass spectra of hydrolysis ginkgolides C in negative mode and the proposed fragmentation pathway. metabolites, were detected by comparison to the blank plasma sample (Fig. 2, Tables 2 and 3).
In our research, one monophenolic-related metabolite (M1) was detected. M1 (t R ¼ 2.59 min) showed a deprotonated molecular ion at m/z 313.0555, and product ions of m/z 137.0233 were obtained. The neutral loss of 176 Da (C 6 H 8 O 6 ) indicated that M1 was the glucuronide conjugate 32,33 (Fig. 3). Based on retention time, accurate quasi-molecular ion, and MS/MS data, as well as the related literature, 32-34 M1 was tentatively identied as the characteristic ion of protocatechuic aldehyde glucuronide.
Ginkgolide-related metabolites were the main possible xenobiotic metabolites in plasma. 35,36 In the present study, 3 ginkgolides and 4 ginkgolide-related metabolites were detected. The 3 ginkgolides (ginkgolide A, ginkgolide C and ginkgolide B) were identied by comparison to standards. The characteristic ions at m/z 351, m/z 383 and m/z 367 were observed in ginkgolide A, ginkgolide C, and ginkgolide B, respectively. The common fragmentation pattern of ginkgolide was the successive loss of 2CO. 37 M2, M3 and M4 have [M À H] À ions at m/z , which have a similar mass schizolysis rule as seen for ginkgolide. According to the above analysis, we tentatively presumed that M2 is a hydrolyzed metabolite of ginkgolide B, M3 is a hydrolyzed metabolite of ginkgolide A, and M4 is a hydrolyzed metabolite of ginkgolide C (Fig. 4). 35,36,38,39 Furthermore, M6 (m/z 443.1534) was tentatively identied as another hydrolyzed metabolite of ginkgolide A (bi-ionized ginkgolide A) by comparing deprotonated molecular [M À H] À and characteristic fragments [M À H-74] to those reported in the literature. 36 Two avonoid-related metabolites were detected. M5 (t R ¼ 10.  5) and kaempferol-O-glucuronide by comparing retention times, accurate quasi-molecular ions, and MS/MS fragmentation data as well as by referring to the related literature. [40][41][42][43][44] One tanshinone-related metabolite was tentatively identied from rat plasma on the basis of accurate deprotonated molecular data, MS/MS fragmentation data, and previous reports. [45][46][47] M7 gives rise to a protonated molecule [M + H] + at m/z 293.0816. This molecule was 2 Da smaller than TSIIA. The neutral loss of m/z 15, m/z 18, and m/z 28 indicated the loss of CH 3 , H 2 O, and CO, respectively (Fig. 6). Corresponding to the deprotonated molecular ion and its fragments, M7 was tentatively identied as dehydrogenated TSIIA, which has been reported in the literature. 34 3.4.2 Linearity. The calibration curves showed good linearity (R 2 > 0.9931) over the concentration ranges of the nine Table 6 Extraction recovery and matrix effect for the analytes or bioactive compounds in rat plasma and brain homogenates (n ¼ 6) Paper constituents in rat plasma. The correlation coefficients of the standard curves and the linear ranges of the plasma are listed in Table 4.

Precision and accuracy.
The precision and accuracy data for six replicates of the QC samples at three concentrations are shown in Table 5. The results demonstrated that the precision (RSD%) of the method was 10.26%.
3.4.4 Extraction recovery rates and matrix effects. The extraction recovery and matrix effects of each QC concentration are listed in Table 6. The extraction recoveries ranged from 76.86% to 95.77% at different concentrations in plasma. All ratios for matrix effects were in the range of 94.56-103.25% in plasma samples. Thus, no signicant matrix effects were observed for the analytes.
3.4.5 Stability. The data for freeze-thaw stability, shortterm temperature stability, long-term stability, and autosampler stability under different storage conditions are summarized in Table 7. The results were well within the acceptable limits, therefore validating the established method for sample extraction, storage and intermittent analysis and indicating its suitability for pharmacokinetic study.

Pharmacokinetics studies
The developed and validated UPLC-QqQ-MS/MS method was successfully applied in the pharmacokinetics study of QCT, ISR, KMF, GA, GB, GC, BB, TSIIA, and SAB in rat plasma aer an oral dose of 4.8 g kg À1 d À1 YDXNT. Fig. 7 shows the mean plasma concentration-time proles of nine constituents in rat plasma. Table 8 shows the main pharmacokinetic parameters of QCT, ISR, KMF, GA, GB, GC, BB, TSIIA, and SAB in rat plasma.
GA and GB have similar pharmacokinetic parameters and may be connected to similar polar constituents, similar to the absorption and elimination processes. 48 The C max of GC was the lowest among the four ginkgolides, indicating that GC may be partially converted to hydrolyzed ginkgolide C, as we detected. The time required to reach the maximum plasma concentration (T max ) was 0.75 h for GA, 1.00 h for GB, 1.50 h for GC, and 0.75 h for BB. The T 1/2 of four ginkgolides in YDXNT were 7.52 h, 7.84 h, 5.45 h, and 5.55 h. Absorption of the four ginkgolides was slower than that of the avonols, TSIIA and SAB, while their distribution occurred relatively faster than that of the avonols, TSIIA and SAB. Additionally, the avonols showed double peaks in the mean plasma concentration curves. This phenomenon has been reported previously, indicating that these components might have enterohepatic recirculation. 49,50 The double peaks suggested that enterohepatic recirculation, as well as intertransformation among the compounds, might have occurred. One possible explanation for this phenomenon is that some of the other compounds with similar structures might have transformed into these compounds. Since drug absorption is a complex process, more detailed adsorption studies are needed to ascertain the mechanism of the double-peak phenomenon. The C max of TSIIA (38.34 ng mL À1 ) is higher than that of SAB (32.00 ng mL À1 ), and the T max and T 1/2 of TSIIA (0.25 h, 7.04 h) are lower than those of SAB (0.75 h, 12.17 h), indicating the absorption and distribution rates of TSIIA are faster than those of SAB in rats.

Conclusion
In this research, we developed the UPLC-Q-TOF-MS method for analyzing and identifying the main absorbed constituents of YDXNT and their possible metabolites in rat plasma. A total of 52 constituents, including 44 prototype compounds and 8 metabolites, were identied or tentatively characterized. Among the prototype constituents, 15 avonoids, 4 ginkgolides, 8 phenolic, 8 ginsenoside, and 7 diterpenoid tanshinones were identied. The metabolic routes of 8 metabolites in rat plasma aer oral administration were hypothesized. The result indicated that the metabolites of YDXNT undergo the phase I metabolite pathway of hydrolysis (M2, M3, M4, and M6), dehydrogenation (M8) and the phase II metabolic route of glucuronide (M1, M5 and M7). Furthermore, a sensitive and reliable method based on UPLC-QqQ-MS was established for simultaneous quantication of nine abundant main compounds in rat plasma. The developed assay was successfully applied to assess the pharmacokinetics of ginkgetin aglycone, terpene lactones, salvianolic acid, and tanshinone from YDXNT in rat plasma. Our research claries for the rst time the absorbed constituents, possible metabolites, and pharmacokinetics of YDXNT in vivo. This work provides more in-depth insight into the main active constituents of YDXNT in vivo as well as helpful information for clinical application. More importantly, our ndings are helpful for better understanding the material foundation and action targets underlying the efficacy of YDXNT.

Author contributions
All the authors have approved the manuscript and agree with submission to your esteemed journal.

Conflicts of interest
There are no conicts of interest to declare.