Analysis of Bioactive Components, Simultaneous Determination and Pharmacokinetic Study of Six Components After Intragastric Administration of Qiling Wenshen Xiaonang Granule.

Background: Qiling Wenshen Xiaonang Granule (QLWSG), a compound preparation of traditional Chinese medicine, which was recently approved by the National Medical Products Administration for clinical trials of polycystic ovary syndrome (PCOS). It is necessary to do further researches about QLWSG for better understanding. Methods: In this study, qualitative and quantitative analysis of QLWSG in rat plasma were conducted by ultra-fast liquid chromatography coupled with triple quadrupole mass spectrometry (UFLC-QqQ-MS). The multi-reaction monitoring mode/electrospray ionization source of the triple quadrupole mass spectrometer was used for detection. The plasma samples were prepared by precipitating protein with organic solvent. Results: 21 prototypes in rat plasma were identied after intragastric administration of QLWSG, and most of the identied components are avonoids and phenolic acids. Furthermore, a sensitive and specic UFLC-QqQ-MS method was also built, which was used for simultaneous determination and pharmacokinetic study of six active components, astragaloside IV, calycosin, calycosin-7-O-β-D-glucoside, icariin, epimedin C and rosmarinic acid in rat plasma after intragastric administration of QLWSG with three doses (8, 16, 24 g/kg). All the calibration curves showed good linearity (r > 0.9942) in the range of concentration measured. The validated method is stable and reliable, meanwhile, successfully applied for pharmacokinetic evaluation of tested compounds. The results showed that there was no signicant difference in half-lives (t 1/2 ) and clearance (CL/F) of six analytes at the three doses were observed. Conclusions: It can be inferred that the pharmacokinetic behavior of QLWSG is positively correlated to dose at the range of 8–24 g/kg. Simultaneous determination of the six compounds in vivo for the rst time by UFLC-QqQ-MS. The validated method and the results of this study would benet therapeutic material basis and clinical


Background
Traditional Chinese medicine (TCM), which has been used by practitioners for thousands of years, plays an important role in clinical treatment and has attracted increasing attention due to its almost no side effects and bene cial therapeutic effect on western medicines [1]. Polycystic ovarian syndrome (PCOS) is a common endocrine disorder with three cardinal features: hyperandrogenism, anovulation, and polycystic ovary (PCO) morphology [2], which affects 5-10% of women of reproductive age and leading to the increased risk of anovulatory infertility and recurrent pregnancy loss [3,4]. However, there are currently no TCMs on the market for PCOS treatment. Qiling Wenshen Xiaosang, formerly known as Bushen Huatan Formula (BHF), was a decoction according to Professor Lihui Hou's (First A liated Hospital, Heilongjiang University of Chinese Medicine) clinical experience in treating PCOS for many years. Tasly Pharmaceutical Co. Ltd revised the decoction to granules during their cooperation with First A liated Hospital of Heilongjiang University of Chinese Medicine. Qiling Wenshen Xiaonang Granule (QLWSG), a compound preparation of traditional Chinese medicine, which was recently approved by the National Medical Products Administration for clinical trials of polycystic ovary syndrome (PCOS). The preclinical trial of QLWSG is being developed, and a good clinical effect is achieved. Moreover, some publications have shown that QLWSG was proved effective to PCOS by improving the in ammatory reaction and oxidative stress [5]. It has also been reported that QLWSG can improve insulin resistance and regulate ovulation function in PCOS patients. [6]. Combined with clinical biochemical data, QLWSG has unique advantages in improving the clinical indicators of PCOS patients. So, this granule is expected to complete the new drug registration and ll a gap in the market for TCM treatment of PCOS.
It is widely accepted that TCMs are mostly used in combination, and the composite formula will produce therapeutic effect [7]. QLWSG formula is composed by Radix Astragli, Epimedii folium, Salvia miltiorrhiza, Atractylodes rhizomes and Poria cocos. Among them, Radix Astragli was used as the sovereign drug, which played a major therapeutic role against PCOS and was reported to improve insulin secretion and lower blood glucose [8]. Epimedii folium, as the minister drug, the extract and active components of which was commonly used for regulating hormone level, enhancing sexual function and treating hyperglycemia [9][10][11][12][13]. It is well known that formulation is a complex mixture containing different chemical constituents which are responsible for therapeutic effects. Previous studies have identi ed 36 chemical constituents about QLWSG, including 19 avonoids, 11 phenolic acids, 4 triterpenoid saponins, etc. [14], and it is demonstrated that the major active components of QLWSG are avonoids such as calycosin-7-O-β-D-glucoside, calycosin,icariin and epimedin C [15]. In particularly, Astragaloside IV and icariin have been recognized as a unique chemical marker for the quality control of Radix Astragli and Epimedii folium in the Chinese Pharmacopoeia, respectively [16]. TCMs combination is used to cure diseases by their holistic effects [17]. Phenolic acids from Salvia miltiorrhiza in QLWSG are also play an important role. Rosmarinic acid is one of phenolic acids which are the main type of hydrophilic components from Salvia miltiorrhiza, and phenolic acids was reported has plenty of biological activities, was used for treatment of coronary arteriosclerosis, angina pectoris and hyperlipaemia [18,19]. Therefore, it is necessary to do further research on the above components.
Pharmacokinetic study on active constituents in herbal preparations,which is a good way to explain and predict correlations between e cacy and toxicity of TCMs. A clear understanding of pharmacokinetics will guide prescriber to choose doses and dose intervals which made the target tissues be exposed to appropriate drug concentrations for a su cient length of time, and will also help for the better application of the drug. [20,21]. As far as we know, the investigations of this formulation were mainly focused on clinical and pharmacological aspects, but no previous study was carried on pharmacokinetics of QLWSG components. Therefore, pharmacokinetics properties of bioactive ingredients in QLWSG remain unclear.
In the present study, 21 prototypes were identi ed from the plasma of QLWSG, and a sensitive, speci c and accurate method was also developed and validated for simultaneous determination of astragaloside IV, calycosin-7-O-β-D-glucoside, calycosin, icariin, epimedin C and rosmarinic acid in rat plasma using chloramphenicol and diazepam as internal standards (IS). The method was successfully applied to evaluate the pharmacokinetics behaviors of astragaloside IV, calycosin-7-O-β-D-glucoside, calycosin, icariin, epimedin C and rosmarinic acid after oral administration of QLWSG to rats, which would provide a solid foundation for further investigation of this TCM formula.

Materials and reagents
QLWSG was supplied by Tasly Pharmaceutical Co. Ltd (Tianjin, China). Astragaloside IV, calycosin-7-O-β-D-glucoside, icariin, rosmarinic acid, chloramphenicol (as IS) and diazepam (as IS) were obtained from the National Institutes for Food and Drug Control (Beijing, China). Calycosin and epimedin C were acquired from Tianjin Yifang science and technology Co. Ltd. According to HPLC/UV analysis, the purity of the standards was > 97.1%. Acetonitrile was acquired from Merck (Merck Serono, Germany, LCMSgrade). Formic acid was acquired from Sigma (LC-grade, Saint Louis, Mo, USA). Deionized water was achieved from a Milli-Q system (Millipore, Billerica, USA). The other reagents used were of analytical grade.
A Q-Trap® 5500 triple quadrupole mass spectrometry allocated with electrospray ionization (ESI) was applied on both in the positive and negative ionization mode (AB Sciex, CA, USA). MS conditions were installed as follows: Ion-spray voltage was set to -4.5 kV in negative ionization mode; Ions-pray voltage was set to 4.8 kV in positive ionization mode; ion source temperature was set to 550 o C; the nebulizer gas and heater gas (both are nitrogen )was set to 45 psi; and curtain gas, was set to 20 psi.
All the operations, the receiving and analysis of data on the instrument were controlled by Analyst software (AB Sciex, USA).

Preparation of various samples
With methanol as the solvent, the required amount of reference standard was added to prepare the mixed reserve standard solution containing astragaloside IV (0.5 mg/mL), calycosin-7-O-β-D-glucoside (0.5 mg/mL), calycosin (0.2 mg/mL), icariin (0.2 mg/mL), epimedin C (0.2 mg /mL), and rosmarinic acid (0.5 mg/mL). The IS solution (1.0 mg/mL) was also prepared with methanol as solvent. A series of working solutions were prepared by diluting reserve solutions with methanol, in which the concentrations of astragaloside IV, calycosin-7-O-β-D-glucoside and rosmarinic acid ranged from 1 to 5000 ng/mL,while the concentrations of calycosin, icariin and epimedin C ranged from 1 to 2000 ng/mL. The IS working solution (500 ng/mL) was also prepared by diluting the reserve solution with methanol. All the above solutions were kept at 4 o C before used.
The samples for standard calibration curves were prepared by gradually adding appropriate amount of the standards solutions to 150 µL blank plasma to nal concentrations of 1-5000 ng/mL for astragaloside IV, calycosin-7-O-β-D-glucoside and rosmarinic acid; and 1-2000 ng/mL for calycosin, icariin and epimedin C. As above, the quality control (QC) samples were prepared with blank plasma at concentrations of 2, 250, 4000 ng/mL for astragaloside IV, calycosin-7-O-β-D-glucoside and rosmarinic acid; 2, 100, 1000 ng/mL for icariin, calycosin and epimedin C. The above spiked samples were treated according to the following parts.

Processing of plasma samples
After thawing the plasma samples at room-temperature, the method of precipitating protein with organic solvent was used for the plasma sample preparation. After 100 µL IS solution (500 g/mL) was added into 150 µL plasma sample, the mixture was swirled and mixed for 1 min to achieve uniform mixing state. Then 500 µL 0.2% pre-cooled formic acid acetonitrile was added. After 1 min of vortex mixing, the mixture was centrifuged at 14,100 rpm (15 o C) for 10 min. The supernatant after centrifugation was removed to an eppendorf tube, 500 µL 0.2% pre-cooled formic acid acetonitrile was added for the second time about protein precipitation. After vortex and centrifugation as above, the supernatant was transferred to a 5 mL tube and evaporated to dryness at 30 o C under a gentle stream of nitrogen. The residue in the tube was redissolved in 80 µL mobile phase, then the supernatant was obtained by vortex mixing and centrifugation under the above conditions. And 2 µL of the supernatant was injected into the UFLC-QqQ-MS system for further analysis. The speci city of method, which was tested by analyzing the chromatograms of six separate blank rat plasma samples, plasma samples mixed with the analytes and IS, and plasma samples after a dosage by gavage. Blank rat plasma samples were used to analyze endogenous interference.

Linearity and lower limits of quanti cation (LLOQ)
Each calibration curve was built by scheming the peak area ratio of the analytes to IS (Y-axis) relative to the nominal concentration(X-axis) of astragaloside IV, calycosin-7-O-β-D-glucoside, calycosin, icariin,epimedin C and rosmarinic acid. Then the above curves were evaluated by weighted least-squares linear regression analysis with a weighed factor (1/x2). The lower limit of quanti cation (LLOQ) was de ned as the lowest concentration, which with an acceptable precision within 20% and accuracy of 80-120%.

Precision and accuracy
The precision and accuracy (intra-day and inter-day) of the method were calculated by analyzing the three QC concentrations (LQC, MQC and HQC) for plasma in six repetitions within a day and three days. Tolerable precision (relative standard deviation, RSD) and accuracy (RE) values were set within ± 15% of the nominal concentration, except LLOQ (within ± 20%).

Extraction recovery and matrix effect
The extraction recovery was determined by calculating the ratio of the amounts of QC samples(LQC, MQC and HQC) nally obtained against those originally spiked in the blank plasma. The effect of matrix was assessed by comparing the peak area acquired from samples where the processed matrix was mixed with standard solutions to those acquired from the unmixed reference standards solutions at the equally concentration.

Stability of samples
The stabilities of low, medium and high QC samples were analyzed under different storage and process states. The short-term stability of QC samples was analyzed by placing them at room-temperature for 24 h. The freeze/thaw stability was evaluated by analyzing QC samples after undergoing three cycles of freeze (-20 °C)-thaw (25 °C). Long-term stability in rat plasma stored at -80 °C was studied for a period of one month employing QC samples.

Identi cation of components in plasma of QLWSG
Two female Wistar rats (190 ± 10 g, age from 9 to 10 weeks) were acquire d from Beijing Vital River Laboratory Animal Technology Co. Ltd. Rats were housed in a 12 h light/12 h dark cycle at an ambient temperature (25 °C) and 60% RH. All animal care and experiments met the requirements of the NIH guidelines for the Care and Use of Laboratory Animals, and were also supported by the Animal Ethics Committee of Tasly academy (Tianjin, China). The rats were adapted to the lab for 1 week and allowed to drink water freely with fasted overnight before experiments. The gavage dose of QLWSG was twice the clinical equivalent dose (16 g/kg). QLWSG was dissolved in water. Plasma samples (400 µL) were obtained from orbital venous plexus of rats at 0.417, 1, 2, 4 h after gavage, and stored in 1.5 mL heparinized polythene tubes. The target plasma sample was supernatant obtained by centrifugation (4900 rpm, 10 min), which was then stored at -80 o C until advanced analysis. After centrifugation at 4900 rpm for 10 min, plasma samples were separated and stored at − 80 until further analysis.

The study of Pharmacokinetic
Eighteen female Wistar rats (190 ± 10 g, age from 9 to 10 weeks) were acquire d from Beijing Vital River Laboratory Animal Technology Co. Ltd. Rats were housed in a 12 h light/12 h dark cycle at an ambient temperature (25 °C) and 60% RH. All animal care and experiments met the requirements of the NIH guidelines for the Care and Use of Laboratory Animals, and were also supported by the Animal Ethics Committee of Tasly academy (Tianjin, China). The rats were adapted to the lab for 1 week and allowed to drink water freely with fasted overnight before experiments. 18 rats were randomly assigned to 3 groups (6 rats /group). Intragastric administration of QLWSG at low dose (clinical equivalent dose, 8 g/kg), medium dose (twice the clinical equivalent dose,16 g/kg), and high dose (three times the clinical equivalent dose, 24 g/kg), respectively. QLWSG was dissolved in water. Blood samples (400 µL) were obtained from orbital venous plexus of rats at 0.167, 0.417, 0.667, 1, 2, 4, 6, 8, 12, 24 h after gavage, and stored in 1.5 mL heparinized polythene tube. The target plasma sample was supernatant obtained by centrifugation (4900 rpm, 10 min), which was then stored at -80 o C until advanced analysis.

Identi cation of components in plasma of QLWSG
Using the MRM (multi reaction monitoring) mode of UFLC-QqQ-MS, the reference standards were analyzed to determine the information of the precursor ion and the product ion, and the parameters were optimized. The plasma samples collected at different time points after mixing was treated in 2.4 terms, and the sample was analyzed. From the reference material information, also based on information of products ions, 21 prototypes were identi ed from the plasma of QLWSG, including 5 components from Radix Astragli, 8 components in Epimedii folium, 1 component in Atractylodes rhizomes, and 9 components in Salvia miltiorrhiza. Most of the above components are avonoids and phenolic acids. The results are shown in Table. 1.

Pharmacokinetic study
On the basis of the identi cation results above, the pharmacokinetic study of six main components (astragaloside IV, calycosin-7-O-β-D-glucoside, calycosin, icariin,epimedin C and rosmarinic acid) were carried out. Their structures and characteristic ion pairs were shown in Fig. 1.

Optimization of UFLC-QqQ-MS method
The responses of calycosin-7-O-β-D-glucoside, calycosin observed in negative ionization mode were lower than that in positive ionization mode. But the responses of astragaloside IV, icariin, epimedin C and rosmarinic acid observed in negative ionization mode were higher than that in positive ionization mode. So, we also use two ionization modes at the same time.
MS parameters of collision energy (CE), declustering potential (DP), entrance potential (EP) and cell exit potential (CXP) were adjusted by injecting the standard solution respectively, so that the response of the precursor and product ions of each component reached the maximum (Table 2). Furthermore, the precursor and product ions of astragaloside IV, calycosin, calycosin-7-O-β-D-glucoside, icariin, epimedin C, rosmarinic acid and IS were shown in Fig. 2.

Speci city and selectivity
Representative chromatograms received from a blank sample, a spiked plasma sample with analytes (at LLOQ) and IS, and a plasma sample after intragastric administration of low-dose QLWSG were showed in Fig. 3. The retention times for astragaloside IV, calycosin-7-O-β-D-glucoside, calycosin, icariin, epimedin C and rosmarinic acid, chloramphenicol (IS) and diazepam (IS) were 8.38, 5.79, 6.18, 7.29, 7.16, 6.56, 7.02 and 9.44 min, respectively. The interference of endogenous peaks was not found in blank plasma sample, it showed that the UFLC-QqQ-MS with high selectivity.
The LLOQ of astragaloside IV, calycosin-7-O-β-D-glucoside, calycosin, rosmarinic acid, icariin and epimedin C were 1 ng/mL, which were su cient for pharmacokinetic studies in rats following the administration of three oral doses of QLWS extract. The RSD values of all analytes for precision were less than 10.61%, and the accuracy of all analytes ranged from 86.91-113.01%, which in accordance with the requirements of the Guideline on bioanalytical method validation (EMEA).

Precision and accuracy
The results of intra-and inter-day precision and accuracy for astragaloside IV, calycosin, calycosin-7-O-β-D-glucoside, rosmarinic acid, icariin and epimedin C are shown in Table 3. The intra-and inter-day precisions (RSD) were all less than 14.74%, and the accuracy ranged from 86.50-114.60%. The results indicated that the precision and accuracy of this method were acceptable. Precision and accuracy of intra-day (n = 6) and inter-day (n = 6)

Extraction recovery and matrix effect
The extraction recoveries and matrix effects of astragaloside IV, calycosin-7-O-β-D-glucoside, calycosin, rosmarinic acid, icariin, epimedin C and IS from rat plasma were shown in Table 4. Extraction recoveries and matrix effects of analytes The average recoveries of the samples were over 65.85% and the mean extraction recovery of IS was 76.27%. The above data indicated that the extraction recoveries of calycosin-7-O-β-D-glucoside, calycosin, astragaloside IV, rosmarinic acid, icariin, epimedin C and IS from the plasma were independent of concentration within the range of concentration assessed and were acceptable.
As displayed in Table 4, matrix effects of the all analytes ranged from 86. 97 to112.6%, and extraction recoveries of all analytes ranged from 65.85-103.44%. The results showed that the interference of matrix in rat plasma was negligible.

Stability
As presented in Table 5, the results of short-term stability, freeze/thaw stability and long-term stability were all within acceptable ranges. The results showed that the six compounds were steady under every tested situation.

Pharmacokinetic study of QLWSG
The validated method was successfully applied to pharmacokinetic study of astragaloside IV, calycosin-7-O-β-D-glucoside, calycosin, rosmarinic acid, icariin, epimedin C in rat plasma after intragastric administration of QLWSG with doses of 8 g/kg, 16 g/kg, 24 g/kg. The pharmacokinetic parameters were computed and analyzed by the DAS Software (version 3.2.6, National Medical Products Administration) using non-compartmental model. The mean plasma concentration-time curves (n = 6) of the six analytes were shown in Fig. 4. The main pharmacokinetic parameters in rats were presented in Table 6,including And data with P value less than 0.05 were judged to have statistical signi cance. Calycosin-7-O-β-D-glucoside and calycosin plasma concentrations were both lower than the limit of quantitation at 24 h for three doses, so only the data no more than 12 hours were adopted for analysis. As results indicated in Table 6, no signi cant differences of t 1/2 and CL/F among three doses (Table 6) after intragastric administration of QLWSG. At low, medium and high doses, the t 1/2 of calycosin-7-O-β-D-glucoside was 1.10 ± 0.07 h, 1.76 ± 0.32 h, 1.09 ± 0.63 h, respectively, suggesting that it was speedily eliminated from rat plasma. And it may also imply that calycosin-7-O-β-glucoside distributed to various tissues or promoted metabolic reactions in a short time that resulting its decline in plasma. The Tmax of astragaloside IV was shorter than 0.417 h. And at low, medium and high doses, astragaloside IV eliminated slower in plasma than the other analytes with the t 1/2 of 6.586 ± 1.64 h, 8.33 ± 2.89 h, 6.97 ± 0.79, respectively. It demonstrated that astragaloside IV absorbed rapidly in rats, and the elimination process is mainly. The results are similar to those reported in oral administration of astragaloside IV in rats [22].  Fig. 4, Rosmarinic acid showed a double-peak phenomenon in the high dose group, it may be due to excessive dose and enterohepatic circulation [23], etc. By normalizing the value of AUC 0−∞ with the corresponding dose, the results showed that the AUC 0−∞ values of the six analytes were proportional to the doses as shown in Fig. 5.. And the results of regression analysis showed that the correlation between the AUC 0−∞ value of Rosmarinic acid and its dose was the highest (correlation coe cient was 0.9981).
Since there was no signi cant difference in t1/2, we could infer that the pharmacokinetic behavior of QLWSG is positively correlated with dose in the range of 8-24 g/kg based on the above results.
Compared with published pharmacokinetic studies of individual compounds, the pharmacokinetic parameters of six compounds have some differences [24][25][26]. It could be the effect of herb-herb interactions in QLWSG. In addition, this is the rst time of the six compounds of QLWSG have been simultaneously determined and analyzed in vivo by an UFLC-QqQ-MS method.

Conclusions
In this study, a total of 21 prototypes of plasma of QLWSG were identi ed by the MRM mode of UFLC-QqQ-MS. These results provide information on the identi cation of bioactive compounds of QLWSG. And combining with the previous studies of the chemical constituents, the results of identi cation in vivo would bene t therapeutic material basis for QLWSG.
In addition, based on identi cation of prototypes and described a simple, sensitive and validated UFLC- Overall, simultaneous determination of these six bioactive components in rat plasma for pharmacokinetic investigations is then required, combined with the subsequent metabolomics research results of the QLWSG, not only to provide information on the identi cation of more bioactive compounds and to promote the clinical applications of QLWSG, but also to receive a better comprehension of the pharmacological action mechanism and to deduce the mechanism of the action of TCMs, explaining various events correlated with the e cacy of TCMs.  The structures and characteristic ion pairs of astragaloside IV, calycosin-7-O-β-D-glucoside, calycosin, icariin, epimedin C and rosmarinic acid. Mean plasma concentration-time curves of six analytes in rat plasma after intragastric administration of QLWSG at different doses (each point represents mean ± S.D., n = 6).