Simultaneous Determination and Pharmacokinetics Study of Three Triterpenes from Sanguisorba officinalis L. in Rats by UHPLC–MS/MS

A selective and rapid ultra-high-performance liquid chromatography–tandem mass spectrometry (UHPLC–MS/MS) method was established and validated for the determination of ziyuglycoside I, 3β,19α-dihydroxyurs-12-en-28-oic-acid 28-β-d-glucopyranosyl ester, and pomolic acid in rats after the oral administration of ziyuglycoside I, 3β,19α-dihydroxyurs-12-en-28-oic-acid 28-β-d-glucopyranosyl ester, pomolic acid, and Sanguisorba officinalis L. extract. The separation was carried out on an ACQUITY UPLC®HSS T3 column (2.1 mm × 100 mm, 1.8 μm), using methanol and 5 mmol/L ammonium acetate water as the mobile phase. The three compounds were quantified using the multiple reaction monitoring mode with the electrospray ion source in both the positive and negative mode. Liquid-liquid extraction was applied to the plasma sample preparation. Bifendate was selected as the internal standard. The intra-day and inter-day precision and the accuracy of the method were all within receivable ranges. The lower limit of quantification of ziyuglycoside I, 3β,19α-dihydroxyurs-12-en-28-oic-acid 28-β-d-glucopyranosyl ester, and pomolic acid were 6.50, 5.75, and 2.63 ng/mL, respectively. The extraction recoveries of analytes in rat plasma ranged from 83 to 94%. The three components could be rapidly absorbed into the blood (Tmax, 1.4–1.6 h) both in the single-administration group or S. officinalis extract group, but the first peak of PA occurred at 0.5 h and the second peak at 4–5 h in the S. officinalis extract. Three compounds were eliminated relatively slowly (t1/2, 7.3–11 h). The research was to establish a rapid, sensible, and sensitive UHPLC–MS/MS method using the multi-ion mode for multi-channel simultaneous mensuration pharmacokinetics parameters of three compounds in rats after oral administration of S. officinalis extract. This study found, for the first time, differences in the pharmacokinetic parameters of the three compounds in the monomer compounds and S. officinalis extract administration, which preliminarily revealed the transformation and metabolism of the three compounds in vivo.


Introduction
Chinese herbal medicine, which can be used as both food and medicine, is widely applied to the prevention and cure of various diseases and clinical presentations, with its low toxicity and other unique advantages. The understanding of natural compounds, toxicity, and side effects, and the safe use of Chinese herbal medicine, can be an effective adjuvant treatment for disease [1,2]. In addition to the efficacy of the drug, understanding the safety of the drug and the pharmacokinetic profile of the bioactive compounds in the drug is crucial for determining the potential for successful treatment [1,2]. Sanguisorba officinalis L. (S. officinalis), belonging to the Rosaceae family and Sanguisorba genus, has detoxification, analgesic [3], and hemostasis [4] functions, which are noted in the Chinese Both the negative and positive ion modes were checked by different mobile phases. Both m/z 784.5 for ZGI and m/z 652.5 for DGE with [M + NH 4 ] + can acquire strong and steady signals. To improve the sensitivity and enhance the response, the negative mode of [M − H] − at m/z 471.2 was more appropriate for PA. To obtain a good resolution, an ACQUITY UPLC ® HSS T 3 (2.1 mm × 100 mm, 1.8 µm) column was adopted to separate the three analytes. In summary, two scanning time intervals with positive and negative modes were adopted: positive (ESI + : 0-5.0 min) and negative (ESI − : 5.0-6.5 min) ion modes. The main reason for segmental detection of the three compounds was to protect the capillary. Finally, the separating conditions were set with methanol (mobile phase A) and 5 mM ammonium acetate (mobile phase B) as the mobile phase at a flow rate of 0.3 mL/min.

Method Validation 2.2.1. Selectivity
The selectivity was appraised by comparing the chromatograms of blank plasma, a plasma sample spiked with LLOQ analytes and IS, and a plasma sample from rats following single oral administration of monomers and S. officinalis extract, as listed in Figure 1. The results suggested that there was no interference of endogenous substances during the retention time of the IS and analyte. The gradient elution program was as follows: 0-4.5 min, 70-90% B; 4.5-6.0 min, 90% B; 6.0-6.3 min, 90-70% B; 6.3-6.5 min, 70% B. The total running time was 6.5 min, and the injection volume was 10 μL.

Selectivity
The selectivity was appraised by comparing the chromatograms of blank plasma, a plasma sample spiked with LLOQ analytes and IS, and a plasma sample from rats following single oral administration of monomers and S. officinalis extract, as listed in Figure 1. The results suggested that there was no interference of endogenous substances during the retention time of the IS and analyte.

Linearity and Lower Limits of Quantification
The typical equations of the calibration curves were Y = 1.522X + 1.390 × 10 −1 for ZGI, Y = 4.734X + 3.194 × 10 −2 for DGE, Y = 0.685X + 6.882 × 10 −3 for PA, where X means the plasma concentration of analytes and Y represents the peak area ratio of analytes to IS. As shown in Table 1, Figure 2, the calibration curves for ZGI, DGE, and PA had good linearity over the concentration ranges of 6.50-2600, 5.75-2300, and 2.63-1050 ng/mL, with all correlation coefficients r > 0.9960. The results showed that the compounds were within the linearity ranges. The LLOQ of ZGI, DGE, and PA were 6.50, 5.75, and 2.63 ng/mL, respectively. plasma concentration of analytes and Y represents the peak area ratio of analytes to IS. As shown in Table 1, Figure 2, the calibration curves for ZGI, DGE, and PA had good linearity over the concentration ranges of 6.50-2600, 5.75-2300, and 2.63-1050 ng/mL, with all correlation coefficients r > 0.9960. The results showed that the compounds were within the linearity ranges. The LLOQ of ZGI, DGE, and PA were 6.50, 5.75, and 2.63 ng/mL, respectively.

Precision and Accuracy
The precision and accuracies of the three analytes in rat plasma at four levels are listed in Table 2. In this experiment, the intra-day and inter-day precision (RSD) of the three analytes were not more than 13%, and the accuracies (RE) for the three analytes ranged from −0.71% to 4.7%. The precision and accuracy were in line with the relevant provisions of the biological sample analysis guidelines. Table 2. Intra-day and inter-day precision and accuracies for the determination of the three analytes in rat plasma (n = 6).

Matrix Effect and Extraction Recovery
The extraction recoveries and matrix effects of the three components in rat plasma are listed in Table 3. The extraction recoveries of the three ingredients in rat plasma were 83-94% at three QC levels. The extraction recovery of IS was 94%. The matrix effects were between 102 and 109% for the three compounds in rat plasma. There were no significant matrix effects for the three analytes in rat plasma. Table 3. Matrix effects and extraction recoveries for analytes and IS in rat plasma (n = 6).

Stability
The stability of rat plasma under different storage conditions was determined. The results (Table 4) showed the three analytes were stable in plasma after three freeze-thaw cycles, and then at room temperature for 4 h. The stability of the analytes at the postpreparative stage also showed that the sample did not degrade significantly when it remained at 4 • C for 12 h. All compounds maintained steady levels for 2 weeks at −20 • C.

Carryover Effect
The blank rat plasma sample was immediately analyzed after the injection of the highest concentration calibration standard sample. The interference in the blank rat plasma sample was 6.0% at LLOQ (<20%) and 0.85% of the response for IS (<5%), which suggested that no obvious carryover effect was found under the described conditions.

Pharmacokinetic Studies
The UHPLC-MS/MS method was successfully used in the pharmacokinetic studies of the three analytes after single dose administration of ZGI, DGE, PA, and S. officinalis extract in rats (0.55 g/kg, equivalent to 0.03, 0.013, and 0.007 g/kg of ZGI, DGE, and PA, respectively). Based on the body surface area calculations of people and the animals and equivalent dose conversion calculations, the dosage of the rats was 0.55 g/kg. The mean plasma concentration-time curves of compounds are listed in Figures 3 and 4. The half-time (t 1/2 ), maximum plasma concentration (C max ), time to reach C max (T max ), and area under concentration-time curve (AUC 0→t and AUC 0→∞ ) were calculated as shown in Tables 5-7.   Notably, as Tables 5-7 show, the pharmacokinetic process of three analytes differed between the monomer groups and S. officinalis extract. The T max of ZGI after consuming the extract was similar to that of after consuming pure ZGI. In contrast, the T max of DGE after consuming extract is higher than that of the pure DGE, suggesting that DGE in extract needs a longer time for absorption. The T max values of ZGI, DGE, and PA were 1.4 ± 0.20, 1.6 ± 0.20, and 4.2 ± 0.41 h in the S. officinalis extract, and 1.4 ± 0.20, 0.92 ± 0.20, 0.92 ± 0.20 h for the ZGI, DGE, and PA groups, respectively. The absorbance velocity of the ZGI, DGE, and PA was rapid; they reached C max between 0.92 and 2.00 h after single dose administration of ZGI, DGE, PA, and S. officinalis extract, but the first peak of PA appeared at 0.5 h and the second peak reached C max at 4.2 h after administration of S. officinalis extract, compared with PA group. The C max of ZGI, DGE, and PA in the S. officinalis extract was 1091 ± 1.1 × 10 2 , 359 ± 51, and 703 ± 96 ng/mL, respectively. The C max of ZGI, DGE, and PA of the monomer groups was 678 ± 47, 166 ± 21, and 262 ± 31 ng/mL. Moreover, the AUC 0→t of the three compounds was 2877 ± 1.1 × 10 2 , 1200 ± 76, and 3315 ± 89 ng·h/mL in the S. officinalis extract, and 1716 ± 3.1 × 10 2 , 394 ± 33, and 1021 ± 55 ng·h/mL from the ZGI, DGE, and PA groups, respectively. The AUC 0→∞ of the three compounds in the S. officinalis extract was 3098 ± 1.2 × 10 2 , 1314 ± 1.5 × 10 2 , and 4026 ± 2.1 × 10 2 ng·h/mL, and the AUC 0→∞ of the three compounds in the monomer groups was 2166 ± 3.3 × 10 2 , 551 ± 47, and 1183 ± 92 ng·h/mL. The increasing C max , AUC 0→t , and AUC 0→∞ showed the better absorption of the three compounds in the S. officinalis extract. The mechanism of the difference in pharmacokinetics between the monomer groups and S. officinalis extract remains unclear. It may be inferred that some ingredients in the S. officinalis extract may enhance the absorption of the three analytes. Furthermore, the t 1/2 values of ZGI, DGE, and PA were 9.3 ± 2.4, 7.3 ± 2.7, and 11 ± 1.4 h in the S. officinalis extract. The t 1/2 values of ZGI, DGE, and PA were 11 ± 0.95, 11 ± 0.79, and 8.8 ± 1.0 h in the monomer groups. It was revealed that ZGI and DGE in the S. officinalis extract were more easily eliminated and metabolized than in the monomer groups. In addition, from these analysis results, we could conclude that PA has a different absorption compared with ZGI and DGE. The t 1/2 of PA after consuming the extract is higher than that of the pure PA, suggesting that PA in extract needs a longer time for elimination. The T max of PA was at 4.2 h, which was longer than the other two compounds. However, the content of PA in S. officinalis was much lower than that of ZGI, yet the AUC 0→t and AUC 0→∞ of PA were greater than those of ZGI and DGE. The reason for these results was probably that ZGI and DGE metabolized PA in vivo.

Discussion
Few previous studies indicated that a method was developed to simultaneously determine ZGI, DGE, and PA from S. officinalis in the rat. Therefore, this research established a rapid, sensible, and sensitive UHPLC-MS/MS method for simultaneous determination and investigation of three compounds in rats after oral administration of S. officinalis extract. Meanwhile, the rats were given the three monomers; respectively, it was found that the three monomers have a transformational relationship and affect their pharmacokinetic parameters, to explore the process of the transformation of the three components in the rat.
For rapid, sensible, and simultaneous determination of three compounds, a UHPLC-MS/MS method was optimized. First, different proportions of mobile phase systems were tested, and ammonium acetate (5 mM) water was opted for, to enhance the peak strength of the three analytes. To realize a high extraction yield and weaken the matrix effect, LLE was chosen as the sample extraction method. As a result of the three compounds with similar polarities, ethyl acetate was the best option for extraction efficiency and repeatability. In addition, it is worth mentioning that to improve the sensitivity and accuracy, the mass spectrometer adopts multi-channel positive and negative ion mode to simultaneously detect a variety of compounds, which greatly optimizes the UHPLC-MS/MS method.
The elimination rate of the three compounds can be reflected from the t 1/2 . The t 1/2 of both ZGI and DGE monomers were lower than those of S. officinalis extract. The reason may be that the presence of ZGI and DGE in the extract may be converted into PA in vivo, increasing the amount of PA and reducing the elimination rate of PA. In contrast, the conversion of ZGI and DGE in S. officinalis extracts reduces the amount in the body, so the elimination rate of these two compounds was improved. From the T max , the absorption rates of the three compounds can be reflected. It is worth noting that the DGE in S. officinalis extract was absorbed for a longer time, but the absorption amount was higher than that of the pure compound. The interactions between compounds of herbal extracts in vivo may affect the absorption of target compounds [29] The reason for the conversion of ZGI to DGE in the extract leads to its increased amount, prolonged absorption time, and increased absorption.
In addition, according to the data in Tables 5-7, in rats after the oral administration of ZGI, DGE, and PA, it is worth noting that pharmacokinetic parameters of three compounds may have some connection. As shown in Figure 5, ZGI may be converted into DGE and PA after oral administration of ZGI, and DGE may be transformed into PA after oral administration of DGE. On the contrary, PA cannot convert into ZGI and DGE after oral administration of PA. The hemiacetal hydroxyls of sugar molecules are active, so the linkage to triterpenes is unstable. This was previously demonstrated by the chemical and gut metabolic characterization of saponins [30]. C-3 and C-28 of pentacyclic triterpenoid saponins are prone to deglycosylation in vivo and in vitro [31]. The pentacyclic saponins can produce deglycosylation reactions in electric fields, as well as in intestinal flora in vitro [32][33][34]. On the other hand, a similar reaction also occurs in saponins in vivo [35]. However, the mechanism for this phenomenon needs further study. As shown in Figures 3 and 4, the double peaks phenomenon was observed in the mean plasma concentration-time distribution curves of the three analytes during the elimination phase. The double peaks phenomenon of compounds may be due to the distribution of reabsorption and entero-hepatic circulation [36]. These results could be conducive to further exploring the mechanism of triterpenes, and provide effective pharmacokinetic information.
In this work, a rapid, sensible, and sensitive UHPLC-MS/MS method was established with the multi-ion mode, multi-channel simultaneous mensuration, and further applied to the pharmacokinetic studies of three compounds in rats after oral administration of S. officinalis extract. The pharmacokinetic and transformation characteristics of the three compounds in S. officinalis were initially revealed in vivo. However, the relevant mechanism of its specific metabolic behaviors needs further study. peaks phenomenon of compounds may be due to the distribution of reabsorption and entero-hepatic circulation [36]. These results could be conducive to further exploring the mechanism of triterpenes, and provide effective pharmacokinetic information.
In this work, a rapid, sensible, and sensitive UHPLC-MS/MS method was established with the multi-ion mode, multi-channel simultaneous mensuration, and further applied to the pharmacokinetic studies of three compounds in rats after oral administration of S. officinalis extract. The pharmacokinetic and transformation characteristics of the three compounds in S. officinalis were initially revealed in vivo. However, the relevant mechanism of its specific metabolic behaviors needs further study.

Chemicals and Reagents
The ZGI, DGE, and PA, with purity of more than 98%, were refined in our laboratory (identified by NMR and MS) [37]. Bifendate (Lot: 73536-69-3; purity > 98%, IS) was purchased from Chengdu Must Bio-Technology (China). Methanol (HPLC-grade) was obtained from J&K Medical (Beijing, China). Ammonium acetate was purchased from Kermel (Tianjin, China). Ultra-pure water was obtained using a Milli-Q water purification system (Millipore, Molsheim, France). All the other reagents, including ethyl acetate, ether, and dichloromethane, were of analytical grade. Sanguisorba officinalis L., which was purchased from the Anguo Traditional Chinese Medicine Market of Hebei, was appraised by Professor Zhenyue Wang of Heilongjiang University of Chinese Medicine.

Chemicals and Reagents
The ZGI, DGE, and PA, with purity of more than 98%, were refined in our laboratory (identified by NMR and MS) [37]. Bifendate (Lot: 73536-69-3; purity > 98%, IS) was purchased from Chengdu Must Bio-Technology (China). Methanol (HPLC-grade) was obtained from J&K Medical (Beijing, China). Ammonium acetate was purchased from Kermel (Tianjin, China). Ultra-pure water was obtained using a Milli-Q water purification system (Millipore, Molsheim, France). All the other reagents, including ethyl acetate, ether, and dichloromethane, were of analytical grade. Sanguisorba officinalis L., which was purchased from the Anguo Traditional Chinese Medicine Market of Hebei, was appraised by Professor Zhenyue Wang of Heilongjiang University of Chinese Medicine.

Animals
The experimental protocol was permitted by the Animal Ethics Committee of Harbin Medical University (Approval Code: AECHMU20150021, Approval Date: 3 June 2015) and conformed to the principles for the Care and Use of Laboratory Animals. Twentyfour male Sprague-Dawley rats (weighting 200 ± 20 g), which were provided by the Laboratory Animal Center at Harbin Medical University (Harbin, China), were used for the pharmacokinetic study. The animals were maintained in a SPF-grade room on a 12 h light/dark cycle with a controlled temperature conditions (21 ± 2 • C) and relative humidity (50 ± 5%) before the experiment. During the adaptation process, all rats were provided with standard food and purified water each morning and evening. Each rat fasted for 12 h before giving the drug and was randomly divided into one of four groups (n = 6).

Preparation of S. officinalis Extract
After smashing the dried root of S. officinalis, it was weighed to 100 g. The S. officinalis was extracted by reflux with 2000 mL of 70% ethanol (1:10, w/v) solution two times at 85 • C, 1 h each time, and was then filtrated. The combined filtrate was evaporated into steam, and the residue was dissolved in water to obtain the concentration of the S. officinalis extract equivalent to 0.025 g/mL.

UHPLC-MS/MS Analytical Conditions
The analytes were evaluated on an Agilent series 1290 UHPLC instrument combined with an Agilent Technologies 6430 mass spectrometer. The eluant was surveyed by applying a triple quadrupole tandem mass spectrometer equipped with an ESI source and operated in positive (ESI + : 0.0-5.0 min) and negative (ESI − : 5.0-6.5 min) ion modes with MRM. The transitions were m/z 784.5→437.4 for ZGI, m/z 652.5→455.4 for DGE, m/z 471.2→453.2 for PA, and m/z 418.9→342.8 for bifendate (IS). The mobile phases were as follows: solvent A was 5 mM ammonium acetate water, and solvent B was methanol, which was delivered at a flow rate of 0.3 mL/min. The gradient elution program was as follows: 0-4.5 min, 70 to 90% B; 4.5-6.0 min, 90% B; 6.0-6.3 min, 90-70% B; 6.3-6.5 min, 70% B. The sample injection volume was 10 µL.
The conditions for MS analysis were as below: drying gas (N 2 ) flow rate, 11 L/min; drying gas temperature, 300 • C; high purity nitrogen (N 2 ) was atomized as the nebulizing gas; capillary voltage, 4000 V. The mass parameters for the three analytes and IS are listed in Table 8. The chemical structure and product ion scan spectra of ZGI, DGE, PA, and IS are presented in Figure 6.

Preparation of Calibration and Quality Control (QC) Samples
Standard stock solutions of the three components were obtained from resolving each compound in methanol to obtain an ideal concentration (0.52 mg/mL for ZGI, 0.23 mg/mL for DGE, 0.21 mg/mL for PA). Standard solutions were prepared by compatible dilutions of the stock solutions with methanol (6.5-2600 ng/mL for ZGI, 5.8-2300 ng/mL for DGE, 2.6-1050 ng/mL for PA). The IS solution (520 ng/mL) was prepared by diluting stock solution of methanol. Calibration standards were prepared by spiking each working stock

Sample Preparation
The supernatant was separated by adding 10 µL IS (520 ng/mL) solution into 100 µL plasma sample, vortexing for 30 s, and mixing with 3 mL ethyl acetate by being vortexmixed for 60 s. After centrifuging, the supernatant was dried by N 2 blowing at 40 • C. After stirring for 4 min at 3800 rpm, the residue was reassembled with 100 µL methanol, then mixed by vortexing for 120 s and filtered by a 0.22 µm organic membrane. Finally, 10 µL sample solution was injected into the UHPLC-MS/MS system [38].

Selectivity
Selectivity evaluation was conducted by comparing chromatograms of blank plasma samples from six individual rats, corresponding blank plasma spiked with LLOQ of the three analytes and IS, and the plasma samples from the rats after oral administration of the three analytes and S. officinalis extract.

Linearity and LLOQ
The calibration curves were constructed by plotting the peak area ratio versus the concentration of the three analytes and IS with a weighted (1/x 2 ) least square linear regression, using standard plasma samples. The lower limit of quantification (LLOQ) was the lowest analytical concentration of the calibration curve, for which the signal-to-noise ratio was >10.

Precision and Accuracy
The intra-and inter-day precision and accuracy were determined by testing the LLOQ sample and QC samples at three concentration levels of ZGI, DGE, and PA in six replicates for three days in a row. Precision and accuracy were shown by the coefficient of variation (RSD) and relative error (RE), respectively. The RSD values should not exceed 15% for the QC samples, except for the LLOQ, which should not exceed 20%. The RE values should be within ±15% of the nominal values for the QC samples, except for the LLOQ, which should be within ±20% of the nominal value.

Matrix Effect and Extraction Recovery
The extraction recovery of analytes was determined by comparing the peak areas of the three analytes from the QC samples with those obtained from blank plasma samples with the three analytes spiked into the post-extraction supernatant at three QC levels in six replicates. The matrix effect was evaluated by comparing the peak areas of analytes spiked after plasma extraction into the post-extraction supernatants at three QC levels in six replicates. The extraction recovery and matrix effects of IS were also determined at one concentration. The RSD values should be within ±15% of the nominal values for the QC samples and IS, except for the LLOQ, which should be within ±20% of the nominal value.

Stability
The stability of the three compounds in rat plasma, including freeze and thaw stability (three freeze-thaw cycles at −20 • C), long-term stability (storage for 2 weeks at −80 • C), room temperature stability (storage for 4 h at ambient temperature), and post-preparation stability (storage for 12 h after sample preparation at 4 • C), was tested at two QC levels with six replicates at each level. All stability testing QC samples were determined using the calibration curve of freshly prepared standard samples.

Carryover Effect
The carryover effect was assessed by injecting a blank rat plasma sample after calibration of the standard sample at the upper limit of quantification. Carryover in a blank rat plasma sample should not be greater than 20% of response at the LLOQ for samples and 5% of response for IS.

Pharmacokinetic Study
A single dose of the S. officinalis extract (0.55 g/kg, group I), ZGI (0.03 g/kg, group II), DGE (0.007 g/kg, group III), and PA (0.013 g/kg, group IV) was administrated to the rats. The S. officinalis extract was dissolved in water. A single dose of 0.55 g/kg of S. officinalis extract was administrated to the rats. Blood was obtained from the retinal venous plexus at 0, 0.083, 0.5, 1.0, 1.5, 2.0, 2.5, 3.0, 4.0, 6.0, 8.0, 12.0, and 24.0 h after dosing. The plasma was immediately separated by centrifugation at 8000 rpm for 4 min at −4 • C.
The maximum concentration (C max ) and the time to attain it (T max ) were observed directly from the measured data. The elimination rate constant (K e ) was calculated by linear regression of the terminal points in a semi-log plot of the plasma concentration against time. The area under plasma concentration-time curve (AUC 0→t ) to the last measurable plasma concentration (C t ) was estimated using the linear trapezoidal rule.

Data Analysis
Calculate the amount of three compounds using linear regression from the standard curve. The pharmacokinetic parameters of the analytes were calculated with DAS 2.0 (Shanghai, China) [40]. All results were expressed as mean ± SD. The comparison of pharmacokinetic parameters was conducted using standard Student's t-test. Differences between groups were assumed statistically significant for p values < 0.05.

Conclusions
This study demonstrated that a simple, rapid, and sensitive LC-MS/MS method was successfully developed for simultaneous quantification of the three representative components from ZGI, DGE, PA, and S. officinalis in rat plasma. This is the first report of a pharmacokinetic study of these three triterpenes together in vivo following the oral administration of ZGI, DGE, PA, and S. officinalis extract. This paper may be useful for further studies on the metabolism and absorption process of S. officinalis extract in vivo, and may also be beneficial for the application of this TCM in clinical therapy.

Conflicts of Interest:
The authors declare no conflict of interest.