Qualitative and Quantitative Analysis of Phenolic Acids, Flavonoids and Iridoid Glycosides in Yinhua Kanggan Tablet by UPLC-QqQ-MS/MS

A simple, rapid and specific ultra-performance liquid chromatography-triple quadrupole mass spectrometry method was developed for the analysis of 29 bioactive components (10 phenolic acids, 16 flavonoids, and three iridoid glycosides) in Yinhua Kanggan tablet (YHKGT), a herbal prescription used for treating upper respiratory infections, fevers, coughs and pharyngalgia. The separation was successfully achieved using a Waters Cortecs UPLC C18 column (50 × 2.1 mm, 1.6 μm) and gradient elution with water-0.1% formic acid and acetonitrile. Polarity switching mode was used in the optimization of multiple reaction monitoring conditions. The analytical method was validated for linearity, precision and accuracy. Calibration curves for the 29 marker compounds showed good linear regression (r > 0.9982). The limits of detection (LOD) and limits of quantification (LOQ) for the 29 analytes were in the range of 0.03–4.99 ng/mL and 0.16–14.87 ng/mL, respectively. The relative standard deviation (RSD) values of intra-day precision, inter-day precision, repeatability, and stability were less than 2.79%, 4.87%, 4.18% and 4.71%, respectively. The recoveries of the 29 marker compounds were in the range of 94.67%–104.78% (RSD ≤ 4.72%). These results have shown that this developed method was efficient for the quality evaluation of YHKGT.

Although herbal medicines are increasingly being understood and accepted by more and more people around the world, the problem of quality control remains one of the major obstacles for their internationalization. A published paper has described a high performance liquid chromatographydiode-array detector (HPLC-DAD) method for the determination of the content of chlorogenic acid, one of the major components in YHKGT [13], but the efficacy of YHKGT should be associated with the synergistic or interactive action of various types of compounds derived from its component herbs rather than only one of them. Actually, the present quality control method severely restricts the clinical applications and in-depth study of YHKGT, therefore, it is of great significance to develop a more sensitive and efficient analytical method for the determination of more bioactive components in YHKGT for its quality assurance. Moreover, the aforementioned three types of active compounds of YHKGT should be selected for the quality control analysis.
In the present study, we have developed and validated for the first time a polarity switching ultra-performance liquid chromatography coupled with triple quadrupole mass spectrometry (UPLC-QqQ-MS) method for the rapid simultaneous determination of 29 active components (10 phenolic acids, 16 flavonoids and three iridoid glycosides) in YHKGT. Thirteen batches of YHKGT were collected for the analysis. Additionally, to ensure the accuracy and the sensitivity of quantification, 2-hydroxycinnamic acid, liquiritin, and albiflorin were employed as internal standards for phenolic acids, flavonoids, and iridoids, respectively.

Optimization of Sample Preparation
In order to achieve optimal extraction efficiency, the variables involved in the extraction, such as extract solvent and extract method, were optimized. Due to the different polarity and water-solubility of some analytes, 70% aqueous methanol was chosen as the extraction solvent. To find the best extraction method, ultrasonic extraction, refluxing and Soxhlet extraction were selected because of their relatively shorter extraction time than percolation and maceration. The results suggested that ultrasonic extraction was simpler and more efficient than refluxing and Soxhlet extraction in extracting typical compounds 1-29 using 70% aqueous methanol as extraction solvent ( Figure S1). Moreover, to obtain the optimal extraction efficiency, the effects of different factors including the different concentrations of methanol-water solution (30%, 50%, 70% and 90%), solvent volume (10, 50, 100 and 150 times) and extraction time (15,30,45 and 60 min) on the extraction performance were evaluated. An almost equal amount of sample (0.2 g) was extracted and analyzed using the described procedure. As a result, compared with the extraction yields of different factors of the extraction solution, volume and extraction time for 29 typical compounds, it was found that ultrasonic extraction with 100 times the volume of 70% methanol-water for 30 min for one time was an optimum method to prepare the sample solution ( Figures S2-S4).

Optimization of Chromatographic Conditions
The chromatographic conditions were optimized to improve the resolution and sensitivity and shorten the analysis time. Different mobile phases including methanol-water and acetonitrile-water were examined. Acetonitrile-water was found to produce better peak shapes than methanol-water. Interestingly, using 5% methanol to water could increase the resolution of several isomers such as schaftoside and isoschaftoside. Moreover, it was found that formic acid was not only beneficial to improving the chromatographic separation, but also in improving the ionization efficiency of analytes. In addition, due to their similar structures, retention time and ionization response in the negative ion mode, 2-hydroxycinnamic acid, liquiritin and albiflorin were chosen as internal standards for phenolic acids, flavonoids and iridoids, respectively.
In order to develop a sensitive and accurate quantitative method, the MS/MS fragmentation for each analyte was investigated by direct infusion of the single standard solution into the mass spectrometer (−)-ESI and (+)-ESI source to optimize MS parameters, including product ion, cone voltage, and collision energy, the product ion of each analytes. The negative ion mode was found to be more suitable for flavonoids and iridoid glycosides analyses, while positive ion mode was found to be more suitable for phenolic acids. Therefore, ion polarity switching mode was used in the optimization of MRM conditions in the quantitative analysis. The optimum results are shown in Table 1 and the MRM chromatograms of the 29 markers are shown in Figure 1A. LC/MS chromatograms of analytes from a real sample is presented in Figure 1B.

Method Validation
(a) Linearity. The calibration curves, plotted with at least six concentrations of standard solutions, were constructed from the peak areas ratios of each standard to IS vs. concentration of each analyte. Acceptable linear correlation at these conditions was confirmed by correlation coefficients (r, 0.9987-0.9998) (Figure 3). (b) LOD and LOQ. Limits of detection (LODs) and quantification (LOQs) are three times and ten times the noise level, respectively. For each target compound, the LODs and LOQs were determined by serial dilution of standard solution under the described UPLC-QqQ MS conditions. The LODs (S/N = 3) and LOQs (S/N = 10) for all standard analytes were in the range of 0.03-4.99 and 0. 16-14.87 ng/mL, respectively, indicating that this method is sensitive for the quantitative determination of major components in YHKGT samples (Table S2, Supplementary Information).
(c) Precision. Intra-and inter-day variations were chosen to determine the precision from standard solutions the developed method. For intra-day precision test, the standards solutions were analyzed for six replicates within 1 day, while for inter-day precision test, the solutions were examined in duplicates for consecutive 3 days. The RSD values of intra-and inter-day precision were in the range of 0.84%-2.79% and 1.07%-4.87%, respectively.
To confirm the precision from real samples (YHKGT), six samples of YHKGT (No. 1301014) were extracted and analyzed on three separate days. The RSD values of 29 standards were within the range of 2.03%-4.18%. In order to investigate the stability of the samples, each sample solution was analyzed within 24 h (0, 8, 12 and 24 h) at room temperature. The RSD values of the 29 analytes were all less than 4.71% within 24 h.
(d) Stability. Meanwhile, the stability of the standards was also investigated at 25 °C standards were analyzed every 4 h within 12 h in triplicate. In conclusion, this developed method had good precision, repeatability and stability ( Table 2).  (e) Accuracy. The recovery was used to evaluate the accuracy of the method and determine by adding the mixed standard solutions with three different concentration levels (low, medium and high) to the known amounts of YHKGT sample. Triplicate experiments were conducted at each level. The percentage recoveries were calculated according to the following equation: (detected amount − original amount) × 100%/spiked amount. As shown in Table 3, the recovery rate of 29 standards varied from 94.67 to 104.78% (RSDs ≤ 4.72%), revealing the acceptable recovery and accuracy of this method. Table 3. Recovery data of the proposed method (n = 3).

Sample Analysis
The validated method was successfully applied for the identification and quantification of 29 target compounds in 13 batches of YHKGT. The contents of the investigated compounds, based on their respective calibration curves, are summarized in Table 4. There were great variations among the contents of 24 compounds in different batches of YHKGT with RSD values exceeding 10%. Among them, chlorogenic acid, isochlorogenic acid A and isochlorogenic acid C, sweroside and agnuside, and schaftoside and rutin were the major phenolic acids, iridoids and flavonoids respectively, which have been found to have antiviral, anti-inflammatory and immunomodulatory properties [27][28][29][30]. Therefore, these three types of compounds should be considered as important bioactive components of YHKGT, and their content variabilities could influence the quality and efficacy of YHKGT.

Preparation of Standard Solution and Samples
Stock solutions of the 29 standards (approx. 1 mg/mL) were prepared individually by dissolving accurately weighted amount of standards in 70% methanol-water. An internal standards stock solution was also prepared in a concentration of 0.32 μg/mL for liquiritin, 0.25 μg/mL for 2-hydroxycinnamic acid and 0.81 μg/mL for albiflorin. Then a mixed solution containing all the 29 standards were prepared and serially diluted with 70% methanol-water (v/v) to obtain seven reference solutions with different concentrations used for plotting standard curves. All prepared solutions were stored at 4 °C before analysis. The 13 batches of YHKGT samples were ground to a fine powder. A powder sample (0.20 g) was accurately weighted and extracted with 100 mL of 70% methanol-water (v/v) in an ultrasonic bath (40 kHz, 500 w) for 30 min. Additional 70% methanol-water was added to make up the lost weight. The extracted solution was centrifuged at 150,000 rpm for 10 min, and the supernatant was obtained as sample solution. The internal standard working solution (500 μL) was added to 500 μL of the mixed standards solution or sample solution, then vortex blended and filtered through a 0.22 μm micropore membrane prior to injection. All the samples were stored at 4 °C before analysis.

Mass Spectrometry
Tandem mass spectrometry was performed on an Xevo TQ QqQ MS triple quadrupole mass spectrometer equipped with an electrospray ion source (ESI) (Waters). The MS spectra were acquired in multiple reaction monitoring (MRM) mode. Polarity switching electrospray ionization was applied. Argon was chosen as collision gas, nitrogen was chosen as nebulizer gas and heater gas. The MS conditions were optimized as follows: capillary voltage, 2.5 kV; source temperature, 200 °C; dwell time, 20 ms.

Conclusions
In this study, a UPLC-QqQ-MS method for the simultaneous determination of 29 major components in YHKGT has been developed and validated for the first time, which greatly improved its quality control. The polarity switching mode facilitated the detection of multiple types of constituents in YHKGT with different ionization responses. Compared with the current published HPLC method [13], this developed method enabled identification of target compounds with high selectivity by comparison with standards, high sensitivity and a rapid analysis was performed within 12 min. The results obtained in this work demonstrated that polarity switching in UPLC-QqQ-MS provides an unsuspected advantage for complex method development in the TCM analytical services industry.