Quantification of Nineteen Bioactive Components in the Ancient Classical Chinese Medicine Formula of Wen-Dan Decoction and Its Commercial Preparations by UHPLC-QQQ-MS/MS

A UHPLC-QQQ-MS/MS method was developed to quantify the significant constituents in Wen-Dan Decoction (WDD), a traditional Chinese medicine. Analysis of 19 compounds was conducted on an ACQUITY UPLC® BEH C18 Column (2.1 × 50 mm, 1.7 μm) using elution with a gradient elution of acetonitrile and 0.05% (v/v) formic acid in water. A triple quadrupole mass spectrometer was operated in negative ionization mode and positive ionization mode by multiple reaction monitoring (MRM), respectively. All calibration curves showed acceptable linearity (r ≥ 0.9950). The RSDs of intra- and inter-day precisions of low, mid and high concentrations were ≤ 8.88%. The repeatabilities (RSDs ≤ 7.17%) and stabilities (RSD ≤ 4.79%) of the samples were qualified. The recoveries were found in the range of 93.07 ± 3.86 to 103.98 ± 2.98% with the RSD varying between 1.30 and 7.86%. The final rapid, sensitive, precise, accurate and reliable UHPLC-QQQ-MS/MS method was used for the simultaneous quantification of 19 constituents in WDD and its commercial preparations. The strategy of combining the contents of the 19 chemicals in a daily dose of the WDD preparations with the hierarchical cluster analysis and the 3D principal component analysis was employed to effectively distinguish the WDD preparations provided by the different suppliers, which represents a contribution to the evaluation and control of the quality of WDD (or other decoctions consisting of the same herbs) and the preparations of WDD in other dosage forms such as tablets and granules.


Optimization of LC Conditions
A gradient elution chromatographic procedure was used to achieve chromatographic separation of the 19 components in a short period of time. An ACQUITY UPLC ® BEH C18 column (1.7 μm, 2.1×50 mm) was employed for better peak symmetries than others, and acetonitrile was applied for its better chromatographic separation than that of methanol. The optimized column temperature of 35 °C and flow rate of 0.3 mL·min −1 were also contributed to the efficient separation. The 0.05% formic acidwater solution provided better peak shapes. After these optimization procedures, all reference compounds could be generally and chromatographically separated within 30 minutes.

Optimization of MS Conditions
Studies on the individual mass spectra of the 19 compounds were performed by injecting the corresponding reference solutions into the mass spectrometer in both positive and negative ion modes. The results of the pilot study showed that the product ion peaks of succinate, liquiritin, eriocitrin, rutin, narirutin, naringin, hesperidin, neohesperidin, liquiritigenin, isoliquiritin, didymin, poncirin and dehydropachymic acid performed well in negative ion modes. Meanwhile, synephrine, 6-gingerol, tangeretin, 8-gingerol, 10-gingerol and pachymic acid gave better product ion results in positive ions mode. The fragmentor voltage (FV, 50-350V) and collision energy (CE, 0-50V) of all ingredients were optimized for greater abundances of precursor and product ion on the mass spectrometer, and the retention times were also determined by the reference solutions. The negative

Optimization of LC Conditions
A gradient elution chromatographic procedure was used to achieve chromatographic separation of the 19 components in a short period of time. An ACQUITY UPLC ® BEH C18 column (1.7 µm, 2.1 × 50 mm) was employed for better peak symmetries than others, and acetonitrile was applied for its better chromatographic separation than that of methanol. The optimized column temperature of 35 • C and flow rate of 0.3 mL·min −1 were also contributed to the efficient separation. The 0.05% formic acid-water solution provided better peak shapes. After these optimization procedures, all reference compounds could be generally and chromatographically separated within 30 min.

Method Validation
The LODs and the LOQs of each compound showed acceptable sensitivities for the assays. The regression equations were constructed by comparing peak areas (Y) versus the concentrations (X) to present the linearity and the r ≥ 0.9950 demonstrated acceptable correlation coefficients for the calibration curves. The RSDs of intra-day precision (≤7.85%) and inter-day precision (≤8.88%) at three concentration levels (low, mid and high) indicated an acceptable precision of the method. The repeatabilities were found with RSDs of ≤7.17%. The components of the sample were stable (RSDs ≤ 4.79%) for 24 h at 4 • C in the autosampler after preparation, which is the acceptable stability during the testing process. The recoveries varied in the range from 93.07 ± 3.86 to 103.98 ± 2.98% with the RSDs ≤ 7.86%. All the results of the method validation mentioned above are summarized in Tables 2 and 3.

Quantification and Analysis of 19 Compounds in WDD and Its Commercial Preparations
Since the contents of the 19 components covered an extensive range, it was convenient for the establishment of the standard curves to dilute the WDD samples to different appropriate concentrations. Twelve replicates of WDD extracted in the lab and its five brands of commercial preparations were diluted according to the method mentioned in Section 4.4 and determined for the contents of 19 components by the validated UHPLC-QQQ-MS/MS method in 30 min. The assay efficiency had significant improvements over the HPLC-based investigation, ensuring that we could test more samples in a day.
The range, the mean and standard deviation of the 19 chemical contents in a daily dose of WDD prepared in the lab and its five brands commercial preparations are shown in Table 4 and the raw data of the chemical contents in a daily dose of WDD prepared in the lab and its five brands commercial preparations are shown in Supplementary Table S1. Among the investigated substances in the 12 batches of WDD, 6-gingerol was the most abundant contents in a daily dose for the most proportion of Zingiberis Rhizoma (15.5 g). The two triterpenoids pachymic acid and dehydropachymic acid were not detected in the WDD samples prepared in the lab. Three flavonoids (liquiritin, isoliquiritin and liquiritigenin) in glycyrrhizae radix et rhizoma provided the larger content RSDs in the 12 batches WDD. The contents of 19 components in a daily dose of the 12 batches of WDD extracted in the lab and five brands WDD commercial preparations were normalized by z-score method. The normalized results were expressed in a heatmap and analyzed by hierarchical cluster analysis ( Figure 4A) and 3D principal component analysis ( Figure 4B Dehydropachymic acid 4.98 * 2.73 * 1.77 ** 4.10 ** 7.85 *** 2.73 *** * Low concentration; ** Medium concentration; *** High concentration.

Quantification and Analysis of 19 Compounds in WDD and Its commercial preparations
Since the contents of the 19 components covered an extensive range, it was convenient for the establishment of the standard curves to dilute the WDD samples to different appropriate concentrations. Twelve replicates of WDD extracted in the lab and its five brands of commercial preparations were diluted according to the method mentioned in Section 4.4 and determined for the contents of 19 components by the validated UHPLC-QQQ-MS/MS method in 30 min. The assay efficiency had significant improvements over the HPLC-based investigation, ensuring that we could test more samples in a day.
The range, the mean and standard deviation of the 19 chemical contents in a daily dose of WDD prepared in the lab and its five brands commercial preparations are shown in Table 4 and the raw data of the chemical contents in a daily dose of WDD prepared in the lab and its five brands commercial preparations are shown in Supplementary Table S1. Among the investigated substances in the 12 batches of WDD, 6-gingerol was the most abundant contents in a daily dose for the most proportion of Zingiberis Rhizoma (15.5g). The two triterpenoids pachymic acid and dehydropachymic acid were not detected in the WDD samples prepared in the lab. Three flavonoids (liquiritin, isoliquiritin and liquiritigenin) in glycyrrhizae radix et rhizoma provided the larger content RSDs in the 12 batches WDD. The contents of 19 components in a daily dose of the 12 batches of WDD extracted in the lab and five brands WDD commercial preparations were normalized by zscore method. The normalized results were expressed in a heatmap and analyzed by hierarchical cluster analysis ( Figure 4A) and 3D principal component analysis ( Figure 4B    As shown in Figure 4A, the WDD prepared in the lab (A-1-A-12) and WDD commercial preparations (B-1-B-3; C-1-C-3; D-1-D-3; E-1-E-3; F-1-F-3) were classified into two clusters. The five commercial WDD formulations are then grouped into five different second-layer clusters, which matched their brands. Additionally, the two kinds of commercial WDD formulations from Kaiser Pharmaceutical Co., Ltd (concentrated particles of D-1-D-3 and concentrated ingots of E-1-E-3) were in one sort of cluster. As shown in Figure 4B, the contribution of PC1-PC3 was 89.1%. According to PCA analysis, the samples are classified into six groups according to their origins, which matched the sources of the WDD preparations.

Selection of Indicators and Preparation of Solutions
Nineteen indicators, including one organic acid, one alkaloid, two triterpenoids, three phenolics and 12 flavonoids, were selected by reviewing previous reports and the Chinese Pharmacopeia. From the negative and positive scan chromatograms of the WDD sample (Figure 2A,B), it can be concluded that the selected 19 chemicals generally cover the main ingredients in WDD.

Quantification and Analysis of 19 Compounds in WDD and Its Commercial Preparations
Among the investigated substances in the 12 batches of WDD, 6-gingerol was the most abundant component in a daily dose due to the larger proportion of Zingiberis Rhizoma (15.5 g) in the preparation. 6-Gingerol can reduce the level of dopamine in vivo to inhibit nerve activity [17], which can be employed to treat insomnia. With regard to 6-gingerol, although the daily dose is the largest among the 19 ingredients, we cannot ignore the potential loss of 6-gingerol during the process of extracting WDD in the lab. It has been reported that heating is adverse for the stability of 6-gingerol, which will rapidly be converted into the dehydration product shogaol at high temperature [18][19][20]. Moreover, organic solvents such as methanol, ethanol, isopropyl alcohol or n-hexane were generally used for better solubility of gingerol than water [16,21,22]. However, during the process of extracting WDD in the traditional way, the solvent (water) and boiling temperature (100 • C) are not conducive to the stability of gingerols. In addition, the extraction using a traditional decoction pot led to the loss of gingerols with the water vapor [23][24][25]. Additionally, compared with the GC-MS which requires heating samples, the UHPLC-MS is more beneficial for the stability of the gingerols in this study [16].
Two triterpenoids of pachymic acid and dehydropachymic acid were not detected in the WDD samples prepared in the lab. The low percentages of pachymic acid and dehydropachymic acid in Poria and their poor water-solubility partly accounted for their absence in the WDD samples prepared in the lab. Furthermore, their loss by the release of water vapor could reduce their contents in the WDD solution.
Three flavonoids (liquiritin, isoliquiritin and liquiritigenin) in glycyrrhizae radix et rhizoma provided the larger content RSDs in the 12 batches WDD. Fick's first law of diffusion (1) was utilized to explore the sources of the large differences in the content RSDs of these three flavonoids in glycyrrhizae radix et rhizoma in the 12 batches WDD prepared in the lab [26,27]: where, dt is diffusion time (s). ds is the amount of diffused compound in dt time (mol). F is the diffusion area and represents the size and surface state of the herbs (m 2 ). dc is the concentration difference of the compounds between the herbs and the solution (mol/m 3 ). dx is the characteristic length scale of the diffusion system (m). D is the diffusion coefficient (m 2 /s). "-" indicates a decrease of concentration difference when the diffusion tends to equilibrium. According to Equation (1), the differences in the diffusion area and the concentration gradient are the main factors causing the variances in the content of ingredients in different batches of traditional Chinese medicine decoctions. In one dosage of the decoction, the surface area of the materials differs due to its size (shown in Supplementary Figure  S4). Meanwhile, some reports have pointed out that growing region, harvest time, germplasm line, growing years and drying process gave rise to content variations of components between individual herbs [28][29][30][31][32]. Even in a single herb, significant differences in the amount of ingredients were discovered in different parts, such as the main root, branch root and fibrous root [33]. All of the above factors might give an explanation for the significant content difference between batches when the only 1.8 g of uneven glycyrrhizae radix et rhizoma was used in one dosage. The herb materials with different content of compounds also gave rise to the various content of ingredients among the batches of WDD. The hierarchical cluster analysis results of the contents of 19 components in one daily dose in the 12 batches of WDD extracted in the lab almost matched the four decoction pots we used in the extraction process (Supplementary Figure S3), which suggest that even if we use the different types of equipment of the same type from the same manufacturer to extract the WDD, they would still lead to the variations in contents of the ingredients in the decoction procedures [34].
In the hierarchical cluster analysis and the 3D principal component analysis, it was found that the WDD granules ( Figure 4D-1-D-3) and WDD ingots ( Figure 4E-1-E-3) both acquired from Kaiser Pharmaceutical Co., Ltd. were divided into the one second-level cluster, which indicated their homology in regards to the materials and the extraction process. However, they were distinguished into two different first-level clusters instead of being mixed together in one, which implied that though the same WDD related preparations from one manufacturer, the daily dose of the ingredients might be distinct depending on the dosage form, which might lead to different therapeutic effects.
In general, the difference in a daily dose of the chemicals between the WDD prepared in the lab and WDD commercial preparations may be caused by the discrepancy of the ingredient contents in the herbs, the shape of the herbs, the size of the herbs, the divergence in the extraction method and equipment, the composition of the prescription (Supplementary Table S2) and the dosage form. Therefore, it is essential to assay the contents of the ingredients in herbal extraction related preparations before confirming the dosage quantity of herbal medicine for better treatment effects [16,35]. The uniform contents of compounds are apparently more conducive to the clinical application of traditional Chinese medicine preparations. In order to obtain batches of uniform contents decoctions or its related preparations, the proportions of various herbs, the contents of the ingredients, the sizes of the herbs and the extraction equipment should be controlled.

Recommendation on Preclinical Research and Clinical Application of WDD
In Section 3.1, it is pointed out that the 19 determined components have been originated from seven pieces of traditional Chinese medicine. Supplementary Table S3 displays the "total daily dosages of measured ingredient" for the seven pieces of traditional Chinese medicine, which is defined as the sum of the average daily dosages of all the measured ingredients tracking back to a certain piece of traditional Chinese medicine. For example, the daily dosage of Zingiberis Rhizoma is expressed as the sum of the mean daily dosages of 6-gingerol, 8-gingerol and 10-gingerol. The z-score normalized data of the "total daily dosage of measured ingredient" of the seven pieces of traditional Chinese medicine is represented by a radar chart showing seven pieces of traditional Chinese medicine from six different WDD preparations. These advantages and disadvantages can further provide a stroma for the selection of the source of WDD preparation, according to the patient condition in the clinical practice under the guidance of traditional Chinese medicine theory. For example, in traditional Chinese medicine theory, Zingiberis Rhizoma has a role in treating vomiting [1]. If the relevant conditions of insomnia and asthenia accompanied by vomiting are to be treated clinically, the WDD prepared in the lab ( Figure 5A) can be given priority in the six WDD preparations involved in this study. owning a required pharmacological activity and a superior average daily dosage in the six WDD preparations. For example, it is shown in Figure 6 that WDD prepared in the lab ( Figure 6A) possesses the superiorities in the average daily dosage of 6-gingerol and 8-gingerol. As it is reported that the gingerols have the pharmacological activity in the treatment of vomiting [36], when it comes to the preclinical research of the antiemetic effect of WDD or the clinical application for the treatment of vomiting associated with insomnia, WDD prepared in the lab could be a better choice.  Furthermore, the average chemicals contents in a daily dose of WDD prepared in the lab and its five brands commercial preparations are shown in Table 4. The z-score normalized data of the average chemicals contents in a daily dose of the above WDD preparations is displayed in another radar chart ( Figure 6), which can be used to preliminarily characterize the advantages and disadvantages of 19 chemicals in six different WDD preparations. These advantages and disadvantages provide a basis for the selection of the source of WDD preparation for preclinical research and clinical application based on the modern pharmacological research theory. Generally, before a single component of the 19 components of the measured WDD exerts its pharmacological activity, the concentration of component should reach an effective level, which is closely related to the daily dose of the component. It can contribute to the preclinical research and the clinical application by selecting the source of the WDD preparation based on selecting a certain chemical owning a required pharmacological activity and a superior average daily dosage in the six WDD preparations. For example, it is shown in Figure 6 that WDD prepared in the lab ( Figure 6A) possesses the superiorities in the average daily dosage of 6-gingerol and 8-gingerol. As it is reported that the gingerols have the pharmacological activity in the treatment of vomiting [36], when it comes to the preclinical research of the antiemetic effect of WDD or the clinical application for the treatment of vomiting associated with insomnia, WDD prepared in the lab could be a better choice.   Table 4.

Reagents and Materials
Reference   Table 4.

UHPLC-QQQ-MS/MS Conditions
The quantitative analysis was carried out on an Agilent 1290 series UHPLC-triple quadrupole mass spectrometry (UHPLC-QQQ-MS/MS) system (Agilent Technologies, Palo Alto, CA, USA). The chromatographic separation was performed on an ACQUITY UPLC ® BEH C18 column (1.7 µm, 2.1 × 50 mm, Waters, Milford, MA, USA) and the column temperature was maintained at 35 • C by a column thermostat. The autosampler temperature was controlled at 4 • C and the sample injection volume was 2 µL. A mixture of 0.05% (v/v) formic acid water (A) and acetonitrile (B) was used as the mobile phase, which was driven by a binary pump at a flow rate of 0.3 mL·min −1 . The gradient elution program was as follows: 0-5 min, 10-13%B; 5-11 min, 13-17%B; 11-14 min, 17-19% B; 14-20 min, 19-47% B; 20-22 min, 47-67% B; 22-26 min, 67-80% B; 26-30 min, 80-90% B and the post time was set as 3 min. The UHPLC system was coupled to an Agilent 6460 triple quadrupole mass spectrometer (Agilent Technologies) equipped with an ESI source. Both negative ionization mode and positive ionization mode were used for the analysis and determination. The MS conditions were set as follows: capillary voltage at 3500 V; gas (N 2 ) temperature at 350 • C; drying gas flow rate at 10 L·min −1 ; nebulizer gas (N 2 ) pressure at 35psi; high purity nitrogen as the collision gas. The multiple reaction monitoring (MRM) was used for quantification. The ionization mode, fragmentor voltage, collision energy and the pair of precursor-product ions for each compound were optimized in the following study. A MassHunter Workstation software (Version B.07.00, Agilent Technologies) was employed for the LC-MS data acquisition and analysis. The acceptable accuracy variation between the measured concentrations and the actual value was set within 20%.

Preparation of Samples Solutions
Pinelliae Rhizoma (3.7 g), Bambusae Caulis in Taenias (3.7 g), Aurantii Fructus Immaturus (3.7 g), Citri Reticulatae Pericarpium (5.4 g), Glycyrrhizae Radix Et Rhizoma (1.8 g), Poria (2.7 g), Zingiberis Rhizoma (15.5 g) and Jujubae Fructus (2.5 g) were put in a decoction pot (2L, Kangyashun Electric Co., Ltd., Guangdong, China) and immersed in 300 mL deionized water for 30 min. The formulation was boiled at 220 V and then evaporated to 150 mL at 175 V. The remaining decoction was filtered with two layers of gauze to obtain one dose of WDD. Appropriate volume of WDD was mixed with the same volume of methanol by sonication (320 W, 40 kHz) with a KQ-400KDE ultrasonic cleaner (Kunshan Ultrasonic Instruments Co., Ltd., Jiangsu, China) for 10 min. Since it was assayed that 25 mL WDD we extracted in the lab included about 0.6685 g of dry-powder, the five brands of WDD commercial preparations were ground into uniform fine powders (composed of some excipients and the extracted dry-powder) and then different weights of the uniform fine powders (containing about 0.5 g extracted dry-powder) were dissolved in 25 mL deionized water, respectively, and then they were mixed with the same volume of methanol by sonication (320 W, 40 kHz) with the KQ-400KDE ultrasonic cleaner for 10 min.

Method Validation
The 19 stock solutions were precisely mixed and further diluted with 50% (v/v) methanol-water to prepare the calibration curves owing different concentration ranges. All calibration curves were established by the ratio of standard reference peak areas (Y) versus their concentrations (X) with the weight of 1/X.
Limit of detection (LOD, signal-to-noise values (S/N) ≥3) and limit of quantification (LOQ, signal-to-noise values (S/N) ≥10) of the 19 chemicals were obtained by determining continuously diluted standard mixture to assess the sensitivity. The stabilities of the compositions were detected by assaying the prepared samples stored at 4 • C for 0, 2, 4, 6, 8, 10, 12 and 24 h. Intra-day (1 day) and inter-day (3 consecutive days) precisions at low, mid and high concentration levels within the calibration curves were examined. Replicates (n = 6 for diluted 4 times and 40 times, respectively) of WDD samples were prepared as the method in Section 2.4 and determined to demonstrate the repeatability. Relative standard deviation (RSD) was employed to evaluate the stabilities, precisions, and repeatabilities.
The recovery was surveyed to evaluate the method accuracy. The spiked samples were prepared by mixing known amounts of the 19 standard references with the known amounts of quantitatively analyzed WDD samples in sextuplicate, and then diluted and analyzed with the same procedures.

Statistics Analysis
The data were presented as "mean ± S.D.". The heatmap-dendrogram application from OriginPro 2018C (v9.5.1) software was employed for the hierarchical cluster analysis and the demonstration of the daily dose of 19 compounds in WDD samples and its commercial preparations. The 3D principal component analysis application from OriginPro 2018C software was used for the 3D principal component analysis of 19 compounds in WDD samples and its commercial preparations. The radar chart application from OriginPro 2018C software was utilized for the comparisons of the total daily dosage of the determined ingredients in the respective traditional Chinese medicines and the daily dosage of each the 19 ingredients in the six WDD preparations.

Conclusions
In this study, a validated and reliable UHPLC-QQQ-MS/MS method was developed for the simultaneous quantification of 19 compounds in WDD and its commercial preparations, including one organic acid, one alkaloid, two triterpenoids, three phenols and 12 flavonoids. In addition, the proportions of various herbs, the contents of the ingredients in the herbs, the Fick's first law of diffusion, the differences between the extraction equipment used and the dosage form were utilized explain for the content variations among batches of traditional Chinese herb preparations and its marketed preparations. Furthermore, it can be used to effectively distinguish the sources of the preparations with combining the contents of the 19 components in WDD in a daily dose with the hierarchical cluster analysis and 3D principal component analysis. Simultaneously, the developed method could contribute to the evaluation and control the quality of WDD (or other decoctions consisting of the same herbs) and the preparations of WDD in other dosage forms such as tablets and granules. The radar charts of the "total daily dosage of measured ingredients" of the seven pieces of traditional Chinese medicine and the 19 chemicals contents in a daily dosage would do some help to the development of the researches on preclinical research and clinical application of WDD.
Supplementary Materials: The following are available online. Figure S1: MRM chromatogram of 19 standard chemicals. Figure S2: The characteristic product ion maps of the 19 chemicals. Figure S3: The four decoction pots in the WDD extraction process. Figure S4: Pinelliae Rhizoma (A), Bambusae Caulis in Taenias (B), Aurantii Fructus Immaturus (C), Citri Reticulatae Pericarpium (D), Glycyrrhizae Radix Et Rhizoma (E), Poria (F), Zingiberis Rhizoma (G) and Jujubae Fructus (H). Table S1: Chemicals contents in a daily dose of WDD prepared in the lab and its five brands commercial preparations (µg). Table S2: The proportion of the herbs and the excipiet in a daily dose of the WDD preparations. Table S3: The "total daily dosages of measured ingredient" for the 7 seven pieces of traditional Chinese medicine.