The use of detection techniques for herbal medicines or phytomedicines has been increasingly studied in recent years. While therapeutic effects of medicinal plants have been discovered and utilized in the past, new potential compounds, ranging from medicinal herbs to new drugs, have been discovered and screened. To investigate the candidate compounds and the efficacy, safety, and quality of medicinal plants that contain hundreds of chemical constituents, separation and detection methods that enable rapid analysis of inherently complex herbs have been developed.
Various detection techniques, including ultraviolet detection (UV), fluorescence detection (FD), photodiode array detection (DAD), and mass spectrometry (MS), can offer excellent selectivity and sensitivity for known or unidentified structures in natural compounds. Liquid chromatography, coupled with mass spectrometry, is the most prevalent because it can collect data instantly from each chromatographic peak, even at low sample concentrations and with relatively short analysis times. In recent years, studies have shown that versatile high performance liquid chromatography (HPLC) coupled with mass spectrometry can be used to analyze complex natural products and metabolites. HPLC in combination with two or more MS experiments (tandem mass spectrometry, MS/MS) provides higher resolution to achieve better quantitative analysis and has become a routine analytical technique for the quantitative identification of herbal medicines.
Quantification analysis of bioactive compounds by LC-MS/MS has been applied to combined herbal medicines. For instance, twelve bioactive compounds were simultaneously determined in a Ge-Gen-Qin-Lian decoction using HPLC-MS/MS [1
]. Ten flavonoids, four alkaloids and four saponins were separated from a Ban-Xia-Xie-Xin decoction under the developed methods of HPLC-MS/MS [2
]. Six bioactive compounds, including rhein, that are found in Rheum palmatum
L., the major herb of San-Huang-Xie-Xin-Tang, were quantified by HPLC-MS/MS for further pharmacokinetic study [3
]. Analysis of the Ling-Gui-Zhu-Gan decoction revealed 90 tentatively identified compounds classified as flavonoids, coumarins, benzofurans, or coumestans by HPLC-hybrid electrospray ionization linear ion trap-Orbitrap mass spectrometry (HPLC-LTQ-Orbitrap-MS/MS) [4
]. Bioactive components make qualitative control of herbal formulations possible due to this highly efficient technique to separate and identify constituents in complex medicinal herb materials.
Thunb (Chinese herbal name, Yin-Cen-Hao), Gardenia jasminoides
Ellis (Zhi-Zi), and Rheum officinale
Baill (Da-Huang) were combined to prepare a formulation to treat jaundice. The formulation was named Yin-Chen-Hao-Tang (Artemisia capillaries
decoction) and was first recorded in the book Shang-Han-Lun (Treatise on Cold Damage Diseases) two thousand years ago. The weight ratio of each herb component recorded in the book was 6:3:2 for Artemisia capillaries
Thunb, Gardenia jasminoides
Ellis, and Rheum officinale
Baill. The therapeutic properties of this formulation have been demonstrated from its herbal composition. Artemisia capillaries
Thunb has been experimentally verified to have anti-hepatic fibrotic, anti-inflammatory, and hepatoprotective activities due to its constituent bioactive compounds, such as scoparone (6,7-dimethylesculetin), β-sitosterol, capillarin, capillarisin, cirsimaritin, quercetin, and chlorogenic acid [5
]. Another herbal component, Gardenia jasminoides
Ellis, has anti-inflammatory, anti-angiogenic, and choleretic effects from its ingredients genipin and geniposide [7
]. The known ingredient Rheum officinale
Baill contains aloeemodin, rhein, emodin, gallic acid, and chrysophano-8-o-d-glucopyranoside. Several pharmacological effects, including hepatoprotective, nephroprotective, anti-inflammatory, anti-oxidant, anticancer, and antimicrobial activities, are derived from the biological activities of rhein and emodin [10
]. The synergetic effects of 6,7-dimethylesculetin, capillarisin, chlorogenic acid, geniposide, and rhein, the major bioactive constituents of Yin-Chen-Hao-Tang, were reported to contribute to the therapeutic effects of the formulation [16
]. Studies on three combinations of these herbal medicines have shown potent anti-hepatic fibrosis and anti-hepatic apoptosis activities and alleviated hepatic oxidative stress effects after oral administration of Yin-Chen-Hao-Tang in rats [18
]. Literature reports have also demonstrated that aqueous extracts of Yin-Chen-Hao-Tang could diminish the infectivities of both herpes simplex virus 1 (HSV-1) and HSV-2 [21
A large body of literature exists on the bioactive compounds of Yin-Chen-Hao-Tang and their synergetic therapeutic effects. However, quality control information for this commonly available herbal formulation is scarce. HPLC-MS/MS has provided rapid and efficient quantification analysis of bioactive compounds among various herbal formulations. Good quality control for herbal formulations entails not only thoroughly validated quantitative methods for HPLC-MS/MS analysis, as per bioanalytical international guidance, but also the physical characteristics of herbal product quality. Therefore, this study aimed to physically and chemically examine the herbal formula Yin-Chen-Hao-Tang in preparations of pharmaceutical herbal products, raw fiber powders, and a decoction by light microscopy inspection with Congo red staining, and to simultaneously quantitatively evaluate the bioactive contents of scoparone, geniposide, and rhein, providing full validation and quality control of this formulation.
Yin-Chen-Hao-Tang has long been used for various liver diseases in clinical practice. In vivo studies have shown that administering Yin-Chen-Hao-Tang can reduce the concentrations of liver serum enzymes and potentially decrease collagen bundles thickening during the fibrosis process [22
]. The formula, which consists only of three medicinal herbs, also induces anti-hepatic apoptosis and alleviates hepatic oxidative stress effects [19
]. Concerning practical applications, therapeutic effects have been shown in animal experiments using Yin-Chen-Hao-Tang, and quality control standards have been created for this herbal formula.
Congo red and iodine staining was performed to examine the physical properties of the pharmaceutical products and the raw herbal powders. Congo red binds to cellulose molecules due to its symmetry and hydrogen bonding sites and, thus, is widely applied in phytochemistry to examine crude fibers [23
]. Alignment between the diphenol backbone of Congo red and cellulose fibers imparts strong affinity to the binding site. Raw herbal powders of herbal medicines were filamentous and irregular in shape, and the samples that contained these cellulose fibers were stained red. The experimental results showed that all brands of pharmaceutical products exhibited granules that were multiform, with rod-like shapes in various distributions. In other words, compared to the raw powder fibers used as references, the final commercial pharmaceutical products contained added grinder-crushed herbal powder.
The iodine-potassium iodine staining experiment was performed to identify the starch contents in various brands of pharmaceutical products. Iodine can detect starch concentrations as low as 1 µg/mL and causes blue staining due to the starch’s amylose content [24
]. The starch granule particles were blue and round compared to the fully stained cornstarch used as a positive control. The raw fiber powders were also stained with iodine but presented no blue color, indicating no starch content in the natural herbs or the decoction-prepared Yin-Chen-Hao-Tang. These results suggest that the prilling procedures likely involved additives that mixed and gelatinized during processing of the pharmaceutical products. A light microscopy study that physically inspected the pharmaceutical powders, the crude fibers and the decoction could provide further purity-indexed data to examine herbal formulations.
To quantify the amounts of ingredients in various commercially available pharmaceutical products and raw herbal powders, the most intense ion of each analyte detected in MRM mode was selected for quantitation. Under the described analytical conditions, scoparone, geniposide, and rhein underwent fragmentation of the parent ion to the product ion in the m/z
ratios 207.0 to 151.0, 406.1 to 227.1, and 282.9 to 240.15, respectively. The selected LC-MS/MS mass ions used for determination of bioactive compounds in commercial pharmaceutical products were consistent with previous reports of scoparone, geniposide, and rhein [25
]. The developed LC-MS/MS method for quantification of pharmaceutical YCHT products was well validated. All calibration curves obtained for scoparone, geniposide, and rhein exhibited good linear ranges from the LLOQ 10 ng/mL to 1000 ng/mL. The estimated R2
coefficients were all greater than 0.995, and all % RSD and % bias values were within 15%, indicating that the LC-MS/MS method provided excellent quantitative analysis of bioactive components in various preparations of Yin-Chen-Hao-Tang.
The quantification results demonstrate that the aqueous extract of geniposide, among the pharmaceutical powder samples from manufacturers A–C, ranged from 6.362 to 8.972 mg/g. However, geniposide in brand D was measured at only 0.840 mg/g, distinct from other manufacturers (Table 4
). The raw powders of Gardenia jasminoides
Ellis extracted in water contained approximately 5.288 mg/g geniposide, close to the values detected in the three pharmaceutical manufacturer products. In addition, the measured amount of geniposide was obviously increased in the extract raw fiber powder in ethanol, which corresponded with a previous study in which ethanol promoted the partition efficiency of geniposide [28
]. However, the extracted amount of geniposide from pharmaceutical products in ethanol was equal to, or less than, the previously published levels, which may be due to the presence of additives during granulating process that decrease the partition capacity of geniposide.
Aqueous-extracted rhein from pharmaceutical manufacturers A and B was detected at 0.092 and 0.093 mg/g, whereas brands C and D had only half as much, at 0.040 and 0.045 mg/g. Compared to grinder-crushed crude powder from Rheum officinale
Baill, the water-insoluble rhein was detected at only 0.014 mg/g, only one-fourth to one-tenth the levels in the other extracted pharmaceutical powders. However, the ethanol-extracted raw fiber powder of Rheum officinale
Baill was detected at 0.996 mg/g, greater than other preparations of Yin-Chen-Hao-Tang. The synergic effects of multicomponent herbal medicines in a combined preparation affect the concentrations of bioactive compounds, distinct from a single preparation [30
According to the quantitative results, although Gardenia jasminoides
Ellis is not considered the major herb in this formula, geniposide was predominant among the three bioactive compounds. Geniposide has reported anti-inflammatory, anti-oxidant and hepatoprotective activities [31
]. Rhein has anti-fibrotic, anti-inflammatory, anti-oxidant, and anti-tumorigenic effects [13
]. Considering the synergetic effects of the three medicinal herbs in the Yin-Chen-Hao-Tang formulation, the prolonged potency of geniposide with rhein from the area under the curve (AUC) of pharmacokinetic parameters exhibited additive properties not observed with geniposide alone [32
]. Based on the therapeutic properties of geniposide, it is supposed that the predominant geniposide, accompanied by rhein, plays an essential role in this formula.
In addition, there was no detectable scoparone in the herbal ingredients of any Yin-Chen-Hao-Tang preparation except the samples from manufacturers A and B. The quantitative results are not consistent with previous studies that identified scoparone in the decoction preparation. First, the organic properties of these compounds make dissolution of scoparone in water difficult. Second, the melting point of the standard 6,7-dimethoxycoumarin is 143–145 °C; thus, an ordinary process of Yin-Chen-Hao-Tang decoction would not easily include scoparone. Based on a review of previous studies on the pharmacokinetics and fingerprints of this formula, the preparation process includes dissolution of the freeze-dried Yin-Chen-Hao-Tang extraction powder with methanol [17
], which simplifies disclosure of the bioactive component by analytical instruments. In addition, only the pharmaceutical products from brand A and B contained detectable scoparone, indicating that some crude fibers may not be fully isolated from the gelatinized pharmaceutical samples during the manufacturing processes.
Our quantification results also offer a geniposide-based translational amount of pharmaceutical product for Yin-Chen-Hao-Tang decoction. According to back-calculations, the 250 mL of prepared decoction contained 36.068 mg/g geniposide, which is approximately four times the content of geniposide in brand C and six-fold higher than the amount in brand A, while the rhein content in brand C was 1% of that in the Yin-Chen-Hao-Tang decoction. However, the amount of marker compound could be influenced by the processing methods for Chinese herbal medicine preparation, such as baking, decoction, frying, soaking, and granulation processes. Crude fiber powder extraction and additives used may also differ in the granulating processes among manufacturers. The experimental results for different preparations provide clinical references for pharmaceutical product dosing that are translational to the Yin-Chen-Hao-Tang decoction.
4. Materials and Methods
All solvents, including LC/MS grade methanol (MeOH), ammonium acetate, and ethanol, were obtained from E. Merck (Darmstadt, Germany). Formic acid (98%–100%) was also purchased from E. Merck. Scoparone (6,7-dimethoxycoumarin) and rhein were both purchased from Sigma-Aldrich Research Biochemicals Inc. (St. Louis, MO, USA). Geniposide was obtained from Nacalai Tesque (Kyoto, Japan). The internal standard carbamazepine was provided from Research Biochemicals International Inc. (Natick, MA, USA). Triply deionized water was purified using a Q-Gard 1 Purification Cartridge water purification system from Millipore (Bedford, MA, USA) with a Millipak 40 Gamma Gold filter (0.22 µm) to produce particulate- and bacteria-free water. The pharmaceutical herbal products of Yin-Chen-Hao-Tang were purchased from Kaiser Pharmaceutical Co., Ltd. (Tainan, Taiwan), Sheng Chang Pharmaceutical Co., Ltd. (Taipei, Taiwan), Chuang Song-Zong Pharmaceutical Co., Ltd. (Kaohsiung, Taiwan), and Koda Pharmaceutical Co., Ltd. (Taoyung, Taiwan). Herbal medicines that comprise Yin-Chen-Hao-Tang, including Artemisia capillaries Thunb, Gardenia jasminoides Ellis, and Rheum officinale Baill, were prepared from Kaohsiung Chang Gung Memorial Hospital, Taiwan.
4.2. Instrumentation and Software
A Shimadzu UHPLC system consisting of a CBM-20A system controller, LC-20AD XR pumps, DGU-20A3 degasser, SIL-20AC XR auto sampler, and CTO-20A column oven coupled with an electrospray ionization (ESI) interface equipped with an LCMS-8030 triple quadrupole mass spectrometer (Shimadzu, Kyoto, Japan) were utilized for the separation and detection of bioactive compounds in Yin-Chen-Hao-Tang and carbamazepine (IS). The remote-controlled software for the Shimadzu UHPLC-MS/MS system (Kyoto, Japan) was LabSolutions v. 5.60 SP1. Statistical calculations were performed using Microsoft Excel.
4.3. High-Performance Liquid Chromatography Separation and Tandem Mass Spectrometry Detection
All samples were subjected to chromatographic separation using the Shimadzu HPLC system with an Acquity HPLC BEH C18 column (1.7 µm, 100 mm × 2.1 mm) (Waters Corp., Milford, MA, USA). Each sample was maintained in the autosampler at 4 °C; 10 µL of each sample was injected into the column. Chromatographic analyses were performed at 35 °C with a flow rate of 0.2 mL/min. The mobile phase consisted of 1 mM ammonium acetate with 0.1% formic acid in H2O (pH = 3) and methanol. A gradient elution was applied from 20% methanol (0–2 min) to 95% methanol and was held constant for 8 min (3–11 min) and decreased to 20% methanol at 12 min, with an overall run time of 20 min.
The MS/MS spectrometer operating conditions were optimized with a set interface voltage of 3.5 kV, a desolvation line temperature of 250 °C, heat block temperature of 400 °C, and a collision gas pressure of 230 kPa. The desolvation gas and drying gas was nitrogen, and the gas flow rates for the two conditions were 3 and 17 L/min, respectively. Argon gas was used for collision-induced dissociation (CID). All samples were detected using the MRM mode.
4.4. Stock and Working Solutions
The stock solutions of scoparone, geniposide, and rhein were prepared in 100% methanol at a concentration of 1 mg/mL. These stock solutions were diluted with 100% methanol to prepare a series of working solutions at 1, 5, 10, 50, 100, 500 and 1000 ng/mL for calibration curves. Quality control (QC) samples were prepared in the same manner at 5, 10, 50, and 100 ng/mL. All solutions were stored at −20 °C and were brought to room temperature for analysis.
4.5. Sample Preparation for Extracts of Pharmaceutical Products and the Decoction
Extract samples of 0.1 g of the above pharmaceutical products were immersed in 1 mL H2O or ethanol to a concentration of 100 mg/mL. After ultrasonication in a water bath for 10 min at room temperature, the samples were centrifuged at 13,000 rpm for 10 min at 4 °C. The supernatant was then collected and filtered through a 0.22-μm filter, and the filtrate was analyzed by LC-MS/MS at an appropriately diluted concentration. The Yin-Chen-Hao-Tang decoction was prepared as recorded in the original report, which consisted of boiling 22.5 g of Artemisia capillaries Thunb in 1 L of water for approximately 60 min until 500 mL water remained. Gardenia jasminoides Ellis (11.3 g) and Rheum officinale Baill (7.5 g) were added to the decoction for an additional 30 min to collect a final Yin-Chen-Hao-Tang decoction (250 mL).
4.6. Analytical Method Validation
The validation of all methods was based on published FDA bioanalytical method validation [34
]. The LOD and the LLOQ were, respectively, defined as signal-to-noise ratios (S/N) of 3 and 10 for each standard. The standard calibration curve was used to determine the QC and unknown sample concentrations from the peak area ratios already obtained for these samples. The acceptance criteria for calibration curves were least-squares linear regression R2
values greater than 0.995. However, LLOQ was accepted within 20% of the nominal concentration, and all back-calculated values of the created standard calibration curves were required to be within 15%.
The intra-day precision was the coefficient of variation (%, CV) within a single day from six replicate analyses, and the inter-day precision was determined across all consecutive days. Accuracy was the deviation (%, bias) of estimated concentration and nominal concentration. The accuracy and precision were calculated as bias (%) = [(Cobs − Cnom)/Cnom] × 100% and RSD (%) = (standard deviation (SD)/Cobs) × 100%. Accuracy and precision were required to be within ±15% of nominal values to be considered acceptable. The developed and validated method was applied to analyze four brands of commercial pharmaceutical herbal products, raw fiber powders, and a decoction. All samples were extracted and analyzed in triplicate.
4.7. Light Microscopy Photographs of Congo Red- and Iodine-Stained Samples
Samples were immersed in a mixture of glycerol and 20% ethanol (1:1) to make a 3% solution (w/w). One to two drops of the suspension were then added to a microslide, stained with 0.1% Congo red, and covered with a coverslip, avoiding bubbles. The iodine staining procedure was performed in the same manner, adding 10% iodine-potassium iodide solution instead of Congo red. Congo red and iodine staining were viewed by light microscopy (Olympus CKX41, Tokyo, Japan) under a total magnification of 100×, and photographs were taken.
In this study, a developed and validated LC-MS/MS method was used to determine the scoparone, geniposide, and rhein concentrations in various preparations of Yin-Chen-Hao-Tang simultaneously. The composition ratios of the Yin-Chen-Hao-Tang decoction were consistently labeled, but different amounts of the three marker ingredients in samples A–D from different manufacturers were present. The marker compound levels were complex and influenced by herbal processes, including growth time, storage, decoction method, granulation process, or product batch. The complexities of herbal medicine formulation, such as synergetic effects between herbs or analogues co-eluting during analysis, also increase the difficulties involved in quantification. Our data revealed detectable scoparone, geniposide, and rhein levels in samples from manufacturers A and B, which could be further selected to investigate the pharmacokinetics of these bioactive compounds. The quantitative results and the available data, despite the absence of scoparone values in some cases, which was not consistent with the previous research, were easily applied and similar under clinical administration.