Introduction

Jinlianhua decoction (JD) is a prescription for the prevention and treatment of severe acute respiratory syndrome (SARS) issued by State Administration of Traditional Chinese Medicine in 2003. It is an aqueous decoction made of Flos Trollii (the flowers of Trollius chinensis), Herba Taraxaci (the whole plant of Taraxacum mongolicum), Folium Isatidis (the leaves of Isatis indigotica), Radix Puerariae Lobatae (the roots of Pueraria lobata), and Folium Perillae (the leaves of Perilla frutescens) in a ratio of 6:15:10:10:6. The entire prescription has efficacies of clearing away heat and toxic material, and dispelling wind and pathogen, and thus can be used for the treatment of Fengwen disease. According to traditional Chinese medicine (TCM), Fengwen is a group of warm diseases caused by exogenous evils of wind and heat, which occurs in the warm and windy spring or in the late winter. It starts with fever, cold intolerance, headache, cough and other mild pulmonary symptoms. In addition to SARS, influenza in western medicine or so-called seasonal cold in Chinese medicine also belongs to the category of warm diseases [1]. In Chinese medicine, the treatment of warm diseases not only focuses on direct elimination of influenza virus, but also highlights the relationship among virus, organism and drug. Thus, dialectical treatment through improving the state of human body and regulating its capacity of disease resistance is preferred. Studies have shown that the abnormal increase of various inflammatory cytokines as the influenza virus infects the body is the basis of immune damage [2]. TCM not only directly inhibits the replication of influenza virus, but also protects the human body from the damage caused by the virus through regulating the expression of inflammatory factors and the immune network in the body [3]. It is well known that the pharmacological actions of TCM are closely related to its chemical composition. In the previous studies, we have confirmed the anti-influenza effect of JD [4], and have quantitatively analyzed a part of the components in this decoction by high-performance liquid chromatography (HPLC) [5]. However, this is not enough for elucidation of its effective substances because the efficacy of TCM is the result of synergistic effects of multiple effective components [6]. Consequently, the comprehensive study of the composition of JD is of great significance for revealing its acting mechanism.

Although there is still technical difficulty in fully understanding the composition of TCM, ultra-high-performance liquid chromatography (UHPLC) shows great advantages in the separation and analysis of complex systems such as TCM due to its ultra-high efficiency, resolution and sensitivity. UHPLC tandem quadrupole-electrostatic field orbitrap high-resolution mass spectrometry (UHPLC-QExactive-MS) system with high-throughput scanning and multiple detection capabilities can provide high-quality mass spectra and accurate molecular weight of compounds. It has been applied to multi-component microanalysis, so as to realize the separation and qualitative detection of the compounds in complex systems [7]. Another technique quadrupole-linear ion trap mass spectrometry (QTrap), which uses a multiple reaction monitoring information-dependent acquisition-enhanced product ion scanning (MRM-IDA-EPI) detection mode [8], can obtain the high-quality MS/MS spectra of parent ions in the corresponding MRM channel to double characterize the unknown compounds and improve the accuracy of the qualitative analysis. Furthermore, it also can use the high-selectivity and high-sensitivity MRM scan to obtain the peak area of the compounds for quantitative analysis. In view of this, we used LC-QExactive-MS and LC-QTrap-MS to qualitatively and quantitatively analyze the constituents of JD, respectively, to provide a basis for the determination of pharmacodynamic substance basis. In addition, we used morphological identification in combination with DNA barcoding technique [9] to identify the crude drugs used in the preparation of JD decoction.

Materials and Methods

Crude Drugs

The five crude drugs including Flos Trollii, Herba Taraxaci, Folium Isatidis, Radix Puerariae Lobatae, and Folium Perillae were purchased from materia medica markets or drugstores in various regions of China (Table 1), and the voucher specimens have been deposited at the Herbarium of School of Life Sciences, Beijing University of Chinese Medicine.

Table 1 Morphological and molecular identification of crude drugs
Full size table

Reagents and Materials

Thermo UHPLC series quadrupole-electrostatic field orbitrap high-resolution mass (Ultimate3000 QExactive plus LC–MS) spectrometer, the data processing software Xcalibur and Compound Discoverer (CD) 2.1 were the products of Thermo Fisher Scientific (Pittsburgh, PA, USA). Acquity ultra-performance liquid chromatograph (UPLC) was from Waters Corporation (Milford, MA, USA). AB Sciex QTrap 4500 triple quadrupole mass spectrometer equipped with an electrospray ionization (ESI) source and the data processing software Analyst 1.6.1 was manufactured by AB Sciex Pte. Ltd. (Framingham, MA, USA). Sigma 1–14 desktop high-speed centrifuge was the product of Sigma-Aldrich, Inc. (St. Louis, MO, USA). Milli-Q integral water purification system was produced by Millipore Corporation (Bedford, MA, USA). METTLER TOLEDO Xp26 one-millionth electronic analytical balance was from METTLER TOLEDO (Zurich, CH). IKA VORTEX GENIUS 3 vortex mixer was made by Janke & Kunkel KG.IKA-werk (Staufen, DE). LongGene thermal cycle analyzer was provided by Hangzhou Langji Scientific Instrument Co., Ltd. (Hangzhou, Zhejiang, China). BG-gdsAUTO510 gel imaging system was the product of Beijing Bay Gene Biotechnology Co., Ltd. (Beijing, China). DYY-6C electrophoresis apparatus was provided by Beijing Liuyi Device Factory (Beijing, China). KQ-500DE CNC ultrasonic cleaner was provided by Kunshan Ultrasonic Instrument Co. Ltd. (Kunshan, Jiangsu, China). Methanol and acetonitrile of LC–MS grade were manufactured by Fisher Scientific (Pittsburgh, PA, USA). Formic acid of chromatographic grade was the product of Tedia Company, Inc (Fairfield, Ohio, USA). DNA rapid extraction kit for broad spectrum plant genome, 2 × Taq PCR Master Mix enzyme, and BM2000 + DNA Marker were purchased from Beijing Bomaide Biotechnology Co., Ltd. (Beijing, China). Deionized water was purified by Milli-Q system. All other chemicals were available products of at least analytical grade.

Reference compounds including genistin, apigenin, daidzin, kaempferol, rutin, scutellarin, vitexin, ferulic acid, 3′-hydroxy puerarin, 2″-O-β-l-galactopyranosylorientin, indirubin and cynaroside were bought from Shanghai Yuanye Biotech Co., Ltd. (Shanghai, China). 3′-Methoxy puerarin was purchased from Chengdu Push Biotech Co., Ltd. (Chengdu, Sichuan, China). Caffeic acid was from Chengdu Pufei De Biotech Co., Ltd. (Chengdu, Sichuan, China). Orientin was from Pharmacodia (Beijing) Co., Ltd. (Beijing, China). Puerarin and rosmarinic acid were purchased from National Institutes for Food and Drug Control (Beijing, China). Trollioside [10], 3,4-dimethoxybenzoic acid [10], trollisin I [11], proglobeflowery acid [10], 2″-O-(2′″-methylbutanoyl)isoswertisin [11], tecomin [12], and 2″-O-(2′″-methylbutanoyl)vitexin [11] were isolated from Flos Trollii in our lab. The individual purity of each reference was confirmed over 98% according to HPLC analysis.

Morphological and Molecular Identification of Crude Drugs

Morphological Identification

The five crude drugs were morphologically identified by comparing their morphological characteristics including shape, color, odor, size, texture and sectional properties with those recorded in Pharmacopoeia of the People’s Republic of China [13, 14].

Molecular Identification

The five crude drugs (0.1 g) were subject to total DNA extraction using DNA rapid extraction kit of broad spectra plant genome. The sequence of forward primer ITS2F was 5′-ATGCGATACTTGGTGTGAAT-3′, while that of the reverse primer ITS3R was 5′-GACGCTTCTCCAGACTACAAT-3′. The reaction system of PCR included 12.5 µL 2 × Taq PCR Master Mix enzyme, 1.0 µL of forward primer and 1.0 µL of reverse primer (5 μmol L−1), 1.0 µL of total DNA and 9.5 µL of ddH2O. The PCR amplification was carried out using the program compose of 94 °C for 5 min, 94 °C for 30 s, 56 °C for 30 s, and 72 °C for 45 s (40 recycles), then 72 °C for 10 min again. The resultant samples were stored at 4 °C. The samples were sent to Beijing Bomaide Biotechnology Co., Ltd. for sequencing after test with 1% agarose gel electrophoresis. Contig Express 3.0 (Informax., Inc, USA) was used for the assembly and sequence checking. The checked ITS2 sequences were uploaded to the GenBank database for Basic Local Alignment Search Tool (BLAST) comparison to determine whether they were the target ones.

Qualitative Analysis of JD by LC-QExactive-MS

Preparation of Test Samples

In accordance with the proportion of the prescription, a quantity of five crude drugs was taken, and a volume of water was added to extract two times, each for 30 min. The extract was filtered and combined, and then concentrated to each 1 mL containing 0.1 g of crude drug. The concentrate was stored at − 20 °C, and centrifuged at 12,000 r min−1 for 10 min before use. The supernatant was taken as the test solution.

Apparatus and Analytical Conditions

Chromatographic separation was performed with a Waters BEH C18 column (100 mm × 2.1 mm i.d.; 1.7 μm). Gradient elution was performed in 60 min using the mobile phase consisting of acetonitrile with formic acid (0.1% v/v) (A) and 0.1% v/v aqueous formic acid (B) from 5 to 95% A at a flow rate of 0.2 mL min−1. The injection volume was 5 μL.

System used in positive and negative ion modes was coupled with heated electrospray source (HESI), and the spray voltages were 3.5 kV for positive and 2.8 kV for negative. The flow rate of sheath gas was 30 arbitrary units (a.u.), and that of auxiliary gas was 10 a.u.. Other conditions included capillary temperature of 320 °C and resolution of 70,000.

Statistical Analysis

In accordance with the precise molecular weight of the compound obtained from the total ion chromatogram (TIC) and the fragment ions generated in the characteristic mode, Xcalibur and CD 2.1 were used to calculate the possible composition (error less than 5 ppm). Then, based on the characteristic fragment ion information of compounds, the attribution information and the existing relevant composition reported, the possible structure was inferred. The compound designation referred to the ChemSpider database, the Pubmed database, the mzVault database, and the mzCloud database.

Quantitative Analysis of 24 Components by LC-QTrap-MS

Apparatus and Analytical Conditions

Chromatographic separation was performed with an ACQUITY UPLC HSS T3 column (100 mm × 2.1 mm i.d.; 1.8 μm). The mobile phase consisted of acetonitrile with formic acid (0.1% v/v) (A) and 0.1% v/v aqueous formic acid (B), and the flow rate was 0.2 mL min−1. Gradient elution was as follows: 0–1 min, 0% A; 1–3 min, 0–20% A; 3–11 min, 20% A; 11–14 min, 20–35% A; 14–17 min, 35–70% A; 17–19 min, 70–100% A; 19–21 min, 100% A; 21–21.1 min, 100–0% A. The injection volume was 5 μL.

System used in negative ion mode was coupled with electrospray ionization source (ESI). The ionspray voltage (IS) was 4500 V, and ionization temperature (TEM) was 500 °C. The nebulizer gas (GS1) and heater gas (GS2) were 50 and 40 psi, respectively. For MRM mode, the precursor ion, product ion, de-clustering potential (DP), entrance potential (EP), collision energy (CE) and collision cell exit potential (CXP) of the 24 measured components are shown in Table 2. Total ion MRM chromatogram of mixed references and samples are shown in Fig. 1.

Table 2 Optimized structure parameters of 24 components in JD
Full size table
Fig. 1
figure 1

Total ion MRM chromatogram of mixed references (left) and samples (right)

Full size image

Preparation of Reference Solutions

Stock solutions of 24 references (each 1.0 mg mL−1) were individually prepared by dissolving accurately weighed reference compounds in methanol. A mixed reference solution containing all 24 reference compounds was prepared and serially diluted with methanol to appropriate concentrations to produce working solutions for quantitative analysis (0.5, 1, 2, 5, 10, 20, 50, 100, 200, 500, and 1000 ng mL−1). The solutions were stored at − 20 °C before use.

Results

Morphological and Molecular Identification of Crude Drugs

The results of morphological identification were consistent with those of molecular identification (Table 1), which showed that 34 samples out of the total of 40 samples of crude drugs used were from the original plant collected in Chinese pharmacopoeia and the remaining six samples including five samples of Radix Puerariae Lobatae and 1 sample of Folium Isatidis were from the alternative original plants Pueraria thomsonii and Polygonum tinctorium, respectively. All identification results were confirmed by Professor Rufeng Wang.

Qualitative Analysis of JD

The TIC of JD analyzed by LC-QExactive-MS system is shown in Fig. 2. Eighty-nine compounds were deduced and their structures are provided in Tables 3, 4 and Fig. 3, respectively.

Fig. 2
figure 2

Total ion chromatogram (TIC) of JD in negative mode (left) and positive mode (right)

Full size image
Table 3 Characterization of constituents of JD by LC-QExactive-MS (negative mode)
Full size table
Table 4 Characterization of constituents of JD by LC-QExactive-MS (positive mode)
Full size table
Fig. 3
figure 3

The structures of 89 assigned compounds

Full size image

Validation of Quantitative Analysis

Linearity, Lower Limit of Detection (LLOD), And lower Limit of Quantification (LLOQ)

For the calibration curve established, different concentrations of reference solution were taken for LC–MS/MS analysis in triplicate. Calibration curves for 24 analytes were generated by plotting the average peak areas versus the corresponding concentrations. Good linearity (R ≤ 0.9919) in the tested concentration ranges of the analytes except indirubin was observed. The minimum concentration of linearity solution was selected and then diluted to series of solution, respectively, and the values calculated at signal to noise (S/N) ratios of 3 and 10 were considered as LLOD and LLOQ, respectively (Table 5).

Table 5 Calibration curves, correlation coefficient (R), linear range, lower limit of detection (LLOD), and lower limit of quantification (LLOQ) of 24 analytes
Full size table

Precision, Repeatability and Stability

The precision of the developed method was validated by determining intra-day and inter-day variations and was expressed in the form of relative standard deviations (RSD). Six replicates were analyzed within a day for the intra-day variation assessment, and the experiments were duplicated on three consecutive days for the inter-day variation assessment. The RSD values of the intra-day and inter-day variations were in the ranges of 2.17–4.79% and 0.91–4.95%, respectively. The repeatability was tested using six test solutions prepared according to the same method, and the RSD values were in the range of 1.20–4.95%. Stability was evaluated by analyzing the sample solutions at room temperature at 0, 2, 4, 8, 12, and 24 h, and the RSD values for 24 references were all less than 4.95%. Thus, the developed method exhibited good precision, repeatability, and stability (Table 6).

Table 6 Precision, repeatability and stability of 24 analytes (n = 6)
Full size table

Recovery

The recovery test was conducted with three concentration levels (low, medium, and high) of the mixed references added to the known amounts of samples. The resulting samples were extracted and analyzed by the proposed method. The whole process was repeated, and the content of each analyte was determined by the corresponding calibration curve. The percentage recoveries were calculated according to the equation: (total detected amount − original amount)/added amount × 100%. The results showed that the recoveries were within the range of 80.35–119.68%, and the RSD value variations were in the range of 0.19–5.08%. Thus, the developed method exhibited good accuracy (Table 7).

Table 7 Method recoveries for 23 analytes (n = 3)
Full size table

Sample Analysis

The validated UPLC–MS/MS analytical method was applied to simultaneous quantification of 24 components in eight batches of JD. The contents of the compounds (n = 3) were calculated with an external standard method based on their respective calibration curves. The results are listed in Table 8.

Table 8 The contents (μg g−1 crude drug) of 24 analytes in samples
Full size table

Discussion

Unlike individual crude drugs, decoction of Chinese medicine contains diversified compounds. These compounds are difficult to be characterized by the common analytical procedure because of their complexity. We established a qualitative and quantitative method for analysis of the main components in JD by UHPLC–MS/MS for the first time. The qualitative analysis was performed using LC-QExactive-MS under high resolution and high sensitivity. Based on the resultant precise molecular weight, and mass spectrometry fragmentation of the compounds, the composition of JD has been comprehensively explained. The quantitative analysis was conducted by LC-QTrap-MS. This method is fast and efficient, and does not require complete separation of the chromatographic peaks of multiple components. The precision, repeatability, stability and recovery for the assigned compounds excluding indirubin are in well compliance with the measurement requirements. And the LLOQ is as low as 0.05–4.67 ng mL−1, which can be used for related studies with low component content.

Qualitative and quantitative analyses of JD prepared from eight batches of crude drugs obtained from various places of China were carried out by the above methods. The qualitative results showed that JD mainly contains 89 compounds, and most of them belong to flavonoids, phenolic acids and alkaloids. After these compounds were screened on the basis of their bioactivities related to the efficacy of JD, 24 compounds including 16 flavonoids, 7 phenolic acids and 1 alkaloid were selected to perform the quantitative analysis. The results showed that the compounds whose contents in the decoction were above 100 μg g−1 crude drug included orientin, 2″-O-β-l-galactopyranosylorientin, puerarin, trollisin I, rosmarinic acid, 2″-O-(2′″-methylbutanoyl) isoswertisin, daidzin, scutellarin, 3′-methoxy puerarin, vitexin, 3′-hydroxy puerarin, 2″-O-(2′″-methylbutanoyl) vitexin, kaempferol, caffeic acid, 3,4-dimethoxybenzoic acid, and cynaroside in descending order. These components are mainly from Flos Trollii, Radix Puerariae Lobatae, and Folium Perillae [15,16,17], and have related bioactivities such as antiviral, antibacterial and anti-inflammatory effects [12, 18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37]. Therefore, these compounds can be considered as the major components of JD.

It was found through statistical analysis that the contents of the 24 components differ greatly in JD prepared from different batches of crude drugs. This might be because the crude drugs of different batches were different in origin, growing environments, preparation process, storage, etc., which resulted in uneven drug quality. For example, based on the results of morphological and molecular identifications, some samples of Radix Puerariae Lobatae used in the experiments were from the roots of P. thomsonii rather than Pueraria lobata. Although the Chinese pharmacopoeia 2005 edition onward has excluded P. thomsonii as the original plant of Puerariae Lobatae, the crude drugs from both original plants are still used indiscriminately in the clinical practice. Thus, the content of puerarin in JD decoction prepared with the roots of Pueraria lobata was significantly higher than that prepared with the roots of P. thomsonii. Our results were also consistent with the findings reported previously [38]. Even for crude drugs of the same origin, their quality is also different due to the diversified growing environment, preparation process and storage condition. The difference of components in the crude drugs and decoctions inevitably affects the consistency of efficacy. The quality control of the chemical components is an important means to guarantee the pharmacological effects of crude drugs and the decoctions [39]. Therefore, the genuine regional crude drugs should be used as much as possible, and the preparation process and storage condition should also be strictly standardized. In addition, indirubin was detected as one of the main components of Folium Isatidis in qualitative analysis; however, it could not be quantified due to its poor ion response. In the methodological test of quantitative analysis, the precision, stability, and repeatability results for indirubin were not qualified. Consequently, the quantitative results for indirubin are for reference only, which may be related to the poor stability of this compound. It has been reported that the content of this compound decreased significantly with the increase of standing time under natural light and room temperature, and slightly reduced even under refrigeration conditions [40].

Conclusions

The qualitative and quantitative analysis methods established in this study are suitable for the analysis and monitoring of the main components in JD. The 16 compounds determined based on the qualitative and quantitative results are the major components of the decoction. This study provides a scientific basis for the determination of pharmacodynamic substances of JD, and lays a foundation for the quality control research of the decoction.