Comparison of Five Prescriptions Composed by Coptidis Rhizoma and Evodiae Fructus Based on Separation and Quantiﬁcation of Eight Active Ingredients and Chemometrics Analysis

Veiw figure View Table

Furthermore, according to the ratio of the amount of GLS, ZYW, and BTW, we hypothesized that the differences in the content of their main ingredients could not be significant. It was difficult to find out the slight changes in main composition of GLS, ZYW, and BTW. It was also difficult to distinguish ZJW, GLS, ZYW, BTW, and FZJW effectively according to the contents of their active ingredients change. Principal component analysis (PCA) is a very popular data exploration and reduction technique for extracting the most important information from the large, and confusing data sets ^{[16, 17, 18]}. The PCA was useful and powerful to evaluate and distinguish ZJW, GLS, ZYW, BTW, and FZJW according to the contents of the detected active ingredients.

In this study, 105 batches of samples were prepared, in which 23 batches of BTW, 14 batches of ZYW, 23 batches of GLS, 20 batches of FZJW, 25 batches of ZJW, and 5 batches of ZJW preparations (ZJWps) were included. Immediately, 8 active ingredients were detected simultaneously by HPLC-DAD. Finally, the PCA method was used to analyze the 105 batches of samples and distinguish ZJW, GLS, ZYW, BTW, and FZJW according to the contents results.

2. Materials and Methods

2.1. Reagents and Materials

Coptidis rhizoma was purchased from Nanjing Pharmacy (Nanjing, P.R. China), Evodiae fructus was purchased from Evodiae implantation depot (Hubei, P.R. China). The specific species of Coptidis rhizoma and Evodiae fructus were Coptidis rhizoma Franch and Evodiarutaecar pa (Juss.) Benth. var. officinalis (Dode) Huang, respectively, which were identified by Prof. Xiaobin Jia (Jiangsu Provincial Academy of Chinese Medicine, Nanjing, P.R. China). Zuojinwan, Fanzuojinwan, Biantongwan, Ganlusan, and Zhuyuwan were all self-made in our laboratory (Table 1). Five batches of ZJWps were purchased from Hubei Ruidesheng pharmaceutical limited company (Hubei, China).

Figure 1. The chemical structures of 8 active compounds

Download as

View current figure in a new window

Figures index

Veiw figure View Figure

View next figure

Chlorogenic acid (1), jatrorrhizine (2), palmatin (4), berberine (5), evodin (6), evodiamine (7), rutaecarpine (8) were purchased from National Institute for the Control of Pharmaceutical and Biological Products (Beijing, China); Coptisine (3) was purchased from Prof. Hao Zhang (West China School of Pharmacy, Sichuan University, Chengdu, P.R. China). Their chemical structures were shown in Figure 1. HPLC-grade acetonitrile, methanol was purchased from Merck Company Inc. (Merck, Darmstadt, Germany). Ultra-pure water was prepared by a Milli-Q50 SP Reagent Water System (Millipore Corporation, MA, USA) for the preparation of samples and mobile phase solutions. Other reagents were of analytical grade.

2.2. HPLC-DAD Conditions

Agilent 1200 HPLC system (Agilent Corporation, USA) coupled with diode array detector (DAD) was used for determination. The chromatographic mobile phase was that 0.1% phosphoric acid and acetonitrile. The gradient program is presented in Table 2. Detection wavelength was set at 345 nm for compounds 1-5, and 210 nm for compounds 6-8.

Table 2. Solvent gradient program of HPLC-DAD analysis

Download as

Veiw figure View Table

2.3. Preparation of Standard Solutions and Sample Solutions

Mix standard stock solution containing chlorogenic acid (34.5 μg/mL), jatrorrhizine (35.0 μg/mL), coptisine (18.5 μg/mL), palmatine (12.9 μg/mL), berberine (42.6 μg/mL), evodin (35.3 μg/mL), evodiamine (41.8 μg/mL) and rutacarpine (32.0 μg /mL) was prepared in methanol. The solutions were stored in refrigerator at 4°C all the time. A set of standard solutions were prepared by the appropriate dilution of the stock solution with methanol.

samples were all extracted with 50 mL, 50% methanol by sonicating for 30 min (200 W, 40 KHz), and then 50% methanol was added to compensate the lost volume. The supernatant solution was filtrated through a syringe filter (0.45 μm) and 20 μL were subjected to HPLC analysis.

2.4. PCA for SO Samples

The PCA was performed to analyze these samples by SPSS 16.0 software package (SPSS, Chicago, IL, USA). In this study, the contents of the 8 compounds were analyzed from 23 batches of BTW, 14 batches of ZYW, 23 batches of GLS, 20 batches of FZJW, 20 batches of ZJW, and 5 batches of ZJWps, which composed a data matrix with 105 rows and 8 columns for PCA analysis after normalization. The first three principal components (PCs) were extracted, and the scatter plot were obtained by plotting the scores of PC 1 versus PC 2 and PC 3.

3. Results and Discussion

3.1. Optimization of Chromatographic Condition

The optimization of chromatographic condition was guided by the requirement of obtaining better resolution of adjacent peaks. The optimal mobile phase, consisting of acetonitrile-0.1% phosphoric acid (H₃PO₄) in mobile phase, which leads to good resolution and satisfactory peak shape. Besides, it was observed that the separation could be affected by column temperature obviously, especially the protoberberine alkoloids in Coptidis rhizoma. When the column temperature was higher than 20°C, the peaks of protoberberine alkoloids were too broad to part. Therefore, the column temperature was set at 20°C during analysis (Figure 2).

Figure 2. Typical HPLC-DAD chromatograms of the standard solution, ZJW and ZJWps

Download as

View current figure in a new window

Figures index

Veiw figure View Figure

View previous figure

View next figure

3.2. Method Validation of HPLC-DAD
3.2.1. Linearity, LOD and LOQ

The stock solution of the eight standards was prepared and diluted to six appropriate concentrations for the establishment of calibration curves. The regression equations were achieved after linear regression of the peak areas versus the corresponding concentrations. The LOD and LOQ for each analyte under the chromatographic conditions were determined at the S/N of 3 and 10, respectively. The results of the regression indicated that all eight reference compounds showed good linearity in a relatively wide concentration range. The LOD and LOQ of the eight compounds were 0.41-2.77 ng and 1.36-9.93 ng, respectively (Table 3).

Table 3. Linear regression data, LOD and LOQ of eight standards

Download as

Veiw figure View Table

3.2.2. Precision, Repeatability, and Stability

Intra-day and inter-day variations were chosen to determine the precision of the developed assay. Three different concentrations of standards were prepared. The intra-day variation was determined by analyzing the six replicates within a single day. Inter-day variation was examined in six consecutive days. Repeatability was confirmed with solutions prepared from BTW, and it was injected into the apparatus at 0, 2, 4, 8, 12, 16, and 24 h, respectively. Variations were expressed by RSD. All the results are summarized in Table 4. It indicated that the intra-day, inter-day, repeatability, and stability RSD values of the eight compounds were all less than 3.0%.

Table 4. Precision, repeatability and stability of eight compounds in sample (Biantongwan) (n=6)

Download as

Veiw figure View Table

3.2.3. Recovery

The recovery was performed by adding known amounts of the eight references. The spiked samples were then extracted, processed, and quantified in accordance with the methods mentioned above. The overall recoveries lay between 98.05 and 100.86% for all reference compounds. The results were summarized in Table 5.

Table 5. Recovery of eight compounds in BTW and ZJWps (n=3)

Download as

Veiw figure View Table

Table 6. Each content of eight active components in various samples (mg/g) (n=3)

Download as

Veiw figure View Table

3.3. Quantitative Analysis of Samples

105 batches of samples were prepared, in which 23 batches of BTW, 14 batches of ZYW, 23 batches of GLS, 20 batches of FZJW, 25 batches of ZJW (including 5 batches of ZJWps) were included (Table 6). Based on the detected results of 105 batches of samples, the content ranges of the 8 components were chlorogenic acid 0.44-5.52 mg/g, jatrorrhizine 1.71-18.46 mg/g, coptisine 6.15-46.38 mg/g, palmatin 1.78-19.61 mg/g, berberine 7.67-96.06 mg/g, evodin 1.43-16.44 mg/g, evodiamine 0.39-4.82 mg/g, rutaecarpine 0.23-3.38 mg/g.

In 105 batches of samples, the contents of alkaloids in Coptidis rhizoma were significantly higher than the ingredients in Evodiae fructus. Meanwhile, ingredients contents of Coptidis rhizoma alkaloids were berberine > coptisine > palmatin > jatrorrhizine, and the content of berberine was the highest than other 7 compounds. With the increase of the amount of Evodiae fructus, the contents of chlorogenic acid, evodin, evodiamine and rutaecarpine also increased in varying degrees. Meanwhile, evodin > chlorogenic acid > evodiamine > rutaecarpine, and the content of rutaecarpine was the lowest than other 7 compounds.

3.4. PCA of the Samples

To evaluate the variation of ZJW, FZJW, GLS, ZYW, and BTW, PCA was performed on the basis of the contents of 8 tested compounds from them. The first three principal components (PC 1, PC 2, and PC 3) with > 89.5% of the whole variance was extracted for analysis. Among them, PC 1 accounted for 79.742% of total variance, whereas PC 2 and PC 3 for 5.452% and 4.379%, respectively (Table 7). The remaining principal components were discarded owing to a minor effect on the model. According to their loadings, PC1 had good correlation with the 8 tested compounds. The above results suggested that most of the compounds might contribute to the classification of the samples.

Table 7. Total Variance Explained for principal component analysis

Download as

Veiw figure View Table

Figure 3. The 3D scatter plots obtained by Principal component analysis (PCA) of 105 samples

Download as