The study of different markers among Chuanxiong Chatiao dosage forms based on bioactive components and antiplatelet aggregation biopotency integrated with chemometrics

2.1 Reagents and chemicals

Standards of vanillic acid, ferulic acid, senkyunolide I, senkyunolide H, coniferyl ferulate, Z-ligustilide, senkyunolide A, butylidenephthalide and levistilide A (Purities ≥ 98 % by HPLC) were obtained from Chengdu Pufei De Biotech Co., Ltd. (Chengdu, China). Chlorogenic acid (96.3 % Pure), caffeic acid (99.7 % Pure), prim-O-glucosylcimifugin (94.6 % Pure), liquiritin (95.0 % Pure), Luteolin-7-O-glucoside (96.6 % Pure), 5-O-Methylvisammioside (97.4 % Pure), quercitrin (93.5 % Pure), hesperidin (95.3 % Pure), quercetin (99.8 % Pure), rosmarinic acid (98.1 % Pure), buddleoside, (98.0 % Pure), pulegone (99.8 % Pure), ammonium glycyrrhizinate (94.4 % Pure), imperatorin (99.0 % Pure), asarinin (100.0 % Pure), notopterol (97.5 % Pure), isoimperatorin (99.2 % Pure) were purchased from national institutes for food and drug control (Beijing, China). Methanol and acetonitrile were HPLC grade, adenosine-5-two sodium phosphate (C₁₀H₁₃N₅Na₂O₁₀P₂, ADP), sodium ferulic acid, and dimethyl sulfoxide (DMSO) were obtained from Sigma-Aldrich Co (St. Louis, MO, USA). Formic acid was analytical grade and purchased from Guangzhou Chemical Reagent Factory (Guangzhou, China). All aqueous solutions were prepared and produced by ultra-Pure water with a Milli-Q system (18.2 MΩ, Millipore, Bedford, MA, USA). Tri Sodium citrate dihydrate C₆H₅Na₃O₇·2(H₂O) were analytical grade and acquired from Guangzhou Chemical Reagent Factory (Guangzhou, China). Sodium chloride injection (0.9 %, W/V) was obtained from Sichuan Cologne Pharmaceutical Co., Ltd. (Chengdu, China).

2.2 Instrumentation

KQ-500VDE double frequency digital ultrasonic cleaning instrument was obtained from Kunshan Ultrasonic Instrument Co., Ltd. (Kunshan, China). All the analytes were performed using a Waters Acquity HPLC-MS/MS system (Waters Corp., Milford, MA, USA). HPLC system which consisted of an online degasser, an autosampler, and a binary pump solvent management system. Mass spectrometry detection was performed using a Xevo Triple Quadrupole MS (Waters Corp.) equipped with an electrospray ionization source (ESI). The ESI-MS spectra were obtained using multiple reaction monitoring (MRM). Separation was achieved on an Agilent Eclipse Plus C_l8 column (5 μm, 4.6 mm × 250 mm). BSA224S Precision electronic balance was acquired from Beijing Sartorius Scientific Instrument Co., Ltd. (Beijing, China). ZNHW-II Intelligent digital display electric hearting set was acquired from Gongyi Yuhua Instrument Co., Ltd. (Tianjin, China). Frontier™ 5000 Multi Pro multi-function centrifuge was obtained from OHAUS Company (Pine Brook, USA); RE-2000B rotatory evaporator was acquired from Shanghai Yarong Biochemical Instrument Factory (Shanghai, China). SC-2000 Platelets aggregated instrument was obtained from Beijing Succeeder Technology Development Co., Ltd. (Beijing, China).

2.3 Sample collection

CX, QH, BZ, XX, JJ, FF, BH, and GC were collected from the Hehuachi Chinese herbal medicine market of Chengdu, Sichuan province, and the traditional Chinese medicine market of Puning, Guangdong province. There were 20 samples and they were further identified as certified goods by chief pharmacist Liu Maogui, director of the quality management department, Kangmei Pharmaceutical Co., Ltd. CX is a dried rhizome of Ligusticum chuanxiong Hort.; QH is the dried rhizome and root of Notopterygium incisum Ting ex H. T. Chang.; BZ is the dried root of Angelica dahurica (Fisch. ex Hoffm.) Benth. et Hook. f.; XX is the dried rhizome and root of Asarum heterotropoides Fr. Schmidt var. mandshuricum (Maxim.) Kitag.; JJ is the dried aerial part of Schizonepeta tenuifolia Briq.; BH is the dried aerial part of Mentha haplocalyx Briq.; and GC is the dried rhizome and root of Glycyrrhiza uralensis Fisch. Twenty samples of Chuanxiong Chatiao powder were prepared with 120 g of CX, 60 g of QH, 60 g of BZ, 30 g of XX, 120 g of JJ, 45 g of FF, 240 g of BH, and 60 g of GC, which had been pulverized to a fine powder, sifted, and mixed well (National Commission of Chinese Pharmacopoeia, 2020a, 2020b). Chuanxiong Chatiao dropping pills, granules, tablets, concentrated pills, and pills were collected from different factories. All sample information is shown in Table 1.

2.4 Determination of 26 constituents by HPLC-MS/MS

2.5 Biopotency assay in vitro

2.6 Statistical analysis

The data were analyzed by SNK with one-way ANOVA using SPSS 26.0 (Palo Alto, CA, USA). The differences were considered significant differences for the different subsets when alpha = 0.05, and there were no significant differences in the same subset. Analyses were further conducted using hierarchical clustering analysis (HCA) and partial least squares (PLS) with SIMCA-P 14.0 (Umetrics AB, Umea, Sweden). The parameters of the modeling (R² and Q² values) explained the accuracy of the fitting model.

3.1 Optimization of chromatographic and MS condition

Various compositions of the mobile phase, different column temperatures and flow rates were compared and tested for obtaining accurate and sensitive quantitative method by HPLC-MS/MS. As a result, acetonitrile was selected as the mobile phase A, and 0.1 % formic acid–0.5 mmol/L ammonium acetate (V/V) were selected as the mobile phase B, the column temperature was held at 30 °C, the flow rate was determined at 1.0 mL/min. In order to obtain better responses, all analytes were determined with the electrospray ionization (ESI) source by a full-scan mass spectrometry (MS) method, and their ion source parameters such as cone voltage (CV) and collision energy (CE) were optimized under both positive and negative modes. In the meanwhile, the most abundant, specific, and stable fragment ions were obtained by multiple reaction monitoring (MRM) from MS/MS spectrum. The detailed information of retention time (t_R), MS information, CV and CE for each analyte was listed in Table 2 and Fig. 1.

Close

Fig. 1

MRM chromatogram of the 26 compounds (1. Chlorogenic acid 2. Caffeic acid 3. Vanillic acid 4. Prim-o-glucosylcimifugin 5. Ferulic acid 6. Liquiritin 7. Senkyunolide I 8. Senkyunolide H 9. Luteolin-7-O-glucoside 10. 5-O-methylvisamidol glycoside 11. Quercitrin 12. Hesperidin 13. Quercetin 14. Rosemarinic acid 15. Buddleoside 16. Pulegone 17. Coniferyl ferulate 18. Glycyrrhizic acid 19. Senkyunolide A 20. Z- ligustilide 21. Butylidenephthalide 22. Imperatorin 23. Asarinin 24. Notopterol 25. Isoimperatorin 26. Levistilide A).

Export to PPT

3.2 Validation of quantitative analysis in HPLC-MS/MS

Linearity was evaluated by developing calibration curves for the 26 analytes in which mixed standard solutions were diluted and six concentrations of the standard solutions were prepared. Calibration curves for the 26 analytes were formed by plotting the corresponding concentration versus the peak area. The results showed that the R² value of the calibration curves were above 0.9997 for all components. The limit of detection (LOD) and limit of quantitation (LOQ) for each compound were determined at signal-to-noise (S/N) ratios of 3.

and 10, respectively. The precision of the method was assessed based on the relative standard deviation (RSD) by the determination of intraday and interday variations, using six replicates within a day for the intraday variations, and the experiments were repeated on three continuous days for the interday variations. The results showed that the RSD values of intraday variations of all analytes were within 5 %, and the RSD values of five analytes, namely, pulegone, coniferyl ferulate, senkyunolide A, Z-ligustilide, and levistilide A, were greater than 5.0 %. It was suggested that those analytes should be used immediately after preparation and not be held for very long before analysis. The RSD values of the other analytes were within 5.0 %. The repeatability of the proposed method was assessed based on six replicate injections for the same sample, and the RSD values of all analytes were within 5.0 %. The stability of the samples was evaluated by analyzing six of the same sample solutions at room temperature at different times (0, 3, 6, 9, 12, or 15 h) within a day, and the RSD values were all within 5.0 %. A recovery experiment was conducted with three different concentration levels (low, medium, or high) of the mixed standard solutions added to known concentrations of a sample, and the rate of recovery of the sample and standard was about 1:0.5 (low), 1:1 (medium), and 1:1.5 (high). The entire process was repeated three times at each concentration level. The method recovery was calculated using the formula: $$re\text{cov}ery\left(\%\right)=\left(\mathit{Ctotal}-Csample\right).Cs\tan dard\times 100\%$$

The results showed that the recoveries were within 5.0 %. The detailed results are shown in Table 3.

3.3 Validation of the methodology in the bioassay of AAB

According to the discriminant requirements of the validity test results of the parallel line assay based on a quantitative response (2∙2) method, the validity test of the results of the quantitative determination of the AAB of sample was conducted. The results showed that the regression was extremely significant (P < 0.01), indicating that while the dose of sodium ferulate and the samples increased, the platelet inhibition rate also increased, further revealing that the dose–effect relationship was linear. As far as the parallelism, there was no significant difference (P > 0.05), indicating that the lines formed by the increased platelet inhibition rate in the high-dose and low-dose sample groups were parallel to the lines formed by the increased platelet inhibition rate in the high-dose and low-dose sodium ferulate groups, respectively. Therefore, the assay result was satisfactory. The rate of fiducial limits indicates the precision and stability of the biological assay method. When the rate of fiducial limits is smaller, the precision is better. On the basis of the recommendation of the traditional Chinese medicine biological assay method on the rate of fiducial limits (Xiao et al., 2014), the rate of fiducial limits of AAB in this study was set as not exceeding 30 %.

3.4 Comparison analysis of AAB among six dosage forms

The results of AAB testing were shown in Table 4. In terms of the average value ranking (Fig. 2), powder (6.7760 U g⁻¹) > pills (5.1737 U g⁻¹) > dropping pills (4.7759 U g⁻¹) > concentrated pills (4.5890 U g⁻¹) > tablets (4.2761 U g⁻¹) > granules (3.9586 U g⁻¹). One-way ANOVA was performed, and SNK was selected to compare the differences in the AAB among Chuanxiong Chatiao powder and the five related dosage forms. The results indicated that there were significant differences (P < 0.05) between groups, and there was no significant difference between the dropping pill and concentrated pill. However, significant differences were observed with the other dosage forms. This indicated that the AAB significantly differed between powder and other dosage forms and could be divided into five subgroups, with the AAB of powder the highest, followed by pills, then followed by concentrated pills and dropping pills, then tablets, and finally granules. We concluded that the efficacy of the powder was the strongest, followed by pills, concentrated pills and dropping pills, tablets, granules, indicating that there were some differences among the six dosage forms.

Close

Fig. 2

The AAB of Chuanxiong Chatiao Powder and five related dosage forms.

Export to PPT

3.5 HCA

To evaluate the difference among Chuanxiong Chatiao powder and five related dosage forms with 26 bioactive ingredients and AAB, HCA was performed. A total of 43 samples were selected as observations while the contents of 26 compounds were selected as independent variables, AAB (Table 4) as dependent variable. In the dendrogram of HCA (Fig. 3), where it can be seen that these samples were sorted into three clusters reflecting the active constituents and AAB differences of three types of samples. Firstly, powder and others were respectively clustered into one category, indicating that the difference was significant between powder (group 1) and others. Secondly, pill and concentrated pill were clustered into one category (group 2), most granules and all tablets were clustered into other category (group 3), indicating that there were significant differences between pill, concentrated pill and granule, tablet, respectively, but there were some similarities between pill and concentrated pill, or granule and tablet. The HCA results demonstrated that the difference is most significant between powder and other dosage forms in the combination of active ingredients and bioactivity, concentrated pills and pills being similar, and tablets and drop pills and granules being similar, there were also significant difference between concentrated pills, pills and tablets, drop pills, granules.

Close

Fig. 3

The result of dendrogram by HCA (Group 1, powder-1 ∼powder-20; Group 2, pill-1 ∼ pill-3, concentrated pill-1 ∼ concentrated pill-7, granule-3; Group 3, granule-1, granule-2, granule-4 ∼granule −7, tablet-1∼tablet-3).

Export to PPT

3.6 The analysis of different markers among six dosage forms by PLS

To further compare and differentiate the difference among Chuanxiong Chatiao powder and five related dosage forms base on 26 active compounds and AAB, PLS was performed. A total of 43 samples were selected for analysis, while the contents of 26 compounds were selected as independent variables, AAB as dependent variable. In the PLS, the first three principal components were selected, R²X (cum) was 0.677, R²Y (cum) was 0.958, Q² (cum) was 0.940, and we generated score scatter plot, permutation, loading scatter plot, coefficient plot and variable importance plot (Fig. 4). In the score scatter plot (Fig. 4A), all samples were mainly divided into four parts, besides it couldn’t be distinguished between granule and tablet, other dosage forms could be better distinguished. It indicated that there was no significant difference between granules and tablets, but the difference were significant among other dosage forms. In the permutation (Fig. 4B), R² was 0.337, Q² was −0.205, the values of left are lower than the right, indicating that the model was accurate and predictive. In the coefficient plot (Fig. 4C), the correlations were negative between Prim-O-glucosylcimifugin, Senkyunolide I, Quercitrin, Quercetin, Buddleoside and AAB, the correlation were positive between other ingredients and AAB, respectively. In the variable importance plot (Fig. 4D), the value of VIP in decreasing order was as follows: Z- ligustilide (1.369) > Senkyunolide A (1.357) > Butylidenephthalide (1.338) > Asarinin (1.312) > Rosmarinic acid (1.270) > Pulegone (1.256) > Senkyunolide H (1.210) > Notopterol (1.201) > Coniferyl ferulate (1.191) > Hesperidin (1.190) > Vanillic acid (1.176) > Luteolin-7-O-glucoside (1.154) > Levistilide A (0.991) > Caffeic acid (0.933) > Isoimperatorin (0.889) > Ferulic acid (0.847) > Prim-O-glucosylcimifugin (0.838) > Quercitrin (0.786) > Liquiritin (0.747) > Chlorogenic Acid (0.658) > Quercetin (0.656) > Buddleoside (0.574) > Senkyunolide I (0.462) > 5-O-Methylvisammioside (0.431) > Glycyrrhizic acid (0.398) > Imperatorin (0.376). The results indicated that the correlations were different between 26 bioactive constituents and AAB, further revealed the efficacy of the six Chuanxiong Chatiao dosage forms was the result which combined action of multiple bioactive constituents.Fig. 5..

Close

Fig. 4

The results of statistical analysis by PLS (A: score scatter plot; B: permutation; C: coefficient plot; D: variable importance plot).

Export to PPT

Close

Fig. 4

The results of statistical analysis by PLS (A: score scatter plot; B: permutation; C: coefficient plot; D: variable importance plot).

Export to PPT

Close

Fig. 5

Total concentration of 12 difference markers in different dosage forms.

Export to PPT

When the value of variable importance was > 1, Z- ligustilide, senkyunolide A, butylidenephthalide, asarinin, rosmarinic acid, pulegone, senkyunolide H, notopterol, coniferyl ferulate, hesperidin, vanillic acid, luteolin-7-O-glucoside were selected as the different markers. All different markers were positively associated with the AAB.

The results of the concentrations of 12 different markers showed that all constituents were detected in 20 samples of powder. However, the number and concentrations of constituents in the other five dosage forms were lower than in the powder. The common active components among six dosage forms were hesperidin and rosmarinic acid, however, other components were not detected in the tablet, granules and dropping pills. In the pills and concentrated pills, notopterol was detected, while other ingredients were not detected.

The total concentration of the 12 different markers were shown in Table 4. In terms of the average value ranking, powder (20.6360 mg g⁻¹) > tablets (7.8653 mg g⁻¹) > pills (2.9006 mg g⁻¹) > concentrated pills (2.7065 mg g⁻¹) > dropping pills (2.5818 mg g⁻¹) > granules (1.7305 mg g⁻¹). To further compare the differences of the total concentration of the12 different markers among the six Chuanxiong Chatiao dosage forms, SNK with one-way ANOVA was used (Table 5). The results indicated that there were significant differences (P < 0.05) between powders and other dosage forms, and there was no significant difference among the other dosage forms.

To compare the differences of hesperidin, rosmarinic acid and notopterol among powder, concentrated pills and pills, SNK with one-way ANOVA was used (Table 5). The results indicated that there were significant differences (P < 0.05) in the content of hesperidin and rosmarinic acid between powders and the other two dosage forms, and there was no significant difference between concentrated pills and pills. In the content of notopterol, there were significant differences (P < 0.05) between concentrated pills and the other two dosage forms, and there was no significant difference between pills and powders. Based on the above discussion, it can be concluded that the pharmacodynamic substances vary among the six Chuanxiong Chatiao dosage forms.

The results revealed that numerous active components were not detected in the new dosage forms of Chuanxiong Chatiao powder, a situation potentially attributable to a variety of factors, these might include the raw materials used in the production and the production process employed. Regarding the raw materials used in the production, it’s one of crucial factors that the quality of medicinal materials varies. For instance, the content of various active ingredients varied significantly among different production regions in CX, such as Z- ligustilide, senkyunolide A, senkyunolide H, coniferyl ferulate and vanillic acid (Wang et al., 2020a, 2020b; Fang et al., 2015; Yi et al., 2016). A considerable difference was observed in the content of notopterol between different sources of QH (Ma et al., 2020). The content of active components in different parts of BH (stem, leave), differed significantly, including pulegone, rosmarinic acid, hesperidin. (Zhong, 2016; Wu et al., 2016). Furthermore, a significant variation was found in the content of luteolin-7-O-glucoside among different production regions of BH (Zhen et al., 2021). The content of pulegone in JJ was noticeably influenced by various factors, including different parts (spikes, stems, leaves), drying method, and harvest time (Yu, 2013).

In relation to the production process, Chuanxiong Chatiao powder is used as powder in medicine, which maximizes the preservation of all components in the medicinal material. However, most dosage forms are prepared by water extraction. In the process of water extraction, the solubility of liposoluble components is one of most important factors that influence the quality and efficacy of the product, because they have limited solubility in water. Furthermore, certain active components are unstable and susceptible to degradation or transformation during the water extraction process, which can result in a lower content of these active components in the final product. For example, coniferyl ferrate is prone to hydrolysis, resulting in the formation of ferulic acid and coniferol, and further oxidation of coniferol can generate vanillic acid. Z-ligustilide is susceptible to oxidation, causing it to transform into senkyunolide H and senkyunolide I, or polymerize into levistilide A. Senkyunolide A can easily undergo isomerization to generate butylphthalide (Mei et al., 2023; Zhang et al., 2019). Hesperidin is unstable and prone to degradation during high-temperature water extraction (Wang et al., 2014). Rosmarinic acid is susceptible to hydrolysis, resulting in the formation of caffeic acid (Li et al., 2017). The content of pulegone consistently decreases with increasing decoction time (Luo et al., 2016) (Fig. 6). Even though the content of asarinin gradually increases with extended decoction time, it remains significantly lower compared to that in the decoction pieces or medicinal material (Huang et al., 2012).

Close

Fig. 6

The transformation relation of active components.

Export to PPT

Furthermore, some dosage forms employ ethanol for the purpose of extracting liposoluble components, such as concentrated pill, which CX, QH, and FF were extracted using 70 % ethanol. This method aims to extract the active ingredients from the medicinal materials to the greatest extent possible. However, it should be noted that certain liposoluble ingredients, such as Z-ligustilide, senkyunolide A, etc., which are susceptible to degradation after extraction or during product storage (Zhu et al., 2013). Currently, the stability of these extracted active ingredients has not been comprehensively considered and controlled. Meanwhile, in these dosage forms, the initial step typically involves the extraction of volatile oil from BH and JJ, followed by a mixture with other extracts, which is similar to approach used to maintain the stability of active ingredient in the product during storage (Wang et al., 2012; Yu et al., 2013). Additionally, in the drying process of products, such as tablets and granules, the drying temperature has a significant impact on certain active ingredients, particularly the volatile components (Wang et al., 2012). Due to the absence of relevant quality control indicators, the product quality can easily become inconsistent. Pills, being similar to powder, but it might be influenced by the raw materials, storage and drying.

This showed that because of the issues with quality standards, the quality of the six dosage forms was not controlled comprehensively, and led to bioactive components that were different among the six dosage forms, which also caused the AABs to be different. This indicates that it is necessary for developing a comprehensive quality control method based on multiple indicators in the six Chuanxiong Chatiao dosage forms, and that optimizing the process of production would better ensure consistency when used as a clinical medication.