Quality Assessment of Artemisia rupestris L. Using Quantitative Analysis of Multi-Components by Single Marker and Fingerprint Analysis

The chromatographic fingerprint of 14 batches of Artemisia rupestris L. samples were established in this study. The constituents of ten components in Artemisia rupestris L. were determined using quantitative analysis of multi-components by single marker (QAMS) and the external standard method (ESM). Due to their stability and accessibility, chlorogenic acid and linarin were used as references to calculate the relative correction factors (RCFs) of apigenin-C-6,8-pentoside-hexoside, apigenin-C-6,8-di-pentoside, luteolin, 3,4-dicaffeoylquinic acid, 3,5-dicaffeoylquinic acid, 4,5-dicaffeoylquinic acid, chrysosplenetin B, and sbsinthin, based on high-performance liquid chromatography (HPLC). The value calculated by QAMS was consistent with that of the ESM, and the reproducibility of RCFs was found to be reliable. In conclusion, simultaneous determination of the ten components by the QAMS method and chromatographic fingerprint analysis were feasible and accurate in evaluating the quality of Artemisia rupestris L. and can be used as reference in traditional Chinese medicine quality control.


Introduction
Artemisia rupestris L. is a commonly used herbal medicine in Xinjiang for reducing fever and other symptoms of cold by acting as an anti-inflammatory and analgesic agent, as well as for detoxification and treating hepatitis [1]. Wild A. rupestris is distributed in Xinjiang, Middle Asian countries, and Northern Europe [2]. A. rupestris extracts contain sesquiterpenoids [3,4], flavonoids [5,6], alkaloids [7,8] and volatile oils [9,10].
A. rupestris has certain antiviral [11,12], anti-inflammatory activities and immune regulation [13][14][15] properties. Many studies have been conducted on the analysis of related components of A. rupestris and its preparations. In a previous study, the quality of Yizhihao capsules was assessed by quantitative analysis of rupestonic acid using the HPLC method [16]. Lan et al. [17] established an HPLC method to simultaneously determine rupestonic acidartesunone, chlorogenic acid, and luteolin in A. rupestris. Zhang Suwan et al. [18] determined the content of rupestonic acidartesunone, 6-dimethoxy-4, methyl artemisinin, and artemisinin in A rupestris using HPLC. Furthermore, Cai Xiaocui et al. [19] simultaneously determined the contents of chlorogenic and rupestonic acid, luteolin, vitexin, and apigenin in A. rupestris by liquid chromatography tandem-mass spectrometry (LC-MS/MS).
In the global market, herbal medicines are treated or dried using different methods, which may result in unstable levels of single components. Fingerprinting can comprehensively reflect the overall chemical information of traditional Chinese medicine (TCM). It is usually used for origin identification, species certification and quality control of herbal medicine, and to evaluate the authenticity, excellence and stability of the quality of traditional Chinese medicine and semi-finished products of traditional Chinese medicine preparations [20,21]. By confirming the main common fingerprint peaks in the HPLC fingerprint, the qualitative and quantitative research of traditional Chinese medicine and its preparations can be carried out to evaluate and control quality [22,23].
The determination of the content of a single component cannot be used to accurately and sufficiently evaluate the quality of A. rupestris. Therefore, an effective strategy in determining the quality of A. rupestris will be to evaluate multiple components. In this study, we established a qualitative fingerprint method for A. rupestris, and ten principal components were confirmed and analyzed quantitatively.
QAMS enables the quantitative analysis of multiple components by using a cheap and easily available standard [24]. The combination of QAMS and fingerprint method showed the convenience and economic advantages of the QAMS method, meanwhile exhibiting the integrity and comprehensive advantages of the fingerprint method. Based on the chromatographic fingerprinting method and QAMS, the qualitative and quantitative determination of A. rupestris was evaluated in the present study. Chlorogenic acid and linarin were used as internal references for phenolic acids and flavonoids, respectively, to calculate the average RCFs. The proposed method of QAMS and fingerprint analysis provides a reliable, comprehensive and efficient way for evaluating A. Rupestris quality.

HPLC Conditions
Due to the complexity of chemical constituents in A. rupestris, it is crucial to separate the target components efficiently by optimizing chromatographic conditions.
The HPLC chromatographic peaks of A. rupestris were most informative with the ultraviolet (UV) wavelength of detection set at 350 nm. Therefore, we chose 350 nm for the assay of the selected components. The mobile phase consisted of acetonitrile-0.2% phosphoric acid at 1.0 mL/min of flow rate. The gradient elution program used is described in Section 3.5 and the favorable column temperature was set at 35 • C. The extraction solvent was ethanol-water (7:3, v/v) solution, and samples were treated for 30 min by ultrasonic extraction. The samples of A. rupestris, and the mixed standard solutions containing the 10 reference substances, were analyzed under the conditions described in Section 3.5. The chromatographic peak position of the standard substances was determined as follows: chlorogenic acid, apigenin-C-6,8-pentoside-hexoside, apigenin-C-6,8-di-pentoside, luteolin, 3,4-dicaffeoylquinic acid, 3,5-dicaffeoylquinic acid, 4,5-dicaffeoylquinic acid, linarin, chrysosplenetin B, and sbsinthin ( Figure 1).

Method Validation
In order to support its application in the quantitative analysis of the ten compounds the HPLC method was validated in terms of linearity, stability, precision, and accuracy.

Method Validation
In order to support its application in the quantitative analysis of the ten compounds, the HPLC method was validated in terms of linearity, stability, precision, and accuracy.

Linearity
Within the setting concentration range, the extent of direct linear relationship between the test results and the concentration of analytes in the samples was investigated. The mixed standard solutions were serially diluted to obtain desired concentrations with methanol.
The 10 standard solutions were analyzed, and the calibration curves were formed by the relationship between the peak area of each component and corresponding concentration. The standard curve of each component was stable and the obtained linear regression equation was suitable for QAMS analysis (Table 1).

Stability
We investigated whether the sample solution was stable for 24 h storage at room temperature. The sample solution stability (S1) at room temperature (22 ± 3 • C) was tested at 0, 4, 8, 12, 16, 20, and 24 h, to obtain the RSDs. The RSD values of the stability tests were <2% (Table 1). The method could be considered stable, suggesting that within 24 h the sample solution was stable.

Accuracy
To verify the accuracy of the study, the mixed standard solutions of the analytes at low, medium, and high concentration levels were added into a sample (S1) of a certain amount (0.5 g), using six replicates. Then the mixed samples were subsequently extracted and analyzed. The average recovery of the 10 components was between 86.1-106%, and the RSDs of the accuracy values are shown in Table 2. a Intra-day precision tested six times during the same day. b Inter-day precision tested on three separate days.

Precision
To evaluate the intra-day and inter-day precision, the sample solution was analyzed within a single assay day and on three separate days for six replicates at each concentration level, respectively. The RSDs of intra-day and inter-day were all less than 3% ( Table 2).

Fingerprints Analysis
Traditional Chinese medicine fingerprint is a comprehensive and quantifiable identification method, and is mainly used to evaluate the authenticity and the quality of traditional Chinese medicine and semi-finished products of traditional Chinese medicine preparations. For establishment of a novel method for multiple components from A. rupestris, high-performance liquid chromatograms of each sample were imported into the software recommended by SFDA, which was called chromatographic fingerprint similarity evaluation system for traditional Chinese medicine (version 2012a). Then, the chromatographic fingerprints were collected. Among the 14 samples, S1 was selected as the reference spectrum, while 18 common fingerprint peaks were recorded in the chromatograms ( Figure 2). Meanwhile, the similarity of the 14 batches of A. rupestris samples was analyzed and obtained (Table 3). S11 gave the lowest similarity among the samples, and the simiarity of S1, S2, S3, S5, S6, S10, S14 were all above 0.8. According to the software requirements, similarity greater than 0.9 is generally required. In the similarity analysis, only 3 batches of samples were greater than 0.9, which were S1, S2 and S5, suggesting that the quality of medicinal materials from different producing areas and growth conditions is quite different.

Quantitative Analysis by QAMS and ESM
Ten components were identified in the fingerprint with reference materials, including chlorogenic acid, apigenin-C-6,8-pentoside-hexoside, apigenin-C-6,8-di-pentoside, luteolin, 3,4-dicaffeoylquinic acid, 3,5-dicaffeoylquinic acid, 4,5-dicaffeoylquinic acid, linarin, chrysosplenetin B, and sbsinthin. ESM and QAMS were used to quantify the ten components in the sample to verify consistency between QAMS and ESM. In ESM, the mixed standard solution and sample solutions were analyzed by liquid chromatography, and the content of each component calculated according to the regression equations listed in Table 1.
Choosing a suitable standard for internal reference using the QAMS method for analyzing multiple components in traditional Chinese medicinal substances is important. The component selected as internal reference should be selected on grounds of ease of acquisition, low price, and stable properties, and should be separable from the other compounds under chromatographic conditions [25]. In the present study, chlorogenic acid and linarin were used as internal references for phenolic acids and flavonoids. respectively. QAMS calculates the RCF between the component which was selected as internal reference and other components in medicinal materials. Furthermore, by calculating the amounts of other components through RCF, the simultaneous determination of multiple components [24,26] can be accomplished. The deviations (RE) between QAMS and ESM were calculated using the following formula (Equation (1)).

RCF = (C i /A i )/(C s /A s )
where A s is the peak area of the internal reference substance, C s is the concentration of the internal reference substance, A i is the peak area of the component to be tested, and C i is the concentration of the component to be tested. To analyze the 10 components simultaneously in A. rupestris by QAMS, the RCFs were calculated based on the ratio of peak area and corresponding concentration between internal references and the other analytes. The RCFs are shown in Tables 4 and 5.  Overall, 14 batches of A. rupestris samples from various production areas were analyzed using the validated ESM and QAMS methods. The deviations (RE) between QAMS and ESM were calculated using the following formula (Equation (1)).

RE = (QAMS
The quantitative results of the 10 compounds in A. rupestris calculated by ESM and QAMS methods are shown in Tables 6 and 7. The REs were less than 5%, which is the requirement of Chinese Pharmacopoeia. It was indicated that there was no significant difference in the content results obtained by QAMS and ESM Table 6. Contents of Phenolic Acids in A. rupestris using different methods (mg/g). Analysis of the 14 batches of A. rupestris samples revealed the contents of chlorogenic acid, apigenin-C-6,8-pentoside-hexoside, apigenin-C-6,8-di-pentoside, luteolin, 3,4dicaffeoylquinic acid, 3,5-dicaffeoylquinic acid, 4,5-dicaffeoylquinic acid, linarin, chrysosplenetin B, and sbsinthin. The determination acquired by QAMS was consistent with that of ESM. Therefore, it was proven that the QAMS method for simultaneous quantitative analysis of the 10 components in A. rupestris was reliable and feasible.

Chlorogenic Acid 3,4-Dicaffeoylquinic Acid 3,5-Dicaffeoylquinic Acid 4,5-Dicaffeoylquinic
Fingerprints analysis and quantitative data showed significant differences among various samples from different sources. According to the results in Tables 6 and 7, it can be seen that the content of ten components fluctuates greatly between batches. It may be related to the production environment, planting methods, growth time and harvest season. Among phenolic acids, the contents of chlorogenic acid and 3,5-dicaffeoylquinic acid were relatively high, and there was positive correlation. Among the flavonoids, the contents of linarin and chrysosplenetin B were relatively high. Next, we studied the relationship between the content of each component and the related efficacy.

Cluster Analysis
All the 14 samples of A. rupestris were selected from different regions or planting methods, and the content of their 10 main compounds might differ. Our QAMS methods could accurately analyze the composition, and the contents of the 10 compounds in the 14 batches of A. rupestris samples selected as clustering variable. The samples were divided into two categories by cluster analysis. S2, S3, and S4 are one category, which were artificially planted in Fuyun County of Altay Region and the other samples formed one category, to ascertain whether artificial planting or origins would change the composition of A. rupestris.

Plant Material
A. rupestris samples used in the study were collected from different production areas, as shown in Table 8.  China). Apigenin-C-6,8-pentoside-hexoside and apigenin-C-6,8-di-pentoside were determined based on their spectral structure, and their purities were calculated by the peak area normalization method (purity HPLC ≥ 98.0%). Acetonitrile and methanol (Thermo Fisher Scientific, Inc. Shanghai, China) were of HPLC grade. Phosphoric acid (YSHC Chemical Company Limited, Tianjin, China) and absolute ethanol (Tianjin Xinbote Chemical Company Limited, Tianjin, China) were analytical grade. Ultrapure water was prepared by a Milli-QAC SP Reagent Water System (Millipore Corporation, Billerica, MA, USA). Other chemicals used in the study were all analytical grade.

Procedure of Sample Solution Preparation
The A. rupestris samples were ground using a high-speed traditional Chinese medicine pulverizer to make powder, and passed through a 10-mesh sieve. Next, 0.5 g of sample was weighed precisely into a bottle with a plug. The sample powder was extracted with 30 mL of the extraction solvent (ethanol:water, 7:3, v/v). The mixed solution was sonicated for 30 min (250 W, 40 kHz) at room temperature (25 ± 5 • C). After cooling, additional extraction solvent was added to the sample solution to compensate for weight loss, followed by thorough shaking. Before injection, the supernatant of the sample solution was filtered through membranes of 0.22 µm.

Instruments and Chromatographic Procedures
An Agilent 1260 series high-performance liquid chromatograph, equipped with a quaternary pump, UV detection, an autosampler, a column temperature controller, and a vacuum degasser (Agilent Technologies, Palo Alto, CA, USA) was used in the HPLC assay of A. rupestris. The samples were separated using a Hypersil GOLD C 18 column (5 µm, 250 × 4.6 mm). The injection volume of sample and standards were all 10 µL, and the column temperature was maintained at 35 • C. A mixture of acetonitrile (A) and 0.2% phosphate solution (B) was used as the mobile phase at a flow rate of 1.0 mL/min. The gradient elution mode was modified as follows: 0-15 min, 5-18% A; 30 min, 20% A; 35 min, 21% A; 40 min, 40% A; 47 min, 45% A; 51-55 min, 80% A; 55.1-65.0 min, 5% A. The detection wavelength was sat at 350 nm.

Conclusions
A novel method to determine multiple components from A. rupestris was established and reported using QAMS in this study, which was reproducible and accurate. The A. rupestris samples from different regions or planting methods were determined by HPLC, and the content determination of ten components (chlorogenic acid, apigenin-C-6,8-pentoside-hexoside, apigenin-C-6,8-di-pentoside, luteolin, 3,4-dicaffeoylquinic acid, 3,5-dicaffeoylquinic acid, 4,5-dicaffeoylquinic acid, linarin, chrysosplenetin B, and sbsinthin in different laboratories) in the A. rupestris samples was simultaneously accomplished by the proposed QAMS, applying an exclusive identification and evaluation method for qualitative and quantitative analysis of A. rupestris.
The chromatographic fingerprint showed the details of A. rupestris chromatographic spectrum, while the cluster analysis identified that region and growth conditions could influence the content of A. rupestris. Therefore, this method might be suitable both for A. rupestris quantitative analysis and for quality examination. The similarity analysis suggested that the quality of medicinal materials from different producing areas and growth conditions is quite different.

Abbreviations
The following abbreviations have been used in this manuscript. HPLC High performance liquid chromatography QAMS Qualitative and quantitative analysis of multi-component by single marker ESM External standard method RSD Relative standard deviation TCM Traditional Chinese medicine RCF Relative correction factor