Study on the Influence of Sulfur Fumigation on Chemical Constituents of Angelicae dahuricae Radix (Baizhi)

Purpose: To study the influence of different sulfur fumigation time and dosage on the chemical constituents of Baizhi (Angelicae dahurica Radix). Methods: The relationship of chemical constituents in Baizhi with different sulfur fumigation time and dosages was evaluated by high performance liquid chromatography (HPLC) fingerprint and chemometrics methods, including similarity analysis (SA), hierarchical clustering analysis (HCA) and principal component analysis (PCA). HPLC was carried out on a Diamonil C18 column; linear gradient elution was performed with acetonitrile and water; column temperature, injection volume of sample, flow rate, detection wavelength and testing time were 30 °C, 20 μL, 1 mL/min, 310 nm and 75 min respectively. Results: The chemical constituents of Baizhi significantly (p < 0.05) decreased after sulfur fumigation though sulfur fumigation time and dosage were at low levels – 2 h and 25 g (sulphur)/10 kg (Baizhi), respectively. Conclusion: desirable method for field processing of Baizhi even when sulfur fumigation time and dosage were short and small.


INTRODUCTION
Angelicae dahurica Radix, a common used Chinese medicine, has extensive pharmacological activities such as spasmolysis, analgesia, relieving asthma, etc. [1]. Currently, the problem about field processing of Baizhi has aroused wide attention. The medicinal part of Baizhi is fleshy root which is hard to timely dry and perishable [2]. Sulfur fumigation, typically used in the field processing of Baizhi all over China, can extend storage time and improve the appearance of traditional Chinese medicines (TCM) due to its preservative and bleaching effects. [3]. While a large number of studies have indicated that sulfur fumigation can not only cause the loss of biological active components but also create harmful residues of sulfur dioxide in TCM [3][4][5][6][7][8], and the negative impact of sulfur fumigation has aroused researchers' attention. In addition, it was found in our earlier study that active ingredients (coumarins) and pharmacological effects of Baizhi were evidently decreased after sulfur fumigation [9,10]. Previous studies mainly focus on the influence of the same sulfur fumigation time and dosages on Baizhi, and neglect the influence of different sulfur fumigation time and dosages on Baizhi. Moreover, there is a heated debate about whether the chemical constituents of Baizhi were decreased by sulfur fumigation with short time and low dosages in the Good Agricultural Practices base (GAPB) of Baizhi, Sichuan Suining, China.
Our team designed a series of Baizhi samples, fumigated at varying sulfur fumigation time and sulfur doses, to study whether the chemical constituents of Baizhi were not damaged by sulfur fumigation with a short time and low dosages via HPLC fingerprint and chemo metrics methods such as similarity analysis (SA), hierarchical clustering analysis (HCA) and principal component analysis (PCA).

Reagents
Methanol, tetrahydrofuran and acetonitrile were purchased from Fisher (HPLC grade, Fisher Scientific, Germany). Ultra-pure water was prepared by Milli-Q Advantage A10 (Millipore Crop., MA, USA). All other reagents were analytical grade.

Extracts of Baizhi
After processed Baizhi samples were ground into powder, twenty batches of samples were accurately weighed (about 0.5 g) and ultrasonicextracted with 45 mL of methanol for 1 h at room temperature, and then the extract was filtered through a 0.22 μm membrane before analysis.

Preparation of mixing standard solution
Appropriate quantities of oxypeucedanin hydrate, byakangelicin, bergapten, isopimpinellin, neobyakangelicol, oxypeucedanin, byakangelicol, phellopterin, imperatorin and isoimperatorin were dissolved in methanol to produce mixed standard solutions, stored at 4 °C. The mixed standard solutions need to be standing to room temperature and then were filtered through a 0.22 μm membrane filter before analysis.

HPLC fingerprint method validation
The feasibility analysis of the HPLC fingerprint method of determining processed Baizhi samples was evaluated by precision, stability and reproducibility experiments which can be conducted though HPLC analysis mentioned above to determine chemical constituents in Baizhi samples at different times. Common peaks in HPLC fingerprint are defined as the chromatographic peaks which are appeared in all the samples. The relative standard deviation (RSD) values of retention time and peak area of common peaks with respect to the reference peak (imperatorin) were calculated to verify the precision, stability and reproducibility of the developed method.

Data analysis
Similarity analysis (SA), performed by Similarity Evaluation System for Chromatographic Fingerprint of TCM (Version 2004A), was used to establish characteristic fingerprint and calculate similarity among samples. Hierarchical clustering analysis (HCA) and principal component analysis (PCA) , performed by Statistical Product and Service Solution 19.0 (SPSS 19.0), were used to establish clusters and comprehensive evaluation of samples of various field processing methods. All peak data were represented as mean ± standard deviation (n = 5), which were analyzed by the two-tailed Student's t-test. Differences were considered statistically significant at p < 0.05.

Optimization of chromatographic conditions
The influence of mobile phases including methanol-water, methanol-tetrahydrofuran-water and acetonitrile-water, columns containing Diamonil (C18 4.6 mm × 250 mm, 5 μm) and Dikma (C18 4.6 mm × 250 mm, 5 μm), column temperatures including 25 °C and 30 °C on theoretical plates, resolution, reproducibility and tailing factor were studied, and the results indicated that acetonitrile-water, Diamonil C18 column (4.6 mm × 250 mm, 5 μm) and 30 °C column temperature could obtain desirable theoretical plates, resolution, reproducibility and tailing factor. In addition, the detection wavelength was identified as 310 nm by DAD three-dimensional map.

Optimization of sample preparation process
In this study, the influences of four types of extraction solvents (ethyl ether, ethyl acetate, 95 % ethanol and methanol) and three types of extraction methods (ultrasonic, reflux and cold soak) on extraction efficiency were investigated, and the results suggested that methanol and ultrasonic extraction was the best extraction solvent and method. In addition, the influences of other extracting conditions including extraction solvent volumes (30, 45 or 60 mL) and extraction time (30, 45, 60 or 75 min) on extraction efficiency also were explored. All the results suggested that ultrasonic with methanol (45 mL) for 75 min was the simplest and most efficient method for extraction of sample.

HPLC fingerprint method validation
The RSD of relative retention time and relative peak areas of common peaks were less than 1 and 5 percent separately, which were in conformity with the guideline of technology requirements about the research of TCM injection fingerprint (provisional, the state administration of TCM), and thus the developed HPLC fingerprint method was feasible in terms of precision, stability and reproducibility.

Identification of chromatographic peaks
Ten chromatographic peaks of Baizhi sample was identified by comparing retention time and UV absorption curve of samples with mixed standard solution. The results are presented in Figure 1 and Table 2.  Figure 2.

Similarity analysis (SA)
The similarities of 18 batches of Baizhi samples were evaluated by the correlation coefficient (median) and time window (0. 5

) in Similarity Evaluation
System for Chromatographic Fingerprint of TCM (Version 2004 A). The results are displayed in Table 3.     (Table  4) show that the contribution of PC1 was maximal among four components, which could represent the main original information and its formula was described in Eq 1 in which BXi stands for standardized variable of i variable. The coefficients of five peaks [peak 8 (isopimpinellin), 9, 11 (oxypeucedanin), 14 (imperatorin) and 15 (phellopterin)] were higher than 0.90, and these indicated that the changes of the five peak areas were the main change among samples. The five peak areas are listed in Table 5.

DISCUSSION
In this work, chromatographic conditions, sample preparation process and HPLC-DAD method were optimized and validated to achieve a good separation of chemical constituents in Baizhi. Nineteen common peaks obtained by HPLC fingerprint analysis of Baizhi samples were used to analyze the influence of different sulfur fumigation time and dosages on chemical constituents' changes of Baizhi by SA, HCA and PCA. In addition, the 10 peaks of 19 common peaks were identified by standard substances.
The results of SA (Table 3) showed that the difference of similarities among 18 batches of Baizhi samples, processed with different sulfur fumigation time and dosages, were rather small, indicating that the influences of different sulfur fumigation time and dosages on the chemical  composition of Baizhi were similar. In addition, the results of HCA (Figure 3) suggested that the 20 batches of samples could be classified into two clusters when Euclid Distance was 20.
Cluster I included 18 samples processed with different sulfur fumigation time and dosages and Cluster II contained two samples without sulfur fumigation (S19 and S20). It implied that the content of chemical constituents of Baizhi could be significantly decreased by sulfur fumigation (p < 0.05) though sulfur fumigation time and dosages were in low levels, 25 g (sulphur)/10 kg (Baizhi) (Sample S1) and 2 h (Sample S9). PCA had further verified the above viewpoint, and the influences of sulfur fumigation on the chemical composition of Baizhi were positively related to sulfur fumigation time and dosages according to the change of the main five peak areas (peak 8, 9, 11, 14 and 15).

CONCLUSION
The findings of this study indicate that the chemical constituents of Baizhi could be significantly decreased by sulfur fumigation even at low sulfur fumigation time and doses, suggesting that the sulfur fumigation is not a desirable method for field processing of Baizhi. It is therefore necessary to explore new field processing procedures that would not only improve the appearance of TCM but also extend the storage time of TCM, as well as prevent a reduction of its effective constituents in Baizhi.