Rapid Classification and Identification of Chemical Components of Schisandra Chinensis by UPLC-Q-TOF/MS Combined with Data Post-Processing

Schisandra chinensis (known in Chinese as WuWeiZi, WWZ) has observable effects such as astringing the lung to stop coughs, arresting sweating, preserving semen and preventing diarrhea. The major components of WWZ include lignans, triterpenoids, organic acids and fatty acids. In this paper, a reliable method for the rapid identification of multiple components in WWZ by their characteristic fragments and neutral losses using UPLC-Q-TOF/MS technology was developed. After review of the literature and some reference experiments, the fragmentation pattern of several compounds were studied and summarized. Then, according to the corresponding characteristic fragments coupled with neutral losses in the positive or negative ion mode produced by different types of substances a rapid identification of target compounds was achieved. Finally, a total of 30 constituents of WWZ were successfully identified, including 15 lignans, nine triterpenoids, three organic acids and three fatty acids. The method established in this study not only provides a comprehensive analysis of the chemical ingredients of WWZ, providing a basis for further phytochemical studies on WWZ but also provides a more efficient way to solve the problem of identification of complex chemical constituents in traditional Chinese medicines.


Introduction
Schisandra chinensis (the dried mature fruits of Schisandra chinensis (Turcz.) Baill or Schisandra sphenanthera Rehd et Wils) is commonly used as a tonic, sedative and an anti-aging drug in clinical practice, for it has good anti-inflammatory, anti-oxidative, immunomodulatory and anti-hepatic injury effects [1,2].The main components in Schisandra chinensis (named WuWeiZi, WWZ) contain lignans, triterpenoids, organic acids and fatty acids, volatile oils, and sugars. Many studies have shown that lignans, which have various biological activities such as antihepatotoxic, antioxidant and detoxificant effects, are the major bioactive constituents of WWZ [3][4][5], able to inhibit hepato-carcinogenesis and enhance human intellectual activity [6]. Triterpenoids show anti-HIV effects and inhibitory activities toward cholesterol biosynthesis [7]. Organic acids show antibiotic pharmacological activities [8], while fatty acids have been associated with anti-inflammatory, immunomodulatory, analgesic and anti-tumour properties [9]. Compared to the single compounds of Western medicine, the efficacy of traditional Chinese medicines (TCMs) is characterized for its integrity and synergy resulting from the various components. However, according to current research results, the qualitative identification of WWZ is mainly concentrated on the lignans [10,11]. Therefore, it is not only necessary to develop multi-component fingerprints by UPLC-Q-TOF/MS, but more importantly to establish a rapid and reliable method for the identification of different types of constituents in WWZ. Many methods for analysing chemical ingredients of WWZ have been reported, including thin layer chromatography (TLC) [12], high-performance liquid chromatography (HPLC) [13], capillary electrophoresis (CE) [14], high speed countercurrent chromatography (HSCCC) [15], gas chromatography-mass spectrometry (GC/MS) [16], among others. Nowadays, ultra-performance liquid chromatography/time-of-flight mass spectrometry (UPLC-Q-TOF/MS) is widely used [17,18]. With the rapid development of modern analysis technology and data post-processing technology, characteristic fragments filter (CFF) and neutral loss filter (NLF) represent two data post-processing technology methods that can be used for the in-depth study of TCM components [19,20]. Under identical MS conditions, compounds with same or similar parent nuclei could be split into different fragments. Characteristic fragment ions (CFs) refer to the fragments that could be used to infer the cleavage types and the classification of substances, which helps to screen the target components and filter congeners. Moreover, neutral fragments (NFs) lost during cleavage processes also play an important role in screening substances, as they reflect the discrepancy of the m/z between the parent ion and the fragment ions in high mass-to-charge ratio portions [21][22][23][24][25][26]. Therefore, it is propitious to realize the rapid screening and qualitative analysis of constituents of TCM combined with data post-processing techniques (CFF and NLF) on the basis of UPLC-Q-TOF/MS.
The rapid identification and analysis of the components are helpful to further develop pharmacodynamic material basis and determine the functional mechanism(s) of TCMs. Therefore, analytical methods should be developed to rapidly classify and identify complex TCM components. This paper is mainly based on the UPLC-Q-TOF/MS method. Firstly, the mass spectrometry information in regard to the components of WWZ extract were classified and summarized after a review of the literature and some reference experiments results. According to the above results, it can be speculated that fragmentation regularity exists in the mass spectra of several different constituents of WWZ. Then, the CF and NL rules of lignans, triterpenoids, organic acids and fatty acids were established and generalized. Then, original mass spectral information was obtained and processed by the Masslynx version 4.1 software (Waters, Milford, UK) to detect and align the peaks. Finally, the identities of these compounds were determined on the basis of mass information in combination with different CFs and NLs. In this study, more abundant mass spectrometry information was acquired by scanning in the positive and negative ion mode, and in combination with a characteristic post-processing technique for information integration, several compounds could be classified and identified. To a certain degree, the method provides an effective way for solving the key issue of rapid classification and identification of complex chemical components in TCMs, and also a foundation for controlling the quality of different batches of original herbs.

The Establishment of Data Post-Processing Technology Based on UPLC-Q-TOF/MS
In order to obtain further material information of the fingerprint, based on the physical structure and chemical characteristics of components in WWZ; a BEH C18 column was eventually used. Good resolution with high and narrow peaks were obtained at a column temperature of 35 • C, flow rate of 0.4 mL·min −1 , injection volume of 2 µL and with a mobile phase consisted of (A) 0.1% formic acid in water and (B) acetonitrile containing 0.1% formic acid. Additionally, the WWZ extract was comprehensively analysed under positive and negative ion modes. Typical BPI chromatograms of the substances in the WWZ extract in both modes are shown in Figure 1. In this study, WWZ extract was selected for the experiments. Firstly, according to the existing literature, the fragmentation information about the lignans, triterpenoids, organic acids and fatty acids as main components in WWZ was collected. Then, the fragmentation rules of CFs and NLs with regard to the above four types of components were inferred and are listed in Table 1, and the ion information of the constituents could be obtained using the Masslynx software. Substances can be classified by NLs, that refers to the mass difference between molecular ion peaks and high mass-to-charge ratio fragment peaks, combined with the CFs of other subtypes to determine the type of compound. Finally, a method was established for rapid identification and classification of chemical compositions in WWZ extract by using CFF and NLF. Lignans are generally divided into biphenyl cyclooctene lignans and open-loop lignans. In line with the fragmentation information obtained in mass spectrometry from the literature and the total ion flow mass spectrometry information, biphenyl cyclooctene lignans are characterised by CFs at m/z 415 In this study, WWZ extract was selected for the experiments. Firstly, according to the existing literature, the fragmentation information about the lignans, triterpenoids, organic acids and fatty acids as main components in WWZ was collected. Then, the fragmentation rules of CFs and NLs with regard to the above four types of components were inferred and are listed in Table 1, and the ion information of the constituents could be obtained using the Masslynx software. Substances can be classified by NLs, that refers to the mass difference between molecular ion peaks and high mass-to-charge ratio fragment peaks, combined with the CFs of other subtypes to determine the type of compound. Finally, a method was established for rapid identification and classification of chemical compositions in WWZ extract by using CFF and NLF. Lignans are generally divided into biphenyl cyclooctene lignans and open-loop lignans. In line with the fragmentation information ] + . Combined with the CFs and NLs of other subtypes, this type of compound could be rapidly determined among the many components. The biphenyl cyclooctene lignans are divided into four categories according to their structure and different connection groups. There are type 1 biphenyl cyclooctene lignans with closed loops (without OH), type 2 biphenyl cyclooctene lignans (with OH), type 3 biphenyl cyclooctene lignans (with OH and benzoyl), type 4 biphenyl cyclooctene lignans (with OH and angeloyl or tigloyl moieties). The loss of neutral molecules occurs through energy collisions. For example, type 3 biphenyl cyclooctene lignans (with OH and benzoyl) have a symmetrical structure with benzoyl groups, so the neutral loss of a benzoyl at 122 Da (C 6 H 5 COOH) are one of the major fragmentation pathways obtained by collision-induced dissociation. According to the loss of molecules including 18 Da (H 2 O), 122 Da (C 6 H 5 COOH), 30 Da (CH 2 O), the unknown component could be further identified as type 3 biphenyl cyclooctene lignans.

Lignans
Lignans are the main active ingredients in WWZ fruits. Based on the parent nucleus structure lignans were divided into biphenyl cyclooctene lignans and open-loop lignans. According to their fragmentions in multi-stage spectra, characteristic dissociation rules were obtained, which can be summarized as follows: firstly, the fragment ions 415 [C 24  ] + . Additionally, biphenyl cyclooctene lignans were divided into four classes consisting of type 1 biphenyl cyclooctene lignans with closed loop (without OH), type 2 biphenyl cyclooctene lignans (with OH), type 3 biphenyl cyclooctene lignans (with OH and benzoyl) and type 4 biphenyl cyclooctene lignans (with OH and angeloyl or tigloyl units) on the basis of their structure and different groups. NLs often reveal the information related to the categories of compounds and also play an important role in the rapid identification of compounds. Type 1 biphenyl cyclooctene lignans easily lose neutral fragments of m/z 70 Da (C 5 H 10 ) and 56 Da (C 4 H 8 ) to produce five membered rings or six membered rings formed from the eight membered ring. While type 1 of the parent nucleus is not connected with hydroxyl groups, based on the characteristic NLs at m/z 18 Da (H 2 O), three other classes and type 1 were distinguished. NLs at m/z 122 Da (C 6 H 5 COOH) and m/z 100 Da (C 4 H 7 COOH) are the key to distinguishing between type 2 and type 3, type 4. Additionally, type 3 and type 4 exhibited benzoyl and angeloyl or tigloyl structures; 122 Da (C 6 H 5 COOH) and 100 Da (C 4 H 7 COOH) peaks are characteristic NLs of type 3 and type 4 respectively. Lignans show adduct ions [M + Na] + in positive mode; however, without signals in the negative ion mode [6,10,11,[27][28][29][30][31][32][33]. Thus, substances in the WWZ extract were rapidly classified and identified using CFF combined with NLF.  [10,27,28] (for more details, see Table 2). The specific fragmentation pathways for deoxyschisandrin are shown in Figure 2. . This fragmentation information combined with reference data allowed the compound to be identified as schisandrol A [11,29,30], and its cleavage pathways are shown in Figure 3.  . This fragmentation information combined with reference data allowed the compound to be identified as schisandrol A [11,29,30], and its cleavage pathways are shown in Figure 3.    The presence of the 415 [M + H − C 6 H 5 COOH] + ion indicated a benzoyloxy group in the structure. The fragmention at m/z 537 [M + H] + was a molecular ion, which resulted in a fragment ion at m/z 560 [M + H + Na] + . Therefore, comparing the fragment ions rules from the literature [10,29], compound 8 was determined to be schisantherin A. The specific fragmentation pathways of schisantherin A are shown in Figure 4.

Triterpenoids
Triterpenoids, according to their chemical structure, can be divided into three types: lanostane-type triterpenoids, cycloartane-type triterpenoids, and schisanra-type triterpenoids. Firstly, lanostane-type triterpenoids usually have carboxyl groups, and easily lose 46 Da (HCOOH) and 45 Da (HCOO − ) molecules. Secondly, based on the relative abundance of the fragments at m/z 312 [C 22 H 32 O] + , 271 [C 19 H 27 O] + in the MS data, the components which belong to cycloartane-type triterpenoids can be identified. Finally, schisanra-type triterpenoids differ from the other two types, and could be identified by their typical NLs, including 60 Da (CH 2 C(OH) 2 ) and 74 Da (CH 3 CH=C(OH) 2 ) that appeared in the spectra. Thus, based on the CFs and NLs we could rapidly determine the components which belong to the different types of triterpenoids [33][34][35].  [35]. A proposed mechanistic pathway for fragments formed in negative ion mode is shown in Figure 5. Thus, compound 15 was determined as sohisanlaotone D by combining the fragmentation rules with literature data [35]. A pathway for the fragments formed in the MS is shown in Figure 6.  Table 1. Therefore, compound 24 was determined as schindilactone A [33]. The proposed fragmentation pathways of schisantherin A are shown in Figure 7.   Table 1. Therefore, compound 24 was determined as schindilactone A [33]. The proposed fragmentation pathways of schisantherin A are shown in Figure 7.

Sample Preparation
The dried WWZ fruits were pulverized and then the coarse powder (10 g) was accurately weighed and extracted twice with four times the amount of 85% ethanol (made up of 34 mL of ethanol and 6 mL of water). Each reflux time lasted 3 h at room temperature (about 25 °C). The combination of the two extracts was filtered, and the filtrate was subsequently passed through a 0.22 µm membrane and then 2 µL of the sample was injected into the UPLC-Q-TOF/MS for constituent analysis [28,39].   [36]. For more details, see Table 3.

Organic Acids
Organic acids play a major role in antibiotics [8]. According to our literature review this kind of compound strongly loses acid groups [37], besides, organic acids easily produce H 2 O and CO 2 groups in negative ion mode. Thus, this kind of compound was identified on the basis of the parention fragments  [37,38], which formula is C 6 H 8 O 7 , shown in Table 3.  [10,11,29]

Sample Preparation
The dried WWZ fruits were pulverized and then the coarse powder (10 g) was accurately weighed and extracted twice with four times the amount of 85% ethanol (made up of 34 mL of ethanol and 6 mL of water). Each reflux time lasted 3 h at room temperature (about 25 • C). The combination of the two extracts was filtered, and the filtrate was subsequently passed through a 0.22 µm membrane and then 2 µL of the sample was injected into the UPLC-Q-TOF/MS for constituent analysis [28,39].

Conclusions
In this study, based on a powerful integrated approach that UPLC-Q-TOF/MS combined with data post-processing technology, 30 compounds were screened successfully (for further information, see Tables 2 and 3). The structures of compounds were confirmed by using their MS fragments and by comparison with reference standards and corresponding reference data on the basis of the typical cleavage pathways of four chemical constituent classes in WWZ. In addition to the main component lignans found in WWZ, it also contains triterpenoids, organic acids and fatty acids. The method described in this report has high resolution and sensitivity that reduces difficulty of identification of complicated and diverse components in WWZ fruits, and laid the foundation for study on pharmacological and pharmacokinetics in WWZ fruits. To some extent, it made up for the deficiency of the existing analytic methods for TCMs. Furthermore, the novel strategy was a powerful tool for the systematic screening and identification of quality control and chemical analysis of TCM, and promoted the development of TCM.