Analysis of the Volatile Components in Selaginella doederleinii by Headspace Solid Phase Microextraction-Gas Chromatography-Mass Spectrometry

Selaginella doederleinii (SD) is a perennial medicinal herb widely distributed in China. In this study, the volatile components of SD from two regions (24 batches), namely Zhejiang and Guizhou, were determined by combining headspace solid phase microextraction and gas chromatography-mass spectrometry (HS-SPME/GC-MS). After investigating different influence factors, the optimal conditions for extraction were as follows: The sample amount of 1 g, the polydimethylsiloxane-divinylbenzene (PDMS-DVB) fiber of 65 µm, the extraction time of 20 min, and the extraction temperature of 100 °C. Based on the above optimum conditions, 58 volatiles compounds, including 20 terpenes, 11 alkanes, 3 alcohols, 6 ketones, 3 esters, 11 aldehydes, 1 ether, 1 aromatic, 1 phenol, and 1 furan, were found and identified in SD. Furthermore, hierarchical cluster analysis and principal component analysis were successfully applied to distinguish the chemical constituents of SD from two regions. Additionally, anethol, zingerone, 2,4-di-tert-butylphenol, ledene, hexyl hexanoate, α-cadinol, phytone, hinesol, decanal, octadecene, cedren, 7-tetradecene, copaene, β-humulene, 2-butyl-2-octenal, tetradecane, cedrol, calacorene, 6-dodecanone, β-caryophyllene, 4-oxoisophorone, γ-nonanolactone, 2-pentylfuran, 1,2-epoxyhexadecane, carvacrol, n-pentadecane, diisobutyl phthalate, farnesene, n-heptadecane, linalool, 1-octen-3-ol, phytane, and β-asarone were selected as the potential markers for discriminating SD from 24 habitats in Zhejiang and Guizhou by partial least squares discrimination analysis (PLS-DA). This study revealed the differences in the components of SD from different regions, which could provide a reference for the future quality evaluation.


Introduction
Selaginella doederleinii (SD) is widely distributed in China and used as a traditional Chinese medicine [1]. Clinically, it is often applied to treat cancer such as nasopharyngeal carcinoma and esophageal cancer. The main components in SD include flavonoids, alkaloids, lignans, organic acids, and volatiles [2]. These components have been shown to have a wide range of activities. For instance: (a) Amentoflavone, robustaflavone, heveaflavone, and 7, 4 , 7 , 4 -tetra-O-methyl-amentoflavone possessed an anti-proliferation effect on human cancer cells [3]; (b) nine kinds of biflavones showed by PDMS/DVB fiber showed the largest peak area. As shown in Figure 1A, the total number of peaks obtained by PDMS/DVB fiber was also more than that obtained by the other three fibers, which proved that PDMS/DVB fiber had good adsorption and retention capacities for the extraction of volatile components from S. doederleinii. Therefore, PDMS/DVB was identified as the optimal fiber in this study. Temperature is an important parameter for the SPME extraction process. Since extraction by this method is an exothermic process, it affects the diffusion rate of the analytes into the fiber coating [23]. One gram samples were extracted with 65 µm PDMS/DVB fiber for 20 min. Volatile components were detected at 80, 90, 100, 110, and 120 • C, respectively. Each sample was repeated three times. As can be seen in Figure 1B, when the temperature reached 100 • C, the total peak area and the number of the volatile oil were the highest. Therefore, 100 • C was the optimal extraction temperature.
The time that the fiber is in contact with the sample has a significant effect on the chromatographic peak areas of the volatile oils [24]. The extraction time was optimized under the conditions of 65 µm PDMS/DVB fiber, 1.0 g sample amount, and 60 • C, and the total peak areas of the volatiles were analyzed at 5, 10, 15, 20 and 25 min (see Figure 1C). The results showed that the total peak area and the number of the volatile oil were the highest at 20 min. Therefore, the optimal extraction time was set to 20 min.
After screening the effects of PDMS/DVB fiber, extraction temperature, and time, the influence of sample amount on the extraction efficiency of the volatile oils is shown in Figure 1D. The extraction time was 20 min, and the extraction temperature was set at 100 • C. Under these conditions, different sample amounts were analyzed and optimized. As shown in Figure 1D, the sample amount of 1.0 g had a significant effect on the extraction efficiency of volatile oil. Therefore, it was appropriate to select 1.0 g as the sample amount.
Whether in SD Guizhou or SD Zhengjiang, terpenes were the highest percentage compounds and accounted for nearly half of all oil samples, which represented the chemical characteristic of SD well. The most abundant terpenes (Table 2) in SD, of which the concentrations were above 1%, were linalool (2.37-3.36%), aromadendrene (3.70-4.80%), zingerone (1.51-5.48%), cedrol (4.66-6.97%), anethol (0.56-12.78%), and α-cadinol (1.15-2.21%). The terpenes were generally considered to be active constituents in natural products [27], which exhibited antibacterial, antitumor, anti-wrinkle, antioxidative, anti-tussive, analgesia, and immune-regulatory effects [28]. For example, Gunaseelan [29] and others found the treatment of linalool prevented acute ultraviolet-B-induced hyperplasia, lipid peroxidation, and anti-oxidation loss of mice skin and could further inhibited overexpression of cyclooxygenase-2 and ornithine decarboxylase in mice skin. Zhang [30] and others first revealed that cedrol improved the level of 5-hydroxytryptamine (5-HT), decreased the ratio of 5-hydroxyindoleacetic acid/5-HT, and had significant anxiolytic effect on female mice via the 5-HT pathways.   However, the pharmacological activity of terpene from SD has not been reported. In our preliminary experiment, the volatile oils have been observed to inhibit proliferation of tumor cell and eliminate free radicals. Further study is needed to explore the bioactivities of the terpene of the volatile oils.

HCA
HCA is an analysis process that groups a collection of physical or abstract objects into multiple classes composed of similar objects [38]. It has developed rapidly in the past 20 years and has been widely used in the fields of commercial pattern recognition, plant classification, etc. [39]. As shown in Figure 4, the results of clustering analysis indicated that the 24 batches of SD samples were mainly divided into two categories: The first type was Zhejiang SD (ZJ-1 to ZJ-11); the second type was Guizhou SD (GZ-1 to GZ-13). There were some differences in the components of SD from various habitats. It might be due to some factors such as soil, climate, and harvesting season. It could be seen that the samples were separated by clustering analysis, and the differences among the samples were accurately described.
Based on the above chemometric analysis, the content of the volatile compounds showed great differences between different habitats of SD. It is deduced that the difference on SD might be related to the harvest season of the plant, the geographic regions, the processing technology of the herbal medicines, etc. [42].

Sample Preparation
The purchased herbs were cut, washed and dried naturally at room temperature. Then, the raw material of SD was crushed at a Chinese herbal medicine crusher (WK-1000A, Ruiqianshun Co. Ltd., Chongqing, China). According to a previous report [43], particle size was controlled through the 100 mesh sieve (36.358-95.719 µm), which was measured by a winner-2006 laser particle diameter analyzer (Weina technology Co. Ltd., Qingdao, China).

Sample Pretreatment
To eliminate the residue on the fiber coating, the four fibers were treated in a GC inlet at 250 • C for 5 min. A certain amount of the medicinal material was pulverized and passed through a 100 mesh sieve. The medicinal material powder was placed in a 10 mL glass vial, which was covered with a polytetrafluoroethylene (PTFE) septum and an aluminum lid. The needle was inserted into the vials. Different fibers were exposed to the headspace above the sample at a certain temperature to extract the medicinal material for a period of time. Next, the needle was transferred to the GC-MS inlet. After 2 min of thermal desorption of the fibers, the volatile oils were measured and analyzed by GC-MS.

GC-MS Analysis
An Agilent Technologies 6890 GC system (Agilent Technologies Inc., Palo Alto, CA, USA) coupled with an Agilent Technologies 5973 mass spectrometer (Agilent Technologies Inc., Palo Alto, CA, USA) was applied for volatility component analysis (Supplementary Materials). A HP-5MS capillary column (30m × 0.25mm × 0.25µm) was used to separate the volatiles. The temperature procedure was as following: 0-5 min, 50-50 • C; 5-20 min, 50-200 • C; 20-23 min, 200-200 • C; 23-27 min, 200-250 • C; 27-32 min, 250-250 • C. The splitless mode was adopted. High-purity helium was applied as the carrier gas, and the rate was 1 mL/min. The working conditions of MS were as follows: The electron ionization energy was 70 eV, the full-scan acquisition was used in the range of 10-800 amu, the ion source temperature was 230 • C, the emission current was 200 µA, and solvent delay was set at 5 min.
Retention indices (RI) of each volatile, a parameter for qualitative indicators of GC, were calculated using the retention value of two adjacent n-alkane standards. The preliminary identification of the volatiles was done by comparing the mass spectra with National Institute of Standards and Technology (NIST) 14.0 MS database and Wiley National Institute of Standards and Technology (Wiley NIST) 14.0 MS database. These volatiles were further confirmed by matching the reference literature records and retention time of authentic standards, as well as contrasting the calculated RI with that which was recorded in the NIST network database (http://webbook.nist.gov/chemistry/) [44]. Identified volatile oils were semi-quantified by comparing the peak areas of each volatile with total peak areas. Relative percentage content of the volatile compounds was calculated via the peak area normalization method using the following formula: where Mi is the percentage content of the measured volatile i; Ai is the total peak area; Ai is the peak area of the measured volatile.

Data Analysis
All experiments were carried out three times in parallel. Statistical significance (p < 0.05) was analyzed by one-way ANOVA analysis of SPSS 23.0 (SPSS Inc., Chicago, IL, USA). Principal component analysis (PCA) and partial least squares-discrimination analysis (PLS-DA) were performed by SIMCA P14 (Umetrics, Umea, Sweden). Hierarchical cluster analysis was proposed by SPSS 23.0 software (SPSS Inc., Chicago, IL, USA).

Concluding Remarks
The HS-SPME-GC-MS technique was applied to study the differences in the volatile oils of S. doederleinii from different habitats. The results showed that the optimization of the significant factors affecting sorption process such as different fibers, extraction temperature, sample amount, and extraction time, was completed by a single factor experiment design. The results showed that the PDMS/DVB fiber of 65 µm was most suitable for the isolation of the volatiles from SD. The other optimum conditions were as follows: Sample amount of 1.0 g, extraction time of 20 min, and extraction temperature of 100 • C, respectively.
Based on the optimal conditions, there were common 58 volatile substances in Guizhou SD and Zhejiang SD. Terpenes were found to comprise the largest chemical class in SD. Phytone (11.41%) and cis-anethol (12.78%) were found to be the most abundant volatile component in SD from Guizhou and Zhengjiang, respectively.
Finally, chemometric analysis indicated that the difference of SD in two provinces was very significant. The acquired data set was submitted to PCA and HCA, and the corresponding Guizhou SD and Zhejiang SD discrimination pattern according to 24 habitats was successfully established.
Moreover, based on SLDA analysis, 34 volatile constituents (VIP > 1) were important markers for the differentiation between Guizhou SD and Zhejiang SD, and quality control of SD.
In this study, it can be concluded that the producing area has an obvious influence on the contents of volatile oils of SD. Attention should be paid to the chemical differences among different habitats of SD, and their main components such as terpenes. The biological activity of components in SD needs to be further revealed. The results in this work provides a basis for the quality evaluation of SD in the future.