Comparison of volatile compounds in different parts of fresh Amomum villosum Lour. from different geographical areas using cryogenic grinding combined HS–SPME–GC–MS

Background The essential oil is one of the main active ingredients of Amomum villosum Lour. However, volatile compounds are easily lost during the drying, storage and even sample preparation procedure. Therefore, using fresh samples can obtain more accurately data for qualitative and comparative analysis. Methods In this study, the volatile compounds in different parts of fresh A. villosum from different origins were systemic analyzed and compared by using cryogenic grinding combined HS–SPME–GC–MS for the first time. GC–MS analyses were performed on a 6890 Series GC instrument coupled to a 5973 N mass spectrometer. The volatile compounds were extracted by the SPME fiber (100 μm PDMS). Analytes separation was achieved on a HP-5MS capillary column. The oven temperature was initially programmed at 70 °C, then raised 4 °C/min to reach 125 °C and then programmed at 0.5 °C/min to 133 °C, then at 6 °C/min to 170 °C and finally, at 20 °C/min to 280 °C held for 2 min. The temperatures of the injection port, ion source and transfer line were set at 250 °C, 230 °C and 280 °C, respectively. Results Forty-eight main compounds were identified in different parts of fresh A. villosum. The most abundant components in fresh fruit samples were camphor (3.91%), bornyl acetate (10.53%), caryophyllene (8.70%), β-bisabolene (11.50%), (E)-nerolidol (14.82%) and cubenol (10.04%). This is quite different with that of dried samples analyzed in our previous work. As different parts of the same plant, many common components with biological activities were detected in fruit and other parts. In principle components analysis (PCA) and hierarchical clustering analysis (HCA), four parts of A. villosum were divided into different groups clearly. Additionally, fruit and root samples also could be divided into two subgroups (HCA) in accordance with their regions. Conclusion The developed method was successfully used for qualitative and comparative analysis of volatile compounds in fresh A. villosum samples. Additionally, using fresh samples can obtain much more information which is helpful for their performance in the fields of functional foods, agriculture and biomedical industry. Furthermore, our research is helpful for comprehensive utilization and quality control of A. villosum.


Introduction
Amomum villosum Lour., belongs to the Zingiberaceae family, is wildly cultivated in southern China, especially in Yangchun, Guangdong province (111° 16′ 27″ to 112° 09′ 22″ E, 21° 50′ 36″ to 22° 41′ 01″ N) [1,2]. The dried ripe fruit of A. villosum, also named Amomi Fructus, has been used as traditional medicine to treat gastritis, stomachache, and digestive diseases for hundreds of years [3]. Recently, Amomi Fructus is also authorized as a food by the China Food and Drug Administration [4]. As the main active ingredient in Amomi Fructus, essential oil possesses many biological activities such as antimicrobial, anti-inflammation and analgesic effects [5][6][7]. Due to the excellent functions and special aroma of its oil, Amomi Fructus presents great economic values in the functional foods and biomedical industry. Nowadays, it has been used as natural raw materials for wine, tea, candies and cosmetics. Additionally, Amomi Fructus is also one of the food spices in China [8]. With the development of Amomi Fructus related industry, the natural resource of Amomi Fructus is hard to meet the increasing demand [1,9]. Generally, different parts of the plant have similar secondary metabolites. Expand the application parts are helpful for making up the shortages of this resource. Therefore, qualitative and comparative analysis of the essential oil in different parts of A. villosum from different geographical areas is very necessary, which is beneficial for comprehensive utilization of this these resources as well as quality control of A. villosum.
The chemical composition of essential oil in A. villosum has been reported by several papers [10][11][12][13][14]. However, due to the fresh A. villosum samples is difficult to collect and store, most of these studies were focused on dried samples. Actually, volatile compounds are easily lost during the drying, storage and even sample preparation procedure. Therefore, using fresh samples can obtain more accurately data for qualitative and comparative analysis. Conventional techniques such as steam distillation and supercritical CO 2 extraction coupled with gas chromatography-mass spectrometry (GC-MS) have been employed to investigate the essential oils in fresh plants [15][16][17][18]. However, these techniques either need a tedious and time-consuming sample preparation procedure or require special and expensive equipment. Headspace solid-phase microextraction (HS-SPME) is a convenient alternative to traditional essential oil extraction methods. It allows extraction and simultaneous concentration of analytes in one single step. Furthermore, no organic solvents were needed in the whole extraction process [19,20]. In recent years, HS-SPME coupled GC-MS has gained wide applications in volatile compounds analysis especially in case on natural plants [21][22][23][24].
The sample preparation methods for HS-SPME-GC-MS analysis of herbs usually contain pulping, milling and so on. However, the frictional forces arising during the milling process result in the temperature increasing which is enhancing the losses of essential oil and thermosensitive components through evaporation and oxidative reactions [25]. As the literature report, the losses of essential oil in nutmeg and cinnamon are about 37% and 17%, respectively [26]. Therefore, a volatile compounds saver sample preparation method is very necessary. Cryogenic grinding technique using liquid nitrogen that provides extremely low temperature needed to pre-cool the sample. During this process, the moisture and oil are solidified so that the sample becomes brittle. Additionally, to maintain the low temperature, the liquid nitrogen vaporized to gaseous state and creates an inert and dry atmosphere which further protects the sample [27,28]. As the literature report, cryogenic grinding obtained 29.5% more volatile oil in cloves than that of ambient grinding [28].
In this work, the volatile compounds in 72 fresh A. villosum samples were analyzed and compared by using cryogenic grinding combined HS-SPME-GC-MS. As far as we are aware, this is the first time for systemic qualitative and comparative analysis of volatile compounds in different parts of fresh A. villosum from different geographical areas. Additionally, this work is helpful for comprehensive utilization and quality control of this plant.

Experimental procedures for cryogenic grinding of fresh A. villosum samples
In order to reduce the loss of volatile compounds in the sample preparation procedure, cryogenic grinding was employed. Fresh A. villosum samples (5 g) were frozen in liquid nitrogen, then volatilization of the liquid nitrogen and quickly crushed the sample in airtight grinding bowl. The procedure was repeated until the sample milling to fine powder (20 mesh). Then 50 mg sample powder was transferred to a 20 mL headspace vials for SPME analysis.

Instrumentation and GC-MS conditions
The incubation, equilibrium, extraction and desorption of volatile components were carried out automatically by a Combi-Pal autosampler (CTC Analytics, Zwingen, Switzerland). Specifically, the headspace vials were incubated and equilibrated at 80 °C for 10 min, under continuous agitation (500 rpm). And then the volatile compounds were extracted by the SPME fiber (100 μm PDMS) for 30 min at the same condition. The desorption was performed in the injector at 250 °C for 5 min.
GC-MS analyses were performed on a 6890 Series GC instrument coupled to a 5973 N mass spectrometer and a ChemStation software (Agilent Technologies, Palo Alto, CA). Analytes separation was achieved on a HP-5MS capillary column (30 m × 0.25 mm i.d.) coated with 0.25 μm film (5% phenyl methyl siloxane). The splitless injection mode was used. High-purity helium was used as carrier gas with flow rate of 1 mL/min. The oven temperature was initially programmed at 70 °C, then raised 4 °C/min to reach 125 °C and then programmed at 0.5 °C/ min to 133 °C, then at 6 °C/min to 170 °C and finally, at 20 °C/min to 280 °C held for 2 min. The temperatures of the injection port, ion source and transfer line were set at 250 °C, 230 °C and 280 °C, respectively. The mass spectrometer was operated in electron-impact mode (EI).

Statistical data analysis
The principal components analysis (PCA) and hierarchical clustering analysis (HCA) analyses were performed in software R for windows based on the 72 A. villosum samples with 48 identified volatile compounds and their Comparative index. Specifically, the PCA was conducted using R package, gmodels. And the analysis results were visualized using R packages, 'ggplot2' and 'scatterplot3d' . The heatmap analysis and HCA were performed using R package, 'pheatmap' .

Optimization of the SPME conditions
For volatile compounds analysis, selected the appropriate SPME fibers is the first step. In this study, 65 μm PDMS/ DVB, 100 μm PDMS, 85 μm polyacrylate and 75 μm CAR/PDMS were used to evaluate the effect of fiber types on the extraction of volatile compounds in fresh A. villosum. (Sample 1). The relative peak area (RPA) was employed to evaluate the content of volatile compounds under different SPME conditions. The RPA was calculated as: RPA (B) % = (B/A) *100%; (A was the total peak area of volatile compounds under the optimum condition; B was the total peak area of volatile compounds under other conditions) [29]. As shown in Fig. 1a, 100 μm PDMS achieved the highest extraction of the volatile compounds than other fibers. This indicated that the retention ability of this fiber for the volatile compounds in fresh A. villosum is much stronger than the other fibers. Additionally, the 65 μm PDMS/DVB absorbed more volatile compounds than 85 μm polyacrylate and 75 μm CAR/PDMS. Actually, the 100 μm PDMS, 65 μm PDMS/ DVB are non-polar fiber and semi-polar fiber, respectively. Otherwise, 85 μm polyacrylate and 75 μm CAR/ PDMS are polar fibers. The main volatile compounds in A. villosum such as bornyl acetate, borneol and camphor are low-polar compounds which are effective absorbed by the non-polar fiber or semi-polar fiber [30]. In addition, the fiber coating thickness also affects the extraction process. Usually, the small molecules are absorbed more efficiently by a thicker fiber [30]. The extraction temperature (60, 70, 80, 90 °C) at four levels were investigated. The results (Fig. 1b) showed that the content of volatile compounds increased steadily with temperature (60-80 °C). Actually, the extraction temperature had a significant influence on the adsorption process because it can influence the distribution coefficients of the volatile compounds among sample, headspace and fiber. When the temperature was higher than 80 °C, the content of volatile compounds was stable. However, higher temperature may also cause compound degradation. Therefore 80 °C was selected as the optimum temperature. The extraction time ranges from 10 to 40 min was investigated (Fig. 1c). When the extraction time increased from 10 to 30 min, the total content of volatile compounds was increased correspondingly. When the extraction time was higher than 30 min, the total content of volatile compounds was stable. Therefore 30 min was selected as the optimum extraction time. Particle size is also an important factor affecting the extraction efficiency. Usually, small particle size is easier for essential oil volatilize. However, conventional sample preparation methods for HS-SPME GC-MS analysis containing cut, pulping, milling and so on. All these methods may cause the loss of volatile compounds. In this study, cryogenic grinding was employed and the particle size (20, 40, 60, 80 mesh) was optimized. As shown in Fig. 1d, when the particle size was smaller than 40 mesh, the total content of volatile compounds was decreased significantly. This result indicated that the volatile compounds are easily lost in sample preparation procedure. The samples with particle size bigger than 20 mesh was not investigated in this study, because the large particle size might affect sampling uniformity. The total content of volatile compounds detected at 20 and 40 mesh was similar, but the particle size at 20 mesh requires less sample preparation time and lower cost. Therefore, 20 mesh was selected as optimum particle size.
In our previous work, the volatile compounds in 12 batches of the dried fruit of A. villosum were evaluated and 11 main components were determined. Camphor, borneol and bornyl acetate were the most abundant compounds, accounting for almost 98% of the determined compounds [1]. This result was in accordance with that previously reported by other authors. Zeng et al. (2010) analyzed the volatile compounds in different varieties of Amomi Fructus and found camphor (27.81%), borneol (2.11%) and bornyl acetate (59.60%) as the major components in dried fruit of A. villosum [31]. However, in fresh fruits of A. villosum, the average content of camphor, borneol and bornyl acetate was 3.91%, 1.06% and 10.53%, respectively. This result is quite different with that of dried fruits of A. villosum. The reason might be that many volatile components are lost or degraded during drying and storage. (E)-nerolidol, the most abundant component in fresh fruits of A. villosum, exhibits many biological functions such as antifungal [32] and antiulcer [33] effects. However, this compound was great lost in dried samples. Therefore, qualitative and comparative analysis of volatile compounds in fresh A. villosum samples can obtain much more information which is helpful for their performance in the fields of functional foods, agriculture and biomedical industry.
The volatile compounds from root samples showed a higher amount of sesquiterpene hydrocarbons (67.21%, 61. 31 For further evaluation, the comparative index was used for comparative analysis of volatile compounds in different parts of A. villosum samples. The comparative index was calculated as: comparative index (A/SI) = P(A/ SI)/P(A Max. )*100; (comparative index (A/SI) was the comparative index of compound A in sample SI; P(A/SI) was the peak area of compound A in sample SI; P(A Max. ) was the maximum peak area of compound A in different parts of A. villosum samples). As shown in Table 2 (Fig. 3). As different parts of the same plant, there are many common components between fruit and other parts (Fig. 3). The bioactive compounds such as (-)-β-pinene, d-limonene and caryophyllene were existed in four parts of A. villosum. Additionally, the main pharmacological active ingredients camphor can be detected in root and stem samples, and bornyl acetate can be detected in  Fig. 4a, three principal components (PC) explained 83.5% of the total variance. Specifically, PC 1, 2 and 3 were account for 44.4%, 22.8% and 16.3, respectively. All the A. villosum samples could be divided into four groups, which were corresponding to sample sets collected from fruit, root, leaf and stem, respectively. The HCA (Fig. 4b) was in accordance with that of PCA, four parts of A. villosum samples were divided into four subgroups clearly. These results indicated that the pattern of the content of volatile compounds detected in different organs of A. villosum were so divergent that a sample could easily be assigned to its sample set basing on this kind of pattern.

Comparison of volatile compounds in fresh A. villosum from different geographical areas
As shown by the results of HCA (Fig. 4b), samples collected from fruits could be divided into two groups although two of the samples (sample 1 and 9) was misassigned. Samples in group I-1 were all from Pingxi village, Yangchun and samples in group I-2 were from Shiwanzai village, Yangchun (exclude sample 1 and 9). This result indicated that the volatile compounds in fruit samples may be affected by the growing region. The average comparative index of five chemical classes (monoterpene hydrocarbons, oxygenated monoterpenes, sesquiterpene hydrocarbons, oxygenated sesquiterpenes and others) of group I-1 and I-2 were shown in Fig. 5a. The results showed that great difference was observed in four chemical classes (oxygenated monoterpenes, sesquiterpene hydrocarbons, oxygenated sesquiterpenes and others). This indicated that these four chemical classes in fruit of A. villosum might be more susceptible to the growing environment. Similarly, many root samples from different regions were also divided into two groups (group II-1 and group II-2). As shown in Fig. 5b great difference was observed in sesquiterpene hydrocarbons and other compounds. Leaf and stem samples could not be clustered into any subgroups that related to the geographic distribution of the corresponding samples. These results indicated that the fruits and roots of A. villosum produced by distinct areas, at least the areas analyzed here, showed a difference in the content of volatile compounds. However, for further investigate, much more samples are needed to investigate the most suitable growth environment, cultivating mode and collection time for A. villosum with high quality.