Figures
Abstract
Metabolomics technology has enabled an important method for the identification and quality control of Traditional Chinese Medical materials. In this study, we isolated metabolites from cultivated Dendrobium officinale and Dendrobium huoshanense stems of different growth years in the methanol/water phase and identified them using gas chromatography coupled with mass spectrometry (GC-MS). First, a metabolomics technology platform for Dendrobium was constructed. The metabolites in the Dendrobium methanol/water phase were mainly sugars and glycosides, amino acids, organic acids, alcohols. D. officinale and D. huoshanense and their growth years were distinguished by cluster analysis in combination with multivariate statistical analysis, including principal component analysis (PCA) and orthogonal partial least squares discriminant analysis (OPLS-DA). Eleven metabolites that contributed significantly to this differentiation were subjected to t-tests (P<0.05) to identify biomarkers that discriminate between D. officinale and D. huoshanense, including sucrose, glucose, galactose, succinate, fructose, hexadecanoate, oleanitrile, myo-inositol, and glycerol. Metabolic profiling of the chemical compositions of Dendrobium species revealed that the polysaccharide content of D. huoshanense was higher than that of D. officinale, indicating that the D. huoshanense was of higher quality. Based on the accumulation of Dendrobium metabolites, the optimal harvest time for Dendrobium was in the third year. This initial metabolic profiling platform for Dendrobium provides an important foundation for the further study of secondary metabolites (pharmaceutical active ingredients) and metabolic pathways.
Citation: Jin Q, Jiao C, Sun S, Song C, Cai Y, Lin Y, et al. (2016) Metabolic Analysis of Medicinal Dendrobium officinale and Dendrobium huoshanense during Different Growth Years. PLoS ONE 11(1): e0146607. https://doi.org/10.1371/journal.pone.0146607
Editor: Mohana Krishna Reddy Mudiam, CSIR- Indian Institute of Toxicology Research, INDIA
Received: September 6, 2015; Accepted: December 18, 2015; Published: January 11, 2016
Copyright: © 2016 Jin et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All relevant data are within the paper and its Supporting Information files.
Funding: The study was supported by Discipline Backbone Cultivated Foundation in Anhui Agricultural University (Grant No.2014XKPY-41, www.ahau.edu.cn), Twelfth Five-Year Plan for Science & Technology Support of Anhui Province (Grant No.1301032139, www.ahkjt.gov.cn/technologi/index.htm), and Collaborative Innovation Center of Agri-forestry Industry in Dabieshan Area in Anhui Agricultural University (dbs.w1.ahau.edu.cn).
Competing interests: The authors declare that they have no conflict of interest.
Introduction
Dendrobium is a perennial herb in the family Orchidacea (Dendrobium Sw.) and is widely distributed in Australasia, Oceania and other tropical and subtropical areas [1,2]. In China, there are 74 Dendrobium species and two varieties [3], and nearly 50 of these species are used in medicine [4]. However, wild Dendrobium resources are threatened by extinction due to slow growth rates, habitat destruction and overexploitation. Thus, artificial, large-scale cultivation of medical Dendrobium has been developed. As valuable Chinese medicinal materials, Dendrobium species play important pharmacological roles with abundant polysaccharides, alkaloids, phenanthrenes, bibenzyls, and other biologically active substances [5,6]. However, the chemical constituents and contents differ significantly among different medicinal Dendrobium species. Some non-genuine Dendrobium is adulterated and many fake species referred to as “Dendrobium” are circulating in the market. This misrepresentation is not conducive to the safety and quality of medicinal Dendrobium, its clinical applications, or the healthy development of the industry. Therefore, an effective comprehensive method of Dendrobium germplasm identification and quality control is urgently needed.
Dendrobium officinale Kimura et Migo and Dendrobium huoshanense C. Z. Tang et S. J. Cheng are both commercially valuable, particularly D. huoshanense [7]. A comprehensive analysis of the chemical compositions of cultivated D. officinale and D. huoshanense and the differences in their metabolic components have not been reported. Metabolomics is the study of all low molecular weight metabolites within an organism or cell during a specific period of time by both qualitative and quantitative methods. Metabolomics has been widely used in the study of medicinal plants, including the identification of medicinal herbs [8], discrimination of origin [9], determination of harvest time [10], method of processing [11] and other factors. Metabolomic studies of Dendrobium metabolites have been limited.
In this study, a metabolic profile of Dendrobium was constructed using gas chromatography-mass spectrometry (GC-MS) combined with multivariate statistical analysis. The changes in the composition and content of metabolites, including sugars, alcohols, organic acids, amino acids and other metabolites, were studied in perennially cultivated D. officinale and D. huoshanense to identify biomarkers as a reference for the identification and quality control of Dendrobium.
Materials and Methods
Plant materials and reagents
The experiment was conducted using one-, two- and three-year-old basin-cultured D. officinale and D. huoshanense seedlings (Fig 1) grown in a greenhouse (Hefei Anhui Mulong Mountain Dendrobium Biotechnology Development Co., Ltd) under the conditions of day 24°C and night 18°C, with natural light. Six replicates of each sample, including one- to three-year-old stems of the two species, were collected from the same pot. Surface soil was removed by washing with water, and the materials were dried with filter paper. Then, the samples were immediately frozen in liquid nitrogen and stored at -80°C until use for metabolomics analysis.
(A) Cultivated D. officinale (B) Cultivated D. huoshanense; 1, 2, and 3 represent one-,two- and three-year-old seedlings.
Methanol and chloroform (HPLC grade) were purchased from TEDIA (Fairfield, OH, USA). Pyridine was obtained from Dr. Ehrenstorfer (Augsburg, Germany). Adonitol and methoxylamine hydrochloride were purchased from Sigma-Aldrich. N,O-bis (trimethylsilyl)-trifluoroacetamide (BSTFA) containing 1% trimethylchlorosilane (TMCS) was purchased from SUPELO (Bellefonte, PA, USA). Ultra-pure water was obtained from Wahaha Group Co., Ltd. (Hangzhou, China).
Metabolite extraction and derivatization
Metabolite extraction was performed according to the reference [12]. Each of the frozen samples (100±5 mg of fresh weight for D. officinale, 10±0.5 mg of fresh weight for D. huoshanense) was ground to a fine powder with liquid nitrogen and transferred to 10-mL centrifuge tubes. Quality control (QC) samples were used by mixing the same mass of each Dendrobium sample and blank samples were also prepared with empty reactions, handling with the same method as that for the real samples. Next, 1.4 mL of cold methanol (-20°C) was added to the tube and vortexed for 1 min. As an internal quantitative standard in the methanol/water phase, 60 μL of adonitol (0.2 mg/mL) was added to the tube and vortexed for 30 s. The mixture was extracted using a supersonic instrument for 30 min (40°C). Next, mixed with 750 μL of chloroform and 1.4 mL of dH2O (4°C) vortexed for 1 min, and centrifuged at 8000 rpm for 15 min. One milliliter of the upper phase was transferred into a fresh 1.5-mL tube and dried under a nitrogen gas stream for derivatization. First, the dried samples were dissolved in 60 μL of methoxylamine hydrochloride (20 mg/mL in pyridine), vortexed for 30 s and heated at 37°C for 120 min. Then, 60 μL of BSTFA was added, followed by heating at 25°C for 90 min. The derivatized samples were transferred into glass vials (Aglient) for GC-MS analysis.
GC-MS analysis
The samples were analyzed using an Agilent 7890A gas chromatograph coupled to an Agilent 5975C mass spectrometer. The chromatographic separation was performed on a DB-5MS column (30 m ×250 μm × 0.25 μm). The injection volume was 1 μL, and the split ratio was 20: 1. The injector and ion source temperatures were 280°C and 250°C, respectively. The interface temperature was 250°C. The helium gas flow rate through the column was 1.0 mL/min. The temperature program began at 40°C and was held for 5 min, then increased at 10°C/min to 280°C and was maintained for 5 min. For MS detection, ions were generated by a 70-eV electron with an electron impact (EI) ionization mass spectrometric detector (MSD). Quadrupole mass spectrometry was performed using the full-scan method from 35 to 780 (m/z).
Data analysis
Raw GC-MS data were exported into CDF format by Agilent GC/MS 6890 data analysis software and subsequently processed by XCMS (V. 1.12.1) running under the R package (V. 2.7.2). The main functions of the XCMS software include matched filtration, peak detection, peak matching and novel nonlinear retention time alignment. Internal standards and any known artefact peaks caused by column pressure, noise, solvent and derivatization procedure, were removed from the matrix. The XCMS output was further processed using Microsoft Excel 2010. The peak area of each metabolite was normalized according to the adonitol internal standard. Finally, the normalized data were imported into Simca-p software (V. 11.5) for multivariate statistical analysis, including unsupervised principal component analysis (PCA) and supervised orthogonal partial least squares discriminant analysis (OPLS-DA), and into SPSS (V. 13.0) for cluster analysis and t-tests. A statistically significant threshold of variable influence on projection values (variable influence on projection values, VIP>1) obtained from the OPLS-DA model was combined Student’s t test (t-test) (P<0.05) to identifiy discriminating metabolites [13]. Metabolites were identified by searching the commercial database NIST 11. Publicly available data from the KEGG pathway database (http://www.kegg.jp/kegg/pathway.html) were used to confirm the relationships between metabolite-metabolite correlations. Heat map analysis was performed using Multi Experiment View 4.9.
Results
Detection of extracts from Dendrobium samples
The GC-MS spectra of methanol/water phase extracts from one-year-old D. officinale and D. huoshanense stems are presented in Fig 2. In total, 544 peaks (Fig 2A) and 249 peaks (Fig 2B) were observed for D. officinale and D. huoshanense, respectively, and pretreated using AMDIS software. A total of 139 metabolites of Dendrobium methanol/water phase extracts were tentatively identified based on similarities of greater than 70% to mass fragments in the NIST 11 standard mass spectral databases. We classified the 139 metabolites into nine main categories: fatty acids, sugars and glycosides, organic acids, amino acids, amines and amides, alcohols, alkanes, ketones, and others (Table 1).
A and B show the spectra of methanol/water phase extracts of D. officinale and D. huoshanense one-year-old stems, respectively.
Metabolic profiling of two medicinal Dendrobium stems during different growth years
Cluster analysis of metabolites in two medicinal Dendrobium stems.
The metabolites in the D. officinale and D. huoshanense stems were subjected to cluster analysis using SPSS 13.0 software. D. officinale and D. huoshanense were clearly distinguished. Thirty-six Dendrobium samples were clustered into two major groups (Fig 3). All D. officinale samples were classified into Class I, and Class II contained 18 collections of D. huoshanense. These results demonstrate that the metabolites in cultivated D. officinale and D. huoshanense were substaintially different. The cluster analysis showed a certain influence on metabolite compositions by growth years. For example, collections of D. officinale could be generally clustered into three subgroups (one, two and three growth years), with some overlap. In order to further distinguish the two Dendrobium species, we need to use multivariate statistical analysis.
“Fe” indicates D. officinale, and “M” indicates D. huoshanense. The numbers 1, 2, and 3 indicate the corresponding growth years of Dendrobium. The letters a, b,…, f indicate repeats of the same sample.
Multivariate statistical analysis of metabolites in two medicinal Dendrobium stems.
Principal component analysis (PCA) and orthogonal partial least squares discriminant analysis (OPLS-DA) of cultivated D. officinale and D. huoshanense were performed using Simca-p 11.5 software (Fig 4 and Fig 5). The PCA analysis showed a substantial difference between the two Dendrobium species (Fig 4), with two principal components explaining 83.8% of the total variability (67.2% and 16.6% for principal component 1 and principal component 2, respectively). Moreover, a clear separation among different growth years (one, two, and three) of Dendrobium samples was observed in the scores plot, and only a few of samples overlapped. In order to find the features with power to distinguish the two Dendrobium species with different growth years, OPLS-DA model (noisy information was removed prior to model building) was established with the scores plot and loadings plot shown in Fig 5. The R2X, R2Y, and Q2 of this model were 0.755, 0.868 and 0.861, respectively, indicating the stability and reliability of this OPLS-DA model. Obviously, in the analysis of the two Dendrobium species metabolites data, PCA and OPLS-DA were more powerful than cluster analysis.
(A) PCA scores plot for 36 Dendrobium collections. (B) PCA loadings plot marked by two Dendrobium species. “Fe” indicates D. officinale, and “M” indicates D. huoshanense. The numbers 1, 2, and 3 indicate the corresponding growth years of Dendrobium.
(A) OPLS-DA scores plot for 36 Dendrobium collections. (B) OPLS-DA loadings plot marked by two Dendrobium species. “Fe” indicates D. officinale, and “M” indicates D. huoshanense. The numbers 1, 2, and 3 indicate the corresponding growth years of Dendrobium.
We further employed OPLS-DA to identify the metabolites contributing significantly to the separation. Table 2 lists the top 11 metabolites (VIPs) influencing cluster formation within the methanol/water phase generated from OPLS-DA of the two Dendrobium species. The identified VIP components include 4 sugars (sucrose, galactose, glucose, fructose), 3 organic acids (1-cyclohexene-1-carboxylic acid, succinate, propanoate), 1 fatty acid (hexadecanoate), 2 alcohols (glycerol, myo-inositol) and 1 volatile substance (oleanitrile). The contents of sucrose, galactose, glucose, fructose, succinate, myo-inositol, and glycerol were much higher in D. huoshanense than in D. officinale (Fig 6), whereas hexadecanoate and oleanitrile levels were much lower in D. huoshanense than in D. officinale. The VIP components from OPLS-DA were combined with the t-test (P <0.05) to identify nine significantly different metabolites (sucrose, glucose, galactose, succinate, fructose, hexadecanoate, oleanitrile, myo-inositol, and glycerol) as potential biomarkers that may discriminate these two Dendrobium species.
“Fe” means D. officinale, and “M” means D. huoshanense. The bars represent metabolite peak areas across the 18 samples of each species. The error bars indicate the standard deviations of six biological repeats (including one-, two-, and three-year-old Dendrobium stem samples, each with six repeats). The nine candidate biomarkers were sucrose, glucose, galactose, succinate, fructose, hexadecanoate, oleanitrile, myo-inositol, and glycerol (**P<0.01, D. officinale compared to D. huoshanense).
Metabolites levels in the stems of two medicinal Dendrobium species during different growth years.
Raw GC-MS data were pretreated, and the peak area of each metabolite was obtained. The software Multi Experiment View 4.9 was used to construct a heat map, displaying the changes in metabolite content between D. officinale and D. huoshanense in one-, two-, and three-year-old stems (Fig 7). The result suggested that the levels of the majority of amino acids were higher in D. officinale than in D. huoshanense. Amino acids decreased in D. officinale during growth (from one to three years). By contrast, amino acids were maintained at low levels during three years of growth in D. huoshanense, with only valine and proline exhibiting higher levels in three-year-old stems. The profiles of changes in sugar and glycoside levels were obvious in both D. officinale and D. huoshanense. The total amount of sugars was higher in D. huoshanense than D. officinale. Sucrose, glucose, mannose, fructose and erythrose maintained constant high levels during the three-year growth period in both Dendrobium species, whereas galactose and trehalose reached their highest levels in D. officinale during the first growth year and then decreased in the next two years. For D. huoshanense, sugar and glycoside levels either remained constantly high or increased from one to three growth years. The organic acids propanoate, succinate, and 1-cyclohexene-1-carboxylic acid all remained at a high level during the three growth years. In D. officinale, 2-keto-l-gluconate, 2-ketoglutarate, glutarate, ribonate and arabino-hexonate decreased throughout the three-year growth period. In D. huoshanense, 2-keto-l-gluconate, gluconate, acetate, 2-butenoate and benzoate increased significantly during the three years. Fatty acids and ketones did not vary during the growth stage. Alcohols other than glycerol and myo-inositol remained at high levels in both Dendrobium species. Other alcohols exhibited different changes between D. officinale and D. huoshanense. For example, arabitol and 1,2-butanediol increased in D. officinale but decreased in D. huoshanense.
All data are provided in S1 Table. The metabolite peak areas in each sample represent the average peak areas. The lowest figures are in green, and the highest figures are in red. “Fe” and “M” refer to D. officinale and D. huoshanense, respectively. The numbers 1, 2, and 3 represent one, two, and three growth years.
Construction of metabolic profiling between two medicinal Dendrobium species
The functions of the identified metabolites in the main plant metabolic pathways network were examined (Fig 8). As compared with those in D. officinale, the contents of sucrose, glucose, myo-inositol, hexane and benzamide in D. huoshanense increased 9.5-fold, 54.4-fold, 12.6-fold, 4.2-fold and 9.7-fold, respectively. Whereas, the contents of piperidine, oxalate, octadecanoate, urea, carbamate, ethane and oleanitrile decreased 0.2-fold, 0.3-fold, 0.5-fold, 0.02-fold, 0.009-fold, 0.3-fold and 0.15-fold, correspondingly. As shown above, most soluble sugars showed significant increases in D. huoshanense, which may suggest much higher freezing tolerance in D. huoshanense [14]. In D. officinale, the content of piperidine was much higher than that in D. huoshanense.
The ratios in the red or green bar indicate high or low relative metabolite peak areas, respectively, of cultivated D. huoshanense compared to D. officinale. The level of significance was set at P<0.05. The metabolites in gray characters were undetectable. G6P, glucose-6-phosphate; F6P, fructose-6-phosphate; 3PGA, 3-phosphoglycerate; PEP, phosphoenolpyruvic acid.
Discussion
Metabolomics has been applied extensively to plants [15–17], such as the identification of metabolite changes, the identification of differences in metabolites between wild type and mutant plants [18–21]. The use of metabolomics to study Dendrobium metabolites has not been reported. We preliminarily constructed a metabolomics platform and analyzed the metabolites in the methanol/water phase of cultivated D. officinale and D. huoshanense stems collected during different growth years.
Analysis of Dendrobium metabolic profiling provides an important basis for species identification
Cluster analysis of the methanol/water phase metabolites of Dendrobium distinguished the two Dendrobium species. D. officinale and D. huoshanense were classified into two clusters, Class I and Class II, respectively. PCA and OPLS-DA, two multivariate statistical analysis methods widely used in metabolomics, both not only clearly separated D. officinale and D. huoshanense but also distinguished different growth years of each Dendrobium species. VIP components combined with the t-test (P<0.05), in which significantly different metabolites were selected as potential biomarkers (sucrose, glucose, galactose, succinate, fructose, hexadecanoate, oleanitrile, myo-inositol, and glycerol), provide an important method for Dendrobium identification. Consistent with our results, Yuan H reported that the relative peak area of glucose could be used as a foundation for Dendrobium identification using a pre-column derivatization HPLC method [22]. Currently, oleanitrile has not been reported in Dendrobium species. In our study, oleanitrile was one of the biomarkers of Dendrobium, which suggests the metabolomics technology platform we constructed is relatively comprehensive.
Analysis of Dendrobium metabolic profiling provides an important basis for quality control of medicinal Dendrobium
The main chemical components in medicinal Dendrobium are polysaccharides and alkaliods, with multiple biological activities [23]. Polysaccharide content has been used to determine the medicinal quality of Dendrobium [24]. The heat map and 11 VIP components confirmed the higher sugar content in D. huoshanense than in D. officinale (Fig 7 and Table 2), which indicates that D. huoshanense exhibits better quality. According to the metabolic profiling of Dendrobium, we suppose that D. officinale may have piperidine alkaliod because of its high piperidine content [25]. However, piperidine alkaliod has not been successfully annotated. It’s probable that we just detected the metabolites in polar phase of Dendrobium. The metabolites from Dendrobium non-polar phase should also be detected since most alkaloids are fat-souble in plants. Besides, most of the secondary metabolites are thermally labile and unsuitable for GC-MS analysis.
During the three growth years, the total sugar content first decreased and then increased. These changes may reflect the basic physical transformation and energy metabolism in the Dendrobium vegetative phase. Amino acids levels decreased over the growth years because basic amino acids are nitrogen-containing precursors that are involved in the biosynthesis of a variety of secondary metabolites in plants. Based on the accumulated metabolites in Dendrobium, the optimal harvest time is in the third year [26]. Cluster analysis and multivariate statistical analysis of Dendrobium metabolites discriminated different Dendrobium species and provided a new basis for identification and quality control.
In our study, the metabolites in Dendrobium stems were analyzed using GC-MS, and the majority of the metabolites examined were primary metabolites. In future work, we will combine metabolomics, transcriptomics and proteomics technologies to further study secondary metabolites, particularly pharmaceutically effective ingredients in Dendrobium and synthesis mechanisms.
Supporting Information
S1 Table. Major metabolites of D. officinale and D. huoshanense during different growth years.
https://doi.org/10.1371/journal.pone.0146607.s001
(XLS)
Acknowledgments
We would like to thank American Journal Experts (AJE) for editing our manuscript. Thanks for the GC-MS analysis platform provided by Zhongliang Zhu from University Of Science And Technology Of China. The study was supported by Discipline Backbone Cultivated Foundation in Anhui Agricultural University (Grant No.2014XKPY-41), Twelfth Five-Year Plan for Science & Technology Support of Anhui Province (Grant No.1301032139), and Collaborative innovation Center of Agri-forestry Industry in Dabieshan Area in Anhui Agricultural University.
Author Contributions
Conceived and designed the experiments: YL YC. Performed the experiments: QJ CJ SS YZ. Analyzed the data: QJ CJ SS CS. Contributed reagents/materials/analysis tools: CJ SS. Wrote the paper: QJ CJ SS HF.
References
- 1. Jin Q, Yao Y, Cai Y, Lin Y. Molecular cloning and sequence analysis of a phenylalanine ammonia-lyase gene from dendrobium. PLOS ONE. 2013; 8: e62352. pmid:23638048
- 2. Xiang X-G, Schuiteman A, Li D-Z, Huang W-C, Chung S-W, Li J-W, et al. Molecular systematics of dendrobium (Orchidaceae, Dendrobieae) from mainland Asia based on plastid and nuclear sequences. Mol Phylogenet Evol. 2013; 69: 950–960. pmid:23811435
- 3. Feng S, He R, Yang S, Chen Z, Jiang M, Lu J, et al. Start codon targeted (SCoT) and target region amplification polymorphism (TRAP) for evaluating the genetic relationship of dendrobium species. Gene. 2015; 567: 182–188. pmid:25936992
- 4. Chen X, Wang F, Wang Y, Li X, Wang A, Wang C, et al. Discrimination of the rare medicinal plant dendrobium officinale based on naringenin, bibenzyl, and polysaccharides. Sci China Life Sci. 2012; 55: 1092–1099. pmid:23233224
- 5. Ng TB, Liu J, Wong JH, Ye X, Sze SCW, Tong Y, et al. Review of research on Dendrobium, a prized folk medicine. Appl Microbiol Biotechnol. 2012; 93: 1795–1803. pmid:22322870
- 6. Gutiérrez RMP. Orchids: a review of uses in traditional medicine, its phytochemistry and pharmacology. J Med Plants Res. 2010; 4: 592–638.
- 7. Lu S, Guo G, Cai Y. Progress in physiological and biochemical characters of dendrobium huoshanense. Chin Tradit Herb Drugs. 2006; 37: 790.
- 8. Duan L-X, Chen T-L, Li M, Chen M, Zhou Y-Q, Cui G-H, et al. Use of the metabolomics approach to characterize Chinese medicinal material Huangqi. Mol Plant. 2012; 5: 376–386. pmid:22138859
- 9. Lee D-K, Yoon MH, Kang YP, Yu J, Park JH, Lee J, et al. Comparison of primary and secondary metabolites for suitability to discriminate the origins of Schisandra chinensis by GC/MS and LC/MS. Food Chem. 2013; 141: 3931–3937. pmid:23993568
- 10.
Wang H, Wang H-H, Ma C-F, Liu B-Y, Xu G-W, Ye H-C. Metabolomics-based studies on Artemisinin biosynthesis. In: Plant metabolomics. Netherlands: Springer; 2015. pp. 245–273.
- 11. Sun H, Ni B, Zhang A, Wang M, Dong H, Wang X. Metabolomics study on Fuzi and its processed products using ultra-performance liquid-chromatography/electrospray-ionization synapt high-definition mass spectrometry coupled with pattern recognition analysis. Analyst. 2012; 137: 170–185. pmid:22030742
- 12. Lisec J, Schauer N, Kopka J, Willmitzer L, Fernie AR. Gas chromatography mass spectrometry–based metabolite profiling in plants. Nat Protoc. 2006; 1: 387–396. pmid:17406261
- 13. Jumtee K, Bamba T, Fukusaki E. Fast GC‐FID based metabolic fingerprinting of Japanese green tea leaf for its quality ranking prediction. Journal of separation science. 2009; 32: 2296–2304. pmid:19569110
- 14. Wanner LA, Junttila O. Cold-induced freezing tolerance in Arabidopsis. Plant Physiology. 1999; 120: 391–400. pmid:10364390
- 15. Noutsos C, Perera AM, Nikolau BJ, Seaver SMD. Metabolomic Profiling of the Nectars of Aquilegia pubescens and A. Canadensis. PLOS ONE. 2015; 10: e0124501. pmid:25933103
- 16. Zhang J, Wang X, Yu O, Tang J, Gu X, Wan X, et al. Metabolic profiling of strawberry (Fragaria× ananassa Duch.) during fruit development and maturation. Journal of experimental botany. 2010; erq343.
- 17. Tan H, Xie Q, Xiang X, Li J, Zheng S, Xu X, et al. Dynamic Metabolic Profiles and Tissue-Specific Source Effects on the Metabolome of Developing Seeds of Brassica napus. PLOS ONE. 2015; 10: e0124794. pmid:25919591
- 18. Lee E-J, Shaykhutdinov R, Weljie AM, Vogel HJ, Facchini PJ, Park S-U, et al. Quality assessment of ginseng by 1(H) NMR metabolite fingerprinting and profiling analysis. J Agric Food Chem. 2009; 57: 7513–7522. pmid:19655726
- 19. Kim N, Kim K, Choi BY, Lee D, Shin Y-S, Bang K-H, et al. Metabolomic approach for age discrimination of panax ginseng using UPLC-Q-Tof MS. J Agric Food Chem. 2011; 59: 10435–10441. pmid:21916514
- 20. Toh D-F, New L-S, Koh H-L, Chan EC. Ultra-high performance liquid chromatography/time-of-flight mass spectrometry (UHPLC/TOFMS) for time-dependent profiling of raw and steamed Panax notoginseng. J Pharm Biomed Anal. 2010; 52: 43–50. pmid:20079593
- 21. Tianniam S, Tarachiwin L, Bamba T, Kobayashi A, Fukusaki E. Metabolic profiling of Angelica acutiloba roots utilizing gas chromatography-time-of-flight-mass spectrometry for quality assessment based on cultivation area and cultivar via multivariate pattern recognition. J Biosci Bioeng. 2008; 105: 655–659. pmid:18640606
- 22. Yuan H, Bai Y, Si J, Zhang A, Jin X. Variation of monosacchride composition of polysacchrides in Dendrobium officinale by pre-column derivatization HPLC method. Zhongguo Zhong Yao Za Zhi=China Journal of Chinese Materia Medica. 2011; 36: 2465–2470. pmid:22256746
- 23. Xu J, Han Q-B, Li S-L, Chen X-J, Wang X-N, Zhao Z-Z, et al. Chemistry, bioactivity and quality control of Dendrobium, a commonly used tonic herb in traditional Chinese medicine. Phytochemistry reviews. 2013; 12: 341–367.
- 24. Lv G, Yan M, Chen S. Review of pharmacological activities of Dendrobium officinale based on traditional functions. Zhongguo Zhong Yao Za Zhi=China Journal of Chinese Materia Medica. 2013; 38: 489–493. pmid:23713270
- 25. Virjamo V, Julkunen-Tiitto R. Shoot development of Norway spruce (Picea abies) involves changes in piperidine alkaloids and condensed tannins. Trees. 2014; 28: 427–437.
- 26. Yaping D, Qingsheng W, Liwen Y. Study on the best time for harvesting dendrobium huoshanense. Chin Pharm J. 1998; 08.