Metabolomic study of water hyacinth exposed to CuCI2, FeCI3, Na2HPO4 solution by GC m

.. .METABOLOMIC STUDY OF WATER HYACINTH EXPOSED TO CuCl2, FeCl3 AND Na2HPO4 SOLUTIONS BY GC- MS HUANG XULEI (M Sc., PEKING UNIVERSITY) A THESIS SUBMITTED FOR THE DEGREE OF MASTER OF SCIENCE... cultured in 0.3 mmol FeCl3, 1.2 mmol CuCl2 and 3.46 mmol Na2HPO4 solution for 15 days, respectively The metabolic profiles of roots, stems and leaves of water hyacinth were determined by GC- MS and further... study, water hyacinth was exposed to 0.3 mmol FeCl3, 1.2 mmol CuCl2 and 3.46 mmol Na2HPO4 concentrations for 15 days, respectively The metabolic profiles of roots, stems and leaves of water hyacinth

METABOLOMIC STUDY OF WATER HYACINTH EXPOSED TO CuCl2, FeCl3 AND Na2HPO4 SOLUTIONS BY GC-MS HUANG XULEI NATIONAL UNIVERSITY OF SINGAPORE 2014 METABOLOMIC STUDY OF WATER HYACINTH EXPOSED TO CuCl2, FeCl3 AND Na2HPO4 SOLUTIONS BY GC-MS HUANG XULEI (M. Sc., PEKING UNIVERSITY) A THESIS SUBMITTED FOR THE DEGREE OF MASTER OF SCIENCE DEPARTMENT OF CHEMISTRY NATIONAL UNIVERSITY OF SINGAPORE 2014 DECLARATION I hereby declare that this thesis is my original work and it has been written by me in its entirety, under the supervision of Prof. Sam Li Fong Yau, Chemistry Department, National University of Singapore, between 12 August 2013 and 12 August 2014. I have duly acknowledged all the sources of information which have been used in the thesis. This thesis has also not been submitted for any degree in any university previously. Huang Xulei Name 8 August 2014 Signature I Date ACKNOWLEDGEMENTS First and foremost, I would like to thank my supervisor Prof. Li Fong Yau for giving me the chance to join his group and for encouraging me to enter into the wonderful world of analytical chemistry. His integral view on research has made a deep impression on me and has helped me out immensely by keeping me and my research focused and on track. I owe him lots of gratitude for having shown me the ways of scientific research. Besides of being an excellent supervisor, Prof Li was as close as a relative and a good friend to all the students. I am really glad that I have come to get know Prof. Li in my life. I would like to thank all the staffs and students in particular Dr.Liu Feng, Dr. Gan Pei Pei, Dr. Li Ping Jing, Dr. Guo Rui, Lin Xuanhao, Lai Linke, Feng Ting, Wu Ye and Liang Xiaojian who were in the same lab with me. Over the last year, I have indeed enjoyed working with them. They are so kind and ready to help me when necessary. We also discussed and shared some knowledge and information with each other freely. Best wishes to all of them. Finally, heartful thanks go to my family for their immense support along the way. II TABLE OF CONTENTS DECLARATION .................................................................................................. I ACKNOWLEDGEMENTS ................................................................................. II TABLE OF CONTENTS .................................................................................... III SUMMARY ......................................................................................................... V LIST OF TABLES .............................................................................................. VI LIST OF FIGURES ........................................................................................... VII 1 INTRODUCTION ......................................................................................... 1 1.1 INTRODUCTION TO METABOLOMICS ........................................................ 1 1.2 PLANT METABOLOMICS................................................................................ 3 1.2.1 PLANT METABOLOMICS IN RESPONSE TO ABIOTIC STRESSES .......... 5 1.3 HEAVY METAL ACCUMULATION IN WATER HYACINTH ...................... 14 1.4 BASIS AND SIGNIFICANCE OF DISSERTATION ....................................... 15 2 MATERIALS AND METHODS ..................................................................17 2.1 MATERIALS..................................................................................................... 17 2.2 INSTRUMENTS AND REAGENTS ................................................................ 17 2.3 SAMPLE EXTRACTION AND DERIVATIZATION ...................................... 18 III 2.4 GC-MS ANALYSIS .......................................................................................... 18 2.5 DATA ANALYSIS ............................................................................................. 18 3 RESULTS AND DISCUSSION ....................................................................20 3.1 GC RESULTS ANALYSIS ............................................................................... 20 3.2 IDENTIFIED COMPOUNDS ........................................................................... 25 3.3 PCA ANALYSIS ............................................................................................... 32 3.4 WEAKNESS AND PROSPECT OF THE STUDY .......................................... 37 4 CONCLUSION ............................................................................................39 BIBLIOGRAPHY...............................................................................................41 IV SUMMARY It is generally accepted that water hyacinth is capable of adsorbing excessive heavy metals, but the metal adsorption mechanism in the metabolic level is unknown. In this study, the water hyacinth plants were cultured in 0.3 mmol FeCl3, 1.2 mmol CuCl2 and 3.46 mmol Na2HPO4 solution for 15 days, respectively. The metabolic profiles of roots, stems and leaves of water hyacinth were determined by GC-MS and further analyzed by PCA method. Results showed that plants suffered severe damage under FeCl3 exposure but were tolerable to Na2HPO4 exposure. Metabolites levels in stems and leaves increased but decreased in roots under CuCl2 exposure. Leaves and stems of the four differently treated plants could be distinctly separated in three-dimensional PCA, while roots could only be separated between control group and the treated groups individually by two-dimensional PCA. Levels of D-glucopyranose, L-threonine, Butanoic acid and 9H-Purin-6-amine significantly increased in treated plants and acted as osmoprotectants. This study provided an overall perspective of metabolites change in water hyacinth for mechanism of metal accumulation. V LIST OF TABLES Table 1 Advantages and disadvantages of NMR and MS for metabolomics study. ... 3 Table 2 (a)Identified compounds in leaves of control and Na2HPO4, FeCl3, CuCl2 exposed samples in sequence of retention time (RT) and the corresponding peak areas. The peak areas were normalized to the mean response calculated for the control of each measured batch (±standard error). ................................................ 25 Table 2 (b)Identified compounds in stem of control and Na2HPO4, FeCl3, CuCl2 exposed samples in sequence of retention time (RT) and the corresponding peak areas. The peak areas were normalized to the mean response calculated for the control of each measured batch (±standard error). ................................................ 25 Table 2 (c) Identified compounds in root of control and Na2HPO4, FeCl3, CuCl2 exposed samples in sequence of retention time (RT) and the corresponding peak areas. The peak areas were normalized to the mean response calculated for the control of each measured batch (±standard error). ................................................... 25 VI LIST OF FIGURES Figure 1. GC results of leaves pretreated by (a) wet grinding and (b) dry grinding in the control group ....................................................................................... 20 Figure 2. GC results of stems pretreated by (a) wet grinding and (b) dry grinding in the control group ........................................................................................ 21 Figure 3. GC results of roots pretreated by (a) wet grinding and (b) dry grinding in the control group ........................................................................................ 22 Figure 4. GC results of leaf in the control group (a) before cultivation and (b) after cultivation .............................................................................................. 23 Figure 5. GC results of (a) leaf, (b) stem and (c) root in the control group after cultivation ...................................................................................................... 24 Figure 6. Three-dimensional PCA of the metabolic profiles of (a)leaf, (b) stem and (c) root in control and the exposed samples. ........................................... 33 Figure 7. Two-dimensional PCA of the metabolic profiles in roots of (a) control vs Na2HPO4 treated group, (b) control vs FeCl3 treated group and (c) control vs CuCl2 treated group ................................................................................... 35 VII 1 Introduction 1.1 Introduction to metabolomics Metabolomics is a science for living systems (including cell, tissue and organism) research through examining metabolic responses or time-course variation of living systems when they are subjected to stimulation or disturbance. Based on group indicators, the goal of metabolomics as a branch of systems biology is information modeling and system integration via high-throughput detection and data processing. Metabolomics is another important research area in systems biology following after genomics, transcriptomics and proteomics. The focus of metabolomics is metabolites variation of small molecules with molecular weight smaller than 1000 in metabolism, reflecting metabolic responses variation of cells or tissues subjected to outside stimuli or genetic modification. The living organism is a dynamic system regulated by multi-factors integratively. In the biological information transport chain from genes to traits, organisms need to constantly adjust their own complex metabolic network to maintain normal dynamic balance within system or between system and external environment. The existence of DNA, mRNA and protein provides a material basis for the biological processes, while metabolic substances and metabolic phenotype reflect a biological event that has happened. The metabolic substances and metabolic phenotype are the comprehensive result of genotype and environment combination, and direct embodiment of physiological and biochemical function status in a biological system. Therefore, as an important component of systems biology, metabolic groups can better reflect the system phenotype. 1 The process of metabolomics analysis includes sample preparation, data collection, data analysis and interpretation. The sample preparation is composed of sample extraction, pretreatment and compounds separation. After extracted by water or organic solvent, samples are commonly pretreated using solid-phase microextraction, solid-phase extraction and affinity chromatographic methods. Then compounds are separated via electrophoresis gas chromatography, methods, etc. Such liquid chromatography separation and analysis and capillary methods as chromatography, mass spectrometry (MS), nuclear magnetic resonance (NMR), infrared spectroscopy, coulometric analysis, ultraviolet absorption, fluorescent scattering, radioactivity detection and light scattering and their combinations are all applied in the metabolomics study. Among them, the NMR technology, especially the hydrogen spectrum (1H NMR), chromatography and MS become the main analysis tools, due to the universality of 1H NMR for hydrogen metabolites, and high resolution and high flux of chromatography, and universality, high sensitivity and specificity of MS. Later period of Metabolomics research is to interpretate the biological significance of data based on data analysis and interpretation with the aid of bioinformatics platform. Constantly used methods by far include multiple regression, discriminant analysis, principal component analysis, hierarchical cluster analysis, factor analysis and canonical analysis, etc. NMR and MS based approaches in metabolomics have their own advantages and disadvantages, respectively (Table 1). The advantages of NMR involve high reproducibility, minimal sample preparation, short analysis time and low cost per sample, while MS has the advantages of high sensitivity, availability for targeted analysis, cheaper instrument cost. Choosing whether NMR or MS for metabolomics study depends on the purpose of the study, the research object and the instrument 2 availability. Table 1 Advantages and disadvantages of NMR and MS for metabolomics study. NMR MS Sensitivity Low High Reproducibility Very high Average Number of detectable metabolites 30-100 300-1000+ (depending on whether GC-MS or LC-MS is used) Targeted analysis Not optimal analysis targeted Better for targeted analysis than NMR Sample preparation Minimal sample preparation required More complex sample preparation required Tissue extraction Not required. Tissues can be analyzed directly Requires tissue extraction Sample analysis time Fast. The whole sample can be analyzed in one measurement Takes longer than NMR. Requires different chromatography techniques for depending on type of metabolites analyzed Instrument Cost More expensive and occupies more space than MS Cheaper and occupies less space than NMR Sample Cost Low cost per sample High cost per sample for 1.2 Plant metabolomics In 1999, Nicholson team put forward the concept of metabonomics [1]. So far they have done a lot of fruitful work in disease diagnosis and drug screening and so on ([2]; [3]; [4]). Fiehn [5] put forward the concept of metabolomics and correlated metabolites to biological gene function for the first time. Afterwards, many plants chemists had been carrying out research on plant metabolomics. 3 The application of metabolomics in plant research mainly included following aspects: (1) plant metabolites of certain species. Such researches usually focused on a certain plant, selected a particular organ or tissue, analyzed the metabolites qualitatively and quantitatively, studied comprehensively on changes of metabolites types and contents in different periods or different parts, and further speculated the corresponding metabolic pathways and metabolic networks through these changes; (2) phenotype Metabolomics of different genotypes plants. Such researches usually studied two or more than two plants (including normal controls and genetically modified plants), comparing and identifying different genotypes plants using metabolomics, as comparison of difference between the mutant or genetically modified plants and normal wild-type plants, or difference between tissue-cultured Metabolomics and the wild-type. This category of studies played an important role in evaluation of the efficacy of genetic modification or tissue culture and screening of good varieties; (3) metabolomics of certain ecotypes plants. Such researches usually chose the same type of plants under different ecological environment, and studied the effect of habitat on plant metabolites; (4) plant autoimmune response after external stimulation. In such researches, changes of plant metabolites were induced by exogenous chemicals stimuli, physical or biological stimuli and were monitored and comprehensively analyzed by metabolomics method; (5) application in gene function research. Metabolic products are the final products of gene expression and tiny changes in gene expression level may lead to massive changes of metabolites. Previous determination of rise and fall of gene expression level through visible phenotypic change takes long time, and sometimes gene expression changes cannot cause phenotypic change, while the content of certain metabolites in plant body has already changed significantly. Using of metabolomics method can judge the change of 4 gene expression level, so as to deduce the function of genes and their metabolic flux. A lot of research results have been made in the plant metabolomics research. Fiehn [5] analyzed the petiole vascular and leaf extract of Cucurbita maxima using GC-MS and obtained more than 400 peaks. By comparison with the mass spectrum database, he preliminarily identified 90 compounds, and compared the differences on metabolites in sugar and amino acid composition between petiole and leaf; Tiessen et al. (2002) [6] conducted Metabolomic analysis of Solanum tuberosum tuber using high performance liquid chromatography (HPLC). They determined the quantity change of a series of substrates, intermediates, enzyme and products in the starch synthesis approach. Then through comparison research between the wild and heterologous adenosine diphosphate glucose focal phosphorylase (AGPase) transgenic potatoes, they proposed a new regulating mechanism in the starch synthesis approach; Maier, etc (1999) [7] studied the effect of Glomus intraradices on Nicotiana tabacum root metabolism, compared the metabolites difference between tobacco roots with and without Glomus intraradices. To sum up, Metabolomics technology is an ideal platform for plant metabolism study. 1.2.1 Plant metabolomics in response to abiotic stresses Recently many scientific research institutions carried out metabolomics studies on abiotic stress responses of plants. Through qualitative and quantitative analysis of plant metabolites under stress environment via modern detection and analysis methods, the variation trend and rule of plant metabolites over time can be monitored. The integration of various omics platforms such as genomics, proteomics and metabolomics is also a powerful toolkit [8]. Combination of all these information 5 helps to study responses of biological systems to genes or environment changes as a whole. For example, one can judge the level where metabolites change happens, helping people uncover the mysterious and complex mechanism of plant stress response. These stress factors include water deficit, excessive high or low temperature, phosphorus and sulfur deficit, excessive salt and heavy metals and so on. (1) Drought stress Water is one of the important factors that affect plant growth and development. The harmful effect on plant due to less environmental moisture is called drought stress. In order to study the contribution of different wine grape (Vitis vinifera) fruit organization to the wine quality and the influence of drought stress on wine quality, Grimplet (2009) [9] determined protein with specific differences in fruit (peel and fruit pulp) and tissue of wine grape planted under condition of enough moisture and dry environment. Using two-dimensional gel electrophoresis (2-d PAGE) technology, 1047 proteins in fruits were detected, among which 90 were differentially expressed in peel and fruit, while 695 proteins were detected in seeds, among which 163 proteins showed almost no difference in the seed and pee expression spectrum. Drought stress changed abundance of about 7% skin protein, but showed little effect on seed protein expression. In the selected 32 small molecule metabolites to be determined, about 50% showed differences in the peel and seeds organizations, while under drought stress condition 7 compounds were affected in accumulation within grape fruits. The metabolic fingerprinting results provided new inspiration for studying the effect of drought on the main compounds related to wine flavor and aroma in grape. Deluc etc. (2009) [10] studied the metabolomics and transcriptome of two different strains of grape Cabernet Sauvignon and Chardonnay under long-term 6 drought and seasonal drought. Studies showed different metabolic pathway changes for two strains of grape in the response to drought stress. For Cabernet Sauvignon, the glutamic acid and proline synthesis pathways and some important intermediate steps in styrene acrylic acid synthesis pathway can be activated by drought stress, while for Chardonnay under drought stress, styrene acrylic acid, carotenoids and isoprenoid synthesis pathways were activated. Both stress responses involved influence on abscisic acid metabolic pathway. These metabolic products changes had a great influence on fruit and wine flavor. Mane etc. (2009) [11] conducted metabolic profile analysis and biomass and yield comparison of two genotypes Andean potato (Solanum tuberosum ssp. Andigena Juz & Buk Hawkes) Sullu and Ccompis. Results showed that although the tuber yield of the two genotypes potatoes was not obviously affected, the aboveground biomass of Ccompis reduced and Sullu biomass was not affected. Sucrose and and trehalose in regulatory molecules accumulated in Sullu blade, while in Ccompis blade, the oligosaccharide family way of cottonseed sugar was activated, and low level change of sucrose and a small amount of stress-related trehalose change. Proline and related gene expression level improved, and the expression amount of which in Sullu is 3 times more than Ccompis. To sum up, the yield of two genotypes plants showed no obvious change under drought condition, but the biomass accumulation and metabolite changes were obviously different. (2) Temperature stress Plant response to temperature in growth and development has three basic points: lowest temperature, optimum temperature and maximum temperature. The harmful effect on plant caused by too low or too high temperature is called temperature stress. 7 Shulaev et al. (2008) [12] reviewed in detail plant metabolomics under temperature stress. The metabolic fingerprinting technology is used to explore response of Arabidopsis thaliana plants (Arabidopsis) to temperature stress. Kaplan et al. (2004) [13] studied metabolic fingerprint of Arabidopsis thaliana plant under high and low temperature environment using GC-MS technology and found a series of small molecule metabolites related to high temperature and low temperature or both. The metabolite changes associated with low temperature were the most significant, but to our surprise, most of the metabolites produced under thermal stress would also be produced under cold stress, among which many metabolites were not considered to be related to the temperature stress in previous studies. In subsequent research work ([14]), these metabolic fingerprint data were integrated in order to study the adaptive mechanism of Arabidopsis thaliana plants to low temperature. Results showed that only part of the metabolites change were related to transcriptomics change, while the rest of metabolites change was not directly related to transcriptomics change. It can be concluded through the above research that in the process of plant response to cold stress, the metabolites not directly related to transcriptomic changes played an important role in temperature response of Arabidopsis thaliana plants. Cook et al. (2004) [15] compared metabolic fingerprint of Arabidopsis thaliana plants with different cold resistance abilities and excessive expression (CBF) (C-repeat/dehydration responsive element-binding factor) using GC-MS technology. Results showed that metabolism of Arabidopsis thaliana obviously changed in the process of cold stress responses, and that CBF pathway played an important role in the adaptation of low temperature environment in Arabidopsis. Morsy et al. (2007) [16] studied the sugar metabolomics of cold stress and high salt stress response for two genotypes rice with different cold resistance abilities. Using 8 HPLC method, the authors quantitatively analyzed the soluble saccharide compounds of cold resistance and cold non-resistance rice, and found that accumulation of soluble saccharide of the two genotypes rice under cold stress was different. For cold resistant rice, galactose and raffinose accumulated under cold stress environment, while the content of these two kinds of sugar showed a downward trend in the other genotype rice. The two genotypes rice also showed different saccharide metabolism characteristics under high salt stress environment. (3) Salt stress The adverse effect on plant caused by too many soluble salts in the soil is called salt stress. Metabolomics technology was used to identify metabolites change of tomato (Solanum lycopersicum) under salt stress. Johnson et al. (2003) [17] selected two tomato strains with different salt sensitivity Edkawy and Simge F1 for research, and found that the relative growth rate of Simge F1 under salt stress significantly decreased, while that of Edkawy was not affected. Using Fourier transform infrared spectrum (FT-IR), the fresh tomato fruit extracts from control group and salt stress group were analyzed. The obtained data was processed by PCA and discriminant function analysis (DFA), respectively. PCA method could not distinguish the fruit difference between control group and high salt treatment group, while DFA method could distinguish between two different genotypes and fruits of different genotypes in control group and high salt treatment group. Genetic algorithm (GA) model was used to identify possible important functional groups in FT-IR spectrugram for salt stress response. These functional groups included saturated nitrile compounds, unsaturated nitriles cyanide compounds and strong NH2 radicals peaks and other nitrogen compounds, etc. 9 More detailed plant research in salt stress response was the time-course metabolomics study by Kim et al. (2007) [18] at Arabidopsis thaliana cells in high salt culture. The metabolic fingerprint metabolomics of Arabidopsis cells after treatment by 100 mmol.L-1 NaCl for 0.5, 1, 2, 4, 12, 24, 48 and 72 h separately were determined using LC-MS and GC-MS. The data was analyzed by PCA and self-organizing map. Results showed that short-term metabolism change of plant cells in salt stress response included induction of methylation cycle which provided methyl, induction of hydroxyl methyl amine circulation which induced lignin synthesis, and 3-armour amino acid synthesis; while long-term metabolism changes included influence on glycolysis and sucrose metabolism, and co-reduction of methylation system. Cramer et al. (2007) [19] compared the difference of grapevine (Vitis vinifera ‘Cabemet Sauvignon’) metabolites change between response to salt stress and drought stress using GC-MS and anion exchange chromatography-ultraviolet detector method. They found that in response to salt stress, the content of sugar, aspartic acid, succinic acid and fumaric acid in plant declined, while the content of proline, asparagine, malic acid and fructose etc. was relatively higher. Compared with response to salt stress, the content of glucose, malic acid and proline was higher in response to drought stress. Gong et al. (2005) [20] also compared metabolites difference between salt-tolerant plants (Thellungiella halophila) and Arabidopsis thaliana by combination of GC-MS metabolic fingerprint with biochip technology. They found obvious differences between metabolites of the two species. Compared with Arabidopsis thaliana, salt mustard maintained higher levels of metabolites whether in high salt environment or general environment. Analysis of Arabidopsis thaliana showed that glucose, proline and possible polysaccharide significantly increased in Arabidopsis thaliana plants under 150 mmol.L-1 salt stress. While in salt mustard, salt stress induced fairly 10 complex metabolic response. Not only many metabolites levels were higher before salt stress, but also the content of many sugars, sugar alcohol, organic acid and phosphate after induction showed apparent change. (4) Sulfur and phosphorus stress In addition to the above abiotic stresses, plants are also subjected to other environmental stresses, such as sulfur stress and phosphorus stress. Recently, many studies involved metabolomics research of plants subjected to sulfur, and phosphorus stress. Nikiforova et al. ([21, 22]) analyzed Arabidopsis thaliana plants under sulfur stress using GC-MS and LC-MS techniques. They detected 134 known compounds and a series of unknown compounds in Arabidopsis thaliana related to sulfur stress, and made dynamic monitoring of these compounds, thus successfully rebuilt metabolic network of sulfur stress response in Arabidopsis thaliana. Then, metabolic network data was consolidated with transcriptomics data of sulfur stress response in Arabidopsis thaliana, therefore the relationship between gene expression and metabolite changes under sulfur stress was obtained. The combination of Metabolomics and proteomics was also applied in the study of leguminous plants under phosphorus stress. Hernandez et al. (2007) [23] made metabolic profile analysis of roots of leguminous plant with sufficient phosphorus and insufficient phosphorus using GC-TOF-MS, and identified a series of metabolic products related to the phosphorus stress, many of which (including amino acids, polyols and sugars) increased in content in response to phosphorus stress. (5) Heavy metal stress Several studies have investigated the metal stress responses in various plants. As a frequent farmland contaminant, cadmium was intensively studied in regards to metal 11 toxicity in plants. Bailey et al. [24] analyzed NMR Metabolomics of bottle wheat grass (Silenecucubalus) cells under high cadmium stress. PCA analysis could distinguish the cadmium stress group and blank group. They found that the content of malic acid and acetic acid salt increased significantly in bottle wheat grass cells under cadmium stresses, while the content of glutamic acid and some branched chain amino acids showed a downward trend in cadmium stress response. Kieffer et al. [25] conducted research of proteome combined with metabolic profile analysis on Poplar (Poplar spp.) under cadmium stress. Results showed that levels of pigment and carbohydrates of Poplar changed in response to cadmium stress. Under cadmium stress, the poplar showed growth inhibition and photosynthesis was also affected. In the process of growth and development, photosynthesis products stored in the form of hexose or other complex carbohydrates in plants, thus adjusting osmotic pressure. Sun et al. [26] elucidated the metabolic responses of Arabidopsis thaliana to different cadmium concentrations (0, 5, 50 µM) for 2 weeks using GC-MS analysis. Results showed that levels of carbohydrates, organic acids, amino acids, and other metabolites changed under cadmium stress. Levels of Ala, b-ala, Pro, Ser, putrescine, Suc, 4-aminobutyric acid, glycerol, raffinose and trehalose increased in the treated plants compared to control group. Concentrations of antioxidants such as alfa-tocopherol, campesterol, beta-sitosterol and isoflavone also significantly increased. Hediji et al. [27] evaluated the long-term response of tomato plants to cadmium exposure through 1H NMR, HPLC-PDA, and colorimetric methods. The plants were 12 cultured in hydroponic conditions (0, 20, and 100 µM CdCl2) for 90 days. Results showed that tomato plants adapted to 20 µm Cd concentration during long-term exposure and were perturbed physiologically leading to limited growth and fruit set abortion. Liu et al. [28] studied the metabolites response of halophyte (Suaeda salsa) exposed to 2, 10 and 50 µg. L-1 cadmium concentration using NMR-based Metabolomics. After cadmium exposures, the levels of amino acids, carbohydrates, intermediates of tricarboxylic acid cycle and osmolyte changed in the samples, indicating increased protein degradation and disturbances in the osmotic regulation and energy metabolism. Grid chara (Scenedesmus) plant is an important research object for study of metal stress and phenolic expression. Dividing the genus plant S. quadricauda into 3 groups, Jozef et al. [29] conducted Metabolomics research after treatment of them using Cu2+, salicylic acid and combination of Cu2+ and salicylic acid, respectively. Results showed that content of chlorophyll, soluble protein and phenolic compounds declined in the Cu2+ treatment group; The alicylic acid treatment group showed opposite trend; In the Cu2+ and salicylic acid treatment group, salicylic acid could not resist the downward trend of the three kinds of compounds under Cu2+ stress. However, the concentration of Cu2+ did not increased in plants body, possibly due to accumulation of benzoic acid which was related to salicylic acid. Brassica species plants have long been regarded as metal ion collector, and widely used in the soil remediation, but the resistance mechanism for metal ions is unclear. Jahangir et al. [30], studied the metabolites change of turnip (Brassica rapa) subjected to metal ions stress. The turnip was exposed to three kinds of metal ions 13 (Cu2+, Fe3+ and Mn2+) with four different concentrations (50, 100, 250 and 500 mmol.L-1). The plants sample after treatment was detected by 1H NMR and 2D-NMR, then the data was analyzed by PCA and partial least squares (PLS). Such primary metabolites as gluconic acid and hydroxy acid conjugate salt compounds, some carbohydrates and amino acids can be used to distinguish plants treated by different kinds of metal ions. Results showed that the metabolites change of plants treated by Cu2+ and Fe3+ was greater than those treated by Mn2+. Moreover, change of metabolic products was not only related to the species of metal ions, but also associated with the concentration of metal ions. 1.3 Heavy metal accumulation in Water Hyacinth Water hyacinth (Eichhornia crassipes) is a perennial, floating aquatic plant with rapid-growing capability. It is widely distributed in warm temperate, subtropical, and tropical regions around the world. Various studies have shown that water hyacinth is capable of adsorbing excessive heavy metals. Research by Misbahuddin and Fariduddin [31] showed that 81% of As with 400 ppb concentration was removed by the roots of living water hyacinth plants, and when stems and leaves of the water hyacinth plant were involved, 100% As was removed in six hours. In Taiwan, Liao and Chang [32] demonstrated that water hyacinth removed large amounts of lead, copper, and zinc in a constructed wetland. The results by Soltan and Rashed [33] showed that water hyacinth can survive in a mixture of heavy metal concentrations (Cd, Co, Cr, Cu, Mn, Ni, Pb and Zn) up to 3 mg.L-1 and in 100 mg.L-1 Pb solution. The uptake of metal ions from aqueous solutions by water hyacinth is a deprotonation reaction explained by a decrease in pH. The mechanism was believed to 14 be founded on the chelation with carboxylic, amino acid and hydroxyl groups of macrocyclic molecules, such as ionophores existed in the mitochondria of water hyacinth. Haider et al. (1983) suggested that the uptake of chemicals by water hyacinth might happen through the cell membrane via osmosis and diffusion. Field studies by Ajmal et al. [34] and Zaranyika and Ndapwadza [35] had shown that metals accumulated in the leaves and roots of water hyacinth. Cooly and Martin [36] found that Cu and Cd accumulated more in roots than least and petioles in leaves. Kelley et al. [37] explained that carboxylic acids were responsible for chelating the intracellular proportion of Eu(III) in the roots. According to the study by Malik [38], almost all heavy metal ions accumulated in the roots rather than in the shoot system of water hyacinth. An exceptional research carried out by Zhu et al. [39] suggested that the metal Se was transported to the upper biomass of water hyacinth. Metal accumulation was found in the sequence of roots > stems > leaves of water hyacinth, and there was a linear correlation between the external metal concentration and internal metal concentration [40] for Cr, Cu, Ni, and As [41]. It was suggested that the capability of water hyacinth in storing most heavy metals enabled the plant to avoid toxicity of photosynthetic tissues caused by heavy metals. 1.4 Basis and significance of dissertation Modern industries have released large amounts of heavy metals into the environment. Among various technologies for heavy metal removal, such as adsorption, reverse osmosis and electrochemical method etc., phytoremediation has been regarded as a promising method due to its low cost. Water hyacinth has the capacity of adsorbing heavy metal during growth, thus is a promising plant for phytoremediation. Although numerous researches had been done on mechanism of 15 heavy metal adsorption by water hyacinth, these studies all focused on reaction in parts of the plant such as roots [33], or certain features of the plant, such as reflectance [42], the complete metabolism response when water hyacinth is subjected to heavy metal stress is unknown. The application of metabolomics method on water hyacinth in response to heavy metal stress may provide incentive on the metabolites pathway of the water hyacinth for heavy metal adsorption from the perspective of the whole plant, thus building an solid scientific foundation for better application of water hyacinth in the phytoremediation. Previously, few studies on plant metabolomics in response to heavy metal stress involved comparison of metabolites change of different parts of plants. This point deserved research as different parts of plant may play different roles in the metabolites change, subdivision of their functions could help us understand plant response to heavy metal stress as a whole. Moreover, previous studies on plant metabolomics mostly described metabolites response to only one kind of stress, such as drought stress, metal stress or salt stress. Few studies involved metabolites response to two or more kinds of stress. The comparison between metabolites responses to different stresses may be significant, as it may help us better understand each stress response of the same plant, so as to understand those shared responses and stress-specific responses of the same plant. In this article, the water hyacinth plants were exposed to CuCl2, FeCl3 and Na2HPO4 solutions. Then we analyzed the metabolites change and differences of roots, stems and leaves of the plants using GC-MS method and PCA analysis. The metabolism pathways of water hyacinth in response to different heavy metal stress were proposed. 16 2 Materials and Methods 2.1 Materials Water Hyacinth seedlings were all in the similar growth condition, purchased from the same batch in an ornamental fish shop in Clementi Station, Singapore. They were washed by deionized water for several times before planting. 32 water hyacinth seedlings were planted in 8 1mL glasses with 4 seedlings in each glass. In the first week, all the plants were cultivated in deionized water which was changed for fresh water every three days. After that, plants in 2 of 8 glasses were kept in deionized water continuingly as the control group. In the rest 6 glasses, 2 glasses were replaced with 0.3 mmol FeCl3 solution, another 2 glasses replaced with 1.2 mmol CuCl2 solution, the last 2 glasses replaced with 3.46 mmol Na2HPO4 solution. All of the solutions were replaced with originally corresponding solutions every three days, respectively. In the whole cultivation process, plants were placed in laboratory with temperature maintained around 23 oC and illumination time from 8:00 am to 18:00 pm every day. 15 days later, the roots, stems and leaves of water hyacinth were cut into segments with length of 1 mm, then put in labelled centrifuge tubes and stored in the freezer with -20 oC. 2.2 Instruments and Reagents Instruments included 7980A GC-QQQ (Agilent Technologies) with HP-5 capillary gas chromatography column, drying machine, centrifuge. Methanol, trichloromethane, pyridine and bis(trimethylsilyl)trifluoroacetamide chromatographically pure purchased from Sigma, Singapore. 17 (BSTFA) were all 2.3 Sample extraction and derivatization Part of the plants samples (including leaves, stems and roots) were directly grinded without drying, others were dried before grinding. Individual samples were extracted with 0.5 mL CHCl3 : CH3OH (3:1) solution, and then centrifuged at 161000 x g for 10 min. The supernatant was lyophilized for derivatization. Dried sample extracts were reconstituted in 300 L pyridine. 100 L BSTFA was added to each dried sample extract for derivatization for 120 minutes. The obtained solution was agitated for 1 minute under room temperature and transferred to an amber vial for GC-MS analysis. 2.4 GC-MS analysis 1.0 L aliquot of the BSTFA-derivatized sample was injected using the splitless mode into a GC/QqQ 7980A system equipped with a HP-5 capillary column. The inlet temperature was set at 250 oC. Helium was used as the carrier gas with a constant flow rate of 1.40 mL/min through the column. The initial temperature was set at 100 o C, 1 minute after injection the GC temperature was increased at a rate of 5 oC/min to 280oC and held for 5 minutes at 300 oC. The transfer line temperature was set at 280 o C. Detection was achieved using MS in electron impact mode and full scan monitoring (m/z 100 to 600). The temperature of the ion source was set at 250 oC. 2.5 Data analysis To identify metabolites, all chromatographic peaks obtained by GC-MS analysis were compared with NIST by masshunter software. The data were then exported to Simca-P+ Software package (Umetrics, Umea, Sweden) for subsequent processing by 18 an unsupervised method. For principal component analysis (PCA), the data was reduced to two latent variables (or principal components, PCs) that will describe maximum variation within the data. The PCs which are obtained from the scores will highlight clustering, trends and outliers in the observation direction of the data set. 19 3 Results and Discussion 3.1 GC results analysis (a) wet grinding (b) dry grinding Figure 1. GC results of leaves pretreated by (a) wet grinding and (b) dry grinding in the control group 20 (a) wet grinding (b) dry grinding Figure 2. GC results of stems pretreated by (a) wet grinding and (b) dry grinding in the control group (a) wet grinding 21 (b) dry grinding Figure 3. GC results of roots pretreated by (a) wet grinding and (b) dry grinding in the control group Figure 1, 2 and 3 showed the effect of grinding on GC results of leaves, stems and roots in the control group, respectively. In the three tissues of plants, the number and peak intensities of wet grinding and dry grinding were different. The GC results of leaves and stems pretreated by dry grinding visually demonstrated more peaks compared to that pretreated by wet grinding, especially in the retention time period of 4 - 10 min. It suggested that more metabolites were detected in the leaves pretreated by dry grinding than by wet grinding. The reason may be related to water inhibition of the followed extraction process by CHCl3 : CH3OH (3:1) solution. In the sample pretreated without drying, hydrophilic metabolites tend to retain in the water rather than extracted by the organic solvent. Thus, fewer metabolites were detected in wet grinding sample. As such, only dried samples were utilized for GC-MS analysis in the following experiments. 22 (a) before cultivation (b) after cultivation Figure 4. GC results of leaf in the control group (a) before cultivation and (b) after cultivation Figure 4 described the GC results of leaved in the control group before and after cultivation by deionized water. The quantity difference of peaks between the two samples was not obvious, suggesting that number and species of metabolites during 15 days’ cultivation maintained. However, clear difference with respect to peak intensity of the two samples occurred, such as the peaks between 6 – 8 min. A general tendency is stronger peak after cultivation than before. This phenomenon may be associated with the growth stage of water hyacinth. The significant quantity of metabolites in the control group benefited the comparison with metal-exposed groups, 23 as tiny alteration in metabolites quantity would be easily distinguished. (a) leaf (b) stem (c) root Figure 5. GC results of (a) leaf, (b) stem and (c) root in the control group after cultivation 24 The comparison among GC results of leaf, stem and root of samples in the control group was shown in figure 5. The number of compounds detected in leaf, stem and root was 65, 45 and 19, respectively. This clear difference might be related with the function of different tissues. The function of leaf in plant is photosynthesis and respiration which involve complicated reaction among significant amounts of compounds. The function of stem is conduction, requiring the participation of comparatively less compounds, while root mainly undertakes the mission of water and nutrient absorption in plant, requiring the least amount of compounds. 3.2 Identified compounds Table 2 (a)Identified compounds in leaves of control and Na2HPO4, FeCl3, CuCl2 exposed samples in sequence of retention time (RT) and the corresponding peak areas. The peak areas were normalized to the mean response calculated for the control of each measured batch (±standard error). 25 RT (min) Compounds Control Na2HPO4 FeCl3 CuCl2 4.441 Astaxanthin 1.00±0.15 0.86±0.03 1.00±0.06 0.95±0.02 4.690 α-D-Galactopyranose 1.00±0.15 0.85±0.10 0.95±0.14 1.03±0.11 4.801 Glycine 1.00±0.15 1.38±0.27 1.33±0.13 1.03±0.12 5.022 Urea tri-TMS 1.00±0.18 1.18±0.15 0.90±0.13 0.87±0.07 5.668 L-Valine 1.00±0.20 1.83±0.35 0.67±0.21 1.41±0.28 6.093 Urea 1.00±0.18 1.40±0.41 0.78±0.23 1.13±0.12 6.663 Silanamine 1.00±0.23 0.98±0.16 0.73±0.28 0.89±0.14 6.722 Phosphate 1.00±0.31 0.72±0.13 0.45±0.08 0.71±0.13 6.787 Glycerol 1.00±0.48 0.91±0.11 2.32±0.30 0.75±0.43 7.196 L-Isoleucine 1.00±0.22 0.87±0.23 0.35±0.26 0.93±0.12 7.287 L-Serinamide 1.00±0.21 1.09±0.30 0.67±0.30 1.00±0.09 7.648 Butanedioic acid 1.00±0.29 1.14±0.25 0.40±0.24 0.45±0.14 8.640 D-glucopyranose 1.00±0.34 1.79±0.26 1.36±0.33 1.48±0.34 9.208 L-threonine 1.00±0.21 1.98±0.42 2.64±0.21 2.10±0.33 9.558 Hexanoic acid 1.00±0.23 1.15±0.15 0.76±0.22 1.06±0.16 12.027 L-Threitol 1.00±0.17 1.28±0.20 0.62±0.15 0.95±0.14 12.205 L-Proline 1.00±0.19 0.92±0.23 0.57±0.18 0.72±0.08 12.469 Butanoic acid 1.00±0.33 2.92±0.49 1.56±0.22 2.22±0.23 1.00±0.21 0.67±0.22 0.47±0.20 1.03±0.17 12.662 3,5-di-tert-Butyl-4-hydroxyphen ylpropionic acid 13.674 Xylulose 1.00±0.19 1.30±0.19 0.75±0.13 1.01±0.11 14.126 Pregn-4-ene-3,11,20-trione 1.00±0.25 1.02±0.13 0.45±0.24 0.72±0.19 14.416 Prosta-5,13-dien-1-oic acid 1.00±0.38 1.21±0.26 0.81±0.26 1.11±0.21 14.578 L-phenylalanine 1.00±0.21 1.30±0.18 1.15±0.36 1.07±0.20 15.589 L-Asparagine 1.00±0.26 1.27±0.38 0.58±0.15 1.13±0.29 16.230 D-(+)-Galactose 1.00±0.36 0.72±0.13 0.73±0.05 0.75±0.13 16.730 L-(-)-Arabitol 1.00±0.28 0.96±0.14 0.57±0.24 0.78±0.21 1.00±0.20 1.54±0.10 0.72±0.22 1.29±0.11 Phosphoric 17.628 acid, bis(trimethylsilyl) 2,3-bis[(trimethylsilyl)oxy]prop yl ester 18.500 D-(-)-Fructofuranose 1.00±0.26 0.99±0.15 0.87±0.17 1.71±0.31 19.194 Cyclohexene 1.00±0.16 1.48±0.11 1.13±0.11 1.24±0.08 26 RT (min) Compounds Control Na2HPO4 FeCl3 CuCl2 19.382 α-D-Glucopyranosiduronic acid 1.00±0.22 1.29±0.09 0.97±0.33 1.46±0.20 1.00±0.18 1.10±0.23 0.44±0.11 0.91±0.13 19.608 3,9-Epoxypregn-16-ene-14,20-d iol 19.689 9H-Purin-6-amine 1.00±0.22 1.58±0.14 1.56±0.35 1.51±0.18 20.060 Ingol 12-acetate 1.00±0.22 0.88±0.11 0.66±0.26 1.46±0.23 20.270 D-(+)-Galactopyranose 1.00±0.28 1.03±0.16 0.50±0.21 1.17±0.21 20.475 Galactopyranose 1.00±0.23 0.98±0.25 1.40±0.40 1.39±0.13 20.991 D-Glucose 1.00±0.23 0.89±0.16 0.42±0.08 1.08±0.19 21.492 D-(+)-Galactose 1.00±0.25 1.36±0.02 0.97±0.11 1.36±0.25 22.046 D-Glucose 1.00±0.27 0.90±0.17 0.63±0.19 1.27±0.18 23.262 Hexadecanoic acid 1.00±0.17 1.10±0.10 1.09±0.09 1.11±0.07 23.999 Myo-Inositol 1.00±0.24 1.24±0.11 0.68±0.23 0.87±0.17 24.602 9H-Purin-6-amine 1.00±0.32 0.76±0.16 0.63±0.31 0.90±0.12 1.00±0.17 1.25±0.17 0.76±0.14 0.78±0.11 Silane, 25.575 [(3,7,11,15-tetramethyl-2-hexad ecenyl)oxy]trimethyl 26.194 9,12-Octadecadienoic acid 1.00±0.19 1.35±0.11 0.69±0.09 0.72±0.16 26.296 α-Linolenic acid 1.00±0.18 1.31±0.07 0.55±0.09 0.49±0.18 26.797 Octadecanoic a 1.00±0.17 1.09±0.11 0.66±0.12 0.83±0.12 28.029 Glyceryl-glycoside 1.00±0.25 1.54±0.16 0.34±0.21 0.80±0.11 (b) Identified compounds in stems of control and Na2HPO4, FeCl3, CuCl2 exposed samples in sequence of retention time (RT) and the corresponding peak areas. The peak areas were normalized to the mean response calculated for the control of each measured batch (±standard error). 27 RT (min) Compounds Control Na2HPO4 FeCl3 CuCl2 4.274 Astaxanthin 1.00±0.09 1.01±0.11 0.88±0.02 0.94±0.12 4.435 β-D-Galactopyranoside 1.00±0.15 1.40±0.11 1.27±0.07 1.56±0.09 4.683 Pentasiloxane 1.00±0.14 1.24±0.19 0.88±0.11 1.00±0.07 5.022 Urea tri-TMS 1.00±0.20 1.68±0.20 0.90±0.19 1.33±0.18 5.657 L-Valine 1.00±0.24 2.78±0.17 0.94±0.26 1.83±0.21 6.093 Urea 1.00±0.20 1.38±0.21 1.05±0.19 1.19±0.19 6.663 Silanamine 1.00±0.14 0.93±0.14 0.89±0.15 0.92±0.12 6.722 Phosphate 1.00±0.09 1.41±0.15 0.51±0.13 0.86±0.10 6.787 Glycerol 1.00±0.15 1.15±0.15 0.90±0.20 1.76±0.07 7.196 L-Isoleucine 1.00±0.11 1.48±0.18 0.88±0.21 1.20±0.10 7.653 Butanedioic acid 1.00±0.15 1.20±0.16 0.87±0.16 1.26±0.12 8.121 5β-Cholestane-3α,7α 1.00±0.19 1.16±0.12 1.15±0.12 1.29±0.15 8.756 Benzoic acid 1.00±0.13 0.93±0.13 0.76±0.15 0.78±0.23 9.208 Glycine 1.00±0.17 1.61±0.08 0.92±0.08 1.53±0.24 12.022 Threitol 1.00±0.12 1.11±0.11 0.85±0.12 1.13±0.13 12.463 Butanoic acid 1.00±0.17 1.35±0.24 0.72±0.18 0.56±0.18 12.668 Tetrahydro 1.00±0.12 0.90±0.15 0.40±0.07 0.89±0.11 13.674 Xylulose 1.00±0.20 1.30±0.24 0.77±0.10 0.85±0.12 16.219 D-(-)-Fructose 1.00±0.14 0.99±0.14 0.71±0.13 1.30±0.16 18.349 1.00±0.33 0.93±0.06 0.62±0.05 1.03±0.21 18.694 D-(-)-Tagatofuranose 1.00±0.26 1.44±0.27 1.00±0.23 1.73±0.32 19.609 D-(-)-Fructopyranose 1.00±0.20 0.86±0.18 0.77±0.15 1.01±0.15 1.00±0.35 1.78±0.32 0.71±0.16 1.82±0.10 22.041 α-D-(+)-Talopyranose 1.00±0.32 1.80±0.33 0.77±0.17 1.90±0.14 23.262 D-Glucose 1.00±0.08 1.13±0.11 0.87±0.10 0.87±0.07 26.194 Hexadecanoic acid 1.00±0.14 1.07±0.12 0.62±0.09 0.54±0.05 26.302 9,12-Octadecadienoic acid 1.00±0.10 1.00±0.11 0.53±0.09 0.43±0.10 D-Glucopyranosiduronic acid (22S)-21-Acetoxy-6β,11β-dihydroxy 20.270 -16α,17α-propylmethylenedioxypreg na-1,4-diene-3,20-dione (c) Identified compounds in roots of control and Na2HPO4, FeCl3, CuCl2 exposed samples in sequence of retention time (RT) and the corresponding peak areas. The peak areas were normalized to the mean response calculated for the control of each 28 measured batch (±standard error). RT (min) Control Na2HPO4 FeCl3 CuCl2 1.00±0.05 1.02±0.07 0.8±0.08 0.72±0.08 1.00±0.16 1.19±0.13 0.54±0.18 0.55±0.33 5.797 Urea 1.00±0.22 0.84±0.25 1.61±0.19 1.91±0.36 6.717 Phosphate 1.00±0.08 1.35±0.15 0.72±0.18 0.70±0.15 6.787 Glycerol 1.00±0.17 1.14±0.20 0.35±0.21 0.22±0.11 8.121 Milbemycin 1.00±0.26 0.70±0.14 0.3±0.15 0.64±0.02 9.009 Benzaldehyde 1.00±0.04 1.27±0.25 0.81±0.27 1.68±0.24 9.321 Glycine 1.00±0.45 1.16±0.23 0.98±0.24 1.24±0.35 1.00±0.11 0.85±0.28 1.40±0.19 0.55±0.16 1.00±0.20 1.52±0.18 1.28±0.38 0.74±0.47 1.00±0.24 0.79±0.25 0.71±0.28 0.37±0.01 1.00±0.21 0.76±0.32 0.69±0.16 0.74±0.25 1.00±0.12 1.26±0.16 0.92±0.23 0.44±0.18 20.663 Anodendroside 1.00±0.28 0.83±0.18 0.29±0.03 0.46±0.16 21.373 Prosta-5,13-dien-1-oic acid 1.00±0.14 0.44±0.16 0.54±0.21 0.73±0.46 22.041 L-Asparagine 1.00±0.07 0.88±0.17 0.83±0.42 0.83±0.20 23.262 Hexadecanoic acid 1.00±0.08 0.97±0.22 0.88±0.09 0.99±0.05 24.350 Pregn-4-ene-3,20-dione 1.00±0.65 1.84±0.32 1.28±0.16 1.21±0.15 26.802 Octadecanoic acid 1.00±0.12 0.82±0.26 0.89±0.10 0.76±0.16 4.435 Astaxanthin 5.022 1-Monolinoleoylglycerol trimethylsilyl ether 12.668 2,4,6-Tri-t-butylbenzenethiol 12.980 1-Monolinoleoylglycerol trimethylsilyl ether 16.445 Glycine 19.431 Olean-12-ene-3,15,16,21,22,28-hexo l, (3β,15α,16α,21β,22α)(22S)-21-Acetoxy-6α,11β-dihydroxy 19.614 -16α,17α-propylmethylenedioxypreg na-1,4-diene-3,20-dione Table 2 showed the identified compounds in (a) leaves, (b) stems and (c) roots of control and Na2HPO4, FeCl3, CuCl2 exposed samples. Given defined significant alteration as the number of normalized average peak area in the treated plants larger than 1.30 or less than 0.70, we found diverse significant alteration in the three kinds of exposure. The number of identified compounds was smaller than preceding number 29 of detected compounds due to the existence of certain unknown compounds. Among 46 identified compounds in leaves (table 2(a)), levels of 13, 7 and 10 metabolites showed up significant alteration while 1, 22 and 2 metabolites showed down significant alteration for Na2HPO4, FeCl3 and CuCl2 treated plants, respectively. Among 27 identified compounds in stems (table 2(b)), levels of 12, 0 and 9 metabolites showed up significant alteration while 0, 5 and 3 metabolites showed down significant alteration for Na2HPO4, FeCl3 and CuCl2 treated plants, respectively. Among 19 identified compounds in roots (table 2(c)), levels of 3, 2 and 2 metabolites showed up significant alteration while 2, 6 and 8 metabolites showed down significant alteration for Na2HPO4, FeCl3 and CuCl2 treated plants, respectively. Generally, the metabolic levels in plants are enhanced under comparatively low-strength pressure, but would decrease under excessively strong pressure due to loss of tolerance capacity. For FeCl3 treated plants, the number of metabolites with down significant alteration was tremendously larger than that with up significant alteration in all three organs of plants, indicating the aborted tolerance under such high concentration. The proportion of down significant alteration occurred especially high in metabolic levels of leaves, suggesting that leaves had been much more severely damaged than stems and roots under FeCl3 exposure. The number of down significant alteration in all parts of Na2HPO4 treated plants was small, while numerous up significant alteration appeared. This phenomenon is a sign of healthy tolerance. For CuCl2 treated plants, the result was of significant interest, as leaves and stems exhibited only a small ratio of down significant alteration but roots showed much greater ratio of down significant alteration. The tendency for up significant alteration was opposite. Detailed explanation was discussed below. We can conclude from the data that after 15 days’ culture, all plants were severely damaged by FeCl3 30 exposure but maintained high tolerance toward Na2HPO4 exposure, and leaves and stems retained tolerance to CuCl2 while roots have suffered severe perturbance. The variation of metabolites displayed diverse tendency with respect to different categories of compounds under different exposure. Levels of many amino acids (typically L-Valine and L-threonine), carbonhydrates (typically D-glucopyranose), inorganic acids (typically Butanoic acid), ketones (typically Pregn-4-ene-3,20-dione) and polyamines (typically 9H-Purin-6-amine) increased in leaves, stems and roots under Na2HPO4 exposure. For CuCl2 exposed leaves and stems, the enhanced metabolites comprised of amino acids (typically L-threonine), carbohydrates (typically D-Fructofuranose) and inorganic acids (typically Butanoic acid). The significantly decreased metabolites under FeCl3 exposure involved amino acids, inorganic acids, carbohydrates, and alcohols. When the data set for intermediates of the tricarboxylic acid cycle was considered, changes in the individual metabolites that did not correlate became apparent 31 3.3 PCA analysis (a) Leaf 32 (b) stem (c) root Figure 6. Three-dimensional PCA of the metabolic profiles of (a)leaf, (b) stem and (c) root in control and the exposed samples. 33 (a) Control vs Na2HPO4 treated group (b) Control vs FeCl3 treated group 34 (c) Control vs CuCl2 treated group Figure 7. Two-dimensional PCA of the metabolic profiles in roots of (a) control vs Na2HPO4 treated group, (b) control vs FeCl3 treated group and (c) control vs CuCl2 treated group Figure 6 showed the three-dimensional PCA of metabolites in (a) leaf, (b) stem and (c) root of control and the exposed samples. It can be seen from figure 6 that both the leaves and stems of the four differently treated plants could be distinctly clustered into four separated groups, while the boundary for roots of the four different plants was not clear. However, in two-dimensional PCA of control group and treated groups in figure 7, the boundary was very clear. We can see from figure 7 that when separately plotted between the roots of control group and the three kinds of treated group, they all fell into two clusters obviously. Given that two-dimensional PCA is much more precise than three-dimensional PCA, the successful separation of roots in two-dimensional PCA demonstrated the clear difference between metabolome of control group and treated groups. As different clusters in PCA means different traits in 35 certain attributes, it can be concluded from figure 6 and 7 that the metabolome of leaves, stems and roots in control plants were entirely different from that in treated plants. 3.1 Metabolic response Our results provided insight into mechanisms of water hyacinth’s adaptation to salt stress at the metabolite level, highlighting the roles of changed metabolites. Na2HPO4, FeCl3 and CuCl2 exposure could all be attributed to salt stress, the concentration difference of which led to diverse levels of plant response. However, metabolites in leaves shared some common characteristics. Levels of D-glucopyranose, L-threonine, Butanoic acid and 9H-Purin-6-amine all significantly increased in response to the three different exposures. These compounds fell into the general types of osmoprotectants: amino acids, quaternary ammonium compounds, and polyols [43-45]. We found that these primary metabolites all have compatible solutes-like property and they act collectively as osmoprotectants. These low molecular organic chemicals accumulated under the three different exposures possibly played the important role of stabilizing proteins and membranes, and contributing to cell osmotic pressure. The accumulation of glycerol under FeCl3 stress and D-Fructofuranose under CuCl2 stress could be considered to undertake this mission. Plants metabolism perturbation by excessive metal may cause a reduction of chlorophyll content, inhibit plant growth and respiration, and alter the activity and quantity of enzymes that are key to various metabolic pathways [46]. Psotova et al. [47] proved that primary and secondary metabolites in plants increased with increasing metal concentrations with a certain point, beyond which a decrease of metabolites concentration could be observed. Our results in down significant 36 alteration of metabolites under FeCl3 stress acted as evidence of this claim. Memon et al. [48] and Jahangir et al. [30] suggested that accumulation of metals is more dependent on the type of metal rather than metal concentration. The fact that metal concentration was in order of P > Cu > Fe but plant damage in order of Fe > Cu > P support this suggestion. Plants may localize the absorbed metals mostly in leaves and roots [48]. We found in this study that roots suffered the severest damage compared to leaves and stems under the same concentration of CuCl2 exposure. It suggested that roots may play the most significant role in detoxification of CuCl2. The survival of leaves and stems may be attributed to metal chelation by roots metabolites in the first line. An alternative explanation is that the number of roots metabolites is too small to protect roots from damaging. This phenomenon required further exploration and explanation. 3.4 Weakness and prospect of the study Although we have obtained valuable information in metabolites change of water hyacinth upon metal exposure, the results of the experiment were limited. We need more experiments to confirm our deduction and better understand the mechanism of metal accumulation in water hyacinth. Firstly, the effect of metal concentrations on stress response of water hyacinth needs to be studied. Although the concentration of metal solutions in this article was set according to literature, under which the water hyacinth may be tolerable to the imposed stress, the results obtained were not sufficient. Metabolomics exposed to a range of metal concentrations could uncover the metabolites pathway of water hyacinth more comprehensively. 37 Secondly, more species of heavy metals need to be introduced in the experiment system. As the toxic effect of different metals may be different, the metabolites change in water hyacinth for detoxification strategy may be different. Understanding the accumulation mechanism of a wider range of metal species in water hyacinth is significant for a wider application of this plant in phytoremediation. Thirdly, the short-term effect of metal stress in water hyacinth needs to be separately studied. In this article, the time period for metal exposure is 15 days when the plants metabolism became stable after adaptation to metal stress. We missed the important information on short-term (i.e. 1 h, 2 h, etc.) metabolites shifts of water hyacinth in response to metal stress. Last and the most important, the function of various metabolites in water hyacinth needs to be confirmed. Unfortunately, it is difficult to find out the function clues of large amounts of metabolites in a complete metabolic pathway. One possible solution may be the combined application of proteomics and transcriptomics, making an connection between genotype and phenotype of water hyacinth. 38 4 Conclusion In this study, water hyacinth was exposed to 0.3 mmol FeCl3, 1.2 mmol CuCl2 and 3.46 mmol Na2HPO4 concentrations for 15 days, respectively. The metabolic profiles of roots, stems and leaves of water hyacinth were constructed by GC-MS analysis and further analyzed by PCA method. The GC results of leaves and stems pretreated by dry grinding visually demonstrated more peaks compared to that pretreated by wet grinding. Stronger peak was observed in the GC results of control leaves after cultivation than before. The number of compounds detected in control leaves, stems and roots was 65, 45 and 19, respectively For FeCl3 treated plants, the number of metabolites with down significant alteration was tremendously larger than that with up significant alteration in all three organs of plants. The number of down significant alteration in all parts of Na2HPO4 treated plants was small, while numerous up significant alteration appeared. This phenomenon is a sign of healthy tolerance. For CuCl2 treated plants, leaves and stems exhibited only a small ratio of down significant alteration but roots showed much greater ratio of down significant alteration. Leaves and stems of the four differently treated plants could be distinctly clustered into four separated groups in three-dimensional PCA, while roots could only be separately clustered into different groups between control group and the treated groups individually by two-dimensional PCA. Levels of D-glucopyranose, L-threonine, Butanoic acid and 9H-Purin-6-amine all significantly increased in response to the three different exposures. These metabolites 39 have compatible solutes –like property and act as osmoprotectants. Accumulation of metals is more dependent on the type of metal rather than metal concentration. Roots may play the most significant role in detoxification of CuCl2. More experiments to confirm our deduction and better understand the mechanism of metal accumulation in water hyacinth, such as metal concentration variation, metal species introduction, short –term stress study and metabolites confirmation. 40 Bibliography [1] J. K. Nicholson, J. C. Lindon, and E. Holmes, "'Metabonomics': understanding the metabolic responses of living systems to pathophysiological stimuli via multivariate statistical analysis of biological NMR spectroscopic data," Xenobiotica, vol. 29, pp. 1181-1189, Nov 1999. [2] J. K. Nicholson, J. Connelly, J. C. Lindon, and E. Holmes, "Metabonomics: a platform for studying drug toxicity and gene function," Nature Reviews Drug Discovery, vol. 1, pp. 153-161, Feb 2002. [3] J. T. Brindle, H. Antti, E. Holmes, G. Tranter, J. K. Nicholson, H. W. L. Bethell, et al., "Rapid and noninvasive diagnosis of the presence and severity of coronary heart disease using H-1-NMR-based metabonomics," Nature Medicine, vol. 8, pp. 1439-1444, Dec 2002. [4] E. Holmes and H. Antti, "Chemometric contributions to the evolution of metabonomics: mathematical solutions to characterising and interpreting complex biological NMR spectra," Analyst, vol. 127, pp. 1549-1557, 2002 2002. [5] O. Fiehn, "Metabolic networks of Cucurbita maxima phloem," Phytochemistry, vol. 62, pp. 875-886, Mar 2003. [6] A. Tiessen, J. H. M. Hendriks, M. Stitt, A. Branscheid, Y. Gibon, E. M. Farre, et al., "Starch synthesis in potato tubers is regulated by post-translational redox modification of ADP-glucose pyrophosphorylase: A novel regulatory mechanism linking starch synthesis to the sucrose supply," Plant Cell, vol. 14, pp. 2191-2213, Sep 2002. [7] W. Maier, J. Schmidt, V. Wray, M. H. Walter, and D. Strack, "The arbuscular mycorrhizal fungus, Glomus intraradices, induces the accumulation of cyclohexenone derivatives in tobacco roots," Planta, vol. 207, pp. 620-623, Feb 1999. [8] K. Shinozaki and H. Sakakibara, "Omics and Bioinformatics: An Essential Toolbox for Systems Analyses of Plant Functions Beyond 2010," Plant and Cell Physiology, vol. 50, pp. 1177-1180, Jul 2009. [9] J. Grimplet, M. D. Wheatley, H. B. Jouira, L. G. Deluc, G. R. Cramer, and J. C. Cushman, "Proteomic and selected metabolite analysis of grape berry tissues under well-watered and 41 water-deficit stress conditions," Proteomics, vol. 9, pp. 2503-2528, May 2009. [10] L. G. Deluc, D. R. Quilici, A. Decendit, J. Grimplet, M. D. Wheatley, K. A. Schlauch, et al., "Water deficit alters differentially metabolic pathways affecting important flavor and quality traits in grape berries of Cabernet Sauvignon and Chardonnay," Bmc Genomics, vol. 10, May 2009. [11] S. P. Mane, C. V. Robinet, A. Ulanov, R. Schafleitner, L. Tincopa, A. Gaudin, et al., "Molecular and physiological adaptation to prolonged drought stress in the leaves of two Andean potato genotypes," Functional Plant Biology, vol. 35, pp. 669-688, 2008 2008. [12] V. Shulaev, D. Cortes, G. Miller, and R. Mittler, "Metabolomics for plant stress response," Physiologia Plantarum, vol. 132, pp. 199-208, Feb 8 2008. [13] F. Kaplan, J. Kopka, D. W. Haskell, W. Zhao, K. C. Schiller, N. Gatzke, et al., "Exploring the temperature-stress metabolome of Arabidopsis," Plant Physiology, vol. 136, pp. 4159-4168, Dec 2004. [14] F. Kaplan, J. Kopka, D. Y. Sung, W. Zhao, M. Popp, R. Porat, et al., "Transcript and metabolite profiling during cold acclimation of Arabidopsis reveals an intricate relationship of cold-regulated gene expression with modifications in metabolite content," Plant Journal, vol. 50, pp. 967-981, Jun 2007. [15] D. Cook, S. Fowler, O. Fiehn, and M. F. Thomashow, "A prominent role for the CBF cold response pathway in configuring the low-temperature metabolome of Arabidopsis," Proceedings of the National Academy of Sciences of the United States of America, vol. 101, pp. 15243-15248, Oct 19 2004. [16] M. R. Morsy, L. Jouve, J.-F. Hausman, L. Hoffmann, and J. M. Stewart, "Alteration of oxidative and carbohydrate metabolism under abiotic stress in two rice (Oryza sativa L.) genotypes contrasting in chilling tolerance," Journal of Plant Physiology, vol. 164, pp. 157-167, Feb 2007. [17] H. E. Johnson, D. Broadhurst, R. Goodacre, and A. R. Smith, "Metabolic fingerprinting of salt-stressed tomatoes," Phytochemistry, vol. 62, pp. 919-928, Mar 2003. 42 [18] J. K. Kim, T. Bamba, K. Harada, E. Fukusaki, and A. Kobayashi, "Time-course metabolic profiling in Arabidopsis thaliana cell cultures after salt stress treatment," Journal of Experimental Botany, vol. 58, pp. 415-424, Feb 2007. [19] G. R. Cramer, A. Ergul, J. Grimplet, R. L. Tillett, E. A. R. Tattersall, M. C. Bohlman, et al., "Water and salinity stress in grapevines: early and late changes in transcript and metabolite profiles," Functional & Integrative Genomics, vol. 7, pp. 111-134, Apr 2007. [20] Q. Q. Gong, P. H. Li, S. S. Ma, S. I. Rupassara, and H. J. Bohnert, "Salinity stress adaptation competence in the extremophile Thellungiella halophila in comparison with its relative Arabidopsis thaliana," Plant Journal, vol. 44, pp. 826-839, Dec 2005. [21] V. J. Nikiforova, C. O. Daub, H. Hesse, L. Willmitzer, and R. Hoefgen, "Integrative gene-metabolite network with implemented causality deciphers informational fluxes of sulphur stress response," Journal of Experimental Botany, vol. 56, pp. 1887-1896, Jul 2005. [22] V. J. Nikiforova, J. Kopka, V. Tolstikov, O. Fiehn, L. Hopkins, M. J. Hawkesford, et al., "Systems rebalancing of metabolism in response to sulfur deprivation, as revealed by metabolome analysis of arabidopsis plants," Plant Physiology, vol. 138, pp. 304-318, May 2005. [23] G. Hernandez, M. Ramirez, O. Valdes-Lopez, M. Tesfaye, M. A. Graham, T. Czechowski, et al., "Phosphorus stress in common bean: Root transcript and metabolic responses," Plant Physiology, vol. 144, pp. 752-767, Jun 2007. [24] N. J. C. Bailey, M. Oven, E. Holmes, J. K. Nicholson, and M. H. Zenk, "Metabolomic analysis of the consequences of cadmium exposure in Silene cucubalus cell cultures via H-1 NMR spectroscopy and chemometrics," Phytochemistry, vol. 62, pp. 851-858, Mar 2003. [25] P. Kieffer, S. Planchon, M. Oufir, J. Ziebel, J. Dommes, L. Hoffmann, et al., "Combining Proteomics and Metabolite Analyses To Unravel Cadmium Stress-Response in Poplar Leaves," Journal of Proteome Research, vol. 8, pp. 400-417, Jan 2009. [26] X. Sun, J. Zhang, H. Zhang, Y. Ni, Q. Zhang, J. Chen, et al., "The responses of Arabidopsis thaliana to cadmium exposure explored via metabolite profiling," Chemosphere, vol. 78, pp. 840-845, Feb 2010. 43 [27] H. Hediji, W. Djebali, C. Cabasson, M. Maucourt, P. Baldet, A. Bertrand, et al., "Effects of long-term cadmium exposure on growth and metabolomic profile of tomato plants," Ecotoxicology and Environmental Safety, vol. 73, pp. 1965-1974, Nov 2010. [28] X. Liu, C. Yang, L. Zhang, L. Li, S. Liu, J. Yu, et al., "Metabolic profiling of cadmium-induced effects in one pioneer intertidal halophyte Suaeda salsa by NMR-based metabolomics," Ecotoxicology, vol. 20, pp. 1422-1431, Aug 2011. [29] J. Kovacik, B. Klejdus, J. Hedbavny, and M. Backor, "Effect of copper and salicylic acid on phenolic metabolites and free amino acids in Scenedesmus quadricauda (Chlorophyceae)," Plant Science, vol. 178, pp. 307-311, Mar 2010. [30] M. Jahangir, I. B. Abdel-Farid, Y. H. Choi, and R. Verpoorte, "Metal ion-inducing metabolite accumulation in Brassica rapa," Journal of Plant Physiology, vol. 165, pp. 1429-1437, 2008 2008. [31] M. Misbahuddin and A. Fariduddin, "Water hyacinth removes arsenic from arsenic-contaminated drinking water," Archives of Environmental Health, vol. 57, pp. 516-518, Nov-Dec 2002. [32] S. W. Liao and N. L. Chang, "Heavy metal phytoremediation by water hyacinth at constructed wetlands in Taiwan," Journal of Aquatic Plant Management, vol. 42, pp. 60-68, Jan 2004. [33] M. E. Soltan and M. N. Rashed, "Laboratory study on the survival of water hyacinth under several conditions of heavy metal concentrations," Advances in Environmental Research, vol. 7, pp. 321-334, Jan 2003. [34] M. Ajmal, R. Khan, and A. U. Khan, "Heavy-metals in water, sediments, fish and plants of river hindon, up india," Hydrobiologia, vol. 148, pp. 151-157, May 20 1987. [35] M. F. Zaranyika and T. Ndapwadza, "Uptake of Ni, Zn, Fe, Co, Cr, Pb, Cu and Cd by water hyacinth (Eichrornia crassipes) in Mukuvisi and Manyame rivers, Zimbabwe," Journal of Environmental Science and Health Part a-Environmental Science and Engineering & Toxic and Hazardous Substance Control, vol. 30, pp. 157-169, 1995 1995. [36] T. N. Cooley and D. F. Martin, "Cadmium in naturally occurring water hyacinths," 44 Chemosphere, vol. 8, pp. 75-78, 1979 1979. [37] C. Kelley, R. E. Mielke, D. Dimaquibo, A. J. Curtis, and J. G. Dewitt, "Adsorption of Eu(III) onto roots of water hyacinth," Environmental Science & Technology, vol. 33, pp. 1439-1443, May 1 1999. [38] A. Malik, "Environmental challenge vis a vis opportunity: The case of water hyacinth," Environment International, vol. 33, pp. 122-138, Jan 2007. [39] Y. L. Zhu, A. M. Zayed, J. H. Qian, M. de Souza, and N. Terry, "Phytoaccumulation of trace elements by wetland plants: II. Water hyacinth," Journal of Environmental Quality, vol. 28, pp. 339-344, Jan-Feb 1999. [40] S. H. Kay, W. T. Haller, and L. A. Garrard, "Effects of heavy metals on water hyacinths (Eichhornia crassipes (mart) solms)," Aquatic Toxicology, vol. 5, pp. 117-128, 1984 1984. [41] X. Lu, M. Kruatrachue, P. Pokethitiyook, and K. Homyok, "Removal of cadmium and zinc by water hyacinth, Eichhornia crassipes," ScienceAsia, vol. 30, pp. 93-103, June 2004. [42] S. W. Newete, B. F. N. Erasmus, I. M. Weiersbye, M. A. Cho, and M. J. Byrne, "Hyperspectral reflectance features of water hyacinth growing under feeding stresses of Neochetina spp. and different heavy metal pollutants," International Journal of Remote Sensing, vol. 35, pp. 799-817, Feb 1 2014. [43] H. M. Shahjee, K. Banerjee, and F. Ahmad, "Comparative analysis of naturally occurring L-amino acid osmolytes and their D-isomers on protection of Escherichia coli against environmental stresses," Journal of Biosciences, vol. 27, pp. 515-520, Sep 2002. [44] E. P. Murakeozy, Z. Nagy, C. Duhaze, A. Bouchereau, and Z. Tuba, "Seasonal changes in the levels of compatible osmolytes in three halophytic species of inland saline vegetation in Hungary," Journal of Plant Physiology, vol. 160, pp. 395-401, Apr 2003. [45] N. S. Sarraf, A. A. Saboury, B. Ranjbar, and M. Nemat-Gorgani, "Effect of some amino acids on the structure and activity of carbonic anhydrase," Asian Journal of Chemistry, vol. 17, pp. 2385-2394, Oct-Dec 2005. [46] T. R. Guo, G. P. Zhang, and Y. H. Zhang, "Physiological changes in barley plants under 45 combined toxicity of aluminum, copper and cadmium," Colloids and Surfaces B-Biointerfaces, vol. 57, pp. 182-188, Jun 15 2007. [47] J. Psotova, J. Lasovsky, and J. Vicar, "Metal-chelating properties, electrochemical behavior, scavenging and cytoprotective activities of six natural phenolics," Biomedical Papers (Olomouc), vol. 147, pp. 147-153, December 2003. [48] A. R. Memon, D. Aktoprakligil, A. Ozdemir, and A. Vertii, "Heavy metal accumulation and detoxification mechanisms in plants," Turkish Journal of Botany, vol. 25, pp. 111-121, 2001 2001. 46 [...]... analysis of plant metabolites under stress environment via modern detection and analysis methods, the variation trend and rule of plant metabolites over time can be monitored The integration of various omics platforms such as genomics, proteomics and metabolomics is also a powerful toolkit [8] Combination of all these information 5 helps to study responses of biological systems to genes or environment... capability of water hyacinth in storing most heavy metals enabled the plant to avoid toxicity of photosynthetic tissues caused by heavy metals 1.4 Basis and significance of dissertation Modern industries have released large amounts of heavy metals into the environment Among various technologies for heavy metal removal, such as adsorption, reverse osmosis and electrochemical method etc., phytoremediation... complete metabolism response when water hyacinth is subjected to heavy metal stress is unknown The application of metabolomics method on water hyacinth in response to heavy metal stress may provide incentive on the metabolites pathway of the water hyacinth for heavy metal adsorption from the perspective of the whole plant, thus building an solid scientific foundation for better application of water hyacinth. .. responses of the same plant In this article, the water hyacinth plants were exposed to CuCl2, FeCl3 and Na2HPO4 solutions Then we analyzed the metabolites change and differences of roots, stems and leaves of the plants using GC- MS method and PCA analysis The metabolism pathways of water hyacinth in response to different heavy metal stress were proposed 16 2 Materials and Methods 2.1 Materials Water Hyacinth. .. spectrum (1H NMR), chromatography and MS become the main analysis tools, due to the universality of 1H NMR for hydrogen metabolites, and high resolution and high flux of chromatography, and universality, high sensitivity and specificity of MS Later period of Metabolomics research is to interpretate the biological significance of data based on data analysis and interpretation with the aid of bioinformatics... on Nicotiana tabacum root metabolism, compared the metabolites difference between tobacco roots with and without Glomus intraradices To sum up, Metabolomics technology is an ideal platform for plant metabolism study 1.2.1 Plant metabolomics in response to abiotic stresses Recently many scientific research institutions carried out metabolomics studies on abiotic stress responses of plants Through qualitative... leaf extract of Cucurbita maxima using GC- MS and obtained more than 400 peaks By comparison with the mass spectrum database, he preliminarily identified 90 compounds, and compared the differences on metabolites in sugar and amino acid composition between petiole and leaf; Tiessen et al (2002) [6] conducted Metabolomic analysis of Solanum tuberosum tuber using high performance liquid chromatography (HPLC)... obtained The combination of Metabolomics and proteomics was also applied in the study of leguminous plants under phosphorus stress Hernandez et al (2007) [23] made metabolic profile analysis of roots of leguminous plant with sufficient phosphorus and insufficient phosphorus using GC- TOF-MS, and identified a series of metabolic products related to the phosphorus stress, many of which (including amino acids,... a rate of 5 oC/min to 280oC and held for 5 minutes at 300 oC The transfer line temperature was set at 280 o C Detection was achieved using MS in electron impact mode and full scan monitoring (m/ z 100 to 600) The temperature of the ion source was set at 250 oC 2.5 Data analysis To identify metabolites, all chromatographic peaks obtained by GC- MS analysis were compared with NIST by masshunter software... sample, while MS has the advantages of high sensitivity, availability for targeted analysis, cheaper instrument cost Choosing whether NMR or MS for metabolomics study depends on the purpose of the study, the research object and the instrument 2 availability Table 1 Advantages and disadvantages of NMR and MS for metabolomics study NMR MS Sensitivity Low High Reproducibility Very high Average Number of