1
NMR SPECTRAL ANALYSIS OF ASCORABTE-DEFICIENT
MUTANTS OF ARABIDOPSIS THALIANA
Submitted by Felicia Charles Johnson
to the University of Exeter
as a dissertation for the degree of
Master of Philosophy
in Biological Sciences,
January, 2011.
This dissertation is available for Library use on the understanding that it is copyright
material and that no quotation from the dissertation may be published without proper
acknowledgement.
I certify that all material in this dissertation which is not my own work has been
identified and that no material has previously been submitted and approved for the
award of a degree by this or any other University.
.........................
2
ABSTRACT
Ascorbate is an important antioxidant, involved in the scavenging of reactive oxygen species
(ROS) such as hydrogen peroxide hydrogen. It is also a cofactor for 2-oxoglutarate-
dependent dioxygenases, which are involved in the biosynthesis of a number of metabolites.
In the current study, Arabidopsis thaliana ascorbate-deficient mutants (vtc1, vtc2, vtc3, and
vtc4) were used to investigate the effect of ascorbate deficiency on the metabolic changes.
NMR was used as an analytical technique for the metabolite profiling of Arabidopsis
thaliana vtc mutants. A total of 45 NMR spectra representing wild type and four mutants of
two different growth stages (rosette and flowering) were collected and spectral processed for
exploratory analysis. Analysis of variance (ANOVA) and multivariate data analysis
(hierarchical clustering) were used to find the differences between different strains and
metabolite clustering patterns. Combinations of different parameters were used to find the
best method for the analysis of NMR-based metabolites data. The method was developed
which allowed making multiple comparison such as two growth stages and five strains. It
was found the vtc mutants grouped separately from WT which contained the low ascorbate
showing ascorbate deficiency might have some effect on the metabolic level.
Two interesting resonances 1.32 and 1.33ppm were separated which were negatively
correlated with the ascorbate levels in the vtc mutants. These resonances (1.32 and 1.33) are
putatively identified as threonine and lactate. Further investigations are needed to
investigate the nature of relationship between ascorbate and threonine and lactate.
3
CONTENTS
PAGE NO.
Chapter 1
INTRODUCTION
ASCORBATE METABOLISM AND FUNCTIONS 7
ASCORBATE BIOSYNTHESIS (D-MAN/L-GAL PATHWAY) 9
ASCORBATE BIOSYNTHESIS OTHER MULTIPLE PATHWAYS 10
CONTROL OF ASCORBATE BIOSYNTHESIS 11
ASCORBATE DEGRADATION 11
ASCORBATE-DEFICIENT MUTANTS 11
ISOLATION OF VTC MUTANTS 11
ASCORBATE LEVELS IN VTC MUTANTS 12
CHARACTERIZATION OF VTC MUTANTS 12
PLANT METABOLOMICS 14
METABOLITE/METABOLIC PROFILING 15
METABOLOME TECHNOLOGIES FOR DATA ACQUISITION 16
BIOINFORMATICS 21
DATA ANALYSIS 21
NMR SPECTRAL PROCESSING 22
STATISTICAL ANALYSIS 23
APPLICATION OF PLANT METABOLOMICS 24
LIMITATIONS OF PLANT METABOLOMICS 26
PRESENT STUDY 27
4
Chapter 2
MATERIALS AND METHODS
GERMINATION OF ARABIDOPSIS THALIANA SEEDS ON AGAROSE 29
TRANSFERRING ARABIDOPSIS THALIANA SEEDLINGS TO SOIL 32
HARVESTING OF ARABIDOPSIS THALIANA PLANTS FOR ANALYSIS 32
PREPARING ARABIDOPSIS THALIANA PLANTS FOR NMR ANALYSIS 33
ANALYSING ARABIDOPSIS THALIANA PLANTS BY NMR 34
NMR SPECTRAL PROCESSING WITH MESTRE-C 36
DATA ANALYSIS 37
Chapter 3
RESULTS AND DISCUSSION
NMR SPECTRAL FEATURES OF ASCORBATE- DEFICIENT MUTANTS 39
NMR DATA PRE-PROCESSING 41
MULTIVARIATE ANALYSIS OF ASCORBATE- DEFICIENT MUTANTS 42
DIFFERENTIALLY EXPRESSED BINS BETWEEN WT AND VTC MUTANTS 50
Chapter 4
CONCLUSION 54
REFERENCES 55
5
LIST OF TABLES
TABLE NO.
PAGE
NO.
Table 1.1 Strategies for metabolomic analysis 16
Table 1.2 Common analytical techniques used in metabolomics 19
Table 2.1 Plant types used for the study 29
Table 2.2 Parameters for NMR Data Collection 35
Table 3.1 Clustering analyses of all plant types in rosette stage with 45
0.01 bin width and different processing combinations
Table 3.2 Clustering analyses of all plant types in rosette stage with 46
0.04 bin width and different processing combinations
Table 3.3 Clustering analyses of all plant types in flowering stage with 47
0.01 bin width and different processing combinations
Table 3.4 Clustering analyses of all plant types in flowering stage with 48
0.04 bin width and different processing combinations
Table 3.5 List of resonances showing discrimination in the spectral regions 51
Table 3.6 Resonances different between strains at spectral level with 51
two bin widths
6
LIST OF FIGURES
FIGURE NO.
PAGE
NO.
Figure 1.1 The proposed pathways for L-ascorbic acid biosynthesis 8
in plants
Figure 1.2 Wild type (Col-0) and four ascorbate - deficient mutants 13
Figure 2.1 Spectral processing with MestRe-C 37
Figure 2.2 Schema of the data analysis 38
Figure 3.1 1H-NMR spectra of the Arabidopsis thaliana (Wild type) 40
extracted within D2O:CD3OD
Figure.3.2a 1H-NMR spectra of Arabidopsis thaliana at rosette stage 40
Figure.3.2b 1H-NMR spectra of Arabidopsis thaliana strains at 41
flowering stage
Figure 3.3 HCA showing the difference in growth stages 44
(rosette and flowering) and WT and the vtc mutants
Figure 3.4 Heatmap for the 1H NMR based metabolic profiles of 50
Arabidopsis thaliana Wild type and four ascorbate-deficient
mutants in rosette
Figure 3.5 The intensity of the resonances 1.32 and 1.33ppm across the strains 52
Figure 3.6 Identification of the resonances 1.32 and 1.33 ppm in the 53
original NMR spectra of one of the vtc mutant
7
Chapter 1
INTRODUCTION
Ascorbate metabolism and functions
L-ascorbic acid (vitamin C) has a key role in prevention of scurvy in humans and also have
shown to fulfil many other essential functions in animals and plants. The biological role of
ascorbate (Asc) depends on its ability as a reducing agent, which acts as an effective anti-
oxidant and a free radical scavenger (Smirnoff,2011).It is the most abundant soluble
antioxidant found in plants in almost all plant tissues, with the exception of dry seeds (De
Gara et al.,1997) and is effective at directly scavenging reactive oxygen species (ROS) such
as super oxide, hydrogen peroxide, singlet oxygen and the hydroxyl radical (Noctor &
Foyer, 1998).It is a cofactor for various enzymes, including Fe2+
/2-oxoglutarate-dependent
dioxygenases and violaxanthin de-epoxidase (De Tullio, 2004).Ascorbate is required for the
regeneration of the lipophilic antioxidant a-tocopherol from the α -chromanoxyl radical
(Fryer, 1992) and is also responsible for the regeneration of the carotenoid pigments:
carotenes and xanthophylls (Eskling and Akerlund,1997).
8
Significant progress has been made in the elucidation of the plant ascorbate biosynthetic
pathway (Wheeler et al.,1998) to understand how ascorbate metabolism is controlled.
These progresses provide a basis for engineering or manipulating its accumulation. The
proposed pathway for L-ascorbate biosynthesis in plants is shown in Figure1, D-
Mannose/L-Galactose (D-Man/L-Gal) pathway is the predominant pathway. This pathway
is different in animals where the last enzyme in the pathway (L-gulonolactone oxidase) is
not expressed due to the lack of the enzyme. There are also other alternative biosynthetic
pathways in plants along with the D-Man/L-Gal pathway and their existence in ascorbate
biosynthesis is still under investigation. The alternative pathways are found to involve
intermediates such as D-glucurono-1, 4-lactone, D-galacturonate, methyl-D-galacturonate
and L-gulono-1,4-lactone (Smirnoff et al.,2001).
9
Figure 1.1 The proposed pathways for L-ascorbic acid biosynthesis in plants.The D-mannose/L-
galactose pathway is shown in bold: D-mannose-1-phosphate is synthesised from D-glucose-1-
phosphate via D-fructose-6-phosphate. It is converted into L-ascorbic acid via a series of steps
catalysed by: 1, GDP-mannose pyrophosphorylase (Arabidopsis thaliana VTC1); 2, GDP-D-
mannose-3, 5-epimerase; 3, GDP-L-galactose phosphorylase (Arabidopsis VTC2); 4, L-galactose-1-
phosphate phosphatase (Arabidopsis thaliana VTC4); 5, L-galactose dehydrogenase; 6, L-galactono-
1,4-lactone dehydrogenase.Alternative pathways and some of the enzymes catalysing the various
reactions are shown:7,GDP-D-mannose-4, 6-dehydratase; 8,GDP-4-keto-6-deoxy-D-mannose-3,5-
epimerase-4-reductase;9,D-galacturonatereductase;10,myo-inositoloxygenase; 11, D-glucuronate
reductase;12, aldonolactonase; 13, L-gulonolactone dehydrogenase (Wheeler et al.,1998).
10
Ascorbate biosynthesis (D-Man/L-Gal pathway)
Ascorbate is present in mill molar (mM) concentration in plant tissues which varies
considerably between tissues. The concentration of ascorbate also depends on the
physiological status of the plant as well and on the environmental factors (Luwe et al.,1993).
Advances in understanding of biosynthetic pathway of Asc in plants have been made by
studying Pisum sativum and Arabidopsis plants (Wheeler et al.,1998).Plants synthesize
ascorbate involving D-glucose-6-P, D-fructose-6-P, D-manose-1-P, GDP-D-mannose, GDP-
L-gulose, GDP-L-galactose, L-galactose-1-P and L-galactose as intermediaries in the
production of L-galactono-1-4-lactone, a precursor, oxidised to ascorbate by the enzyme L-
galactose 1-4 lactone dehydrogenase. There are many enzymes involved which catalyses
various steps in Asc biosynthesis. Some of those key enzymes are briefly described
below.Phosphomannose isomerase (PMI), is involved in the conversion of fructose 6-P to
mannose 6-P and its activity is readily detected in Arabidopsis leaf extracts and also in other
species (Smirnoff and Wheeler, 2000). Phosphomannose mutase (PMM) catalyses the
interconversion of mannose 6-P and Mannose 1-P. Studies show increase in ascorbate when
PMM was over expressed in Nicotiana benthamiana,Arabidopsis and virus-induced gene
silencing caused a decrease in ascorbate (Smirnoff, 2011). GDP-mannose pyrophosphorylase
(GMP) catalyses the formation of GDP-mannose 1-P and GTP, GDP-mannose-3, 4-epimerase
(GME) catalyses the double epimerisation of GDP-mannose with the production of GDP-L-
galactose.
11
GDP-L-galactose phosphorylase/guanylyl transferase was identified in pea seedling extracts
(Dowdle et al.,2007;Ishikawa et al.,2006a and b) and in Arabidopsis by (Dowdle et al.,2007;
Laing et al.,2007 and Linster et al.,2007, 2008) catalyse the phosphorylytic breakdown of
GDP-L-galactose to L-Galactose 1-P.L-galactose 1-P phosphatase, hydrolyses L-galactose 1-
P to L-Gal. L-galactose dehydrogenase (L-GalDH) catalyses NAD+-dependent oxidation of L-
Gal at C1 to produce L-galactono lactone (Gatzek et al.,2002) and L-galactono-1,4 lactone
dehydrogenase (L-GalLDH) catalyses the last step which is the oxidation of L-GalL to
ascorbate.
Ascorbate biosynthesis Other Multiple Pathways
It has been proposed that plants not only synthesize ascorbate from D-Man/L-Gal but also
from D-galacturonic acid (D-GalUA),L-gulose and from myo-inostiol/D-glucuronic acid.
External application of uronic acid derivatives, labelling studies, gene expression studies,
over expression studies have shown to increase in ascorbate providing evidence for the
existence of other pathways involved in ascorbate biosynthesis(Smirnoff 2011).
Control of Ascorbate biosynthesis
The rate of biosynthesis depends on the amount of each enzymes and kinetic properties in
relation to substrate concentrations and other factors. Ascorbate concentration is regulated
by cell type and environmental conditions. Ascorbate concentration is high in leaves, peel of
fruits (Davey et al.,2000; Li et al.,2008), meristem cells (Cordoba-Pedregosa et al.,2003)
germinating seeds (Pallanca and Smirnoff 2000) and also in high light (Gautier et al.,2009;
12
Li et al.,2009;Dowdle et al.,2007),low temperature (Schoner and Krause,1990),jasmonates
Wolucka etal.,2005) mechanical wounding (Suza et al.,2010) and also when over expressing
biosynthesis genes (Smirnoff,2011) however,a very little is known about how the
biosynthesis or breakdown is controlled.
Ascorbate degradation
Asc is generally involved as an electron donor transferring electrons by oxidation as
monodehydroascorbate (MDHA) and dehydroascorbate (DHA).MDHA does not readily
react with other molecules and is relatively harmless.DHA undergoes spontaneous and
irreversible hydrolysis to 2,3-diketogulonic acid (Washko et al.,1992).Ascorbate and DHA
are catabolised in plants and give rise to a number of end products, including L-threonate,
oxalate and L-tartrate (Loewus,1999).
Ascorbate-deficient mutants
To get a better understanding of the biosynthetic pathways mutants have been a valuable
tool and are utilized to uncover a number of plant primary and secondary metabolic
pathways and to identify the pathways of ascorbate catabolism. Gene function can be
discovered through the use of mutants that are altered in gene of interest and the
physiological and biochemical changes are observed comparing to the Wild type. Conklin et
al.,(2000) identified a number of ascorbate-deficient mutants (Figure 2a,2b).
13
Isolation of vtc mutants
The vtc (vitamin-c) mutants were produced through ethyl methane sulfonate mutagenesis of
the Col-0 wt seeds, which is the Wild type plants of Arabidopsis and the plants were
screened by nitro blue tetrazolium (NBT) based assay for an Asc-deficient phenotype and
the produced mutants were also used for the analyses of genetic segregation and allelism.
Asc deficiency in the mutants vtc1-2, vtc2-1, vtc2-2, vtc3-1 and vtc4-1 is conferred by single
monogenic recessive traits. vtc mutants represent five different loci; VTC1, VTC2, VTC3,
VTC4 AND VTC5.Two of the mutants (vtc1-1 and vtc2-1) were isolated through their
increased ozone sensitivity. In total there are two vtc1 mutant alleles, three vtc2 mutants as
well as vtc3-1 and vtc4-1.vtc5-1 and vtc5-2 were produced by T-DNA insertional
mutagenesis (Linster and Clarke,2008).
The VTC1 locus has been cloned and it encodes GDP-mannose pyrophosphorylase (GMP)
(Conklin et al.,1999)which is an enzyme that catalyses the conversion of D-mannose-1-
phosphate to GDP-mannose in the ascorbate biosynthesis pathway. The vtc1 mutant has a
decreased ability to convert glucose and mannose to ascorbate.
VTC2, VTC5 and VTC4 have been identified as genes of the D-Mannose/L-Galactose
ascorbate biosynthesis pathway, encoding respectively GDP-L-galactose phosphorylase;
GDP-β-L-Gal: orthophosphate guanyltransferase (Dowdle et al.,2007;Ishikawa et al.,2006a
and b) and L-galactose 1-P phosphatase (Conklin, et al.,2006).
14
Ascorbate levels in vtc mutants
The vtc mutants contain 1/3 to 1/2 the total Asc present in the wild type Col-0(Conklin et al.,
2000).They range from 50-70 % of wild-type levels in vtc2-3, vtc3 and vtc4, to 25-30% in
vtc1-1, vtc1-2 and 10-20% in vtc2-1 and vtc2-2. The Asc levels increases upon the transition
from vegetative to reproductive state (Gander 1982).The reproductive tissues contain twice
the amount of Asc found in mature leaves of both wt and mutants.
Characterization of vtc mutants
There has been many studies that provides strong evidence that low ascorbate reduces
growth and increases susceptibility to a range of stresses. Characterization of vtc1-1 and
vtc2-1 has revealed that both the mutants have slow growth rate (Pavet et al.,2005;Conklin
et al.,2000).vtc1-1 and vtc2-1 have been widely studied in relation to the stress responses,
while there is little information on other mutants. The Wild type Arabidopsis is quite
tolerant to ozone (O3),probably because these plants mount an effective antioxidant response
(Sharma and Davis,1994;Conklin and Last, 1995).vtc1-1 mutant is extremely O3 sensitive
with visible injury including lesion formation, enhanced chlorosis, and/or tissue
collapse.vtc2 mutants have varied O3 sensitive phenotypes.vtc2-1 was similar to vtc1-1,
vtc2-3 appeared somewhat sensitive but vtc2-2 was not visibly injured (Conklin et al.,2000).
All the four mutants are more salt-sensitive (Huang et al., 2005;Smirnoff, 2000). Both vtc1
and vtc2 has reduced basal thermo tolerance (Larkindale et al.,2005). vtc2-1 has greater
thermo induced photo emission indicating increased lipid peroxidation at high temperature
(Havaux,2003). In vtc 1-1,171 genes were found differentially expressed and identified by
transcriptomic analysis of which 12.9% genes were involved in cell defence (Pastori,2003).
15
vtc1-1 and vtc2-1 have increased resistance to infection by virulent pathogens (Barth et
al.,2004,Pavet et al.,2005).Ascorbate deficient mutants also show increased salicylic acid,
increased expression of pathogenesis-related proteins, peroxidise activity and accumulation
of the phytoalexin camalexin (Kleibenstein, 2004).
Figure 1.2 Wild type (Col-0) and four ascorbate-deficient mutants (vtc 1, vtc 2, vtc 3 and vtc 4)
(vtc= vitamin c mutants).
Plant metabolomics
Metabolic networks are very complex as there is an enormous chemical diversity of
compounds. In the plant kingdom there are over approximately 200,000 metabolites and
about 5000 individual low molecular weight compounds (Fiehn,2002).Metabolomics is the
"systematic study of the unique chemical fingerprints that specific cellular processes leave
behind" - specifically, the study of their small-molecule metabolite profiles (Bennett,2005).
Metabolomics has emerged as an important functional genomics tool to aid in the
comprehensive analysis of all the metabolites in the cellular system adding to the
understanding of the complex molecular interactions in the biological systems (Sumner et
al.,2003; van der Greef et al.,2003;Saito and Matsuda,2010).Metabolomics currently in
16
plants are more complex due to large number of metabolites and the presence of
differentiated tissues, including specialist storage organs, with different metabolite
compliments(Ward et al.,2002).The field of plant metabolomics is a complicated
interdisciplinary research field that requires bioscience, analytical chemistry, organic
chemistry, chemometrics, and informatics knowledge. Genome and transcript or protein
profiling offers only limited information about the chosen system but profiling metabolome
actually provides the most “functional” information (Sumner et al.,2003).The ultimate aim
of metabolomics is an unbiased and comprehensive monitoring of all the metabolites in a
cellular system and to nearly complete the molecular picture of the state of a particular
biological system at a given time (Fiehn 2002).There appear to be differences of opinion as
how best to define the comprehensive profiling of the metabolome. Sumner (2003) proposed
that any technology whose output is processed with pattern recognition software and without
differentiation of individual metabolites should be termed metabolic finger printing not
metabolomics or metabonomics. Metabolic finger printing which is also the chemometric
approach where the compounds are not initially identified only their spectral patterns and
intensities are recorded, statistically compared and used to identify the relevant spectral
features that distinguish sample classes (Wishart, 2008;Xia et al.,2009)
17
Metabolite/metabolic profiling
Metabolite/metabolic profiling is the measurement of hundreds or potentially thousands of
metabolites in other words a quantitative approach which aims to formally identify and
quantify all detectable metabolites from the spectra, that requires the compounds of interest
to be known from a spectral reference library obtained from authentic standards prior to
subsequent data analysis(Xia et al.,2009).There is yet another approach of profiling
metabolome called targeted analysis which aims to measure the concentration of a limited
number of known metabolites precisely after knowing the structure of the target metabolite
which also requires the compounds of interest to be known a priori in purified form
(Shulaev, 2006).The strategies for metabolomic analysis given by Dunn and Ellis (2005) is
presented in Table 1.1below.
18
Table 1.1 Strategies for metabolomic analysis (Dunn and Ellis, 2005)
Metabolomics
Non-biased identification and quantification of all metabolites in a
biological system.
Sample preparation must not exclude metabolites, and selectivity
and sensitivity of the analytical technique must be high.
Metabolite profiling
Identification and quantification of a selected number of pre-
defined metabolites, generally related to a specific metabolic
pathway(s).
Sample preparation and instrumentation are employed so to isolate
those compounds of interest from possible matrix effects prior to
detection, normally with chromatographic separation prior to
detection with MS. Widely used in pharmaceutical industry
Metabolic fingerprinting
High-throughput, rapid, global analysis of samples to provide
sample classification.
Quantification and metabolic identification are generally not
employed. A screening tool to discriminate between samples of
different biological status or origin.
Sample preparation is simple and, as chromatographic separation is
absent, rapid analysis times are small (normally 1 min or less)
Metabolite target
analysis
Qualitative and quantitative analysis of one or a few metabolites
related to a specific metabolic reaction.
Extensive sample preparation and separation from other
metabolites is required and this approach is especially
employed when low limits of detection are required. Generally,
chromatographic separation is used followed by
sensitive MS or UV detection
Metabonomics
Evaluation of tissues and biological fluids for changes in
endogenous metabolite levels that result from
disease or therapeutic treatments
19
Metabolome technologies for data acquisition
The central to metabolomic processes are a variety of chemical profiling technologies. The
gross chemical composition of various biological fluids has been investigated by a variety of
chromatographic (separation methods) and spectroscopic techniques(detection methods),
notably gas and liquid chromatography (Fiehn et al.,2000b; Sauter et al.,1991;Roessner et
al.,2001), Mass spectrometry (Matsumoto and Kuhara, 1996; Wolfender and Hostettmann,
1996; Aharoni et al.,2002),NMR (Ward et al.,2002,Kikuchi et al.,2004)and Infrared
spectrophotometry (Jackson and Mantsch, 1996).Various technologies used are discussed
and their advantages and disadvantages (Shulaev,2006) are shown in Table 2 below:
Mass spectrometry (MS): Due to its high sensitivity and wide range of covered
metabolites, MS has become the technique of choice in many metabolomics studies (Halket
et al.,2005). MS can be used to analyse biological samples either directly via direct-injection
MS or following chromatographic or electrophoretic separation. Direct-injection MS,
especially when using a high-resolution mass spectrometer, provides a very rapid technique
to analyse large number of metabolites, and therefore is extensively used for metabolic
fingerprinting and metabolite profiling. Mass spectrometry has a serious drawback, known
as ionization suppression (which results from the presence of less volatile compounds that
can change the efficiency of droplet formation or droplet evaporation, which in turn affects
the amount of charged ion in the gas phase that ultimately reaches the detector) which might
diminish quantitative reproducibility. To avoid this problem and to decrease the complexity
of the sample, MS is often used as a hyphenated technique, i.e. it is coupled with gas
chromatography (GC), liquid chromatography (LC) or capillary electrophoresis (CE) where
20
the sample mixture is separated by chromatography or electrophoresis, and individual
compounds or less complex mixtures of compounds (which cannot be separated due to
similar properties) that are eluted from the column or capillary are analysed by MS.
Gas chromatography: is one of the most widely used and is a powerful method especially
when interfaced with electron impact (EI) quadruple or time-of flight (TOF) mass
spectrometry (GC-MS). GC-MS is viewed as the gold standard of metabolomic techniques
due to the high separation power of GC combined with better deconvolution. The advantage
of this coupled approach is that, it is possible to profile several hundred chemically diverse
compounds including organic acids, most amino acids, sugars, sugar alcohols, aromatic
amines and fatty acids (Roessner,2000).It offers very high chromatographic resolution, but
requires chemical derivatization (chemical conversion of sample compounds into their
derivatives which makes them suitable for separation and analysis) for many bio molecules
and hence some large and polar metabolites cannot be analysed by GC (Schauer et al.,
2005).
Liquid chromatography/mass spectrometry (LC-MS): LC-MS is being increasingly used
due to its high sensitivity and a range in analyte polarity and molecular mass wider than GC-
MS. Liquid chromatography, and specifically high-performance liquid chromatography
(HPLC), combines both high resolution and analytical flexibility. It can be tailored for the
analysis of a specific metabolite or class of compounds, or it can be used for the analysis of
a broad range of compound classes. LC-MS has one advantage over GC-MS, in that; there is
largely no need for chemical derivatization of metabolites (Shulaev,2006).
21
Capillary electrophoresis-mass spectrometry (CE-MS): CE-MS provides several
important advantages over other separation techniques. It has a very high resolving power;
very small sample is required for analysis and has short analysis time. CE has been used for
both targeted and non-targeted analysis of metabolites. One of the significant advantages of
the CE-MS is the ability to separate cations, anions and uncharged molecules in a single
analytical run, and therefore can be used for simultaneous profiling of many different
metabolite classes(Soga et al.,2002).This feature makes it a very attractive and promising
analytical technique for high-throughput non-targeted metabolomics.
Fourier transform infrared spectroscopy (FTIR): is useful in the field of large-scale
exhaustive profiling as an easy and high-throughput method (Gidman et al.,2003;Johnson et
al.,2003).It is used for the quantification of compounds that have specific functional groups
although it is not optimal for complicated metabolite profiling due to the lack of a separation
procedure. Fourier transformation ion cyclotron resonance mass spectrometry (FT-ICR-
MS), is the latest mass spectrometry technology. Infusion FT-ICR-MS analysis without pre-
separation by chromatography has been achieved for exhaustive metabolic profiling
(Aharoni, et al.,2002).
22
Table 1.2 Common analytical techniques used in metabolomics (Shulev 2006)
Analytical method Advantage Disadvantage
NMR Rapid analysis Low sensitivity
High resolution Convoluted spectra
No derivatization needed More than one peak per
component
Non-destructive Libraries of limited use due
to complex matrix
GC/MS Sensitive Slow
Robust Often requires
derivatization
Large linear range
Many analytes thermally-
unstable or too large for
analysis
Large commercial and public
libraries
LC/MS
No derivatization required
(usually) Slow
Many modes of separation
available
Limited commercial
libraries
Large sample capacity
CE/MS High separation power
Limited commercial
libraries
Small sample requirements
Poor retention time
reproducibility
Rapid analysis
Can analyse neutrals, anions
and cations in single run
No derivatization (usually)
FTIR Rapid analysis
Extremely convoluted
spectra
Complete fingerprint of sample
chemical composition
More than one peak per
component
Metabolite identification nearly
impossible No derivatization needed
Requires sample drying
23
Nuclear magnetic resonance (NMR): can uniquely identify and simultaneously quantify a
wide range of organic compounds in the micromolar range (Viant et al.,2003; LeGall et
al.,2003; Krishnan et al.,2005).In recent times, high-resolution NMR well-suited to
metabolomics studies has become a key technology for the elucidation of biosynthetic
pathways and metabolite flux via quantitative assessment of multiple isotopologues
(Eisenreich and Bacher,2007).NMR spectroscopy is the only detection technique which does
not rely on separation of the analytes, since the sample preparation for NMR is
straightforward and also is non-destructive the sample can thus be recovered for further
analyses(Bailey et al.,2003). All kinds of small molecule metabolites can be measured
simultaneously - in this sense, NMR is close to being a universal detector.
1HNMR is also a rapid technique that characterizes and quantifies a broad range of
metabolites in a non- targeted way, since in principle any chemical species that contains
protons gives rise to signals.NMR signals (i.e. resonance) are observed when a sample is
irradiated with pulses of radiofrequency electromagnetic radiation in a strong magnetic field.
Each nucleus within a molecule experiences a slightly different magnetic field because of its
distinct chemical environment and absorbs energy at a slightly different frequency. The
separation of these resonance frequencies from an arbitrarily chosen reference is called the
chemical shift (ppm). The result of an NMR analysis of a tissue is a spectrum, that is, a plot
of intensity (area) against chemical shift in which each signal occurs at a characteristic
energy. These signals are profiled as NMR spectrum comprises peaks which, for a given
solvent appear at characteristic frequencies. The major limitation of NMR for
comprehensive metabolite profiling is its relatively low sensitivity, making it inappropriate
for the analysis of large number of low-abundance metabolites (Sumner et al.,2003).Some of
the studies using NMR has shown to cover most of the ‘‘organic’’ compounds such as
24
carbohydrates, amino acids, organic and fatty acids, amines, esters, ethers and lipids, which
are present in plant tissues (Ward et al.,2002) and when coupled with statistical analysis
serves as an excellent screen to analyse metabolome in a non-destructive way. NMR has
also provided information on biosynthesis (Lutterbach and Stockigt,1995; Prabhu et
al.,1996;Fiehn and Weckwerth, 2002),on metabolism (Ratcliffe and Shachar-Hilt, 2001),
and on the effects of herbicides on metabolism (Lutterbach and Stockigt, 1994 and1995) and
mode-of-action (Hole et al.,2000, Hadfield et al.,2001), or used in investigations of whole
plants (Schneider,1997;Pope et al.,1993).Although NMR and MS are most often used for
large-scale analysis, metabolomics is not limited to these techniques. Other alternatives
include thin-layer chromatography(Tweeddale et al.,1998 and1999).HPLC with UV/visible
absorbance, photodiode array (PDA) (Fraser et al., 2000) or electrochemical detectors
(Gamache et al., 2004,Kristal et al.,2002) FT-IR (Gidman et al.,2003), phenotype
microarrays (Bochner et al.,2001) and variety of other enzymatic assays. Combined use of
multiple techniques or multiple detectors in online or parallel analysis can significantly
increase the metabolite coverage, increase quantification limits and improve identification of
metabolites from a single biological sample.
25
Bioinformatics
Data Analysis
The major challenge of metabolomics studies is data processing and data interpretation
(Shulaev,2006). Database management systems for metabolomics are required to collect
both metadata (metadata is the data about the experiment), raw and processed experimental
data. Storing metadata, covering experimental design, the nature of the samples and their
treatment prior to the analysis and information about the analytical technique and data-
processing details are important to be able to reproduce the experimental conditions and
compare results obtained in different laboratories(Bino etal.,2004).All the metabolomic
approaches heavily relies upon bioinformatics for the storage, retrieval, and analysis of
larger datasets and many efforts have been made on software development (Katajamaa and
Oresic, 2007).There are many tools used to align, visualize and differentiate components in
datasets and the individual components are then correlated and placed in metabolic networks
or pathways. Changes in metabolite levels may be drastic or subtle and those subtle changes
will require statistical processing to determine whether the observed changes are significant
or not (Miller and Miller, 2000).
For NMR-based metabolomics approach, after NMR experimental data collection, post
experimental data handling including NMR spectral processing, data pre-processing and
data analysis, is critical for obtaining good results. To assist these procedures, several
commercial software tools have been released, such as: AMIX (Bruker Biospin, Germany),
KnowItALL (BIO-RAD, USA),Chenomx NMR Suite (Chenomx, Canada) and Hi-Res
(Zhao,2006). NMRPipe is a widely used traditional NMR data processing software tool
(Delaglio et al.,1995).These software tools provide plenty of functions involving spectral
26
processing, comprehensive identification and quantification of metabolites. However, due to
the complexity of NMR data and different application purposes, further data analysis
procedures, such as filtering out unwanted variations (e.g. background noise, uncorrelated
variation in data model, etc.) in a dataset, generating and applying predictive classification
or regression models, are usually required. To complete these tasks, researchers usually
invoke other advanced chemometrics tools or statistical tools. Commercial software
packages such as MATLAB (Mathworks, USA), SIMCA-P (Umetrics, Sweden) and SPSS
(SPSS, USA) are frequently used by researchers.
NMR spectral processing
Spectral data often include some disturbance according to each specificity and
environmental perturbation. Such disturbance needs to be corrected for by appropriate data
pre-processing. Several tactics, such as spectral difference, noise reduction, baseline
correction, normalization, mean centre, and scaling are often performed in common pre-
processing as for the metabolomics studies, spectral processing may have significant impact
on data analysis (Defernez and Colquhoun, 2003).The present study uses a commercial
software MestRe-C (Magnetic Resonance Companion) Cobas and Sardina, 2003) designed
specifically for metabolomics, will be evaluated.
Statistical Analysis
Statistical analyses must be performed to ensure good analytical rigor, but unfortunately are
often a burden when working with large datasets. Tools are therefore used for high
throughput statistical analyses of all components in a dataset to provide sound evidence for
the relevance of changes in a metabolite level contained in the raw data. Computer based
applications are used that can differentiate whether or not samples are statistically similar or
different and what the exact differences/similarities are. A single metabolite profile can yield
27
hundreds of distinct components. This wealth of information that needs to be interpreted,
leads to significant challenges in processing the data. To simplify the task, many researchers
have used techniques to reduce the dimensionality of the data set and to visualize the data
(Sumner et al.,2003)
Principle Component Analysis (PCA) is the choice for metabolomics for many researchers
as the data is complex (Le Gall,2001;Ward et al.,2002;Defernez and Colquhoun, 2003).PCA
provides a way to summarize the information contained in large sets of spectra. It transforms
the initial variables (or ‘measurements’) into a much smaller set of variables or PC scores.
The main purpose of PCA is to eliminate the collinear problem and then reduce the
dimensionality of the original feature (variable) space. It is an unsupervised method used to
reveal the internal structure of datasets in an unbiased way. PLS is a supervised method for
regression.PCA and PLS reveal the variable contribution for the separation between
different groups by loadings or regression coefficients (Fukusaki and Kobayashi
2005).Machine learning methods such as self-organising maps (SOM) and artificial neural
networks (ANN) are also promising in research (Fukusaki and Kobayashi, 2005).A variety
of approaches have been applied for the statistical analysis of spectral data. One approach is
to gather information (e.g. integrate individual peak areas) for as many constituents as
possible and study the occurrence of each of these compounds separately, for example by
calculating coefficients of variation or carrying out an Analysis of variance (ANOVA) (Le
Gall, 2001;Smilde et al.,2005).ANOVA estimates the factor effects and tests significance by
splitting the variations in orthogonal and independent parts (Searle, 1971).Another data
analysis approach, Hierarchical cluster analysis (HCA), is a method of cluster analysis based
on the multivariate distance between every pair of data points. In HCA, the data are not
partitioned into a particular cluster in a single step. Instead, a series of partitions takes place,
28
which may run from a single cluster containing all objects to n clusters each, containing a
single object (Mochida et al.,2009).For the present study the statistical analysis was carried
out and visualized in an application tool called Multi Experiment Viewer (MeV) (Saeed et
al.,2006).
In plant metabolomics bioinformatics tools provide methods of determining differences or
similarities in datasets. The metabolomic data sets must then be integrated and correlated in
a global manner with genetic and enzymatic data, pathways assembled into systems, and
literature references incorporated as learning tools to annotate existing data to yield in silico
biological information (Palsson,2000).
Application of Plant metabolomics
Although relatively young, and still very much in development, plant metabolomics is now
being widely applied and is already considered as a technology, which is a ‘maturing
science’ and is ‘established and robust’(Dixon et al.,2006;Schauer and Fernie 2006).Some
of the general applications of plant metabolomics are studied in plant tissues using NMR
(Kim, et al.,2010;Kruger et al.,2008) GC-MS (Lisec et al.,2006), LC-MS (De Vos et
al.,2007). Hall et al.,(2008), has elaborated the use of plant metabolomics and its potential
application for human nutrition. Metabolite profiling has been performed on a diverse array
of plant species, including tomato (Schauer et al.,2005;Tunali et al.,2003), potato (Roessner
et al.,2000, Roessner et al.,2001),rice (Sato et al.,2004),wheat(Hamzehzarghani et
al.,2005),strawberry(Aharoni, et al.,2002), medicago(Chen et al.,2003),cucumber(Tagashira
et al.,2005),lettuce(Garratt et al.,2005)tobacco(Blount et al.,2002) poplar(Robinson et
al.,2005,Brosche et al.,2005) and eucalyptus (Merchant et al.,2006). Metabolite profiling
has also been carried out widely in Arabidopsis thaliana, in a tissue specific level (Schad et
29
al.,2005), in transgenic plants (Le Gall et al.,2005) and also for comparison of ecotypes of
Arabidopsis thaliana (Ward et al.,2002).Hirai et al., (2004) linked the genomic data and the
function of metabolite under the deficiency of sulphur and nitrogen in Arabidopsis thaliana.
Metabolomics has been useful to ascertain, for example, the metabolic response to herbicide
(Ott et al., 2003), the equivalence of genetically modified (GM) and conventional crops
(Catchpole et al.,2005;Defernez et al.,2004;Barros et al.,2010;Matsuda et al.,2010) and the
classification of plant genotypes (Roessner et al.,2000;Roessner et al.,2001, Fiehn et
al.,2000a,Tikunov, et al.,2005,Kim, et al.,2010),Quality control of medicinal plants
viz.,Gingko(Agnolet et al.,2010), St. John’s wort (Roos et al.,2004),Rasmussen et al.,2006),
Ginseng (Yang et al.,2007), and to find activity-related compounds in medicinal plants
(Cardoso-Taketa et al.,2008,Modarai et al.,2010) and to study the plant interaction with
other organisms(Jahangir et al.,2008, Abdel-Farid et al.,2009,Leiss et al.,2009).Also there
have been significant recent successes in gaining new insights into the pathways of lignin
biosynthesis (Anterola and Lewis,2002)and isoflavone synthesis (Jung et al.,2000;Liu et
al.,2002) and progression in metabolite-profiling methods for isoprenoids (Lange et
al.,2001), oxylipins (Weichert et al.,2002), taxane diterpenoids(Ketchum et al.,2003),
alkaloids(Yamazaki,2003), flavonoid glycosides (Le Gall et al.,2003) and volatile
compounds (Verdonk et al.,2003;Flamini et al.,2003;Brown,2002).
More recently, metabolomics has been much used in descriptions of the response of plants to
a wide range of biotic or abiotic stresses, to decipher gene function, to investigate metabolic
regulation and (in combination with analysis of other molecular entities of the cell) as part of
integrative analyses of the systemic response to environmental or genetic perturbations
(Schauer and Fernie,2006).
30
Limitations of plant metabolomics
There are several critical factors in metabolomics studies, since the metabolome is sensitive
to perturbation. Plant cultivation, sampling, extraction and derivatization must all be
carefully standardised and strategies need to be incorporated to minimize variations
(Fukusaki and Kobayashi, 2005).It is generally accepted that a single analytical technique
will not provide sufficient visualization of the metabolome and therefore, multiple
technologies are needed for a comprehensive view and sometimes a combination of
techniques are to used (Le Gall et al.,2005).The selection of the most suitable technology is
generally a compromise between speed, selectivity and sensitivity (Sumner et
al.,2003).Metabolic profiling is complex and it involves various techniques to cover a wide
range of metabolites. Each metabolite should be considered according to its characteristics
in the following categories: hydrophilic, hydrophobic, small molecule, large molecule,
charged, uncharged and combinations of these (Schauer and Fernie, 2006). Metabolites vary
due to volatility, polarity, solubility and chromatographic behaviour mean that multiple
methods will need to be deployed to analyse different subsets of metabolites (Ward et al.,
2002).A major technological challenge encountered in metabolomics is dynamic range.
Dynamic range defines the concentration boundaries of an analytical determination over
which the instrumental response as a function of analyte concentration is linear. The
dynamic range of many techniques can be severely limited by the sample matrix or the
presence of interfering and competing compounds. In other words, the presence of some
excessive metabolites can cause significant or severe chemical interferences that limit the
range in which other metabolites may be successfully profiled .For example, high levels of
primary metabolites such as sugars often interfere with the ability to profile secondary
31
metabolites such as flavonoids. This is one of the most difficult issues to address in
metabolomics (Sumner et al.,2003).There is another drawback where metabolomics is a
snapshot and thus is not particularly suited to studying the rapid kinetics of metabolic
processes and requires measurement of fluxes (a series of snapshots), which represents a
major challenge (Kim et al.,2011).A truly comprehensive analysis of the metabolome is
currently not feasible and the number of the primary and secondary metabolites in any given
plant species is still uncertain. This is the current inability and limitation of metabolomics.
Metabolomics is not a goal in itself; it is a tool for improving our understanding of the
metabolism and biochemistry of organisms. To provide a deeper biological meaning,
metabolomic studies should deal with identified metabolites instead of unknown signals: the
number of signals is less important than the number of identified metabolites. To facilitate
identification, an important issue and a great challenge will be to build a common public
database to share information within scientific communities (Kim et al.,2011)
Present study
The recent progress about the pathways, control of ascorbate metabolism and the potential
roles of intracellular and long distance transport as factors determining ascorbate
concentration has helped in understanding plant ascorbate to a certain extent. There are also
certain areas of ascorbate metabolism which require a deeper understanding. Much remains
to be learned about the contribution of alternative ascorbate biosynthetic pathways in each
plant tissue and the regulation of metabolic fluxes, and many important genes have not yet
been identified. There are many important questions that still needs to be explored about
ascorbate metabolism and the ascorbate-deficient mutants provide an opportunity to
32
investigate the following questions such as-What is the effect of each specific mutation on
metabolism and can the function of VTC3 identified? Ascorbate has proposed roles as an
antioxidant and as a cofactor for dioxygenase enzymes that are involved in synthesis of a
variety of plant hormones and secondary compounds (e.g. phenolic compounds involved in
defence against pathogens).Therefore can the effect of ascorbate deficiency on plant
function, as visualised by the metabolome, provide information on the function(s) of
ascorbate? Ascorbate mutants are more sensitive to environmental stresses and, if so, can
this be detected by changes in their metabolome compared to wild-type plants. Profiling of
all the metabolites contained in the ascorbate deficient mutants will offer much useful
information about metabolism and manipulation. The main objectives of the present study
are: 1.to use NMR as an analytical tool for profiling the vtc mutants and Wild type plants to
produce NMR data 2. to use software packages MestRe-C to perform spectral processing to
study the ascorbate NMR data 3.to look for differences in spectral processed data at the
metabolite levels across the plant types (Wild type/ mutants and alleles), growth stages
(rosette and flowering) 4.to verify the discriminating bins from the statistical analysis with
the original spectra for the presence of peaks.
33
Chapter 2
MATERIALS AND METHODS
Arabidopsis thaliana ecotype Columbia-0 seeds (Wild type) and ascorbate-deficient mutants
namely vtc1, vtc2, vtc3 and vtc4 were chosen for the present study (Table 2.1). All the
mutants have the same genetic background.
Table 2.1 Plant types used for the study
The seeds were grown, seedlings transplanted to soil, harvested, prepared and extracted for
NMR analysis in The National Centre for Plant and Microbial Metabolomics, at Rothamsted
Research. U.K. The protocols followed for each of the above stages are given below:
Germination of Arabidopsis thaliana seeds on agarose:
Before starting the procedure Laminar Flow Hood was turned on and the working surface
were wiped with 70% Ethanol.The hood was left running in LAF purge mode for 10 minutes
and then turned to red to normalise the flow for working.
Step 1: Preparing 1 litre of Arabidopsis growth media: 1Litre Duran bottle was labelled
as M+S + 3% Sucrose, and with the date and the name of the operator and 4.4g of
Murashige-Skoog media 30g sucrose (3% final concentration), 7g Agarose Type PGP,
supplied by Park Scientific Ltd (0.7% final concentration) was weighed into a 1L Duran
Wild type plant
Ascorbate-deficient Mutants
Wt
vtc1
vtc2
vtc3
vtc4
34
bottle.1 litre of deionised polished water was added to the bottle. The growth media was
mixed well in the solution by inverting the bottle three times. The solution was adjusted to
pH 5.6 with 1M KOH. The lid of the Duran bottle was slightly loosened and a strip of
autoclave tape was place over the top of the lid. The bottle was autoclaved using the 2100
Classic Autoclave supplied by Jencons-PLS. (Autoclaving takes approximately 30 minutes
but the programme was automatic and timing depended on whether the machine was starting
from cold, or has recently been used).Once the autoclave programme finished, the Duran
bottle was removed, the lid was tightened and placed in the Laminar Flow Hood to cool.
(The following steps 2, 3 and 4 were carried out in a Laminar Flow Cabinet)
Step 2: Pouring Agarose plates: Using a black marker pen the outside of the bottom half of
sterile petri dishes were labelled with details: M+S+3% Sucrose, name of the seed to be
plated, date and name of the operator. Once the agarose was cooled enough to handle
(usually 30-60 minutes after removal from the autoclave) it was poured directly from the
bottle into the labelled triple vent 90mm petri dishes until agarose 2/3rds fill the bottom of
the dishes. The petri dishes were pushed to the back of the flow hood and the lids of the petri
dishes were left slightly open (offset by approximately 5mm), with the opened lid facing the
filter at the back of the flow hood, to prevent condensation while minimising contamination.
The plates were allowed to stand for one hour. Before carrying out sterilisation of seeds,
10% bleach solution was made in a 100ml Duran bottle using Parazone Jeyes Active Power
Thick Bleach and deionised water (v/v). All the equipment needed viz., seeds, eppendorf
tubes, bleach solution, sterile water, pipettes (P1000 and P20, pipette tips (1ml and 10-100µl
Barky Ultipette Capillary tips), parafilm and scissors - were transferred to the Laminar Flow
Hood.
35
Step 3: Sterilising Arabidopsis seeds: The seeds to be plated were transferred to an
eppendorf tube labelled with the seed name.1ml of 10% bleach (v/v in water) was added to
sterilise the seeds. Each eppendorf tube was inverted five times in 3 minutes, to ensure all
the seeds are exposed to the bleach. The seeds in bleach were left not for longer than 3
minutes as it would cause seed coat breakdown. The bleach solution was pipetted off and
1ml of sterile deionised water was added to remove residual bleach from the seeds. The tube
was inverted five times to ensure thorough washing of the seeds. The water was removed
with a pipette and repeated the sterile water wash. The majority of the water was removed,
leaving a small volume in the base of the tube to facilitate plating of seeds.
Step 4: Plating out of Arabidopsis seeds: The lid of the petri dish was lifted, of the agarose
plate to be plated, and angled towards the back of the flow hood, to help minimise
contamination. Using capillary tips and a P20 Gilson Pipette, a 20µl aliquot of seeds was
taken from the appropriate eppendorf tube. The pipette tip was inserted with the seeds under
the lid to distribute the seeds individually at regular intervals over the plate. A maximum of
50 seeds were plate per plate. The plates were sealed by wrapping a 1cm strip of parafilm
around the petri dishes where the top and bottom of the dishes meet. The petri dishes were
place flat, in a cabinet where ensuring the conditions were set at 24 hours light and
22°C.The plates were left in the tissue culture cabinet 3 for 10 days or until the seedlings
have 2-4 leaves.
Step 5: Recording Information: The seeds plated were listed on a record of Arabidopsis
thaliana plant growth sheet. The sheet was completed with the following sections for
referencing: description of the ecotype, seed name and number for identification and date
plated.
36
Transferring Arabidopsis thaliana seedlings to soil
The seedlings were transferred 10 days after plating to the soil (75% medium grade peat,
12% screened sterilised loam, 3% medium grade vermiculite, 10% grit (5mm screened, lime
free), 3.5kg osmocote plus ¾ month per m3, 0.5kg PG mix per m3, Lime to pH 5.5-6.0 and
wetting agent.).The plants are usually at the 2-4 leaf stage. On day of transferral the soil
trays were made up according to number required. The soil is watered from above (from a
height of approximately 50cm) using a hose with a rose attached, so as not to cause breakup
of the soil surface. Once the surface looks wet some water was added to the bottom of the
trays. Each tray was labelled with: the seed number, date of plating, and date of soil transfer,
tray number and the name of the operator. The petri dishes containing seedlings were
collected from cabinet where they were kept for germination. The seedlings were carefully
picked off of the agarose plates with a sterilised cocktail stick and laid on the soil surface.
Using the cocktail stick they were then gently pushed so that all the roots under the surface
of the soil, leaving the seedling’s leaves resting on top of the soil. The trays were covered
with an incubation lid, ensuring all vents are closed; the trays then were transferred to
controlled environment growth room with growth conditions specific for Arabidopsis
growth viz., Long day conditions: 16 hour day(Day- Light: 00:00-16:00 hrs, Humidity 75
%,Temperature 23ºC; Night16:00-00:00 hrs, 80 %,18ºC).
The incubation lids were left on for 2-3 days, until seedlings have grown and look healthy,
and removed when the plants were grown the required developmental stage for harvest. The
date of transferral of seeds to soil was recorded on the relevant record of Arabidopsis
thaliana growth sheet.
37
Harvesting of Arabidopsis thaliana plants for analysis
50ml centrifuge tubes were labelled with sample number, date of harvest and the operator's
name. Separate tubes were used for different plant types and different trays of the same plant
type. A minimum of three holes were made in the centrifuge tube caps with a pair of
scissors. The plant samples were harvested at two growth stages namely rosette (5.10) and
flowering (6.1-6.5) (Boyes et al.,2001).Harvesting of the plants were started at 14 hours into
the growth day (2pm). A labelled 50ml centrifuge tube was dipped into a flask containing
liquid. The plants were cut, using scissors, immediately below the rosette, as close to the soil
surface as possible. The harvested plant was then put immediately using tweezers, into the
centrifuge tube containing liquid nitrogen, so that it was frozen instantly, thus minimising
any alteration of metabolites after harvest. Three to six plants were put in each tube. The cap
of the tube was replaced, ensuring it has holes in it and immerse it in the flask of liquid
nitrogen to ensure samples are kept frozen. The numbers of plants harvested were recorded
for reference. The tubes of plant material was then transferred to a plastic bag labelled with
date and operator name and stored at –80°C until preparation for analysis can be carried out.
Preparing Arabidopsis thaliana plants for NMR analysis
A pestle and mortar was chilled in a –80°C freezer for 30 minutes. A polystyrene box was
filled with ice, a well was made in the ice and the mortar was placed in the well. Little
Liquid Nitrogen was poured into the mortar and allowed to evaporate. The plant material
from the –80°C freezer was removed and the plant material of the same ecotype and tray
number were batched by emptying all centrifuge tubes containing the specified material into
the pre chilled mortar. The liquid nitrogen was poured into the mortar and the pestle was
used to grind the plant material. Grinding was continued until the material was fine and
38
homogenous. Any excess liquid nitrogen was allowed to evaporate off. A glass vial was
taken, respectively labelled and the lid of vial was removed and the ground material was
transferred to the vial, with a spatula. The material was retained at –80°C until further
analysis.
Analysing Arabidopsis thaliana plants by NMR
Step 1: Plant Extraction: A solution of 0.05% TSP w/v 80:20 D2O:CD3OD (pH typically
around 6.5) was made up. Three 1.5ml eppendorf tubes were labelled with the name of the
plant material and the replications to be analysed. The material stored at -80°C was removed
from storage and allowed stand on the laboratory bench for one hour, to allow the plant
material to return to room temperature.15mg ± 0.03mg of plant material were weighed out,
into each labelled eppendorf tube. The water bath was turn set to 50°C and 90°C and
allowed to reach the required temperature. The extraction method was a hot ethanol
extraction and hence two temperatures were used for effective extraction of samples. To
each eppendorf tube of plant material 1ml of the solvent, 0.05% TSP w/v 80:20
D2O:CD3OD, made in step 1.1 was added. Each tube was mixed, using a bench top whirl
mix, until all the plant material was held within the solvent. The samples were then heated at
50°C in a water bath for 10 minutes. A second set of eppendorf tubes were labelled with the
same descriptions as the first set in the water bath. The samples from the water bath were
removed and centrifuged for 5 min in a bench top centrifuge. The supernatant were then
transferred, using a 1ml Gilson pipette and a clean tip for each sample, to the clean,
appropriately labelled eppendorf tube. The supernatant was heated at 90°C in a water bath
for 2 minutes and the samples were removed from the water bath and transferred to a
39
refrigerator and store for 30 minutes. Clean NMR tubes were labelled with the same
descriptions as those on the eppendorf tubes. The samples were then removed from the
refrigerator and centrifuged for 5 minutes, in a bench top centrifuge. 750µl of the
supernatant was transfer, using a 1ml Gilson Pipette and a clean tip for each sample, from
each eppendorf tube to the appropriately labelled NMR tube and cap the tube. To get good
quality NMR spectra, there should be enough solution in the NMR tube to span the “active
volume” region. For a 5mm NMR tube the active volume was normally in excess of
500microlitres.750 µl were used because that was approximately the amount of clean
supernatant that can reproducibly collect from the extraction protocol.
Step 2: NMR Analysis: The samples were loaded on to the NMR instrument which was run
to the program set, selecting the appropriate solvent, (D2O) and experiment (Watersup.lars
2K) and giving each sample a title. The data was collected by NMR acquisition Software
and then was used for further analysis.The parameters set for data collection is tabulated as
shown in Table 4.
40
Table 2.2 Parameters for NMR Data Collection
Instrument Parameters
Instrument Bruker Biospin, Advance 400
Probe 5mm broad band multinuclear probe (BBO Z8248/0048)
Proton Frequency 400.1299232 MHz
Sample Introduction System BACS
Automation Control Software ICON-NMR
NMR acquisition Software X win-NMR 9 (v3.5)
NMR tube size 5mm
Acquisition Parameters
Parameter Set Watersup.roth
Pulse Sequence zgpr (low power pre-saturation pulse followed by high power 90˚ pulse)
Presaturation pulse 50.0 dB for 5 s
90˚ pulse -6 dB for 7.9 µs
Sweep width 12.1084 ppm
Time Domain size 32k data points
Temperature 300 K
Irradiation Frequency/
transmitter offset
4.7ppm
Number of Scans 2048
Lock Automated locking on D2O
Sample Shimming tune xyz (all under automation)
Sample Spinning 20 Hz
Receiver Gain Optimized for each sample by the automation program au_zg (using
command RGA)
other details
sample temperature 27ºC
T1 values not measured for the sample
Time Delay 5s( combines with 3.4s acquisition is enough for adequate relaxation)
spectral processing 0.5Hz line broadening
41
NMR Spectral processing with MestRe-C
A commercial processing software, MestRe-C, which is particularly designed to process
NMR data is used (Figure 3).MestRe-C can display and manipulate single or multiple 1D
spectrum and allows for detailed analysis and interpretation of NMR spectra. It is able to
handle data formats from a wide variety of instrument vendors and it automatically
recognises and converts data to the MestRe-C format when opened.
The data file taken from the spectrometer called the raw free induction decay (fid) were
imported in the MestRe-C and all the fids for all the samples involved in the present study
and their respective biological replications and technical replications were given a new
appropriate name for identification(e.g. WTR1 refers to wild type, rosette, biological
replication number 1.)Full Fourier Transformation (FT) was selected to get the proton
spectrum onto the interface and to convert NMR signals from time domain to frequency
domain which was done automatically with the help of option button on the software. Phase
correction was done next on to the spectra on the screen to get the phase corrected by
Automatic phase correction (Global method). Baseline correction was performed next on the
phase corrected spectra. Two baseline corrections are used with a) multipoint baseline
correction with linear picking and b) Auto baseline correction, to smooth the outliers to
obtain a smooth baseline. Referencing was done to the spectra with Trimethylsilane (TMS)
which was used as internal standard and was referenced to a tsp value of zero by expanding
the spectrum and dragging the TMS icon manually and the spectra was calibrated
(referenced). Bucket/binning, a commonly used technique for digitizing a spectrum into a
row vector was done with a bucket width of 0.01 ppm and 0.04 ppm was applied, the spectra
42
was split into 1000 and 250 integrated buckets respectively and the resultant data (as
absolute normalized values).The generated ASCII files was then imported to Microsoft
EXCEL for the addition of labels saved for further analysis. All the samples were subjected
to similar data processing procedures and all the resultant files are copied to a single
Microsoft EXCEL spread sheet and all the three technical replications for each single
sample are averaged up to give rise to biological replications. The processed data was also
saved as text files for statistical analysis.
Figure 2.1 Spectral processing with MestRe-C. (A) Raw NMR data exported into MestRe-C as
FIDs;(B) Fourier transformation;(C,D) Phase correction before and after;(E)Auto-baseline correction
(F,G) Baseline correction by linear correction before and after;(H)Calibration/Referencing with TMS
shift value 0;(I) Bucketing Integration and integrals imported as ASII files.
Data Analysis
The pre-processed from MestRe-C are then further altered by removing water
regions(bins)450-560 from 0.01ppm bucketed data and regions(bins)116-154 from 0.04 ppm
bucketed spectra to eliminate any variability in suppression of the water signal.
43
The statistical data analyses and visualisation were performed using MeV. Two types of
statistical analysis were done to the data; Analysis of Variance (ANOVA) and Hierarchical
clustering (HCA).The statistical analysis ANOVA in MeV was carried out by uploading the
data file and the numbers of groups was assigned followed by the P-value parameters and
Bonferroni correction which gives a stringent condition. No correction gives a least stringent
condition. The results are displayed in an open window which gives the values for the
significant bins and non-significant bins. The significant bins after from ANOVA were
chosen for performing clustering analysis of the data.HCA was chosen with Pearson's
correlation with average linking clustering as linkage method. The results are clusters which
were visualized as Gene Tree and heat maps. The schema of data processing and data
analysis is given in Figure 2.2.
Figure 2.2 Schema of the data analysis. Wild type and four vtc mutants have two stages (rosette
and flowering), two baseline correction (auto, linear), two band width (0.01 and 0.04 ppm), two
normalizations with control (Total area and HPV (HPV=High Peak value) and statistically analyzed
(HCA=Hierarchical clustering analysis, ANOVA=analysis of variation)
44
Chapter 3
RESULTS and DISCUSSION
NMR SPECTRAL FEATURES OF ASCORBATE- DEFICIENT MUTANTS
Four ascorbate-deficient mutants (vtc1, vtc2, vtc3 and vtc4) were employed in the current
study. All mutants contained different levels of ascorbate as described in the Chapter 1 as
these mutants have genetic defect in ascorbate biosynthesis pathways (Conklin, 1999;
Conklin, 2000; Dowdle et al, 2007).These mutants are phenotypically different on the basis
of their size, as previously found ascorbate is an important regulator of growth (Kercheve et
al, 2011).The aim of the current study is to investigate the effect of ascorbate deficiency on
metabolic changes analysed by NMR.NMR was used as an analytical tool to analyse the
differences at metabolite level in two growth stages such as rosette and flowering in WT and
vtc mutants. There are different techniques used for plant metabolic profiling as described in
Chapter 1, but NMR spectroscopy is a very powerful technique which gives the structural
elucidation of metabolites which are useful for identification. For the past many years, NMR
has been found very useful technique because of its unique advantages viz., non-destructive,
non-biased, easily quantifiable, requires no or little separation and needs no derivatization
(Wisahart, 2008).NMR has played a central role in understanding the metabolic changes in
different strains. The limitation of NMR is that it is unable to detect the low abundant
metabolites and also provide no information about retention time (RT) and mass spectral.
45
Figure 3.1 represents the one dimensional H-NMR spectra from Arabidopsis thaliana. A
total of 45 NMR spectra representing Wild type and four mutants of two different growth
stages (rosette and flowering) were collected and spectral processed in a commercial
software MestRe-C for exploratory analysis. The spectrum of both rosette and flowering
shows a dominance of signals in the carbohydrate region and well defined signals both in
aromatic and aliphatic regions (Fig 3.2a and 3.2b).
Figure 3.1 1H-NMR spectra of the Arabidopsis thaliana (Wild type) extracted with in
D2O:CD3OD. The bottom spectrum represents the original spectra and the spectra in the box shows
the extended region of the original spectra.
46
Figure.3.2a 1H-NMR spectra of Arabidopsis thaliana at rosette stage. Spectra here represent
different vtc mutants. A: Wild type B: vtc1, C: vtc2, D: vtc3, E: vtc4.
Figure.3.2b 1H-NMR spectra of Arabidopsis thaliana strains at flowering stage. Insert represents
different vtc mutants. A: Wild type B: vtc1, C: vtc2, D: vtc3, E: vtc4.
47
NMR DATA PRE-PROCESSING
Pre-processing of data is a major challenge for metabolomics study as spectral processing
influences the outcome for the data (Schleif, 2007). The strains in the study were subjected
to different steps normally involved in data pre-processing i.e., Fourier transformation,
phase correction, baseline correction, peak alignment (referencing to internal standard) and
bucketing (binning). Many of these steps are necessary, because the crude NMR data show
artefacts due to physicochemical differences (Torgrip et al.,2006).(1) Two baseline
correction in MestRe-C namely the auto baseline and linear baseline corrections were used
in the present study.(2) Binning is another crucial step in data pre-processing, as it is an
approach to solve the unmanageable dimensionality of the data and the inter individual
differences in peak locations caused by slight variations in the sample environment (Meyer,
2008).Two different bin widths (0.04 ppm and 0.01ppm) were applied to divide the spectral
regions into equal sized bins. The intensity values of each bin were integrated and annotated
to the bin. The bins were then compared across the different spectra.(3)All the bin intensity
were then normalised. Normalization is a row operation that is applied to the processed data
from each sample and comprises methods to make the data from all samples directly
comparable (Craig et al., 2006). (a) The binned data were normalized with total area where
the bin values were converted as a fraction of the total spectrum area. (b) They were also
normalized using highest peak value where the integrals were individually divided by the
highest peak value (here the reference peak).
All the above mentioned parameters were applied in the present study while processing the
NMR spectra and also after obtaining the processed data to see if there are any influences of
these different parameters on the outcome of the data and to choose the best combination for
further analysis.
48
MULTIVARIATE ANALYSIS OF ASCORBATE- DEFICIENT MUTANTS
Many approaches have been used for the analyses of NMR data. The most common methods
used for the analysis of metabolites data are the multivariate data analysis such as principle
component analysis (PCA) and the hierarchical clustering (HCA) clustering(Deferenez and
Colquhoun, 2003). In the current study, the analysis of variance (ANOVA) was performed
to analyse the variability of the data between different strains. Each strain contained three
biological replicates and three technical replicates. The ANOVA was calculated on all the
strains against WT (control) in flowering and rosette separately. Bonferroni correction was
used, which kept the false positives to a minimum as hundreds of bins (integrated buckets)
were being analysed. The p < 0.05 was selected as the significant values. The significant
data (p < 0.05) from both stages of growth used in the current study was combined and
subjected to HCA.
HCA clusters calculated the differences between biological replicates and strains. It also
clustered features together which are behaving in a similar way on the basis of their
chemical nature. HCA was performed using Pearson's correlation and average linkages, the
most commonly used methods for the metabolomic and genomic analysis. The resulted
dendrogram (rosette and flowering), which separates the data into branches and showed the
variability between strains and the biological replicates. The horizontal (x-axis) represents
the biological replicates of each strain and two growth stages while vertical (y-axis) shows
the features (metabolites). The features which are similar in nature were clustered together
as shown in Figure 3.3. The dendrogram showed all the biological replicates from rosette
were clustered together but unfortunately, the biological replicates from flowering were far
away from each other and not clustered properly (Figure 3.3).
49
This may be due to variability in biological replicates from flowering which needs further
investigation. Interestingly, we observed a prominent difference in metabolites intensity
between WT and the vtc mutants. The differences were also observed in the flowering and
the rosette in the accumulation of metabolites. The results indicated that spectra produced by
NMR for metabolites intensity are able to primarily show difference between vegetative and
flowering stages.
Sixty four combinations from two different growth stages were used, including two different
baseline corrections, two bin widths, two different normalisations methods along with the
non-normalised. The data were obtained after performing data pre-processing and post
processing using different parameters as described earlier are presented in Tables 3.1-3.4.
As described before we observed variation between biological replicates in flowering
compared to rosette. We are unable to investigate the variability between biological replicate
in flowering in the current study which needs further work. So, for the further analysis here,
we selected the rosette data which showed very close behaviour in biological replication.
50
Figure 3.3 HCA showing the difference in growth stages (rosette and flowering) and WT and
the vtc mutants. Boxes and ellipses show the clustering of biological replicates in rosette stage the
flowering stage respectively. M=mutants, R=rosette, F=flowering, WT= wild type. The scale shows
intensity of colours from highest to lowest. Red = Highest intensity, Green = Lowest intensity. The
significant difference was calculated by ANOVA (p < 0.05).
51
Table 3.1 Clustering analyses of all plant types in rosette stage with 0.01 bin width and
different processing combinations. Cluster No = number of clusters where all the three biological
replicates within and between strain clustered together, Strains = represents the strains in which are
all the biological replicates are clustered close to each other.
STAGE
BIN
WIDTH
(PPM)
BASELINE
CORRECTION NORMALIZATION
CLUSTERING
ANALYSIS
CLUSTERS
(No) STRAINS
ROSETTE
0.01
linear
non-normalized
HCA 1 wt
HCA/ ANOVA
/no correction
2 wt,vtc1
HCA/ANOVA
/Bonferroni 3 wt, vtc 1, vtc 2
HPV
HCA
1 wt
HCA/ ANOVA
/no correction
4 wt, vtc 1, vtc 2, vtc
3
HCA/ANOVA
/Bonferroni 5
w, vtc 1, vtc 2, vtc
3, vtc 4
Total Area
HCA 2 wt, vtc 1
HCA/ ANOVA
/no correction
2 wt, vtc 1
HCA/ANOVA
/Bonferroni 5
wt, vtc 1, vtc 2, vtc
3, vtc 4
auto
non-normalized
HCA 2 wt, vtc 1
HCA/ ANOVA
/no correction
4 wt, vtc 1, vtc 2, vtc
3,
HCA/ANOVA
/Bonferroni 3 wt, vtc 1, vtc 4
HPV
HCA 2 wt, vtc 1
HCA/ ANOVA
/no correction
4 wt, vtc 1, vtc 2,
vtc 3,
HCA/ANOVA
/Bonferroni 2 wt, vtc 1
Total Area
HCA 2 wt, vtc 1
HCA/ ANOVA
/no correction
2 wt, vtc 1
HCA/ANOVA
/Bonferroni
5 wt, vtc 1,v2, vtc 3,
vtc 4
52
Table 3.2 Clustering analyses of all plant types in rosette stage with 0.04 bin width and
different processing combinations. Cluster No = number of clusters where all the three biological
replicates within and between strain clustered together, Strains = represents the strains in which are
all the biological replicates are clustered close to each other
STAGE
BIN
WIDTH
(PPM)
BASELINE
CORRECTION NORMALIZATION
CLUSTERING
ANALYSIS
CLUSTERS
(No) STRAINS
ROSETTE 0.04
linear
non-normalized
HCA 2 vtc 1, vtc 2
HCA/ ANOVA
/no correction
4 wt, vtc 1, vtc 2,
vtc 3
HCA/ANOVA
/Bonferroni
correction
3 wt, vtc 1, vtc 4
HPV
HCA 2 vtc 1, vtc 2
HCA/ ANOVA
/no correction
4 wt, vtc 1, vtc 2,
vtc 3
HCA/ANOVA
/Bonferroni
correction
2 wt, vtc 2
Total Area
HCA 2 vtc 1, vtc 2
HCA/ ANOVA
/no correction
2 vtc 1, vtc 2
HCA/ANOVA
/Bonferroni
correction
5 wt, vtc 1, vtc 2,
vtc 3, vtc 4
auto
non-normalized
HCA 2 vtc 1, vtc 2
HCA/ ANOVA
/no correction
4 wt, vtc 1, vtc 2,
vtc 3
HCA/ANOVA
/Bonferroni
correction
3 wt, vtc 1, vtc 4
HPV
HCA 2 vtc 1, vtc 2
HCA/ ANOVA
/no correction
4 vtc 1, vtc 2, vtc 3,
vtc 4
HCA/ANOVA
/Bonferroni - 5
wt, vtc 1, vtc 2,
vtc 3, vtc 4
Total Area
HCA 2 vtc 1, vtc 2
HCA/ ANOVA
/no correction
4 vtc 1, vtc 2, vtc 3,
vtc 4
HCA/ANOVA
/Bonferroni - 5
wt, vtc 1, vtc 2,
vtc 3, vtc 4
53
Table 3.3 Clustering analyses of all plant types in flowering stage with 0.01 bin width and
different processing combinations. Cluster No = number of clusters where all the three biological
replicates within and between strain clustered together, Strains = represents the strains in which are
all the biological replicates are clustered close to each other
STAGE
BIN
WIDTH
(PPM)
BASELINE
CORRECTION NORMALIZATION
CLUSTERING
ANALYSIS
CLUSTERS
(No) STRAINS
FLOWERING 0.01
linear
non-normalized HCA 1 vtc 3
HCA/ ANOVA /no
correction
1 vtc 3
HCA/ANOVA
/Bonferroni 1 wt
HPV HCA 1 vtc 3
HCA/ ANOVA /no
correction
1 vtc 3
HCA/ANOVA
/Bonferroni 0 -
Total Area HCA 1 vtc 3
HCA/ ANOVA /no
correction
1 vtc 3
HCA/ANOVA
/Bonferroni 1 wt
auto
non-normalized HCA 1 vtc 3
HCA/ ANOVA /no
correction
2 vtc 1, vtc 3
HCA/ANOVA
/Bonferroni 0 -
HPV HCA 1 vtc 3
HCA/ ANOVA /no
correction
1 vtc 1
HCA/ANOVA
/Bonferroni 0 -
Total Area HCA 1 vtc 3
HCA/ ANOVA /no
correction
1 vtc 3
HCA/ANOVA
/Bonferroni 2 wt, vtc 1
54
Table 3.4 Clustering analyses of all plant types in flowering stage with 0.04 bin width and
different processing combinations. Cluster No = number of clusters where all the three biological
replicates within and between strain clustered together, Strains = represents the strains in which are
all the biological replicates are clustered close to each other. *Equal refers to the clusters which
presented no hierarchy.
STAGE
BIN
WIDTH
(PPM)
BASELINE
CORRECTION NORMALIZATION CLUSTERING ANALYSIS
CLUSTER
S (No) STRAINS
FLOWERING
0.04
linear
non-normalized HCA 1 vtc 2
HCA/ ANOVA /no
correction
2 vtc 1, vtc
3
HCA/ANOVA
/Bonferroni - -
HPV HCA 1 vtc 2
HCA/ ANOVA /no
correction
1 vtc 2
HCA/ANOVA
/Bonferroni 5
EQUAL
*
Total Area HCA 1 vtc 2
HCA/ ANOVA /no
correction
2 vtc 2, vtc
3
HCA/ANOVA
/Bonferroni 0 -
auto
non-normalized HCA 1 vtc 2
HCA/ ANOVA /no
correction
2 vtc 1, vtc
3
HCA/ANOVA
/Bonferroni 0 -
HPV HCA 1 vtc 2
HCA/ ANOVA /no
correction
2 vtc 1, vtc
3
HCA/ANOVA
/Bonferroni 5
EQUAL
*
Total Area HCA 1 vtc 2
HCA/ ANOVA /no
correction
2 vtc 1, vtc
3
HCA/ANOVA
/Bonferroni 0 -
55
HCA was also performed on non-significant data (no ANOVA was performed here) as listed
in the Tables 3.1 to 3.4. We observed the biological replicates were not clustered together,
which makes sense because ANOVA is able to calculate the variability of the data and
filtered out all the features which might be noise. Good clustering shows good experimental
replication and consistency of the replicates (Ward et al., 2003).In general normalised data
showed better clusters than non-normalized data and both the normalizations on the data
were comparable in terms of clusters showing that the two different types of normalizations
did not outperform each other, although total area normalization gave slightly increase in the
number of clusters.
Definite differentiation was observed in the rosette stage where the separation between wild
type and four mutants were obvious which also shows the reliability of all the three
biological replicates and hence the rosette stage was chosen than flowering stage for further
analysis to determine the influence of different baseline correction and bin widths on the
data.
Four combinations from rosette stage were used to carry on further analysis: (1) Total area
normalisation, (2) ANOVA (Bonferroni corrected bins), (3) auto and (4) linear baseline
correction. The baseline correction was performed with 0.01 and 0.04 ppm bin widths. The
results were listed in Tables 3.1 and 3.2.Auto baseline correction combination in 0.01 ppm
bin width although comparable discriminated few more bins than linear baseline correction
whereas they were equal in 0.04 ppm. It was concluded, that bin width 0.01 ppm
discriminated more metabolites than 0.04 ppm, as bin width 0.01 ppm separated two to four
times more integrated buckets (12-29 in numbers) than 0.04ppm (7 in numbers).
56
Wild type and mutants classified themselves as five crisp clusters and in all the four
combinations studied. In all four combinations chosen as mentioned before, the vtc2
grouped itself as a separate group from the other strains (Figure 3.4). The separate grouping
of vtc2 suggested it shows the different behaviour in metabolic level, this may be due to
lowest ascorbate level. As mentioned Chapter 1 the ascorbate levels in vtc 2 mutants are 10-
15 % of Wild type as compared to other which contained about 50 -70 % for vtc 3 and 4
(Conklin et al 2000; Dowdle et al, 2007).
Figure 3.4 Heatmap for the
1H NMR based metabolic profiles of Arabidopsis thaliana Wild type
and four ascorbate-deficient mutants in rosette. The scale shows the colour intensity from highest
to lowest. Red = Highest, Green = Lowest. Different colours represent different metabolites intensity
which is significantly different between different strains. The wildtype cluster and its corresponding
discriminating resonances (1.32, 1.33) are shown in ellipses.
57
DIFFERENTIALLY EXPRESSED BINS BETWEEN WT AND VTC MUTANTS.
The four combinations from rosette stage the bins separated by ANOVA and HCA were
traced to their resonances and the list of all the resonances responsible for the separation of
classes and groups at spectral level and are tabulated in Table 3.5. Major differences occurred
in the regions 0-6 ppm which corresponds to the aliphatic and carbohydrate regions in the
spectra. The regions where major resonances are discriminated correspond to the regions of
amino acids and fatty acids (Ward et al 2003).The unique resonances that showed visual
differences across the strains from two bin widths and the HCA heatmaps were extracted and
tabulated in Table 3.6. There are also some resonances which clearly might be from the same
peak area. The bin width 0.01 ppm has also picked up resonances from carbohydrate regions
than 0.04 ppm (e.g. 5.25 and 5.80 ppm).
The resonance 1.32, 1.33 that discriminated WT and other mutants were traced in the
original spectra shows the presence of peaks (Figure 3.6). The intensity of both the
resonances shows negative correlation to the ascorbate concentration in the vtc mutants and
was proportional to the ascorbate deficiency with across strains (Figure 3.5). Identifying the
spectral patterns forms the main focus rather identifying and quantifying metabolites. These
resonances (features) can then used to identify the corresponding metabolites from the NMR
database.
58
Table 3.5 List of resonances showing discrimination in the spectral regions.
REGIONS(ppm) RESONANCES(ppm)
0.00-2.00 0.96,0.97,0.98,1.00,1.04, 1.32,1.33,1.36, 1.48,1.81,1.88, 1.89 and 1.90
2.00-3.00
2.10, 2.11, 2.12,2.13, 2.14, 2.15, 2.39,2.40, 2.43, 2.45,2.48,2.46,
2.62,2.71,2.72, 2.91, 2.92,2.93 and 2.94
Table 3.6 Resonances different between strains at spectral level with two bin widths. (Wt=Wild
type, vtc=ascorbate-deficient mutants)
STRAINS 0.01ppm 0.04ppm
wt 1.32,1.33 1.36
vtc1 2.12,2.13,2.14,2.15,2.43,2.45,2.46,2.47,2.72, 5.80 0.99,2.48,2.72
vtc2 2.39,2.72 2.40, 2.48,2.72
vtc3 5.18,5.80 2.40
vtc4 0.96,0.97,0.98,1.00,1.33,2.39 and 2.94 0.99,1.04
The resonances (1.32 and 1.33) were identified using the information from the previous
literature. These resonances (1.32 and 1.33) putatively correspond to threonine and lactate
respectively (Urenjak et al., 1993).Threonine is an amino acid in plants which is involved in
defence mechanism with other defence related metabolites (Gonzales-Vigila, 2011). In plants,
pyruate, the end product of glycolysis, is converted to lactate by an enzyme called lactate
dehydrogenase when oxygen is absent or in short supply (anoxia).Lactate provides the signal
triggering for ethanol production and helps for prolonged survival in the absence of oxygen
(Xia and Saglio 1992, Tadege, et al 1999).We are unable to investigate the nature of
relationship between ascorbate and threonine and lactate at this stage. It needs further
investigation.
59
Figure 3.5 The intensity of the resonances 1.32 and 1.33ppm across the strains. Each dot
represents the intensity of resonance. WT shows the lowest intensity than vtc mutants.
Figure 3.6 Identification of the resonances 1.32 and 1.33 ppm (shown as ellipses) in the
original NMR spectra of one of the vtc mutant. The expanded regions are shown in the box. The
insert is the NMR spectra of Wild type. The 1H-NMR-spectra shows very low signal in WT than vtc
mutants.
Wildtype
vtc1vtc2
vtc3vtc4
Wildtype
vtc1vtc2
vtc3 vtc4
0
5000
10000
15000
20000
25000
30000
Strains
Intensity
1.32ppm
1.33ppm
60
Chapter 4
CONCLUSION
1 H-NMR spectroscopy has proved to be a valuable tool for unbiased metabolite profiling of
Arabidopisis ascorbate-deficient mutants. ANOVA and HCA highlighted differences
between the growth stages and strains. Combinations of different parameters were used to
find the best method for the analysis of NMR-based metabolites data. The method was
developed which allowed making multiple comparison such as two growth stages and five
strains. The vtc2 mutants grouped separately from other strains which contained the lowest
ascorbate suggested ascorbate deficiency might have some effect on the metabolic level.
Two interesting resonances 1.32 and 1.33ppm were separated which were negatively
correlated with the ascorbate levels in the vtc mutants. These resonances (1.32 and 1.33) are
putatively identified as threonine and lactate which needs further validation by running
standards.
61
REFERENCE
Aharoni, et al. 2002
Aharoni A, De Vos CHR, Verhoeven HA, Maliepaard, CA, Kruppa G, Bino R,
Goodenowe DB. 2002. OMICS 6, 217-234.
Abdel-Farid et al.2009
Abdel-Farid IB, Jahangir M, van den Hondel CAMJJ, Kim HK,Choi YH, Verpoorte
R.2009. Fungal infection-induced metabolites in Brassica rapa. Plant Sci 176,608-
615.
Agnolet et al.2010
Agnolet S, Jaroszewski JW, Verpoorte R, Staerk D.2010.1H NMR-based
metabolomics combined with HPLC-PDA-MSSPE-NMR for investigation of
standardized Ginkgo biloba preparations. Metabolomics 6, 292-302.
Anterola and Lewis 2002
Anterola AM and Lewis NG.2002.Trends in lignin modification: a comprehensive
analysis of the effects of genetic manipulations/mutations on lignification and vascular
integrity.Phytochemistry 61,221-294.
Bailey et al.2003
Bailey NJC, Oven M, Holmes E, Nicholson JK, Zenk MH.2003. Metabolomic
analysis of the consequences of cadmium exposure in Silene cucubalus cell cultures
via 1H NMR spectroscopy and chemometrics. Phytochemistry 62, 851-858.
Barros et al.2010
Barros E, Lezar S,Anttonen MJ,Dijk JPV,Röhlig RM,Kok EJ,Engel K-
Heinz.2010.Comparison of two GM maize varieties with a near-isogenic non-
GM variety using transcriptomics, proteomics and metabolomics. Plant
Biotechnology Journal 8(4), 436-451.
Barth et al. 2004
Barth C, Moeder W, Klessig DF, Conklin PL. 2004.The timing of senescence and
response to pathogens is altered in the ascorbate-deficient Arabidopsis mutant vitamin
vtc-1-1.Plant Physiology 134,1784-1792.
Bino et al.2004
Bino RJ, Hall RD, Fiehn O, Kopka J, Saito K, Draper John, Nikolau BJ, Mendesh P,
Roessner-Tunalii U, Beale MH, Trethewey RN, Langel BM, Wurtele ES, Sumner
LW. 2004. Potential of metabolomics as a functional genomics tool. Trends in Plant
Science 9(9), 418-425.
62
Bochner et al.2001
Bochner BR, Gadzinski P, Panomitros E. 2001. Phenotype microarrays for high-
throughput phenotypic testing and assay of gene function.Genome Res 11(7), 1246-
55.
Blount et al.2002
Blount JW, Masoud S, Sumner LW, Huhman D and Dixon RA.(2002).Over-
expression of cinnamate 4-hydroxylase leads to increased accumulation of
acetosyringone in elicited tobacco cell-suspension cultures. Planta 214, 902-910.
Brosche et al.2005
Brosche M,Vinocur B, Alatalo ER, Lamminmaki A, Teichmann T, Ottow EA,
Djilianov D, Afif D, Bogeat-Triboulot MB, Altman A.2005.Gene expression and
metabolite profiling of Populus euphratica growing in the Negev desert. Genome Biol
6, R101.
Brown 2002
Brown K.2002. Something to sniff at: unbottling floral scent. Science 296, 2327-2329.
Cardoso-Taketa et al. 2008
Cardoso-Taketa AT, Pereda-Miranda R, Choi YH, Verpoorte R, Villarreal ML.
2008. Metabolic profiling of the Mexican anxiolytic and sedative plant Galphimia
glauca using nuclear magnetic resonance spectroscopy and multivariate data analysis.
Planta Med 74, 1295-1301.
Catchpole et al.2005
Catchpole GS, Beckmann M, Enot D P, Mondhe M, Zywicki B, Taylor J, Hardy N,
Smith A, King R D, Kell DB, Fiehn O, Draper J,Van Montagu MCE.2005.
Hierarchical Metabolomics Demonstrates Substantial Compositional Similarity
between Genetically Modified and Conventional Potato Crops. PNAS 40, 14458-
14462.
Chen, et al.2003
Chen Z, Young T E, Ling J, Chang S-C, Gallie DR. 2003. Increasing vitamin C
content of plants through enhanced ascorbate recycling. PNAS 100(6), 3525-3530.
Cobas and Sardina 2003
Cobas JC, Sardina FJ. 2003. Nuclear magnetic resonance data processing. MestRe-C:
A software package for desktop computers. Concepts in Magnetic Resonance Part A
19A (2), 80-96.
Colville and Smirnoff 2008
Colville L and Smirnoff N. 2008. Antioxidant status, peroxidase activity, and PR
protein transcript levels in ascorbate-deficient Arabidopsis thaliana vtc mutants J Exp
Bot 59 (14), 3857-3868.
63
Conklin, et al. 2006
Conklin PL, Gatzek S, Wheeler GL, Dowdle J, Raymond MJ, Rolinski S,
Isupov M, Littlechild JA, Smirnoff N.2006.Arabidopsis thaliana VTC4 Encodes L-
Galactose-1-P Phosphatase, a Plant Ascorbic Acid Biosynthetic Enzyme.The Journal
Of Biological Chemistry 281(23),15662-15670.
Conklin and Last 1995
Conklin PL, Last RL.1995.Differential accumulation of antioxidant enzyme mRNAs
in Arabidopsis thaliana exposed to ozone. Plant Physiol 109, 203-212.
Conklin et al.1999
Conklin PL, Norris SR, Wheeler GL, Williams EH, Smirnoff N and Last
RL.1999.Genetic evidence for the role of GDP-mannose in plant ascorbic acid
(Vitamin C) biosynthesis. PNAS 96(7), 4198 - 4203.
Conklin et al. 2000
Conklin PL, Saracco SA, Norris S R,Last RL. 2000. Identification of Ascorbic Acid-
Deficient Arabidopsis thaliana Mutants. Genetics 154, 847-856.
Cordoba-Pedregosa et al. 2003
Córdoba-Pedregosa MC,Córdoba F, Villalba JM, Gonzalez-Reyes JA.2003. Zonal
changes in ascorbate and hydrogen peroxide contents, peroxidase, and ascorbate-
related enzyme activities in onion roots. Plant Physiology131, 697-706.
Craig et al. 2006
Craig A, Cloarec O, Holmes E, Nicholson JK, Lindon JC.2006.Scaling and
normalization effects in NMR spectroscopic metabonomic data sets. Anal Chem 78,
2262-2267.
Davey et al.2000
Davey MW,Montagu MV,Inzé D,Sanmartin M,Kanellis A,Smirnoff N,Benzie IJJ,
Strain JJ,Favell D,Fletcher J.2000. Plant L-ascorbic acid: chemistry, function,
metabolism, bioavailability and effects of processing.Journal of the Science of Food
and Agriculture 80(7), 825-860.
Bennett 2005
Bennett D.2005. Growing pains for metabolomics. The Scientist 19 (8), 25-28.
Defernez and Colquhoun 2003
Defernez M, Colquhoun IJ.2003. Factors affecting the robustness of metabolite
fingerprinting using 1H NMR spectra. Phytochemistry 62, 1009-1017.
Defernez et al.2004
Defernez M, Gunning YM, Parr AJ,Shepherd LVT, Davies HV and Colquhoun
IJ.2004.NMR and HPLC–UV profiling of potatoes with genetic modifications to
metabolic pathways. Journal of Agricultural and Food Chemistry 52, 6075-6085.
64
De Gara et al.1997
DeGara L, De Pinto MC, Arrigoni O.1997. Ascorbate synthesis and ascorbate
peroxidase activity during the early stage of wheat germination. Physiol
Plant 100, 894-900.
Delaglio et al.1995
Delaglio F, Grzesiek S, Vuister GW, Zhu G, Pfeifer J, Bax A.1995.Nmrpipe – a
Multidimensional Spectral Processing System Based on Unix Pipes. Journal of
Biomolecular NMR 6(3), 277-293.
De Tullio 2004
De Tullio, M.2004. How does ascorbate prevent scurvy? A survey of the non-
antioxidant functions of vitamin C. in Vitamin C. Function and Biochemistry in
Animals and Plants (Asard, H., May, J.M. and Smirnoff, N. eds). Oxford:Bios
Scientific Publishers,159-171.
De Vos et al. 2007
De Vos RCH, Moco S, Lommen,Keurentjes JJB, Bino RJ,Hall
RD.2007.Untargeted large-scale plant metabolomics using liquid chromatography
coupled to mass spectrometry. Nature Protocols 2, 778 - 791.
Dixon et al.2006
Dixon RA, Gang DR, Charlton AJ, Fiehn O, Kuiper HA, Reynolds TL, Tjeerdema RS,
Jeffery EH, German JB, Ridley WP, Seiber JN.2006. Applications of Metabolomics in
Agriculture.J Agric Food Chem 54, 8984-8994.
Dowdle et al 2007
Dowdle J, Ishikawa T, Gatzek S, Rolinski S, Smirnoff N.2007. Two genes
in Arabidopsis thaliana encoding GDP-l-galactose phosphorylase are required for
ascorbate biosynthesis and seedling viability. The Plant Journal 52(4), 73-689.
Dunn and Ellis 2005
Dunn WB, Ellis DI.2005.Metabolomics: Current analytical platforms and
methodologies TrAC. Trends in Analytical Chemistry 24(4), 285-294.
Eisenreich and Bacher 2007
Eisenreich W, Bacher A.2007.Advances of high-resolution NMR techniques in the
structural and metabolic analysis of plant biochemistry. Phytochemistry 68, 22-24.
Eskling and Akerlund 1997
Eskling M, Akerlund HE.1997.Changes in the quantities of violaxanthin de-epoxidase,
xanthophylls and ascorbate in spinach upon shift from low to high light.
Photosynthesis Research 57(1), 41-50.
Fiehn, et al.2000a
Fiehn O, Kopka J, Dörmann P, Altmann T, Trethewey RN, Willmitzer L. 2000.
Metabolite profiling for plant functional genomics. Nat Biotechnol 18, 1157-1161.
65
Fiehn et al.2000b
Fiehn O, Kopka J, Trethewey RN and Willmitzer L (2000) Identification of
uncommon plant metabolites based on calculation of elemental compositions using gas
chromatography and quadrupole mass spectrometry. Anal Chem 72, 3573-3580.
Fiehn 2002
Fiehn O.2002. Metabolomics – the link between genotype and phenotype. Plant Mol
Biol 48, 155-17.
Fiehn and Weckwerth 2002
Fiehn O, Weckwerth W.2002.Can we discover novel pathways using metabolomic
analysis? Curr. Opin Biotechnol 13, 156-160.
Flamini et al.2003
Flamini G, Cioni PL, Morelli I.2003.Volatiles from leaves, fruits, and virgin oil
from Olea europaea cv. Olivastra Seggianese from Italy.Journal of Agricultural and
Food Chemistry 51, 1382-1386.
Fraser et al.2000
Fraser PD, Pinto EMS, Holloway DE, Bramley PM.2000.Application of high-
performance liquid chromatography with photodiode array detection to the metabolic
profiling of plant isoprenoids. The Plant Journal 24(4), 551-558.
Fryer 1992
Fryer MJ.1992.The antioxidant effects of thylakoid vitamin E (α-tocopherol).Plant
Cell environment 15, 381-392.
Fukusaki and Kobayashi 2005
Fukusaki E and Kobayashi A.2005. Plant metabolomics: potential for practical
operation. Journal Of Bioscience And Bioengineering 100(4), 347-354.
Gamache et al.2004
Gamache PH, Meyer DF, Granger MC.2004.Metabolomic applications of
electrochemistry/mass spectrometry. J Am Soc Mass Spectrom 15, 1717-26.
Gander1982
Gander JE.1982. Polyhexose acids: relation to hexose phosphate metabolism. Plant
Carbohydrates I: Intracellular Carbohydrates,edited by F. A. Loewus and W.
Tanner.Springer-Verlag, Berlin.77-102.
Garratt et al.2005
Garratt LC, Linforth R, Taylor AJ, Lowe KC, Power JB and Michael RD. 2005.
Metabolite fingerprinting in transgenic lettuce.Plant Biotechnology Journal.165-174.
66
Gatzek et al. 2002
Gatzek S, Wheeler GL, Smirnoff N.2002.Antisense suppression of l-galactose
dehydrogenase in Arabidopsis thaliana provides evidence for its role in ascorbate
synthesis and reveals light modulated l-galactose synthesis.Plant J 30, 541-553.
Gautier et al. 2009
Gautier H, Massot C, Stevens R, Sérino S, Génard M.2009. Regulation of tomato fruit
ascorbate content is more highly dependent on fruit irradiance than leaf irradiance.
Annals of Botany 103(3), 495-504
Gidman et al. 2003
Gidman E, Goodacre R, Emmett B,Smith AR, Gwynn-Jones D.2003.Investigating
plant–plant interference by metabolic fingerprinting. Phytochemistry 63, 705-71.
Gonzales-Vigila 2011
Gonzales -Vigila E, Bianchetti CM, Phillips GN, Howea GA.2011.Adaptive evolution
of threonine deaminase in plant defense against insect herbivores. PNAS 5(108),
5897-5902.
Hadfield et al.2001
Hadfield ST, Hole SJW, Howe PWA, Stanley PD.2001. Metabolite profiling by NMR
for high-throughput mode of action identification of screen hits. Weeds 2, 551-556.
Hall et al 2008
Hall RD, Brouwer ID, Fitzgerald MA.2008.Plant metabolomics and its potential
application for human nutrition. Physiologia Plantarum132(2), 162-175.
Halket et al. 2005
Halket JM, Waterman D, Przyborowska AM, Patel RKP, Fraser PD, Bramley
PM.2005.Chemical derivatization and mass spectral libraries in Metabolic profiling by
GC/MS and LC/MS/MS. J Exp Bot 56, 219-243.
Hamzehzarghani et al.2005
Hamzehzarghani H, Kushalappa AC, Dion Y, Rioux S, Comeau
A,Yaylayan V.2005.Metabolic profiling and factor analysis to discriminate
quantitative resistance in wheat cultivars against fusarium head blight.Physiological
and Molecular Plant Pathology 66, 119-133.
Havaux 2003
Havaux M.2003.Spontaneous and thermo induced photon emission:New methods to
detect and quantify oxidative stress in plants.Trends Plant Sci. 8,409-413.
Hirai et al.2004
Hirai MY, Yano M, Goodenowe DB, Kanaya S, Kimura T, Awazuhara M, Arita M,
Fujiwara T and Saito K. 2004.Integration of transcriptomics and metabolomics for
understanding of global responses to nutritional stresses in Arabidopsis thaliana.
PNAS 6(101), 10205-10210.
67
Hole et al.2000
Hole SJW, Howe PWA, Stanley PD and Hadfield ST.2000. Pattern recognition
analysis of endogenous cell metabolites for high throughput mode of action
identification: removing the post screening dilemma associated with whole-organism
high throughput screening. J Biomol Screen 5, 335-342.
Huang et al. 2005
Increased sensitivity to salt stress in an ascorbate-deficient Arabidopsis mutant
Huang C, He W, Guo J, Chang X, Su P, Zhang L.2005.J Exp Bot 56(422),3041-3049.
Ishikawa et al 2006a, b
Ishikawa T, Dowdle J, Smirnoff N.2006a.Progress in manipulating ascorbic acid
biosynthesis and accumulation in plants. Physiol.Plant.126, 343-355.
Ishikawa T, Masumoto I, Iwasa N, Nishikawa H, Sawa Y,Shibata, H, Nakamura A,
YabutaY, Shigeoka,S.2006b.Functional characterization of d-galacturonic acid
reductase, a key enzyme of the ascorbate biosynthesis pathway, from Euglena
gracilis. Biosci Biotech Biochem 70, 2720-2726.
Jackson and Mantsch 1996
Jackson M and Mantsch HH.1996. The medical challenge to infrared spectroscopy.
Journal of Molecular Structure 408-409, 105-111.
Jahangir et al.2008
Jahangir M, Kim HK, Choi YH, Verpoorte R.2008.Metabolomic response of Brassica
rapa submitted to pre-harvest bacterial contamination.Food Chemistry 107(1), 362-
368.
Johnson et al.2003
Johnson HE, Broadhurst D, Goodacre R,Smith AR.2003 Metabolic fingerprinting of
salt-stressed tomatoes Phytochem 62, 919-928.
Jung et al.2000
Jung W, Yu O, Lau SMC, O'Keefe DP, Odell J, Fader G, McGonigle
B.2000.Identification and expression of isoflavone synthase, the key enzyme for
biosynthesis of isoflavones in legumes.Nature Biotechnology 18, 208-212
Katajamaa and Oresic 2007
Katajamaa M and Oresic M.2007. Data processing for mass spectrometry-based
metabolomics. Journal of Chromatography A 1158, 318-328.
Kercheve et al. 2011
Kerchev PI, Fenton B, Foyer CH, Hancock RD.2011.Plant responses to insect
herbivory: Interactions between photosynthesis, reactive oxygen species and hormonal
signalling pathways. Plant Cell Environ doi:10.1111/j.1365-3040.2011.02399.
68
Ketchum et al. 2003
Ketchuma REB, Rithnerb CD,Qiua D, KimaYS,Williams RM,Croteau RB.2003.
Taxus metabolomics: methyl jasmonate preferentially induces production of taxoids
oxygenated at C-13 in Taxus x media cell cultures.Phytochemistry 62(6), 901-909.
Kikuchi et al.2004
Kikuchi J, Shinozaki K, Hirayama T.2004. Stable Isotope Labeling of Arabidopsis
thaliana for an NMR-Based Metabolomics Approach. Plant Cell Physiol 45(8), 1099-
1104.
Kim, et al. 2010
Kim HK, Choi YH, Verpoorte R.2010.NMR-based metabolomic analysis of
plants.Nature Protocols 5,536-549.
Kim et al. 2011
Kim HU, Sohn SB, Lee SY.2011.Metabolic network modeling and simulation for
drug targeting and discovery.Article first published online: 29 NOV2011.
Kliebenstein 2004
Kliebenstein DJ.2004.Secondary metabolites and plant/environment interactions: a
view through Arabidopsis thaliana tinged glasses. Plant,Cell& Environment 27(6),
675-684.
Krishnan et al 2005
Krishnan P, Kruger NJ, Ratcliffe RG.2005. Metabolite fingerprinting and profiling in
plants using NMR. J Exp Bot 56(410), 255-265.
Kristal et al. 2002
Kristal BS, Callahan KV, Matson WR.2002. Simultaneous Analysis of Multiple
Redox-Active Metabolites from Biological Matrices. Methods in Molecular Biology
186, 185-194.
Kruger et al. 2008
Kruger NJ, Troncoso-Ponce MA, Ratcliff e RG. 2008.1H NMR metabolite
fingerprinting and metabolomic analysis of perchloric acid extracts from plant
tissues. Nat Protoc 3, 1001-1012.
Lange et al. 2001
Lange BM, Ketchum REB, Croteau RB.2001.Isoprenoid Biosynthesis. Metabolite
Profiling of Peppermint Oil Gland Secretory Cells and Application to Herbicide
Target Analysis.Plant Physiology 127(1), 305-314.
Laing et al. 2007
Laing WA,Wright MA,Cooney J, Bulley SM.2007.The missing step of the l-galactose
pathway of ascorbate biosynthesis in plants, an l-galactose guanyltransferase,increases
leaf ascorbate content. Proc Natl Acad Sci. USA 104, 9534-9539.
69
Larkindale et al. 2005
Larkindale J, Hall JD, Knight MR, Vierling E.2005.Heat stress phenotypes of
Arabidopsis mutants implicate multiple signaling pathways in the acquisition of
thermo tolerance. Plant Physiol 138, 882-897.
Le Gall et al.2001
Le Gall G, Puaud M and Colquhoun IJ.2001.Discrimination between orange juice and
pulp wash by 1H nuclear magnetic resonance spectroscopy: identification of marker
compounds, J Agric Food Chem 49, 580-588.
Le Gall et al. 2003
Le Gall G, Colquhoun IJ, Davis AL, Collins GJ, Verhoeyen ME.2003.Metabolite
profiling of tomato (Lycopersicon esculentum) using 1H NMR spectroscopy as a tool
to detect potential unintended effects following a genetic modification. J Agric Food
Chem 51, 2447-2456.
Le Gall et al.2005
Le Gall G, Metzdorff SB, Pedersen J, Bennet RN,Colquhoun IJ.2005.Metabolite
profiling of Arabidopsis thaliana (L.) plants transformed with an antisense chalcone
synthase gene. Metabolomics 1(2), 181-198.
Leiss et al.2009
Leiss KA,Maltese F,Choi YH,Verpoorte R, Klinkhamer PGL.2009. Identification of
Chlorogenic Acid as a Resistance Factor for Thrips in Chrysanthemum.Plant
Physiology 150, 1567-1575.
Li et al. 2008
Li MJ, Ma FW, Zhang M, Pu F.2008.Distribution and metabolism of ascorbic acid in
apple fruits (Malus domestica Borkh cv. Gala.)Plant Sci 174, 606-612.
Li et al. 2009
Li S, Lauri A, Ziemann M, Busch A, Bhave M, Zachgo S.2009.Nuclear activity of
ROXY1, a glutaredoxin interacting with TGA factors, is required for petal
development in Arabidopsis thaliana. Plant Cell 21, 429-441.
Linster et al. 2007, 2008
Linster CL,Gomez TA,Christensen KC,Adler LN, Young BD, Brenner C, Clarke
SG.2007.Arabidopsis VTC2 encodes a GDP-l-galactose phosphorylase, the last
unknown enzyme in the Smirnoff–Wheeler pathway to ascorbic acid in plants. J Biol
Chem 282, 18879-18885.
Linster 2008
Linster CL, Adler LN, Webb K, Christensen KC, Brenner C, Clarke SG.2008.A
second GDP-l- galactose phosphorylase in Arabidopsis en route to vitamin C: covalent
intermediate and substrate requirements for the conserved reaction. J Biol Chem 28,
18483-18492.
70
Linster and Clarke 2008
Linster CL,Clarke SG.2008.Ascorbate biosynthesis in higher plants: the role
of VTC2.Trends in Plant Science 13(11), 567-573.
Lisec et al.2006
Lisec J, Schauer N, Kopka J, Willmitzer L, Fernie AR (2006) Gas chromatography
mass spectrometry-based metabolite profiling in plants. Nat Protoc 1, 387-396
Liu et al. 2002
Liu CJ, Blount JW, Steele CL, Dixon RA.2002.Bottlenecks for metabolic engineering
of isoflavone glycoconjugates in Arabidopsis.Proc Natl Acad Sci USA 99,14578-
1458.
Loewus 1999
Loewus FA.1999. Biosynthesis and metabolism of ascorbic acid in plants and of
analogs of ascorbic acid in fungi. Phytochemistry 52, 193-210.
Lutterbach and Stockigt 1994
Lutterbach R and Stöckigt J. 1994. In vivo Investigation of Plant-Cell Metabolism by
means of natural-abundance 13C-NMR spectroscopy Helv Chim Acta 77, 2153-2161.
Lutterbach and Stockigt 1995
Lutterbach R and Stöckigt J. 1995.Dynamics of the biosynthesis of methylursubin
in plantcells employing in vivo 13C NMR without labelling.Phytochemistry 40, 801-
806.
Luwe et al. 1993
Luwe MWF, Takahama U,Heber U.1993.Role of Ascorbate in Detoxifying Ozone in
the Apoplast of Spinach (Spinacia oleracea) Leaves Plant Physiol 101, 969-976.
Matsuda et al.2010
Matsuda F, Hirai MY, Sasaki E, Akiyama K, Sakakibara KY, Provart NJ, Sakurai T,
Shimada Y, Saito K.2010.AtMetExpress Development: A Phytochemical Atlas of
Arabidopsis Development. Plant Physiology 152(2),566-578.
Matsumoto and Kuhara 1996
Matsumoto I and Kuhara T. 1996. A new chemical diagnostic method for inborn
errors of metabolism by mass spectrometry-rapid, practical and simultaneous urinary
metabolites analysis. Mass Spectrometry Reviews 5(1), 43-57.
Mochida et al. (2009)
Mochida K, Furuta T, Ebana K, Shinozaki K and Kikuchi J.2009.Correlation
exploration of metabolic and genomic diversity in rice BMC. Genomics 10, 568.
Modarai et al.2010
Modarai M, Yang M,Suter A,Kortenkamp A, Heinrich M.2010.Metabolomic
Profiling of Liquid Echinacea Medicinal Products with In Vitro Inhibitory Effects on
Cytochrome P450 3A4 (CYP3A4). Planta Med 76(4), 378-385.
71
Merchant et al.2006
Merchant A, Adams MA, Richter A and Popp M. 2006. A metabolite approach
provides functional links among eucalyptus taxonomy, physiology and evolution.
Phytochemistry 67,402-408.
Miller and Miller 2000
Miller JN, Miller JC.2000.Statistic and Chemometrics for Analytical Chemistry.4th
Ed. Prentice Hall, New York.
Noctor and Foyer 1998
Noctor G and Foyer CH.1998.Ascorbate And Glutathione: Keeping Active Oxygen
under Control. Annu Rev Plant Physiol. Plant Mol Biol 49, 249-279.
Pallanca and Smirnoff 2000
Pallanca JE, Smirnoff N.2000.The control of ascorbic acid synthesis and turnover in
pea seedlings. J Exp Bot 51, 669-674.
Palsson 2000
Palsson BO.2000.The challenges of in silico biology. Nat Biotechnol18, 1147-1150.
Pastori 2003
Pastori GM,Kiddle G, Antoniw J, Bernard S, Jovanovic SV, Verrier PJ, Noctor
G,Foyer C.2003.Leaf Vitamin C Contents Modulate Plant Defense Transcripts and
Regulate Genes That Control Development through Hormone Signaling. The Plant
Cell 15,939-951.
Pavet et al. 2005
Pavet V, Olmos E, Kiddle G, Mowla S, Kumar S, Antoniw J, Alvarez ME, Foyer
CH.2005.Ascorbic acid deficiency activates cell death and disease resistance
responses in Arabidopsis. Plant Physiol 139, 1291-1303.
Pope et al.1993
Pope JM, Jonas D,Walker RR.1993. Applications of NMR micro imaging to the study
of water, lipid, and carbohydrate distribution in grape berries. Protoplasma, 173, 177-
186.
Prabhu et al.1996
Prabhu V, Chatson KB, Abrams GD,King J.1996.13C Nuclear Magnetic Resonance
Detection of Interactions of Serine Hydroxymethyltransferase with C1-
Tetrahydrofolate Synthase and Glycine Decarboxylase Complex Activities in
Arabidopsis. Plant Physiology 112(1), 207-216.
Ott et al.2003
Ott KH, Aranibar N,Singh B and Stockton GW.2003.Metabonomics classifies
pathways affected by bioactive compounds. Artificial neural network classification of
NMR spectra of plant extracts. Phytochemistry 62, 971-985.
72
Rasmussen et al.2006
Rasmussen B, Cloarec O, Tang H, Stærk D, Jaroszewski JW.2006.Multivariate
Analysis of Integrated and Full-Resolution1H-NMR Spectral Data from Complex
Pharmaceutical Preparations: St. John's Wort.Planta Med 72(6), 556-563
Ratcliffe and Shachar-Hilt 2001
Ratcliffe RG, and Shachar-Hill Y, 2001.Probing plant metabolism with NMR.Annual
review of Plant Physiology and Plant Molecular Biology 52, 499-526.
Robinson et al. (2005)
Robinson, AR, Gheneim R, Kozak RA, Ellis DD, and Mansfield SD.2005.The
potential of metabolite profiling as a selection tool for genotype discrimination
in Populus. J Exp Bot 56, 2807-2819.
Roessner et al.2001
Roessner U, Luedemann A,Brust D,Fiehn O,Linke T,Willmitzer L,Fernie
AR.2001.Metabolic profiling allows comprehensive phenotyping of genetically or
environmentally modified plant systems.The Plant Cell 13,11-29.
Roessner et al.2000
Roessner U, Wagner C,Kopka J,Trethewey RN,Willmitzer L.2000.Simultaneous
analysis of metabolites in potato tuber by gas chromatography-mass spectrometry. The
Plant Journal 23, 131-142.
Roos et al.2004
Roos G, Roeseler C, Bueter KB, Simmen U.2004.Classification and correlation of St.
John's wort extracts by nuclear magnetic resonance spectroscopy, multivariate data
analysis and pharmacological activity. Planta Medica 70, 771-777.
Saito and Matsuda 2010
Saito K, Matsuda F.2010.Metabolomics for functional genomics, systems biology, and
biotechnology. Annu Rev Plant Biol 61,463-489.
Saeed et al. 2006
Saeed AI, Bhagabati NK, Braisted JC, Liang W, Sharov V, Howe EA, Li J,
Thiagarajan M, White JA, Quackenbush J .2006.TM4 microarray software suite. Meth
Enzymol 411, 134-193.
Sato et al.2004
Sato S, Soga T, Nishioka T, Tomita M. 2004. Simultaneous determination of the main
metabolites in rice leaves using capillary electrophoresis mass spectrometry and
capillary electrophoresis diode array detection. Plant Journal 40, 151-163.
Sauter et al.1991
Sauter H, Lauer M and Fritsch H, 1991. Metabolic profiling of plants a new diagnostic
technique. In: Baker DR, Fenyes JG. and Moberg WK, Editors, American Chemical
Society Symposium Series No. 443, American Chemical Society, 288-299.
73
Schad et al.2005
Schad M, Mungur R, Fiehn O,Kehr J.2005.Metabolic profiling of laser micro
dissected vascular bundles of Arabidopsis thaliana. Plant Methods 1, 2.
Schauer 2005
Schauer N, Steinhauser D, Strelkov S, Schomburg D, Allison G, Moritz T, Lundgren
K, Roessner-Tunali U,Forbes MG, Willmitzer L, Fernie AR, Kopka J. 2005. GC-MS
libraries for the rapid identification of metabolites in complex biological
samples. FEBS Lett 579 (6), 1332-7.
Schauer and Fernie 2006
Schauer N and Fernie AR.2006.Plant metabolomics: towards biological function and
mechanism. Trends Plant Sci 11, 508-16.
Schleif 2007
Schleif FM, Hammer B, Villmann T.2007.Margin based Active Learning for LVQ
Networks. Neuro Computing 70(7-9), 1215-1224.
Schoner and Krause 1990
Schöner S, Krause GH.1990.Protective systems against active oxygen species in
spinach: response to cold acclimation in excess light. Planta 180(3), 383-389.
Sharma and Davis 1994
Sharma YK, Davis KR.1994.Ozone-induced expression of stress related genes in
Arabidopsis thaliana. Plant Physiol 105, 1089-1096.
Shulaev 2006
Shulaev.2006. Metabolomics technology and bioinformatics. Brief bioinformatics
7(2), 128-139.
Smilde 2005
Smilde AK, Jansen JJ, Hoefsloot HCJ, Lamers R-JAN, van der Greef J,Timmerman
ME.ANOVA-simultaneous component analysis (ASCA): a new tool for analyzing
designed metabolomics data. Bioinformatics 21 (13), 3043-3048.
Smirnoff 2000
Smirnoff,N.2000. Ascorbic acid: Metabolism and functions of a multifaceted
molecule. Curr Opin Plant Biol 3,229-235.
Smirnoff et al.2001
Smirnoff N, Conklin PL, Loewus FA.2001.Biosynthesis of Ascorbic Acid in Plants: A
Renaissance. Annu Rev Plant Physiol Plant Mol Biol 52, 437-67.
Smirnoff and Wheeler 2000
Smirnoff N, Wheeler GL.2000.Ascorbic Acid in Plants: Biosynthesis and Function.
Critical Reviews in Biochemistry and Molecular Biology 35(4), 291-314.
Smirnoff 2011
Smirnoff, N.2011.Vitamin C: The Metabolism and Functions ofAscorbic Acid in
Plants.Advances in Botanical Research 59
Soga et al 2002
Soga T, Ueno Y, Naraoka H.
intermediates for Bacillus subtilis
electrospray ionization mass spectrometry.
Sumner et al.2003
Sumner, L W, Mendes PJ,Dixon RA.2003.
phytochemistry in the functional genomics era.
Suza et al 2010
Suza WP, Rowe ML,
synthesizes(+)-7-iso-jasmonoyl
728.
Tadege et al. 1999
Tadege M, Dupuis I,
old pathway. 4(8), 320
Tagashira et al.2005
Tagashira N, Plader W
O,Hoshi Y, Kondo K
cucumber lines vary with different chromosomal locations of the transgene.
and Molecular Biology
Tikunov et al.2005
Tikunov Y, Lommen A, de Vos CHR, Verhoeven HA, Bino RJ, Hall RD, Bovy
AG.2005.A novel approach for non targ
scale profiling of tomato fruit volatiles. Plant Physiol
Torgrip et al.2006
Torgrip RJO, Aberg KM
normalization of bio fluid 1D
Tunali et al.2003
Tunali UR, Hegemann B, Lytovchenko A, Carrari F, Bruedigam C, Granot D, Fernie
AR.2003.Metabolic profiling of transgenic tomato plants
reveals that the influence of hexose phosphorylation diminishes during fruit
development. Plant Physiol. 133(1), 84
74
Vitamin C: The Metabolism and Functions ofAscorbic Acid in
Advances in Botanical Research 59,108-177.
Naraoka H.2002.Simultaneous determination of anionic
Bacillus subtilis metabolic pathways by capillary electrophoresis
electrospray ionization mass spectrometry. Anal Chem 74, 2233-9.
Sumner, L W, Mendes PJ,Dixon RA.2003.Plant Metabolomics: Large
phytochemistry in the functional genomics era. Phytochemistry 62
ML, Hamberg M,Staswick PE. 2010. A tomato enzyme
jasmonoyl-L:-isoleucine in wounded leaves. Planta
I, Kuhlemeier C.1999.Ethanolic fermentation: new
320-324.
Plader W, Filipecki M , Yin Z, Wiśniewska A, Gaj P
Kondo K, Malepszy S.2005.The metabolic profiles of transgenic
cumber lines vary with different chromosomal locations of the transgene.
Biology Letters 10(4), 697-710.
Y, Lommen A, de Vos CHR, Verhoeven HA, Bino RJ, Hall RD, Bovy
A novel approach for non targeted data analysis for metabolomics: large
scale profiling of tomato fruit volatiles. Plant Physiol 139, 1125-1137.
KM, Alm E, Schuppe-Koistinen I, Lindberg J
normalization of bio fluid 1D 1H-NMR data. Metabolomics 4(2), 14
Tunali UR, Hegemann B, Lytovchenko A, Carrari F, Bruedigam C, Granot D, Fernie
Metabolic profiling of transgenic tomato plants overexpressing hexokinase
reveals that the influence of hexose phosphorylation diminishes during fruit
Plant Physiol. 133(1), 84-99
Vitamin C: The Metabolism and Functions ofAscorbic Acid in
determination of anionic
metabolic pathways by capillary electrophoresis
9.
Plant Metabolomics: Large-scale
62 (6), 817-836.
enzyme
Planta 231, 717-
Ethanolic fermentation: new functions for an
Gaj P, Szwacka M, Fiehn
2005.The metabolic profiles of transgenic
cumber lines vary with different chromosomal locations of the transgene. Cellular
Y, Lommen A, de Vos CHR, Verhoeven HA, Bino RJ, Hall RD, Bovy
abolomics: large-
1137.
J.2006.A note on
14-121.
Tunali UR, Hegemann B, Lytovchenko A, Carrari F, Bruedigam C, Granot D, Fernie
overexpressing hexokinase
reveals that the influence of hexose phosphorylation diminishes during fruit
75
Tweeddale et al. 1998,1999
Tweeddale H, Notley-McRobb L, Ferenci T.1998.Effect of slow growth on
metabolism of Escherichia coli, as revealed by global metabolite pool
(‘‘metabolome’’) analysis. J Bacteriol 180, 5109-16.
Tweeddale H, Notley-McRobb L, Ferenci T.1999.Assessing the effect of reactive
oxygen species on Escherichia coli using a metabolome approach. Redox Rep 4, 237-
41.
Verdonk et al. 2003
Verdonk JC, Vosb CHR, Verhoeven HA, Harin MA, Tunena AJ, Schuurinka R
C.2003. Regulation of floral scent production in petunia revealed by targeted
metabolomics. Phytochemistry 62(6), 997-1008,
Urenjak et al.1993
Urenjak J, Williams SR, Gadian DG, Noble M.1993.Proton Nuclear Magnetic
Resonance Spectroscopy Unambiguously Identifies Different Neural Cell Types.The
Journal of Neuroscience 13(3), 981-989.
van der Greef et al.2003
van der Greef J, Davidov E, Verheij ER, van der Heijden R, Adourian AS, Oresic M,
Marple EW, Naylor S. 2003.The role of metabolomics in Systems Biology.In Harrigan
GG. and Goodacre R(Eds.). Metabolic Profiling: Its Role in Biomarker Discovery and
Gene Function Analysis. Boston/Dordrecht/London Kluwer Academic Publishers,
170-198.
Viant et al, 2003
Viant MR, Rosenblum ES, Tieerdema RS.2003.NMR-based metabolomics: a powerful
approach for characterizing the effects of environmental stressors on organism health.
Environ Sci Technol 37, 4982-9.
Ward et al.2002
Ward JL, Harris C, Lewis J,Beale M H.2003.Assessment of 1H NMR spectroscopy
and multivariate analysis as a technique for metabolite fingerprinting of Arabidopsis
thaliana. Phytochemistry 62(6), 949-957.
Wang et al. 2009
Wang T, Shao K, Chu Q, Ren Y, Mu Y, Qu L, He J, Jin C and Xia B.2009.
Automics:an integrated platform for NMR-based metabonomics spectral processing
and data analysis. BMC Bioinformatics 10, 83.
Washko et al. 1992
Washko PW,Welch RW, Dhariwal KR, Wang Y, Levine M.1992.Ascorbic acid and
dehydroascorbic acid analysis in biological samples.Anal Biochem 204,1-14.
76
Weichert et al.2002
Weichert,H. Kolbe A, Kraus A, Wasternack C,Feussner I.2002.Metabolic profiling of
oxylipins in germinating cucumber seedlings -lipoxygenase-dependent
degradationof triacylglycerols and biosynthesis of volatile aldehydes. Planta 215, 612-
619.
Wheeler et al.1998
Wheeler GL, Jones MA and Smirnoff N. 1998.The biosynthetic pathway of vitamin C
in higher plants, Nature 393, 365-369.
Wolfender and Hostettmann 1996
Wolfender J-L.,Hostettmann K. 1996. Importance of LC/MS in plant analysis.
Spectroscopy Europe 8, 7-12.
Wolucka et al 2005
Wolucka BA,Van Montagu M.2003.GDP-mannose-3, 5-epimerase forms GDP-L-
gulose,a putative intermediate for the de novo biosynthesis of vitamin C in plants. J
Biol Chem 278,47483–47490.
Wishart et al. 2008
Wishart DS, Tzur D, Knox C, Eisner R ,Guo ACi ,Young N, Cheng D, Jewell K,
Arndt D , Sawhney S ,Fung C , Nikolai L , Lewis M , Coutouly M-A , Forsythe
I,Tang P , Shrivastava S, Jeroncic K, Stothard P,Amegbey G, Block D, Hau
DD, Wagner J, Miniaci J,Clements M, Mulu G, Guo N, Zhang Y,Duggan
GE, MacInnis GD , Weljie AM. , Dowlatabadi R ,Bamforth F, Clive D , Greiner R , Li
L , Marrie T, Sykes BD, Vogel HJ and Querengesser .2008. HMDB: the Human
Metabolome Database. Nucl. Acids Res. 35 (suppl1)
Xia et al. 2009
Xia J, Nick P, Young N, Wishart D.2009.MetaboAnalyst: a web server for
metabolomic data analysis and interpretation. Nucl Acids Res 37 (2), 652-660.
Xia and Saglio 1992
Xia JH and Saglio PH.1992.Lactic Acid Efflux as a Mechanism of Hypoxic
Acclimation of Maize Root Tips to Anoxia.Plant Physiol 100, 40-46.
Yamazaki 2003
Yamazaki Y, Urano A, Sudo H, Kitajima M, Takayama H.2003.Metabolite profiling
of alkaloids and strictosidine synthase activity in camptothecin producing plants.
Phytochemistry 62,461-70.
Yang et al.2007
Yang L, Deng Y, Xu S, Zeng X.2007.In vivo pharmacokinetic and metabolism
studies of ginsenoside Rd. J Chromatogr B Analyt Technol Biomed Life Sci 854,77-
84.
77
Zhao et al.2006
Zhao Q, Stoyanova R, Du SY, Sajda P, Brown TR.2006.HiRes- a tool for
comprehensive assessment and interpretation of metabolomic data. Bioinformatics
22(20), 2562-2564.
+
78