.1; I. Esau:

. .aikvéMrmwm

. h.

£94.” 4.3,“
a 1L. .. u

a.

=

.2 3.:
it an .3

.u

»

z...
.1...”
. 1%...
z. 51
“in.

 

 

 

 

, u .51..

.. £2.23. a... x

i i . .

1.31,...3wuifisnufif .9,
i Emit-2». :1.-: ill"
{SL3}: {SI I :3...
1... .m... E

. 3.

“Burn;

.4.
3L- 10>
Vt...

:5 1'
{av 33‘

«a

 

is...

 

 

LIBRARY
Michigan State
University

 

 

 

This is to certify that the
dissertation entitled

LIQUID CHROMATOGRAPHY/MASS SPECTROMETRY
BASED METABOLITE PROFILING IN REVERSE
GENETIC INVESTIGATIONS OF WOUNDIN G AND
PATHOGEN STRESS RESPONSES IN ARABIDOPSIS
THALIANA

presented by

Xiaoli Gao

has been accepted towards fulfillment
of the requirements for the

PhD. degree In Genetics

 

 

roam

Major Profié or’ s Signature

,1? MQFJOO?

Date

MSU is an Affirmative Action/Equal Opportunity Employer

 

— -.-._.—.-._.—.-c—-u-

 

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

5/08 K:lProj/Acc&Pres/ClRClDateDue.indd

 

 

LIQUID CHROMATOGRAPHY/MASS SPECTROMETRY BASED
~ METABOLITE PROFILING IN REVERSE GENETIC INVESTIGATIONS 0F
WOUNDING AND PATHOGEN STRESS RESPONSES IN ARABIDOPSIS
THALIANA
By

Xiaoli Gao

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY
Genetics

2009

 

ABSTRACT

LIQUID CHROMATOGRAPHY/MASS SPECTROMETRY BASED METABOLITE
PROFILING IN REVERSE GENETIC INVESTIGATIONS OF WOUNDING AND
PATHOGEN STRESS RESPONSES IN ARABIDOPSIS THALIANA
By

Xiaoli Gao

Investigations of gene functions have relied on two strategies known as forward
genetics and reserve genetics. Both strategies are most powerful when phenotypes are
described explicitly, and global phenotype profiling has emerged as a set of vital
methodologies for ftmctional genomics. This dissertation presents development and
application of liquid chromatography/mass spectrometry (LC/MS) methods for
metabolomic profiling for investigations of plant responses to wounding and pathogen
suess. In the first study, liquid chromatography/time-of-flight mass spectrometry was
used for nontargeted profiling of metabolites in leaves of wild type and 0-
methyltransferase (OMT) knockout mutants. 0-methyltransferase (OMT) methylates the
hydroxyl groups on phenols and produces methylated products. One of these mutants,
omtl, exhibited metabolic phenotypes distinct from wild type plants, with notable
accumulation of S-hydroxyferuloyl malate in the mutant. After inoculation with pathogen
Pseudomonas syringae DC3000, wild type plants exhibited five-fold greater bacterial
counts relative to the amt! mutant. Profiling of metabolites in control and pathogen-
infected omtI mutant and wild type plants pointed to several aspects of stress
biochemistry associated with this specific mutation. Most notable was evidence of

conversion of S-hydroxyferuloyl malate to a reactive quinone metabolite in the pathogen-

 

infected omtI mutant. The quinone metabolite was converted to glutathione (GSH)
conjugates in vivo in mutant plants. Further support for antimicrobial properties of the
quinone metabolites came from in vitro screening of synthetic substances against P.
syringae. Inhibition of bacterial growth was observed at low concentrations for 5-
hydroxyferulic acid, its quinone, and the glutathione conjugate of the quinone, but not for
the methylated analog Sinapic acid. Though glutathione conjugation of quinones has
been considered a detoxification step, the persistence of antagonism to P. syringae
growth with the quinone-glutathione conjugate suggests this class of metabolites may
undergo redox cycling, catalyzing formation of superoxide and hydrogen peroxide, as has
been observed in studies of quinone toxicity in animals. Though some polyphenols have
been documented in earlier studies to exhibit antimicrobial properties, roles for specific
metabolites or metabolic genes have remained elusive. Furthermore, the amt] mutant
displayed a more pronounced initial burst in levels of jasmonic acid, suggestive of
crosstalk between polyphenol metabolism and jasmonate signaling.

The second part of this dissertation research has focused on development of
ultraperformance LC (UPLC)-MS/MS methodology for rapid screening of multiple
phytohormones in a single analysis. A fast, sensitive, and selective UPLC MS/MS
method was developed for quantifying an assortment of jasmonic acid metabolites
including bioactive amino acid conjugates, and an assortment of related phytohormones
including salicylates, auxin, abscisic acid, and other oxylipins. These experimental
approaches have been applied to study interactions among different metabolic pathways

involved in gene regulation and signal transduction in wounding and pathogen stress.

.- l'- " .'_E.J

 

Copyright by
Xiaoli Gao
2009

 

Dedicated to my beloved: my husband and my parents

 

ACKNOWLEDGMENTS

I would like to express my earnest gratitude to my advisor and mentor, Dr. A.
Daniel Jones for the guidance during these years. Dr. Jones has always been an admirable
and brilliant mentor to me, has Shown substantial support for my study and work with his
inspiration and encouragement, and has strengthened my determination in pursuit of my
dream. I still remember clearly the day when I went to his laboratory for the first time; he
was very supportive to a genetics student who really wanted to study metabolomics.
Luckily, I was accepted as one of his graduate students later. His persistent efforts and
patience have assisted me through the most challenging but rewarding experiences to
become a professional researcher. I will be grateful for his mentoring forever.

I also wish to show sincere appreciation to members of my guidance committee, Dr.
Christoph Benning, Dr. Gregg Howe, and Dr. Lee Kroos for their constructive comments
and insightful suggestions on my work. I have been extensively benefited from their
marvelous instruction and treasured their friendship during my entire doctoral study.

I am grateful for the Plant Science Fellowship and Dissertation Completion
Fellowship offered by the Michigan State University. I would like to thank Dr. Abraham
J. K. Koo, Dr. Weiqing Zeng, and other colleagues for the collaboration. I would also like
to thank the colleagues from Dr. Christoph Benning and Dr. Dean Della Penna’s groups
in Biochemistry second floor for their help.

I would like to express my special gratitude to Dr. Barbara Sears, the director of
Genetics Program, who is like a big mum for every student in our program. I also would
thank Genetics Program for enrolling me into Michigan State University, which gave me

such a remarkable experience in my life.

vi

 

I would like to acknowledge my friends and my colleagues in our laboratory who
always bring joy and happiness to me. I would also like to thank their friendships and
scientific interactions.

I owe countless thanks to my husband, Yanfeng Zhang, my parents, my parents-in-
law, and my fiiends for years of unconditional love and support. Their encouragement
and belief have pushed me to finish this work.

Finally, I would like to thank all of the people above. This work would not be

possible without criticism, suggestion, support, and help from them.

Vii

 

TABLE OF CONTENTS

 

 

 

 

LIST OF TABLES x
LIST OF FIGURES xi
ABBREVIATIONS xiv
CHAPTER 1

INTRODUCTION 1
1.1 The significance of plant disease control .................................................................... 2
1.2 Plant-pathogen interaction studies .............................................................................. 4
1.3 Responses of plant to wound stress .......................................................................... 12
1.4 Current knowledge and knowledge gaps pertaining to mechanisms of plant response
to stress and disease ........................................................................................................... 17
1.5 Metabolomics as a new tool to close the gap between gene annotation and gene
function... .. ................................................................................................... 29
1.6 Research Objectives .................................................................................................. 40
CHAPTER 2

REVERSE GENETIC INVESTIGATION OF FUNCTIONS OF 0-
METHYLTRANSFERASES IN ARABIDOPSIS T HALIANA RESPONSES TO
PSEUDOMONAS S YRINGAE INFECTION USING LC/MS FOR

 

NONTARGETED METABOLITE PROFILING 41
2.1 Introduction ............................................................................................................... 42
2.2 Materials and methods .............................................................................................. 50
2.3 Results and discussion .............................................................................................. 74
2.4 Conclusion and perspective .................................................................................... 145
CHAPTER 3

UPLC MS/MS ASSAY OF JASMONATES AND RELATED PHYTOHORMONES
FOR LARGES SCALE SCREENING OF PLANT METABOLIC

 

PHENOTYPES.... ............................................................................... 148
3.1 Abstract ................................................................................................................... 149
3.2 Introduction ............................................................................................................ 149
3.3 Materials and methods ............................................................................................ 153
3.4 Results and discussion ............................................................................................ 164
3.5 Conclusion .............................................................................................................. 188
CHAPTER 4

CONCLUSIONS AND PERSPECTIVES 189
4.1 Conclusion of thesis project .................................................................................... 190
4.2 Opportunities for future research ............................. ' ............................................... 1 91

viii

 

 

APPENDIX A. Levels and chemical structures of metabolites 195

 

REFERENCES 214

 

ix

LIST OF TABLES

Table 2.1 Affinities and activities of purified Arabidopsis recombinant OMTl towards
phenolic substrates ........................................................................................................... 108

Table 3.1 Mobile phase gradient in HPLC separation ..................................................... 161

Table 3.2 MRM transitions, optimized source cone voltages, collison cell voltages, and
analyte retention times ..................................................................................................... 179

Table 3.3 Matrix Effect (ME %), Recovery Efficiency (RE %), Process Efficiency (PE
%) of the three internal standards .................................................................................... 183

Table 3.4 Levels of phytohormones in extracts of control, wounded, and P. syringae
infected Arabidopsis leaves ............................................................................................. l 86

 

LIST OF FIGURES

Figure 1.1 Disease symptoms in Arabidopsis leaves caused by P. syringae DC3000

infection at third day after inoculation with 1 x 106 cfu/ml DC3000 ........................... 5
Figure 1.2 Model for the evolution of bacterial resistance in plants ................................... 9
Figure 1.3 Systemic signaling in the wound response ....................................................... 14
Figure 2.1 Proposed 0-methylation from 5-hydroxyferuloyl malate and sinapoyl malate

to 3,4,5-trimethoxycinnamoyl malate, based on work by Hanley ..................................... 44
Figure 2.2 Chemical structure of OPC:4 malate (C13H2607) with accurate molecular mass
of 354.1679 ........................................................................................................................ 45
Figure 2.3 Diagram of primer design for SALK T-DNA insertion lines .......................... 51

Figure 2.4 Identification of 4-methylthiobutyl glucosinolate (C12H23N0983) by accurate
mass measurement ............................................................................................................. 56

Figure 2.5 Identification of sinapoyl malate (C15H1609) based upon fragment ions from
multiplexed CID ................................................................................................................ 58

Figure 2.6 Fragmentation reactions for sinapoyl malate (C15H1609) in negative ion mass
spectra .- ............................................................................................................................... 59

Figure 2.7 Relative quantification of S-hydroxyferuloyl malate (SHFM) in two leaf
extracts ............................................................................................................................... 61

Figure 2.8 Phylogenetic tree with annotation of the sixteen OMT T-DNA insertion mutant
lines with 13 amt mutant genes in the phylogenetic tree of Arabidopsis OMT family. . ..75

Figure 2.9 PCA scores plot of metabolite data from wild type (WT) and 16 omt T-DNA
insertion mutants (MT) leaves measured using LC/MS and showing the first two

principal components ......................................................................................................... 76
Figure 2.10 PCR products separated on 1.2% agarose gel ................................................ 78
Figure 2.11 RT-PCR products separated on 1.2% agarose gel .......................................... 79
Figure 2.12 T-DNA insertion position for omtI gene (At5g54160) .................................. 80
Figure 2.13 Bacterial growth on the wild type and the amt] mutant plant leaves ............. 81

xi

 

Figure 2.14 Statistical analysis of LC/MS metabolite data (using Umetrics SIMCA-P 11.0
sofiware) from the wild type (w) and comtl mutant (m) leaves without stress. ............... 85

Figure 2.15 Relative quantification of (A) 5-hydroxyferuloyl malate (SFHM) and
sinapoyl malate (SM) (B) 5-hydroxyferuloyl glucose (SHFG) and sinapoyl glucose (SO)
in leaves of wild type (WT) and comtI (AT5G54160) mutant (MT) Arabidopsis thaliana

 

.................................................................................................................. 93
Figure 2.16 Proposed O-methylations catalyzed by OMT] .............................................. 94
Figure 2.17 Purification of OMT] ..................................................................................... 97
Figure 2.18 Molecular mass of OMTl determined by ESI-TOF MS and MaxEntl
spectrum deconvolution ..................................................................................................... 98
Figure 2.19 Size exclusion chromatography (with Superdex 75 column) Showed elution
of COMTI as 8015 kDa, consistent with a dimeric native state ..................................... 101
Figure 2.20 Structures of substrates tested for OMT] activity in this study ................... 103
Figure 2.21 Recombinant OMTl enzyme kinetics towards caffeic acid ......................... 106
Figure 2.22 Levels and chemical structure of SHFM (m/z 325.06) ................................ 114
Figure 2.23 Levels and chemical structure of SM (m/z 339.07) ............................ 115
Figure 2.24 Levels and chemical structure of SHFM-GS (m/z 630.13) .......................... 116
Figure 2.25 Levels and chemical structure of SHF M quinone-GS (m/z 628.11) ............ 117
Figure 2.26 Levels and chemical structure of GSH (m/z 306.08) ................................... 118
Figure 2.27 Proposed pathway for production of SHFM-GS and SHF M quinone-GS and
catalytic formation of superoxide and H202 via redox cycling ....................................... 119
Figure 2.28 Levels and chemical structure of SHFM-OH (m/z 341.06) ......................... 121
Figure 2.29 Levels and chemical structure of SHFM-OH-GS (m/z 646.12) ................... 124
Figure 2.30 Levels and chemical structure of coronatine (m/z 318.17) .......................... 128
Figure 2.31 Levels and chemical structure of camalexin (m/z 199.04) ........................... 130
Figure 2.32 Levels and chemical structure of JA (m/z 209.12) ....................................... 132

xii

 

Figure 2.33 Concentration dependence of antimicrobial potential of H202, 5-
hydroxyferulic acid quinone conjugate (SHFQ), 5-hydroxyferulic acid quinone with
glutathione conjugate (SHFQ-GS), 5-hydroxyferulic acid (SHFA), and Sinapic acid (S)
against P. syringae DC3 000 ............................................................................................ 136

Figure 3.1 Diagram for triple quatrupole and multiple reaction monitoring (MRM) ..... 152

Figure 3.2 Chemical structures of the 24 compounds and the internal standards used in

this study .......................................................................................................................... 154
Figure 3.3 Extracted ion MRM chromatograms of 27 standard compounds .................. 166
Figure 3.4 Quantification of JA lie in mixtures of J A Ile and JA Leu ............................ 173
Figure 3.5 MRM transition of m/z 291.2 > m/z 165.1 fi'om OPDA standard (3A) and

OPDA in pathogen stressed leave tissue (3B) ................................................................. 175
Figure 3.6 Distinguishing SA glucoside isomers through base hydrolysis ..................... 177

Figure 3.7 LC/MS detection and quantification of SA glucoside isomers before and after
base hydrolysis ................................................................................................................. 178

(Images in this dissertation are presented in color)

xiii

[13C6]-JA Ile
COMT

Cor

dJA

ESI

FT-ICR MS:
GC/MS
GSA

SI-IF

SI-IFG
SHFM
SHFMQ
SHFMQ-GS
SHFQ
SI-IFQ-GS
[2H5]-OPDA
1AA

IS

JA

ABBREVIATIONS
ABA abscisic acid
ANOVA analysis of variance
CCoAOMT caffeoyl-COA O-methyltransferase

[”Cd-jasmonyl-isoleucine

caffeic acid O-methyltransferase

coronatine

dihydrojasmonic acid

electrospray ionization

Fourier transform-ion cyclotron mass spectrometry
gas chromatography/mass Spectrometry

B-O-D glucosylsalicylic acid

5-hydroxyferulic acid

5-hydroxyferuloyl glucose

S-hydroxyferuloyl malate

S-hydroxyferuloyl malate quinone
5-hydroxyferuloyl malate quinone glutathione conjugate
5-hydroxyferulic acid quinone

5-hydroxyferulic acid quinone glutathione conjugate
[2H5]-12-oxo-phytodienoic acid

indole-3-acetic acid

internal standard

jasmonic acid

xiv

 

P Emu—:2 r

LA

LC/MS

MC

MeJA

OMT

OPDA

OPLS

PCA

PLS-DA

SA

SAG

SAM

SIM

SG

SGE

SM

UPLC-
MS/MS

WC

linolenic acid

liquid chromatography/mass spectrometry
mutant without any treatment

mutant control, with the infiltration of Mng buffer
methyl jasmonate

multiple reaction monitoring

mutant with the infiltration of P. syringae
nuclear magnetic resonance
O-methyltransferase

lZ-oxo—phytodienoic acid

orthogonal partial least squares

principal component analysis

partial least squares - discriminant analysis
salicylic acid

salicylic acid glucosides

S-adenosyl methionine

selected ion monitoring

sinapoyl glucose

salicylic acid glucose ester (acylglucoside)
sinapoyl malate

ultraperforrnance liquid chromatography-tandem mass
spectrometry

wild type without any treatment

wild type control, with the infiltration of Mng buffer

XV

 

wild type with the infiltration of P. syringae

xvi

CHAPTER 1

Introduction

1.1 The significance of plant disease control

As the world population grows, the world places increasing demands on natural
resources to sustain supplies of food, energy, and materials. Photosynthetic organisms, in
particular land plants, serve as the foundation for nearly all of the food supply for humans
and animals. Plants can convert energy from sunlight into chemical energy, and these
biochemical intermediates flow ultimately into macronutrients including carbohydrates,
proteins, and fats. In this manner, the survival of all animals including humans depends
on plants, but the ability of these resources to meet the demands of a growing human
population is threatened by diseases and challenged by the desire to use plants as a source
of bioenergy. Plant diseases are classified as infectious (or biotic) disease and
noninfectious (abiotic disease). Biotic diseases in plants are caused by biotic pathogenic
microorganisms, such as viruses, bacteria, fiIngi, protozoa, nematodes, and threats from
insect damage. Abiotic diseases are caused by environmental conditions, such as lack or
excess of nutrients, water or light, and the presence of toxic chemicals in air or soil
(Agrios 2005).

Productivity of diseased plants can show dramatic decreases relative to healthy plants,
and uncontrolled plant diseases result in economic losses or foods of poor quality that can
threaten human health (Agrios 2005). In the United States, annual crop losses have been
estimated at $9.1 billion lost to diseases, $7.7 billion to insects, and $6.2 billion to
competition by weeds. Therefore, crop protection becomes even more important than
increasing fertilization, irrigation, and other methods to avoid the losses (Agrios 2005).
Diseased plants may also contain poisons that make them unfit for consumption and

affect the environment. Many plant products, like grains, hay, and fruits are often

infected with fungi that produce toxic secondary metabolites known as mycotoxins
(Blumenthal 2004; Serra et a1. 2005). Animals and humans consuming such foods may
develop severe diseases of organs, the nervous system, and the circulatory system that are
sometimes lethal.

Improved control of plant disease can result in more food, better quality, lower
production costs, and less environment contamination. Therefore, finding
environmentally friendly ways to control plant disease serves as an important avenue of
research. Over the last 100 years, the control of plant diseases has depended increasingly
on the extensive use of pesticides, with associated risks of toxicity and environmental
contamination (Blain 1990; Mattsson 2000). Modern research in plant control aims at
finding other environmentally fi‘iendly approaches to reduce toxicity and negative
environmental impacts. The most promising approaches employ biological approaches
that enhance plant resistance to disease including conventional breeding, genetic
engineering of disease-resistant plants, RNAi gene-silencing to silence some genes, and
enhanced production of agents antagonistic to microorganisms that cause plant disease.
New biotechnological products are currently being developed based on stimulation of the
plant defense response, and on the use of plant-beneficial bacteria for biological control
of plant diseases (biopesticides) and for plant growth promotion (biofertilizers)
(Montesinos et al. 2002).

In this study, the long-term goal is to increase our knowledge about how plants might
better resist disease and use the knowledge to manipulate the combinations of plant
genetics with environment to better avoid or control those diseases. Specifically, I am

studying how plants generate biochemical responses to biotic stress (pathogen stress) and

 

abiotic stress (wounding stress) using Arabidopsis thaIiana-Pseudomonas syringae as a
model plant and pathogen system. To establish the context for my efforts, this section
explores and explains recent applications of LC/MS based metabolomics and systems

biology to the area of plant stress response.

1.2 Plant-pathogen interaction studies
1.2.1 Arabidopsis-Pseudomonas syringae model system.

In this study, we investigated plant-pathogen interactions using a model system based
on the model plant Arabidopsis thaliana and model pathogen Pseudomonas syringae. In
this model interaction, the genomes of both the host and the pathogen have been
sequenced, and plant-pathogen interactions are well-studied in this model System.
Arabidopsis thaliana is a small flowering weed of mustard family (Brassica) and has a
short life cycle of Six weeks (Meinke et al. 1998; Dennis and Surridge 2000). Its genome
is one of the smallest plant genomes and was the first plant genome to be sequenced in
2000 (Arabidopsis Genome Initiative, 2000). The genome of P. syringae pv. tomato
DC3000 has also been sequenced (Buell et al. 2003) (http://www.pseudomonas-
syringae.org/). P. syringae strains had been observed to be a natural pathogen of soybean,
tomato, and bean (Keen 1990). Examples of disease symptoms caused by P. syringae

DC3000 infection in Arabidopsis leaves are shown in Figure 1.1.

 

 

Figure 1.1 Disease symptoms in Arabidopsis leaves caused by P. syringae DC3000
infection at third day after inoculation with 1 x 106 cfu/ml DC3000. The infected leaves
were in yellow.

However, some researchers have not been convinced that Arabidopsis—P. syringae
provides a good model system for two reasons. First, P. syringae is not known as a
natural pathogen for Arabidopsis in the wild. Second, artificial inoculation methods,
either infiltration or use of surfactant, were required to develop infection in the laboratory
(Katagiri et al. 2002). Despite these disadvantages, there are substantial advantages that
make this a good model system, since no other systems have been studied for such a long
time with well-developed information on genomes and mechanisms of the pathogenesis
from the pathogen side and resistance from the plant side.

Current wisdom suggests that plant resistance to pathogens is correlated with the
activation of a network of broad-spectrum defense mechanisms. The response involves
initial bacterial invasion, recognition of the pathogen effectors, opening of ion channels,
signal transduction, activation of protein kinases, transcriptional activation of defense-

related genes, and activation of preformed enzymes to modify primary and secondary

 

metabolism (Buchanan et a1. 2001). The most frequently encountered defense and
Signaling mechanisms are described below to give an overall picture how plants respond
to pathogens. These include stomata-based defense against bacterial invasion, two-
branched innate immune system: nonhost resistance (basal defense) and pathogen-
specific resistance (gene-for-gene model) (Jones and Dang] 2006), and systemic acquired

resistance (SAR).

1.2.2 Stomata-based defense against bacteria] invasion

Stomata, which are microscopic surface openings on the leaf, provide the first
entrance for microbial invasion of plants. They serve as the critical first step for bacteria
to cause infection. Stomata have long been assumed to be passive ports of bacterial entry,
but recent studies in the laboratory of Sheng Yang He at Michigan State University
found that stomatal closure is part of the plant immune response and acts as an innate
immrmity barrier to restrict bacterial invasion. However, some P. syringae strains
produce a polyketide toxin, coronatine, that triggers reopening of stomata, and

circumvent this host defense (Melotto et al. 2006; Thilmony et al. 2006)

1.2.3 Innate immune system: nonhost resistance (basal defense)

When bacteria invade plant tissues into the apoplast through the stomata, the innate
immune system presents the first line of defense. A key function of innate immunity is
recognition of pathogen-associated molecular patterns (PAMP) produced by infectious
agents such as bacterial flagellin. Such recognition events activate a cascade of mitogen-

activated protein kinases (MAPKs) and production of WRKY transcription factors that

 

 

confer resistance by altering gene expression (Mizoguchi et al. 1997; Asai et al. 2002;
Chishohn et al. 2006). A diagram representing these processes is shown in Figure 1.2A
(Chisholm et al. 2006). Alterations of plant cell walls at infection sites and deposition of
electron dense callose-rich papillae have also been reported as a major component of

non-host resistance (Truman et al. 2006).

.8385
88 8888 83 .683 .8880 .338 .80 .8588 a .m as 80 .m .880 .0 .8 .m .8880 .38 .888...
08.88 :88 05 mo 80:30.6 05 3838 ”8288088 3088-828 808 00888008 .Goom 8 8 88828 88:08.8 0888M
80mm8¢08080 .«0 82888 038080 8 82.8 82000 8.883 80 888383008 @0885 8880888 0888_ “0080388080880
888 008888 0888 08 .3388 8808.8 088008 888008 080m 5 880808 008888 883 0% 888% 880 8.8 830.8
8.883 3528 £88088 0888 :83 80888 8 880808 83 085:8 8mg 88 A8808 080w 83 880808 88080 .8383 8
Amp—.5 808.8 88888 E 083 05 8: 8.883 088388 Am: .8088 8088888.: >E 8308 8888880800 88888888
88 8008880 8880— 8.32 8:88 858me 808388 :08? 0.8888 :83 8035 38808 8:08: 88083 3 0880 82.8%
8088 8800—08 00888880888 80 80880038 A3 .888 8 008888 88083 .8 88:05 08 .88 8—002 «4 08mm...—

 

880%“ 308080304

 

88008 _
8802088808

89880 C

8883 0880.032 .

888M 0

 

 

m08<h2wm¢

88:083—
088m
088th
-880tm

mmdumziz.

 

woS—>So< Houoommm
088000m 880808 m

 

. ..r

33.0.2303

          

5838:: 8.088%
8088.5 mm:

fwd.-a:;1 ....

...!

 

 

u02<h2wuc

moan—080m

0888~
008mmuh [I ll

dug/E

 

 

5888: 888:.

82888 .828 l

 

 

1.2.4 Innate immune system: pathogen-specific resistance (Gene-for-gene model)
The first well-developed pathogen-specific resistance mechanism is described as the
gene-for-gene model (Flor 1955), the specific recognition of the resistance gene (R gene)
product in the plant by the pathogen avirulence (avr) gene product. Afier pathogenic
bacteria enter plant tissue through stomata or wounding, they proliferate in the apoplast
(the intercellular spaces between plant cells) (Jones and Dangl 2006). The Avr products,
specific virulence proteins, are secreted across both the inner and outer bacterial
membranes and get into the cyt0plasm of the host plant cell. The secretion mechanism
utilized by the pathogen, termed a type III secretion system (TTSS), involves a protein
complex that forms a needle-like structure that spans both the inner and outer bacterial
membranes and allows bacterial proteins to enter plant cells. The major function of these
effector proteins (avr gene products) is to suppress the host innate immune response,
allowing pathogens to grow and cause disease (Chisholm et al. 2005; Chisholm et al.
2006). The mechanism is shown in Figure 1.2B. The R proteins, operate as receptors and
signal transducers. They first recognize any Avr gene-dependent ligand and then activate
downstream signaling that leads to rapid induction of various defense responses
(Buchanan et al. 2001). The mechanism is shown in Figure l.2C. This recognition sets
off a hypersensitive response (HR), a rapid and localized cell death with hydrogen
peroxide concomitantly produced which effectively prevents further spread of the
pathogen. The HR is generally recognized by the presence of brown and dead cells at the
infection site and by “defense gene” expression (Heath 2000). Two mechanisms underlie

HR formation. Either the attacked cell initiates a regulated cell death program (apoptosis)

10

 

or the responding cells are rapidly poisoned by the toxic compounds and free radicals

they have synthesized, and thus die as a result of necrosis.

1.2.5 Systemic Acquired Resistance (SAR)

Within minutes of pathogen attack, plant defense responses are activated locally (as
described above). Within hours, defense responses are sometimes activated in tissues far
from the invasion site and even in neighboring plants to protect non-infected tissues from
secondary infection. This mechanism of induced defense that confers long-lasting
protection against a broad spectrum of microorganisms is called systemic acquired
resistance (SAR) (Durrant and Dong 2004). SAR in plants is analogous to the innate
immune system found in animals, in that both produce antimicrobial substances such as
peptides and metabolites in response to infection (Traw et al. 2007). The combination of
accumulation of the phytohorrnone salicylic acid (SA) and systemic activation of a
specific subset of pathogenesis related (PR)-type genes is essential for the expression of
SAR. (Hunt et al. 1996; Fobert and Despres 2005). This SA-mediated signal transduction
cascade is modulated by interactions between a redox-sensitive protein known as the non
expressor of pathogenesis-related genes 1 (NPRl) and the TGA family of transcription
factors. These interactions regulate expression of an assortment of PR defense genes.
The mechanisms underlying the influence of cellular redox state are not fully established,
but it has been proposed that hydrogen peroxide may serve as a second messenger of SA.
(Lebel et al. 1998; Zhang et al. 1999; Fobert and Despres 2005; Weigel et a1. 2005; van

Verk et al. 2008). Some PR proteins are chitinases and glycanases, enzymes that degrade

ll

structural polysaccharides of pathogen cell walls and may reduce bacterial growth
(Metzler et al. 1991; Rogers and Ausubel 1997).

The mechanisms of plant activation of defenses against pathogens using Arabidopsis
thaliana-Pseudomonas syringae as a model system were reviewed above. However, the
extent to which genetic diversity in plants contributes to resistance to pathogens remains
a body of knowledge with several gaps. To understand exactly how these defense
mechanisms work, we need to broaden our knowledge about functions of plant genes
including aspects of protein localization, protein-protein interactions, endogenous
substrates for enzymes and transporters, and the biological roles of these genes in plant

interactions with microbes and other organisms.

1.3 Responses of plants to wound stress

Introduction of pathogens to plant tissues is often accompanied by mechanical
wounding. Such wounding is an important aspect of plant responses to herbivory by
insects. To dissect the contributions of plant responses to various stressors, plant
responses to mechanical wounding have been investigated in detail.

Leaf wounding is well characterized for its stimulation of rapid activation of jasmonic
acid (JA) synthesis (Weber et al. 1997; Schilmiller and Howe 2005; Chung et al. 2008;
Glauser et al. 2008; Koo et al. 2009), and IA is widely considered as the main
phytohorrnone produced as a wound response and an important regulator of gene
expression following wounding. The precursors of JA, linolenic acid (C183) and its
metabolite lZ-oxo-phytodienoic acid (OPDA), are released from plastid membrane

galactolipids by the catalytic action of lipases (Ishiguro et al. 2001; Buseman et al. 2006;

12

Kourtchenko et al. 2007; Hyun et al. 2008), and some basal levels of OPDA may be
present prior to wounding. Conversion of OPDA and related intermediates to JA is
believed to occur in the peroxisome, first through reduction to the intermediate OPC28,
followed by chain ‘shortening by B-oxidation. JA can be further metabolized to assorted

derivatives such as JA He and MeJA (Figure 1.3) that differ in biological activity.

13

 

 

18:2
FAD3 FAD7
l EAQB.
18:3 2
FACs(+) L LOX g
— E
AOS
OGAs(+/—) y A—O C
Systemin (+) OPDA
Eth lene +/— 9553
y ( ) V AC5 :1)
ABA (+) OPCB-CoA E
8
5
o.

 

 

t3?

HOOC ,

 

Local and systemic
gene expression COR

 

Current Opinion in Plant Biology

 

 

Figure 1.3 Systemic signaling in the wound response (Schilmiller and Howe 2005).
Reprinted from Systemic signaling in the wound response, 8/4, Schilmiller, A. L. and G.
A. Howe, Current opinion in plant biology, 369-377, Copyright (2005), with permission
from Elsevier.

14

While JA and related oxylipins have attracted much attention for their role in plant
responses to wounding, the early events that plants use to recognize and generate
biochemical responses to wounding have remained a subject of intensive study and
debate. Release of linolenic acid from membranes following wounding has been
proposed as one of the early steps leading to increased J A levels following wounding.

Hydrolysis of membrane lipids to generate signaling molecules other than JA has
been reported as a consequence of wounding in plants. Accumulation of phosphatidic
acid (PA) has been observed in leaves following wounding (Lee et al. 1997), and PA is
now recognized to activate an assortment of signaling pathways in plants. Though it has
been proposed that wound-activated phospholipase D (PLD) could catalyze formation of
PA upstream of JA biosynthesis (Wang et al. 2000), a recent study of pld mutants showed
no significant differences in IA or related oxylipin levels following wounding (Bargmann
et al. 2009).

Early events in responses to wounding have been associated with release of calcium
from damaged cells, membrane depolarization, formation of systemin peptides, and
activation of mitogen activated protein kinase (MAP kinase) activity (Ryan 2000).
Within minutes, measurable increases in levels of IA and its amino acid conjugate,
jasmonoyl-isoleucine (JA Ile), can be observed, and these jasmonates play important
regulatory roles. Most members of the JASMONATE ZIM-domain (JAZ) transcriptional
repressors were highly expressed in mechanical wounding within 5 min of mechanical
tissue damage, coincident with a large rise in IA and JA Ile levels (Chung et al. 2008). JA

and J A Ile activate wound-induced defense responses by promoting SCFCOIl-dependent

15

degradation of JAZ proteins that repress the expression of early response genes (Chung et
al. 2008; Katsir et al. 2008; Koo et al. 2009).

The jasmonic acid (JA) pathway in plant responses to wounding is regulated by other
signaling molecules which have been proposed to play important roles in wound
signaling as well (Figure 1.3). Wounded plants also accumulate abscisic acid (ABA)
(Birkenmeier and Ryan 1998), ethylene (O'Donnell et al. 1996), oligosaccharides (Doares
et al. 1995), and the oligopeptide systemin (Constabel et al. 1995; Howe and Ryan 1999;
Ryan 2000). Wound signal transduction cascades also involve secondary messager
calcium which is released from damaged cells (Knight et al. 1993), fatty acid conjugates
(FACs) (Ryu and Wang 1998), and hydrogen peroxide (Orozco-Cardenas and Ryan
1999; Guan and Scandalios 2000; Orozco-Cardenas et al. 2001). Nitric oxide (NO) may
down regulate the expression of wound-inducible defense genes during pathogenesis
(Orozco-Cardenas and Ryan 2002). Besides jasmonate-dependent signal transduction
cascades, jasmonate-independent cascades also exist including gene expression activated
by oligogalacturonides (Titarenko et al. 1997; Leon et al. 1998). Coronatine, a microbial
metabolite and structural mimic of IA, exerts its effects by activating the JA singaling
pathway through COIl. A summary of wound-induced signaling is shown in Figure 1.3
(Schilmiller and Howe 2005).

Tissue damage is associated with decompartrnentalization, release of cellular
contents, and a loss of water, similar to those occurring in water-stressed intact plants
(Reymond et al. 2000). Therefore, wounding shares many biochemical consequences

with osmotic stress (Denekamp and Smeekens 2003). The overlap was also observed

16

between wounding and pathogen response (Cheong et a1. 2002), since mechanical

wounding not only damages plant tissues, but also facilitates pathogen invasion.

1.4 Current knowledge and knowledge gaps pertaining to mechanisms of plant
responses to stress and disease
1.4.1 Genetics in plant response to stress and disease: plant gene function studies
Association of individual plant genes with specific functions serves as a key step
toward development of plants with improved productivity and resistance to pathogens
and other environmental stresses. In order to determine biological function of a gene, a
well-used genetics approach investigates the phenotypic effects of the perturbation of the
gene by introducing genetic alterations. It is helpful to compare two genotypes: the wild
type that can express a functional gene, and a mutant that lacks this gene. Identification
of mutants and their corresponding phenotypes is achieved either by ‘forward’ or
‘reverse’ genetics (Alonso and Ecker 2006). Forward genetics involves working from a
phenotype to identify a corresponding gene and its functions, following the approach
developed by Mendel, the father of genetics, in his legendary experiments on garden pea
plants in the mid-19th century. It is the traditional approach where groups of randomly
generated mutants are screened based on their phenotype, and the gene responsible for
the phenotype is identified by gene mapping (J ander et al. 2002). The advantage of
forward genetics is that no prior assumptions need be made about the mutant genes and
phenotypes, making this unbiased approach powerful for identifying roles for genes of
unknown fimctions. Forward genetic approaches encounter challenges in gene mapping

and discovery, particularly for those plants for which genomic sequence information is

17

 

limited. Reverse genetics approaches explore gene functions by working from specific
mutant genes to identification of phenotype. Mutants with the depletion of specific genes
are analyzed with a number of phenotypic assays (McCallum et al. 2000). The great
advantage of reverse genetics is that more facile association of mutant phenotype with the
affected gene can be achieved, and a broader array of phenotypes can be measured
against mutants than in a forward genetics screen (Lahner et al. 2003; Messerli et al.
2007). A major disadvantage of the reverse genetics approach arises because most
common phenotypic assays of mutants lacking specific genes have failed to associate
functions with those genes. As a result, large-scale phenotypic profiling of numerous
mutants may be necessary for gene function discovery.

In the model plant Arabidopsis thaliana, it has been estimated that the genome codes
for about 25,000 genes (Poethig 2001). The Arabidopsis 2010 program, funded by the US
National Science Foundation, has a stated aim of identifying all Arabidopsis gene
functions by the year 2010. However, at this date, many genes remain with unidentified
functions. Approximately one thousand predicted genes encode proteins assumed to
function in secondary metabolism in Arabidopsis (von Roepenack-Lahaye et al. 2004).
Sequence similarities have suggested >250 cytochrome P450, >100 acyl transferase,
>300 glycosyl transferase, >300 glycoside hydrolase genes, and a large number of other
putative genes encoding enzymes such as dioxygenases, O-methyltransferases, terpene
synthases, or polyketide synthases. Compared with the huge number of these genes, the
number of known reactions catalyzed by these types of enzymes is relatively low in

Arabidopsis.

18

A large number of metabolites have yet to be identified as well (von Roepenack-
Lahaye et al. 2004). Despite well-developed approaches of forward and reverse genetics,
why is it so difficult to determine the functions of these genes? The answers come first
from realization that gene annotation is most often based on similarity of DNA sequence
to a gene of known function. However, sequence similarities do not always lead to
identical functions, and single mutations can have dramatic effects on protein functions.
Second, redundant genes often display no overt phenotypes. When a given biochemical
function is redundantly encoded by two or more genes (so called genetic redundancy),
mutations in one of these genes may have a small or negligible effect on the loss of
phenotypes. Therefore, it is challenging to establish the function of a single gene with
overlapping functions with other genes. Third, regardless of whether a study pursues
forward or reserve genetic approaches, it is essential that the phenotypes be described as
explicitly as possible, so that the functions of genes responsible for the phenotypes can be
assigned accurately. The molecular phenotypes include all the levels of mRN As, proteins,
and metabolites. More commonly, however, the approaches that have been used to study
cellular responses have focused on the transcript or the protein levels (transcriptomics
and proteomics, respectively) because technologies and databases have been more
advanced in these areas than for metabolites.

The traditional forward genetic and reverse genetic strategies have suffered from the
interrogation of each mutant with a limited number of phenotypic assays, typically
targeting only one or a limited number of pathways. This approach limits the likelihood
that the full effects of a mutation will be discovered, and hinders researchers from

discovering unexpected relationships between genes. Examples of comprehensive

19

screening of biochemical phenotypes have been few until recent years. In one prominent
and recent study, previously characterized Arabidopsis mutants that have primary
physiological defects were shown to have unexpected secondary phenotypes and
syndromes based upon parallel assays of metabolites, physiology, and morphology (Lu et
al. 2008). An important aspect of the novelty of this study lies in the depth of phenotypic
screening of large numbers (~ 100) of mutants, and comparisons of numerous phenotypes
including levels of numerous fatty acids and 25 amino acids and related metabolites. The
findings of this study suggest that greater efiiciencies in gene fimction discoveries result
from parallel measurements of multiple phenotypes in a single study.

In view of these findings, this dissertation emphasizes development and application of
approaches capable of profiling metabolites on a global level to accelerate the precise
linkage of genes to their functions in plants. Most investigations of this kind have been
centered on measurements of biochemical phenotypes under control environmental
conditions, but the research described below aims to extend metabolite profiling to

investigate gene functions in plant responses to stress.

1.4.2 Biochemistry of plant responses to stress and disease
1.4.2.1 Stress and metabolic homeostasis

Stress in a plant can be defined as any change that requires adjustments to metabolic
fluxes to sustain homeostatic control (Shulaev et al. 2008), and these adjustments are
usually referred to as acclimation (Mittler 2006). To thrive despite an assortment of
environmental stresses, plants must be able to sense changes in environmental conditions

and respond to these changes through regulation of gene expression and protein activity.

20

Because plants are immobile, they cannot avoid stress by moving to less stressful
locations. Their successful resistance to pathogens depends on their ability to detect
pathogens and generate biochemical responses that provide a defense against pathogens.
These biochemical defenses will be described in greater detail below. Pathogen detection
by plants is followed by a complex system of signal transduction events that ultimately
result in dramatic changes in profiles of gene expression (Thilmony et al. 2006) and
metabolites. Since metabolite levels change in response to levels and activities of protein
gene products including enzymes and transporters, metabolic responses often lag behind
changes in gene expression. Plant responses to pathogens often involve changes in
expression of hundreds of genes and levels of metabolites, most of which remain
unidentified and unmeasured. As a consequence, attribution of resistance functions to
individual metabolic genes or metabolites remains challenging, and it has been difficult
to prove roles for individual metabolites in resistance. One group of specialized
metabolites, the phytoalexins, is defined as a group of antimicrobial metabolites
synthesized by plants in response to pathogen infection, and these will be discussed in
more detail below.

The plant kingdom produces a diverse suite of specialized secondary metabolites and
other natural products that can confer disease resistance (Dixon 2001) including
numerous metabolites with antimicrobial properties (Cowan 1999). Most specialized
metabolites have functions that often remain unclear, although their modulations are
likely to be the major result of the numerous induced changes in transcriptional activity
occurring in infected plant cells (Scholthof 2001). Plant responses to pathogen infection

and other environmental stress often involve increased production of secondary

21

metabolites (Bednarek and Osbourn 2009), but the extent of these changes has not been

addressed in a comprehensive manner.

1.4.2.2 Classes of secondary metabolites associated with plant stress responses
Based on current knowledge of secondary metabolites involved in plant responses to
pathogen stress, the metabolites can be classified based on their dynamics and functions

in plant response and their chemical structures.

Metabolites with different dynamics in plant responses to stress

Several stages are thought to be involved in acclimation of plants to stress. In the
initial stages, the change or attack is sensed by the plant, and a network of signaling
pathways is activated. These signaling cascades involve changes in gene expression,
protein function, and metabolite levels. Details of the early mechanisms in conversion of
molecular stress to a biochemical signal are often unclear. In later stages, the activated
signal transduction pathways trigger the production of different genes, proteins and

metabolites that restore or achieve a new state of homeostasis (Mittler et al. 2004).

Metabolites with different functions in the plant response to stress

At least three different types of metabolites with different functions have been touted
as important for acclimation: (1) compounds involved in the acclimation process such as
antioxidants including ascorbic acid, glutathione, tocopherols, anthocyanins, and
carotenoids, or osmoprotectants including the polyols mannitol and sorbitol, betaines and

dirnethylsulfonium compounds, and the amino acids proline and ectoine (Shulaev et al.

22

2008); these compounds offer protection against damage to cellular functions in the form
of inactivation of reactive intermediates including free radicals, maintenance of cellular
redox state, and stabilization of three-dimensional protein structures; (2) metabolic by-
products of stress that result from the disturbance of normal homeostasis and resulting
changes in gene expression and protein functions; these include compounds with
suspected antimicrobial properties, classified as phytoalexins (Hammerschmidt 1999);
and (3) signal transduction molecules involved in mediating the stress response (Shulaev
et al. 2008).

The production of phytoalexins such as the indole alkaloid camalexin in Arabidopsis,
tomatine (steroidal glycoalkaloid) in tomato, and avenacin (triterpene glycoside) in oat
are phytochemicals with antimicrobial properties (Bednarek and Osbourn 2009). In 1994,
the first report of a mutant Arabidopsis plants deficient in camalexin exhibited enhanced
sensitivity to P. syringae, but one camalexin-deficient mutant (pad3) exhibited behavior
in pathogen growth similar to wild type plants (Glazebrook and Ausubel 1994). The
study was unable to reach definitive conclusions about the role of camalexin in these
plant-pathogen interactions, having shown that levels of camalexin alone did not explain
differences in pathogen susceptibility, but suggested that another metabolite in the
camalexin biosynthetic pathway might show similar antimicrobial effects. Such findings
point to the drawbacks of attributing antimicrobial fimctions to a single targeted
metabolite.

One unanticipated finding, reported in 1988, identified the tripeptide glutathione
(GSH) as a metabolite that stimulated induction of plant defense genes in cultured bean

cells (Wingate et al. 1988). While GSH has long been recognized as a redox buffering

23

metabolite, mechanistic understanding has yet to explain these observations. One of the
camalexin-deficient mutants (pad2) described above was recently used to demonstrate
that the PADZ gene codes for y—glutamylcysteine synthetase, a key enzyme in GSH
biosynthesis. The padZ-I mutants showed enhanced susceptibility to additional
pathogens, and these findings suggested an important, but mechanistically undefined role
for GSH in pathogen resistance (Parisy et al. 2007).

A variety of studies have suggested that signaling metabolites involved in plant
responses to pathogens could be newly synthesized or released from their conjugated
forms. These substances could be by-products of stress-induced metabolism (similar to
the second type of metabolites described above), degradation of membrane lipids and
release of precursors of bioactive metabolites, an assortment of reactive oxygen species
(ROS), oxidized metabolites such as oxidized phenolic compounds, or even some
antioxidants (Mittler 2002). Salicylic acid (SA), methyl salicylate (Me SA), and abscisic
acid (ABA) are induced and involved in the signal transduction for activating systemic
defense and acclimation response under pathogen stress (N awrath et al. 2002; Guo and
Stotz 2007; Kachroo and Kachroo 2007). Jasmonic acid (JA), JA amino acid conjugates,
methyl jasmonate (Me JA), 12-oxophytodienoic acid (OPDA) and other oxylipins and
their conjugates that are produced as a result of wounding stress can also serve as
signaling molecules. Crosstalk among those signals makes the study of signaling
mechanisms more challenging (Gupta et al. 2000; Gutjahr and Paszkowski 2009), and it
is in these studies where reverse genetics approaches offer prospects for identifying roles

of specific individual genes.

24

Classes of metabolites important in plant stress response

Numerous classes of specialized plant metabolites have important roles in plant
responses to stress including pathogen infection. Polyphenols are some of the most
abundant secondary metabolites in plants and their induced production indicates change
in relevant metabolic steps in plants under pathogen stress (Dixon and Paiva 1995). In
addition, glucosinolates, oxylipins, indole conjugates, terpenes, and alkaloids are also
thought to play important roles in plant responses to pathogens (Kliebenstein 2004), and
these are discussed in greater detail below.

Polyphenols are specialized metabolites defined by the presence of at least one
phenolic hydroxyl group per molecule. They are generally subdivided into tannins and
phenylpropanoids such as lignins and flavonoids that are derived from simple
polyphenolic units produced by the shikimate pathway (Dewick 1995). Polyphenols are
important to the plant cell and have been assigned functions as phytoalexins, UV
sunscreens, pigments, signaling molecules, and are precursors of major structural
components including lignin and hemicelluloses (Winkel 2004). Polyphenols are also
potent inhibitors of free radical oxidations owing to their ease of forming phenoxy
radicals. Oxidative burst is a phrase that describes the rapid generation of superoxide and
hydrogen peroxide that follows initial recognition of pathogen infection. Many
polyphenols are potent scavengers of oxidants that are generated during this process, and
may be oxidized to various forms by nonenzymatic and enzyme-catalyzed reactions with
such oxidants.

Polyphenols are derived fi'om the amino acid phenylalanine, which is transformed

into a variety of important secondary products, including lignin, sinapate esters, stilbenes,

25

and flavonoids. Flavonoids are derived from a C15 skeleton, which is synthesized via the
condensation of p-coumaroyl-CoA and three molecules of malonyl-CoA (Harbome
1988). Flavonoids occur widely in plants and have diverse subgroups that can be divided
into anthocyanins, flavonols, flavones, flavanones, chalcones, dihydrochalcones and
dihydroflavonols. (Treutter 2005; Treutter 2006). The distribution of specialized
metabolites among these forms depends on the presence and expression of metabolic
enzymes that catalyze key steps in their biosynthesis. The induced formation of
flavonoids after injury by pathogens or pests is a well-known phenomenon (Barry 2002;
Gallet 2004). The phenolic unit can often be modified through esterification,
glycosylation, methylation, and in some plants, prenylation, and these modifications can
influence reactivity and sequestration. Dimerization or further polymerization can create
additional classes of polyphenols.

Some polyphenols are reported to be antifeeding compounds that inhibit herbivory by
insects and other animals. However, roles of most polyphenols in microbial pathogen
infection have often been postulated, but have remained uncertain (Harbome 1994;
Treutter 2005).

Glucosinolates are also well-characterized antimicrobial and anti-insect metabolites
(Grubb and Abel 2006; Mewis et al. 2006) that may exert their antimicrobial activity
after metabolic conversion to reactive isothiocyanate metabolites. Glucosinolates share a
common core structure of a fi-D-thioglucose group linked to a sulfonated aldoxirne
moiety and variable aglycone side chains (Fahey et al. 2001). In A. thaliana, the source of
the side chain defines three major classes of glucosinolates: aliphatic glucosinolates are

derived principally from methionine via elongation of the alkyl chain, indolyl

26

glucosinolates are from tryptophan, and aromatic glucosinolates are derived from
phenylalanine (Mewis et al. 2006). It was reported that aphid attacks increased indole-
glucosinolate levels and suppressed methylsulfmylpropyl and propenyl-glucosinolate
synthesis (Kusnierczyk et al. 2008). The metabolite 4-methylsulphinylbutyl
isothiocyanate has been shown to have antimicrobial activity against DC3000 in vitro
(Tierens et al. 2001). This activity may result from enzymatic transformation of
glucosinolates to isothiocyanates, which are electrophilic intermediates reactive toward
nucleophiles such as certain proteins and low molecular weight thiols.

Biosynthesis of secondary metabolites following pathogen infection and stress is
ofien regulated by oxylipin metabolites that are central regulators of plant gene
expression, including genes involved in responses to stress and pathogen infection.
Among oxylipins, jasmonic acid (JA), its methyl ester (MeJA) and 12-oxophytodienoic
acid (OPDA) are important mediators of signal transduction (Browse 2005; Koo et al.
2009). A growing research effort focuses on the roles of oxylipin conjugates including
glycolipid conjugates and bioactive amino acid conjugates of jasmonates in signal
transduction. Much work is underway and is needed to improve understanding of the
roles of oxylipin metabolites in control of plant responses to stress (Staswick and Tiryaki
2004).

Accumulation of soluble and wall-bound indolic metabolites in Arabidopsis leaves
has also been found to follow infection by Pseudomonas syringae (Hagemeier et al.
2001). Most prominent among the accumulating soluble substances were tryptophan, B-
D—glucopyranosyl indole-3-carboxylic acid, 6-hydroxyindole-3-carboxylic acid 6-0—B-D-

glucopyranoside, and the indolic phytoalexin camalexin.

27

Terpenes comprise another class of abundant specialized metabolites; many are
volatile compounds that give plants their odors, and some are chemical insect attractants
and repellents. They are based upon isoprene (Cng) building blocks, with an enormous
number of isomers possible within the class. Diterpenes (C20), triterpenes (C30) and
tetraterpenes (C40), hemiterpenes (C5), and sesquiterpenes (C15) are commonly found in
plants. Terpenoids are terpenes with the addition of oxygen-containing functional groups
(Cowan 1999). Within the Arabidopsis genome, over 30 genes have been identified that
potentially encode terpene synthases (TPSs) (Chen et al. 2004). Volatile terpenoids have
been implicated in plant defenses, particularly toward herbivory by insects, but
demonstration of their direct role in pathogen resistance has remained elusive.
Attribution of roles to specific terpenoids is complicated by concomitant changes in
numerous other metabolites during the course of pathogen infection. In the case of
volatile terpenoids, levels in Arabidopsis are considered low compared to other plants. A
possible role of terpenoid metabolism was found in the induced defense of A. thaliana
plants against P. syringae. Elevated emission of the volatile terpenoid (E,E)-4,8,12-
trimethyl-l,3,7,1l-nidecatetraene (TMTI') was suggested to be a by-product of activated
J A signaling, rather than an effective defense response that contributes to resistance
against P. syringae (Attaran et al. 2008).

Alkaloids are specialized metabolites having a nitrogen atom within a heterocyclic
ring, as is the case for pharrnacologically important compounds such as morphine
(Cowan 1999). As was the case for other classes of metabolites, roles for alkaloids in
pathogen resistance have been proposed but difficult to prove. In one recent paper (Kim

and Sano 2008), transgenic tobacco plants developed to produce the alkaloid caffeine,

28

showed enhanced resistance to P. syringae and tobacco mosaic virus. Direct application
of caffeine to wild-type plant leaves generated similar resistance. Since the transgenic
plants constitutively expressed defense-related genes, the study concluded that a
bifunctional mechanism was at work, involving a combination of direct antagonism
toward the pathogen and indirect activation of host defense systems. Alkaloids are not
known to be significant metabolites in A. thaliana.

Though many gene functions have been suggested by recent research, the roles of
individual genes in plant defense responses and resistance to effects of stress and
pathogens remain mysterious. Defining those genes whose functions are important from
, those whose expression is more an effect than a cause of stress resistance remains an
important area for research. To answer such questions, combinations of genetic and
biochemical efforts ought to take advantage of new technologies capable of providing
rich information about gene functions. The study of plant biochemistry using
comprehensive metabolite profiling, known as metabolomics, is a promising approach for

discovery of new bioactive metabolites important for disease resistance.

1.5 Metabolomics as a new tool to close the gap between gene annotation and gene
functions
1.5.1 The feasibility and utility of metabolomics in gene annotation

As biology moves increasingly away from reductionism and toward the more holistic
systems biology approach, researchers have tools to explore comprehensive ‘omics’
studies, including genomics, transcriptomics, proteomics, and metabolomics. Like the

other ‘omics’ studies in systems biology, metabolomics is, in theory, the study of all the

29

metabolites and metabolic pathways in an organism (Guy et al. 2008). Of course, no
single analytical approach can measure all metabolites, but the spirit of this approach
dictates that investigators should design analyses to measure as many metabolites as is
feasible. Metabolomic approaches seek to profile metabolites in a nontargeted way, i.e.
to reliably separate and detect as many metabolites as possible in a single analysis.
Combined with information on transcript and protein abundance, metabolite profiles
ideally lead to a complete molecular picture of the state of a particular biological system
at a given time (von Roepenack-Lahaye et al. 2004).

In some of its earliest applications, metabolite profiling was used as a diagnostic tool
to measure metabolite composition to ascertain metabolic responses to herbicide, and the
classification of plant genotypes or treatments. Recently it has found growing application
to decipher gene functions, investigate metabolic regulation, and as part of integrative
analyses of systemic responses to environmental or genetic perturbations (Schauer and
Fernie 2006). The real advantage of metabolomics lies not only in the characterization of
the metabolic response to stress, but also provides utility in using metabolite profiles for
gene annotation (Schauer and Fernie 2006). Since metabolites are the end products of
cellular processes, their levels can be regarded as the ultimate responses of biological
systems to genetic and/or environmental changes (Gu et al., 2009). Understanding the
metabolome of each plant species is critical to understanding gene function and coupling
each of these functions to plant traits (Shulaev et al. 2008). The effects of a single genetic
mutatiOn are often not limited to a single biochemical pathway. Indeed, levels of
metabolites derived from seemingly unrelated biochemical pathways may be altered

owing to pleiotropic effects. To understand the effects of loss of a metabolic enzyme,

30

comprehensive identification and quantification of all metabolites would be helpful

(Fiehn 2001).

1.5.2 Methodology of metabolomics

Global metabolite analyses have taken one of two dominant approaches termed
metabolic fingerprinting and metabolic profiling. Metabolic fingerprinting involves the
analysis of the total composition of metabolites for comparison of patterns or fingerprints
among samples (as opposed to identification and quantification of specific compounds)
generated by analytical instruments. In this approach, the magnitudes of several
hundreds to thousands of discrete analytical signals are used to compare different
samples. These signals are usually generated using Nuclear Magnetic Resonance (NMR)
spectroscopy (Ward et al. 2007) or Fourier Transform-Ion Cyclotron Resonance Mass
Spectrometry (FT-ICR MS), the latter of which is a high resolution mass spectrometry
method. Each metabolite may give several signals (e.g. peaks in NMR spectra or mass
spectra), and the magnitude of each signal serves as a measure of the quantity of a
metabolite. Profiles of these signals are quantitatively compared to identify signals that
can define phenotypic differences, but it is uncommon to identify all of the metabolites
responsible for the phenotype. Metabolite fingerprinting is considered a nontargeted
analysis; the instrument detects signals from as many metabolites as possible. In
contrast, metabolite profiling, a term first used in the 1970s, is defined as the qualitative
and quantitative analysis of complex mixtures of physiological origin (Sumner et al.
2003). It consists of analyzing a known subset of metabolites, and is a technology

increasingly used in functional genomics. Most work has focused on optimization of Gas

31

Chromatography/Mass Spectrometry (GC/MS) and Liquid Chromatography/Mass
Spectrometry (LC/MS) analysis and demonstrations of their potential applications. The
former is a more developed technology because extensive libraries of mass spectra are
available to aid metabolite identification. Once optimized, these methods can be sensitive
and quantitative, and are capable of detecting picomole quantities of numerous
metabolites.

In addition to its role as a quantitative tool, mass spectrometry provides the first step
toward metabolite identification owing to its ability to generate molecular and fragment
mass information for nearly all compounds. The nontargeted LC/MS analyses of
metabolites reported in this work were conducted on a time-of-flight mass spectrometer
that is capable of measuring ion masses with great accuracy, usually to within 10 parts—
per-million (ppm) of their true masses. For a metabolite ion mass of 400 Da,
measurements accurate to :b 0.004 Da or better are common with this approach. Nearly
all observed ions are derived from molecules that are mixtures of different combinations
of naturally occurring stable isotopes. Common practice calls for reporting masses of
monoisotopic ions, more specifically, the ions corresponding to only the most abundant
isotopes (e.g. 12C, 1H, 160). It is possible for molecules of different elemental formulas
to have the same nominal mass, which is the mass rounded to the nearest integer. This
phenomenon arises because each element has a unique mass defect, which is the
difference between the atomic mass and the nominal mass. For example, the nominal
mass of 1H is 1 Da, whereas the exact mass of this isotope is 1.0078. In this case, the
mass defect is 0.0078 Da, or 7.8 milliDaltons (mDa). Substances that differ in their

elemental formulas will have slight differences in their mass defects. As a result,

32

elemental formulas of compounds can often be determined from molecular mass
measurements alone, providing that the mass measurement accuracy is sufficient.
Additional information that supports elemental formula assignments comes from ratios of
ion abundances for ions that contain naturally occurring stable isotopes such as 13C and
34$.

To date, both GC and LC in combination with various MS detection technologies and
NMR have been employed to analyze complex rrrixtures of extracted metabolites. The
use of GC allows separation of metabolites with high separation efficiency and
sensitivity. In combination with MS, it provides very accurate, high resolution, sensitive
and selective identification and quantification of chemical compounds (F iehn 2008)! In
the past, GC-MS has been applied and optimized for simultaneous analyses of
metabolites in many different plant species such as Arabidopsis thaliana (F iehn et al.
2000; Fiehn 2006), Solanum tuberosum (Roessner et al. 2000), and Cu'curbita maxima
(Fiehn 2003). Despite the many advantages of GC/MS, GC can only be used for low
molecular weight (<1000 Da) compounds, that are either volatile at relatively low
temperatures or can be chemically transformed into volatile derivatives. Thus, for a
comprehensive analysis of a greater range of plant metabolites, LC/MS offers more
advantages. First, compounds do not have to be chemically altered by derivatization prior
to analysis and second, highly polar, thenno-unstable and high-molecular weight
compounds, such as oligosaccharides, glycoconjugates, and complex lipids, can be

separated and quantified (Roessner 2007).

1.5.3 Challenges of metabolomics

33

The biggest challenge facing researchers wishing to perform metabolomic analyses
lies in the identification of all the metabolites. Nowadays the significant challenge for
plant metabolomics is the lack of a fully described and annotated metabolome for any
plant species. The magnitude of such tasks is immense; it has been estimated that the
plant kingdom produces 90,000-200,000 metabolites (Fiehn 2002), but these estimates
are based on speculation. Arabidopsis, with a relatively simple secondary metabolism,
has been estimated to produce more than 5,000 metabolites (Roessner et al. 2001).
Because of the diverse chemical nature of metabolites and the lack of schemes that can
predict metabolites structures from genetic information, the analysis of the metabolome is
particularly challenging. Typical efforts have rarely identified more than 100 metabolites
in a single analysis. The study of CBF cold response genes and metabolome in cold
adaptation in Arabidopsis (Cook et al. 2004) is a strong example of the successes and
limitations of GC/MS metabolite profiling. In this study, 434 metabolites were detected
and 325 (75%) were found to increase in Arabidopsis plants in response to low
temperature. However, the large majority (84 of 114) of metabolites increasing by > 5
fold remained unidentified despite the availability of libraries of metabolite mass spectra.

It is frequently the case that metabolites that are novel in stressed plants or specific
mutants are present in tiny quantities that often preclude structure elucidation by NMR
spectroscopy or x-ray crystallography. For this reason, information fiom mass spectra
often provides the only information upon which metabolites can be identified. Perhaps
the most important information obtained from a metabolite mass spectrum is the
molecular mass of the compound. For LC/MS analyses, metabolite ionization is a gentle

process. Most mass spectra contain only molecular mass information, with minimal

34

formation of fragment ions that are common in GC/MS analyses. Additional information
about structure can be generated by inducing the molecular ions to fragment. This is
accomplished using collision induced dissociation (CID), in which the molecular ion
collides with gas molecules, and the collision leads to formation of fragment ions whose
masses are also measured. This information can be used to construct a proposed
chemical structure for the metabolite. Since CID spectrum libraries are limited and are
not standardized, metabolite identification is a time- and labor-intensive process.

Although there are growing discussions and numbers of publications on the
processing of metabolomics in biology, there are few examples of open-ended
metabolomic approaches leading to novel biological insights. Most plant biologists are
still using targeted metabolite profiling to discover functions for genes and their
regulation (Gu et al., 2009). Furthermore, most metabolomic studies on plants have used
plants grown under control environmental conditions using wild type plants, and have
measured only a limited number of metabolite classes. There are great opportunities for
research that employs metabolomic approaches for investigation of responses of wild
type and mutant plants to various stresses including pathogen infection.

The second great challenge posed in metabolome analysis involves finding improved
methods for interpretation of complex metabolite data sets that will allow for robust
comparisons of metabolite profiles. One may anticipate that important changes in
metabolite profiles may involve appearance of novel metabolites at low levels, and these
are easily obscured by more abundant metabolites. Statistical evaluation and visualization
software tools have been developed recently and enable researchers to identify those

metabolites that exhibit significant differences between treatments or genotypes. These

35

tools enable investigators to recognize those metabolites that are highest priority for
identifying, but confirmation of structure requires convincing evidence of structure. In
addition, knowledge of metabolite structure often fails to reveal the substance’s
biological function. Linking changes in metabolite levels to meaningful biological
interpretation often requires challenging follow-up experiments.

Most early metabolite profiling studies of plants demonstrated the capabilities of
metabolomic methodologies in the form of distinguishing different plant species or
tissues. Before 2006, most published reports largely demonstrated the capabilities of the
technology for distinguishing plants or tissues without using metabolite profiles for
hypothesis generation or testing. More exciting applications have appeared during the
past five years. Profiling of metabolites in transgenic plants using GC/MS (Cook, Fowler
et al. 2004) revealed functions of the CBF cold response pathway genes in regulating
levels of amino acids, carbohydrates, and organic acids in cold stress. A somewhat
different approach, used by Georg J ander and colleagues (Jander et al. 2004), employed
high-throughput LC/MS/MS profiling of amino acids in thousands of mutant Arabidopsis
lines to discover a role for threonine aldolase in controlling seed amino acid levels. Both
targeted and nontargeted metabolite analyses were used by Saito and coworkers (Tohge
et al. 2005) to generate results that documented participation of two genes as
glycosyltransferases in anthocyanin biosynthesis.

Another challenge faced during the application of metabolite profiling results because
the inherent variation in biological behavior usually exceeds analytical variance in
metabolite measurement. It has been estimated that the biological variability exceeds

analytical variability of gas chromatography-mass spectrometry by a factor of ten

36

(Roessner et al. 2000; Sumner et al. 2003). Quality control is important for any analysis,
as it allows analysts to judge the reliability of analytical measurements. Some groups
have reported spending comparable efforts on analysis of quality control samples as on
biological extracts (F iehn et al. 2008).

Absolute quantification of all metabolites presents enormous challenges. A shortage
of reference standards makes comprehensive and absolute quantification of
phytochemicals in plant extracts improbable (Bednarek and Osbourn 2009). As a result,
most metabolite profiling efforts are usually limited to relative quantitative analyses,
reporting analytical signals relative to an internal standard or relative to the sum of all
metabolite signals. Phenotypes are thus defined in terms of differences in relative levels

of individual metabolites between treatments, time points, or genotypes.

1.5.4 Introduction to my project

Metabolite profiling provides a powerful link between genotypes and phenotypes.
LC/MS is just emerging as a tool for plant metabolite profiling, and is being used in this
study as a primary screening method to determine changes in plant metabolite profiles
associated with different genotypes or treatments in Arabidopsis. Such changes would
provide powerful evidence for functions of specific genes or mutations. Specifically, I
have used liquid chromatography coupled to electrospray ionization time-of-flight mass
spectrometry (ESI-TOF—MS) for rapid identification and measurement of metabolites
with high mass accuracy (to ~ 5 ppm). This analytical tool can measure hundreds of
metabolites in a single analysis, and can discriminate between metabolites that have the

same integer mass (and slightly different molecular masses) but different elemental

37

formulas. Under biotic and abiotic stress, plants may produce a wide variety of secondary
metabolites that may differ both qualitatively (presence or absence) or quantitatively
relative to control plants.

The impetus for this research came from preliminary studies performed by graduate
student John Hanley in the Jones laboratory, then at Penn State University (Hanley JC,
PhD Dissertation, 2005). His tentative metabolite identifications, performed without the
benefit of high resolution mass spectrometry, suggested that O-methyltransferases
(OMTs) involved in polyphenol metabolism might be induced in Arabidopsis leaves
upon P. syringae infection, and to a lesser extent, following herbivory by two insects. To
explore the functions of OMTs during pathogen infection, the studies described in
Chapter 2 report assessment of leaf metabolic phenotypes in infected and control mutant
and wild type Arabidopsis using LC/l' OF MS.

The first step for analyzing metabolite profile data involves extraction and
organization of LC/MS data and determination of whether there are meaningful statistical
changes in metabolite levels. In this study, automated processing of LC/MS data using
Masslynx sofiware was used for chromatographic peak detection, alignment, and
integration. Statistical analyses using Principal Component Analysis (PCA), Partial Least
Squares for Discriminant Analysis (PLS-DA), Orthogonal Partial Least Squares (OPLS),
and two-way AN OVA were used to aid recognition of those metabolites most affected by
genetic or environmental changes (Eriksson et al. 2004; Trygg et al. 2007). PCA is an
unsupervised data reduction technique that is used to reduce the dimensionality of a
multi-dimensional dataset while retaining the characteristics of the dataset that contribute

most to data variance. PLS-DA is a supervised regression extension of PCA that takes

38

advantage of class information to attempt to maximize the separation between groups of
observations. OPLS is a modification of the usual PLS method that filters out variation
that is not directly related to the response.

Identification of novel metabolites relevant to pathogen stress represents a central
aspect of this work, but metabolite identification remains a challenging activity. Ten
years ago, a summary of A. thaliana metabolism reported 36 secondary metabolites fiom
four major classes (Chapple et al. 1994). This list now includes more than 170 secondary
metabolites from seven major classes (D'Auria and Gershenzon 2005), and this still
represents only a small fraction (about 3%) of expected Arabidopsis metabolites. Many
important metabolites and novel pathways still need to be identified and studied. The
research described in this dissertation will help expand understanding of Arabidopsis
secondary metabolism and develop the research tools for analyzing metabolic networks,
cross-talk between different pathways (Bostock 1999), and identifying novel pathways.
Furthermore, pathogens and other stresses can induce extensive changes in gene
expression and thereby reprogramming of cellular metabolism. The roles of and
connections between these responses remain mysterious to a great extent. It is often not
clear which defense reactions are fundamental to plant survival and which are simply an
effect of the perturbation of an interconnected network. One promising approach to
address these questions involves large-scale profiling of metabolites in mutant plants to
identify genes and their functions (Last et al. 2007). In this study, I have been
investigating the functions of OMT genes using metabolite profiling. It is expected this
work will be helpful to determine whether OMT and plant resistance pathways are

coordinately regulated or are activated by independent mechanisms.

39

In this study, I applied metabolite profiling to expanding functional annotation of
genes to help develop a more comprehensive understanding of cellular responses to
biological conditions. In Chapter 2, I used a reverse genetic approach to investigate OMT
gene function using metabolite profiling to recognize genes important in leaf secondary
metabolism and explore their broader biological functions under conditions of pathogen
infection. In Chapter 3, a comprehensive and fast analytical method was developed to

support studies of the dynamics of multiple phytohorrnones in wounding stress.

1.6 Research objectives
The primary purpose was to develop high-throughput LC/MS methods to measure
metabolite profiles, and apply them to help us to understand gene functions and gene

regulation in plant responses to wounding and pathogen stress.

40

CHAPTER 2

Reverse genetic investigation of functions of 0-methyltransferases in
Arabidopsis thaliana responses to Pseudomonas syringae infection
using liquid chromatography/mass spectrometry for nontargeted metabolite
profiling

41

2.1 Introduction

Discoveries of functions of plant genes hold central importance in efforts to improve
global food supplies and develop biological energy resources. Investigations of gene
functions have relied on two strategies known as forward genetics and reserve genetics.
Both tactics are most powerful when phenotypes are described explicitly, and global
phenotype profiling has emerged as a set of vital methodologies for functional genomics.
Plants synthesize a huge repository of metabolites, but most conventional approaches
have employed a small ntunber of phenotypic assays that cover a limited number of
biochemical pathways. Recent years have seen emergence of a research area known as
metabolomics, which aims to describe the biological state of an organism by measuring
the entire set of metabolites.

This dissertation presents development and application of LC/MS methods for
metabolomic profiling for investigations of plant responses to wounding and pathogen
stress. In the first study, LC/time-of-flight MS was used for nontargeted profiling of

metabolites in leaves of wild type and O-methyltransferase (OMT) knockout mutants.

2.1.1 Preliminary studies in our laboratory

The impetus for this research came from preliminary studies performed by graduate
student John Hanley in the Jones laboratory at Penn State University (Hanley 2005).
When Arabidopsis plants were infected by Pseudomonas syringae, one of the most
striking changes in metabolite profiles was a sharp decrease in the abundant metabolite
sinapoyl malate (SM) and the less abundant 5-hydroxyferuloyl malate (SHFM), with

concomitant increases in a metabolite with molecular mass 14 Da greater than SM. Since

42

this mass was consistent with methylation, and the fragmentation pattern was indicative
of a malate ester, the report identified the new metabolite as 3,4,5-trimethoxycinnamoyl
malate (TMCM) (Hanley 2005). The chemical structure of TMCM has the unusual trait
among Arabidopsis metabolites as being methylated in the 4-position (Figure 2.1), and
methylation in this position posed a potential effect on lignin formation. These enzymatic
O-methylations are proposed to occur via catalysis by OMTs. There are 31 loci with 45
distinct gene models encoding members of the OMT family in Arabidopsis. Therefore,
the original aim of this project was to identify which genes code for the enzyme that

methylates sinapoyl malate to 3,4,5-trimethoxycinnamoyl malate.

43

.Amoom >250—d >25“: .3 #53 no
000.3 .3208 Sofiaéfixéfiofitfiflim 8 830E 789050 98 8038 323033636 Eob cows—£0086 comaoi mm 053m

3020.022? 30505965055000 5080205 $00050 $83225 32303233
0
0:00 MU:00 :00
:08 00m: :08 0: :08 0:
/ m 0 0 : 0 /
o / Com: 0 I / Com: _ O:
:08 0 :08 0 :08 .

44

[lit

a.
ill.
'

(L)
. 0-,
I, n

2.1,;

However, while my early experiments replicated detection of a metabolite with this
mass that appeared following P. syringae infection, its assignment as TMCM was called
into question as described in more detail in the Results section below. High resolution
mass measurements determined the mass of the ionized metabolite to be m/z 353.1600.
This experimental result disproved the identification of 3,4,5-trimethoxycinnamoyl
malate (C16H1309) with its calculated monoisotopic ion mass of m/z 353.0873, and
suggested instead a formula consistent with an oxylipin conjugate OPC:4 malate of

formula C13H2607 with the structure shown in Figure 2.2.

o
m COOH
0%

HOOC

Figure 2.2 Chemical structure of OPC:4 malate (C13H2607) with accurate molecular mass
of 354. 1 679.

An unexpected turn in this research came about a short time later, as amt] showed
distinct differences in metabolite profile compared to wild type plants and showed
evidence of slower development of pathogen colonization than wild-type plants. More

extensive investigation of these processes is presented in the research described below.

2.1.2 Introduction to O-methyltransferase (OMT)

45

Metabolic methylation plays important roles in secondary metabolism in animals,
microbes, and plants, and methylation of hydroxyl groups is catalyzed by members of a
family of O-methyltransferases. O-methylation catalyzed by OMTs [EC 2.1.1.6.x]
involves the transfer of the methyl group from S—adenosyl-L-methionine (SAM) to the
hydroxyl group (-OH) of an acceptor molecule, with the formation of its methylated
product (-OCH3) and S-adenosyl-L-homocysteine (SAH) (Ibrahim et al. 1998).
Methylation of polyphenols by OMTs is considered a critical step for several biological
functions, since methylation of a phenolic hydroxyl can block glycosylation and other
chemical transformations, particularly those that convert phenolic groups to phenoxy
radical intermediates. For example, lignin, one of the polymeric products of the general
phenylpropanoid pathway, is an abundant biopolymer that adds mechanical strength to
plant tissues and is thought to be an important component of some disease resistance
mechanisms. Monomeric precursors of lignin include two methylated polyphenols,
sinapyl alcohol and coniferyl alcohol, which contribute syringyl and guaiacyl groups
respectively. Two classes of “lignin-specific” OMTs have been found (Joshi and Chiang
1998; Burga et al. 2005). Class I consists of low molecular weight (23-27 kDa subunits),
Mg2+—dependent OMTs. Class I enzymes were initially identified as caffeoyl-CoA OMTs
(CCoAOMTs), and these enzymes catalyze methylation involved in formation of the
guaiacyl moiety of phenolic CoA esters in lignin biosynthesis. Class II enzymes are ,
larger OMTs (3 8—43 kDa subunits) that methylate the phenolic hydroxyl groups of
caffeic acid, caffeoyl aldehyde, and caffeoyl alcohol (COMTs) independent of Mg”, and
were considered to methylate primarily 5-hydroxyconiferaldehyde residues during lignin

biosynthesis (Joshi and Chiang 1998; Burga et al. 2005). COMT-deficient plants exhibit

46

altered lignin structures and properties. In this mutant, lignins are deficient in syringyl (S)
units and contain more S-hydroxyguaiacyl units (5-OH-G), which are S-units lacking a
single methyl group. Therefore, the enzymatic activity of COMT appears to be involved
in lignin S-unit biosynthesis. However, COMT] is not a limiting enzyme for the lignin
synthesis since the over-expression of poplar COMT] in wild-type Arabidopsis did not
increase the S-lignin content (Goujon et al. 2003; Do et al. 2007).

The Arabidopsis thaliana genome contains 31 loci that encode members of the OMT
family (TAIR 2007), but only a few of these OMT genes have been studied so far (Zhang
et al. 1997; Muzac et al. 2000; Goujon et al. 2003; Deem et al. 2006; Pontier et al. 2007).
Caffeic acid O-methyltransferase 1 (OMTI) (At5g54160) is one of the most studied
OMTs in Arabidopsis. OMT] was originally thought to be a bifunctional enzyme that
methylates caffeic acid and 5-hydroxyferulic acid (Zhang et al. 1997), based on high
sequence similarity to COMT from other species. Transgenic studies revealed that
another predominant role of COMT is the methylation of 5-hydroxyconiferaldehyde and
5-hydroxyconiferyl alcohol to sinapaldehyde and sinapyl alcohol, respectively, which are
also metabolites in the lignin synthesis pathway (Goujon et al. 2003; Raes et al. 2003). In
contrast in a report from the year 2000, OMT] was cloned and reported to use the
flavanol quercetin as the preferred substrate, but no significant activity was observed
toward the hydroxycinnamic acids, caffeic or 5-hydroxyferulic acid (Muzac et al. 2000).
More recently, it was reported that stem extracts of the OMTl T—DNA knockout mutant
exhibited only about 7% of activity toward caffeic acid and 5-hydroxyferulic acid relative
to that of wild type, and stem extracts from an overexpressing line gave a 5-fold increase

in activity toward these two substrates relative to wild type (Goujon et al. 2003). Since

47

the finding from the T-DNA mutant is inconsistent with the biochemical observations
regarding OMTl activities, a more complete assessment of the kinetic properties of
OMTl is needed to resolve these differences.

In a related study, the group of Kazuki Saito (Tohge et al. 2007) recently used
targeted profiling of flavonoids using LC/MS and photodiode array detection, and
transcript analysis to propose that OMT] in Arabidopsis root extracts is also involved in
the methylation of the flavonol quercetin to form isorhamnetin. Quercetin and its
derivatives are not abundant in leaves of the Col-0 ecotype of Arabidopsis, where
kaempferol derivatives, which lack a hydroxyl group in the position methylated for
quercetin, are the dominant flavonols.

Minor sequence modifications are known to result in significant changes in substrate
affinity for the enzymes (Gauthier et al. 1998; Wang and Pichersky 1999). As a result,
enzymes may be somewhat promiscuous in their use of substrates. OMTs are involved in
many different biosynthetic pathways. Thus, comparisons of sequence similarity/identity
are not sufficient to determine the functions of OMT gene family members. To fill in the
gaps between gene annotations and proven gene functions, metabolite profiles of leaves
from wild-type plants and T-DNA insertion mutants were used in the research described
below to provide evidence for the metabolic role of different OMT genes.

In Arabidopsis leaves, sinapoyl malate is one of the most abundant metabolites. This
polyphenolic compound has two methyl ether groups, and presence of these groups is
attributed to OMT genes. The reduction of sinapoyl malate levels was observed in several
mutants in Arabidopsis phenylpropanoid pathway for lignin synthesis. An Arabidopsis

mutant defective in the conversion of ferulate to 5-hydroxyferulate exhibited lower levels

48

of sinapoyl esters in leaves, a trait recognized by a characteristic red fluorescence under
UV light, compared with blue-green color in the wild type plants (Chapple et al. 1992);
(Ruegger et al. 1999; Ruegger and Chapple 2001). One mutant, fahI-Z, was unable to
produce sinapoyl malate but exhibited growth indistinguishable from wildtype under
laboratory conditions. This finding demonstrated that these compounds are not essential
for normal development, but a separate study (Landry et al. 1995) demonstrated that
sinapoyl esters confer protection against damage by ultraviolet radiation. An Arabidopsis
mutant deficient in the expression of a specific O-methyltransferase (AtOMTI) increased
S-hydroxyguaiacyl groups in lignin and decreased, but did not eliminate, levels of
sinapoyl esters in leaves and stems (Goujon et al. 2003). This finding implicated more
than one O-methyltransferase in sinapoyl malate biosynthesis. These studies preceded
comprehensive metabolite profiling, and none explored the responses of mutants
deficient in polyphenolics to wounding or pathogen infection.

During the course of the research described in this dissertation, it was reported that
OMT] functions important in biosynthesis of sinapoyl esters and lignin are somewhat
redundant with methylation catalyzed by caffeoyl CoA O-methyltransferase
(CCoAOMTl). The development of homozygous double mutants was arrested at the
plantlet stage, with the absence of sinapoyl malate and isorhamnetin (Do et al. 2007).
This report concluded that the methylation functions of these enzymes play crucial roles

in plant development.

2.1.3 Specific aims

49

The goal of this research project has been to develop and apply metabolite profiling
approaches to learn about biochemical and biological functions of OMT genes and

related metabolites in plant-pathogen interactions.

2.2 Materials and methods
2.2.1 Chemicals

All commercially available compounds and mobile phase additives were purchased
from Sigma (St. Louis, MO, USA), and HPLC-grade solvents were purchased from VWR

Scientific.

2.2.2 Confirmation of homozygous T-DNA insertion lines using PCR (DNA level)
The seeds of the A. thaliana T-DNA insertion mutant lines of At5g54160

(SALK_135290 and SALK_002373), and all other mutants, were purchased from the

Arabidopsis Biological Resource Center (Ohio State University, Columbus, OH). All of

the mutants used in this study were derived from the Col-0 ecotype.

Genomic DNA isolation
Genomic DNA was extracted as a template for PCR for amplification and isolation of
the OMT gene. Around 0.5 cm2 of leaf tissue was cut and ground in 20 ul 0.5 N aqueous

NaOH and 180 pl 0.1 M Tris buffer (pH 8.0). The tissue was centrifirged (8000><g, 5

mins) and 5 ul supernatant containing genomic DNA was used for PCR.

Primer Design

50

SALK T-DNA primer is designed using the protocol from Salk Institute Genomic

Analysis Laboratory as shown in Figure 2.3 (http://signal.salk.edu/tdnaprimers.2.html).

 

 

 

 

 

LB
T-DNA —l

 

 

Genome

LP fit]

 

 

 

 

 

 

 

 

 

Figure 2.3 Diagram of primer design for SALK T-DNA insertion lines. LP and RP are
left and right genomic primers. LB is the left T-DNA border primer.

Two sets of primers (LP + RP and LB + RP) were used in PCR. The primers were LP
5’-GAACGTGATTGATTGTAATCGG-3 ’, RP 5’-AATTCTTGATGGTGGGATTCC-3 ’ ,
and LB 5’- GCGTGGACCGCTTGCTGCAACT-3’). For heterozygous T-DNA insertion,
two bands in PCR were observed: one for LP+RP and the other one fiom RP+LB. For

homozygous T-DNA insertion, only one band was observed from RP+LB.

PCR reaction
Amplification using LP and RP was performed using l>< PCR buffer without MgClz,
2 mM MgC12, 0.2 mM for each dNTP, <0.5 pg genomic DNA, 1.25 unit Taq polymerase,
0.5 uM of primers LP and RP. H20 was added to adjust to final volume of 50 ul.
Amplification using LB and RP was performed in the same PCR reaction as

described above, except 0.5 uM of LB and RP were used as primers.

51

PCR program was run at 94 °C for 5 min to denature, followed by 30 cycles as
follows: 94 °C for 45 seconds, 53 0C for 45 seconds, and 72 °C for 80 seconds, and 72 °C
for 10 rrrin. At the end of 30 cycles, the solutions were kept at 4 °C until analysis by gel

electrophoresis.

2.2.3 RT-PCR for gene expression analysis

Total RNA was extracted using RNeasy plant mini kits (QIAGEN, Valencia, CA,
USA) from both wild type and mutant plants (five weeks seedlings). Primers
corresponding to OMTl cDNA sequence were designed according to the T-DNA
insertion position on genomic sequence. Since the T-DNA is inserted at the second exon,
one primer is set at the 5’UTR sequence (5’- CCCATCACA CAATCCTCAA AACAG-
3’) and the other one is set in the second exon (5’- GCATAGTT TTGGAACCGT
GGAAGG-3’). Reverse transcription (RT)-PCR was performed using SuperScriptTM III
One-Step RT-PCR System with Platinum® Taq High Fidelity kits (Invitrogen, Carlsbad,
California). The wild type will have a piece of 490 bp PCR product while omtl mutant is

not expected to express this product.

2.2.4 Plant culture and pathogen stress

For pathogen stress treatment, 5-week-old non-bolting wild type and mutant plants
(12 h light/ 12 h dark, 22 °C) were inoculated with Pseudomonas syringae DC3000 (~106
cfu/ml in 10 mM MgClz). About 20 pl of the bacterial suspension was inoculated with a
syringe via the stomata into the lower surface of half of the leaf. Control leaves were

treated similarly with a sterile solution of 10 mM MgCl; Local ”leaf tissues were

52

harvested from control and stressed plants for both the wild type and the mutant at 0
hour, 1 hour, 24 hours, 48 hours, and 72 hours after inoculation. Five biological replicates

were performed.

2.2.5 Quantification of bacterial growth procedure

The classic phytopathological technique for quantifying bacterial virulence is to
measure bacterial growth within the host tissue (plant leaves in this study) (Katagiri et al.
2002). A standard enumeration procedure involves pathogen inoculation (syringe
injection at ~106 cfu/ml) followed by assaying bacterial populations present within host
tissues daily. Briefly, leaves were harvested and surface sterilized with 70% ethanol for 1
min. Leaf disks were excised from leaves with a cork borer and ground with plastic pestle
in a microfuge tube with 100 pl sterile distilled water. The pestle was rinsed with 900 pl
of water, with the rinse being collected in the original tube such that the sample was now
in a volume of 1.0 ml. An aliquot of 100 pl sample was removed and diluted in 900 pl
sterile distilled water. A serial 1:10 dilution series was created for each sample by
repeating this process. The samples were plated on King’s medium B with antibiotics to
select for the inoculated bacterial strain. The plates were placed at 28 0C for 2 days and

then the colony forming units (cfu) for each dilution of each sample were counted.

2.2.6 LC/MS and LC MS/MS methods
In this study, fast LC/MS and LC MS/MS methods were developed for the
identification and quantification of metabolites. The former employed high resolution

time-of-flight mass spectrometry analysis with multiplexed CID, whereas the latter used a

53

triple quadrupole mass spectrometer to ensure that observed fragment ions were derived
from specific molecular ion species.

Leaves were collected, weighed, frozen in liquid nitrogen, and extracted with
methanol/0.15% aqueous formic acid (50:50 v/v) at 4 °C for 24 h. An aliquot (10 pl of 20
pM) of internal standard solution of 2 pM (final concentration) propyl 4-
hydroxybenzoate was added as internal standard into 90 pl leaf extracts. 10 pl of leaf
extract was injected and separated using a Therrno BetaBasic C18 column with a gradient
of 0.15% aqueous formic acid (solvent A), methanol (solvent B), and acetone (solvent C)
over 25 mins with a mobile phase flow rate of 0.1 ml/min. The gradient consisted of a
linear increase from 5% A/95% B to 100% B in 17 min, and maintenance of 100% C for
4 min. The gradient was then adjusted to 5% A/95% B in 1 min. The column was re-
equilibrated for 3 min prior to running the next sample. The column, which was
maintained at 30°C, was interfaced to a Waters LCT Premier time-of-flight mass
spectrometer (Waters Corporation, Milford, MA) operated using electrospray ionization.
Negative mode electrospray ionization was applied for all analyses. Accurate mass data
were collected using multiplexed Collision Induced Dissociation (CID) acquisition
through modulation of the Aperture 1 potential. Spectra were acquired in quasi-
simultaneous acquisition with five different CID conditions (10, 30, 50, 70, and 90 volts)
in a single analysis to generate fragmentation patterns useful for metabolite identification.
Peak integration and alignment were performed using Markerlynx and Quanlynx from

Masslynx 4.1 software.

54

Additional support for metabolite identifications was generated using LC-MS/MS
using a Quattro Premier XE mass spectrometer, employing LC gradients similar to those

described above.

2.2.6.1 Identification of secondary metabolites

Leaf metabolites were identified based on accurate molecular masses, fragment ion
masses, isotopic patterns of naturally occurring stable isotopes, and mass defects. An
example is shown in Figure 2.4 about how accurate mass measurement was used to help

identify a secondary metabolite.

55

.80.. wd 3:0 .3 @96de NE. 00.? 080.0200
2.0 EB. 20.0.0 @25de 0S5 :0. .54): 05 8.. 00.08 083008 05 .0000 0.5 E .0008 00000008 0.00. :0 o. MEcaoamotoo
0020.02.88 30080.0 03.002. .0 owqfi 2.. 3.8.. 38:80 0000.— 5»... 2.... 0806000000 0008 05 3:. 0003:. 300005.08
m. .05 0:009:00 00:20.0.— 0 20.? 0.00008 03.3808 .0 0085088 032.0 2.033 b80080 .3538: 0 mafia 003300
003 58:80 0008 ami 0008200008 0008 800080 .3 A0m002m~msov 20.00.0803 .b:no.£.b..08-v ..0 02.03.0003 YN 00%.".

 

 

 

NV 1 an. amok. .2015".
D m: 03v Dmmv D.Dm¢ ohmv Damv Dru]. S: 0.0:.
Narr....4.|rihelrrLi—_ ts!»— :—. 4 4. t .F l.._qw .Llru.k4.tqkr_rt\_clf.rel..t. .4..4..<..4..4_....4.. D
:0— mvv 01mm. 03. m.mmo mm.‘ owe: ..mw mmmo wmv m memmfl . ammo. m3. memo m 3.
mmwo an
00
000: 0040.03. 00.
-ww 92 ...OF ”P fine too rmmnhmvv E0 800.8 mm mlmmv. m<m mlleJX

mmv. 9:6 20.0

 

wO M7. 02.. ONU MK” m6- mé- SVOde

 

 

0 0 0. 00 0

0 0 _ 0N Na 0 0m 00 z 00: NC 0...... 0.0- 0.0. $00.00..
N 0 N 0. 00 N mm 00 N: 0:: 06 0H 0.0 0.0 0000.09 005.00..
_ m _ o _ z _ I _ u — 28m . 0.35.0”. _ moo _ Ema — nos _ “no: 6.00 _ nuns.

 

730:. £000 .3. 3.500. guano—o am 3 a... 2.5: 55.3 3.500. n 5.3 tong—.6)“. Tau—:52: ohm
ace. segue—m :u>m 1:0 300 $qu 033020.33-

oxoo... n 00:00:30.: 00 n c0:0:000w “9000:0060 00.020 000.00.

Dom n x9: .0 0- n EE Hmmfl. . 20.0. Om n 00000900

2029.3 nuns. 0.050

 

 

 

___:: ALA—2“: . _..~..L—._.-__

 

56

An example to demonstrate the importance of accurate mass measurement for
metabolite identification comes from my correction of the previous identification of
3,4,5-trimethoxycinnamoyl malate (C16H1309) to OPC:4 malate (C13H2507). Measured
mass of this deprotonated compound in negative ion mode is m/z 353.1610, which
disproved the previous identification of 3,4,5-trimethoxycinnamoyl malate (C16H1309)
(theoretical m/z 353.0873 for [M-H]'), but suggested instead a new oxylipin conjugate
OPC:4 malate (C13H2507) with theoretical m/z 353.1600. (Figure 2.2). This finding may
hold significance to jasmonate signaling during pathogen infection because OPC:4 is an
intermediate in IA biosynthesis. Its conjugation to malate presents a competing pathway
to JA formation and a potential mechanism for sequestration of a potential JA precursor.
The biological properties of OPC:4 malate have yet to be established, and await synthesis
or isolation of sufficient quantities of this compound.

An example is shown in Figure 2.5 about how our laboratory identifies secondary

metabolites using multiplexed in-source collision induced dissociation (CID) in LC/MS.

57

XLG_B7_P96 390 (13.799) 5: TOF MS ES-

 

 

100 Aperture 1 voltage 90v ”7
115 4.150223 3
71\ ' 24 233 3 9
0 ,
XLG_B7_P96 313003.791) A rt 1 lta 7O 4: TOF ms ES-
208
100 _149 223 pe ure vo ge v
136 164
93 ,
0 71 \ 224292 339 , 679
XLG_B7_P96 390 (13.783) 3: TOF MS ES-
00 149 223 Aperture 1 voltage 50v 410
1 164
133 224
XLG 87 P96 380 13.776 22 TOP MS ES-
— - ( ) 223 Aperture 1 voltage 30v 12.7133
100
0 7911:3133 164 24 309 339 399 445 677 695
XLG_B?_P96 380 (13.769) Cm (3811381) Aperture 1 voltage 10v 1: TOP MS ES-
339 2.9593
100
I 223 340
0 7529 97 239 / 423437469 ,593 679 m

50 100 150 200 250 300 350 400 450 500 550 600 650

Figure 2.5 Identification of sinapoyl malate (C15H1609) based upon fiagment ions from
multiplexed CID. Five different functions were used to generate these spectra:

Function 1: Cone voltage 40v; Aperture 1 voltage 10v

Function 2: Cone voltage 40v; Aperture 1 voltage 30v

Function 3: Cone voltage 40v; Aperture 1 voltage 50v

Function 4: Cone voltage 40v; Aperture 1 voltage 70v

Function 5: Cone voltage 40v; Aperture 1 voltage 90v

The major fragments are observed at m/z 223 and 133. The assignments of these

fragments are shown in the following fragmentation reactions (Figure 2.6).

58

.38on @38 :2 gramme E $02328 BEE: Rooms? 8m 28:82 noufiaofiwfim 9N oSmE

5.6de Hwees. Umxm
m"OmIvo ӣ250... _wo_Eoco
mmnodmm Hewes. Homxm
IOOO

-mOmFIBo 63:5“. 52820
Jlo .. foo

1000 .000 o:
1A -1 1 o
J1 / 00m...
985$ 9622 .6me 1000

-mOFVISO ”£250“. _8_Eoco

«.1 00
CI

com:

59

Some characteristic fragment ions can be used for identifying compounds with
common functional groups. For example, the following fragment ions can help the
analyst recognize the presence of specific classes of metabolites: m/z 97 for
glucosinolates (fragment ion of the sulfate group HSO4') or phosphorylated metabolites
(as H2PO4'), m/z 291 (deprotonated OPDA anion) for conjugates of the oxylipin OPDA,
and m/z 306 (deprotonated glutathione) for various glutathione conjugates. Metabolite
structures can also be assigned through the assistance of fragmentation through loss of

group-specific neutral fragments.

2.2.6.2 Quantification of secondary metabolites

Metabolite levels were normalized to the integrated peak area for a constant amount
(2 uM) of added internal standard propyl 4-hydroxybenzoate, then normalized to fresh
leaf weight. Such normalization is essential for relative quantification, because
measurements of absolute amounts were not possible owing to the unavailability of
authentic metabolite standards. Here is an example (Figure 2.7) to show how to quantify

a secondary metabolite S-hydroxyferuloyl malate (SHFM).

60

AQUA Mona: 2: .2953 333 352m 59:8

86:32:20 5 692.3888 a 35% .5363 :2 sea 65 B 82% 55 a 62? 95 Ba .8: £5 982826629
.3qu @3283 8 Ammm NEG 2min mo 0:8 83 xwoa 3 onEuoBu fl 2min .«o cozwomwcmnc 3:23. 65882: .«o 330:8 05 mo
23on a m_ x3e EX an .595 38 xaoa 6283:: .mmm NE.— uom «on xdoq @8835 2: mm vwmm 28 AfiEV can." amaze 3 $2: 2:
9 am .253 62392 65 a £6 .6? 66% 65 E .Qx .632 .35 953 25 25 26 6R 6%: .3): Essa :5 86¢ 6:
n}: 53> Beonaonxxeuxné 1305 c3283 658E 98 mmm ES 53» SEIm mo GEO guwoamfiofio com 38988 -mm ”m oSmE
.omm NE.— fi 58 Q6 mo 2:: cot—~88 a “a Banana czmv

BEE: 3093mm .mmm N): 3 £5 £6 mo 08: 8:5on m an 8.83% ASEES own—m8 Ro—Eommxohimfi .621 332 .323 mam—m
cab Em? use 6:. Bag: .33: E82: 25o Sea 8838 0:33 (:8— E meow—3808 .«o 6:» EmoEEoEo com 38 -mm— H< Pawn
.3235 .32 95 E Czar—C 6338 _>o_ao§xo€>:-m we cosmouccgc gun—om 5m oSwE

61

8.3 8.9 8.9 8d 86 8d 8.“

 

 

 

 

 

 

 

 

 

05:....r_.[._.b.p;__.._._p._..b ....p..L._..FL....._.p.p_...b_...._..F.D
N3 fl
5m .2 .%
8N fl
omm©__ .oo_
0: mom
.3 m2 m9 H_ m to 333.0?
8.3 8.9 8.2 8m a; 2; Sn
..:L...._.. | - - ....+._.......r...p.._..:kL..._.r.._..r.o
i .
«8 2w
3.2 3.x _ x .
at 53 .% .
o: 8;... _ M
an 5.. fl
38m . .8—
0: won cam now
-mm 92 “*9 u_ N to gzamgx

<

62

8.3 8.2 8.2 8m 23 8e 8N

 

 

 

 

 

 

 

 

0E: p..___<..»F_.p...h..—...r_g... rw._...L_..pL_..pr_._E-D
/ H
m?
mum .% -
3m fl
mm? H
Em: -8—
mtémm I u
.3 m2 m9 H_ E? mm mix
8.3 8.9 8.2 8m 23 9: 8d
...._»..-__...—..P._.-.._»_LL »»__...._..FL_...._LEID
/
m .
mmm H
mom .../m
_ Ema llv mos xxom vBEwBE .
$3. mNm Ilv ASEV 0:8 amaze 8 $34 .
3mm? . Ilv 2:: coucogom -2:
Egg o3 .. -
-mm m2 LE. U_ E2 mm 9x
m

3:538 EN oSwE

63

Automatic peak measurement and alignment were performed using Markerlynx fiom
Masslynx 4.1 software. The integrated peak area values (without scaling) were exported,
then normalized to peak area of internal standard propyl 4-hydroxybenzoate and fresh '
leaf weight.

By using LC/MS analyses performed on the Waters LCT Premier instrument,
multiplexed collision induced dissociation (CID) with accurate mass (<5 ppm) was used
to collect the molecular ions and fragment ions together in a single analysis using
experimental parameters common to all analyses. This approach, developed in the Jones
laboratory at Michigan State University, provides a powerful tool to discover, identify,
and quantify metabolites in nontargeted analyses. By using LC MS/MS with the Waters
Quattro Premier triple quadrupole mass spectrometer, daughter ion MS/MS scans were
used to ensure that observed fragment ions were generated by fragmentation of specific
molecular ions, which helps further identify and confirm the structures of metabolites
after LC separations. Daughter ion MS/MS scans were performed by selecting one
specific precursor ion (the molecular ion of a selected metabolite) in the first mass
analyzer and inducing it to fragment through ion-molecule collisions in the collision cell.
A spectrum of the fragment ions derived from the selected molecular ion was generated
by scanning the second mass analyzer, and these product ion spectra were used for
annotating the structure of the metabolite. The advantage of using LC MS/MS daughter
scan is that the precursor ion can be selected specifically without interference from
fragment ions generated from other coeluting metabolites. For more precise
quantification of target metabolites, multiple reaction monitoring (MRM) was used. This

approach uses the first mass analyzer to filter all but the selected molecular ion, and the

64

second mass analyzer allows only a selected and characteristic fragment ion to reach the
detector. This approach yields chromatograms selective for the target metabolite(s).
Details regarding MRM methodology and its performance in analysis of phytohormone

metabolites generated during wounding stress are discussed in Chapter 3.

2.2.7 Statistical analysis

Statistical analyses were performed on multivariate data using SIMCA-P (Umetrics,
San Jose, CA, USA) for Principal Component Analysis (PCA), Orthogonal Partial Least
Squares (OPLS) and Partial Least Squares for Discriminant Analysis (PLS-DA), and SAS

software for two-way Analysis of Variance (AN OVA) was applied as appropriate.

2.2.8 Chemical synthesis
Chemical synthesis of 5-hydroxyferulic acid (SHF)

S-Hydroxyferulic acid was prepared by dissolving 1 g of 4,5-dihydroxy-3-methoxy-
benzaldehyde with 2.9 g of diethyl malonate in 6 ml of dry pyridine and 600 pl piperidine
(as a base) and allowed to sit overnight. 30 ml of ice water was added and then
evaporated off using a rotary evaporator. This step was repeated once more using room
temperature water to ensure the removal of pyridine. A 10:1 ratio of 10% NaOH was
added to hydrolyze the ester groups, and the solution incubated for 2 days. After 2 days,
the solution was acidified using 88% formic acid to a pH of 2.0-3.0 and incubated again
for 3 days. The S-HFA was then extracted using methylene chloride. Methylene chloride
was evaporated under a stream of nitrogen, and product was redissolved in a 1:1

methanol/water mixture. Evidence for the structure of the 5-hydroxyferulic acid product

65

was confirmed using negative mode electrospray ionization mass spectrometry ([M-H]' at

m/z 209).

Chemical synthesis of 5-hydroxyferuloyl malate (SHFM)

SHFM was prepared by heating 1 g of malic acid with 5 g of acetic anhydride, and
then adding 25 ml of water to quench the acetylation. The mixture was then evaporated
using a rotary evaporator to remove any leftover acetic acid and acetic anhydride. This
was repeated using 10 ml of water. The acid was then dissolved in 150-200 ml of diethyl
ether in a 500 ml three neck round bottom flask. 1 ml of H2804 was added and mixed
thoroughly into the solution. The flask was placed in a dry ice bath with acetone and
cooled. Isobutylene gas was charged into the flask for approximately 5 minutes or until
the liquid volume increased sufficiently to guarantee that an excess amount was dissolved
into the ether, while maintaining the flask at cold temperature using the dry ice bath. The
flask was incubated for 2 days while maintaining the dry ice bath. The mixture was
recharged with isobutylene gas if the volume visibly decreased. After 2 days, the flask
was allowed to warm to room temperature, and the reaction was quenched with 25 m1 of
ice water and solid Na2C03 was added until bubbling ceased. The water layer was
removed, and 25 ml of water was again added and drawn off. The ether was evaporated
to dryness using the rotary evaporator, and the di-tert butyl ester of malic acid was then
re-dissolved in 50 ml of THF. Evidence to support the structure of the product was
confirmed using GC/MS at the MSU Mass Spectrometry Facility.

In the next step of the synthesis, solid 5-hydroxyferulic acid was incubated with 0.25

ml of SOC12 and 2 m1 of THF. Once all residues were completely dissolved, the reaction

66

was left at room temperature for 5-10 minutes. The di-tert-butyl malate was then added
as a THF solution, and mixed thoroughly. Reaction was incubated for 3 days and then
pH tested. TFA was added to hydrolyze the tert-butyl groups. Evidence for the structure
of the 5-hydroxyferuloyl malate product was confirmed using negative mode electrospray

ionization mass spectrometry ([M-H]’ at m/z 325).

Chemical synthesis of SHF quinone and SHF quinone-GS

Synthesis of SHF quinone was modified from previous methods on the synthesis of
caffeic acid quinone (Fulcrand et al. 1994). One milliliter of a 10 mM SHF solution (50%
acetonitrile in water) and 1.0 ml of a 20 mM NaIO4 solution (50% acetonitrile in water)
with 1% acetic acid were mixed and then incubated for 10 min at room temperature. Low
pH (pH < 4) was essential for this reaction and to minimize side reactions. The formation
and disappearance of quinones from phenols are dependent on solvent and pH. Quinones
disappear more rapidly in aqueous solution, because water could act as a nucleophile
towards the quinones and produce hydroxylated adducts. The quinone is stable in the pH
range 3-5 (Fulcrand et al. 1994). The product was passed through a C18 solid phase
extraction cartridge which retained the product while eluting inorganic salts. Following
loading on to the column, the cartridge was washed with 10%, 35%, 65%, and 100%
acetonitrile respectively and eluted within 65% acetonitrile, with each fraction analyzed
by mass spectrometry to determine product identity. The solution containing purified
SHF quinone was characterized using LC/MS to determine product pm'ity, and the
reaction product extracted twice into dichloromethane (v:v = 1:2). The combined lower

dichloromethane layers were collected, solvent was removed by rotary evaporation, and

67

the residue was kept dry for storage. Confirmation of chemical structure was by
electrospray ionization mass spectrometry. SHF quinone was dissolved in dimethyl
sulfoxide (DMSO) for introduction into bacterial cultures.

SHF quinone-GS was synthesized by incubating GSH with SHF quinone. A 1 mM
solution of SHF quinone in 50% acetonitrile in water was mixed with 10 mM GSH in
potassium phosphate (K2HPO4) buffer at pH 8 for 10 min at room temperature. The
mixture was passed through a C18 solid phase extraction cartridge for purification,
washed with 10%, 35%, 65%, and 100% methanol respectively. Flow injection analysis
mass spectrometry showed the desired product eluted with 35% methanol. The structure
and purity of the SHF quinone-GS product was confirmed by LC/MS. The purified
product was evaporated to dryness under a stream of nitrogen and finally dissolved in

dimethyl sulfoxide (DMSO) for bacterial culture.

2.2.9 Cloning, Expression and Characterization of OMTl
2.2.9.1 Cloning of OMT] gene

The DNA encoding OMTl was amplified from an Arabidopsis Columbia-0 ecotype
(Col-O) cDNA library by PCR using forward primer: CATG CCATG GGT TCA ACG
GCA GAG ACA CA (NcoI site in bold), and reverse primer: CCG CTC GAG CTT CTT
GAG TAA CTC AAT AAG (XhoI site in bold). An extra glycine near the N-terminus
and one extra leucine and glutamate near the C-terminus were created together with the
C-terminal 6 X His-tag during cloning. The PCR amplification consisted of 35 cycles of
denaturing at 94 °C for 45 sec, annealing at 53 0C for l min, and elongation at 72 0C for 1

min and 30 sec, followed by 72 °C for 10 min. The PCR products were purified by

68

QIAquick PCR purification kit (QIAGEN, CA, USA) and digested with N001 and XhoI.
The pLWOl expression vector (Bridges, Gruenke et al. 1998) digested with NcoI and
XhoI, and ligated with the digested PCR product, was transformed into competent E. coli
DHSa cells. Positive clones growing on LB plates containing 100 ug/ml arnpicillin were

picked, and plasmid DNA was isolated and sequenced.

2.2.9.2 Expression and purification of GMT]

The sequenced plasmid was transformed into expression host E. coli BL21 (DE3)
competent cells. A single colony from a freshly-streaked selection plate was used to
inoculate 100 ml of LB medium containing 100 ug /ml ampicillin and was incubated
overnight at 37 °C with shaking at 200 rpm. Twenty ml of this culture was transferred
into 1L of fresh LB medium containing antibiotics and the cells were grown at 37 °C to
an A600 of 0.6-0.8. The cells were then induced by adding 0.1 mM IPTG and incubated
with shaking at room temperature for 17 hours. Cells were harvested by centrifugation
and stored at —80 °C.

To purify OMT], the cell pellets were resuspended in Buffer A (50 mM sodium
phosphate, 300 mM sodium chloride, 250 mM sucrose, 10% glycerol, 10 mM [3-
mercaptoethanol, and 0.1 mM EDTA, pH 8.0). After sonication, the crude cell extract
was centrifuged at 4 0C for 20 min at 12,000 g. The supernatant was loaded onto a pre-
equilibrated column containing 20 ml Ni-NTA agarose slurry. The column was washed
with buffer B (50 mM sodium phosphate, 300 mM sodium chloride, 10% glycerol, 10
mM B-mercaptoethanol, and 20 mM imidazole, pH 8.0). Protein was eluted from the

column using Buffer C (50 mM sodium phosphate, 300 mM sodium chloride, 10%

69

glycerol, 10 mM B-mercaptoethanol, and 200 mM imidazole, pH 8.0). The protein-
containing fractions were pooled and concentrated to 1 ml using an Amicon 30000
MWCO ultrafilter. To further purify the OMT], the protein was loaded onto a 1 ml
Hightrap Q ion exchanger, eluted with a gradient based upon buffer D (20 mM Tris-HCl,
pH 8.5) and buffer B (20 mM Tris-HCl, 1 M sodium chloride, pH 8.5) at a rate of 1
ml/min. To determine the oligomeric state of OMTl , size exclusion chromatography was
performed. A Superdex 75 10/30 GL column was equilibrated with buffer F (20 mM
Tris-HCl, 150 mM sodium chloride, and 10% glycerol, pH 7.5). The protein was loaded
and eluted at a flow rate of 0.5 ml/min. Fractions containing homogenous OMTl were
combined and concentrated as above. Protein concentration was determined by the
method of Bradford (Bradford, 1976) with bovine serum albumin as the standard. Precast
NuPAGE 4-12% Bis-Tris polyacrylamide gels (Invitrogen, CA, USA) were used for

SDS-PAGE. Proteins were visualized with Coomassie blue staining.

2.2.9.3 Mass spectrometry analysis of OMT] molecular weight (MW)

The molecular weight of purified OMTl was analyzed using positive mode
electrospray ionization on a Waters LCT Premier time-of-flight mass spectrometer
coupled with Shimadzu LC-20AD HPLC pumps and a SIL-SOOO autosampler. Separation
was performed using a Thermo BetaBasic cyano column (1 x 10 mm) with a gradient of
0.15% aqueous formic acid and 75% acetonitrile over 10 min for online desalting and
elution. MassLynx data system (version 4.1) (Waters Ltd, Manchester, UK) provides

instrument control, data acquisition and data processing. The protein molecular mass is

70

calculated based on spectrum deconvolution to a zero charge state spectrum using the

MaxEntl algorithm.

2.2.9.4 Enzymatic assay

The activity of recombinant OMT] was determined by measuring the decrease of
substrate concentrations using LC/MS. Assays were carried out in Buffer G (100 mM
Tris-HCl, 150 mM sodium chloride, pH 7.5) at 30 °C. The 500 pl reaction volume
contained 1 mM SAM, various concentrations of substrate and purified enzyme (IO-25
ug/ml). The reaction was stopped by adding 50 ul quench solution containing 1% formic
acid in methanol: water (1 :1, v/v). The control sample containing all assay components
was stopped at time 0. Fifty ul was withdrawn from the reaction after 60 seconds and
tranfered to autosampler vial for LC/MS analysis. The decrease of substrate
concentrations and the formation of enzymatic reaction products were identified and
quantified using the LCT Premier time-of-flight mass spectrometer (Waters, Milford,
MA, USA). Negative mode electrospray ionization was used for all samples. Fast
separation of analytes as performed by a Therrno BetaBasic C18 (1 x 150 mm, 5 um
particles) column using a gradient of 0.15% aqueous formic acid and methanol over 10
min for every substrate. Accurate mass (<5 ppm) LC/MS data were collected using
multiplexed collision induced dissociation (CID) acquisition through modulation of the
Aperture 1 potentials. Spectra were acquired under two different CID conditions (10 and
50 V) in a single analysis to get fragmentation information for structure confirmation.
Accurate mass, fragmentation pattern, and retention time were used for identifying the

substrates and products. Substrates were also quantified on the basis of integrated areas of

71

the peaks in extracted ion chromatograms of [M-H]' ions, using linear calibration curves
generated using each standard compound for known concentrations. Peak integration and

quantitative analyses were performed using Quanlynx from Masslynx 4.1 software.

2.2.9.5 Determination of kinetic properties

Substrates at different concentrations, ranging from 1 uM to 100 uM, were prepared
for the determination of Km values of purified OMTl. LC/MS based assays were carried
out with the same amount of enzyme against initial velocity of the purified enzyme. Vmax
and Km were obtained through nonlinear curve fitting model with the Michaelis—Menten
kinetics equation of V=Vmax*[S]/(K"fi[S]) using Origin 7 software (OriginLab,

Northampton, MA, USA).

2.2.10 Determination of antibacterial activity

The procedure is modified from previous report (Franklin et al. 2009). Both the plate
and liquid media were augmented with 100 ug/ml rifarnpicin. For the disc diffusion
assay, sterile paper discs loaded with 10 ul of two different concentrations (2 uM and 20
pM) of 5-hydroxyferulic acid quinone conjugate (SHFQ), 5-hydroxyferulic acid quinone
glutathione conjugate (SHFQ-GS), 5-hydroxyferulic acid (SHFA), and sinapic acid (S)
were placed onto Luria-Bertani (LB) media spread with 100 pl P. syringae DC3000
suspension (5 x 107 cfu/rnl). DMSO was used as the negative control and H202 was used
as the positive control for antibacterial activity. For each plate, 10 ul of DMSO, 10 mM
H202, 2 uM and 20 uM chemicals were added to each comer on one single plate and

cultured at 28 °C for 24 h. A series of different concentrations of H202 (100 mM, 10 mM,

72

1 mM, 0.1 mM, 0.01 mM) was loaded on one plate to see which concentration of H202
can inhibit bacterial growth successfully.

For suspension bacterial culture, 100 pl of chemicals or controls were added into 2m]
media and maintained in a shaking incubator at 200 rpm at 28 °C for 12 h. After 8 h, 100
pl aliquot from each sample was serially diluted to 10'6 and spread on plates. The colony
forming units (cfu) on each plate were counted after incubation for 24 h and percentage

of viability was calculated.

2.2.11 Extraction of apoplast sap by centrifugation

The plant above ground level was cut and soaked in buffer (50 mM NaPO4, 150 mM
NaCl, pH 7.0). The apoplast was infiltrated with the buffer using vacuum infiltration
(Katagiri et al. 2002). The plant tissue was centrifuged at 1,000 g for 15 min. Repeated
centrifiigation was occasionally required to remove all visible liquid from the leaves. The
residual liquid removed during centrifugation contained a solution of Arabidopsis

apoplast sap.

2.2.12 H202 measurement

The standard curve of H202 (0-25 MA) was made using FOX assays (Delong et al.
2002). The FOX reagent contained 90 ml MeOH (90% (v/v), 10 ml of 250 mM H2304
(10%), 88 mg BHT (4 mM final), 9.8 mg ferrous ammonium sulfate hexahydrate (250
M final), and 7.6 mg Xylenol Orange (100 M final). The reagent was mixed well with
different concentration of H202 (0-25 pM) and measured under OD560nm. A standard

curve was drawn based on OD values measured. The water extracted leaf samples were

73

also mixed with FOX reagent and measured under ODséonm, 0D values were used to

quantify H202 based on the standard curve.

2.3 Results and discussion
2.3.1 Identify the principal OMT genes responsible for phenolic metabolism in
Arabidopsis thaliana leaves

Before exploring OMT gene functions in pathogen stress, amt mutant plants were
studied to understand basic functions of OMTs without stress in this study. To identify
OMT genes that significantly affect phenolic metabolism in Arabidopsis, amt T-DNA
insertion mutants were screened using nontargeted metabolite profiling to identify
differenes in metabolite profiles associated with each mutation. Sixteen different amt T-
DNA insertion mutant plants (13 OMT genes), as shown in phylogenetic tree of
Arabidopsis OMT family in Figure 2.8, were cultured in a growth chamber (12 h light/ 12
h dark, 105 uE m'2 s'1 light, 22 °C) for five weeks. The metabolites were extracted from
leaf tissues of individual plants and analyzed using LC/MS. The metabolite profiles in
leaves of all mutants were compared to the metabolite profiles of wild type plants after
chromatographic peak integration and retention time alignment using MarkerLynx
software. Principal Component Analysis (PCA) showed that a cluster of the amt]
(At5g54160) mutants, including both alleles tested, separated from wild type and other
mutants (Figure 2.9). Therefore, OMT], which appears to have the most profound
difference of metabolite profiles compared to wild type plants, was chosen for fiirther

study.

74

 

AT3053140.1 (omt13)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Armsmr (omt14)
, 1m Anorarwr
‘ 11120410401
{.00 Ammo]
AT5650240.1
100 151150083361
AT5G23320.1
" 9" ATdGm.1(°mt15)
r AT4G340£111(omt16)
68 AT1024735.1
._ 100 72 1016579811
TRICI-DTOMY A11667990.1
r AT3661990.1
1 1m ATBGGZJGJ.1(°mt4)
* Ancramr
—L100 T 11110211001
1m * l—— Atrcznsm
Atscszmnr (omt5)
"fin Atmmr
'— AT1C£31101
— 100 AT56311?U.1(omt6,7)
7‘ m 14116775201 (omt8,9)
, AT1G77530.1
. * AT16311301 (omt3)
‘LFBZ Amalgam (omt2)

 

 

AT5654160.I(omt1 or omt1-1, omt10 or omt1-2)

[—1.0]- ATlB515U.1(°mt11)
'-— AT4635160.1(omt12)

Figure 2.8 Phylogenetic tree with annotation of the 16 OMT T-DNA insertion mutant
lines with 13 amt mutant genes in the phylogenetic tree of Arabidopsis OMT family.

75

.392 E E38 a Amhmmoonmémv 8:03. Ea 3e: 3 8an a sammnrvfiév 8:33.
mo 22? ESE: SE85 2:8 BE. 68: 803 “53:5 i=8 mo v.22? ESE: 5983 03,—. 6:538 $50 E“ was 093 BE 58m
@8838 £qu 05 mo 02m c2 2: @833 .6530 353 E82: 2:8 95. .mEosogEoo REEF—Q 0E «Em o5 mitt/2m 93 £30:
mi? 3532: meta“: 2.36 3:32: :oEomE <75? 38 m: was C53 093 2:5 Set 8% 823805 mo SE 3.88 <Um ad DSwE

2:6 owed coed owed- cord. 02.0. oomd- ommd. oomd- ommd- cows...
EchQEDO.L:......L::T:.w...._:.;._:~, . a

:___..

 

,._._._.:___.::.LT:,._.._._:» [b.culLLL._____._ , __::...._
. . . . 4 4 4 l J1 . . _
. , . _ , . .

      

 

    

 

” 0
u 0
, m
_ 0
u U
_ a
_ _ w
_ a _ . V Z
m n .1. 9-”: at? .1 .5ng 3
. _ _ . W ..... ..... ..... ..... U ..... _.,...._..~.._2...H .....
..U ......... ..... _ ..... . ..... m ..... . ..... m ..... H. ..... m ..... a ..... m ..... H. ..... m ..... h. ..... m ..... sad

N 228800 - FEocanEoo 62on

76

2.3.2 Confirmation of homozygous T-DNA insertion lines using PCR, RT-PCR, and
sequencing

Genomic DNA (as the template) extracted from the leaves of A. thaliana T-DNA
insertion mutant line of At5g54160 (SALK_135290) was subjected to PCR using specific
primers (primer design is shown in Method 2.2.2). PCR products were separated on 1.2%
agarose gel (Figure 2.10). Only one band for PCR products (~ 500 bp) amplified using
primers of RP and LB was observed, which confirmed the homozygous T-DNA insertion
of the SALK line (Figure 2.10A). By using the primers of LP and RP for the wild type, a
PCR product with the size of 1091 bps (the size of OMT] gene) was observed (Figure
2.10B). However, for the homozygous T-DNA insertion mutant, there is no product since
the normal Taq enzyme cannot amplify the template with size of more than 2 kb for

OMT] gene with T-DNA insertion.

77

1a 1b 2a 2b Mw WT MW

1000
800
650
500
400
300
200
100

   

OMT1 (a, LP+RP; b, RP+LB)

Figure 2.10 PCR products separated on 1.2% agarose gel. (A) Lane a, PCR products
amplified using primers of LP and RP for the homozygous T-DNA insertion of the SALK
line. Lane b, PCR products amplified using primers of RP and LB the homozygous T-
DNA insertion of the SALK line. MW, 1 kb DNA ladder. (B) Lane WT, PCR products
amplified using primers of LP and RP for the wild type. MW, 1 kb DNA ladder.

Total RNA extracted from the leaves was submitted to reverse-transcription (RT)
PCR, so that the absence of functional transcripts was confirmed. RT-PCR products are
separated on 1.2% agarose gel (Figure 2.11). The positions of primers for RT-PCR are
shown in Figure 2.12. The absence of functional transcripts was confirmed since no PCR
product was produced for the mutant. The product is only amplied fi'om cDNA, but not
genomic DNA since the size of the product from DNA is much larger than 448 bp (more

than lkb) with the intron.

78

WT WT OMT1 MW

 

Figure 2.11 RT-PCR products separated on 1.2% agarose gel. Lane WT, PCR products
(448 bp) for the wild type. Lane OMT1, no PCR product was produced for the mutant.
MW, 1 kb DNA ladder.

The PCR amplification product generated using primers of LB and RP was purified
with QIAquick Gel Extraction Kit (Qiagen, CA, USA) and sequenced in the Research
Technology Support Facility (RTSF) at Michigan State University. The sequence was
BLASTed with genomic DNA and the T-DNA insertion position was found at second

exon 1299-1609 bp with position of 1319 bp, as shown in Figure 2.12.

79

OMT1 salk 135290 (At5g54160)

 

LB
ATG TAG
—£:—
‘— — 100bp
—> RP
RT-PCR primer ‘—

RT-PCR primer

Figure 2.12 T-DNA insertion position for amt] gene (At5g54160). T-DNA insertion
position was found at second exon 1299-1609 bp with position of 1319 bp.

2.3.3 Phenotypes of amt mutant compared with the wild type

When homozygous amtI(At5g 54160) mutant plants were cultivated under control
conditions, there was no discemable morphological phenotypic difference between the
wild type and mutant plants; their growth and fertility were comparable to those of wild
type Arabidopsis. To quantify bacterial virulence, bacterial growth within the host tissue
was assayed based on a bacterial pathogen enumeration procedure (Katagiri et al. 2002).
Under pathogen stress, a 5-fold decrease in colony forming units was observed for the
amt] mutant plant leaves relative to the wild type after three days infection (Figure
2.13A), and amt] mutant plants showed slower infection rate than the wild type (Figure

2.133).

80

5.00E+08 j

l

 

 

A
4.00E+08 .
3.00E+08 —
or *+wr
\E +omr1-1
a + omt1-2
° 2.00E+08 -
1.00E+08 4
o.ooe+00 : . .
0 1 2 3

i

 

Figure 2.13 Bacterial growth on the wild type and the amt] mutant plant leaves. omt1-1
stands for insert line of SALK_135290 and omt1-2 stands for insert line of
SALK_002373. (A) Leaves were inoculated with 1 x 106 cfir/mL of P. syringae DC3000
and in planta bacterial populations were determined daily. The error bars indicate the
standard deviation within the 3 replicate samples for each treatment. (B) Leaf

appearances at 3 day infection with P. syringae DC3000.

81

To study OMT] gene function in pathogen stress, employing a virulent strain of P.
syringae offers advantages relative to use of an avirulent strain. Many resistance genes
are induced in the interaction between plant and avirulent pathogen strain, but the virulent
pathogen avoids activation of the plant’s hypersensitive response. Therefore, this report
studied the responses of amt] mutant with virulent pathogen stress to avoid the

complexity of activation of resistance factors.

2.3.4 Comparison of amt mutant lines with two different mutant alleles in At5g54160

Mutant lines of two amt] alleles of At5g54160 (SALK_135290 and SALK_002373)
showed indistinguishable phenotypes of growth, bacterial virulence, and metabolite
profiles from one another. Both displayed no recognizable morphological phenotypic
difference relative to wild type without stress. Both mutant alleles have similar bacterial
virulence after pathogen inoculation, as shown in Figure 2.13. PCA display of metabolite
profiles is shown in Figure 2.9, with the amt] insertion mutant line of At5g54160
(SALK_135290) marked as MT1_1 and At5g54l60 (SALK_002373) marked as MT1_2

grouping together but separated from wild type and other mutants.

2.3.5 LC/MS based metabolic profiling to discover new metabolic phenotypes for
gene function prediction
2.3.5.1 Data processing and statistical analysis for discovery and visualization of
interesting metabolites

Automatic peak measurement and alignment were performed using Markerlynx from

Masslynx 4.1 software. The integrated peak areas (after Pareto scaling) were then

82

exported to statistical software SIMCA-P or SAS. Statistical analyses based on Principal
Component Analysis (PCA), Partial Least Squares for Discriminant Analysis (PLS-DA),
Orthogonal Partial Least Squares (OPLS), and two-way ANOVA were used to aid
recognition of those metabolites are most affected by genetic or environmental changes.
PCA is an unsupervised data reduction technique that is used to reduce the
dimensionality of a multi-dimensional dataset while retaining the characteristics of the
dataset that contribute most to data variance. PLS-DA is a supervised regression
extension of PCA that takes advantage of sample class information to maximize the
separation between groups of observations from different sample classes. OPLS is a
modification of the usual PLS method that filters out variation that is not directly related
to the response (Umetrics SIMCA-P 11.0 software website).

As shown in the Figure 2.14A, PCA scores plot showed segregation of metabolite
profiles of different genotypes of the wild type and the mutant. The PCA analysis is an
unsupervised statistical method, and the separation between groups in the scores plot is
weighted heavily on abundant metabolites. An alternative supervised method such as
OPLS-DA can draw attention to less abundant metabolites whose levels provide effective
discrimination between sample classes. For example, a comparison between the wild
type and the mutant using OPLS-DA (Figure 2.14B) generates a 3D plot using the first
three components that shows clearer separation of wild type and mutant, since OPLS uses
sample class information to sharpen the separation and minimize variation that is not
directly related to the response. An OPLS S-plot (Figure 2.14C) was used to highlight
those metabolites whose levels most distinguish wild type and mutant plants. The S-plot

is a loading plot to show the weighting coefficients of X-variable, here referring to the

83

influence of one single metabolite with a specific retention time and mass to charge ratio
(m/z) value (such as 5.88_325 .0 on the S-plot means signal with retention time of 5.88

min and m/z 325.0).

84

Figure 2.14 Statistical analysis of LC/MS metabolite data (using Umetrics SIMCA-P 11.0
software) from the wild type (w) and camtI mutant (m) leaves without stress. (A) PCA
shows segregation of metabolite profiles of w and m. (B) Orthogonal partial least squares
(OPLS) 3D plot is displayed using the first three components and showed better
separation of classes of the wild type and mutant, since OPLS takes the advantage of
class information to sharpen the separation and filters out variation that is not directly
related to the response. (C) OPLS S plot was used to rank those metabolites whose levels
changed most between wild type and mutant plants. The S plot is a loadings plot that
shows the weighting coefficients of X-variable, here referring to the influence of one
single metabolite with a specific retention time and mass/charge (m/z) value (such as
5.88_325.0 on the S plot means signal with the retention at 5.88 mins and m/z 325.0). X-
axis shows the p score for covariances and y-axis shows the p(corr) score for correlations.
Signals with the greatest y-axis values have high p(corr) scores which are measures of
confidence in differing between the m and w samples. 0n the top-right side, it shows the
signals which are more abundant in mutant (m) than the wild type (w). On the bottom-left
side, it shows the signals that are more abundant in w than m. (D) Loading list of top 35
spectra sorted by descending p(corr) scores. The list at the right shows the spectra from
the top-right side in (C). The list at the left shows the spectra from the bottom-left side in
(C). Signals with m/z 325 and m/z 339, m/z 371 and m/z 385 are different in molecular
mass by 14 Da (the mass change upon adding a methyl group) and are significantly
different in mutant and wild type.

85

00—.

23.8 we mafiaouom "03.3.6
:60de H HNHme mmvmmmé u 72me

r:

om ow on ow om 0v om ON 0.. o 9.. om- om- 0v- om- om- ON- ow. om-

 

0v.

ON-

 

 

 

 

ON

0v

ow

 

 

fl .3085? .anou
Ax.<on= :20: 9 03

[z]:

86

 

new. ..

 

m
82:38 3N aSwE

87

mnmmomé n 72me

 

       

 

 

 

 

 

aEa
to md Nd to 0.0. Pd- Nd- md- v.0. md-
. C.Fl
8 «-3 m. 8 33;. 8039.9
Seaman.
2.5 2 9. ad-
95: Baffle m.
.. .. .wfiwfiodwo. g-
...-V... . ..
. v.0l
“ .. .ImVIm.
... .8m 8:
._ Nd-
N
.3- m
w
83 am rm
No
884.30. I 8020 3
38. i. m . to
Smmwmummm.
9 Ed” m. ed
image”.
:.m8.®.§m-8.o.
2 83: m. w o
. - 30.002.03-39
8.83m m. 5 m2 8% o;
= ...sooxtooxs. .28: 0

8:580 2 .N BEE

88

Figure 2.14 continued
D

More abundant in the wild type (w) More abundant in mutant (m)

 

 

 

 

 

 

 

 

89

Var ID (Primary) M3. p(corr)[1]P Var ID (Primary) M3. p(corr)[1]P

[6.12_339.0696 -0.91875 5.89_209.0506 0.987803
6.10_223.0650 0.91824 5.8§_ 325.0564 0.98465]
5.25_359.0989 0.91385 5.96_133.0182 0.959144
5.88_350.0869 0.86152 [ 5.53_371.0958 0.9410661
6.19_677.1227 0.82063 6.03_665.1243 0.923777
5.87_175.0338 0.7639 6.59_503.1150 0.912133
6.02_817.1290 0.75747 5.78_563.1331 0.818733
l 5.70 385.1152 0.68327 I 6.50_533.1233 0.737952
5.79_352.1071 0.65813 5.98_803.1141 0.72801
6.25_709.1529 0.64474 5.98_542.1432 0.648108
5.25_405.1070 0.63319 6.34_640.1421 0.64545
6.26_917.2190 0.58419 5.81_889.1912 0.631097
5.99_172.1000 0.57378 5.8o_711.2322 0.626145
6.18_179.0755 0.55862 5.82_773.1082 0.583353
6.o1_556.1589 0.55171 5.52_792.1418 0.524987
2.61_565.0407 0.54219 6.59_313.0980 0.508955
6.47_353.0961 0.53471 5.15_343.071o 0.457385
5.38_323.1334 0.53214 5.25_345.0823 0.45498
5.82_389.1770 0.5229 1.16_102.9928 0.451123
6.53_677.1285 0.51223 7.42_801.4026 0.440745
5.87_244.0652 0.5096 6.31_903.2083 0.418297
5.51_241.0873 0.46678 5.89_630.1342 0.416202
6.2o_437.0452 0.48299 5.56_819.1454 0.414853
6.18_149.0265 0.48124 5.86_323.0583 0.411867
5.77_739.1946 0.45576 6.58_675.2071 0.40975
6.17_164.0520 -0.42867 7.43_873.4580 0.40743
7.68_836.4308 0.42053 7.01_462.0874 0.384537
7.67_819.4406 -0.42016 5.90_649.0941 0.381557
8.71_226.0635 0.41494 7.67_61.9897 0.38131
5.77_785.1991 0.41151 5.77_165.0279 0.380935
7.65_854.4417 0.41015 7.65_531.2759 0.371233
8.87_99.9298 0.40828 5.41_420.0425 0.366732
2.67_224.0602 0.39984 6.27_289.0948 0.362378
6.23_338.0662 0.39349 6.47_402.0892 0.358752
5.77_775.1730 0.39118 7.40_785.3805 0.354099

To calculate the probability (p) values that metabolite levels differ between sample
classes, two-way ANOVA with mixed model and student t-test was performed using SAS

software.

2.3.5.2 Interpretation of metabolite data

More than 3000 metabolite signals were detected in LC/TOF analyses of most leaf
extracts. These signals were thought to come from more than 500 distinct compounds. A
total of about one hundred compounds with five hundreds peaks were identified by
annotation with chemical structures. Major groups are polyphenols (flavonoid and
phenolic acid derivatives), lipids (oxylipins), glucosinolates, phytoalexins, and the
bacterial metabolite coronatine. Metabolites showing significant differences in their

levels in the various genotypes or treatments (p<0.05) are discussed below.

Result 1: Comparison of leaf metabolite profiles in WT and amt mutants without
treatments (W vs M) to characterize the OMT function in absence of stress

One early question to be answered was whether the amt mutants showed evidence of
OMT substrate accumulation or analogs of metabolites downstream from methylation
events that might accumulate in the mutants. Nontargeted metabolite analysis using
LC/T OF MS was performed for 16 OMT T-DNA insertion mutant lines with 13 mutant
genes (Figure 2.8) to help recognize metabolites influenced by OMT activities (Figure
2.9), to establish differences between two alleles of amt] mutants (Figure 2.9), and to
assess whether other pleiotropic effects of T-DNA mutations on metabolite profiles could

be observed (Figure 2.14).

90

To identify putative substrates and products of OMT1, data analysis of metabolites
aimed to identify metabolites whose spectra are different in molecular mass by 14 Da (the
mass increase upon replacing a hydrogen with a methyl group) across sample classes. For
example, a pair of signals at m/z 325 and m/z 339 was selected from the complete
metabolite lists with p(corr) and p scores for each signal list after calculation using
OPLS-DA, and also a prominent pair with m/z 371 and 385 (Figure 2.14D). For each pair
of metabolites, the lower mass metabolite (m/z 325 and m/z 371) were more abundant in
amt] mutants, whereas the higher mass metabolites (m/z 339 and m/z 385) as prevalent
in wild type. Signals with m/z 325 and m/z 339 have the highest OPLS-DA p(corr)
scores on the loading lists that exhibit the most significant differences between the wild
type and mutant. Compounds with molecular ions [M-H]' of m/z 325 and m/z 339 were
identified as S-hydroxyferuloyl malate (SHFM) and its methylation product sinapoyl
malate (SM), respectively, based on accurate mass measurements and fragmentation
patterns. The measured mass for SHFM is m/z 325.0564 (calculated mass is m/z
325.0560). The major fragment is m/z 209 for deprotonated 5-hydroxyferulic acid
fragment ion and m/z 133 for deprotonated malate fragment ion. These fragments lend
support to structure assignment as SHFM. The measured mass for SM is m/z 339.0696
(calculated mass is m/z 339.0716), with major fragments at m/z 209 for deprotonated
sinapic acid and m/z 133 for deprotonated malate. Relative quantification of SHFM and
SM in the wild type and the OMT1 mutant is shown in Figure 2.15A. In the amt] mutant,
SHFM showed a ten-fold increase and SM showed a three-fold reduction compared with
the wild type. The continued presence of SM in the amt] mutant is indicative of an

alternative methylation pathway. This result suggests that the loss of OMT1 activity

91

decreases, but does not eliminate, methylation in the 5-position that yields SM, and points
to the possibility that OMT1 may play an important role in methylation of 5-
hydroxyferuloyl metabolites. OMT activity toward malate esters has not previously been
determined, in part due to the unavailability of malate ester standards. Proposed 0-
methylation by OMT1 from SHFM to SM is shown in Figure 2.16A. In addition to the
malate esters, compounds with molecular ions [M-H]° of m/z 371 and m/z 385 were
identified as 5-hydroxyferuloyl glucose (5HF G), present in amt] mutants, and sinapoyl
glucose (SG), abundant in wild type, respectively. Relative amount of SHFG and SG in
the wild type and the OMT1 mutant is shown in Figure 2.15B. In the amt] mutant, SHFG
showed a ten-fold increase and SG showed a five-fold reduction compared with the wild
type. These results are consistent with a relationship between 5HFG and SG being
substrate and product catalyzed by OMT1. A scheme depicting O-methylation by OMT1

from 5HFG to SG is also shown in Figure 2.16B.

92

I>

   

SHFM SM
500
1800
400
1200
son
zoo 000
100 400
0
WT MT °
B
20 18
1a SHFG 12 SG
12
a
o
4 4
o o
wr MT wr MT

Figure 2.15 Relative quantification of (A) 5-hydroxyferuloy1 malate (5FHM) and
sinapoyl malate (SM) (B) 5-hydroxyferuloyl glucose (5HFG) and sinapoyl glucose (SG)
in leaves of wild type (WT) and amt] (AT5G54160) mutant (MT) Arabidopsis thaliana.
Y-axis is integrated XIC peak areas / mg fi'esh leaf weight, which is obtained by dividing
the peak areas of metabolites to the internal standard propyl 4-hydroxybenzoate and
normalizing to wet leaf weight. Data are means 1; SE (N=8).

93

.6@ 882m 328:6 2 Gummy 883w _>o_=._£>xo._cb_-m Sea Amv
can 92% 229: 3092.6 8 ASEIS 223: So—Eomzxohiné Sea 9b 2.20 3 BSA—Emu mucus—>5080 88395 EN PSME

 

 

3:00 ...:00
I0 OI A £20 I 0 OI
/ m
_._o o : T_o O
o / com: o / o...
:o o :0 0
I0 I0 m
£00 £00
I000 CI :20 1000 OI
A / m O
#00 com: o I J\ / 0:
I000 I000 0

94

Biochemical characterization of OMT1: Cloning of the omt1 gene and assessment of
its biochemical activities using an assortment of polyphenolic substrates

To confirm the catalytic activity of OMT1 toward various polyphenol substrates, the
omt1 gene was cloned from cDNA and over-expressed in E. coli as a C-terminal His-
tagged protein. OMT1 was purified and characterized using 12 polyphenolic substrates
including the putative substrate, SHFM, which was synthesized in our laboratory.

OMT1 was annotated as caffeic acid/S-hydroxyferulic acid OMT based on its high
sequence similarity to Populus tremuloides CAOMT (Zhang et al. 1997), but current
annotation (http://www.uniprot.org/) for accession #Q9FK25 describes it as quercetin 3-
O-methyltransferase. Muzac, et al. (2000) reported that OMT1 catalyzed methylation of
quercetin and myricetin when it was expressed as a recombinant protein in E. coli, but
activities toward the hydroxycinnamic acids caffeic acid and 5-hydroxyferulic acid were
reported as minimal (< 5% of activity toward quercetin). Values were not provided for
VW or Km. In contrast, support for the earlier annotation came from analyses of crude
enzyme activity in the stern extracts from an omt1 mutant, which found that the activity
for methylating caffeic acid and S-hydroxyferulic acid was decreased to about 7% of the
wild type for both compounds (Goujon et al., 2003). This suggested that OMT1 may be
involved in the catalysis of caffeic acid/S—hydroxyferulic acid. However, neither of these
metabolites are known to be abundant in leaves of Arabidopsis ecotype Col-0. To
resolve these differences regarding substrate specificity and better understand the
functions of OMT1, this gene was cloned and expressed in E. coli and purified. The
recombinant protein was used to measure catalytic activities and affinities for caffeic

acid, S-hydroxyferulic acid, and luteolin in the current study.

95

Cloning and over-expression of the amt] gene

The OMT1 gene was amplified from cDNA of Arabidopsis Col-0 using PCR, and was
directly cloned into the 6)< His-tagged pLWOl vector. OMT1 was expressed in E. coli
BL21 (DE3) and purified using Ni-NTA column chromatography to obtain >600 mg/L of
pure protein (Fig. 2.17A). The protein was further purified using a 1 ml Hightrap Q ion
exchange column. OMT1 was eluted at 15% - 20% sodium chloride and was purified to
homogeneity based on SDS-PAGE (Fig. 2.17B). Using ESI-TOF MS, the mass of OMT1
was identified to be 40,436 Da (Fig. 2.18), which is consistent with that predicted for
His-tagged protein (40,441 Da). Size exclusion chromatography showed elution of
OMT1 as 80 d: 5 kDa, consistent with a dimeric native state (Fig. 2.19). This finding
stands in contrast with the report by Muzac et al. (2000) that native recombinant OMT1
existed in solution as a monomer. The solved crystal structure of OMT from alfalfa
suggested that it forms a dimer in solution and the dimeric arrangement is critical for its
activity (Zubieta et (al. 2001; Zubieta et al. 2002). The dimeric form of the protein is

thought to exclude solvent from its active site, which is critical for substrate binding.

96

 

kDa
7!

H511
£0111
“PC
was
N

   

M W“ W“ W E1 52 E3 E4 E5 Retentionvol 13 14 15 16 17 18 19 20
)

Figure 2.17 Purification of OMT1. (A) SDS-PAGE of OMT1 purified from Ni-NTA
column. M, marker BioRad precision plus protein standards; Wh, whole cell lysate; Lsn,
supernatant of low speed centrifugation; Ft, flow through from Ni-NTA column; Wa,
wash fraction from Ni-NTA column; E1-E5, elution 1-5 from Ni-NTA column. (B) SDS-
PAGE of fractions containing OMT1.

97

48320 was wcufimoc 25:0 .80 £8 3 ~96 252803 {own was Eon 288m 38:3 $26 mo 333% a 53» A58 2 x 3 5:38
958 ommamSom BEBE. a mafia uogoton $3 538.80% d0 omvdv mm 8.20 .«o mane BEE—Bow baggage 2E. .8th8
3:982 mama: 30202—38 85.50% 88% ombfio PEN Amv .836 owbfio 03:38 wag/0% _ .520 wafiémm .00 83.58% mmmE
83363 >8%oboo_m A3 dou£o>=ooov 55.30% 3:982 98 m2 "5.75m .3 35::83 _ 520 no 358 315202 ”IN an

98

  

898mg: x
mammm
+mmmzmob

$55:

52.29 .5 “I 883 a 383 3% 2.5.3

<.

 

99

—I—_vo-l—.\o {nu-w
a ,1— ...

 

 

 

 

 

$2. . .1. . _ . . hammer. . _ . . .mommvt . L . [0.8%. . . _ . . .82.... t _ . . .83... . L . . .pommmt I... 531%.... . . .o.o__fim.i;...8m_m.mii. 20.08%
t (.E fl 4531 4 1 1
age
388 I
\
mmvoov
I
0
... w
\ w
ménvov
m
8.38.851 2x ommvov .
E: a. - < 3:23:00 M: N oezmmm -9:

And/xiv I:

GIN.

l.‘

!\'\:III‘II<!II“

D

2450 '1

1950 '

 

 

 

 

A

D

3

I

C

8 1450--

N

4.)

'° 1

o 950

U

G

.2

1.1 450~~

O

U!

a A

< -50* r
o 5

Retent ion Volume (ml)

Figure 2.19 Size exclusion chromatography (with Superdex 75 column) showed elution
of OMT1 as 80:5 kDa, consistent with a dimeric native state. The blue peak is OMT1.
The green and red peaks are protein standards with sizes of 67 kDa, 43 kDa, 25 kDa and
13 kDa.

The optimum pH value for the activity of the enzyme was determined to be in the
range of 7-8 and the enzyme is stable when stored at -80 °C with 10% glycerol and 1 mM
DTT. Thirteen different putative substrates for OMT1 were screened for activity (Figure
2.20 and Table 2.1). OMT1 catalyzes the transfer of the methyl group of SAM to the 3’-
position OH group of the phenolic ring of caffeic acid, 5—hydroxyferulic acid, quercetin
and luteolin. The enzyme also methylates two hydroxyl groups in myricetin, the positions
of the hydroxyl groups are most likely to be the 3’- and 5’-positions based on structural
similarities of the phenolic acids. No methylation was detected at the 4’-position in
apigenin, the 4-position in sinapic acid, or the 2-position in salicylic acid. No activity was
detected toward compounds such as protocatechuic acid, gallic acid or the flavonol

glycosides substituted in the 3-position as in quercetin 3-L-rhamnoside and rutin. From

the structures of active substrates (Figure 2.20), it is concluded that the position of the

101

OH group and the size of the substrate are both critical for OMT1 activty. The enzyme
also specifically methylates 3- and 5-position phenols of p-hydroxycinnamic acid
derivatives (e.g. caffeic acid and 5-hydroxyferulic acid) or the 3’- and 5’-positions with
C15 flavonoid skeleton (quercetin, myricetin, and luteolin), but not smaller
hydroxybenzoic acids (gallic acid and protocatechuic acid) lacking the three-carbon side
chain, or flavonol glycosides (quercetin 3-L-rhamnoside and rutin) which are conjugated

to carbohydrate at the 3-position.

102

59% £5 5 £38m _ :20 Sm 338 8:533 00 mogosbm omd oSwE

Away—55-53.5300
53M oEmocfiaf-m-::oo._o=O 822: ..Ao—EouzxegméAmv

 

:o
:0
Bow 230:3 Eon 0:30 Eon 235083805 Bus 298% Eon ozaumzxouvxmfi Bus 20th
o o o
o o

o :o _ o: _ o: _ o:

O»... O .. o .. Q O O
:o
o: :o :o :o :0
IO IO n
:00 :o :o

=EowE< 538.5 £80132 £82050

 

o
_ o o
nzoo
O. ..
z
:o o
:0

103

Label-free 0M T activity assays using LC/MS

Before this study, several methods have been used to measure methyltransferase
activity, but since spectroscopic differences between substrate and product are minimal,
spectrophotometric assays are not normally used. One of the commonly used methods
was described by Dumas et al. (1998) and is based on radiolabeled S-adenosylmethionine
as substrate. There are two approaches that can be used - one is to measure the amount
of residual SAM (by titration), but this is not sensitive when conversion of SAM is low.
The second approach would physically separate radiolabeled product from SAM, and
measure the radiolabel in the product. Both cases are time- and labor-consuming. A
recent report described a new LC/MS based assay for detection and quantitation of
methyltransferase activity, based on monitoring the conversion of S-adenosylmethionine
(SAM) to S-adenosylhomocysteine (SAH) (Salyan et al. 2006). In the current study, a
new assay method was developed to measure the OMT activity using LC/MS. This
approach allows direct measurements of the decrease of phenolic substrates and increase
0f products by LC/MS. The enzymatic assay method is described in the Materials and
Methods section. Briefly, the LCT Premier time-of-flight mass spectrometer was used to
measure the decrease of substrate in the reaction. A Therrno Betabasic C18 column
allOWS fast separation (10 min) of analytes with limits of detection less than 1 uM for all
SubStrates tested. The linear range of LC/MS detection is substance-dependent, and
typically ranges from about 0.1 uM to 50 pM. For enzymatic reactions involving higher
concentrations, the reaction was diluted by buffer G to be in the linear range when the

Concentration of substrate is above 50 pM.

104

vs II
temp:

cafibr

fi mg
i

nu 1

| t

mil:
5

Taking caffeic acid as an example, after the enzyme reaction, caffeic acid substrate
was identified (with m/z 179.04) and the decrease of caffeic acid was quantified by
comparing integrated peak areas for the extracted ion chromatogram for m/z 179 with a
calibration curve generated using standard solutions of varied concentrations. Values for
Vmax (0.795 pmol/min/mg) and Km (27 uM) were obtained through nonlinear curve
fitting model with the Michaelis—Menten kinetics equation of V=Vmax*[S]/(Km+[S])

using Origin 7 software (OriginLab, Northampton, MA, USA) (Figure 2.21).

105

f.".:\f.u

'.'.\-"E I I

Fig
in
p10

0.8.

A 0.6 1
3’
E
E 0.41
3-
>
0.2 a
0-0 I r r I I . I ' I

 

 

 

o 20 4o 60 80'160'

[S] (HM)

Figure 2.21 Recombinant OMT1 enzyme kinetics towards caffeic acid. Nonlinear curve
fitting model with the Michaelis—Menten kinetics equation of V=Vmu*[S]/M+[S]) to

plot V again [S].

106

em:
0535':
2‘ ‘ 1
11:0.

R:

labic'

were
math
Euteo

l‘Ole

The same procedure was used for other substrates with the mass for generating the
extracted ion chromatogram differing for each substrate. S-Hydroxyferulic acid was
measured using m/z 209.04, quercetin with m/z 301.04, myricetin with m/z 317.03, and
luteolin with m/z 285.03.

Results from kinetics studies of OMT1 with different phenolic substrates are shown in
Table 2.1. The best substrates of recombinant OMT1 are quercetin, myricetin and
luteolin, better than caffeic acid and S-hydroxyferulic acid, based on their low Km high
Vmax, and Kan/Km. The flavonol aglycones with hydroxyl groups in 3’ positions gave Km
values ranging from 7-8 pM, slightly lower than the hydroxycinnamic acids which were
27 and 35 uM for caffeic and 5-hydroxyferulic acids, respectively. Values for Kan/Km
were about 10- to 20-fold greater for the flavonols. Muzac et al. (2000) reported minimal
methyltransferase activity of OMT1 toward caffeic acid and 5-hydroxyferulic acid and
luteolin, and activity was so low that no values were reported for Vmax or K".. Since they
found that the protein was a monomer, a dimer may be critical for certain functions of the
enzyme. Therefore, we cannot discount the possibility that the OMT1 they purified either
failed to fold into a dimer during expression, or the activity was partially lost during the
purification. Since Muzac et al. (2000) did obtain activity with quercetin and myricetin,
which indicates the enzyme they purified was at least partially active, it could be that
their enzyme assay was not sensitive enough to measure activities toward caffeic acid and

5- hydroxyferulic acid.

107

a,

Sum
S-Hydr
(here:
llm'ce
luteol

—

Table
Arno]

rm.

prone

All
mice
scum
LC ll
comm
uh I
(e.g..
(e.g.,
OMT
gluco
sinap
lore]

math

 

Substrate Km(p.M) Vmax Km(SJ) Keat/Km(PM-IS.I)

 

(umol/min/mg)
Caffeic acid 27 0.795 0.53 0.020
5-Hydroxyferulic acid 35 1.525 1.24 0.035
Quercetin 7.7 6.818 4.55 0.491
Myricetin 7.5 3.044 2.03 0.270
Luteolin 7.0 3.884 2.59 0.369

 

Table 2.1 Affinities and activities of purified Arabidopsis recombinant OMT1 toward
phenolic substrates. No activities toward sinapic acid, salicylic acid, gallic acid,
protocatechuic acid, quercetin 3-L-rhamnoside, rutin, and apigenin were detected.

Although high OMT1 activities toward caffeic acid, 5-hydroxyferulic acid, quercetin,
myricetin and luteolin were detected in vitro, none of these, or their methylated products
accumulate to significant levels in wild type or mutant Arabidopsis leaves based on
LC/MS metabolite profiles. It may be because some substrates are intermediates and are
consumed in the cell, or some substrates are not present in the same organelle or tissue
with the enzyme. It also may be because the free acids are usually conjugated to sugars
(e.g., glucose conjugates), cell wall carbohydrates (e.g., ferulate esters), or organic acids
(e.g., sinapate esters, chlorogenic acid). The main evidence for in viva functions of
OMT1 in Arabidopsis leaves comes from levels of malate esters SHFM and SM, and
glucose esters SHF G and SG. It has been reported that SG can be converted to SM by
sinapoyl-glucose:L-malate O-sinapoyltransferase in Arabidopsis (Mock et al. 1992;
Lorenzen et al. 1996; Lim et al. 2001). Therefore, SM may be synthesized either from the
methylation of SHFM, or from $6, the product of the methylation of SHFG.

Neither SHFM nor SHFG is commercially available, therefore undergraduate student
Kristina Knight and I worked together to synthesize SHFM. Specific activities of purified
recombinant OMT1 against SHFM is 0.21 U/mg : 0.02. One unit of enzyme activity is

defined as the conversion of 1 umol of SHFM to SM per min at 30°C. Relative activity of

108

5111
0111

of 51

1008
p811
Base

may

Res:
boll

chal

infe
1631'
smn
also

8151'

SHFM to SHFA is about 10%. One finding of note is that incubation of the recombinant
OMT1 with a crude metabolite extract from omt1 mutant leaves gave 100% conversion
of SHFM to SM and 5HFG to SO in less than one hour, suggesting that the enzyme is
capable of methylating malate and glucose esters of hydroxycinnamic acids. The malate
esters are some of the dominant polyphenols in Arabidopsis leaves. To our knowledge
this is the first direct demonstration of OMT activity toward malate esters.

OMT1 can use the flavonoids quercetin, myricetin, and luteolin as substrates.
Quercetin and its glycosidic and rhamnosidic derivatives are major flavonoids found in
Arabidopsis thaliana (Kerhoas et al. 2006; Stobiecki et al. 2006) in certain tissues such as
roots, but are uncommon in leaves. The induced production of flavonoids under stress of
pathogens and pests is a well-known phenomenon in plants (Barry 2002; Gallet 2004).
Based on these earlier findings, we anticipated that quercetin glycosides or rhamnosides

may be impacted in roots and other tissues of omtl mutant plants under pathogen stress.

Result 2: Comparison of metabolite profiles between control and pathogen stress
both in the wild type plants (WC vs WP) and in the mutant (MC vs MP) to
characterize the biomarkers in plant-pathogen interaction

One of the goals of this study is to identify the biomarkers, the indicators of pathogen
infection or plant defense-involved secondary metabolites. In this experiment, the control
leaves were inoculated using a syringe with a sterile solution of 10 mM MgClz via the
stomata into the lower surface of half of the leaf. Therefore, the control samples were
also under wounding stress. Some metabolites were induced both in control and pathogen

samples due to the wounding stress. However, the biomarkers in pathogen stress are the

109

metabolites induced only after pathogen stress but not in control samples, both in the wild
type and the mutant. The amounts of the metabolites discussed below were calculated as
integrated extracted ion current (XIC) peak areas divided by fresh leaf weight (mg). The
amount the metabolites and their corresponding mass to charge ratios (m/z) are shown
either in the content or in the Appendix A.

Some biomarkers are pathogen produced metabolites such as coronatine (m/z
318.17), or induced plant metabolites such as the phytoalexin camalexin (m/z 199.04).
Both of them were induced gradually after pathogen stress and reach their highest levels
at three days (Figure 2.29 and 2.30).

Some biomarkers of interest in pathogen stress are signaling metabolites such as
salicylic acid (SA) (m/z 137.04), two isomeric forms of salicylic acid glucoside (SAG)
(m/z 299.07), ABA (m/z 263.13), jasmonic acid (JA) (m/z 209.12), 12-OH JA (m/z
225.15), JA Ile (m/z 322.20), 12-oxo-phytodienoic acid (OPDA) (m/z 291.20), OPC:8
(m/z 293.2), linolenic acid (LA) (m/z 277.22), and oxylipin glycolipid conjugates
Arabidopsides A-E (either in the content or in the Appendix A).

Metabolite profiling also detects an assortment of candidate biomarkers including
glucosinolates. Quantification. of glucosinolates is shown in Appendix A. Methylsulfinyl
glucosinolates have sulfoxide groups, e.g. 4-methylsulfinylbutyl glucosinolate (m/z 436)
and 8-methylsulfinyloctyl glucosinolate (m/z 492), their reduced methylthio
glucosinolate analogs, e.g. 4-methylthiobutyl glucosinolates (m/z 420) and 8-
methylthiooctyl glucosinolate (m/z 476), and phenethyl glucosinolate 3-
methylsulfinylpropyl glucosinolate (m/z 422) decreased after pathogen stress, which

corresponds to previous transcription results showing down-regulation of glucosinolate

110

biosynthesis genes (Tierens et al. 2001). Indole glucosinolates, e.g. 4-methoxyindolyl or
l-methoxyindolyl glucosinolates (m/z 477.07) did not exhibit obvious increases as shown
in the previous transcription results (Tierens et a1. 2001).

Polyphenols and quinone metabolites were changed after pathogen stress such as the
decrease of SHFM (m/z 325.06) and SM (m/z 339.07) and increase of SHFM-GS (m/z
630.13) and SHFM quinone-GS (m/z 628.11). The measured masses for SHFM-GS
(630.13) and SHFM quinone-GS (628.11) are within 5 ppm of theoretical values. The
major fragment ion for SHFM-GS (630.13) is m/z 325.06 for deprotonated SHFM and
m/z 306.08 for deprotonated GSH. The major fragment ion for SHFM quinone-GS
(628.11) is m/z 323.04 for deprotonated SHFM quinone and m/z 306.08 for deprotonated
GSH.

Some biomarkers were present in different amounts in the wild type and the mutant,
such as coronatine, camalexin, SA (m/z 137.04), ABA (m/z 263.13), JA (m/z 209.12),
12OH-JA (m/z 225.15), JA Ile (m/z 322.20), SHFM (m/z 325.06), SM (m/z 339.07),
SHFM-GS and SHFM quinone-GS. Some of them showed no differences between the
wild type and the mutant, like OPDA, OPC:8, and LA. The metabolites that were

different between the wild type and the mutant are discussed in the next section.

Result 3: Comparison of leaf metabolite profiles between WT and the amt] mutant

after pathogen attack (WP vs MP) to characterize the OMT function under stress
Comparing metabolite profiles of the amt] mutant under pathogen stress with wild

type plants was used to investigate the functions of this gene in pathogen response. Plant

defense response is a time-dependent event; therefore metabolome changes were studied

111

at 1 hour, 24 hours, 48 hours, and 72 hours after inoculation to monitor the progress of
early to late induced responses. Metabolites with significantly different between WT and
the amt] mutant after pathogen attack (WP vs MP) were identified and quantified to
interpret the role of OMT in modulating metabolic responses in Arabidopsis during P.

syringae infection, as discussed below.

SHFM, SM, SHFM-GS, SHFM Quinone-GS, and GSH

The levels of SHFM, the proposed substrate of OMT1, were decreased gradually after
pathogen stress in the mutant (Figure 2.22), and so did the level of SM, the product of
OMT1 (Figure 2.23). The structurally related compounds 5HF M glutathione conjugate
(SHFM-GS) (Figure 2.24) and SHFM quinone glutathione conjugate (SHFM Quinone-
GS) (Figure 2.25) were found to increase gradually after pathogen stress in the mutant,
with the depletion of GSH (Figure 2.26). SHFM quinone-GS was only induced in the
mutant, but not in the wild type. In the mutant, it increased two-fold more during
pathogen stress than in the control. SHFM-GS was also induced mostly in the mutant. In
the mutant, it increased two-fold more during pathogen stress than in the control. SHFM-
GS was present at one time point in wild type plants. We have no explanation for the
appearance in the wild type. Depletion of GSH was observed following pathogen
inoculation. In the mutant, GSH was completely consumed after pathogen stress (relative
amount dropped from 30 to 0), while the decrease of GSH in control was less (relative
amount dropped from 30 to 10). The control sample had wounding stress during the

infiltration of Mg2C12 at 0 h. That is may account for why SHFM-GS and SHFM quinone-

112

GS were also induced in control samples, though two-fold less than during pathogen
stress.

The correlated relationship of the decrease of SHFM and the increase of SHFM
quinone-GS and SHFM-GS, as well as the depletion of GSH is consistent with the
suggestion that SHFM was oxidized to quinone following pathogen infection, and the
quinone was then conjugated to GSH to produce SHFM-GS. SHFM-GS could be
oxidized to SHFM quinone-GS by various oxidants including H202, which could be
reduced to SHFM-GS by cellular reducing agents such as GSH as well. These quinones
and hydroquinones may undergo redox cycling, giving catalytic generation of superoxide
and H202 from 02 as shown Figure 2.27 and discussed below. GSH is consumed both in
the conjugation and redox cycling, and such biochemical events can lead to depletion of
other cellular reducing agents. This assumption is made based on previous redox cycling
studies in animals (Wilson et al. 1986; Galati et al. 1999; Stephensen et al. 2002). One
notable finding was that the oxidized form of GSH, GSSG, was not observed in
substantial amounts in any of the metabolite extracts.

The abbreviation of samples with different genotypes or treatments are: WC for the
wild type control with the infiltration of Mng buffer, MC for the mutant control with
the infiltration of MgClz buffer, WP for the wild type with the infiltration of P. syringae,

and MP for the mutant with the infiltration of P. syringae.

113

600

500

400

 

 

Integrated XIC peak areas / mg fresh leaf weight

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

-I-WC
300 4 +WP
-I- MC
+MP
200 .
100 «
h ‘I— \r
o l . . . . . . j .
0 1o 20 so 40 so so 70 so
T-test analysis on metabolite levels Time (11)
1h 1d 2d 3d
WC vs MC * * * *
WP vs MP * * *
WC vs WP N.S. N.S. N.S. N.S.
MC vs MP * N.S. * *
“*” significant, p < 0.05
“N.S.” not significant

COOH

”W“

OCH3

Figure 2.22 Levels and chemical structure of SHFM (m/z 325.06)

114

 

Integrated XIC peak areas / mg fresh leaf weight

1800

1500

1200]

900+

 

l
i\w.

 

 

 

i
4

 

 

600
300
0 WW I
0 10

 

40 50

:‘kfir/I.
fl

 

:I- WC
+ WP
+ MC
+ MP

 

fl

 

 

 

 

 

 

 

 

 

 

 

 

30 7o 30
T-test analysis on metabolite levels Time (h)
1h 1d 2d 3d

WC VS MC * * at a:

WP vs MP * * ... ...

WC vs WP N.S. N.S. * ...

MC vs MP N.S. N.S. N.S. N.S.
66*” Significant, p < 0.05
“N.S.” not significant

COOH
\ O
HO COOH

Figure 2.23 Levels and chemical structure of SM (m/z 339.07)

OCH.

115

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

E
.9.” .
“3’ 25 . w
9.4
(U
2
>‘ 1
-° 20
v
“:3:
'7'? +wc
E 15 i “ 1 +WP
o ‘ +MC
g " + MP
(’1
\O
N 10 4
E
:2
26’ 5 ‘.
a, ,
33
fi 1
33 0 f * I * . - . . I .
o 10 20 so 40 so so 70 so
T-test analysis on metabolite levels Time (h)
WC vs MC N.S. * * a:
WP VS MP * ‘1‘ * 1:
WC vs WP N.S. * N.S. N.S
MC vs MP * * a: ...
66*” Sigflificant, p < 0.05
“N.S.” not significant
0 OOH
HO
\ O
HO COOH
S
OCH3
H
N '1,”
HO ’N OH
H
0 NH2

Figure 2.24 Levels and chemical structure of SHFM-GS (m/z 630.13)

116

 

25~

201

 

 

Integrated XIC peak areas / mg fresh leaf weight

 

 

0 10 20 30 40 50 60

 

70 80

 

 

-I-WC
+WP
-I-MC
+MP

 

 

 

 

 

 

 

 

 

 

 

 

 

Time (h)
T-test analysis on metabolite levels
1h 1d 2d 3d
WC vs MC N.S. * * *
WP vs MP N.S. N.S. N.S. N.S.
WC vs WP N.S. N.S. N.S. N.S.
MC vs MP N.S. N.S. * *
“*” significant, p < 0.05
“N .8.” not significant
CO OH
oi
COO H
O O

 

NH2

0
Figure 2.25 Levels and chemical structure of SHFM quinone-GS (m/z 628.11)

117

 

Integrated XIC peak areas / mg flesh leaf weight

 

 

0 10 20 3O 40 50 60 70 80

Time (h)

T-test analysis on metabolite levels

 

 

 

 

 

 

 

 

1h 1d 2d 3d
WC vs MC N.S. N.S. * *
WP vs MP N.S. N.S. * *
WC vs WP N.S. N.S. N.S. N.S.
MC vs MP N.S N.S. * *

 

 

 

 

66*”

significant, p < 0.05
“N.S.” not significant

HS

)k/H
N
HO ” OH

O NH2
Figure 2.26 Levels and chemical structure of GSH (m/z 306.08)

I2

118

.wezozo x82 a;
NON: Ea ocmxeogsm mo coves—.8.“ £588 28 mwécofizc SEE Ea mgzmmm .wo cocoa—08a 8m $353 88205 had Emmi

60-8855 size 0
SEE: mcoE:030::okxofiféaéofifiiO-N

   

O
1000
O
V / com:
IOOO O Om
1: <2 :mO NON:
Nv§~UV~ 3mm 0:3 COS—Lug Cm ...
A XESSSD EQEE ozo:._~:o_Eom Atmov QExOSQDV
. n a i in N
: ~D<Z rvvvrv IO DA V
m0-_>_n_Im
OI 939:
1000 _>o_amsxo§;¢_>co_ess_o-m
O
/V\ / OOmI
IOOO O Om
:me x :2“.sz
93$. Schum.
AWMMWV «25.22350 SENS _>o_:._m+>xo.__u>I-m
ococ_zc_>o_2o§xofi>1.m 0 IO
0 o:

 

1000 A 1000

o %cw:\w.§...§5 o
/ OOmI / OOmI
o

1000

119

SHFM-OH and SHFM-OH-GS

Evidence to support that quinones and hydroquinones undergo redox cycling and give
catalytic generation of H202 is that more oxidized metabolites were induced in the mutant
after pathogen stress such as SHFM-OH (Figure 2.28) and SI-IFM-OH-GS (Figure 2.29).
The precise nature of the additional oxidation step is not yet clear, but the results are
consistent with formation of reactive oxygen species that could introduce an additional
oxygen atom into these acids and GSH conjugates. It suggests the mutant was under more

oxidative stress than the wild type.

120

Figure 2.28 Levels and chemical structure of SHFM-OH (m/z 341.06). Two possible
structures are proposed.

121

60

5O
40
—I—WC
30 +WP
-I- MC
+ MP
20

Integrated XIC peak areas / mg fresh leaf weight

 

 

 

 

 

 

 

Time (h)
T-test analysis on metabolite levels
1h 1d 2d 3d
WC vs MC N.S. N.S. * *
WP vs MP N.S. N.S. * *
WC vs WP N.S. N.S. N.S. N.S.
MC vs MP N.S. N.S. * *

 

 

 

 

 

 

 

66*99

significant, p < 0.05
“N.S.” not significant

122

Figure 2.28 continued

H 0 COOH
HO
\ O
H
HO coo
OCH3
0 COOH
0
HO
0
H
HO coo

OCH3

123

Figure 2.29 Levels and chemical structure of SHFM-OH-GS (m/z 646.12). Two possible
structures are proposed.

124

4.5 1

 

Integrated XIC peak areas / mg fresh leaf weight

 

 

j

0 1O 20 30 40 50

 

 

-I—WC
+WP
-I-MC
+MP

 

 

 

 

 

 

7o 80
Time (h)
T-test analysis on metabolite levels
1h 1d 2d 3d
WC vs MC N.S. N.S. N.S. N.S.
WP vs MP N.S. N.S. * *
WC vs WP N.S. N.S. N.S. N.S.
MC vs MP N.S. N.S. * *

 

 

 

 

 

 

 

“*9,

significant, p < 0.05
“N .8.” not significant

125

 

Figure 2.29 continued

 

H COOH
HO
\ O
HO COOH
s
OCH3
H
0”],
HO ”u OH
O NH2
OOH
COOH
OCH3
H
0”],
HO ” OH

N
H

O NH2

126

Con

1101'

all:

5le

51m

int

tie

Coronatine

Coronatine levels (Figure 2.30) increased in both mutant and wild type after
inoculation, but this increase was about two-fold less in the mutant than in the wild type
after three days of pathogen treatment. Increases in the mutant were also less for SA,
ABA, JA, IZOH JA, and JA Ile. Coronatine is an important virulence factor for P.
syringae DC3000 infection in Arabidopsis plants and a pathogen-derived functional and
structural mimic of the phytohormone jasmonic acid (JA) and it is thought to be involved
in the reopening of stomata (Cui et al. 2005; Melotto et al. 2006). It is a metabolite which
the pathogen secretes and is a biomarker for pathogen stress. In the mutant, the slower
increase in coronatine levels may be explained by the inhibition of bacterial grth

relative to wild type plants.

127

..L
N

...:
O
1

Integrated XIC peak areas / mg fresh leaf weight

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

+wc
+WP
-I-MC
+MP
4 i
2 4
0 . . a . I. . + .
0 1o 20 30 40 so so 70 so
Time (h)
T-test analysis on metabolite levels
1h 1d 2d 3d
WC vs MC N.S. N.S. N.S. N.S.
WP vs MP N.S. N.S. N.S. *
WC vs WP N.S. N.S. *
MC vs MP N.S. N.S. *
“*” significant, p < 0.05
“N.S.” not significant
0 .
O
HN

§\ HOOC

Figure 2.30 Levels and chemical structure of Coronatine (m/z 318.17).

128

 

Camalexin

Camalexin was induced two-fold more in the mutant than in the wild type after two
days of pathogen stress (Figure 2.31). Camalexin is a phytoalexin which plants produce
and has anti-microbial properties (Zook 1998). It is also a candidate biomarker for
pathogen stress. Therefore, the higher levels of camalexin in the mutant are consistent
with stronger activation of plant defense in the mutant despite slower progression of

infection in the mutant than the wild type.

129

14 -

.5
N
i

A
o
i

 

 

Integrated XIC peak areas / mg fresh leaf weight

 

 

 

 

—I—WC
+WP
-I-MC
+MP

 

 

 

 

 

 

 

 

 

 

 

 

o "/ . I7— . . F .
0 10 20 30 40 50 60 70 80
Time (h)
T-test analysis on metabolite levels
1h 1d 2d 3d
WC vs MC N.S. N.S. N.S. N.S.
WP vs MP N.S. N.S. * N.S.
WC vs WP N.S. N.S. * *
MC vs MP N.S. N.S. * *
“*” significant, p < 0.05
“N .8.” not significant
H
N

/

\9/

Figure 2.31 Levels and chemical structure of camalexin (m/z 199.04).

130

 

 

SA, ABA, JA, lZOH JA and JA Ile

At 1 h post inoculation, SA, ABA, JA, 12-OH JA and JA Ile were induced suddenly.
Levels of JA are presented in Figure 2.32. Levels of SA, ABA, 12-OH JA and JA Ile are
shown in Appendix A. They were increased at least two-fold more (with the exception of
ABA) in the mutant relative to the wild type both in pathogen inoculated and control
samples at 72 hours. Taking JA as an example, the mutant (both control and pathogen-
infected) produces a quick surge in J A that is about 2-fold greater than the wild-type
(both control and pathogen-infected) at l h. It is noted that SHFM-GS was observed as
early as at 1 h treatment, but not before treatment. The enhanced JA burst in the mutant
suggests participation of OMT functions, manifested in the form of higher levels of
quinone or other non-methylated metabolites in enhancement of the oxidative burst
following pathogen inoculation.

Though levels of J A increased at 1 h post-inoculation, these levels dropped after 24
hours and increased again from 48-72 hours. The early increase is thought to result
largely from wounding stress during inoculation, since IA is known to be the major
signal compound induced after wounding. The late increase is attributed to the
consequences of pathogen infection, which is greater for wild-type infected plants than
for mutant infected plants. JA and JA Ile exhibited same behavior. SA, ABA, JA, 12-OH
JA, and JA Ile exhibited similar behavior at lh and 72 hour. Interestingly, JA and its
precursors in the oxylipin pathway were observed in different patterns. Oxylipins OPDA,
OPC:8, and LA showed no difference between the wild type and the mutant (Appendix

A).

131

—l
N
_J

 

 

 

 

 

 

H
g
.29

“a? 10 .

c...

:3

.2

A:

8 M

cl:

00 -

E

\

s 6—

0 ‘ l.
a , .-
.M .

«3

0 1!

CL 4 1'.

U l]

H ,.

x 1

“o l

0 r I r r T r I fit
0 10 20 30 40 50 60 70 80

 

 

—l—WC
+WP
-I-MC
+MP

 

 

 

 

 

 

 

 

 

 

 

 

 

Time (h)
T-test analysis on metabolite levels
1h 1d 2d 3d
WC vs MC N.S. N.S. N.S. N.S.
WP vs MP N.S. N.S. N.S. N.S.
WC vs WP N.S. N.S. N.S. *
MC vs MP N.S. N.S. N.S. *
“*” significant, p < 0.05
“N.S.” not significant
0

COOH

Figure 2.32 Levels and chemical structure of JA (m/z 209.12).

132

 

Flavonoids

The major flavonoids in leaf extracts, 3-O-,7-O-dirhamnosyl kaempferol and 3-0-
glucosyl, 7-O-rhamnosyl kaempferol, accumulated slightly more (less than 10% of total
amount) in mutant relative to wild type (Appendix A). Those flavonoids are not
substrates of OMT1, as they lack hydroxyl groups in the 3’- and 5’- positions. Levels
showed slight decrease both in pathogen stress and control conditions, while flavonoid 3-
O-glucosylkaempferol (Appendix A) was increased gradually afier pathogen stress,
perhaps due to increased hydrolysis of the 7-position glycosides. Flavonoid synthesis is
part of the phenylpropanoid pathway and is a competing biosynthetic branch of
phenylpropanoid biosynthesis with the synthesis of phenolic acids, the latter of which
serve as precursors of lignins (Nair et al. 2004; Rogers et al. 2005). Therefore, defects in
omt1, which may affect phenols in the phenylpropanoid pathway and lignin synthesis,
may also impact the levels of flavonoids. However, the results of this study do not
provide definitive information in this regard.

Minor amounts of oxidized 3-O-glucosyl, 7-O-rhamnosyl kaempferol, appearing as a
product with an additional hydroxyl group at position undetermined, were elevated
slightly in the mutant relative to the wild type. This finding supports the suggestion that
mutant plants are capable of generating more ROS than wild type plants in both control

and pathogen stressed plants.

OPDA-GS and dinor OPDA-GS

GSH depletion was observed after pathogen stress, dropping to undetectable levels in

the mutant by the third day after inoculation. To assess whether GSH depletion arises

133

largely via conjugations to quinone metabolites, other GSH conjugates such as oxylipin
GSH conjugates were also quantified. Levels of OPDA-GS (m/z 598) and dinor OPDA-
GS (m/z 570) decreased after one day pathogen stress, but did not change further after
two days (Appendix A). There was no difference between mutant and the wild type, so it
is concluded that GSH depletion was not accelerated in the mutant owing to more
conjugation to oxylipins.

Taken in total, metabolite profiling has revealed a substantial decrease in GSH levels
in infected mutant plants, relatively modest increases in the sum of all glutathione
conjugates, and the absence of evidence for an increase in GSSG. These findings provide
indirect evidence to suggest that some GSH is not accounted for. One explanation that
deserves further study is that oxidant stress in infected mutants leads to increased
formation of mixed protein-GSH disulfides, as these would not be detected using the

current methodology.

2.3.5.3 Anti-microbial properties of quinones, quinone-GS conjugates, and H202

The metabolite data suggest that SHFM quinone metabolites and their GSH
conjugates may provide direct toxicity to pathogenic P. syringae. To probe whether
polyphenolic quinone and quinone-GS conjugates pose chemical mechanisms leading to
direct toxicity to P. syringae, in vitro cultures of P. syringae were dosed with
polyphenols, quinone, quinone-GSH conjugate, and H202 (Figure 2.33). SHF was used
instead of SHFM because adequate supplies of SHF M were not available and SHF has the

same functional group and similar physical properties.

134

SHF quinone and SHF quinone-GS were synthesized and used in bacterial culture,
both in liquid and solid plates with the concentrations chosen to mimic the estimated in
viva concentrations of the malate ester analogs based on the LC/MS peak areas of the
quinonoid compounds in extracts of the pathogen stressed leaves (Figure 2.33). Inhibition
of bacterial growth was observed at low uM concentrations (~ 2 11M) for 5-
hydroxyferulic acid, its quinone, and the GSH conjugate of the quinone, but not for the
methylated analog sinapic acid. Though GSH conjugation of quinones has been
considered a detoxification step, the persistence of antagonism to P. syringae grth in
the quinone-GSH conjugate suggests this class of metabolites may undergo redox
cycling, catalyzing formation of superoxide and hydrogen peroxide, as has been observed
in studies of quinone toxicity in animals (Monks and Lau, 1998).

To probe whether H202 leads to direct toxicity to P. syringae, a series of different
concentrations of H202 (100, 10, 1, 0.1, and 0.01 mM) were incubated in bacteria culture.
Growth of P. syringae was inhibited by the highest H202 concentrations (100 mM and 10

mM) but inhibition was not observed at 1 mM or lower concentrations (Figure 2.33A).

135

Figure 2.33 Concentration dependence of antimicrobial potential of H202, 5-
hydroxyferulic acid quinone conjugate (5HFQ), 5-hydroxyferulic acid quinone with
glutathione conjugate (5HFQ-GS), 5-hydroxyferulic acid (5HFA), and sinapic acid (S)
against P. syringae DC3000. Higher concentration (100 mM and 10 mM) of H202, 2 uM
and 20 uM SHFQ, SHFQ-GS, and SHFA showed clear zones of P. syringae DC3000
inhibition in disc diffusion method (A). When SHFQ, SHFQ-GS, or 5HFA was
incorporated into P. syringae DC3 000 suspension, it showed a concentration dependent
colony forming unit reduction (B). In figure (B), all data points represent the average of
three independent experiments.

136

 

5 ul of a series of different concentration of H202 (100 mM, 10 mM, 1 mM, 0.1 mM,
0.01 mM)

 

-: negative control: 10 ul of DMSO

+: positive control: 10 ul of 10 mM H202

5HFQ1: 10 ul of ~2 11M 5-hydroxyferulic acid quinone conjugate
5HFQ2: 10 ul of ~20 uM S-hydroxyferulic acid quinone conjugate

137

Figure 2.33 continued

 

-: negative control: 10 u] of DMSO

+: positive control: 10 u] of 10 mM H202

SHFQ-GSl: 10 ul of ~2 uM 5-hydroxyferulic acid quinone and glutathione conjugate
5HFQ2-GSZ: 10 ul of ~20 uM 5-hydroxyferu1ic acid quinone and glutathione conjugate

 

-: negative control: 10 ul of DMSO

+: positive control: 10 ul of 10 mM H202
SHFAI: 10 ul of ~2 uM 5-hydroxyferulic acid
5HFA2: 10 ul of ~20 11M 5-hydroxyferulic acid

138

Figure 2.33 continued

% cfu reduction

 

-: negative control: 10 ul of DMSO

+: positive control: 10 1.11 of 10 mM H202
$1: 10 ul oflO 11M sinapic acid

82: 10 pl of 20 11M sinapic acid _

120 -

 

O I l l l

 

 

0 10 20 30 40

Concentration (11M)

139

I

50

l

60

 

 

+ 5HFA
—I— 5HFQ
+ 5HFQ-GS

 

 

2.3.5.4 Quantification of H202 in leaf tissue

To see if H202 was produced in Arabidopsis leaves after pathogen stress. Hydrogen
peroxide was quantified using the Fenton reaction with FOX assays (Delong et al. 2002;
Franklin et al. 2009). However, no H202 was detected in extracts of the leaf tissues after
pathogen stress. It is suspected that H202 was degraded by enzymes such as catalases or
H202 reacted with metabolites in leaves. Many microbes are capable of producing H202-
degrading enzymes including catalase and peroxidases as well. Inthe experiment to test
anti-microbial properties of H202, H202 had no effect to inhibit the growth of P. syringae
DC3000 when the concentration was lower than 1 mM in this study. (Figure 2.33A).
Therefore, the results do not allow us to conclude whether enough H202 was produced in

leaves of mutant plants to inhibit the growth of P. syringae DC3000.

2.3.5.5 Localization of SHFM-GS and SHFM Quinone-GS in apoplast

Plant apoplast, the intercellular space, is the first site of contact with a pathogen.
(Bolwell et a1. 2001). Sometimes metabolites will be secreted from the plant cell to the
apoplast. One consequence of pathogen infection is that the cell may die when the
resistant response fails or becomes overwhelmed. When cell death occurs, membrane
integrity is lost, and the metabolites in the cell may flow into the apoplast. To determine
whether SHFM-GS and SHFM quinone-GS could diffuse into the apoplast to suppress
pathogen growth, apoplast extracts were prepared by centrifugation, and these extracts
were analyzed for metabolites. After 1 day post-inoculation, the extract from the apoplast
was collected before evidence of cell death was visible. Apoplast extracts were analyzed

using LC/MS. SHFM-GS and SHFM Quinone-GS were not found in the apoplast. At two

140

days post-inoculation, cell death was evident fi'om leaf discoloration. The extraction from
the apoplast was therefore expected to sample from not only live cells but also damaged
dead cells. LC/MS analyses for this time point also showed low levels of SHFM quinone
glutathione conjugate were released into the sampled volume, which may explain why

the bacteria were inhibited in planta after two days.

Discussion
Polyphenols, quinones and glutathione conjugates

The polyphenols are important specialized metabolites of plants, known for their role
in browning in plants (Antolovich et al. 2004). Many are oxidized to their corresponding
quinones by polyphenol oxidase when plant tissue is damaged. Quinones are reactive
substances that have structural similarities to aromatic compounds, being based upon a
six-membered ring with two double bonds and two ketone substitutions, arranged either
ortho (in the 1,2- positions) or para (in 1,4 positions). Some quinones are further
polymerized by reactions with quinones or amines to form a brown pigment in fi'uit and
vegetables (Murata et a1. 2002). Enzymatic browning involves several steps, including
the hydroxylation of monophenols into o-diphenols and further oxidation of o-diphenols
into quinones. The mechanism of nonenzymatic browning is not well understood
(Antolovich et al. 2004), in large part because the structures of the browning products are
not well characterized.

Quinones are electrophiles and oxidants, with both properties arising fi'om their
electron deficient nature. Quinones are electrophilic (seeking electrons) because the

addition of two electrons can convert them into more stable aromatic molecules

141

containing benzene ring structures. They undergo rapid reactions with the nucleotide
groups in DNA and with amino or thiol groups in proteins and peptides to form covalent
adducts. They are also oxidants, capable of oxidizing other substances that have lower
redox potentials, including some other phenols. The electrophilic and oxidant properties
of quinones can result in the destruction of molecules essential for cell survival including
the tripeptide glutathione, which is an important cellular redox buffer (Fulcrand et al.
1994; Monks and Lau 1998). Quinones are also mutagenic and carcinogenic (Ames 1983;
Chesis et a1. 1984; Patrineli et al. 1996; Comwell et al. 2002) and may provide direct
toxicity in the cell through redox cycling (Lilienblum et al. 1985).

Metabolic conjugation of GSH to reactive electrophilic metabolites is well
documented as a detoxification mechanism in animals (Chroust et a1. 2001; Coles and
Kadlubar 2003; Pastore et al. 2003), but our understanding of GSH conjugation in plants
is still an emerging field of study. The potential importance of GSH conjugation is
suggested by 54 glutathione S-transferase genes in the genome of Arabidopsis thaliana
(Marrs 1996; Neuefeind et al. 1997; Dixon et al. 2009), but their importance has
remained unclear aside from having roles in herbicide resistance and accumulation of
anthocyanins in vacuoles (Alfenito et al. 1998; Mueller et al. 2000; Weisel et al. 2006).
Several electrophilic oxylipin metabolites have been observed to form GSH conjugates
(Davoine et al. 2005), but their significance remains unclear. Glutathione (GSH) is the
major non-protein sulfhydryl in cells, and conjugation of potentially toxic electrophiles
with GSH is usually associated with detoxication and excretion in animals. When GSH
conjugates to quinones, the reaction is generally considered cytoprotective because the

thiol function in GSH serves as a “sacrificial” nucleophile. However, polyphenolic-GSH

142

conjugates still possess a variety of biological activities. For example, quinone-thioether
conjugates may participate in redox cycling, as substrates or inhibitors of enzymes, or as
DNA-reactive mutagens, and may contribute to quinone-mediated carcinogenicity and
neurotoxicity in animals (Monks and Lau 1992; Monks and Lau 1998).

In this study, glutathione conjugates of polyphenolic quinones were discovered in
plants, with levels in pathogen-infected leaves of omt1 mutant plants estimated to reach
low- to mid- uM levels. Based on in vitro assays, 5HFA, 5HFA quinone, and its GSH
conjugate showed marked inhibition of P. syringae growth at comparable concentrations.
To our knowledge, these metabolites represent a novel finding, not just for Arabidopsis,
but for the plant kingdom in general. Glutathione conjugation of quinones generates
thioether-quinol adducts that are more reactive toward oxidation than the precursor
polyphenol, and are more efficient at generating superoxide (Bratton et al. 1997; Bratton
et al. 2000). The studies in animals showed that cytotoxicity of quinones is enhanced
after GSH conjugation, owing in part to increased formation of reactive oxygen species
(ROS). Studies of benzoquinone and its GSH conjugate demonstrated that the quinol-
thioether conjugates readily form semiquinone radicals and undergo redox cycling,
converting 02 to superoxide by one-electron reduction, and providing a path to other
ROS including H202. ROS are known to induce phytoalexin in Arabidopsis and other
plants (Rusterucci et al. 1996; Thoma et al. 2003; Zhao et al. 2007).

After P. syringae inoculation, camalexin was induced two-fold more in the mutant as
shown in the results above, which may be toxic the bacteria. H202 may be directly toxic
to bacteria, but many microbes are capable of producing H202-degrading enzymes

including catalase and peroxidases. After two days of pathogen treatment, some of the

143

plot

GS

101

de

plant cells started to die and lyse; the quinone and other anti-microbial compounds
entered the apoplast and presumably killed the bacteria, or H202 produced through redox
cycling may have killed the bacteria. Fewer bacteria in the mutant plant may explain why

less coronatine was found than in the wild type.

GSH depletion - an alternative explanation for inhibition of pathogen growth

In humans, GSH depletion is linked to a number of disease states including cancer,
neurodegenerative, and cardiovascular diseases (Pastore et al. 2003) as well as liver
toxicity of drugs (Hentze et al. 2000; Jimenez-Lopez et al. 2008). In this study, GSH
depletion was shown in the mutant after pathogen infection. During oxidant stress, GSH
may be transformed to its oxidized dimer GSSG, and the latter may form mixed
glutathione-protein disulfides (not measured in this study), or it can be conjugated to
metabolites such as quinones and electrophilic oxylipins such as OPDA. GSSG could not
be detected in leaf extract. OPDA-GS and dinor OPDA-GS were observed in a small
amount at all the time points under both control and pathogen stress, but differences in
oxylipin glutathione conjugates between mutant and the wild type were minimal.
However, substantial differences in conjugation of GSH to SHFM quinone distinguished
pathogen-infected omt1 mutant and wild type plants. Therefore, GSH depletion was due

to the mutation of the amt] gene and may cause the inhibition of pathogen growth.

Function of OMT1 and its relevance to the metabolome and pathogen resistance

This study has demonstrated a role for OMT1 in the methylation of one of the most

abundant secondary metabolites, SHFM, converting it to SM. This function of OMT1

144

was proposed from the results of metabolite profiling without stress and the function was
confirmed by biochemical characterization. OMT1 was found to methylate 3’ or 5’OH
group on the phenolic ring of polyphenols, including caffeic acid, 5-hydroxyferulic acid,
quercetin and luteolin, and our findings are consistent with one other report (Goujon,
2003). SM is one of the most abundant secondary metabolites in leaves of the Col-0
ecotype. In this study, we found that one of the important functions of OMT1 is its
methylation during its biosynthesis, and experiments with recombinant OMT1
demonstrate, in contrast with an earlier study (Muzac, 2000), that SHFM is a substrate for
this enzyme. In the absence of the amt] gene, SHFM accumulates in leaf tissue, and this
catechol derivative is oxidized to a reactive quinone upon pathogen stress, and to a lesser
extent, following wounding in inoculated controls. Levels of SHFM quinone and its GSH
conjugate, in both oxidized and reduced forms, increased after pathogen stress. These are
some of the few metabolic differences observed between amt] and wild type plants and

their anti-microbial properties were observed.

2.4 Conclusion and perspective

Metabolite profiling of amt1(At5g54l60) mutant and wild type plants cultivated in
control conditions gave no significant morphological phenotype and the only pronounced
metabolic phenotype in the mutant was a ~3 5% reduction in SM and a stoichiometrically
similar increase in SHFM, a metabolite containing a catechol (quinol) moiety. These
findings are in agreement with methylation of SHFM by recombinant OMT1, and
confirm the in viva function of OMT1 in SM biosynthesis. The functional consequences

of the lack of near-complete methylation were only evident under conditions of

145

inoculation with P. syringae, where bacterial counts were higher (p < 0.05) in the wild
type by about 5-fold. In vitro assessment of the antibacterial properties of SHF, its
quinone, and quinone glutathione conjugate, all showed inhibition of bacterial growth at
low p.M concentrations, which are in the range of levels observed in leaf extracts using
LC/MS. In addition, the amt] mutants exhibited enhanced yields of IA at 1 hr past
inoculation, suggestive of crosstalk between polyphenol metabolic pathways and
jasmonate signaling.

The mutation of one single gene, amt], led to changes in metabolite profiles both in
control and pathogen stress conditions, but these are most evident after pathogen stress.
This research has presented an original approach in conducting a wide-range of
simultaneous analyses of specialized metabolites on genotypes with well-defined
differences (a specific T-DNA mutation). When coupled with assessment of pathogen
infection, the findings suggest that nontargeted metabolomic profiling, with emphasis on
specialized metabolites, may draw more useful functional conclusions by studying
stressed rather than control growth environments. The new methodology developed also
provides a broad-based analysis of plant metabolic responses to pathogen infection.

The broader importance of this work will best be assessed by using this information
in the study of other plant and pathogen systems. Leaves of the Col-0 ecotype of wild
type Arabidopsis thaliana have low levels of catechol-containing metabolites owing to
OMT enzymes which convert metabolic intermediates into methylated polyphenols such
as SM. The work described in this chapter points to opportunities to manipulate
expression of OMTs and genes in the phenylpropanoid and glutathione pathways to

enhance resistance to pathogens, and perhaps herbivores, without substantial detriment to

146

the host plant. The results also highlight the biochemistry of quinones and GSH
conjugation, which are well—studied topics in mammalian toxicology, and suggest that

greater awareness of their importance in plant systems could provide fruitful areas for

future research.

147

CHAPTER 3

UPLC MS/MS assay of jasmonates and related phytohormones for large-scale
screening of plant metabolic phenotypes

Portions of this Chapter describe a collaborative work that has been published, to which I contributed by

performing LC MS/MS method development and validation.

Chung, H. S., A. J. Koo, X. Gao, S. Jayanty, B. Thines, A. D. Jones and G. A. Howe (2008). "Regulation
and function of Arabidopsis JASMONATE ZlM-domain genes in response to wounding and
herbivory." Plant Physiol 146(3): 952-64.

Koo, A. J ., X. Gao, A. Daniel Jones and G. A. Howe (2009). "A rapid wound signal activates the systemic
synthesis of bioactive jasmonates in Arabidopsis." Plant J. In press.

148

3.1 Abstract

This report presents a selective and sensitive ultraperformance liquid chromatography
tandem mass spectrometry (UPLC-MS/MS) method for simultaneous analysis of
jasmonates and related phytohormones for large-scale screening of plant metabolic
phenotypes. Twenty-four jasmonates and related phytohormones are measured within 5
min using multiple reaction monitoring and a fast reversed phase gradient. Multiple
reaction monitoring (MRM) transitions were developed to distinguish leucine and
isoleucine conjugates of jasmonic acid (J A) even when chromatographic resolution is
incomplete. Isotope-labeled or dihydro analogs of these compounds were used as internal
standards to yield accurate quantification. Sample preparation requires no derivatization,
and the method gives detection at low frnol levels for all target metabolites. Endogenous
levels of JA, 12-hydroxy JA, ten different JA amino acid conjugates, and methyl ester
metabolites were measured. In the same analysis, the oxylipins OPDA, OPC:8 and
OPC:6 were also quantified, along with their precursor linolenic acid (LA) and
phytohormones salicylic acid, abscisic acid, indole-34acetic acid and the bacterial
phytotoxin coronatine. This method demonstrated presence of some J A metabolites in
Arabidopsis leaves for the first time, and was applied to the analysis of plant hormones
and metabolites from Arabidopsis thaliana leaves induced by wounding and the pathogen

Pseudamanas syringae DC3000 infection.
3.2 Introduction

Plants regulate grth and developmental processes by producing a diverse array of

specialized metabolite hormones (phytohormones). Since plant responses to biotic and

149

abiotic stress rely on changes in plant chemistry, phytohormones play an essential role in
evoking physiological responses that are essential for survival. Phytohormones act at
micromolar or lower concentrations that have presented challenges to the study of plant
hormones, and only since the 19705 have researchers made great strides in clarifying their
effects and importance (Srivastava 2002).

Among phytohormones, jasmonates are oxylipins that hold central importance in plant
responses to wounding and stress (Chiwocha et al. 2003). Jasmonates regulate plant
responses to biotic stresses including insect herbivory and pathogen infections as well as
abiotic stresses including high UV light intensity, air pollutants including ozone, and
osmotic stress. Shifts between growth- and defense-oriented metabolisms regulated by
jasmonates enable plants to adapt to rapid environmental change (Chung et al. 2008).
During pathogen infection, microbial metabolites also contribute to related signaling
networks. Coronatine, a bacterial toxin with structural and functional similarity to J A
and its amino acid conjugates, functions both as a signal mimic (Weiler et a1. 1994) and
inducer of J A biosynthesis (Laudert and Weiler 1998)

Plants convert jasmonic acid (JA) to a suite of modified metabolites, collectively
known as jasmonates, including JA methyl ester (MeJA), glycosyl esters, and amide-
linked conjugates with various amino acids (Staswick and Tiryaki 2004). Many
jasmonates regulate gene expression upward or downward in a regulatory network with
other plant hormones including salicylic acid (SA), auxin (indole-3-acetic acid, IAA), and
abscisic acid (ABA) (Wastemack 2007). Recent findings show that amino acid

conjugates of JA are more potent signaling compounds than JA itself (Staswick and

150

Tiryaki 2004; Thines et al. 2007), and these findings have stimulated a growing
recognition of the importance of JA metabolism in plant signaling.

Few analytical approaches have assessed crosstalk among these signaling
mechanisms through simultaneous and comprehensive measurements across a broad
range of phytohormones at multiple time points. Low throughput in sample processing
and metabolite analysis has limited such efforts, particularly with regard to the spatial and
temporal responses of plants to biotic and abiotic stresses. Since plants need to produce
hormones at specific times and locations, a comprehensive approach that yields
simultaneous quantification of numerous plant metabolites is essential for elucidating
their dynamics and functions. A GC/MS-based approach for measuring multiple
phytohormones was described recently (Schmelz et al. 2004), but GC/MS analyses often
require solvent evaporation and chemical derivatization steps (Wilbert et al. 1998;
Glassbrook et al. 2000; Chiwocha et al. 2003; Engelberth et al. 2003) that present barriers
to large-scale investigations.

An efficient analytical method is needed to provide selective and comprehensive
quantification of phytohormones with low detection limits, high precision and accuracy,
and experimental simplicity compatible with analyses for many treatments and time
points. While GC/MS methods have provided outstanding resolution, dynamic range, and
low susceptibility to matrix effects, LC methods offer advantages in simplicity and speed
of sample preparation coupled with short analysis times. Recent development of sub-2
um particles for ultraperformance LC (UPLC) provides separations that give superior

throughput with chromatographic peak widths competitive with capillary GC.

151

LC/MS/MS methods add selectivity of analyte detection, yielding final detection limits
for many analytes.

LC/MS/MS methods have been reported for a limited number of jasmonates such as
JA, JA Leu and JA Ile (Tarnogarni and Kodama 1998), and UPLC/T OF MS has been
employed to monitor an assortment of phytohormones and metabolites including several
hydroxylated JA derivatives (Glauser et al. 2008). Both of these reports describe LC
separations running more than 30 min/sample.

For more precise quantification of target metabolites, multiple reaction monitoring
(MRM) has emerged as the standard approach for quantitative LC/MS analyses, and was
used in the experiments described below. This approach uses the first mass analyzer
(M81) to filter all but the selected molecular ion, and the second mass analyzer (M82)
allows only a selected and characteristic fiagment ion to reach the detector. This
approach yields chromatograms selective for the target metabolite(s), filtering out signals
from all other substances that cannot satisfy the requirements for both molecular and
fragment ion masses and for chromatographic retention time. The specific experiment is
known as a. "transition" and can be written as parent ion > fragment ion. For example,

m/z 209 > m/z 59 is MRM for jasmonic acid (Figure 3.1).

 

 

 

M81 209 Collision cell 59 M82

 

 

 

 

 

 

 

 

 

 

 

Figure 3.1 Diagram for triple quatrupole and multiple reaction monitoring (MRM).
M81 stands for first mass analyzer and M82 stands for second mass analyzer. For
example, m/z 209 > m/z 59 is MRM for jasmonic acid. m/z 209 is the parent ion, which
was allowed to pass through M81. m/z 59 is the fragment ion which was allowed to pass
through M82.

152

In this study, we report a fast, sensitive, and selective UPLC MS/MS method for
quantifying JA metabolites, additional oxylipins, salicylates, abscisic acid, and auxin
without need for derivatization. We demonstrate the most comprehensive profiles of
jasmonates and related phytohormones using LC/MS/MS, including measurements of
some metabolites in Arabidopsis for the first time. Selective MRM transitions were
optimized to distinguish two coeluting isomers. Performance of the method is evaluated

for measurements of phytohormone levels in leaves upon wounding and pathogen stress.

3.3 Materials and methods
3.3.] Chemicals

All commercially available compounds and mobile phase additives were purchased
from Sigma (St. Louis, MO), and HPLC-grade solvents were purchased from VWR
Scientific. Structures of the phytohormones measured in this study are presented in

Figure 3.2.

153

Figure 3.2 Chemical structures of the 24 compounds and the internal standards used in
this study.

154

. .

COOH COOH
JA (”A
o O
OH
COOCH3 COOH
Me JA 120H JA

 

Figure 3.2 continued

 

 

 

O O
O N C CH 0
H
Hzc—CHz
JA ACC
0 JA Ala
0
0
O N—CH—C—OH
H 0
CH2
CH,
C=O
NH2 JA 116
JA Gln 0
O
O
O N_CH—C—OH
H
CH2
CH—CH3
CH,
JA Leu JA Met

156

N_TH—C_'OH
H

CH,

N—‘CH_C_OH
H

CH—CH,

CH,

 

CH,

O

N—CH—C_OH
H

 

Figure 3.2 continued

0
0 N—TH—C—OH o n— H—c—OH
H H T

“2 TH—OH

CH,

JAThr
JA Phe
O N—TH—C—OH
H
Tit—CH,
CH,
IA Val
O
O
OH
1 ll
0 — H— —
H T C 0H 0 u—TH—c—o—CH,
H
12 1*”
H
C 3 CH,
12-OH JA Ile IA Ile Me

157

Figure 3.2 continued

0
0 OH
OH
/// ///
0
ABA 0 NH
COOH
COOH “000
Cor
N OH
H
1AA. SA
H
o
H OH

0 O OH
0 OOH H0
H0
HO O OH
S

B-O-D glucosylsalicylic acid alicylic acid glucose ester

SA glucosides

158

3.3.2 Plant Material, Growth Conditions and Plant Treatments

Arabidopsis thaliana ecotype Columbia-0 (Col-0) was used for all experiments. Soil-
grown plants were maintained in a growth chamber at 22°C under 12 h light (100 uE rn'2
s") and 12 h dark.

For mechanical wounding treatments, fully expanded rosette leaves on 5-week-old
plants were wounded three times by crushing the leaf across the midrib with a hemostat.
This wounding protocol, which resulted in damage to approximately 40% of the leaf area,
was administered to approximately six rosette leaves per plant (Chiwocha et al. 2003).
For the pathogen stress treatment, 5-week-old plants were inoculated with Pseudamanas
syringae DC3000 (~10‘5 cfu/ml in 10 mM MgC12) via the stomata into the lower surface
of half of the leaf. About 20 uL of the bacterial suspension was injected with a syringe
via the stomata into the lower surface of half of the leaf. Control leaves were treated
similarly with a sterile solution of 10 mM MgC12 (Katagiri et al. 2002). Local and
systemic leaf tissues were harvested from control and stressed plants at 1 hour, 12 hours,
one day and two days after inoculation. Tissues were immediately frozen in liquid

. nitrogen. Five biological replicates were performed.

3.3.3 Sample Preparation

Around 300 mg (IO-14 leaves) of Arabidopsis thaliana leaves were harvested,
immediately fiozen in liquid nitrogen, ground and extracted with 2 ml methanol/water
(80:20 v/v) containing 0.1% formic acid and 0.1 g/L butylated hydroxytoluene (BHT) at
41] for 24 h. An aliquot of each internal standard solution was added to deliver 1.0 nmol

each of internal standards dihydro-JA, [13C6]-JA-isoleucine and [2H5]-12-oxo-

159

phytodienoic acid ([2H5]-OPDA). Homogenates were mixed and centrifuged at 12,000g
for 10 min at 4°C. Supematants were filtered through a 0.2 uM PTFE membrane
(Millipore, Bedford, MA) and transferred to autosampler vials without additional

processing.

3.3.4 Liquid Chromatography and Mass Spectrometry

Plant extracts (5 uL extract per injection) were separated on a UPLC BEH C18
column (2.1 X 50 mm, 1.7 um particles, Waters, Milford, MA) installed in the column
oven of an Acquity Ultra Performance Liquid Chromatography (UPLC) binary pump and
autosampler (Waters Corporation, Milford, MA). The performance of two additional LC
columns was tested: a fused core Ascentis Express C18 column (2.1 X 50 mm, 2.7 pm;
Supelco, Bellefonte, PA) and BetaBasic C18 column (1 X 150 mm, 5 pm; Therrno,
Bellefonte, PA). A gradient of 0.15% aqueous formic acid (solvent A) and methanol
(solvent B) was applied in a 5 min program with a mobile phase flow rate of 0.4 ml/min
(0.15 ml/min for the 1 mm ID column) as shown in Table 3.1. The column, which was
maintained at 50°C, was interfaced to a Quattro Premier XE tandem quadrupole mass
spectrometer (Waters Corporation, Milford, MA) operated using electrospray ionization.
The capillary, cone, and extractor potentials were set at 3 kV, 30 V, and 3V respectively.
Flow rates of cone gas and desolvation gas were 100 and 800 L/h, respectively. Ion
source and desolvation temperatures were 120 and 350 °C, respectively. Collision-
induced dissociation employed argon as collision gas at a manifold pressure of 3 x 10'3
mbar, and collision energies and source cone potentials were optimized for each

transition using Waters QuanOptimize software. This method was composed of two

160

negative ion functions (0-1.5 and 1.5-5.0 min) and one positive ion mode function
covering full run time to allow for adequate dwell time for each analyte. Data were
acquired under the control of MassLynx 4.0 software and processed for calibration and

for quantification of the analytes using QuanLynx software.

 

 

Time (min) Mobile phase A Mobile phase B
0.0 60% 40%

1.0 40% 60%

3.0 25% 75%

3.5 0% 100%

4.51 60% 40%

5.0 60% 40%

 

Table 3.1 Mobile phase gradient in HPLC separation. Mobile phase A =O.15% aqueous
formic acid, Mobile phase B= Methanol
3.3.5 Standard solutions

J asmonoyl-amino acid conjugates were prepared and the structures were verified by
GC-MS as described previously (Krarnell et al. 1997; Staswick and Tiryaki 2004).
OPC:6, OPC:8, OPDA, and dJA were prepared as previously described (Koo et al. 2006).
[13C6]-JA Ile was synthesized by conjugation of cis-(dz)-JA (Sigma, St Louis, MO, USA)
to ['3C6]-JA Ile (Cambridge Isotope Laboratories, Andover, MA, USA) as previously
described (Krarnell et al. 1997; Staswick and Tiryaki 2004). [2H5]-OPDA was prepared

according to Delker et al. (Delker et al. 2007).

3.3.6 Quantification of JA Ile in mixtures of JA Ile and JA Leu.

Two positive mode MRM transitions established capability to distinguish isomeric

Leu and Ile conjugates of IA based upon transitions of [M+H]+ to m/z 30 and 69,

respectively, even when chromatographic resolution is incomplete. To quantify of J A Ile

161

in mixtures of IA Ile and JA Leu, quantification curves were generated by plotting the

percentage of JA Ile verses the peak area ratios of the transition 324 > 69 to 324 > 30.

3.3.7 Annotation of SA glucoside (SAG) isomers.

To confirm assignments of the two isomers of SAG, extracts of pathogen-infected
leaves with high SAG levels were subjected to base hydrolysis by adding 10% aqueous
NaOH to a final pH of 12.5). The reaction was stopped after 24 hours by adding formic
acid to adjust the pH to 7. Levels of SAG and SA were assessed by LC MS/MS using the

protocol described above.

3.3.8 Method Validation.
3.3.8.1 Limit of Detection (LOD)

The limit of detection (LOD) for each analyte was determined, using method
calibration standard solutions, as three times the signal-to-noise ratio (SIN) as calculated
using MassLynx software. Noise levels were determined for a l-minute wide region of

each MRM transition chromatogram.

3.3.8.2 Linearity and calibration curve

For the creation of calibration curves, 20 11M stock solutions of each of the 22
unlabeled compounds were used to prepare different concentration of standard solutions
range from 2.5 nM to 5.0 nM. Dihydro-JA, [13C6]-JA Ile and [2H5]-OPDA were added as
internal standards to give fixed concentration of 0.5 uM each, as was the case for all plant

extracts. Data from analyses of these standards were used to create the calibration curves

162

for each compound by plotting the known concentration of each unlabeled compound as
x-axis and peak area ratio of unlabeled compound/fixed concentration of labeled internal

standard as y-axis using linear regression.

3.3.8.3 Calculations of matrix effect (ME %), recovery efficiency (RE %), and
process efficiency (PE %)

Preextracts. To measure the efficiency of recovery of the target compounds from plant
leaf extracts, dihydro-J A, ['3C6]-JA Ile and [2H5]-OPDA internal standards were added to
extraction buffer to make final preextracts with internal standards in each sample.
Postextracts. To measure matrix effects for dihydro-JA, [13C6]-JA Ile and [2H5]-OPDA
internal standards and provide a comparison with the preextracts, we evaluated the effects
of adding internal standards directly to filtered leaf extract to make final postextracts with

internal standards in the sample.

3.3.8.4 Validate intra- and interday accuracy and reproducibility.

The accuracy of the method was examined by analyzing quality control samples in
Arabidopsis leaf extracts using the method of standard additions. Four different
concentrations of three unlabeled oxylipin conjugates standards JA, JA Ile and OPDA
mix (0, 1, 3, or 5 pM) with fixed concentration of [13C6]-JA Ile internal standards (0.5
,uM) were spiked in 1x, 0.5x, and 0.25x extracts for four replicates. Standard curves were
generated by plotting the added concentration versus peak area ratio of phytohormone

standards/ internal standards, and the intra- and inter-day accuracies were calculated by

163

the following equation: accuracy (%) = 100 x calculated concentration/nominal

concentration.

3.4 Results and discussion

The purpose of this study was to develop a method capable of large-scale
measurements of target jasmonate and related phytohormones compatible with the needs
of comprehensive studies of plant responses to wounding and disease. Development of
the method was achieved by optimizing extraction, LC separation, and MS ionization and
fragmentation to resolve and distinguish all target metabolites. At the end, the potential of

this method as an aid to modeling plant metabolite profiles is evaluated.

3.4.1 Extraction procedure optimization

Six alternative suitable extraction systems were compared to optimize peak areas and
recoveries of the phytohormones from Arabidopsis leaves after pathogen infection.
Around 300 mg of leaves (10-14 leaves) from wounded Arabidopsis thaliana plants were
frozen in liquid nitrogen, ground, and extracted with 2 ml of different extraction solvents
at 4C] for 24 h. The six extraction solvents were (A) 50% methanol in water, (B) 80%
methanol in water, (C) isopropanol, (D) ethyl acetate, (E) 3:2 hexanezisopropanol (v/v),
and (F) 2:1 chloroform: methanol (v/v), with all solvents containing 0.1% formic acid
and 0.1 g/L BHT. Target phytohormones were quantified in each extract using
LC/MS/MS. Solvents B (80% methanol) and D (ethyl acetate) gave consistently higher

phytohormone levels (IO-20% more) compared with other solvents. Solvent B was

164

chosen for this study owing to its miscibility with water — a trait desirable for HPLC

separations.

3.4.2 Development of LC separations for the simultaneous quantification of
jasmonates and related phytohormones.

The method was developed and applied using a BEH C18 column (2.1 X 50 mm, 1.7
pm) (Waters, Milford, MA). Both the Waters and Supelco sub-3 um particle columns
gave exceptional separation efficiencies that allowed for steep solvent gradients and
improved chromatographic resolution relative to the longer BetaBasic C18 column (5 am
particles) while yielding narrow chromatographic peaks (wt/2 = 2.2 sec for JA Ile).

A mixture of 24 target analyte standards and 3 internal standards were separated using
a 5-minute LC separation and detected by MSIMS as shown in Figure 3.3. Many
synthetic JA amino acid conjugates and J A itself elute as two peaks because they are

mixtures of isomers.

165

Figure 3.3 Extracted ion MRM chromatograms of 27 standard compounds. A mixture of
24 standards for the target analytes and 3 internal standards were separated and detected
by HPLC-MS/MS using the transitions described in Table 3.2. The X-axis shows the
retention time, and Y-axis indicates the relative intensity being set as 100% with the

height of the highest peak of each measurement.

166

0.89

,2. SA

‘ L
c1 .. -.. A 4 .A...

' 0.50 ' 1.00 ' 1.50 ' 2.00 ' 2.50 3.00 3.50 4.00 4.50 5.00
1.27

 

 

 

*j J A

‘ L
1
c “I“:IIIVI‘YI ‘

.....,... .....2...,....,....,.........,....,....,......1..,....,....,.........wm
0.5 1.00 1.50 2.00 2.50 3.00 3.50 4.00 4.50 5.00

100- 0.?2

 

 

 

IZOH JA

 

 

A A A

 

 

‘ -
c IIIIIII U'UIIYYYT'II".

r. ..............,.........,....,....,.:.......,.........,.........,....,....,
0.50 1.00 1.50 2.00 2.50 3.00 3.50 4.00 4.50 5.00
0.98

 

 

0.50 1.00 1.50 2.00 2.50 3.00 3.50 4.00 4.50 5.00

1.10

1 1.03
,2. JA Ala

 

 

0.50 1.00 1.50 2.00 2.50 3.00 3.50 4.00 4.50 5.00

167

 

 

 

 

 

 

 

100
JA ACC
' 0.50 ' 1i0'0""'i'.50""'2'.00'""2'.50""'s'.'0'0""'3".5'0""'ai'.'0'0"' 4.50 ' 5.00
100- 0.40
SAG
3“ 0.50
c I I
""""""' 'I""l I I I 'l" I I 'I' I I l u u I I I
0.30 0.50 1.00 150 200 250 3.00 350 400 450 500
1001 1.01
0.96
at- JAThr
0 l I I I I I I I I I I I I I I I I I" 'I ' I
0.50 1.00 1.50 2.00 2.50 3.00 3.50 4.00 4.50 5.00
100 1.39
Cor
as
' 0'50 ' 1.00 "'1".50""'2".00' ' 2.50 ' 5'00 ' 5.50 ' 4.00 ' 4.50 ' 5'00
100 0 76
at? JAGIn
' 0.50 1.00 ' 1.50 ' 2.00 ' 2.50 ' 3.00 ' 3.50 ' 4.00 ' 4.50 ' 5.00

168

0.96

as. lZOH JA Ile

 

 

I
D 'II'lfl'fl'YUUIVI Y YY'U

I’l [TIIYIIVII'IVIVIYVIIIIVI-IAIIVll'IIII'IIIIIIIIIIIItI'IVIIIIUII—ITWI'UIII'ri'fl
0.50 1.00 1.50 2.00 2.50 3.00 3.50 4.00 4.50 5.00
1.51

 

.2. MA

 

 

0.00 0.50 1.00 1.50 2.00 2.50 3.00 3.50 4.00 4.50 5.00

2.29

OPC:6

l
c g A A k A _..
'I'

'I'V'IIIIIVUII I I I' Y Y I I'II'

,.........,. ,....,....,.... ....,:...,...‘......,....,.....:... .. , .2 , ...
0.50 1.00 1.50 2.00 2.50 3.00 3.50 4.00 4.50 5.00
m 4.03

 

 

 

 

 

 

 

0.30 0.50 1.00 1.50 2.00 2.50 3.00 3.50 4.00 4.50 5.00

OPDA

 

 

 

0.50 1.00 1.50 2.00 2.50 3.00 3.50 4.00 4.50 5.00

169

 

3.11

(IJIF'fifsz

 

 

 

.... ...... .........,.........F...,...:,...., ...,....,....,.........,.........,.....
.50 1.00 1.50 2.00 2.50 3.00 3.50 4.00 4.50 5.00

1.39

JA Val
1.51

 

A A A A

 

 

 

.......,....,....,...... ..,....,....,.:.......,..:.,....,....,....,....,.........,.........,....,
0.50 1.00 1.50 2.00 2.50 3.00 3.50 4.00 4.50 5.00

1.77

JA Ile

 

1.53
‘\

0.50 1.00 1.50 2.00 2.50 3.00 3.50 4.00 4.50 5.00

1.80
1.67.

JA Leu

K A
'I'IIIVIIVV'IUUIUII IIIIIIIIIII

 

 

 

 

I ' IUTY—IUIIII'IIIIIllI'IIII[VIII—[Ttil'l'lI]!!!I'V'II'IIII'VIII'IWUI
0.50 1.00 1.50 2.00 2.50 3.00 3.50 4.00 4.50 5.00

JA Met

 

11

 

 

VIII 'I', U‘ 'II" '1"

, ..., , ,.........,.........l.........,....,..............,....,....,....,.........,
0.50 1.00 1.50 2.00 2.50 3.00 3.50 4.00 4.50 5.00

170

1.67

 

 

100-
1 1.77
32; JAPIIC
0.1
0.50 1.00 1.50 2.00 2.50 3.00 3.50 4.00 4.50 5.00
m 1.58
‘ 1.53

 

 

0.50 1.00 1.50 2.00 2.50 3.00 3.50 4.00 4.50 5.00

0.72

3,, 1AA

i L
0.2. “ * ‘4A

.....,........................;"fifiup........,.:.......,.......‘......z......t...,.........,
0.50 1.00 1.50 2.00 2.50 3.00 3.50 4.00 4.50 5.00

 

 

 

 

 

 

 

m 1.70

I Me JA

32.

1

0-

0.50 1.00 1.50 2.00 2.50 3.00 3.50 4.00 4.50 5.00

m 2.03

. 1.92

,2; JAlle Me

U-

 

0.50 1.00 1.50 2.00 2.50 3.00 3.50 4.00 4.50 5.00

171

3.4.3 Development of multiple reaction monitoring (MRM) transitions for
phytohormone analysis.

All target compounds were detectable using electrospray ionization. Positive mode
ESI was employed to quantify IAA, the methyl esters Me JA, Me JA Ile, and for the
purpose of distinguishing JA Leu and JA Ile based on isomer-selective fragmentation. All
other phytohormones were measured using negative mode ESI. A pilot study
demonstrated that IAA yielded about lO-fold greater MRM signal in positive mode than
in negative mode. Despite the lack of a clearly basic functional group, the methyl esters
Me JA and Me JA Ile yielded abundant [M+H]+ ions in positive mode but [M-H]' ions
were not detected in negative mode.

All of the J A amino acid conjugates were detected in both positive and negative
modes. Negative mode has more than 10 times higher transition efficiency than the
positive mode owing to fewer competitive fragmentation pathways. Product ion spectra
of deprotonated J A amino acid conjugates were dominated by deprotonated amino acid
ions (e.g. m/z 130 for JA Ile) as the dominant fragment ions. In positive mode, CID
produced signal distributed over multiple fragment ions with none being dominant

JA Ile and IA Leu have similar structures, differing only the position of the side chain
branch point. A recently reported LC/MS/MS method does not distinguish J A Ile from JA
Leu, and values reported represent the sum of IA Ile plus JA Leu (Wang et al. 2007). In
the current study, these isomers were chromatographically resolved during analysis of
standard solutions (see Figure 3.3), but the minor isomer (JA Leu) could not be resolved
fiom the more abundant JA Ile in the plant extracts using an MRM transition common to

both metabolites. The fast UPLC separation did not resolve these isomers, and both gave

172

indistinguishable product ion spectra in negative mode. To solve this problem, collision
induced dissociation (CID) spectra were generated in positive mode for [M+H]+ ions.
Selective transitions were discovered: m/z 324 > 69 and 324 > 30 for JA Ile and JA Leu,
respectively. These transitions are analogous to those recently reported as selective
product ions from protonated Ile and Len respectively (Gu et al. 2007). To quantify J A
Ile in mixtures of J A Ile and J A Leu, calibration curves were generated using standards
with varying JA Ile/JA Leu ratios, and by plotting the percentage of IA Ile against the

peak area ratios of the transitions m/z 324 > 69: 324 > 30 (Figure 3.4).

A —- N N
O 01 0 0|
L 1 1

Peak area ratio of tI'anattlon 324->69
to 324>30
OI

 

O

 

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
Peconhge of JA Ile In JA Ito and JA Lou mixture

Figure 3.4 Quantification of IA Ile in mixtures of IA Ile and JA Leu. Two positive mode
MRM transitions established capability to distinguish isomeric Leu and Ile conjugates of

jasmonic acid based upon transitions of [M+I-I]+ to m/z 30 and 69, respectively, even
when chromatographic resolution is incomplete.

Oxylipins OPC:6, OPC:8, and precursor fatty acid LA yielded strong [M-H]' ions in
negative ion mode, but none yielded abundant and analytically useful fragments.
Although deprotonated ions fragmented by loss of C02, the high collision voltages (~60

V) needed to generate these fragments resulted in low ion transmission efficiencies (<

173

10%) in the mass spectrometer. Therefore, to provide low detection limits for these three
analytes, MRM channels were set up with collision energy set to zero and both analyzers
transmitted [M-H]', yielding selected ion monitoring (SIM) of these ions instead of
MRM. It is noteworthy that while JA and OPDA fragment efficiently, the analogs OPC:6
and OPC:8 are resistant to fragmentation. These differences are attributed to the greater
distances between the carboxylic acid group (where the charge is likely to be localized)
and carbonyl or double bonds. For these oxylipins and LA, SIM analysis yielded nearly
100-fold lower detection limits for standards (~ 1 final injected) than could be achieved
using MRM.

Analyses of Arabidopsis leaf extracts revealed several chromatographic peaks in the
MRM channel for m/z 291 > 165 (Figure 3.5). This transition detects the oxylipin OPDA
but only two isomers are commonly formed in plants, which result from two
stereochemical configurations at the 13-position. The additional peaks are attributed to
OPDA-containing glycerolipids that undergo partial fragmentation in the mass

spectrometer ion source (Buseman et al. 2006).

174

 

 

 

 

 

 

 

 

 

XLG_B75_0PDA 2: MRM of 13 Channels ES-
291.2>165.1
100a 2'71 3.5386
A .
aka
1
0 ....r....,..-.r.........,....,..............,.........,... ,....,....,-...,..............,....,....,
0.50 1.00 1.50 2.00 2.512 11.00 3.50 4.00 4.50 5.00
XLG_BG2_WP1_1d 2: MRM 01' 13 Channels ES-
291.2>165.1
100- 3'79 1.3455
B 1
. 3.90
$4
1.97 2.18 2.71 4.08
1.78 331 4.17
0 ..“14......",....,..-.,.--.,.........,....,....,....,... ,.724,fi-.1..5.,....,.m,....,....,....‘,Time
0.50 1.00 1.50 2.00 2.512 .00 3.50 4.00 4.50 .

 

Figure 3.5 MRM transition of m/z 291.2 > m/z 165.1 from OPDA standard (A) and
OPDA in pathogen stressed leave tissue (B). More than ten peaks showed up in Figure B.
The peak with same retention time (2.71min) to OPDA standard is the free OPDA in the
leave tissue. The other peaks are from the fi'agments of glycerolipids with OPDA
conjugates.

This method provides for simultaneous measurements of the phytohormone salicylic
acid (SA) and its two isomeric glucosides (SAGs): B-O-D glucosylsalicylic acid (GSA)
and the acylglucoside salicylic acid glucose ester (SGE) (Figure 3.3). To distinguish SGE
from GSA in an extract of Arabidopsis leaves, the extract was treated with base to effect
selective hydrolysis of the acylglucoside, as the phenolic glucoside is resistant to base
hydrolysis (Figure 3.6). The LC/MS/MS analyses of the untreated extract showed two
SAG peaks with the minor peak (tR = 0.50 min) giving a peak area 19% of the area of the
major isomer (tR = 0.40 min). After 24 h of base treatment, the peak for the minor isomer

disappeared (from 3703 to 0), the major SAG peak remained largely unchanged (from

19637 to 20780), and the signal corresponding to the hydrolysis product 8A increased

175

(from 7023 to 11643). Based on this result, we assign the early-eluting major isomer as
GSA and the minor isomer is identified as SGE. LC/MS detection and quantification of
SA glucoside isomers before and after base hydrolysis are shown in the Figure 3.7.
Additional evidence for the SAG annotation came from product ion MS/MS spectra upon
collision induced dissociation of [M-H]' for each isomer. The early eluting isomer gave
products at m/z 137 and 93 corresponding to the salicylate anion and decarboxylated
fragment respectively, whereas the later eluting isomer showed products at m/z 179
(deprotonated glucose), 161 (deprotonated anhydroglucose), 137, 93, and a radical ion at
m/z 136 that is attributed to homolytic cleavage of the glucose ester bond. These results
support the assignments made based on the hydrolysis result. Since the fiagment at rn/z
137 dominated the product ion spectra, the transition of m/z 299>137 was employed for

both isomeric SAGs.

176

2928?»: 83 swap—6 80.20% oEmoofiw <m wgmmswfiuma 9m oafi

9532.8 383.0

IO
IO OI

I0 0 IO
IOaZ 38:3

IO
CI

:0 o Iofo

30$ .98 88:3 28 232.8

...—28$me 02
3024 38:3
:0
O 0 IO
0...
o...
o
Ioazo

2mg 28 e__§_ee_>a8=_m-o-o-n

177

Area=19637

     

 

 

 

 

 

1005
l
A I
*j Area=3703
1
0.00 0.25 0.50 0.75 1.00 1.25 1.50
Area=20780
100-
B l
32-
i Area=0
0- Time

 

0.111 0.25 0.50 0.75 1.00 1.25 1.50

Figure 3.7 LC/MS detection and quantification of SA glucoside isomers before and after
base hydrolysis. (A) Before base hydrolysis. (B) After base hydrolysis. MRM transition
of m/z 299.1 > m/z 137.0 was used to select SA glucoside isomers.

After optimization of ion source cone and collision cell potentials, the data acquisition
was split between two negative ion mode functions, with the first monitoring 11 MRM
transitions from 0 to 1.5 min and the second monitoring 12 MRM transitions from 1.5 to
5.0 min. Eleven MRM transitions were included in function 1 and twelve in function 2.

The HPLC retention times and optimized MRM parameters for these 24 compounds are

summarized in Table 3.2.

178

220-ch :8 280692.:

5on : 3i .8 was; 0 Ea 0: S 3”; é €53 m {783% é was; < .83 203 amcmEmEsm Bees 85 .m
AL x389... 3 3x88 8m omega :o:mo::§c :8 Ammcmgfléfim BEBE 05. .m

.388 NS 8:838 can Eng 2: mo 88: 8:282 EBNtE 2: Java one :9: 0.88 mon :qufim Emma one 555 ._ ”8:02

.88: 5:232 83.98 98 wows—Q, :8 «52:8 wows—3 028 0958 358:3 .mcomzmcflu 322 NM 03am.

179

 

m N wwimmA NN wN ZEN A N.mmm 9H 3.
m N 551$; NN cm 3.2 A Nomm 2E 3.
m N 32.3”; NN ov 3x; A N53. 82 3.
B N 55.18.: NN 3. :62 A N.wNm 0: 3. on...
m N Kine; NN 3 :92 A N.NNm 0: 3.
m N cm. 3.0.— NN en :2: A N.NNm =3 3.
m N Siam.— NN em fie: A Ndom 3r 3.
m N 26. o N N.SN A N.maN «Ono
E N EN N we IVS A NéaN <95 mu...
0 N KN NN cw ~63 A N.SN <95
m N no...V o 9» N.S.N A NSNN <3
m N aN.N o um N.SN A N.m:N 9055
E N RA 3 vm 9% A : N 3:...
m _ cad NN :v :62 A Ndmm 0: 3. ION:
m _ cud NN em ed: A NNmm £0 <~
m 3 am..— wN NN :63 A N.w_m 80
m 3 53.36 E 3. N.SN A N3: m 5:. 3.
m _ omd .93 NN em oi: A _.aoN 838020 <m
m ~ no; 2 N ode A N.NmN 00¢. 3.
m _ 24.8.— NN MN :3 A Nde a? 3.
m _ wad o: N _.mm_ A :ch <m<
< _ de 2 N can A N.mNN 3. $0-2
< _ RNA 3 ”N 9% A :aoN 3.
m 3 35 NN em 93 A osfl <m
2.85% REES .o: Afizflofi: CO ow§o> CCowS—g AEEV com 8389 2.20980 -mm
8:05: cog—80¢ "86:30 280 A8: “8.58.5

 

180

 

mmmmmi’lm

v—iv—tv—IF-Iv-nv—flv-Il

mafia.—

mn. :8;

mm. 3.5.18:
mm. 3.5460.—
8.155.—

5;

N56

wN
wN
Vm
var.
Vm
3
2

Vm
wN
om
om
om
NN
NN

odw A Ndmm
:3 A Ndmm
odw A NéNm
ode A NéNm
odm A NéNm
_.~m_ A fimNN
99: A fie:

02 o: «3.
o: .3. 02...
2:50.“ «3.
o: <—.
:0..— (A.
43. 02
<5

+mm

3:328 am 2%:

181

3.4.4 Method validation
3.4.4.1 LOD
Detection limits ranged from 0.4 frnol for SA to 10 fmol for JA Val, which enable

quantification of the relative lower level plant hormones (Table 3.3).

3.4.4.2 Linearity and calibration curve

For JA Ile quantification, ['3C6]-JA Ile, obtained fi'om chemical synthesis, is six Da
heavier than JA Ile and can be used as an IS using isotope dilution methodology. The
same method can be used to quantify OPDA with [2H5]-OPDA as IS. For JA and 12-OH
.IA analysis, dihydro-IA (dJA) has often been used. dJA is easily prepared compared with
isotope labeled JA and is two mass units heavier than JA. For the other phytohormones,
standards that are stable isotope-labeled versions or structurally-related analogs are
scarce, therefore ['3C6]-JA Ile was used as their IS instead.

Solutions containing varying amounts of each unlabeled analyte compound and a
known, fixed amount of the corresponding internal standard (IS) were used to create
calibration curves. Calibration curves were linear over the range of 2.5 nM to 5 uM.
Average correlation coefficient limits (R2 values) are from 0.9800999.

The precision of the method is also dependent on the ratio of IS and analyte compound
and yields best precision when standard and analyte levels are within an order of
magnitude of one another. Thus, the amount of IS added to samples should be in the
range of the expected amount of the target compound. Though it is difficult to predict the
change of levels of phytohormones, 0.5 pM dihydro-JA, [13C6]-JA Ile and [2H5]-OPDA

internal standards were used in this study according to our previous experience.

182

3.4.4.3 Matrix effect (ME %), recovery efficiency (RE %), and process efficiency
(PE %)

Matrix effect (ME %), recovery efficiency (RE %), and process efficiency (PE %)
were calculated according to Matuszewski et al [Matuszewski, 2003 #92] with ME (%) =
100 (areaspiked after exuaetior/areastandard solution): RE (%) = 100(areaspiked before extraction/areaspiked
an“ extraction), and PE (%) = 100(areaspiked berm extraction/areasmm solution). The overall method
efficiencies ranged fiom 60%-90% (Figure 3.4), which indicates that ion suppression by
matrix constituents was not severe. The results suggest that the method is sufficiently

robust for reliable quantification at levels found in wounded Arabidopsis leaves.

 

 

ME (%) RE (%) PE (%)

mean CV(%) mean CV(%) mean CV(%)
dJA (1 pM) 89.3 4.3 71.2 5.7 63.6 6.5
130 JA Ile (0.5 M) 61.2 5.9 83.9 3.7 51.4 7.3
d5 OPDA (1 pM) 63.3 13.2 80.8 10.1 52.0 15.4

 

Table 3.3 Matrix Effect (ME %), Recovery Efficiency (RE %), Process Efiiciency (PE
%) of the three internal standards. (n=6)

3.4.4.4 Intra- and interday accuracy and reproducibility

The retention times only shifted 0.01 min for all of the phytohormones, and the
interday retention time precision (CV) ranged from 0.5 to 2%. These results indicate that
the HPLC conditions are reproducible fi'om run to run. Next, the intra- and interday
precisions of peak area ratios of JA, JA Ile and OPDA versus their individual stable
isotopically labeled internal standard were examined to evaluate the reproducibility of the

system. The intra- and inter-day CVs were less than or equal to 5%.

183

3.4.5 Quantification of 24 plant hormones in wounding and pathogen stress samples.

To test the feasibility of using this method to analyze plant samples, Arabidopsis
leaves were either mechanically wounded or infected with bacterial pathogen P. syringae,
followed by UPLC MS/MS analysis of leaf extracts.

In the control samples without any treatment, most of the phytohormones were at
levels less than 100 finol/mg fresh weight except GSA, OPDA and LA. Eight (JA Ala, JA
ACC, JA Met, JA Trp, Me JA, JA Ile Me, Cor, and IAA ) were not present at detectable
levels. After wounding stress, JA, 120H JA, 12OH JA Ile, LA, OPDA, OPC:6, and
OPC:8 were induced to more than IOO-fold greater levels than basal levels in unwounded
plants. JA-Ala, JA-Gln, JA—Phe, JA-Thr, and JA-Val were also induced but only 10- to
several hundred-fold. The biological functions of these signaling compounds, especially
JA and JA Ile, were described in a recent report resulting from our collaborative efforts
that employed a variation on this method (Koo et al. 2009). In Arabidopsis leaves, we
found J A Ile is more abundant than JA Leu afler wounding stress, accounting for 85-
100% of the total amount of JA Ile and JA Leu. Previous reports docurhenting the steep
rise and fall of JA levels in response to wounding imply the existence of an efficient
mechanism to metabolize them, which may involve hydroxylation, conjugation with
amino acids, and glycosylation. In this study, we reported levels of hydroxylated and
amino acid conjugates ran parallel with the induction of JA such as 12OH JA and JA Ile.

After pathogen stress, the major induced compounds in leaf extracts were SA, ABA,
GSA, and SGE, increasing by as much as several thousand-fold. Cor was produced de
novo by the pathogen, and its levels may provide a measure of the extent of pathogen

infection. The SGE was barely above limits of detection in control samples, but levels

184

increased by 50-fold after pathogen stress. However, GSA was shown as an abundant
compound both in control and pathogen stress samples. LA, OPC:6, OPC:8, and OPDA
were also induced, but less than 10-fold.

From Table 3.4, we conclude all phytohormones in this study can be detected in
stressed plants, with JA-Ala and JA-Phe only observed in wounding stress. Although the
mechanisms of induction of jasmonates and related compounds remain to be elucidate,
the changes in the levels of these compounds provides substantial evidence that they play
an important role in regulating gene expression and metabolite profiles. An in-depth
analysis of the dynamics of these compounds and their functional consequences is of
utmost importance in order to give an accurate description of metabolic responses in

plant-pathogen interactions.

185

 

888.8880 :305. .8 88.888 8:888 @888053 «c 83288:: 8: 8 836 858383 80: 88> :8880@ Mo 8850
8x8 8: 8 @3586 me 8528 :0»: 5:5 @8829: 803 8:2: 38:00 :08—:88

$3 _ 88.8 @838: @§ 885.5. N we :85 8885 05 8;.» @8858: 88> 8:8: 830882 mo 883 88888 :omofiwm e
.w8@::0>>

:88 a _ @888: 83 8.8: @3883 3004 88880: a 8:? 82,28 8: $88 $8: 3:: @3853 895 $83 .w:_@::o>> o
.m 8 8:3 3on $22 053808 2.: 89m 8088 _o>8 0888838808 8: 8 388.8% 88: :85 0:8 88:

-8488; a 3mm 85 :828 :o @8028 8888 2.: mm @o:_8:80@ 803 :8880@ .8 8:8: @otoaom .:o=o80@ .8 :85" DOA e

883838 308205 3:: .8.“ 3:885 :88 w8\_o~:b Gm H :88 mm @088me 8: £38 0:080:032m a

«888 8.86:3wa @8688 83:59. .& @5 @8883 38:8 no 8888 8 888083;: mo 2264 v. m 88d.

186

 

 

 

 

 

 

 

N_HoVN _Hm_ _Hm DOAV _._ <5
SHRN VH2 mHN_ mHS m.m <m<
QQEKSKQVAELNNQQ
hm H mi. @N H Eh 02 H 02 m @N H Kb 9— <4
2 H0: mHac M: Ham mHam c; oHUmO
bmH S» mev wNHONc mHom N._ %va
NVHawa 2 Ho; 00wHooN2 2 Hmmm 0; <95
3 H 00m GO..— v GOA v DOA v

a. O~Q§ mufitbsadfi
004v QOAV N H m _ QOAV v; 82 o: s.
N H mN GOAV m H 3. QOAV 0; S. 02
_ H v 004v _ H m 004v 2. E S
_ H n DOAV _ H w GOAV md 82 S.
_ H e QOAV N H w 904v 0.x 00¢. s.

QOAV Dogv N H mN no.7 in a? S.
904v DOAV N H VN _ H N. m6 85 5.
NH: mHo_ NH: NH” m6 .893.
NH: _H@ N_Hmm~ _H~. v.» 503.
mHm mHm hHmE VHm m.N :35.
wHNS .H5 2 HEN _H> 0.3 B> S.
N H 5‘ m H 3. 2: H owmm m H m». N.m 0: .3. EON—
VHaa NH3 02 H00: NH» NS 5. ZON—
_Hm_ mHS $850— VHN_ m.N 0: S.
mHoafl VH8 02: HcomoN mHZ. 0.0 5.
thGEQEhQK.
v H 00» _ H m _ ~ H 0 _ H h 085388 80: 5%
82 H cow: an H 03N 3 H omwN o: H 0va 03:83.38 80: <mO
08 H omgV cm H 2N NN_ H 02. 3. H NVN v.0 <m
g
e @8088 :0w050m 10.800 08804 038800 @800 w: 8809800

88385 8883 8883 3:8 .00.:

187

3.5 Conclusions

A fast UPLC MS/MS method was used to generate quantitative profiles of 24
jasmonate-related metabolites. Six different extraction procedures were compared and the
procedure with the most consistently high efficiency involved extraction with
methanol/water (80:20). Twenty-three oxylipin and salicylate phytohormones are
measured within 5 min using a UPLC BEH C18 column and rapid gradient multiple
reaction monitoring (MRM) transitions selective for each compound. Most compounds
were analyzed using negative mode electrospray ionization. Positive mode ESI was
chosen for methylated conjugates methyl jasmonate, JA-isoleucine methyl ester, JA Leu
and JA Ile. Two positive mode MRM transitions established capability to distinguish
isomeric Leu and Ile conjugates of jasmonic acid based upon transitions of [M+H]Jr to
product ions at m/z 30 and 69, respectively, even when chromatographic resolution is
. incomplete. For acidic lipids (OPC:6, OPC:8, linolenic acid) that are ionized well but are
difficult to fragment, collision energy is set to zero and selected ion monitoring is
performed within the same method. Compared to derivatization-based GC/MS methods,
this UPLC-MS/MS approach involves minimal sample processing and avoids sample
losses during evaporation and derivatization, requires only 5 min instrument time per
sample, and yields precise quantification of a broad range of jasmonates and related
metabolites down to less than five femtomoles. The method is well-suited for monitoring
the dynamics of signaling metabolites in studies of wounding, pathogen stress, and other
plant stress, and is recommended to laboratories as a practical alternative to

measurements of single phytohormones.

188

CHAPTER 4

Conclusions and perspectives

189

4. Conclusions and perspectives
4.1 Conclusion of thesis project

At the outset of this project, nearly all published reports on metabolomics relied on
mature GC/MS technologies. These experimental approaches provided extensive data
sets, but were limited to detection of small primary metabolites such as amino acids,
organic acids, and simple sugars. Prior to my dissertation research, published reports of
nontargeted profiling of metabolites using LC/MS were rare and were limited by the lack
of LC/MS spectrum databases and long HPLC separations that precluded analyses of
large sample sets. For these reasons, LC/MS was viewed as a powerful method for
measuring known secondary metabolites, but successes in identifying. new and
unanticipated metabolites were few. In addition, developments of LC/MS metabolomic
approaches for identifying gene fiinctions were in their infancy when these projects were
initiated.

Despite the slow start, metabolite profiling has emerged a fast-growing technology
for phenotyping and diagnostic analyses of plants (Schauer and Fernie 2006). In this
study, my efforts have focused on developing and applying two complementary
approaches for using metabolite profiling to expand functional annotation of genes and to
build a more comprehensive understanding of the cellular responses to changing
environments. In Chapter 2, nontargeted LC/MS metabolite profiling of wild type and
omt1 mutant Arabidopsis plants was performed under control and pathogen stress
conditions to reveal mechanisms of direct and pleiotropic influences of the OMT1 gene.
The findings identified roles for specific polyphenols, their quinone metabolites, and

glutathione in pathogen resistance. The results of this study suggested potential

190

mechanisms by which changing the extent of polyphenol methylation could modulate
levels of metabolites capable of redox cycling, generating ROS, and influencing
phytohormone and phytoalexin signaling networks. In Chapter 3, a fast, accurate, and
targeted analytical method was developed to measure a wide assortment of
phytohormones in regulation of gene expression and signal transduction in wounding
stress. These two approaches bear similarities to nontargeted and targeted measurements
of gene expression, in the manner that microarrays and quantitative real-time PCR are
used in modern research.

Most previous publications on the subject of nontargeted metabolite profiling in
plants have employed plants grown under carefully controlled environments to minimize
unintended phenotypic variations. Though a few concurrent studies have broken with
this tradition to study stress responses (Boccard et al. 2007; Kim et al. 2007; Alvarez et
a1. 2008; Glauser et al. 2008), there remain great opportunities to pursue this kind of
approach to learn about roles for specific genes and environmental conditions on plant

phenotypes.

4.2 Opportunities for future research
4.2.1 Designer plants for control of OMT function

Since the amt] mutant showed the ability to inhibit bacterial growth, metabolic
engineering of plants for optimal production of o-dihydroxylated polyphenols may be

useful to increase plant resistance to pathogen infection.

4.2.2 More studies on quinones and glutathione conjugates

191

It is hoped that the finding of polyphenol quinone metabolites that exhibit anti-
microbial properties will raise awareness of their potential in enhancing plant resistance
to pathogens, and perhaps, other stress conditions. Since there are such big pools of
polyphenols (flavonoids) which have the potential to be oxidized to quinones, it will be -
interesting to identify more quinones and study the functions of these compounds.
Metabolic conjugation of GSH to reactive electrophilic metabolites is well-documented
as a detoxification mechanism in animals (Chroust et al. 2001; Coles and Kadlubar 2003;
Pastore et al. 2003), but our understanding of GSH conjugation in plants is still an
emerging field of study. The potential importance of GSH conjugation is suggested by 54
glutathione S-transferase genes in the genome of Arabidopsis thaliana (Marrs 1996;
Neuefeind et al. 1997; Dixon et al. 2009).

Furthermore, several electrophilic oxylipin metabolites have been observed to form
GSH conjugates (Davoine et al. 2005), and this may represent a competing sink for
quinone metabolites. In this study, some polyphenolic-glutathione conjugates were
discovered. My preliminary studies that have employed LC/multiplexed CID for
metabolite profiling have suggested the presence of dozens of glutathione conjugates in
extracts of plant leaves. Understanding the interaction of secondary metabolite
biosynthetic pathways with glutathione S-transferase functions may lead to novel

discoveries in the area of plant biochemistry.

4.2.3 New trends in plant metabolomics

Integration of ‘omics’ studies

192

Currently, most plant stress response studies use one or two “omics” approaches of
those ‘omics’ studies (Weckwerth 2008). Integrated analysis of metabolite, transcript,
and protein levels in plant systems identified several important features of plant
metabolic regulation (Hirai et al. 2005; Le Lay et al. 2006). In the future, using a
combination of all three “omics” approaches or more will unleash the potential of
systems biology to uncover mechanisms regulating plant stress response on various levels

(Shulaev et al. 2008).

Plant micrometabolomics: the analysis of single cells or tissues

Most metabolomic studies have been conducted at the whole tissue level.
Metabolomics within specific cell types or single cells is of particular interest for natural
product chemistry, chemical ecology, and biochemistry (Moco et al. 2009) because
functions are organized at cellular and subcellular levels. Plants are considered to have
about 40 different cell types, therefore, they are likely to have different metabolites
profiles in different cell types (Martin et al. 2001). Localization of particular metabolites
in certain organelles is important to interpret functions of those metabolites.

Several technical challenges confront investigators hoping to use metabolomics to
discover new gene functions and enhance understanding of cellular physiology. As
mentioned above, metabolomic studies would be most useful if they integrated
nontargeted and targeted metabolite data for multiple genotypes, treatments, time points,
and tissue or cell types. For this to become practical, further improvements in pipeline
throughput and analytical limits of detection will be needed. Ongoing efforts in these

areas promise to make important advances over the coming years, and we may expect

193

metabolome analyses to have more prominence in functional genomic studies of plants

and other organisms.

194

Appendix A

Levels and chemical structures of metabolites
WC: wild type control, with the infiltration of MgClz buffer
MC: mutant control, with the infiltration of MgClz buffer
WP: wild type with the infiltration of P. syringae
MP: mutant with the infiltration of P. syringae.

4-Methylsulfinylbutyl glucosinolate (m/z 436.04)

160 .
14o «
120 .

100 r

 

 

-I-WC
+WP
-I-MC

 

 

 

 

\
N , . +MP
60 \- 'I

404 \‘

20*

 

 

Integrated XIC peak areas / mg fresh leaf weight

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Time (h)
T-test analysis on metabolite levels
1h 1d .2d 3d
WC vs MC N.S. N.S. N.S. N.S.
WP vs MP N.S. N.S. N.S. N.S.
WC vs WP N.S. N.S. * *
MC vs MP N.S. N.S. * *
“*” significant, p < 0.05
“N. S.” not significant

ENJFW; Him

195

8-Methylsulfinyloctyl glucosinolate (m/z 492.11)

300~

250*

200

 

50*

 

Integrated XIC peak areas / mg fresh leaf welght

 

 

 

 

-I-WC
+WP
-I-MC
+MP

 

 

 

 

30

4O 50

80

 

 

 

 

 

 

 

 

 

 

 

0
Time (h)
T-test analysis on metabolite levels
1h 1d 2d 3d

WC vs MC N.S. N.S. N.S. N.S.
WP vs MP * N.S. N.S. N.S.
WC vs WP N.S. N.S. * N.S.
MC vs MP * N.S. * N.S.

“*” significant, p < 0.05

“N.S.” not significant

0303 H
\$ 8

196

RH

 

4-Methylthiobutyl glucosinolates (m/z 420.05)

400 -

350 J

 

300 ~

50

Integrated XIC peak areas / mg fresh leaf weight

 

80

 

T—test analysis on metabolite levels

 

 

-I-WC
+WP
-I-MC
+MP

 

 

 

 

 

 

 

 

 

 

 

 

 

1h 1d 2d 3d
WC vs MC N.S. N.S. N.S. N.S.
WP vs MP N.S. N.S. N.S. N.S.
WC vs WP N.S. N.S. * N.S.
MC vs MP N.S. N.S. * N.S.
“*” significant, p < 0.05
“N.S.” not significant
N0803 HO
H0

197

 

8-Methylthiooctyl glucosinolate (m/z 476.11)

350

 

 

H
..r:
.9.”
Q)
3 300
Q...
CU
.33
'5 250'
Q
d: V.
E” \
\ 20° \ ‘
CD
8 \
3; \\
o) \
0. \
9 100+ \
X
@
3
a 50~
Q) ‘
H
:
H
0 j T Y
0 1o 20 30 4o 50 60 7o

80

 

 

—I—WC
+WP
+MC
+MP

 

 

 

 

 

 

 

 

 

 

 

 

 

Time (h)
T-test analysis on metabolite levels
1h 1d 2d 3d
WC vs MC N.S. N.S. N.S. N.S.
WP vs MP N.S. N.S. N.S. N.S.
WC vs WP N.S. N.S. * *
MC vs MP N.S. N.S. * *
“*” significant, p < 0.05
“N .8.” not significant
030314

\3 3
H0

198

 

3-Methylsulfinylpropyl glucosinolate (m/z 422.06)

167

14«

12~

10‘

 

 

Integrated XIC peak areas / mg fresh leaf weight

 

 

 

-I-WC
+WP
+MC
+MP

 

 

Time (h)

T-test analysis on metabolite levels

 

 

 

 

 

 

 

 

 

 

 

 

1h 1d 2d 3d
WC vs MC N.S. N.S. N.S. N.S.
WP vs MP N.S. N.S. N.S. N.S.
WC vs WP N.S. N.S. * *
MC vs MP N.S. N.S. * *
“*” significant, p < 0.05
“N.S.” not significant
HO

NOSO3
/é\/\/IL3

H0

199

HH

 

4-Methoxyindolyl or l-Methoxyindolyl glucosinolates (m/z 477.07)

70

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

E

.2?

d.)

3

Q—l

E

.:

58

cl:

DO

8 +wc

E +WP

8 +Mc

F: 30 q +MP

a H

B 20 «

><

@

9:; 1o-

80

Q)

E

0 I T . 1
0 40 60 70 80
Time (h)
T-test analysis on metabolite levels
1h 1d 2d 3d

WC vs MC N.S. N.S. * N.S.

WP vs MP N.S. * N.S. *

WC vs WP N.S. * * *

MC vs MP N.S. N.S. *
“*” significant, p < 0.05
“N.S.” not significant

HO
we”
N0803H H
OH
\ S H
HO
Nx
OCH3

200

SA (m/z 137.04)

100*

(O
O
l

 

 

 

Integrated XIC peak areas / mg fresh leaf weight

 

 

0 10 20 30 40 50 60 70 80

 

 

~I—WC
+WP
-I-MC
+MP

 

 

 

 

 

 

 

 

 

 

 

 

Time (h)
T-test analysis on metabolite levels
1h 1d 2d 3d
WC vs MC * N.S. N.S. N.S.
WP vs MP N.S. N.S. N.S. *
WC vs WP * * *
MC vs MP N.S. * *
“*” significant, p < 0.05
“N.S.” not significant

COOH

OH

201

 

 

ABA (m/z 263.13)

1.4~

1.2 -

0.8 a

0.6 ~

0.4 “ /
-

 

 

 

 

Integrated XIC peak areas / mg fresh leaf weight

 

 

 

—I—WC
+WP
-I-MC
+MP

 

 

 

 

 

 

 

 

 

 

 

 

 

o {0 2‘0 3'0 40 5‘0 61) 7‘0 8‘0
Time (h)
T-test analysis on metabolite levels
1h 1d 2d 3d

WC vs MC N.S. N.S. N.S. N.S.

WP vs MP N.S. N.S. N.S. *

WC vs WP N.S. * * *

MC vs MP N.S. * *
“"‘” significant, p < 0.05
“N.S.” not significant

0 Q
/
OH / COOH

202

 

120H JA (m/z 225.15)

8

18*

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

in
Q)
3
“-u
i3
.c:
8
d:
D0
5 —I—wc
8 +WP
a) -I—MC
i3 +MP
% 41"
Q)
CL
B
X
'e
8
C6
&
Q)
E
Ffi 0 r, T r
0 10 20 30 40 50 60 70 80
Time (h)
T-test analysis on metabolite levels
1h 1d 2d 3d
WC vs MC * N.S. N.S. N.S.
WP vs NIP *. N.S. N.S. N.S.
WC vs WP N.S. N.S. N.S. *
MC vs MP * N.S. N.S. *
“*” significant, p < 0.05
“N.S.” not significant
0

OH

COOH

203

 

JA Ile (m/z 322.20)

6a

 

 

—I—WC
+WP
-I-MC
+MP

 

 

 

 

 

 

 

 

 

60 70 80

Integrated XIC peak areas / mg fresh leaf weight

 

 

 

 

 

 

 

T-test analysis on metabolite leve
1h 1d 2d 3d
WC vs MC * N.S. N.S. N.S.
WP vs MP *. N.S. N.S. *
WC vs WP N.S. N.S. N.S. *
MC vs MP N.S. N.S. N.S.

 

 

 

 

 

 

“*” significant, p < 0.05
“N.S.” not significant

 

0
ll
0 N—CH_C_OH
H
CH—CH3
CH2
CH3

204

OPDA (m/z 291.20)

100 ~
90 ~
80 .

70‘

 

 

 

—I—WC
+WP
+MC
+MP

 

 

 

 

 

 

 

Integrated XIC peak areas / mg fresh leaf weight

80

 

 

 

 

 

 

 

 

 

 

 

 

 

Time (h)
T-test analysis on metabolite levels
1h 1d 2d 3d
WC vs MC N.S. N.S. N.S. N.S.
WP vs MP N.S. N.S. N.S. N.S.
WC vs WP N.S. N.S. N.S. *
MC vs MP N.S. N.S. N.S. *
“*” significant, p < 0.05
“N.S.” not significant

0

COOH

205

OPC:8 (m/z 293.2)

250 *

200 ~

150 4

100 *

 

50*

 

 

Integrated XIC peak areas / mg fresh leaf weight

 

 

70 80

 

 

—I-WC
+WP
-I-MC
+MP

 

 

 

 

 

 

Time (h)
T-test analysis on metabolite levels
1h 1d 2d 3d
WC vs MC N.S. N.S. N.S. N.S.
WP vs MP N.S. N.S. N.S. N.S.
WC vs WP N.S. N.S. N.S. *
MC vs MP N.S N.S. N.S. *

 

 

 

 

 

 

 

“*” significant, p < 0.05
“N.S.” not significant

0

COOH

206

 

LA (m/z 277.22)

120 7

100 7

80*

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

3:"
.29
Q)
3
Q—t
i3
a
8
d:
00
E +wc
2 so . +WP
8 -I—MC
:3 +MP
§
9.. 40 -
B
X
B 20 +
C6
Eb
Q)
E o '
T-test analysis on metabolite levels Time (h)
1h 1d 2d 3d
WC vs MC N.S. N.S. N.S. N.S.
WP vs MP N.S. N.S. N.S. N.S.
WC vs WP N.S. N.S. N.S. *
MC vs MP N.S. N.S. N.S. *
“*” significant, p < 0.05
“N.S.” not significant
0

OH

 

3-0-, 7-O-dirhamnosyl kaempferol (m/z 577.16)

 

20*

10«

Integrated XIC peak areas / mg fresh leaf weight

 

 

 

0 1 0 20 30 40 50 60
Time (h)

T-test analysis on metabolite levels

70

80

 

 

—I—WC
+WP
+MC
+MP

 

 

 

 

 

 

 

 

 

 

 

 

 

1h 1d 2d 3d
WC vs MC * * * *
WP vs MP N.S. N.S. * N.S.
WC vs WP N.S. N.S. N.S. N.S.
MC vs MP N.S. N.S. N.S. N.S.
“*” significant, p < 0.05

“N.S.” not significant

H0

208

 

 

 

3-O-glucosyl, 7-O-rhamnosyl kaempferol (m/z 593.15)

400

 

 

 

 

 

 

 

H
'50
-5 350 4
3
“a
2 300
A:
m
o
d: 250 '
OED —I—WC
\ ,' + WP
200
a -I- MC
‘1’ +MP
a
‘fi 150 ‘l
o
a.
u 100 «
#—
X
en
*5 o .
"‘ o 10 20 30 4o 50 60 7o 80

 

 

 

 

 

 

 

 

 

 

 

 

Time (h)
T-test analysis on metabolite levels
1h 1d 2d 3d
WC vs MC * * * *
WP vs MP N.S. N.S. * *
WC vs WP N.S. * N.S. N.S.
MC vs MP N.S. N.S. N.S. N.S.
“"‘” significant, p < 0.05
“N.S.” not significant

HO

 

209

3-O-glucosylkaempferol (m/z 431.10)
45 .
4o «
35 4
30 ~
25 «
20 ~

15*

 

Integrated XIC peak areas / mg fresh leaf weight

 

0 1o 20 3o 40 50 60
Time (h)

T-test analysis on metabolite levels

70

80

 

 

-I—WC
+WP
-I-MC
+MP

 

 

 

 

 

 

 

 

 

 

 

 

 

1h 1d 2d 3d
WC vs MC N.S. * * *
WP vs MP N.S. N.S. N.S. N.S.
WC vs WP N.S. N.S. * *
MC vs MP N.S. * * *
“*” significant, p < 0.05
“N.S.” not significant
HO
OH
OH

 

210

 

 

3-O-glucosyl, 7-O-rhamnosyl kaempferol with addition of OH (m/z 609.15) (the

postion of -OH group is not identified)

1.47.

1.2 -

 

 

 

 

0.4

0.2 2

 

 

 

 

 

Integrated XIC peak areas / mg fresh leaf weight

 

80

 

 

 

 

 

 

 

 

 

 

 

 

Time (h)
T-test analysis on metabolite levels
1h 1d 2d 3d
WC vs MC N.S. * * *
WP vs MP N.S. * * *
WC vs WP N.S. * N.S. N.S.
MC vs MP N.S. N.S. N.S. N.S.
“*” significant, p < 0.05
“N.S.” not significant

HO

 

211

 

 

—I—WC
+WP
+MC
+MP

 

 

OPDA-GS (m/z 598.29)

4.5 -

 

-I-WC
+WP
-I-MC
+MP

 

 

 

 

 

Integrated XIC peak areas / mg fresh leaf weight

 

 

80

 

 

 

 

 

Time (h)
T-test analysis on metabolite levels
1h 1d 2d 3d
WC vs MC N.S. N.S. N.S. N.S.
WP vs MP N.S. N.S. N.S. N.S.
WC vs WP * * N.S. N.S.
MC vs MP N.S. N.S. N.S. N.S.

 

 

 

 

 

 

 

“*9,

significant, p < 0.05
“N .8.” not significant

0

COOH
GS

212

 

dinor OPDA-GS (m/z 570. 26)

2.5 1

 

+WC
+WP
+MC
+MP

 

 

 

 

 

 

 

 

 

Integrated XIC peak areas / mg fresh leaf weight

 

 

 

 

 

 

 

 

 

 

 

 

 

Time (h)
T-test analysis on metabolite levels
1h 1d 2d 3d

WC vs MC N.S. N.S. N.S. . N.S.

WP vs MP N.S. N.S. N.S. N.S.

WC vs WP N.S. N.S. N.S. N.S.

MC vs MP N.S. N.S. N.S. N.S.
“*” significant, p < 0.05
“N.S.” not significant

0
COOH

GS

213

References
Agrios, G. N. (2005). Plant pathology” Elsevier academic press.

Alfenito, M. R., E. Souer, C. D. Goodman, R. Buell, J. Mol, R. Koes and V. Walbot

(1998). "Functional complementation of anthocyanin sequestration in the vacuole
by widely divergent glutathione S-transferases." Plant Cell 10(7): 1135-49.

Alonso, J. M. and J. R. Ecker (2006). "Moving forward in reverse: genetic technologies
to enable genome-wide phenomic screens in Arabidopsis." flat Rev Genet 7(7):
524-36.

Alvarez, S., E. L. Marsh, S. G. Schroeder and D. P. Schachtman (2008). "Metabolomic
and proteomic changes in the xylem sap of maize under drought. " Plant Cell and
Environment 31(3): 325-340.

Ames, B. N. (1983). "Dietary carcinogens and anticarcinogens. Oxygen radicals and
degenerative diseases." Science 221(4617): 1256-64.

Antolovich, M., D. R. Bedgood, Jr., A. G. Bishop, D. Jardine, P. D. Prenzler and K.
Robards (2004). "LC-MS investigation of oxidation products of phenolic
antioxidants." Aggie Food Chem 52(4): 962-71.

Arabidopsis Genome Initiative. (2000). "Analysis of the genome sequence of the
flowering plant Arabidopsis thaliana." Nature 408(6814): 796-815.

 

Asai, T., G. Tena, J. Plotnikova, M. R. Willmann, W. L. Chiu, L. Gomez-Gomez, T.
Boller, F. M. Ausubel and J. Sheen (2002). "MAP kinase signalling cascade in
Arabidopsis innate immunity." Nature 415(6875): 977-83.

Attaran, E., M. Rostas and J. Zeier (2008). "Pseudomonas syringae elicits emission of the
terpenoid (E,E)-4,8,12-trimethyl-1,3,7,1 l-tridecatetraene in Arabidopsis leaves
via jasmonate signaling and expression of the terpene synthase TPS4." Mol Plant
Microbe Intera_c_t 21(11): 1482-97.

Bargmann, B. 0., A. M. Laxalt, B. ter Riet, C. Testerink, E. Merquiol, A. Mosblech, A.
Leon-Reyes, C. M. Pieterse, M. A. Haring, I. Heilmann, D. Bartels and T. Munnik
(2009). "Reassessing the role of phospholipase D in the Arabidopsis wounding
response." Plant Cell E_nviron 32(7): 837-50.

Barry, K. M. (2002). "Effect of season and different fungi on phenolics in response to
xylem wounding and inoculation in Eucalyptus nitens. (vol 32, pg 163, 2002)."
Forest Pathology 32(6): 403-403.

Bednarek, P. and A. Osbourn (2009). "Plant-microbe interactions: chemical diversity in
plant defense." Science 324(5928): 746-8.

 

214

Birkenmeier, G. F. and C. A. Ryan (1998). "Wound signaling in tomato plants. Evidence

that aba is not a primary signal for defense gene activation." Plant Physiol 117(2):
687-93.

Blain, P. G. (1990). "Aspects of pesticide toxicology." Adverse Drug React Acute
Poisoning Rev 9(1): 37-68.

Blumenthal, C. Z. (2004). "Production of toxic metabolites in Aspergillus niger,
Aspergillus oryzae, and Trichoderma reesei: justification of mycotoxin testing in
food grade enzyme preparations derived from the three fungi." Re Toxicol
Pharmacol 39(2): 214-28.

Boccard, J., E. Grata, A. Thiocone, J. Y. Gauvrit, P. Lanteri, P. A. Carrupt, J. L.
Wolfender and S. Rudaz (2007). "Multivariate data analysis of rapid LC-TOF/MS

experiments fi'om Arabidopsis thaliana stressed by wounding." Chemometrics and
Intelligent Laboratory Systems 86(2): 189-197.

Bolwell, P. P., A. Page, M. Pislewska and P. Wojtaszek (2001). "Pathogenic infection
and the oxidative defences in plant apoplast." Protoplasma 217(1-3): 20-32.

Bostock, R. M. (1999). "Signal conflicts and synergies in induced resistance to multiple
attackers." Physiological and Molecular Plant Pathology 55(2): 99-109.

Bratton, S. B., S. S. Lau and T. J. Monks (1997). "Identification of quinol thioethers in
bone marrow of hydroquinone/phenol-treated rats and mice and their potential
role in benzene-mediated hematotoxicity." Chem Res Toxicol 10(8): 859-65.

Bratton, S. B., S. S. Lau and T. J. Monks (2000). "The putative benzene metabolite 2,3,
5-tris(glutathion-S-yl)hydroquinone depletes glutathione, stimulates
sphingomyelin turnover, and induces apoptosis in HL-60 cells." Chem Res
Toxicol 13(7): 550-6.

Browse, J. (2005). "Jasmonate: an oxylipin signal with many roles in plants." Vitam
Horm 72: 431-56.

Buchanan, B. B., W. Gruissem and R. L. Jones (2001). Biochemi§try & Molecular
Biology of Plants. Wiley.

Buell, C. R., V. Joardar, M. Lindeberg, J. Selengut, I. T. Paulsen, M. L. Gwinn, R. J.
Dodson, R. T. Deboy, A. S. Durkin, J. F. Kolonay, R. Madupu, S. Daugherty, L.
Brinkac, M. J. Beanan, D. H. Hafi, W. C. Nelson, T. Davidsen, N. Zafar, L. Zhou,
J. Liu, Q. Yuan, H. Khouri, N. Fedorova, B. Tran, D. Russell, K. Berry, T.
Utterback, S. E. Van Aken, T. V. Feldblyum, M. D'Ascenzo, W. L. Deng, A. R.
Ramos, J. R. Alfano, S. Cartinhour, A. K. Chatterjee, T. P. Delaney, S. G.
Lazarowitz, G. B. Martin, D. J. Schneider, X. Tang, C. L. Bender, 0. White, C.
M. Fraser and A.‘ Collmer (2003). "The complete genome sequence of the

215

Arabidopsis and tomato pathogen Pseudomonas syringae pv. tomato DC3000."
Proc Natl Acad Sci U S A 100(18): 10181-6.

Burga, L., F. Wellmann, R. Lukacin, S. Witte, W. Schwab, J. Schroder and U. Matem
(2005). "Unusual pseudosubstrate specificity of a novel 3,5-dimethoxyphenol O-

methyltransferase cloned from Ruta graveolens L." Arch Biochem Biophys
440(1): 54-64.

Buseman, C. M., P. Tamura, A. A. Sparks, E. J. Baughman, S. Maatta, J. Zhao, M. R.
Roth, 8. W. Esch, J. Shah, T. D. Williams and R. Welti (2006). "Wounding
stimulates the accumulation of glycerolipids containing oxophytodienoic acid and

dinor-oxophytodienoic acid in Arabidopsis leaves." Plant Physiology 142(1): 28-
39.

Buseman, C. M., P. Tamura, A. A. Sparks, E. J. Baughman, S. Maatta, J. Zhao, M. R.
Roth, S. W. Esch, J. Shah, T. D. Williams and R. Welti (2006). "Wounding
stimulates the accumulation of glycerolipids containing oxophytodienoic acid and
dinor-oxophytodienoic acid in Arabidopsis leaves." Plant Physiol 142(1): 28-39.

Chapple, C., B. W. Shirley, M. Zook, R. Harnmerschmidt and S. S. C., Eds. (1994).
Secondary metabolism in Arabidopsis. . Cold Spring Harbor Laboratory Press.

Chapple, C. C., T. Vogt, B. E. Ellis and C. R. Somerville (1992). "An Arabidopsis mutant
defective in the general phenylpropanoid pathway." Plant Cell 4(11): 1413-24.

Chen, F., D. K. Ro, J. Petri, J. Gershenzon, J. Bohlmann, E. Pichersky and D. Tholl
(2004). "Characterization of a root-specific Arabidopsis terpene synthase
responsible for the formation of the volatile monoterpene 1,8-cineole." Plan;
Physiol 135(4): 1956-66.

Cheong, Y. H., H. S. Chang, R. Gupta, X. Wang, T. Zhu and S. Luan (2002).
"Transcriptional profiling reveals novel interactions between wounding, pathogen,
abiotic stress, and hormonal responses in Arabidopsis." Plant Physiol 129(2): 661 -
77.

Chesis, P. L., D. E. Levin, M. T. Smith, L. Emster and B. N. Ames (1984). "Mutagenicity
of quinones: pathways of metabolic activation and detoxification." Proc Natl
Acad Sci U S A 81(6): 1696-700.

Chisholm, S. T., G. Coaker, B. Day and B. J. Staskawicz (2006). "Host-microbe
interactions: shaping the evolution of the plant immune response." Q11 124(4):
803-14.

Chisholm, S. T., D. Dahlbeck, N. Krishnamurthy, B. Day, K. Sjolander and B. J.
Staskawicz (2005). "Molecular characterization of proteolytic cleavage sites of

216

the Pseudomonas syringae efi‘ector Aerpt2." Proc Natl Acad Sci U S A 102(6):
2087-92.

Chiwocha, S. D. S., S. R. Abrams, S. J. Ambrose, A. J. Cutler, M. Loewen, A. R. S. Ross
and A. R. Kermode (2003). "A method for profiling classes of plant hormones
and their metabolites using liquid chromatography-electrospray ionization tandem
mass spectrometry: an analysis of hormone regulation of thermodormancy of
lettuce (Lactuca sativa L.) seeds." Plant Journal 35(3): 405-417.

Chroust, K., T. Jowett, M. F. Farid-Wajidi, J. Y. Huang, M. Ryskova, R. Wolf and I.
Holoubek (2001). "Activation or detoxification of mutagenic and carcinogenic

compounds in transgenic Drosophila expressing human glutathione S-
transferase." Mutat Res 498(1-2): 169-79.

Chung, H. S., A. J. Koo, X. Gao, S. Jayanty, B. Thines, A. D. Jones and G. A. Howe
(2008). "Regulation and function of Arabidopsis JASMONATE ZIM-domain
genes in response to wounding and herbivory." Plant Physiol 146(3): 952-64.

Coles, B. F. and F. F. Kadlubar (2003). "Detoxification of electrophilic compounds by
glutathione S-transferase catalysis: determinants of individual response to
chemical carcinogens and chemotherapeutic drugs?" Biofactors 17(1-4): 115-30.

Constabel, C. P., D. R. Bergey and C. A. Ryan (1995). "Systemin activates synthesis of
wound-inducible tomato leaf polyphenol oxidase via the octadecanoid defense
signaling pathway." Proc Natl Acad Sci U S A 92(2): 407-11.

Cook, D., S. Fowler, 0. Fiehn and M. F. Thomashow (2004). "A prominent role for the
CBF cold response pathway in configuring the low-temperature metabolome of
Arabidopsis." Proc Natl Acad Sci U S A 101(42): 15243-8.

Comwell, D. G., M. V. Williams, A. A. Wani, G. Wani, E. Shen and K. H. Jones (2002).
"Mutagenicity of tocopheryl quinones: evolutionary advantage of selective
accumulation of dietary alpha-tocopherol." Nutr Cancer 43(1): 111-8.

Cowan, M. (1999). "Plant Products as Antimicrobial Agents." Clinical microbiology
reviews 12: 564-582.

Cui, J ., A. K. Bahrami, E. G. Pringle, G. Hemandez-Guzman, C. L. Bender, N. E. Pierce

and F. M. Ausubel (2005). "Pseudomonas syringae manipulates systemic plant
defenses against pathogens and herbivores." Proc Natl Agag Sci U S A 102(5):
1791 -6.

D'Auria, J. C. and J. Gershenzon (2005). "The secondary metabolism of Arabidopsis
thaliana: growing like a weed." Curr Opin Plant Biol 8(3): 308-16.

217

Davoine, C., T. Douki, G. Iacazio, J. L. Montillet and C. Triantaphylides (2005).
"Conjugation of keto fatty acids to glutathione in plant tissues. Characterization
and quantification by HPLC-tandem mass spectrometry." Anal Chem 77(22):
7366-72.

Deem, A. K., R. L. Bultema and D. N. Crowell (2006). "Prenylcysteine methylesterase in
Arabidopsis thaliana." Gene 380(2): 159-66.

Delker, C., B. K. Zolman, O. Miersch and C. Wastemack (2007). "Jasmonate
biosynthesis in Arabidopsis thaliana requires peroxisomal beta-oxidation enzymes
- Additional proof by properties of pex6 and aim] ." Phytochemistfl 68(12): 1642-
1650.

Delong, J. M., R. K. Prange, D. M. Hodges, C. F. Fomey, M. C. Bishop and M. Quilliam
(2002). "Using a modified ferrous oxidation-xylenol orange (FOX) assay for

detection of lipid hydroperoxides in plant tissue." J. Agric. Food Chem. 50: 248-
254.

Denekamp, M. and S. C. Smeekens (2003). "Integration of wounding and osmotic stress

signals determines the expression of the AtMYB102 transcription factor gene."
Plant Physiol 132(3): 1415-23.

Dennis, C. and C. Surridge (2000). "Arabidopsis thaliana genome. Introduction." Nature
408(6814): 791.

Dewick, P. M. (1995). "The Biosynthesis of Shikimate Metabolites." Natural Product
Remus 12(6): 579-607.

Dixon, D. P., T. Hawkins, P. J. Hussey and R. Edwards (2009). "Enzyme activities and
subcellular localization of members of the Arabidopsis glutathione transferase
superfamily." Exp Bot 60(4): 1207-18.

Dixon, R. A. (2001). "Natural products and plant disease resistance." Nature 411(6839):
843-847.

Dixon, R. A. and N. L. Paiva (1995). "Stress-Induced Phenylpropanoid Metabolism."
Plant Cell 7(7): 1085-1097.

Do, C. T., B. Pollet, J. Thevenin, R. Sibout, D. Denoue, Y. Barriere, C. Lapierre and L.
Jouanin (2007). "Both caffeoyl Coenzyme A 3-O-methyltransferase 1 and caffeic
acid O-methyltransferase 1 are involved in redundant functions for lignin,
flavonoids and sinapoyl malate biosynthesis in Arabidopsis." Planta 226(5): 1117-
29.

218

Doares, S. H., T. Syrovets, E. W. Weiler and C. A. Ryan (1995). "Oligogalacturonides
and chitosan activate plant defensive genes through the octadecanoid pathway."
Proc Natl Acad Sci U S A 92(10): 4095-8.

Durrant, W. E. and X. Dong (2004). "Systemic acquired resistance." Annu Rev
Phflopathol 42: 185-209.

Engelberth, J ., E. A. Schmelz, H. T. Albom, Y. J. Cardoza, J. Huang and J. H. Tumlinson
(2003). "Simultaneous quantification of jasmonic acid and salicylic acid in plants
by vapor-phase extraction and gas chromatography-chemical ionization-mass

spectrometry." Analflical Biochemistg 312(2): 242-250.

Eriksson, L., H. Antti, J. Gottfries, E. Holmes, E. Johansson, F. Lindgren, 1. Long, T.
Lundstedt, J. Trygg and S. Wold (2004). "Using Chemometrics for navigating in

the large data sets of genomics, proteomics, and metabonomics (gpm)." Anal
Bioanal Chem 380(3): 419-29.

Fahey, J. W., A. T. Zalcmann and P. Talalay (2001). "The chemical diversity and
distribution of glucosinolates and isothiocyanates among plants." Lhytochemim
56(1): 5-51.

Fiehn, O. (2001). "Combining genomics, metabolome analysis, and biochemical
modelling to understand metabolic networks." Comp Funct Genomica 2(3): 155-
68.

Fiehn, O. (2002). "Metabolomicsuthe link between genotypes and phenotypes." Plant
Mol Biol 48(1-2): 155-71.

Fiehn, O. (2003). "Metabolic networks of Cucurbita maxima phloem." Phflochemisfl
62(6): 875-86.

Fiehn, O. (2006). "Metabolite profiling in Arabidopsis." Methods Mol Biol 323: 439-47.

Fiehn, O. (2008). "Extending the breadth of metabolite profiling by gas chromatography
coupled to mass spectrometry." Trends Analyt Chem 27(3): 261-269.

Fiehn, 0., J. Kopka, P. Dormann, T. Altrnann, R. N. Trethewey and L. Willmitzer (2000).
"Metabolite profiling for plant functional genomics." Nat Biotechnol 18(11):
1 157-61.

Fiehn, 0., G. Wohlgemuth, M. Scholz, T. Kind, Y. Lee do, Y. Lu, S. Moon and B.
Nikolau (2008). "Quality control for plant metabolomics: reporting MSI-
compliant studies." Plant J 53(4): 691-704.

Flor, A. H. (1955). "Host-parasite interactions in flax rust-its genetics and other
implications. ." Phytonathology 45: 680-85.

219

Fobert, P. R. and C. Despres (2005). "Redox control of systemic acquired resistance."
Curr Opin Plant Biol 8(4): 378-82.

Franklin, G., L. F. Conceicao, E. Kombrink and A. C. Dias (2009). "Xanthone
biosynthesis in Hypericum perforatum cells provides antioxidant and
antimicrobial protection upon biotic stress." Phylochemistry 70(1): 60-8.

F ulcrand, H., A. Cheminat, R. Brouillard and V. Cheynier (1994). "Charaterization of
compounds obtained by chemical Oxidation of caffeic acid in acidic conditions."
Phytochemistry 35(2): 499-505.

Galati, G., T. Chan, B. Wu and P. J. O'Brien (1999). "Glutathione-dependent generation
of reactive oxygen species by the peroxidase-catalyzed redox cycling of
flavonoids." Chem Res Toxicol 12(6): 521-5.

Gallet, C., Despr’es, L., Tollenaere, C. (2004). "Phenolic response of Trollius europaeus
to Chiastocheta invasion. ." Polyphenol Commun: 759—760.

Gauthier, A., P. J. Gulick and R. K. Ibrahim (1998). "Characterization of two cDNA
clones which encode O-methyltransferases for the methylation of both flavonoid
and phenylpropanoid compounds." Arch Biochem Biophys 351(2): 243 -9.

Glassbrook, N., C. Beecher and J. Ryals (2000). "Metabolic profiling on the right path."
Maire Biotechnology 18(11): 1142-1143.

Glauser, G., E. Grata, L. Dubugnon, S. Rudaz, E. E. Farmer and J. L. Wolfender (2008).
"Spatial and temporal dynamics of jasmonate synthesis and accumulation in
Arabidopsis in response to wounding." J Biol Chem 283(24): 16400-7.

Glauser, G., D. Guillarme, E. Grata, J. Boccard, A. Thiocone, P. A. Carrupt, J. L.
Veuthey, S. Rudaz and J. L. Wolfender (2008). "Optimized liquid
chromatography-mass spectrometry approach for the isolation of minor stress
biomarkers in plant extracts and their identification by capillary nuclear magnetic
resonance." Journal of Chromatography A 1180(1-2): 90-98.

Glauser, G., D. Guillarme, E. Grata, J. Boccard, A. Thiocone, P. A. Carrupt, J. L.
Veuthey, S. Rudaz and J. L. Wolfender (2008). "Optimized liquid
chromatography-mass spectrometry approach for the isolation of minor stress
biomarkers in plant extracts and their identification by capillary nuclear magnetic
resonance." J Chromatogr A 1180(1-2): 90-8.

Glazebrook, J. and F. M. Ausubel (1994). "Isolation of phytoalexin-deficient mutants of

Arabidopsis thaliana and characterization of their interactions with bacterial
pathogens." Proc Natl Acad Sci U S A 91(19): 8955-9.

220

Goujon, T., R. Sibout, B. Pollet, B. Maba, L. Nussaume, N. Bechtold, F. Lu, J. Ralph, I.
Mila, Y. Barriere, C. Lapierre and L. Jouanin (2003). "A new Arabidopsis

thaliana mutant deficient in the expression of O-methyltransferase impacts lignins
and sinapoyl esters." Plant Mol Biol 51(6): 973-89.

Grubb, C. D. and S. Abel (2006). "Glucosinolate metabolism and its control." Trends in
Plant Science 11(2): 89-100.

Gu, L., A. D. Jones and R. L. Last (2007). "LC-MS/MS assay for protein amino acids and
metabolically related compounds for large-scale screening of metabolic
phenotypes." Afllytical Chemisgy 79(21): 8067-8075.

Guan, L. M. and J. G. Scandalios (2000). "Hydrogen-peroxide-mediated catalase gene
expression in response to wounding." Free Radic Biol Med 28(8): 1182-90.

Guo, X. and H. U. Stotz (2007). "Defense against Sclerotinia sclerotiorum in Arabidopsis
is dependent on jasmonic acid, salicylic acid, and ethylene signaling." Mol Plant
Microbe Interaat 20(11): 1384-95.

Gupta, V., M. G. Willits and J. Glazebrook (2000). "Arabidopsis thaliana EDS4
contributes to salicylic acid (SA)-dependent expression of defense responses:

evidence for inhibition of jasmonic acid signaling by SA." Mol Pm Microbe
Interact 13(5): 503-11.

Gutjahr, C. and U. Paszkowski (2009). "Weights in the balance: jasmonic Acid and
salicylic Acid signaling in root-biotroph interactions." Mol Plant Microbe Intera_ct
22(7): 763-72.

Guy, C., J. Kopka and T. Moritz (2008). "Plant metabolomics coming of age." Physiol
Plant 132(2): 113-6.

Hagemeier, J ., B. Schneider, N. J. Oldham and K. Hahlbrock (2001). "Accumulation of
soluble and wall-bound indolic metabolites in Arabidopsis thaliana leaves
infected with Virulent or avirulent Pseudomonas syringae pathovar tomato
strains." Proceedings of the National Academy of Sciences of the United States of
America 98(2): 753-758.

Harnmerschmidt, R. (1999). "PHYTOALEXINS: What Have We Learned After 60
Years?" Annu Rev Phytopathol 37: 285-306.

Hanley, J. C. (2005). Liquid chromatography/mass spectrometric methodologies for
metabolomics, The Pennsylvania State University. Ph.D.: 282.

Harbome, J. 3., Ed. (1988). The fl_avonoid:Advances in Research since 1980. New York,
Chapman and Hall.

221

Harbome, J. B. (1994). Plant polyphenols aad their role in plant defence mechanism_s_.,

Heath, M. C. (2000). "Hypersensitive response-related death." Plant Mol Biol 44(3): 321-
34.

Hentze, H., F. Gantner, S. A. Kolb and A. Wendel (2000). "Depletion of hepatic
glutathione prevents death receptor-dependent apoptotic and necrotic liver injury
in mice." Am J Pathol 156(6): 2045-56.

Hirai, M. Y., M. Klein, Y. Fujikawa, M. Yano, D. B. Goodenowe, Y. Yamazaki, S.
Kanaya, Y. Nakarnura, M. Kitayama, H. Suzuki, N. Sakurai, D. Shibata, J.
Tokuhisa, M. Reichelt, J. Gershenzon, J. Papenbrock and K. Saito (2005).
"Elucidation of gene-to-gene and metabolite-to-gene networks in arabidopsis by
integration of metabolomics and transcriptomics." J Biol Chem 280(27): 25590-5.

Howe, G. A. and C. A. Ryan (1999). "Suppressors of systemin signaling identify genes in
the tomato wound response pathway." Genetics 153(3): 1411-21.

Hunt, M. D., U. H. Neuenschwander, T. P. Delaney, K. B. Weymann, L. B. Friedrich, K.
A. Lawton, H. Y. Steiner and J. A. Ryals (1996). "Recent advances in systemic
acquired resistance research--a review." Gene 179(1): 89-95.

Hyun, Y., S. Choi, H. J. Hwang, J. Yu, S. J. Nam, J. Ko, J. Y. Park, Y. S. Seo, E. Y. Kim,
S. B. Ryu, W. T. Kim, Y. H. Lee, H. Kang and 1. Lee (2008). "Cooperation and

functional diversification of two closely related galactolipase genes for jasmonate
biosynthesis." Dev Cell 14(2): 183-92.

Ibrahim, R. K., A. Bruneau and B. Bantignies (1998). "Plant O-methyltransferases:
molecular analysis, common signature and classification." Plant Mol Biol 36(1):
1-10.

Ishiguro, S., A. Kawai-Oda, J. Ueda, I. Nishida and K. Okada (2001). "The DEFECTIVE
IN ANTHER DEHISCIENCE gene encodes a novel phospholipase A1 catalyzing
the initial step of jasmonic acid biosynthesis, which synchronizes pollen
maturation, anther dehiscence, and flower opening in Arabidopsis." Plant Cell
13(10): 2191-209.

Jander, G., S. R. Norris, V. Joshi, M. Fraga, A. Rugg, S. Yu, L. Li and R. L. Last (2004).
"Application of a high-throughput HPLC-MS/MS assay to Arabidopsis mutant

screening; evidence that threonine aldolase plays a role in seed nutritional
quality." Plant J 39(3): 465-75.

Jander, G., S. R. Norris, S. D. Rounsley, D. F. Bush, I. M. Levin and R. L. Last (2002).

"Arabidopsis map-based cloning in the post-genome era. " Plant Physiol 129(2):
440-50.

222

Jimenez-Lopez, J. M., D. Wu and A. I. Cederbaum (2008). "Synergistic toxicity induced
by prolonged glutathione depletion and inhibition of nuclear factor-kappaB
signaling in liver cells." Toxicol In Vitro 22(1): 106-15.

Jones, J. D. and J. L. Dangl (2006). "The plant immune system." Nature 444(7117): 323-
9.

Joshi, C. P. and V. L. Chiang (1998). "Conserved sequence motifs in plant S-adenosyl-L-
methionine-dependent methyltransferases." Plant Mol Biol 37(4): 663-74.

Kachroo, A. and P. Kachroo (2007). "Salicylic acid-, jasmonic acid- and ethylene-
mediated regulation of plant defense signaling." Genet Eng1N Y) 28: 55-83.

Katagiri, F., R. Thilmony and S. Y. He (2002). "The Arabidopsis Thaliana -
Pseudomonas Syringae Interaction." The Arabidopsis Book.

Katsir, L., H. S. Chung, A. J. K00 and G. A. Howe (2008). "Jasmonate signaling: a
conserved mechanism of hormone sensing." Curr Opin Plant Biol 11(4): 428-35.

Keen, N. T. (1990). "Gene-for-gene complementarity in plant-pathogen interactions."
Annu Rev Genet 24: 447-63.

Kerhoas, L., D. Aouak, A. Cingoz, J. M. Routaboul, L. Lepiniec, J. Einhom and N.
Birlirakis (2006). "Structural characterization of the major flavonoid glycosides
from Arabidopsis thaliana seeds." Journal of Agg'cultural and Food Chemiag
54(18): 6603-6612.

Kim, J. K., T. Bamba, K. Harada, E. Fukusaki and A. Kobayashi (2007). "Time-course
metabolic profiling in Arabidopsis thaliana cell cultures afier salt stress
treatment." Journal of Experimental Botany 58(3): 415-424.

Kim, Y. S. and H. Sano (2008). "Pathogen resistance of transgenic tobacco plants
producing caffeine." Phytochemisg 69(4): (882-8.

Kliebenstein, D. J. (2004). "Secondary metabolites and plant/environment interactions: a
view through Arabidopsis thaliana tinged glasses." Plant Cell and E_nvironme_nt_
27(6): 675-684.

Knight, M. R., N. D. Read, A. K. Campbell and A. J. Trewavas (1993). "Imaging calcium
dynamics in living plants using semi-synthetic recombinant aequorins." J Cell
M 121(1): 83-90.

Koo, A. J., X. Gao, A. Daniel Jones and G. A. Howe (2009). "A rapid wound signal
activates the systemic synthesis of bioactive jasmonates in Arabidopsis." Plant J.

223

Koo, A. J. K., H. S. Chung, Y. Kobayashi and G. A. Howe (2006). "Identification of a
peroxisomal acyl-activating enzyme involved in the biosynthesis of jasmonic acid
in arabidopsis." Journal of Biological Chemist_ry 281(44): 33511-33520.

Kourtchenko, 0., M. X. Andersson, M. Hamberg, A. Brunnstrom, C. Gobel, K. L.
McPhail, W. H. Gerwick, I. Feussner and M. Ellerstrom (2007). "0x0-
phytodienoic acid-containing galactolipids in Arabidopsis: jasmonate signaling
dependence." Plant Physiol 145(4): 1658-69.

Kramell, R., O. Miersch, B. Hause, B. Ortel, B. Parthier and C. Wastemack (1997).
"Amino acid conjugates of jasmonic acid induce jasmonate-responsive gene
expression in barley (Hordeurn vulgare L.) leaves." Febs Letters 414(2): 197-202.

Kusnierczyk, A., P. Winge, T. S. Jorstad, J. Troczynska, J. T. Rossiter and A. M. Bones
(2008). "Towards global understanding of plant defence against aphids--timing
and dynamics of early Arabidopsis defence responses to cabbage aphid
(Brevicoryne brassicae) attack." Plant Cell Environ 31(8): 1097-115.

Lahner, B., J. Gong, M. Mahmoudian, E. L. Smith, K. B. Abid, E. E. Rogers, M. L.
Guerinot, J. F. Harper, J. M. Ward, L. McIntyre, J. I. Schroeder and D. E. Salt

(2003). "Genomic scale profiling of nutrient and trace elements in Arabidopsis
thaliana." Nat Biotechnol 21(10): 1215-21.

Landry, L. G., C. C. S. Chapple and R. L. Last (1995). "Arabidopsis Mutants Lacking
Phenolic Sunscreens Exhibit Enhanced Ultraviolet-B lnjury and Oxidative
Damage." Plant Physiology 109: 1159-1166.

Last, R. L., A. D. Jones and Y. Shachar-Hill (2007). "Towards the plant metabolome and
beyond." Nat Rev Mol Cell Biol 8(2): 167-74.

Laudert, D. and E. W. Weiler (1998). "Allene oxide synthase: a major control point in
Arabidopsis thaliana octadecanoid signalling." Plant J 15(5): 675-84.

Le Lay, P., M. P. Isaure, J. E. Sarry, L. Kuhn, B. Fayard, J. L. Le Bail, 0. Bastien, J.
Garin, C. Roby and J. Bourguignon (2006). "Metabolomic, proteomic and
biophysical analyses of Arabidopsis thaliana cells exposed to a caesium stress.
Influence of potassium supply." Biochimie 88(11): 1533-47.

Lebel, E., P. Heifetz, L. Thorne, S. Uknes, J. Ryals and E. Ward (1998). "Functional
analysis of regulatory sequences controlling PR-l gene expression in
Arabidopsis." Plant J 16(2): 223-33.

Lee, S., S. Suh, S. Kim, R. C. Crain, J. M. Kwak, H.-G. Nam and Y. Lee (1997).

"Systemic elevation of phosphatidic acid and lysophospholipid levels in wounded
plants." Plant Journal 12: 547-556.

224

Leon, J., E. Rojo, E. Titarenko and J. J. Sanchez-Serrano (1998). "Jasmonic acid-
dependent and -independent wound signal transduction pathways are
differentially regulated by Ca2+/calmodulin in Arabidopsis thaliana." Mol Gen
Genet 258(4): 412-9.

Lilienblum, W., B. S. Bock-Hennig and K. W. Bock (1985). "Protection against toxic
redox cycles between benzo(a)pyrene-3,6-quinone and its quinol by 3-

methylcholanthrene-inducible formation of the quinol mono- and diglucuronide."
Mol Pharrnacol 27(4): 45 1-8.

Lim, E. K., Y. Li, A. Parr, R. Jackson, D. A. Ashford and D. J. Bowles (2001).
"Identification of glucosyltransferase genes involved in sinapate metabolism and
lignin synthesis in Arabidopsis." Journal of Biological Chemistry 276(6): 4344-
4349.

Lorenzen, M., V. Racicot, D. Strack and C. Chapple (1996). "Sinapic acid ester

metabolism in wild type and a sinapoylglucose-accmnulating mutant of
Arabidopsis." fl_ant Physiology 112(4): 1625-1630.

Lu, Y., L. J. Savage, 1. Ajjawi, K. M. Imre, D. W. Yoder, C. Benning, D. Dellapenna, J.
B. Ohlrogge, K. W. Osteryoung, A. P. Weber, C. G. Wilkerson and R. L. Last
(2008). "New connections across pathways and cellular processes: industrialized
mutant screening reveals novel associations between diverse phenotypes in

Arabidopsis." Plant Physiol 146(4): 1482-500.

Marrs, K. A. (1996). "The Functions and Regulation of Glutathione S-Transferases in
Plants." Annu Rev Plant Physiol Plant M01 Biol 47: 127-158.

Martin, C., K. Bhatt and K. Baumann (2001). "Shaping in plant cells." Curr Qpin Plant
Big] 4(6): 540-9.

Mattsson, J. L. (2000). "Do pesticides reduce our total exposure to food borne toxicants?"
Neurotoxicology 21(1-2): 195-202.

McCallum, C. M., L. Comai, E. A. Greene and S. Henikoff (2000). "Targeted screening
for induced mutations." Nat Biotechnol 18(4): 455-7.

Meinke, D. W., J. M. Cherry, C. Dean, S. D. Rounsley and M. Koomneef (1998).
"Arabidopsis thaliana: a model plant for genome analysis." Science 282(5389):
662, 679-82.

 

Melotto, M., W. Underwood, J. Koczan, K. Nomura and S. Y. He (2006). "Plant stomata
function in innate immunity against bacterial invasion." Call, 126(5): 969-80.

Messerli, G., V. Partovi Nia, M. Trevisan, A. Kolbe, N. Schauer, P. Geigenberger, J.
Chen, A. C. Davison, A. R. Fernie and S. C. Zeeman (2007). "Rapid classification .

225

of phenotypic mutants of Arabidopsis via metabolite fingerprinting." Plant
Physiol 143(4): 1484-92.

Metzler, M. C., J. R. Cutt and D. F. Klessig (1991). "Isolation and Characterization of a
Gene Encoding a PR-l-Like Protein from Arabidopsis thaliana." Plant Physiol
96(1): 346-348.

Mewis, 1., J. G. Tokuhisa, J. C. Schultz, H. M. Appel, C. Ulrichs and J. Gershenzon
(2006). "Gene expression and glucosinolate accumulation in Arabidopsis thaliana

in response to generalist and specialist herbivores of different feeding guilds and
the role of defense signaling pathways." Phytochemistly 67(22): 2450-2462.

Mewis, I., J. G. Tokuhisa, J. C. Schultz, H. M. Appel, C. Ulrichs and J. Gershenzon
(2006). "Gene expression and glucosinolate accumulation in Arabidopsis thaliana

in response to generalist and specialist herbivores of different feeding guilds and
the role of defense signaling pathways." Phytochemistry 67(22): 2450-62.

Mittler, R. (2002). "Oxidative stress, antioxidants and stress tolerance." Trends Plant Sci
7(9): 405-10.

Mittler, R. (2006). "Abiotic stress, the field environment and stress combination." Trends
Plant Sci 11(1): 15-9.

Mittler, R., S. Vanderauwera, M. Gollery and F. Van Breusegem (2004). "Reactive
oxygen gene network of plants." Trends Plant Sci 9(10): 490-8.

Mizoguchi, T., K. Ichirnura and K. Shinozaki (1997). "Environmental stress response in
plants: the role of mitogen-activated protein kinases." Trenda Biotechnol 15(1):
15-9.

Mock, H. P., T. Vogt and D. Strack (1992). "Sinapoylglucose - Malate
Sinapoyltransferase Activity in Arabidopsis-Thaliana and Brassica-Rapa."

Zeitschrift Fur Naturforschung C-a Journal of Biosciences 47(9-10): 680-682.

Moco, S., B. Schneider and J. Vervoort (2009). "Plant Micrometabolomics: The Analysis
of Endogenous Metabolites Present in a Plant Cell or Tissue." J Proteome Res.

Monks, T. J. and S. S. Lau (1992). "Toxicology of quinone-thioethers." Crit Rev Toxicol
22(5-6): 243-70.

Monks, T. J. and S. S. Lau (1998). "The pharmacology and toxicology of polyphenolic-
glutathione conjugates." Annu Rev Pharmacol Toxicol 38: 229-55.

Montesinos, E., A. Bonaterra, E. Badosa, J. Frances, J. Alemany, I. Llorente and C.

Moragrega (2002). "Plant-microbe interactions and the new biotechnological
methods of plant disease control." Int Microbiol 5(4): 169-75.

226

Mueller, L. A., C. D. Goodman, R. A. Silady and V. Walbot (2000). "AN9, a petunia
glutathione S-transferase required for anthocyanin sequestration, is a flavonoid-
binding protein." Plant Physiol 123(4): 1561-70.

Murata, M., M. Sugiura, Y. Sonokawa, T. Shimamura and S. Homma (2002). "Properties
of chlorogenic acid quinone: relationship between browning and the formation of

hydrogen peroxide fiom a quinone solution." Biosci Biotechnol Biochem 66(12):
2525-30.

Muzac, 1., J. Wang, D. Auzellotti, H. Zhang and R. K. Ibrahim (2000). "Functional
expression of an Arabidopsis cDNA clone encoding a flavonol 3 '-O-
methyltransferase and characterization of the gene product." Archives of
Biochemistry and Biophysics 375(2): 385-388.

Nair, R. B., K. L. Bastress, M. O. Ruegger, J. W. Denault and C. Chapple (2004). "The
Arabidopsis thaliana REDUCED EPIDERMAL FLUORESCENCE] gene

encodes an aldehyde dehydrogenase involved in ferulic acid and sinapic acid ,
biosynthesis." Plant Cell 16(2): 544-554.

Nawrath, C., S. Heck, N. Parinthawong and J. P. Metraux (2002). "EDSS, an essential
component of salicylic acid-dependent signaling for disease resistance in
Arabidopsis, is a member of the MATE transporter family." Plant Cell 14(1): 275-
86.

Neuefeind, T., P. Reinemer and B. Bieseler (1997). "Plant glutathione S-transferases and
herbicide detoxification." Biol Chem 378(3—4): 199-205.

O'Donnell, P. J ., C. Calvert, R. Atzom, C. Wastemack, H. M. O. Leyser and D. J. Bowles
(1996). "Ethylene as a Signal Mediating the Wound Response of Tomato Plants."
Science 274(5294): 1914-7.

 

Orozco-Cardenas, M. and C. A. Ryan (1999). "Hydrogen peroxide is generated
systemically in plant leaves by wounding and systemin via the octadecanoid
pathway." Proc Natl Agd Sci U S A 96(11): 6553-7.

Orozco-Cardenas, M. L., J. Narvaez-Vasquez and C. A. Ryan (2001). "Hydrogen
peroxide acts as a second messenger for the induction of defense genes in tomato

plants in response to wounding, systemin, and methyl jasmonate." Plant Cell
13(1): 179-91.

Orozco-Cardenas, M. L. and C. A. Ryan (2002). "Nitric oxide negatively modulates
wound signaling in tomato plants." Plant Physiol 130(1): 487-93.

Parisy, V., B. Poinssot, L. Owsianowski, A. Buchala, J. Glazebrook and F. Mauch
(2007). "Identification of PAD2 as a garnma-glutarnylcysteine synthetase

227

highlights the importance of glutathione in disease resistance of Arabidopsis."
Plant J 49(1): 159-72.

Pastore, A., G. Federici, E. Bertini and F. Piemonte (2003). "Analysis of glutathione:
implication in redox and detoxification." Clin Chirn Acta 333(1): 19-39.

Patrineli, A., M. N. Clifford and C. Ioannides (1996). "Contribution of phenols, quinones
and reactive oxygen species to the mutagenicity of white grape juice in the Ames
test." Food Chem Toxicol 34(9): 869-72.

Poethig, R. S. (2001). "Life with 25,000 genes." Genome Res 11(3): 313-6.

Pontier, D., C. Albrieux, J. Joyard, T. Lagrange and M. A. Block (2007). "Knock-out of
the magnesium protoporphyrin IX methyltransferase gene in Arabidopsis. Effects

on chloroplast development and on chloroplast-to-nucleus signaling." J Biol
Chem 282(4): 2297-304.

Raes, J ., A. Rohde, J. H. Christensen, Y. Van de Peer and W. Boerjan (2003). "Genome-
wide characterization of the lignification toolbox in Arabidopsis." Plant
Physiology 133(3): 1051-1071.

Reymond, P., H. Weber, M. Damond and E. E. Farmer (2000). "Differential gene
expression in response to mechanical wounding and insect feeding in
Arabidopsis." Plant Cell 12(5): 707-20.

Roessner, U. (2007). Uncovering the plant metabolome: current and future challenges.
Concepts in plant metabolomics

Roessner, U., A. Luedemann, D. Brust, O. Fiehn, T. Linke, L. Willmitzer and A. R.
Fernie (2001). "Metabolic profiling allows comprehensive phenotyping of
genetically or environmentally modified plant systems." Plant Cell 13(1): 11-29.

Roessner, U., C. Wagner, J. Kopka, R. N. Trethewey and L. Willmitzer (2000).
"Technical advance: simultaneous analysis of metabolites in potato tuber by gas

chromatography-mass spectrometry." Plant J 23(1): 131-42.

Rogers, E. E. and F. M. Ausubel (1997). "Arabidopsis enhanced disease susceptibility
mutants exhibit enhanced susceptibility to several bacterial pathogens and
alterations in PR-l gene expression." Plant Cell 9(3): 305-16.

Rogers, L. A., C. Dubos, C. Sunnan, J. Willment, I. F. Cullis, S. D. Mansfield and M. M.

Campbell (2005). "Comparison of lignin deposition in three ectopic lignification
mutants." New Phflologist 168(1): 123-140.

228

Ruegger, M. and C. Chapple (2001). "Mutations that reduce sinapoylmalate accumulation
in Arabidopsis thaliana define loci with diverse roles in phenylpropanoid
metabolism." Genetics 159(4): 1741-9.

 

Ruegger, M., K. Meyer, J. C. Cusumano and C. Chapple (1999). "Regulation of ferulate-
5-hydroxylase expression in Arabidopsis in the context of sinapate ester
biosynthesis." Plant Physiol 119(1): 101-10.

Rusterucci, C., V. Stallaert, M. L. Milat, A. Pugin, P. Ricci and J. P. Blein (1996).
"Relationship between Active Oxygen Species, Lipid Peroxidation, Necrosis, and

Phytoalexin Production Induced by Elicitins in Nicotiana." Plant Physiol 111(3):
885-891.

Ryan, C. A. (2000). "The systemin signaling pathway: differential activation of plant
defensive genes." Biochim Biophys Acta 1477(1-2): 112-21.

Ryan, C. A. (2000). "The systemin signaling pathway: differential activation of plant
defensive genes." Biochimica Et Biophysica Acta-Protein Structure and
Molecular Enzymology 1477(1-2): 112-121.

Ryu, S. B. and X. Wang (1998). "Increase in free linolenic and linoleic acids associated
with phospholipase D-mediated hydrolysis of phospholipids in wounded castor
bean leaves." Biochim Biophys Acta 1393(1): 193-202.

Salyan, M. E., D. L. Pedicord, L. Bergeron, G. A. Mintier, L. Hunihan, K. Kuit, L. A.
Balanda, B. J. Robertson, J. N. Feder, R. Westphal, P. A. Shipkova and Y. Blat
(2006). "A general liquid chromatography/mass spectroscopy-based assay for
detection and quantitation of methyltransferase activity." Anal Biochem 349(1):
112-7.

Schauer, N. and A. R. Fernie (2006). "Plant metabolomics: towards biological function
and mechanism." Trends Plant Sci 11(10): 508-16.

Schilmiller, A. L. and G. A. Howe (2005). "Systemic signaling in the wound response."
C_urr Opin Plant Biol 8(4): 369-77.

Schmelz, E. A., J. Engelberth, J. H. Tumlinson, A. Block and H. T. Albom (2004). "The
use of vapor phase extraction in metabolic profiling of phytohormones and other
metabolites." Plant Journal 39(5): 790-808.

Scholthof, H. B. (2001). "Molecular plant-microbe interactions that cut the mustard."
P_lant Physiolcgy 127(4): 1476-1483.

Serra, R., A. Braga and A. Venancio (2005). "Mycotoxin-producing and other fungi
isolated from grapes for wine production, with particular emphasis on ochratoxin
A." Res Microbiol 156(4): 5 15-21 .

229

 

Shulaev, V., D. Cortes, G. Miller and R. Mittler (2008). "Metabolomics for plant stress
response." Physiol Plant 132(2): 199-208.

Srivastava, L. M. (2002). Plant Growth and Development: Hormones and Environment
Amsterdam: Academic Press.

Staswick, P. E. and I. Tiryaki (2004). "The oxylipin signal jasmonic acid is activated by
an enzyme that conjugates it to isoleucine in Arabidopsis." Plant Cell 16(8): 2117-
27.

Stephensen, E., J. Sturve and L. Forlin (2002). "Effects of redox cycling compounds on
glutathione content and activity of glutathione-related enzymes in rainbow trout
liver." Comp Biochem Physiol C Toxicol Pharmacol 133(3): 435-42.

Stobiecki, M., A. Skirycz, L. Kerhoas, P. Kachlicki, D. Muth, J. Einhom and B. Mueller-
Roeber (2006). "Profiling of phenolic glycosidic conjugates in leaves of
Arabidopsis thaliana using LC/MS." Metabolomicg 2(4): 197-219.

Sumner, L. W., P. Mendes and R. A. Dixon (2003). "Plant metabolomics: large-scale
phytochemistry in the fimctional genomics era." Phytochemistry 62(6): 817-36.

Sumner, L. W., P. Mendes and R. A. Dixon (2003). "Plant metabolomics: large-scale
phytochemistry in the functional genomics era. " Phflochemisgy 62(6): 817-836.

TAIR. (2007). " The Arabidopsis Information Resource. http://www.arabidopsis.org/."

Tamogami, S. and O. Kodama (1998). "Quantification of amino acid conjugates of
jasmonic acid in rice leaves by high-performance liquid chromatography
turboionspray tandem mass spectrometry." Journal of Chromatoggphy A 822(2):
310-315.

Thihnony, R., W. Underwood and S. Y. He (2006). "Genome-wide transcriptional
analysis of the Arabidopsis thaliana interaction with the plant pathogen

Pseudomonas syringae pv. tomato DC3000 and the human pathogen Escherichia
coli 0157:H7." Plant J 46(1): 34-53.

Thines, B., L. Katsir, M. Melotto, Y. Niu, A. Mandaokar, G. H. Liu, K. Nomura, S. Y.
He, G. A. Howe and J. Browse (2007). "JAZ repressor proteins are targets of the
SCFCOll complex during jasmonate signalling." Nature 448(7154): 661-U2.

Thoma, 1., C. Loeffler, A. K. Sinha, M. Gupta, M. Krischke, B. Steffan, T. Roitsch and
M. J. Mueller (2003). "Cyclopentenone isoprostanes induced by reactive oxygen

species trigger defense gene activation and phytoalexin accumulation in plants."
Plant J 34(3): 363-75.

230

Tierens, K. F ., B. P. Thomma, M. Brouwer, J. Schmidt, K. Kistner, A. Porzel, B. Mauch-
Mani, B. P. Cammue and W. F. Broekaert (2001). "Study of the role of
antimicrobial glucosinolate-derived isothiocyanates in resistance of Arabidopsis
to microbial pathogens." Plant Physiol 125(4): 1688-99.

Titarenko, E., E. Rojo, J. Leon and J. J. Sanchez-Serrano (1997). "Jasmonic acid-
dependent and -independent signaling pathways control wound-induced gene
activation in Arabidopsis thaliana." Plant Physiol 115(2): 817-26.

Tohge, T., Y. Nishiyama, M. Y. Hirai, M. Yano, J. Nakajima, M. Awazuhara, E. Inoue,
H. Takahashi, D. B. Goodenowe, M. Kitayama, M. Noji, M. Yamazaki and K.
Saito (2005). "Functional genomics by integrated analysis of metabolome and
transcriptome of Arabidopsis plants over-expressing an MYB transcription

factor." Plant J 42(2): 218-35.

Tohge, T., K. Yonekura-Sakakibara, R. Niida, A. Watanabe-Rakahashi and K. Saito
(2007). "Phytochemial genomics in Arabidopsis thaliana: A case study for
functional identification of flavonoid biosynthesis genes." Pure and Applied
Chemest_ry. 79(4): 81 1-823.

Traw, M. B., J. M. Kniskem and J. Bergelson (2007). "SAR increases fitness of
Arabidopsis thaliana in the presence of natural bacterial pathogens." Evolution
61(10): 2444-9.

Treutter, D. (2005). "Significance of flavonoids in plant resistance and enhancement of
their biosynthesis." Plant Biol (Stuttg) 7(6): 581-91.

Treutter, D. (2006). "Significance of flavonoids in plant resistance: a review."
Momental Chemistry Letters 4(3): 147-157.

Truman, W., M. T. de Zabala and M. Grant (2006). "Type III effectors orchestrate a
complex interplay between transcriptional networks to modify basal defence
responses during pathogenesis and resistance." Plant J 46(1): 14-33.

Trygg, J., E. Holmes and T. Lundstedt (2007). "Chemometrics in metabonomics." J
Proteome Res 6(2): 469-79.

van Verk, M. C., D. Pappaioannou, L. Neeleman, J. F. B01 and H. J. Linthorst (2008). "A
Novel WRKY transcription factor is required for induction of PR-la gene

expression by salicylic acid and bacterial elicitors." Plant Physiol 146(4): 1983-
95.

von Roepenack-Lahaye, E., T. Degenkolb, M. Zerjeski, M. Franz, U. Roth, L.

Wessjohann, J. Schmidt, D. Scheel and S. Clemens (2004). "Profiling of
Arabidopsis secondary metabolites by capillary liquid chromatography coupled to

231

 

electrospray ionization quadrupole time-of-flight mass spectrometry." Plant
Physiol 134(2): 548-59.

Wang, C. X., C. A. Zien, M. Afitlhile, R. Welti, D. F. Hildebrand and X. M. Wang
(2000). "Involvement of phospholipase D in wound-induced accumulation of
jasmonic acid in Arabidopsis." Plant Cell 12(11): 2237-2246.

Wang, J. and E. Pichersky (1999). "Identification of specific residues involved in
substrate discrimination in two plant O-methyltransferases." Arch Biochem
Biophys 368(1): 172-80.

Wang, L., R. Halitschke, J. H. Kang, A. Berg, F. Harnisch and I. T. Baldwin (2007).
"Independently silencing two JAR family members impairs levels of trypsin
proteinase inhibitors but not nicotine." Planta 226(1): 159-67.

Ward, J. L., J. M. Baker and M. H. Beale (2007). "Recent applications of NMR
spectroscopy in plant metabolomics." Febs J 274(5): 1126-31.

Wastemack, C. (2007). "Jasmonates: an update on biosynthesis, signal transduction and
action in plant stress response, growth and development." Ann Bot (Lond) 100(4):
681-97.

Weber, H., B. A. Vick and E. E. Farmer (1997). "Dinor-oxo-phytodienoic acid: a new
hexadecanoid signal in the jasmonate family." Proc Natl Acad Sci U S A 94(19):
10473-8.

Weckwerth, W. (2008). "Integration of metabolomics and proteomics in molecular plant
physiology--coping with the complexity by data-dimensionality reduction."
Physiol Plant 132(2): 176-89.

Weigel, R. R., U. M. Pfitzner and C. Gatz (2005). "Interaction of NIMINl with NPRl
modulates PR gene expression in Arabidopsis." Plant Cell 17(4): 1279-91.

Weiler, E. W., T. M. Kutchan, T. Gorba, W. Brodschelm, U. Niesel and F. Bublitz
(1994). "The Pseudomonas phytotoxin coronatine mimics octadecanoid signalling
molecules of higher plants." FEBS Lett 345(1): 9-13.

Weisel, T., M. Baum, G. Eisenbrand, H. Dietrich, F. Will, J. P. Stockis, S. Kulling, C.
Rufer, C. Johannes and C. Janzowski (2006). "An anthocyanin/polyphenolic-rich

fruit juice reduces oxidative DNA damage and increases glutathione level in
healthy probands." Biotechnol J 1(4): 388-97.

Wilbert, S. M., L. H. Ericsson and M. P. Gordon (1998). "Quantification of jasmonic
acid, methyl jasmonate, and salicylic acid in plants by capillary liquid
chromatography electrospray tandem mass spectrometry." Analflical
Biochemistry 257(2): 186-194.

232

Wilson, 1., P. Wardrnan, G. M. Cohen and M. D. Doherty (1986). "Reductive role of
glutathione in the redox cycling of oxidizable drugs." Biochem Pharmacol 35(1):
21-2.

Wingate, V. P., M. A. Lawton and C. J. Lamb (1988). "Glutathione Causes a Massive
and Selective Induction of Plant Defense Genes." Plant Physiol 87(1): 206-210.

Winkel, B. S. J. (2004). "Metabolic channeling in plants." Annual Review of Pl_a_n_t
Biology 55: 85-107.

Zhang, H., J. Wang and H. M. Goodman (1997). "An Arabidopsis gene encoding a
putative 14-3-3-interacting protein, caffeic acid/S-hydroxyferulic acid 0-

methyltransferase." Biochimica Et Biophysica Acta-Gene Structure and
Expression 1353(3): 199-202.

Zhang, H., J. Wang and H. M. Goodman (1997). "An Arabidopsis gene encoding a
putative 14-3-3-interacting protein, caffeic acid/S-hydroxyferulic acid 0-
methyltransferase." Biochim Biophys Acta 1353(3): 199-202.

Zhang, Y., W. Fan, M. Kinkema, X. Li and X. Dong (1999). "Interaction of NPR1 with
basic leucine zipper protein transcription factors that bind sequences required for
salicylic acid induction of the PR-l gene." Proc Natl Acad Sci U S A 96(11):
6523-8.

Zhao, J ., K. Fujita and K. Sakai (2007). "Reactive oxygen species, nitric oxide, and their

interactions play different roles in Cupressus lusitanica cell death and phytoalexin
biosynthesis." New Phytol 175(2): 21 5-29.

Zook, M. (1998). "Biosynthesis of camalexin fi'om tryptophan pathway intermediates in
cell-suspension cultures of Arabidopsis." Plant Physiol 118(4): 1389-93.

Zubieta, C., X. Z. He, R. A. Dixon and J. P. Noel (2001). "Structures of two natural
product methyltransferases reveal the basis for substrate specificity in plant 0-
methyltransferases." Nature Structural Biology 8(3): 271-279.

Zubieta, C., P. Kota, J. L. Ferrer, R. A. Dixon and J. P. Noel (2002). "Structural basis for
the modulation of lignin monomer methylation by caffeic acid/S-hydroxyferulic
acid 3/5-O-methyltransferase." Plant Cell 14(6): 1265-77.

233

lllllllllllllllll

l

3062 8592

lllllllll

0
3
9
2
1
3