Molecular Responses

BIOTECHNOLOGY I N T E L L I G E N C E U N I T 1
Kazuo Shinozaki and

Kazuko Yamaguchi-Shinozaki
Molecular Responses to
Cold, Drought, Heat
and Salt Stress in Higher Plants
R.G. LANDES
C OM PA N Y
BIOTECHNOLOGY
INTELLIGENCE
UNIT 1
Molecular Responses
to Cold, Drought, Heat
and Salt Stress in Higher Plants
Kazuo Shinozaki, Ph.D.

Laboratory of Plant Molecular Biology
Tsukuba Life Science Center
The Institute of Physical and Chemical Research (RIKEN)
Tsukuba, Japan
Kazuko Yamaguchi-Shinozaki, Ph.D.

Biological Resources Division
Japan International Research Center
for Agricultural Sciences (JIRCAS)
Tsukuba, Japan
R.G. LANDES
COMPANY
AUSTIN, TEXAS
U.S.A.
BIOTECHNOLOGY INTELLIGENCE UNIT
Cold, Drought, Heat and Salt Stress in Higher Plants
R.G. LANDES COMPANY

Austin, Texas, U.S.A.
Copyright ©1999 R.G. Landes Company
All rights reserved.
No part of this book may be reproduced or transmitted in any form or by any means,
electronic or mechanical, including photocopy, recording, or any information storage and
retrieval system, without permission in writing from the publisher.
Printed in the U.S.A.
Please address all inquiries to the Publishers:
R.G. Landes Company, 810 South Church Street, Georgetown, Texas, U.S.A. 78626
Phone: 512/ 863 7762; FAX: 512/ 863 0081
ISBN: 1-57059-563-1
While the authors, editors and publisher believe that drug selection and dosage and the
specifications and usage of equipment and devices, as set forth in this book, are in accord with
current recommendations and practice at the time of publication, they make no warranty,
expressed or implied, with respect to material described in this book. In view of the ongoing
research, equipment development, changes in governmental regulations and the rapid accumulation of
information relating to the biomedical sciences, the reader is urged to carefully review and evaluate
the information provided herein.
Library of Congress Cataloging-in-Publication Data

Cold, drought, heat and salt stress in higher plants / [edited by] Kazuo Shinozaki,
Kazuko Yamaguchi-Shinozaki.
p. cm. -- (Biotechnology intelligence unit)
ISBN 1-57059-563-1 (alk. paper)
1. Plants, Effects of stress on—Molecular aspects. 2. Plant molecular genetics.
I. Shinozaki, Kazuo. II. Yamaguchi-Shinozaki, Kazuko. III. Series.
QK754.C65 1999
571.9'52—dc21 99-33224
CIP
CONTENTS
1. Genetic Approaches to Abiotic Stress Responses .................................... 1
M. Koornneef and A.J.M. Peeters
The Genetic Approach in Stress Physiology—General Principals ........ 2
Genetic Differences in the Response to Low Temperatures .................. 4
ABA Related Mutants .............................................................................. 5
Conclusions ............................................................................................. 7
2. Molecular Responses to Drought Stress ................................................ 11

Kazuo Shinozaki and Kazuko Yamaguchi-Shinozaki
A Variety of Functions of Drought-Inducible Genes .......................... 12
Regulation of Gene Expression by Drought ........................................ 14
Signal Perception and Signal Transduction in Drought
Stress Response .................................................................................. 18
Conclusion and Perspectives ................................................................ 25
3. Molecular Mechanisms of Salinity Tolerance ....................................... 29
Hans J. Bohnert, Hua Su and Bo Shen
Osmolytes, Osmoprotectants, Compatible Solutes, Osmotic
Adjustment ........................................................................................ 30
Cellular Mechanisms of Salt Tolerance—the Fungal Model .............. 31
Molecular Mechanisms of Salt Tolerance in Plants ............................. 37
Metabolic Engineering of Glycophytic Plants for Increased
Salt Tolerance .................................................................................... 47
Perspectives ............................................................................................ 48
4. Plant Cold Tolerance ............................................................................... 63

Michael F. Thomashow and John Browse
Chilling tolerance .................................................................................. 63
Freezing Tolerance ................................................................................ 69
Conclusions and Perspectives ............................................................... 77
5. Molecular Responses to Heat Stress ....................................................... 83
Fritz Schöffl, Ralf Prändl and Andreas Reindl
Heat Shock Proteins and Thermotolerance ......................................... 84
Links to Other Abiotic Stresses ............................................................. 87
Transcriptional Regulation ................................................................... 89
The Regulation of HSF .......................................................................... 90
Conclusions and Perspectives ............................................................... 93
6. Cellular Responses to Water Stress ...................................................... 101

Michael R. Blatt, Barbara Leyman and Alexander Grabov
The Stomatal Situation ........................................................................ 102
Transport Mechanics ........................................................................... 103
Transport Coordination in the Face of Stress .................................... 105
Interaction of Signaling Elements ....................................................... 114
Initial Events in ABA Stimulus Perception ........................................ 115
Perspectives and Conclusion .............................................................. 117
Acknowledgments ............................................................................... 118
7. Role of Glycine Betaine and Dimethylsulfoniopropionate
in Water-Stress Tolerance ..................................................................... 127
Douglas A. Gage and Bala Rathinasabapathi
Stress Protection by Glycine Betaine and DMSP In Vivo
and In Vitro ..................................................................................... 128
Biosynthesis of DMSP ......................................................................... 134
Conclusion ........................................................................................... 147
8. Osmotic Stress Tolerance in Plants: Role of Proline and Sulfur
Metabolisms ........................................................................................... 155
Desh Pal S. Verma
Osmoregulation in Microorganisms .................................................. 155
Osmosensing and Signal Transduction Machinery ........................... 156
Osmotic Stress Tolerance in Plants .................................................... 158
Accumulation of Other Osmolytes ..................................................... 160
Accumulation of Proline in Transgenic Plants Expressing Elevated
Levels of P5CS ................................................................................. 161
Proline Accumulation Confers Osmoprotection............................... 163
Role of Sulfur metabolism in Osmotic stress Tolerance ................... 164
A Possible Role of DPNPase in Salt Tolerance .................................. 166
Overexpression of Plant HAL2 Gene Confers Reduction in Free Radical
Production and in Heavy Metal Toxicity ....................................... 166
EDITORS
Kazuo Shinozaki, Ph.D.
Laboratory of Plant Molecular Biology
Tsukuba Life Science Center
The Institute of Physical and Chemical Research (RIKEN)
Tsukuba, Japan
Chapter 2
Kazuko Yamaguchi-Shinozaki, Ph.D.
Biological Resources Division
Japan International Research Center
for Agricultural Sciences (JIRCAS)
Tsukuba, Japan
Chapter 2
CONTRIBUTORS
Michael R. Blatt, Ph.D. Alexander Grabov, Ph.D.
Laboratory of Plant Physiology and Laboratory of Plant Physiology and
Biophysics Biophysics
University of London, Wye College University of London, Wye College
Wye, England, U.K. Wye, England, U.K.
Chapter 6 Chapter 6
Bo Shen, Ph.D. Hua Su, Ph.D.

Departments of Plant Sciences Departments of Plant Sciences
The University of Arizona The University of Arizona
Tucson, Arizona, U.S.A. Tucson, Arizona, U.S.A.
Chapter 3 Chapter 3
Hans J. Bohnert, Ph.D. M. Koornneef, Ph.D.

Departments of Biochemistry, Laboratory of Genetics
Molecular and Cellular Sciences Wageningen Agricultural University
The University of Arizona Wageningen, The Netherlands
Tucson, Arizona, U.S.A. Chapter 1
Chapter 3
Barbara Leyman, Ph.D.
John Browse, Ph.D. Laboratory of Plant Physiology and
Institute of Biological Chemistry Biophysics
Washington State University University of London, Wye College
Pullman, Washington, U.S.A. Wye, England, U.K.
Chapter 4 Chapter 6
Douglas A. Gage, Ph.D. A.J.M. Peeters, Ph.D.

Department of Biology Laboratory of Genetics
Michigan State University Wageningen Agricultural University
East Lansing, Michigan, U.S.A. Wageningen, The Netherlands
Chapter 7 Chapter 1
Ralf Prändl, Ph.D.
Lehrstuhl Allgemeine Genetik
Universität Tubingen
Tubingen, Germany
Chapter 5
Bala Rathinasabapathi, Ph.D.

Hort Sciences Department
University of Florida
Gainesville, Florida, U.S.A.
Chapter 7
Andreas Reindl, Ph.D.

Tubingen, Germany
Chapter 5
Fritz Schöffl, Ph.D.

Tubingen, Germany
Chapter 5
Michael F. Thomashow, Ph.D.

Department of Crop and Soil Sciences,
Department of Microbiology
Michigan State University
East Lansing, Michigan, U.S.A.
Chapter 4
Desh Pal S. Verma, Ph.D.

Department of Molecular Genetics and
Plant Biotechnology Center
Ohio State University
Columbus, Ohio, U.S.A.
Chapter 8
PREFACE
T he genetic improvement of tolerance of crops to environmental stresses, such
as drought, high salinity, low temperature and heat, is an important
problem for the future of agriculture. Classical breeding methodologies to
select stress tolerant cultivars have already made some progress. Biotechnology
has the potential to improve environmental stress tolerance of crops using
transgenic plant technology. The limiting factor for developing this technology
is the isolation of genes that can improve drought tolerance and the precise
understandings of molecular process of stress tolerance and plants’ responses
to environmental stresses. Plants respond to environmental stresses, such as
drought, high salinity, low temperature and heat, through a number of
physiological and developmental changes. Recently, higher plants respond to
these stresses at gene expression level. A variety of stress-inducible genes have
been cloned and analyzed concerning to their expression and function in stress
tolerance and stress responses. Recently, many mutants have been isolated that
are resistant or hypersensitive to environmental stresses, and cloning of their
genes is now in progress. Molecular and genetic analyses of the regulation of
gene expression and signal transduction cascades proceed extensively, and will
give us more precise insight on the plants' responses to environmental stresses
and their adaptation processes. These stress-related genes are thought to
become useful resources to produce stress tolerant crops using gene
manipulation. In this book, recent progresses on molecular mechanisms of
plant responses and tolerance to drought, salt, cold and heat stresses are
reviewed by active researchers in this field. I hope that this book stimulates
young students and researchers to become interested in new plant science based
on molecular biology and new plant biotechnology.
CHAPTER 1
Genetic Approaches
to Abiotic Stress Responses
M. Koornneef and A.J.M. Peeters
P lants grow in almost any part of the world and under a wide variety of nutrient and
climatic conditions differing in temperature, light quantity and quality and availability
of water. Plants that grow in a specific environment are adapted to these different local
conditions, and can also cope with changes in these conditions which might be adverse for
their growth and development. Adaptation is required because plants cannot escape
unfavorable conditions due to their sessile growth habit. This implies that species differ
genetically in their adaptation and resistance to abiotic stresses. Although more restricted,
genetic variation for the adaptation to abiotic stresses can also be present within species and
has been used for plant breeding practice. Examples of how plants deal with extreme
temporarily adverse conditions are the so-called resurrection plants, which can lose more
than 90% of their water content, but still are able to revive when supplied again with water.
Examples of plants that can grow under extreme low temperatures are those that grow in
arctic regions or at high altitudes.
Plants can be preadapted to stress conditions but often various protection mechanisms
are induced by the stress treatments itself. This implies that plants are able to perceive stress
signals and that after perception signal transduction events take place. As a consequence,
these lead to changes in gene expression, as indicated by the many situations where
upregulation of genes is observed after the application of various types of abiotic stress
(reviewed by Zhu et al1). Ultimately various cellular mechanisms are set in place, which
allow the plant to cope with the stress imposed. These mechanisms are for instance osmo-
adjustment and osmo-protection, changes in pathways affecting ion and water fluxes,
production of protection proteins etc.2,3 In case of osmo-adjustment the osmotic potential
of the cell is lowered to favor water uptake and maintenance of turgor. Osmoprotectants
stabilize proteins and membranes when present in high concentrations and include a variety of
compounds such as amino acids (proline), quaternary ammonium compounds
(betaines), polyols (pinitol, mannitol), sugars such as fructans2 and specific proteins such as
dehydrins.5 The introduction of genes leading to increased levels of such compounds in
transgenic plants has resulted in increased stress tolerance.2,4 The genes used for this were
often of microbiological origin. Certain gene products might also be involved in the repair
of damage caused by the stress.
In addition to the cellular content, membranes also play an important role in
adaptation. Especially, the degree of saturation of the membrane lipids is an important
factor in this.6 When studying the response to stress, one should take into account that
organs can differ in this respect. As an example seeds, and often pollen also, can survive
extreme desiccation, whereas the vegetative parts and flowers are susceptible to such
Molecular Responses to Cold, Drought, Heat and Salt Stress in Higher Plants, edited by Kazuo
Shinozaki and Kazuko Yamaguchi-Shinozaki. ©1999 R.G. Landes Company.
2 Molecular Responses to Cold, Drought, Heat and Salt Stress in Higher Plants
conditions. This response allows the seeds to survive in a dry state and is also present in
those species that grow under favorable conditions. The acquisition of this desiccation
tolerance during seed maturation is very similar to vegetative responses to water deficits.
In addition to these cellular mechanisms, plants can control their water status by
controlling water uptake and water loss. Water uptake can be regulated by the architecture
and physiology of the root system, whereas water loss is regulated not only by controlling
morphological modifications to avoid excessive loss of water such as specific surface
structures found in succulent plants but also by the strict control of stomatal aperture. When
such adaptations result in stress resistance the mechanism has been called “avoidance.”
Despite variation in the nature of adverse conditions, it should be emphasized that
abiotic stresses can have components in common. Insufficient water supply can result from
an excessive loss of water, or an insufficient uptake of water. The latter can also result from
a high concentration of osmotic material in water, which usually is salts. Chilling and freezing
may also lead to osmotic stress due to reduced water absorption and cellular dehydration.
It is likely that for coping with other types of abiotic stresses such as UV light, heat, touch,
wounding and hypoxia, plants have different mechanisms available. As an
example Reactive oxygen species (ROS) are involved in the damage due to ozone but also
have been implicated in the damage that occurs from drought and chilling stress.7 Plant
possess a number of mechanisms and enzyme systems to scavenge ROS. However, the
protection of ROS targets is also a mechanism to deal with such damage and recently Shen et
al7 showed that mannitol can play such a role by protecting the enzyme phosphoribulokinase
against oxidative inactivation.
Another type of abiotic stress, which has its specific mechanisms and genetics, deals with
heavy metals. The latter topic is beyond the scope of this review.
The Genetic Approach in Stress Physiology—General Principals

The genetic approach in stress biology is based on finding genetic differences in stress
responses and to relate this to the structure and function of the genes involved. The use
genetic variation with a “detectable” phenotype is an important tool to identify genes
related to stress and stress tolerance, because it allows the identification of the respective
genes by various techniques such as map based cloning or tagging. The feasibility of these
techniques in model species such as Arabidopsis explains why much of the genetic analysis
focuses on this species.
Genetic variation can be generated by mutations, but also exists in nature or, in case of
crop plants, among cultivated varieties. Genetic variation within species is often of a different
type than that found in mutant screens. In contrast to mutants, which can be grown in
protective environments to survive even when they are weak growing plants, natural variants
need to survive under normal growth conditions and therefore in any case should be well
growing plants.
In order to understand the underlying biochemical and molecular bases of genetic
variation present in nature, it is important that the genetic differences in stress tolerance
can be correlated with traits and genes that confer this difference in tolerance. For this it is
not sufficient to see a correlation in the parents, but this correlation between traits should
be analyzed in segregating populations. However, a situation of close linkage still does not
prove that these traits are due to the same gene (pleiotropism). Pleiotropism cannot be
distinguished easily from close linkage by segregation analysis, unless very large populations
are used.
Since stress tolerance genetically behaves as a quantitative trait with large environmental
effects on the parameters to be analyzed and often is under polygenic control, detailed
genetic analyses are difficult. However, the advent of molecular markers, and the use of
Genetic Approaches to Abiotic Stress Responses 3
specific mapping populations and developments in statistical methods, have improved the
so called QTL (quantitative trait loci) analysis very much and make this also suitable for the
genetic analysis of stress tolerance.8 The more genes with known functions are placed on
the various genetic maps, the more candidate genes for stress tolerance can be identified by
the co-localisation of QTLs and candidate genes.9 The relation between the map location of
the various dehydrin genes with the map position of a number of genes related to stress
tolerance and other physiological traits in cereals is an example of this candidate gene
approach.10 However, confirmation of a causal relationship should preferentially come from
gene transfer experiments with alleles cloned from the various parental genotypes. In cases
where individual QTLs have an effect that is large enough, it may be possible to identify the
respective gene by map-based cloning approaches.
Additional genetic variation can be generated by the introduction of specific genes
which originate from other plant species or even from completely different organisms. The
latter has opened completely new ways to modify stress tolerance in crop plants.2,4
Furthermore, genetics and gene transfer technology not only allow the modification of
stress resistance but also are important tools as functional tests of components involved in
the response of plants to abiotic stresses.2
Since most mutations reflect damages within the gene or its controlling elements, one
expects that mutants lacking a function are most frequent. However, even loss of function
mutants can result in an increase in the functioning of pathways when repressors of such
pathways are mutated. Not all genes can be identified by looking for mutant phenotypes. A
reason for this is that for some traits, either the phenotype is relatively subtle, or the
phenotype is too general, e.g., reduction of plant size or vigor. To solve the problem of
subtle phenotypes more sophisticated screens, for instance by using reporter genes, are
being developed and will be described hereafter. Another reason for not finding specific
mutants is that for many genes genetic redundancy exists. This means that mutations in
such genes do not result in an obviously visible phenotype since the redundant counterpart
produces enough product to (partially) substitute the function of the mutated gene. Recently a
number of additional methods which include enhancer or gene trapping and reverse
genetics11,12 have become available and allow the analysis of phenotypes associated with
the (loss of) function of specific genes. Reverse genetics uses gene sequences that are identified
by “molecular” approaches and in large scale sequencing projects. When collections of plants
with insertions of T-DNA or transposable elements are available, one can use these to
identify insertions in the gene of interest. Plants with insertions in such genes are
identified by the ability to amplify DNA fragments in polymerase chain reaction (PCR).
One PCR primer is based on the tagging-DNA, whereas the second is based on the sequence
for which an insertion mutant is sought. In case no mutant phenotype is observed when the
respective gene is disrupted, due to the redundancy mentioned above, one can expect
mutant phenotype, when (insertion) mutations in duplicated genes are combined by
cROSsing. In addition to loss of function mutations, gain of function mutants can be
isolated by introducing enhancer sequences next to relevant genes resulting in activation of
such genes (activation tagging).13 The study of transgenic plants in which cloned genes
are over-expressed often has given clues about the function of the respective genes.
The development of appropriate mutant screens is a crucial aspect in any genetic
approach. One can focus on the stress trait itself, which might be considered as the
end point of a signal transduction chain. However, one can also pay attention to specific
components known to be involved in stress responses. Examples of the latter are the search
and analysis of mutants affected in abscisic acid (ABA) biosynthesis or ABA response and
the analysis of mutants affecting specific stress related genes, either by finding mutations in
those genes or by finding mutations that modify the expression of such genes. The ease of
finding mutants also depends on the effectiveness of the screening system. For this, resistance
to, for instance, salt seems a more attractive screen than the isolation of salt susceptible geno-
types. However, screens of the latter type have been successfully applied in case of salt14 and
frost.15
A number of these direct and indirect approaches to select mutants affected in stress
responses will be described, together with the results obtained in analyzing “natural”
genetic differences. In what way these analyses have led and are expected to lead to a
better understanding of the responses to abiotic stress will also be discussed.
Genetic Differences in Salt Tolerance

Salt overly sensitive (sos) Arabidopsis mutants were isolated by the inability of their roots
to grow on 50 mM NaCl.14 It was shown that these mutants were defective in the high-
affinity K+ uptake system. These mutants are hypersensitive to salt stress because relatively
high Na+ concentrations inhibit the residual low affinity system of Na+ and K+, which
results in potassium deficiency, which is the cause of the growth defect. The mutants also
affect Na+ uptake and consequently accumulate less.15 In contrast to sos1 mutant, the sos3
mutant can be rescued by high external Ca2+.16 Attempts to isolate salt tolerant mutants in
Arabidopsis resulted in the rss mutants, that express this tolerance only at the seed-germination
level.17,18 NaCl tolerance at the germination level was also found to be a characteristic of
ABA deficient mutants, which trait in these mutants was considered as reflecting the
reduced seed dormancy characteristic of ABA deficient mutants.19 No seed dormancy or
ABA related phenotype was reported for the rss mutants, which locus maps at a different
position than the three known ABA deficient (aba) mutants.
Genetic variation for salt tolerance has been described in many crop plants and their
wild relatives. In a number of cases, these have been associated with selective ion uptake.20
Hexaploid wheat (Triticum aestivum) is more salt tolerant than tetraploid durum wheat
(T. turgidum). It was shown that in hexaploid wheat a single gene (Kna1) is responsible for
accumulating less Na+ and more K+ in expanding and young leaves. This gene is located on
chromosome 4D, which is lacking in tetraploid wheats.21
The examples mentioned indicate that ion uptake and ion transport can affect stress
tolerance. However, in addition, genetic differences in salt tolerance have also been
associated with properties that relate to other osmotic mechanisms. For example, Saneoka
et al20 found that near-isogenic maize lines differing in glycine betaines also differed in
salt tolerance. Moons et al23 showed that ABA accumulation upon salt stress was much
more in a salt tolerant rice cultivar as compared with a sensitive variety. The accumulation
of a number of ABA induced LEA dehydrin type proteins was also higher in the tolerant
varieties. Whether the latter correlations are causal could not be established, since the various
traits were not analyzed in segregating populations.
Salt tolerance seemed an attractive system for selection at the cell and tissue culture
level. However, success of this approach appeared limited to a few successful examples where
salt tolerant plants could also be obtained.24
Genetic Differences in the Response to Low Temperatures

The ability of plants to tolerate low temperatures differs much between species.
Distinctions are often made between chilling (temperatures <10˚C) and frost at subzero
temperatures. Exposure to low non-freezing temperatures can strongly increase the freezing
tolerance of a species, which process is called acclimation. This interaction between these
two temperature regimes complicates the analysis of cold tolerance.
Screens for cold sensitive mutants have been applied in Arabidopsis, where a number of
loci were identified. In a direct screen for lack of growth at 10˚C-13˚C chilling sensitive (chs)
mutants were found. The chs1 mutant has been described in some detail and was shown to
be defective in the accumulation of newly synthesized chloroplast localized polypeptides at
low temperatures.25 However, the primary defect of this mutant is not known.
It has also been well established that in plants the level of unsaturated fatty acids in the
glycerolipids of membranes changes with changes in growth temperature.6 The alteration
of the level of enzymes controlling this trait, in transgenic plants, modifies cold tolerance.6
Furthermore, it was expected that genotypes with reduced levels of acyl-lipid desaturases
might be more sensitive to low temperatures. This was shown for various Arabidopsis
mutants.26,27 An exception to this appeared to be the fab1 mutant of Arabidopsis (defective in
palmitoyl-ACP elongase).28 However, since chlorosis was observed at 2˚C in continous light,
it appears that damage depends on the light conditions in which the plants were tested. At
high light intensities the unsaturated species of phosphatidylglycerol accelerate the
recovery process from the low temperature inactivated state of the photosystem II complex.6
This indicates the complexity of the various physiological interactions and thereby
the difficulty in designing the proper test conditions to identify genetic differences.
Frost sensitive mutants were isolated in direct screens for such mutants and resulted in
the finding of seven different complementation groups, named sfr1-sfr7 (sensitivity to
freezing). 15 Mutants at the loci sfr1, sfr2, sfr3 and sfr5 had no obvious pleiotropic effects
such as slow growth at the seedling stage, which was observed in the sfr4 and sfr7 mutants.
The sfr6 mutant had several additional pleiotropic defects.15 Thus far the biochemical
defect or the nature of the genes is not known for any of these genes.
Relatively large genetic differences have been reported, especially in the grass family,
where, as expected, winter cereals are clearly more frost tolerant than summer cereals. This
led to the identification of some single gene differences for winter hardiness, which map at
similar genetic positions in wheat, barley and rye.29 Although some recombinants have been
found, thereby excluding pleiotropism, it is of interest that responsiveness to vernalization,
which affects flowering time and thereby winter versus spring habit, and traits such as ABA
accumulation30 and some dehydrin genes map to a similar position.10 Therefore, causal
relationships of some properties can not be excluded but need confirmation.
ABA Related Mutants

The Isolation of ABA Related Mutants
The plant hormone abscisic acid (ABA) seems a crucial component of the response of
plants to stress. ABA levels increase rapidly, especially upon water stress, and result in
stomatal closure. This is caused by changes in gating properties of ion channels, which lead
to changes in the turgor of the two surrounding guard cells. It is assumed that rapid changes
in cell turgor (cell shrinkage) lead to this increase in ABA biosynthesis. The effect of cold
and osmotic stress on ABA levels may be due to the effect on the water status of the cells
caused by these stresses. This signal induces gene expression of the cleavage enzyme gene of
ABA biosynthesis,31 but how this is achieved is not known. However, the role of ABA in the
acclimation process, which is associated with changes in gene expression may be slower
than the effects on stomatal closure, as was shown by the ability of ABA to mimic the
acclimation treatment.32,33 Recent progress on the signal transduction pathway of ABA has
been reviewed by Leung and Giraudat34 and Shinozaki and Yamaguchi-Shinozaki.35
In addition to effects on water relations and stress tolerance, ABA also plays a crucial
role during seed development and germination. This property allowed the efficient
screening for ABA related mutants at the seed germination level.
The lack of seed dormancy due to ABA deficiency is shown by a high percentage of
germination of freshly harvested seeds and of seeds in darkness.36 Furthermore, the gibberellin
(GA) requirement for Arabidopsis seeds to germinate is abolished or strongly reduced in
ABA-deficient and ABA-insensitive mutants. This allows seeds to germinate in the presence
of inhibitors of GA biosynthesis such as tetcyclacis and paclobutrazol. These properties
have been the basis for the selection of ABA-deficient mutants at the ABA1, ABA2 and ABA3
loci in Arabidopsis19,36 ABA-insensitive (abi1-abi5) mutants were selected as seeds that
germinated at ABA concentrations that normally inhibit germination.37,38 These mutants
affect many ABA responses (abi1 and abi2) or only seed germination (abi3, abi4, and abi5).31,38
The cool mutant of barley39 is an example of a mutant in which ABA insensitivity is
specific for stomatal closure upon drought stress. ABA-insensitive mutants for which
insensitivity is restricted to the growth response were isolated as mutants for which root
growth is not inhibited by ABA. This class of mutants is called gca (growth control via ABA)
and comprises at least 8 loci.40 However, the gca1 and gca2 mutants also show defects
in stomatal closure. Mutants which are hypersensitive to ABA are the era1-era3 (enhanced
response to ABA) mutants, which do not germinate at ABA concentrations that normally
permit germination of the wild type.41 The era1 mutants are also affected in several
adult plant responses.
The Molecular Function of ABA, ABI and ERA Genes

Several procedures allow the cloning of genes based on mutants. Map-based cloning
has been used to clone the ABI142,43 and ABI3 loci.44 T-DNA tagging was used to clone
ERA1.41 Mutants tagged with transposable elements were used to clone the ABA2 gene in
N. plumbaginifolia45 and the VP14 gene in maize,31 which both encode steps in ABA
biosynthesis. Through homology with the Nicotiana ABA2 gene, the Arabidopsis ABA1 gene,
encoding zeaxanthin epoxydase was isolated.45 VP14 controls the cleavage of carotenoids,
which is considered to be the key regulatory step in ABA biosynthesis.31,46 The ABI1 gene
encodes a serine/threonine protein phosphatase 2C enzyme42,43,47 and by homology with
ABI1, the ABI2 gene was isolated.48 The similarity between ABI1 and ABI2 indicates that
these genes have overlapping functions. The null mutants at either ABI1 and ABI2 may have
no, or a very mild, phenotype and only mutants with a dominant negative effect would
block the phosphatases encoded by both genes. The dominance of both mutations and the
very similar amino acid substitution present in both mutants are in accordance with this
hypothesis.48
The amino acid sequence of ABI344 revealed that this gene is homologous to the maize
Viviparous-1 (Vp1) gene,49 which is a transcription factor that is specific for seed development.
The amino acid sequence of ERA1 showed that this gene encodes the β subunit of a farnesyl
transferase. The era1 mutant lacks peptide farnesylation activity and it is suggested that the
ERA1 gene acts as negative regulator by modifying ABA signal transduction proteins for
membrane localization.41
The Isolation of Mutants Affected in the Response to Stress

and/or ABA Induced Genes and Processes
Novel ways to identify genes interacting with stress induced genes apply the expression of
reporter genes which are under the control of promoters of these genes.
Mutants with an altered expression of these reporters are expected to also be affected
also in the expression of the endogenous promoter. When this has been confirmed the
changes in physiological properties of the mutant should indicate the relevance of the
expression of this and related genes that might be down or upregulated. Reporter genes
that are most efficient are those where the reporter gene assay is non-destructive, as, e.g.,
in the case of luciferase. That this approach can yield numerous mutants of which many have
no dramatic phenotypic effects was shown by Ishitani et al,50 who used the rd29A promoter
driving luciferase. The promoter of this gene contains both an ABA independent drought
responsive (DRE) and an ABA dependent (ABRE) element. Screens based on this approach
were very productive and resulted both in mutants with constitutive expression, and in
mutants with low and high expression of osmotically responsive genes named respectively
cos, los and hos mutants.50 Within the los and hos class, subclasses could be distinguished
according to defects in their responses to one or a combination of stress and ABA signals.
An approach that also makes use of reporter genes is the use of enhancer or gene trap
procedures, where insertions of reporter genes with a minimal or without a promoter might
result in expression specifically after stress induction.12 This approach might also allow the
identification of genes that are redundant and therefore cannot be found in mutant screens.
Furini et al51 used activation tagging to identify a gene in ABA signaling. In this system
the T-DNA contains enhancers of gene expression and was used to select inserts that would
confer dehydration tolerance to callus of Craterostigma plantagineum, which normally
requires ABA to obtain this property. The cloning of the sequences involved revealed
unusual features and resembled transposon-like sequences.51
Genetic Variation for Morphological Traits Related to Stress Tolerance

Morphological adaptations can also be the basis of genetic variation in both water loss
and water uptake. With respect to the latter, root morphology and rooting patterns are
important. A genetic analysis of these root characters in a segregating population derived
from a cross between a drought resistant upland rice cultivar and a drought sensitive lowland
variety is described by Champoux et al.52 The same material allowed also the detection of
QTLs controlling drought tolerance in a shoot specific way through osmotic adjustment.53
Following a similar QTL approach, Price et al54 investigated leaf rolling, stomatal
behavior and heading date as examples of morphological and physiological traits related to
stress tolerance in rice. The correlation found in rice between smaller leaves and ABA
accumulation observed in similar crosses was not confirmed to be pleiotropic in a refined
QTL analysis, but due instead to linkage.55
Conclusions
Genetic variants tell which intrinsic properties, such as chemical composition of cells,
are important for stress tolerance. Both mutants and transgenic plants have confirmed a
number hypotheses based on earlier observations and physiological experiments. Mutants
affected in specific regulatory factors such as ABA showed that this compound mediates the
expression of many but not all stress induced genes. This indicated that different pathways
are involved.35 The complexity of this stress-induced signal transduction pathway is also
suggested by the large number of mutant classes identified by Ishitani et al.50 Such mutants
will be useful to identify the various steps in this pathway and will complement the
molecular approaches.1,3,4,35 A further challenge of genetics is to identify, up to the
molecular level, the genes controlling natural differences in stress tolerance. Although the
effects of individual genes in some cases may be limited, it can be expected that these
differences are useful in application of cloned genes because they operate with limited
pleiotropic effects. This approach to the identification of the genes for which variation is
present in nature will complement the mutant approaches and molecular approaches. It
can be expected that in a number of cases the same genes will be identified. However, in
other situation they are likely to be different. Together, these genes will tell us how plants
cope with and respond to abiotic stresses.
Acknowledgments
Our research program is supported by the Human Frontier Science Program
(RG-303/95) and the BIOT4 program of the European Union (BIO4-CT96-0062).
References
1. Zhu JK, Hasegawa PM, Bressan RA. Molecular aspects of osmotic stress in plants. Crit Rev
Plant Sci 1997; 16:253-277.
2. Bohnert HJ, Nelson DE, Jensen RG. Adaptations to environmental stresses. Plant Cell 1995;
7:1099-1111.
3. Bray EA. Plant responses to water deficit. Trends Plant Sci 1997; 2:48-54.
4. Ingram J, Bartels D. The molecular basis of dehydration tolerance in plants. Annu Rev
Plant Physiol Plant Mol Biol 1996; 47:377-403.
5. Close TJ. Dehydrins: A commonalty in the response of plants to dehydration and
low temperature. Physiol Plant 1997; 100:291-296.
6. Nishida I, Murata N. Chilling sensitivity in plants and cyanobacteria: The crucial contri-
bution of membrane lipids. Annu Rev Plant Physiol Plant Mol Biol 1996; 47:541-568.
7. Shen B, Jensen RG, Bohnert HJ. Mannitol protects against oxidation by hydroxyl radicals.
Plant Physiol 1997; 115:527-532.
8. Koornneef M, Alonso-Blanco C, Peeters AJM. Genetic approaches in plant physiology. New
Phytol 1997; 137:1-8.
9. Prioul JL, Quarrie S, Causse M et al. Dissecting complex physiological functions through
the use of molecular quantitative genetics. J Exp Bot 1997; 48:1151-1163.
10. Campbell SA, Close TJ. Dehydrins: Genes, proteins, and associations with phenotypic traits.
New Phytol 1997; 137:61-74.
11. Azpiroz-Leehan R, Feldmann KA. T-DNA insertion mutagenesis in Arabidopsis: Going back
and forth. Trends Genet 1997; 13:152-156.
12. Sundaresan V. Horizontal spread of transposon mutagenesis: New uses for old elements.
Trends Plant Sci 1996; 1:184-190.
13. Walden R, Fritze K, Hayashi H et al. Activation tagging: A means of isolating genes
implicated as playing a role in plant growth and development. Plant Mol Biol 1994;
26:1521-1528.
14. Wu SJ, Ding L, Zhu JK. SOS1, a genetic locus essential for salt tolerance and potassium
acquisition. Plant Cell 1996; 8:617-627.
15. Warren G, McKown R, Marin A et al. Isolation of mutations affecting the development of
freezing tolerance in Arabidopsis thaliana (L.) Heynh. Plant Physiol 1996; 111:1011-1019.
16. Liu J, Zhu JK. An Arabidopsis mutant that requires increased calcium for potassium
nutrition and salt tolerance. Proc Natl Acad Sci USA 1997; 94:14960-14964.
17. Saleki R, Young PG, Lefebvre DD. Mutants of Arabidopsis thaliana capable of germination
under saline conditions. Plant Physiol 1993; 101:839-845.
18. Werner JE, Finkelstein RR. Arabidopsis mutants with reduced response to NaCl and
osmotic stress. Physiol Plant 1995; 93:659-666.
19. Leon-Kloosterziel KM, Gil MA, Ruijs GJ et al. Isolation and characterization of abscisic
acid-deficient Arabidopsis mutants at two new loci. Plant J 1996; 10:655-661.
20. Foolad MR. Genetic basis of physiological traits related to salt tolerance in tomato,
Lycopersicon esculentum Mill. Plant Breed 1997; 116:53-58.
21. Dubcovsky J, Santa MG, Epstein E et al. Mapping of the K+/Na+ discrimination locus Kna1
in wheat. Theor Appl Genet 1996; 92:448-454.
22. Saneoka H, Nagasaka C, Hahn DT et al. Salt tolerance of glycinebetaine-deficient
and-containing maize lines. Plant Physiol 1995; 107:631-638.
23. Moons A, Bauw G, Prinsen E et al. Molecular and physiological response to abscisic acid
and salts in roots of salt-sensitive and salt-tolerant indica rice varieties. Plant Physiol 1995;
107:177-186.
24. Winicov I. Characterization of salt tolerant alfalfa (Medicago sativa L.) plants regenerated
from salt tolerant cell lines. Plant Cell Rep 1991; 10:561-564.
25. Schneider JC, Nielsen E, Somerville C. A chilling-sensitive mutant of Arabidopsis is
deficient in chloroplast protein accumulation at low temperature. Plant Cell Environment
1995; 18:23-32.
26. Hugly S, Somerville C. A role for membrane lipid polyunsaturation in chloroplast biogenesis
at low temperature. Plant Physiol 1992; 99:197-202.
27. Miquel M, James DJ, Dooner H et al. Arabidopsis requires polyunsaturated lipids for low
temperature survival. Proc Natl Acad Sci USA 1993; 90:6208-6212.
28. Wu J, Browse J. Elevated levels of high-melting-point phosphatidylglycerols do not induce
chilling sensitivity in an Arabidopsis mutant. Plant Cell 1995; 7:17-27.
29. Galiba G, Quarrie SA, Sutka J et al. RFLP mapping of the vernalization (Vrn1) and frost
resistance (Fr1) genes on chromosome 5A of wheat. Theor Appl Genet 1995; 90:1174-1179.
30. Quarrie SA, Gulli M, Calestani C et al. Location of a gene regulating drought-induced
abscisic acid production on the long arm of chromosome 5A of wheat. Theor Appl Genet
1994; 89:794-800.
31. Tan BC, Schwartz SH, Zeevaart JAD et al. Genetic control of abscisic acid biosynthesis in
maize. Proc Natl Acad Sci USA 1997; 94:12235-12240.
32. Heino P, Sandman G, Lang V et al. Abscisic acid deficiency prevents development of freezing
tolerance in Arabidopsis thaliana (L.) Heynh. Theor Appl Genet 1990; 79:801-806.
33. Gilmour SJ, Thomashow MF. Cold acclimation and cold-regulated gene expression in ABA
mutants of Arabidopsis thaliana. Plant Mol Biol 1991; 17:1233-1240.
34. Leung J, Giraudat J. Abscisic acid signal transduction. Annu Rev Plant Physiol Plant Mol
Biol 1998; 49:199-222.
35. Shinozaki K, Yamaguchi-Shinozaki K. Gene expression and signal transduction in water-
stress response. Plant Physiol 1997; 115:327-334.
36. Koornneef M, Jorna ML, Brinkhorst-van der Swan DLC et al. The isolation of abscisic
acid (ABA) deficient mutants by selection of induced revertants in non-germinating
gibberillin sensitive lines of Arabidopsis thaliana (L.) Heynh. Theor Appl Genet 1982; 61:385-
393.
37. Koornneef M, Reuling G, Karssen CM. The isolation and characterization of abscisic acid-
insensitive mutants of Arabidopsis thaliana. Physiol Plant 1984; 61:377-383.
38. Finkelstein RR. Mutations at two new Arabidopsis ABA response loci are similar to the
abi3 mutations. Plant J 1994; 5:765-771.
39. Raskin I, Ladyman JAR. Isolation and characterization of a barley mutant with abscisic-
acid-insensitive stomata. Planta 1988; 173:73-78.
40. Benning G, Ehrler T, Meyer K, Leube M, Rodriguez P, Grill E. Genetic analysis of
ABA-mediated control of plant growth. In: Abscisic acid signal transduction in plants.
Madrid: Juan March Foundation, 1996:34.
41. Cutler S, Ghassemian M, Bonetta D et al. A protein farnesyl transferase involved in ABA
signal transduction in Arabidopsis. Science 1996; 273:1239-1241.
42. Leung J, Bouvier-Durand M, Morris PC et al. Arabidopsis ABA response gene ABI1:
Features of a calcium-modulated protein phosphatase. Science 1994; 264:1448-1452.
43. Meyer K, Leube MP, Grill E. A protein phosphatase 2C involved in ABA signal transduction
in Arabidopsis thaliana. Science 1994; 264:1452-1455.
44. Giraudat J, Hauge BM, Valon C et al. Isolation of the Arabidopsis ABI3 gene by po-
sitional cloning. Plant Cell 1992; 4:1251-1261.
45. Marin E, Nussaume L, Quesada A et al. Molecular identification of zeaxanthin epoxidase
of Nicotiana plumbaginifolia, a gene involved in abscisic acid biosynthesis and corresponding
to the ABA locus of Arabidopsis thaliana. EMBO J 1996; 15:2331-2342.
46. Schwartz SH, Tan BC, Gage DA et al. Specific oxidative cleavage of carotenoids by vp14 of
maize. Science 1997; 276:1872-1874.
47. Bertauche N, Leung J, Giraudat J. Protein phosphatase activity of abscisic acid insensitive 1
(ABI1) protein from Arabidopsis thaliana. Eur J Biochem 1996; 241:193-200.
48. Leung J, Merlot S, Giraudat J. The Arabidopsis ABI2 gene is a homologue of ABI1 and
implicates redundant protein serine/threonine phosphatases 2C in abscisic acid signal
transduction. Plant Cell 1997.
49. McCarty DR, Hattori T, Carson CB et al. The Viviparous-1 developmental gene of maize
encodes a novel transcriptional activator. Cell 1991; 66:895-906.
50. Ishitani.M., Xiong L, Stevenson B et al. Genetic analysis of osmotic and cold stress signal
transduction in Arabidopsis: Interactions and convergence of abscisic acid-dependent and
abscisic acid-independent pathways. Plant Cell 1997; 9:1-16.
51. Furini A, Koncz C, Salamini F et al. High level transcription of a member of a repeated
gene family confers dehydration tolerance to callus tissue of Craterostigma plantagineum.
EMBO J 1997; 16:3599-3608.
52. Champoux MC, Wang G, Sarkarung S et al. Locating genes associated with root morphology
and drought avoidance in rice via linkage to molecular markers. Theor Appl Genet 1995;
90:969-981.
53. Lilley JM, Ludlow MM, McCouch SR et al. Locating QTL for osmotic adjustment and
dehydration tolerance in rice. J Exp Bot 1996; 47:1427-1436.
54. Price AH, Young EM, Tomos AD. Quantitative trait loci associated with stomatal
conductance, leaf rolling and heading date mapped in upland rice (Oryza sativa). New
Phytol 1997; 137:83-91.
55. Quarrie SA, Laurie DA, Zhu JH et al. QTL analysis to study the association between leaf
size and abscisic acid accumulation in droughted rice leaves and comparisons across cereals.
Plant Mol Biol 1997; 35:155-165.
CHAPTER 2
Molecular Responses to Drought

Stress
Kazuo Shinozaki and Kazuko Yamaguchi-Shinozaki
P lant growth is greatly affected by environmental abiotic stresses, such as drought, high
salinity and low temperature. Plants respond and adapt to these stresses in order to sur-
vive against abiotic stress. Among these abiotic stresses, drought or water deficit is the most
severe limiting factor of plant growth and crop production. Drought stress induces various
biochemical and physiological responses in plants. Recently, a number of genes have been
described that respond to drought at the transcriptional level.1-4 Their gene products are
thought to function in stress tolerance and response (Fig. 2.1). Recently, stress-inducible
genes were used to improve stress tolerance of plants by gene transfer. It is important to
analyze functions of stress-inducible genes not only for further understanding of molecular
mechanisms of stress tolerance and response of higher plants but also for improvement of
stress tolerance of crops by gene manipulation.
The plant hormone abscisic acid (ABA) is produced under water deficit conditions
and plays important roles in response and tolerance to dehydration. Most of the genes that
have been studied to date are also induced by ABA.5 It appears that dehydration triggers
the production of ABA, which, in turn, induces various genes. Several reports have
described genes that are induced by dehydration but are not responsive to exogenous ABA
treatments. These findings suggest the existence of ABA-independent as well as ABA-
dependent signal-transduction cascades between the initial signal of drought stress and
the expression of specific genes.1-4 To understand the molecular mechanisms of gene
expression in response to drought stress, cis- and trans-acting elements that function in
ABA-independent and ABA-responsive gene expression by drought stress have been
precisely analyzed. A variety of transcription factors are involved in stress responsive gene
expression, which suggests the involvement of complex regulatory systems in molecular
responses to drought stress.
Expression and functions of stress-inducible genes have been studied at molecular
level as described in this chapter. Complex mechanisms seem to be involved in gene
expression and signal transduction in response to drought stress. However, genetic analyses of
drought-resistant or drought-sensitive mutants have not been extensively performed. There-
fore, details of molecular mechanisms of regulating plant genes to drought stress remain
to be solved concerning signal transduction cascades. These include the sensing mechanisms
of osmotic stress, modulation of the stress signals to cellular signals, transduction of the
cellular signals to the nucleus, second messengers involved in stress signal transduction,
roles of ABA in the signaling process, transcriptional control of stress-inducible genes, and
the function and cooperation of stress-inducible genes allowing drought stress tolerance
(Fig. 2.1). In this article we describe recent progress mainly on gene expression and
Fig. 2.1. Schematic representation of molecular responses to drought stress in plant cells.
Molecular and cellular responses to drought stress include perception of dehydration signal,
signal transduction to cytoplasm and nucleus, gene expression, and responses and tolerance to
drought stress.
signal transduction in drought stress response. Promoter analysis of several drought

inducible genes suggests that there are at least four signaling cascades in drought stress
responses. Several approaches to improve stress tolerance of plants by gene transfer of
stress-inducible genes are also described.
A Variety of Functions of Drought-Inducible Genes

Two Classes of Drought-Inducible Genes
Various genes respond to drought stress in various species, and functions of their
gene products have been predicted from sequence homology with known proteins. Many
drought-inducible genes are also induced by salt stress (see chapter 3) and low temperature
(see chapter 4), which suggests the existence of similar mechanisms of stress responses.
Genes induced during drought-stress conditions are thought to function not only in
protecting cells from water deficit by the production of important metabolic proteins but
also in the regulation of genes for signal transduction in the drought stress response.1,2,4
Thus, these gene products are classified into two groups (Fig. 2.2). The first group includes
proteins that probably function in stress tolerance, such as chaperones, LEA (late embryo-
genesis abundant) proteins, osmotin, antifreeze proteins, mRNA binding proteins, key
enzymes for osmolyte biosynthesis, water channel proteins, sugar and proline transporters,
detoxification enzymes and various proteases. LEA proteins, chaperones and mRNA binding
proteins have been analyzed biochemically and shown to be involved in protecting macro-
molecules like enzymes, lipids and mRNAs from dehydration. Proline, glycine betaine and
sugars function as osmolytes and in protecting cells from dehydration (see chapter 8). Key
enzymes of several osmolytes have been cloned and analyzed biochemically. Water channel
Molecular Responses to Drought Stress 13
Fig. 2.2. Drought stress-inducible genes and their possible functions in stress tolerance and
response. Gene products are classifed into two groups. The first group includes proteins that
probably function in stress tolerance (function proteins; open boxes), and the second group
contains protein factors involved in further regulation of signal transduction and gene expression
that probably function in stress response (regulatory proteins; shadowed boxes).
proteins, sugar transporters and proline transporters are thought to function in transport
of water, sugars and proline through plasma membranes and tonoplast to adjust osmotic
pressure under stress conditions. Detoxification enzymes such as glutathione S-transferase,
superoxide dismutase, and soluble epoxide hydrolase are involved in protection of cells
from active oxygens. Proteases including thiol proteases, Clp protease, and ubiquitin are
thought to be required for protein turnover and recycle of amino acids.
The second group contains protein factors involved in further regulation of signal
transduction and gene expression that probably function in stress response: protein
kinases, transcription factors and enzymes in phospholipid metabolism.1,2,4 Genes for a
variety of transcription factors that contain typical DNA binding motifs, such as bZIP,
MYB, MYC, EREBP/AP2 and zinc fingers, have been demonstrated to be stress inducible.4
These transcription factors function in further regulation of various functional genes under
stress conditions. Various protein kinases, such as MAP kinases, calcium dependent protein
kinases (CDPK), SNF1 related protein kinase and ribosomal S6 kinase, were demonstrated
to be induced or upregulated by dehydration.4,7 Stress-inducible genes for protein phos-
phatases are reported.8 These protein kinases and phosphatases may be involved in
modification of functional proteins and regulatory proteins involved in stress signal trans-
duction pathways. Phospholipid, such as inositol-1,4,5-triphosphate, diacylglycerol and
phospahtidic acid are believed to be involved in stress signaling processes in plants.
Enzymes involved in phospholipids metabolism whose genes are stress-inducible may play
important roles in stress signaling as well.
Existence of a variety of drought-inducible genes suggests complex responses of plants

to drought stress. Their gene products are involved in drought stress tolerance and stress
responses.
Improvement of Stress Tolerance Using Gene Transfer

Recently, several different approaches were attempted to improve stress tolerance of
plants by gene transfer of stress-inducible genes.9 Stress-inducible genes for functional
proteins such as key enzymes for osmolyte biosynthesis, LEA proteins and detoxification
enzymes were overexpressed in transgenic plants to produce stress tolerant phenotype of
the plants, which indicates that their gene products really function in stress tolerance. Genes
used for transformation were those encoding enzymes required for biosynthesis of various
osmoprotectants, such as Escherichia coli mannitol 1-phosphate dehydrogenase for
mannitol,10 mothbean ∆1-pyrroline-5-carboxylate synthetase for proline,11 Arthrobacter
globiformis choline dehydrogenase for glycine betaine,12 barley LEA protein13 and Nicotiana
plumbaginifolia detoxification enzyme.14 In all these experiments, a single gene for a protective
protein or an enzyme was overexpressed under the control of the CaMV 35S constitutive
promoter in transgenic plants, although a number of genes have been shown to function in
environmental stress tolerance and response. In the next step, I think it important to
manipulate several genes to achieve strong stress tolerance for the application of this
technology to the development of stress tolerant transgenic crops.
Overproduction of genes for stress-induced transcription factors in transgenic plants
activated the expression of many target genes involved in stress tolerance under unstressed
normal conditions and significantly improved stress tolerance to drought and freezing.15,16
These results suggest that regulatory genes involved in stress response can be also used for
the improvement of stress tolerance by gene transfer.
Regulation of Gene Expression by Drought

Complex Regulatory Systems for Gene Expression
The expression patterns of genes induced by drought were analyzed by RNA gel-blot
analysis. Results indicated broad variations in the timing of induction of these genes under
drought conditions. All the drought-inducible genes are induced by high salinity stress.
Most of the drought-inducible genes also respond to cold stress but some of them do not,
and vice versa. Many genes respond to ABA whereas some others do not.1,2,4 ABA-deficient
mutants were used to analyze drought-inducible genes that respond to ABA. Several genes
were induced by exogenous ABA treatment, but were also induced by cold or drought in
ABA-deficient (aba) or ABA-insensitive (abi) Arabidopsis mutants. These observations
indicate that these genes do not require an accumulation of endogenous ABA under cold
or drought conditions, but do respond to ABA. There are ABA-independent as well as
ABA-dependent regulatory systems of gene expression under drought stress. Analysis of
the expression of ABA-inducible genes showed that several genes require protein biosynthesis
for their induction by ABA, suggesting that at least two independent pathways exist
between the production of endogenous ABA and gene expression under stress conditions.
As shown in Figure 2.3, it is now hypothesized that at least four independent signal
transduction pathways function in the activation of stress-inducible genes under dehydration
conditions: Two are ABA-dependent (Pathways I and II) and two are ABA-independent
(Pathways III and IV).4 One of the ABA-dependent pathways requires protein biosynthesis
(Pathway I). Cis- and trans-acting factors involved in ABA-induced gene expression have
been extensively analyzed in one of the ABA-dependent pathway that does not require de
novo protein biosynthesis (Pathway II). One of the ABA-independent pathways overlaps
Fig. 2.3. Signal transduction pathways between the perception of drought stress signal and gene
expression. At least four signal transduction pathways exist (I-IV): Two are ABA-dependent (I
and II) and two are ABA-dependent pathways (III and IV). Protein biosynthesis is required in
one of the ABA-dependent pathways (I). In another ABA-dependent pathway, ABRE functions
as an ABA-responsive element and does not require protein biosynthesis (II). In one of the ABA-
independent pathways, DRE is involved in the regulation of genes not only by drought and salt
but also by cold stress (IV). Another ABA-independent pathway is controlled by drought and
salt, but not by cold (III).
with that of the cold response (Pathway IV). There are several drought-inducible genes
that do not respond to either cold or ABA treatment, which suggests that there is a fourth
pathway in the dehydration stress response (Pathway III). Recently, based on genetic analysis of
Arabidopsis mutants with the rd29A promoter—luciferase transgene, the existence of
drought-, salt- and cold- specific signaling pathways in stress-response was suggested, but
crosstalks between these signaling pathways were also observed (see chapter 1).
Major ABA-Independent Regulatory System of Gene Expression during

Drought and Cold Stress (Pathway IV): Important Roles of DRE/CRT Cis-Acting
Element and its DNA Binding Proteins
A number of genes are induced by drought, salt, and cold in aba (ABA-deficient) or
abi (ABA-insensitive) Arabidopsis mutants. This suggests that these genes do not require
ABA for their expression under cold or drought condition.2,4,17 Among these genes, the
expression of a drought-inducible gene for rd29A/lti78/cor78 was extensively analyzed.18
At least two separate regulatory systems function in gene expression during drought and
cold stress; one is ABA-independent (Fig. 2.3, Pathway IV) and the other is ABA-dependent
(Fig. 2.3, Pathway II). A 9bp conserved sequence, TACCGACAT, named the dehydration
responsive element (DRE), is essential for the regulation of the induction of rd29A under
Fig. 2.4. A model of the induction of the rd29A gene and cis- and trans-acting elements involved
in stress-responsive gene expression.15 Two cis-acting elements, DRE/CRT and ABRE, are
involved in the ABA-independent and ABA-responsive induction of rdnaA, respectively. Two
different DRE-binding proteins, DRBE1 and DREB2, separate two different signal transduction
pathways in response to cold and drought stresses, respectively. ABRE-binding proteins encode
bZIP-transcription factors.
drought, low-temperature, and high-salt stress conditions, but does not function as an
ABA-responsive element (Fig. 2.4). The rd29A promoter contains ABRE, which functions
in ABA-responsive expression. DRE-related motifs have been reported in the promoter
regions of many cold- and drought-inducible genes.3,4,17 These results suggest that DRE-
related motifs including C-repeat (CRT) and low temperature responsive element (LTRE),
which contain a CCGAC core motif, are involved in drought- and cold-responsive but
ABA-independent gene expression (see chapter 5).
Protein factor(s) that specifically interact with the 9bp DRE sequence were detected
in nuclear extract prepared from either dehydrated or untreated Arabidopsis plants.18
Recently, five independent cDNAs for DRE/CRT-binding proteins have been cloned using
the yeast one hybrid screening method.15,19 All the DRE/CRT binding proteins (DREBs
and CBFs) contain a conserved DNA binding motif that has also been reported in EREBP
and AP2 proteins (EREBP/AP2 motif) that are involved in ethylene-responsive gene
expression and floral morphogenesis, respectively. These five cDNA clones that encode
DRE/CRT binding proteins are classified into two groups, CBF1/DREB1 and DREB2.
Expression of the DREB1A gene and its two homologs (DREB1B = CBF1, DREB1C) was
induced by low-temperature stress, whereas expression of the DREB2A gene and its single
homolog (DREB2B) was induced by dehydration.15,64 Overexpression of the DREB1A cDNA
in transgenic Arabidopsis plants not only induced strong expression of the target genes
under unstressed conditions but also caused dwarfed phenotypes in the transgenic plants.
These DREB1A transgenic plants also revealed freezing and dehydration tolerance, which
was also shown in the CBF1 transgenics.16 In contrast, overexpression of the DREB2A
cDNA induced weak expression of the target genes under unstressed conditions and caused
slight growth retardation of the transgenic plants. These results indicate that two independent
families of DREB proteins, DREB1 and DREB2, function as transacting factors in two
separate signal transduction pathways under low-temperature and -dehydration conditions,
respectively (Fig. 2.4).15
Overproduction of the DREB1A and CBF1/DREB1B cDNAs driven by the 35S CaMV
promoter in transgenic plants significantly improved stress tolerance to drought and
freezing.15,16 However, the 35S-DREB1A transgenic plants revealed severe growth retardation
under normal growth conditions. The DREB1A cDNA driven by the stress-inducible rd29A
promoter was expressed at low level under unstressed control conditions and strongly
induced by dehydration, salt and cold stresses. The rd29A promoter minimized negative
effects on growth of plants, whereas the 35S-CaMV promoter caused severe growth
retardation under normal growth conditions.15, 20 Moreover, this stress-inducible promoter
enhanced tolerance to drought, salt and freezing at higher levels than that of the 35S-CaMV
promoter.
Drought-Specific ABA-Independent Regulatory System (Pathway III)

There are several drought-inducible genes that do not respond to either cold or ABA
treatment, which suggests the existence of another ABA-independent pathway in the
dehydration stress response (Fig. 2.3, Pathway III). These genes include rd19 and rd21 that
encode different thiol proteases, and erd1 that encodes a Clp protease regulatory subunit.21,22
The ERD1 protein is targeted to chloroplasts whereas the RD19 and RD21 proteins seem to
function in cytoplasm. The catalytic subunit of the Clp protease (Clp P) is encoded on the
chloroplast genome. The erd1 gene is not only induced by dehydration but also upregulated
during natural senescence and dark-induced senescence.23 The erd1 and rd21 genes were
also identified as senescence-associated genes.24 Promoter analysis of the erd1 gene in
transgenic plant indicates that erd1 promoter contains cis-acting element(s) involved in not
only ABA-independent stress responsive gene expression but also senescence-activated gene
expression.23 Further promoter analysis of these genes will give us more information on
Pathway III.
Major ABA-Dependent Regulatory System (Pathway II): Important Roles of

ABRE Cis-Acting Element and its bZIP DNA Binding Proteins
Most drought-inducible genes are upregulated by exogenous ABA treatment. The levels
of endogenous ABA increase significantly in many plants under drought and high salinity
conditions.1,2,4 In one of the ABA-dependent pathways (Fig. 2.3, Pathway II), drought-stress
inducible genes do not require protein biosynthesis for their expression. 4,5 These
dehydration-inducible genes contain potential ABA-responsive elements (ABREs;
PyACGTGGC) in their promoter regions. ABRE functions as a cis-acting DNA element
involved in ABA-regulated gene expression.5 ABRE was first identified in wheat Em and rice
rab genes, and its DNA-binding protein EmBP1 was shown to encode a bZIP protein. The
G-box resembles the ABRE motif and functions in the regulation of plant genes in a variety
of environmental conditions, such as red light, UV light, anaerobiosis, and wounding. cDNAs
for ABRE and G-box binding proteins have been isolated and shown to have a basic region
adjacent to a leucine zipper motif (bZIP) and constitute a large gene family. Nucleotides
around the ACGT core motif have been shown to be involved in determining the binding
specificity of bZIP proteins. Furthermore, a coupling element (CE) is required to specify
the function of the ABRE, constituting an ABA-responsive complex in the regulation of the
HVA22 gene.25 However, it has not been resolved how ABA activates bZIP proteins to binds
to ABRE and initiate transcription of ABA-inducible genes. Further studies are necessary
for the precise understanding of the molecular mechanisms of ABA-responsive gene
expression that requires ABRE as a cis-acting element.
Several bZIP transcription factors from rice, maize and Arabidopsis plants respond to
cold, dehydration, and to exogenous ABA treatment.4 These bZIP proteins bind to G-box-
like sequences. These results suggest that ABA-inducible bZIP proteins are also involved in
one of the ABA-dependent pathways (Fig. 2.3., Pathway I) or in the enhancement of the
ABA-dependent gene expression (Fig. 2.3., Pathway II).
There are several cis-acting elements other than ABRE that function in ABA-responsive
gene expression, not only under drought conditions but also in seed desiccation. The Sph
box and GTGTC motifs regulate ABA- and VP1-dependent expression of the maize C1
gene, whose product is a MYB-related transcription factor and functions as a controlling
element in anthocyanin biosynthesis during seed.26 VP1 encodes a transcriptional activator
and is thought to cooperate with bZIP proteins. Arabidopsis ABI3 has sequence and
functional similarity with maize VP1. Recently VP1 was demonstrated to have a DNA-
binding activity to Sph box. EmBP1 and VP1 were shown to interact with 14-3-3 proteins
and form a transcription complex.27 This complex also interacted with ABRE on the Em
promoter. A similar system is thought to function in ABA-responsive gene expression in
drought stress response as well as in seed maturation.
Roles of MYC and MYB Homologs in ABA-Dependent Gene Expression

that Requires Protein Biosynthesis (Pathway I)
Biosynthesis of novel protein factors is necessary for the expression of ABA-inducible
genes in one of the two ABA-dependent pathways (Fig. 2.3, Pathway I). The induction of an
Arabidopsis drought-inducible gene, rd22, is mediated by ABA, and requires protein
biosynthesis for its ABA-dependent expression.28 A 67bp region of the rd22 promoter is
essential for this ABA-responsive expression, and contains several conserved motifs of DNA-
binding proteins, two MYC and one MYB recognition sequences, but this region has no
ABREs. First MYC and MYB recognition sequences are essential for the ABA- and drought-
responsive expression of the rd22 gene.29 A cDNA for a transcription factor MYC homo-
logue, named rd22BP1, was cloned by the DNA-ligand binding method, using the 67bp
DNA as a probe. The rd22BP1 gene is induced by drought and salt stress. These results
suggest that a drought- and salt-inducible MYC homologue function in the ABA-inducible
expression of rd22 (Fig. 2.5). The ATMBY2 gene that encodes a MYB-related protein is
induced by dehydration and ABA treatment.30 Recombinant ATMYB2 protein binds to the
MYB recognition sequence in the 67bp region of the rd22 promoter. Moreover, these MYC
and MYB proteins transactivate the rd22 promoter GUS fusion gene in transient expression
system using leaf protoplasts.29 Therefore, the ATMYB2 protein might also cooperatively
function with the rd22BP1 protein as a transcription factor that controls the ABA-dependent
expression of the rd22 gene (Fig. 2.5).
Many stress- and ABA-inducible genes encoding various transcription factors have
now been reported. These contain conserved DNA binding motifs, such as MYB, MYC,
bZIP and zinc finger. These transcription factors are thought to function in the regulation
of ABA inducible genes, which respond to drought stress rather slowly after the production of
ABA-inducible transcription factors (Fig. 2.3., Pathway I).
Signal Perception and Signal Transduction in Drought Stress

Response
Complex Signal Transduction Pathways
Signal-transduction pathways, from the sensing of dehydration or osmotic change to
the expression of various genes, and the signaling, molecules that function in stress signaling
have not been extensively studied in plants. Signal transduction pathways in drought stress
response have been studied in yeast and animal systems (Fig. 2.6). Two component
systems function in sensing osmotic stress in bacteria and yeast. Plants as well as
cyanobacteria contain many genes encoding sensor histidine kinases and response regulator
Fig. 2.5. A model of the ABA-independent induction of the rd22 gene by drought stress. Drought
stress triggers the biosynthesis of ABA, which induces the expression of two genes for transcription
factors, MYC (rd22BP1) and MYB (ATw2) homologues. These transcription factors then
activate the expression of the rd22 gene.
homologues, which suggests the involvement of similar osmosensing mechanisms in higher

plants. Of course, other sensing mechanisms may function during drought stress responses,
such as mechanical sensors in cytoskeltons and sensors for superoxides produced by stress.
Stomatal closure is well characterized as a model system in the responses of plant cells to
dehydration stress and ABA treatment.4,6 During stomatal closure, the level of cytoplasmic
Ca2+ is increased, which suggests that Ca2+ functions as a second messenger in the osmotic
stress response. In animal cells, inositol-3-phosphate (IP3) is involved in the release of
Ca2+ into the cytoplasm from intracellular stores, and it may play a similar role in plant
cells. Ca2+ and IP3 are the most probable candidates as second messengers in drought-
stress responses in plant cells.31
ABA plays important roles in drought stress responses. ABA is involved in not only
stomatal closure but also induction of many genes.6 Several mutants in ABA signaling
have been identified and their genes encode protein phosphatases and farnesyl transferase.
These suggest that protein dephosphorylation and protein farnesylation are involved in
ABA signaling. However, various signaling molecules seem to be involved in ABA signal-
ing, such as phosphatidic acid and cyclic ADP ribose. Various protein kinases and enzymes
involved in phospholipid metabolism have been reported in plants and are thought to func-
tion in signal-transduction pathways including drought-stress and ABA responses (Fig. 2.6).4
MAP kinase cascades and calcium-dependent protein kinase were suggested to be involved
in drought stress response and ABA signaling. Complex signaling cascades are thought to
function in molecular responses to drought stress. Molecular analysis of the signaling process is
in progress based on genetics and gene cloning. In this chapter we will describe recent progress
of signal transduction cascades from sensing of dehydration stress to gene expression.
Fig. 2.6. Second messengers and factors

involved in the signal perception and the
signal transduction in drought stress
response. Two-component histidine
kinase is thought to function as an
osmosensor in plants. Ca2+ and IP3 are
most probable second messengers of the
dehydration signal. Phosphorylation
functions in water stress and ABA signal
transduction pathways. PI turnover is
also involved in drought stress response.
ABA plays inportant roles in the regulation
of gene expression and in physiological
responses during water stress. Several
ABA signal transduction pathways are
reported.
Sensing of Osmotic stress: The Two-Component System

In bacteria, the two-component system functions in sensing and response to osmotic
stress.32 The two-component system is composed of two types of proteins, a sensory histidine
kinase and a response regulator. The osmosensing signaling pathway in Escherichia coli is
composed of one of the two-component system, EnvZ and OmpR. The EnvZ protein
functions as an osmosensor, transmits a signal to the histidine kinase domain and activates the
kinase. The activated histidine kinase autophosphorylates the histidine residue. The phosphate
on the histidine is then transferred to an aspartate residue in the receiver domain of OmpR,
and activates the OmpR transcription factor. The activated OmpR regulates the transcription
of the OmpF and OmpC genes. OmpF and OmpC are porin proteins and form different
pores in the outer membrane. These two porin proteins control cellular osmotic pressure
in E. coli. Cyanobacteria also contain numbers of two-component systems, one of which is
thought to be involved in osmosensing.
In yeast, exposure to high osmolarity activates a MAPK cascade that includes Ssk2/
Ssk22 (MAPKKK), Pbs2 (MAPKK) and Hog1 (MAPK), and then activates several genes
involved in the biosynthesis of glycerol, which is an important osmoprotectant (Fig. 2.7).32
Three gene products (Sln1, Ypd1, and Ssk1) that act in an early phase of the hyperosmolarity
stress response encode signaling molecules that constitute a prokaryote-type two-component
regulatory system.33 Sln1 is thought to act as a sensor protein, phosphorylating Ypd1 and
Ssk1 response regulator proteins under conditions of high osmolarity. The three protein
factors perform a four step phosphorelay (His-Asp-His-Asp). At high osmolarity,
unphosphorylated Ssk1 activates Ssk2 or Ssk22 (MAPKKKs), which results in the activation of
Pbs2 (MAPKK) by Ser/Thr phosphorylation.34 Then, phosphorylated Pbs2 activates Hog1
(MAPK) by Thr/Tyr phosphorylation (Fig. 2.7).
Recently, we isolated an Arabidopsis cDNA (ATHK1) encoding the two-component
histidine kinase, a yeast osmosensor Sln1 homologue, by PCR. ATHK1 has a typical
histidine kinase domain and a receiver domain like Sln1, and has a different structure in
Fig. 2.7. Possible roles of the two-component

system and MAP kinase cascades in plants in
comparison with yeast osmosensing system. In
yeast, Sln1p is thought to act as a sensor protein
phosphorylating Ypd1 and Ssk1p, response
regulator proteins under conditions of high
osmolarity. A MAP kinase cascade (Ssk2p/
Ssk22p-Pbs2p-Hog1p) functions downstream
of the two-component system in osmotic stress
response. In Arabidopsis ATHK1, a Sln1p
homologue, functions the same as an
osmosensor in yeast, which suggests that a
similar two-component system and MAP
kinase cascade are involved in drought stress
signal transduction (see text).
the N-terminal domain from that of ETR1, an ethylene receptor (Fig. 2.7, Urao, Yamaguchi-
Shinozaki, Hirayama and Shinozaki, submitted). Overexpression of cATHK1 suppressed
the lethality of a temperature-sensitive osmosensing-defective yeast sln1 mutant (sln1-ts).
By contrast, cATHK1, with substitution of conserved His or Asp residues, failed to complement
the sln1-ts mutant, indicating that ATHK1 functions as a histidine kinase. Introduction of
cATHK1 into a yeast mutant lacking two osmosensors (sln1∆sho1∆) suppressed its lethality in
high salinity media, and activated the HOG1 MAP kinase. The ATHK1 transcript was more
abundant in roots than other tissues under normal growth conditions and accumulated in
conditions of high salinity and low temperature. Histochemical analysis of β-glucuronidase
(GUS) activities driven by the ATHK1 promoter further indicates that the ATHK1 gene is
responsive to changes in external osmolarity at the transcriptional level. These results
suggest that ATHK1 might function in signal perception during drought stress in Arabidopsis
(Fig. 2.7). A similar osmosensing mechanism might operate in higher plants in response
to a water deficit. In higher plants, a two-component histidine kinase, ETR1, is a receptor
in ethylene signal transduction,35 and another two-component histidine kinase, CKI1, is
involved in cytokinin signaling.36 Two-component histidine kinases may function as sensors
or receptors in various signal transduction pathways in plants.
In Saccharomyces cerevisiae, Sln1 acts as an osmosensory protein, sequentially phos-
phorylating a phosphorelay intermediator, Ypd1, and a response regulator, Ssk1, under
conditions of normal osmolarity.32 These three protein factors perform a four-step
phosphorelay (His-Asp-His-Asp). Recently, we have cloned four cDNAs encoding a
response regulator,37 and three cDNAs encoding a phosphorelay intermediator containing
an HPt domain65 from Arabidopsis. The existence of response regulators was also reported
in Arabidopsis and maize.38, 39 These observations suggest that plants have an osmosensing
and signaling system similar to that of yeast (Fig. 2.7).
Roles of MAP Kinase Cascades in Stress Response

Mitogen-activated protein kinase, or MAP kinase (MAPK), is involved in the signal
transduction pathways associated not only with growth factor-dependent cell proliferation but
also with environmental stress responses in yeast and animals. Many cDNAs for proteins
involved in MAPK cascade—MAPK, MAPKK, MAPKKK—have been cloned in plants (Fig.
2.7). We have isolated more than nine genes for MAPK in Arabidopsis. There are at least
four subfamilies of MAPK based on phylogenetic analysis.7 One of the Arabidopsis MAPK
genes, ATMPK3, is induced at the mRNA level by drought, low temperature, high salinity
and touch.40 Moreover, two genes for protein kinases involved in the MAPK cascade,
MAPKKK (ATMEKK1) and ribosomal S6 kinase (RSK; ATPK19), are induced by similar
stresses. We demonstrated rapid and transient activation of MAP kinase activities of
ATMPK4 in Arabidopsis plants by low temperature, humidity change, wounding and touch
(Ichimura, Mizoguchi, Shinozaki, unpublished data). Recently, alfalfa MAPK, MMK4, was
demonstrated to be activated at posttranslational levels by a variety of stresses including
drought, low temperature and mechanical stimuli.41 The MMK4 gene is also induced by
these stresses at the transcriptional level. These observations indicate that certain MAP
kinase cascades might function in the signal transduction pathways in drought stress
response (Fig. 2.5). Recently, a protein phosphatase, 2C, has been suggested to function as
a negative regulator of MAP kinase pathways in plants.42 Their roles in drought stress
response have not yet been elucidated.
Roles of Phospholipids Metabolism and Calcium in Drought Stress Signaling

In animal systems, a variety of phosphoinositides that are metabolrites of membrane
lipids function as second messengers in various signaling processes. Phospholipase C (PLC)
digests phosphatidyl-inositol-4, 5-bisphosphate (PIP2) to generate two second messengers,
inositol-1,4,5-triphosphate (IP3) and diacylglycerol (DG). IP3 induces the release of Ca2+
into the cytoplasm, which in turn causes various responses in the cytoplasm. DG and PIP2
also function as second messengers and control various cellular responses. In plants, similar
systems are thought to function in drought stress response.43 A gene for phospholipase C,
AtPLC1s, is rapidly induced by drought and salt stresses in Arabidopsis.44 Recently, a gene
for phosphatidylinositol-4-phosphate-5-kinase (PIP5K) was demonstrated to be induced
by drought stress like the AtPLC1s gene.45 In animals, PIP5K catalyzes the production of
phosphatidylinositol-4,5-bisphosphate (PIP2) from phosphatidylinositol-4-phosphate
(PIP). The stress-inducible PLC and PIP5K might function in the signal-transduction
cascade under drought stress (Fig. 2.8).
Two genes for calcium-dependent protein kinases (CDPK), ATCDPK1 and ATCDPK2,
were induced by drought and high salt.46 These CDPKs may be activated by calcium in
response to osmotic stress (Fig. 2.8). Recently, coexpression of the constitutively active
catalytic domain of a stress-inducible CDPK, ATCDPK1, has been demonstrated to
induce the expression of an ABA-inducible HVA1 promoter-reporter fusion gene in maize
protoplasts.47 The HVA1 promoter is also activated not only by cold, high salt and ABA
treatment but also by calcium in protoplasts. These observations also support the idea that
Ca2+ might function as a second messenger and that ATCDPK1 functions as a positive
regulator in the ABA signal-transduction pathways under drought-stress conditions in
plants.
Recently we isolated two genes for diacylglycerol kinase (DGK) and phosphatidic phosphatase
(PAP) that are involved in PI turnover (Fig. 2.8).48 These two genes were demonstrated to be
upregulated by dehydration (Katagiri, Shinozaki, unpublished observation). PAP synthesizes
diacyglycerol (DG) by dephosphorylating phosphatidic acid (PA). DGK converts DG into PA.
PA is produced from phosphatidylcholine (PC) by phospholipase D (PLD), whereas DG is
Fig. 2.8. Possible roles of phospholipids metabolism in drought stress and ABA signal transduc-
tion pathways. Many genes involved in phospholipids metabolism (PI turnover and PC turn-
over), such as PI-PLC, PIP5K DGK and PAP, have been identified and shown to be upregulated
by drought and salt stress. IP3, PIP2, DG and PA are functions as second messengers. IP3 is involved
in release of calcium into cytoplasm to activate calcium dependent signaling molecules
including CDPK.
produced from PIP2 by PLC. However, the roles of DG and PA as second messengers have not
been extensively studied in plants. Recently, PLD was shown to be activated by ABA treatment. PA
also increased transiently and is involved in triggering the subsequent ABA response of
aleurone cells.49 These results also support that phospholipids metabolism including IP3 and
PA as second messengers is involved in drought stress response.
ABA Signal Transduction

The role of ABA in drought-stress signal transduction has been analyzed genetically
with ABA-insensitive mutants in various species. Maize vp1 and Arabidopsis abi1, abi2, abi3,
abi4 and era have been extensively characterized and their genes have been cloned (see chapter
1).50 Among them, ABI1 and ABI2 gene products function mainly in vegetative tissues, and
also participate to some extent in seed development. Because of the wilty phenotype of abi1
and abi2 mutants, ABI1 and ABI2 are thought to have important roles in ABA-dependent
signal-transduction pathways during drought stress. The ABI1 and ABI2 genes have been
cloned and shown to encode proteins that are related to type 2C protein serine/threonine
phosphatases (PP2Cs).51, 52, 53 The ABI1 gene product functions in stomatal closure, and the
abi1 plant reveals the wilty phenotype.54 ABI1 was demonstrated to function as a negative
regulator in ABA dependent gene expression in a transient expression experiment using
maize protoplasts.47 By contrast, the dehydration-inducible ATCDPK1, encoding calcium-
dependent protein kinase functions as a positive regulator. These results indicate that a
protein phosphorylation and dephosphorylation process might be involved in ABA-
responsive signaling during water deficit. ABI3 and ABI4 were shown to encode transcription
factors. Another Arabidopsis mutant, era, that confers an enhanced response to exogenous
ABA, has mutations in the ERA1 gene encoding the β subunit of farnesyl transferase.55 This
suggests that a negative regulator of ABA sensitivity may requires farnesylation to function.
Several different signal transduction pathways are suggested to be involved in ABA
response. ABA was demonstrated to induce a rapid and transient activation of MAPK in
barley aleurone protoplasts.56 Correlation between ABA-induced MAPK activation and
ABA-induced gene expression implicates that MAPK might be involved in ABA signal
transduction (Fig. 2.3). Recently, cyclic ADP ribose (cADPR) is shown to be involved in
ABA signal transduction as a second messenger and activate ABA responsive gene expression.57
cADPR is thought to function in the release of Ca2+ in plants. Several reports suggest that
specific protein kinases, including MAP kinase, are activated in response to ABA treatment in
aleulone cells, guard cells and cultured cells.7 Recently, ABA was shown to activate the
enzyme PLD to produce PA, which is involved in triggering the subsequent ABA responses
of barely aleulone cells.47 These suggest the existence of several different signal transduction
cascades in ABA response. Identification of ABA receptor(s) is important to further
understanding of the signaling process and to understand which pathways are essential in
ABA signaling.
Regulation of ABA Biosynthesis

ABA is produced under drought stress conditions de novo, which requires protein
biosynthesis. As mentioned above, this process is important for drought-inducible gene
expression. ABA is thought to be synthesized from xanthophylls via violaxanthin, xanthoxin
and ABA-aldehyde (C40 pathway) and the conversion of violaxanthin to xanthoxin is rate
limiting in the ABA biosynthesis under drought stress.58 The C15 pathway found in fungi
may also function in plants. Genetic evidence supports only the C40 pathway, and bio-
chemical studies suggest that the cleavage of 9-cis-xanthophylls is the key regulatory step.
Many ABA-deficient mutants that do not produce ABA have been isolated in various
plants. An ABA-deficient tobacco mutant, aba2, was isolated by transposon-tagging using
the maize Ac transposon.59 ABA2 cDNA encodes a chloroplast-imported protein that
exhibits zeaxanthin epoxidase activity, which functions in the first step of the ABA biosyn-
thesis pathway. The tobacco ABA2 gene corresponds to the Arabidopsis ABA1 gene. Two
steps in the conversion of xanthoxin to ABA-aldehyde and oxidation of ABA-aldehyde to
ABA are defined by Arabidopsis aba2 and aba3 mutants, respectively.
An ABA-deficient viviparous mutant of maize, vp14, has been isolated and its
corresponding gene has been cloned by transposon mutagenesis.60 VP14 encodes
neoxanthin cleavage enzymes which are involved in the production of xanthoxin from
9-cis-xanthophyll in ABA biosynthesis. The VP14 gene is expressed in embryo and roots
and is induced in leaves by drought stress.61 There are several VP14 related genes in maize.
We also isolated a VP14 homologue as a drought-inducible gene in cowpea (Iuchi,
Yamaguchi-Shinozaki, Shinozaki, unpublished data). These genes are likely to play key
roles in the ABA synthesis in seed development and stress response. Promoter analysis of
the VP14 gene will give us the precise molecular mechanism of the regulation of ABA
biosynthesis under stress conditions.
Conclusion and Perspectives

Molecular mechanisms of drought stress response and tolerance have been actively
studied over the past ten years. Many genes that are regulated by drought stress have been
reported in a variety of plants. Analyses of stress-inducible gene expression have revealed
the presence of multiple signal transduction pathways between the perception of drought
stress signal and gene expression. At least four different transcription factors have been
suggested to function in the regulation of dehydration-inducible genes; two are ABA-re-
sponsive and two are ABA-independent. This variety explains the complex stress response
observed after exposure of plants to drought stress. Genetic analysis of Arabidopsis mutants
with the rd29A promoter—luciferase transgene also suggests complex signaling pathways
in drought, salt and cold stress responses (see chapter 1).62 Some genes are rapidly induced
by drought stress in 10 minutes, whereas others are slowly induced in a few hours after the
accumulation of endogenous ABA. Several genes for various transcription factors are
induced by drought stress and ABA at transcriptional levels, which might be involved in the
regulation of slowly expressed genes whose products function in stress tolerance and
adaptation. In addition, many genes for factors involved in the signal transduction cascades,
such as protein kinases and enzymes involved in PI turnover, are upregulated by a drought
stress signal.4,7 These signaling factors might be involved in the amplification of the stress
signals and the adaptation of plant cells to drought stress conditions. Molecules that function
as osmosensors and ABA-receptors have not been identified. Based on the knowledge of
osmosensors in yeast and bacteria, cloning of homologues of the two-component histidine
kinase as an osmosensor is in progress in higher plants. Molecular analyses of these factors
should provide a better understanding of the signal transduction cascades during drought
stress. Transgenic plants that modify the expression of these genes will give more information
on the function of their gene products.
Sequencing of the Arabidopsis genome is now in progress and will be completed by
the year 2004, which means that structures of all 20,000 Arabidopsis genes will be determined in
a few years.63 All the stress-inducible genes will be identified by systematic analysis of gene
expression. In the next decade, it will be important to develop novel methods to analyze
complex networks of stress responses of higher plants. We are now constructing insertion
mutant lines using T-DNA and transposons to analyze functions of disrupted genes in
plants. Reverse genetics approaches as well as classical genetics will become more important to
understanding not only functions of stress-inducible genes but also the complex signaling
process in environmental stress responses. An efficient gene disruption method as well as
transgenic approaches using antisense or sense constructs will also contribute to more
precise understanding of the molecular mechanisms of stress response.
Acknowledgments
This work was supported by the Program for Promotion of Basic Research Activities
for Innovative Bioscience, the Special Coordination Fund of the Science and Technology
Agency of the Japanese Government, a Grant-in-Aid from the Ministry of Education,
Science and Culture of Japan, and The Human Frontier Science Program.
References
1. Ingram J, Bartels D. The molecular basis of dehydration tolerance in plants. Annu. Rev.
Plant Physiol. Plant Mol Biol 1996; 47:377-403.
2. Bray EA. Plant responses to water deficit. Trends Plant Sci 1997; 2:48-54.
3. Shinozaki K, Yamaguchi-Shinozaki K. Molecular responses to drought and cold stress.

Current Opinion in Biotech 1996; 7:61-167.
4. Shinozaki K, Yamaguchi-Shinozaki K. Gene expression and signal transduction in water
5. Giraudat J, Parcy F, Bertauche N, Gosti F, Leung J, Morris P-C, Bouvier-Durand M,
Vartanian N. Current advances in abscisic acid action and signaling. Plant Mol Biol 1994;
26:1557-1577.
6. Bohnert HJ, Nelson DE, Jensen RG. Adaptations to environmental stresses. Plant Cell 1995;
7:1099-1111.
7. Mizoguchi T, Ichimura K, Shinozaki K. Environmental stress response in plants: The role
of mitogen-activated protein kinases (MAPKs). Trends Biotech 1997; 15:15-19.
8. Xu Q, Fu H-H, Gupta R, Luan S. Molecular characterization of a tyrosine-specific protein
phosphatase encoded by a stress-responsive gene in Arabidopsis. Plant Cell 1998; 10:849-857.
9. Holmberg N, Bülow L. Improving stress tolerance in plants by gene transfer. Trends Plant
Sci 1998; 3:61-66.
10. Tarczynski M, Bohnert H. Stress protection of transgenic tobacco by production of the
osmolyte mannitol. Science. 1993; 259:508-510.
11. Kavi Kishor PB, Hong Z, Miao G-U, Hu C-AH, Verma DPS. Overexpression of ∆1-
pyrroline-5-carboxylate synthetase increases proline production and confers osmotolerance
in transgenic plants. Plant Physiol. 1995; 108:1387-1394.
12. Hayashi H, Mustardy L, Deshnium P, Ida M, Murata N. Transformation of Arabidopsis
thaliana with the codA gene for choline oxidase; accumulation of glycine betaine and
enhanced tolerance to salt and cold stress. Plant J. 1997; 12:1334-142.
13. Xu D, Duan X, Wang B, Hong B, Ho T-HD, Wu R. Expression of a late embryogenesis
abundant protein gene, HVA1, from barley confers tolerance to water deficit and salt stress
in transgenic rice. Plant Physiol. 1996; 110:249-257.
14. McKersie BD, Bowley SR, Harjanto E, Leprince O. Water-deficit tolerance and field
performance of transgenic alfalfa overexpressing superoxide dismutase. Plant Physiol. 1996;
111:1177-1181.
15. Liu Q, Sakuma Y, Abe H, Kasuga M, Miura S, Yamaguchi-Shinozaki K, Shinozaki K. Two
transcription factors, DREB1 and DREB2, with an EREBP/AP2 DNA binding domain,
separate two cellular signal transduction pathways in drought and low temperature-
responsive gene expression, respectively, in Arabidopsis. Plant Cell 1998; 10:1391-1406.
16. Jaglo-Ottosen KR, Gilmour SJ, Zarka DG, Schabenberger O, Thomashow MF. Arabidopsis
CBF1 overexpression induces coe genes and enhances freezing tolerance. Science 1998;
280:104-106.
17. Thomashow MF. Arabidopsis thaliana as a model for studying mechanisms of plant cold
tolerance. In E. Meyrowitz and C. Sommerville, eds., Arabidopsis. 1994; Cold Spring
Harbor Laboratory Press, Cold Spring Harbor, pp. 807-834.
18. Yamaguchi-Shinozaki K, Shinozaki K. A novel cis-acting element in an Arabidopsis gene is
involved in responsiveness to drought, low-temperature, or high-salt stress. Plant Cell 1994;
6:251-264.
19. Stockinger EJ, Glimour SJ, Thomashow MF. Arabidopsis thaliana CBF1 encodes an AP2
domain-containing transcription activator that binds to the C-repeat/DRE, a cis-acting DNA
regulatory element that stimulates transcription in response to low temperature and water
deficit. Proc. Natl. Acad. Sci. USA 1997; 94:1035-1040.
20. Kasuga M, Liu Q, Miura S, Yamaguchi-Shinozaki K, Shinozaki K. Improving drought, salt,
freezing tolerance by gene transfer of a single stress-inducible transcription factor.
Submitted.
21. Koizumi M, Yamaguchi-Shinozaki K, Tsuji H, Shinozaki K. Structure and expression of
two genes that encode distinct drought-inducible cysteine proteases in Arabidopsis thaliana.
Gene 1993; 129:175-182.
22. Kiyosue T, Yamaguchi-Shinozaki K, Shinozaki K. Characterization of cDNA for a
dehydration-inducible gene that encodes a Clp A, B like protein in Arabidopsis thaliana.
Biochem Biophys Res Comm. 1993; 196:1214-1220.
23. Nakashima K, Kiyosue T, Yamaguchi-Shinozaki K, Shinozaki K. A nuclear gene, erd1,

encoding a chloroplast-targeted Clp protease regulatory subunit homolog is not only
induced by water stress but also developmentally upregulated during senescence in
Arabidopsis thaliana. Plant J. 1997; 12: 851-861.
24. Weaver LM, Gan S, Quirino B, Amasino M. A comparison of the expression of the pattern
of several senescence-associated genes in response to stress and hormone treatment. Plant
Mol Biol 1998: 37:455-469.
25. Shen Q, Ho THD. Functional dissection of an abscisic acid (ABA)-inducible gene reveals
two independent ABA-responsive complexes each containing a G-box and a novel cis-
acting element. Plant Cell 1995; 7:295-307.
26. McCarty DR. Genetic control and integration of maturation and germination pathways in
seed development. Annu Rev Plant Physiol Plant Mol Biol. 1995; 46:71-93.
27. Schultz TF, Medina J, Hill A, Quatrano RS. 14-3-3 proteins are part of an abscisic acid-
VIVIPAROUS1 (VP1) response complex in the Em promoter and interact with VP1 and
EmBP1. Plant Cell 1998; 10:837-847.
28. Yamaguchi-Shinozaki K, Shinozaki K. The plant hormone abscisic acid mediates the
drought-induced expression but not the seed-specific expression of rd22, a gene responsive
to dehydration stress in Arabidopsis thaliana. Mol Gen Genet. 1993; 238:17-25.
29. Abe H, Yamaguchi-Shinozaki K, Urao T, Iwasaki T, Hosokawa D, Shinozaki K. Role of
Arabidopsis MYC and MYB homologs in drought- and abscisic acid-regulated gene
expression. Plant Cell 1997; 9:1859-1868.
30. Urao T, Yamaguchi-Shinozaki K, Urao S, Shinozaki K. An Arabidopsis myb homologue is
induced by dehydration stress and its gene product binds to the conserved MYB recognition
sequence. Plant Cell 1993; 5:1529-1539.
31. McAinsh MR, Hetherington AM. Encoding specificity in Ca2+ signaling systems. Trends
Plant Sci. 1998; 3:32-36.
32. Wurgler-Murphy SM, Saito S. Two-component signal transducers and MAPK cascades.
Trends Biochem Sci. 1997; 22:172-176.
33. Posas F, Wurgler-Murphy SM, Maeda T, Witten TC, Thai TC, Saito H. Yeast HOG1 MAP
kinase cascade is regulated by a multistep phosphorelay mechanism in SLN1-YPD1-SSK1
'two-component' osmosensor. Cell 1996; 86:865-875.
34. Maeda T, Takehara M, Saito H. Activation of yeast PBS2 MAPKK by MAPKKKs or by
binding of an SH3-containing osmosensor. Science 1995; 269:554-558.
35. Chang C. The ethylene signal transduction pathway in Arabidopsis: An emerging paradigm?
Trends Biol Sci. 1996; 21:129-133.
36. Kakimoto T. CKI1, a histidine kinase homologue implicated in cytokinin signal trans-
duction. Science 1996; 274:982-985.
37. Urao T, Yakubov B, Yamaguchi-Shinozaki K, Shinozaki K. Stress-responsive expression of
genes for two-component response regulator-like proteins in Arabidopsis thaliana. FEBS
Lett. 1998; 427:175-178.
38. Imamura A, Hanaki N, Umeda H, Nakamura A, Suzuki T, Ueguchi C, Mizuno T.
Response regulators implicated in His-to-Asp phosphor transfer signaling in Arabidopsis.
Proc Natl Acad Sci USA. 1998; 95:2691-2696.
39. Sakakibara H, Suzuki M, Takei K, Deji A, Taniguchi M, Sugiyama T. A response-regulator
homologue possibly involved in nitrogen signal transduction mediated by cytokinin in
maize. Plant J 1998; 14:337-344.
40. Mizoguchi T, Irie K, Hirayama T, Hayashida N, Yamaguchi-Shinozaki K, Matsumoto K,
Shinozaki K. A gene encoding a MAP kinase is induced simultaneously with genes for a
MAP kinase and an S6 kinase by touch, cold and water stress in Arabidopsis thaliana. Proc.
Nalt. Acad. Sci. USA 1996; 93:765-769.
41. Jonak C, Kiegerl M, Ligterink W, Barker PJ, Huskisson NS, Hirt H. Stress signaling in
plants: A MAP kinase pathway is activated by cold and drought. Proc. Natl. Acad. Sci.
USA 1996; 93:11274-11279.
42. Meskiene I, Boerge L, Glaser W, Balog J, Brandstoetter M, Zwerger K, Ammerer G, Hirt

H. MP2C, a plant protein phosphatase 2C, functions as a negative regulator of mitogen-
activated protein kinase pathways in yeasts and plants. Proc Natl Acad Sci USA. 1998;
95:1938-1943.
43. Munnik T, Irvine RF, Musgrave A. Phospholipid signaling in plants. Biochem Biophys Acta
1998; 1389:222-272.
44. Hirayama T, Ohto C, Mizoguchi T, Shinozaki K. A gene encoding a phosphatidylinositol-
specific phospholipase C is induced by dehydration and salt stress in Arabidopsis thaliana.
Proc. Nalt. Acad. Sci. USA 1995; 92:3903-3907.
45. Mikami K, Katagiri T, Iuchi S, Yamaguchi-Shinozaki K, Shinozaki K. A gene encoding
phoaphatidylinositol-4-phosphate 5-kinase is induced by water stress and abscisic acid in
Arabidopsis thaliana. Plant J 1998; in press.
46. Urao T, Katagiri T, Mizoguchi T, Yamaguchi-Shinozaki K, Hayashida N, Shinozaki K. Two
genes that encode Ca2+-dependent protein kinases are induced by drought and high-salt
stresses in Arabidopsis thaliana. Mol Gen Genet 1994; 224:331-340.
47. Sheen J. Ca2+-dependent protein kinase and stress signal transduction in plants. Science
1996; 274:1900-1902.
48. Katagiri T, Mizoguchi T, Shinozaki K. Molecular cloning of a cDNA encoding
diacylglycerol kinase (DGK) in Arabidopsis thaliana. Plant Mol Biol 1996; 30:647-653.
49. Ritchie S, Gilroy S. Abscisic acid signal transduction in the barley aleurone is mediated
by phospholipase D activity. Proc Natl Acad Sci USA 1998; 95:2697-2702.
50. Bonetta D, McCourt P. Genetic analysis of ABA signal transduction pathways. Trends
Plant Sci 1998; 6:231-235.
51. Leung J, Bouvier-Durand M, Moris PC, Guerrier D, Chefdor F, Giraudat J. Arabidopsis
ABA response gene ABI1: Features of a calcium-modulated protein phosphatase. Science
1994; 264:1448-1452.
in Arabidopsis thaliana. Science 1994; 264:1452-1455.
53. Leung J, Merlot S, Giraudat J. The Arabidopsis ABSCISIC ACID-INSENSITIV2 (ABI2)
and ABI1 genes encode homologous protein phosphatase 2C involved in abscisic acid
signal transduction. Plant Cell 1997; 9:759-771.
54. Armstrong F, Leung J, Grabov A, Brearley J, Giraudat J, Blatt MR. Sensitivity to abscisic
acid of guard-cell K+ channels is suppressed by abi1-1, a mutant Arabidopsis gene encoding a
putative protein phosphatase. Proc Natl Acad Sci USA 1995; 92:9520-9524.
55. Culter S, Ghassemian, Bonetta D, Cooney S, McCourt P. A protein farnesyl transferase
involved in abscisic acid signal transduction in Arabidopsis. Science 1996; 273:1239-1241.
56. Knetsch MLW, Wang M, Snaar-Jagalska BE, Heimovaara-Dijkstra S. Abscisic acid
induces mitogen-activated protein kinase activation in barley aleurone protoplasts. Plant Cell
1996; 8:1061-1067.
57. Wu Y, Kuzma J, Maréchal E, Graeff R, Lee CH, Foster R, Chua N-H. Abscisic acid
signaling through cyclic ADP-ribose in plants. Science 1997; 278:2126-2130.
58. Kende H, Zeevaart JAD. The five "classical" plant hormones. Plant Cell. 1997; 9:1197-1210.
59. Marin E, Nussaume L, Quesada A, Gonneau M, Sotta B, Hugueney P, Frey A, Marion-Poll
A. Molecular identification of zeaxanthin epoxidase of Nicotiana plumbaginifolia, a gene
involved in abscisic acid biosynthesis and corresponding to the ABA locus of Arabidopsis
thaliana. EMBO J 1996; 15:2331-2342.
60. Schwartz SH, Tan BC, Gage DA, Zeebaart JAD, McCarty DR. Science 1997; 276:1872-1874.
61. Tan BC, Schwartz SH, Zeevaart JAD, McCarty DR. Genetic control of abscisic acid biosyn-
thesis in maize. Proc. Natl. Acad. Sci. USA 1997; 94:12235-12240.
62. Ishitani M, Xiong L, Stevenson B, Zhu JK. Genetic analysis of osmotic and cold stress signal
63. Bevan M, Ecker J, Theologis S, Federspiel N, Davis R, McCombie D, Martienssen R, Chen E,
Waterston B, Wilson R, Rounsley S, Ventor C, Tabata S, Salanoubat M, Quetier F, Cherry
JM, Meinke D. The complete sequence of a plant genome. Plant Cell 1997; 9:476-478.
CHAPTER 3
Molecular Mechanisms of Salinity

Tolerance
Hans J. Bohnert, Hua Su and Bo Shen
P lants have evolved complex mechanisms allowing for adaptation to osmotic stress caused
by drought and to osmotic and ionic stress caused by high salinity. These mechanisms
can be classified into two categories: One includes developmental, morphological, and
physiological mechanisms; the other includes biochemical mechanisms. Developmental,
morphological, and physiological mechanisms are usually complex and require the functions
of many gene products. Examples of complex changes initiated by stress are the switch
from the C3 photosynthetic pathway to Crassulacean acid metabolism (CAM) in
Mesembryanthemum crystallinum following salt stress,1 the development of salt glands in
Limonium sp.,2 salt-storing epidermal bladder cells in Mesembryanthemum crystallinum3,4
and changes leading to increased water use efficiency in the development of the C4
photosynthetic pathway.5
Biochemical mechanisms, in contrast, are relatively simple, typically involving the
action of only a few gene products. For example, the accumulation of compatible solutes,
such as glycine betaine, proline, ectoine or polyols, only requires one to three enzymes for
extending a main metabolic pathway into the branch pathway of metabolite accumulation.6-9
Similarly, adjustments in ion uptake seem to be controlled by an equally small number of
gene products.10,11 With the current knowledge of plant genetics and biochemistry, the
genetic engineering of biochemical mechanisms is possible, but the engineering of more
complex traits is still beyond our capabilities. Once all relevant genes are known and
functionally characterized, it should be possible to manipulate complex developmental
mechanisms, such as flower development, vegetative growth or seed formation. This goal
is within reach for the genetic make-up of Arabidopsis thaliana at least,12 but the understanding
of how the 21,000 Arabidopsis genes function and their biochemical and physiological
interactions lies far in the future. In this review, we will focus mainly on biochemical
mechanisms that lead to cellular and whole-plant adaptations caused by the combined
osmotic and ionic disturbance of metabolism resulting from salt stress.
Over evolutionary time, plants colonized most places on earth, being excluded only
from high latitudes, the highest mountains and true deserts, which are cold and/or lack
water completely. Xerophytic and halophytic adaptations evolved in response to long-term
climate changes which allowed plants to tolerate all but the most extreme habitats. Plants
which have adapted to stressful environments provide paradigms for biochemical and
physiological tolerance mechanisms, and they provide genes for pathways that could
become incorporated into crop plants, which are typically stress-sensitive, having originated
from species in subtropical or tropical areas.13-15 Species adapted to extreme habitats are
not equally distributed among all orders or families of the angiosperm lineage. They
appear more frequently in orders which include few crop species but many species
restricted to stressful environments. These orders can be considered quarries for obtaining
novel genes for alternative biochemical pathways, and paradigms for understanding how
these pathways interact physiologically.
What constitutes tolerance or resistance to salinity stress has many facets, but is
surprisingly simple in principle. For both growth and development of reproductive
organs, plants must have water for photosynthesis to continue under stress; each one of
the many diverse mechanisms which evolved in an order-, family- or species-specific fashion
must be subordinate to this essential goal. We will review the molecular mechanisms for
which the evidence is clear:
1. Scavenging of radical oxygen species,
2. Controlled ion uptake,
3. The “burning” of accumulated reducing power, and
4. Adjustments in carbon/nitrogen allocation.
From the confusing multitude of physiological data, a few principles emerge (for
recent reviews, see refs. 11, 14-20, and other articles in this volume). Biophysical and
biochemical principles that govern stress and plant stress responses are outlined by Levitt.21
Osmolytes, Osmoprotectants, Compatible Solutes, Osmotic

Adjustment
High salinity disturbs uptake and conductance of water. Salt stress and other environ-
mental factors that affect water supply lead to changes in stomatal opening which can, if
stress persists, set in motion a chain of events originating from a decline in the leaf-internal
CO2 concentration, consecutively inhibiting the carbon reduction cycle, light reactions,
energy charge, and proton pumping.22 Other pathways are affected by increased shuttling of
carbon through the photorespiratory cycle.14,15 Eventually, carbon and nitrogen allocation and
storage require readjustment, reactions that lead to the consumption of reducing power
become favored, and development and growth may become altered. During the past years,
the complex interrelationship of biochemical pathways that change during salt stress has
become appreciated, although we are far from understanding this complexity.
The accumulation of metabolites, acting as osmolytes, in response to external changes
in osmolarity is probably universal.23-25 The generally accepted view is that osmolytes must
be compatible,26,27 not inhibiting normal metabolic reactions, and that their accumulation
leads to “osmotic adjustment” as the major element in accomplishing tolerance.6,7,24,28
Typically, compatible solutes are hydrophilic, giving rise to the view that they could
replace water at the surface of proteins, protein complexes, or membranes—we might call
them osmoprotectants in this case. The terms carry physiological meaning, but do not
explain the biochemical function(s) such solutes carry out. There may be more than one
function for a particular solute29,30 and, based on results from in vitro experiments,31-34
different compatible solutes seem to have different functions.
The main function of compatible solutes may be stabilization of proteins, protein
complexes or membranes under environmental stress. In in vitro experiments, compatible sol-
utes at high concentrations have been found to reduce the inhibitory effects of ions on
enzyme activity,24,35-37 to increase thermal stability of enzymes,38-40 and to prevent
dissociation of the oxygen-evolving complex of photosystem II.41 One argument often
raised against these studies is that the effective concentration necessary for protection in
vitro is very high, approximately 500 mM, concentrations which are usually not found in
vivo. However, when we consider the high concentration of proteins in cells, the concentration
necessary for protection can, we think, be much lower than that required for protection in
in vitro assays. In addition, it may not be the solute concentration in solution that is
Molecular Mechanisms of Salinity Tolerance 31
important. Glycine betaine (which may be present in high or low amounts), for example,
protects thylakoid membranes and plasma membranes against freezing damage or heat
destabilization,42-44 indicating that the local concentration on membranes or protein
surfaces may be more important than the absolute concentration.
Two theoretical models have been proposed explaining protective or stabilizing
effects of compatible solutes on protein structure and function. The first is termed the
“preferential exclusion model”45 which assumes the solutes are largely excluded from the
hydration shell of proteins. Exclusion leaves a water shell around proteins which stabilizes
protein structure, or promotes or maintains protein/protein interactions. In this model,
the solutes would not disturb the native hydration shell of proteins, but would interact
with the bulk water phase in the cytosol. The “preferential interaction model”, in contrast,
emphasizes interactions between solute and proteins.46 During water deficit, compatible
solutes may interact directly with hydrophobic domains of proteins and prevent their
destabilization, or they may substitute for water molecules in the vicinity of such regions.
While the two models seem to be mutually exclusive at first, the actual function may in fact
be explained by both models. The structures of different solutes could accommodate
hydrophobic, van-der-Waals interactions, as well as electrostatic interactions, but
additional biophysical studies will be necessary to gain a better insight into the stabilizing
effects documented by in vitro experiments.
Cellular Mechanisms of Salt Tolerance—the Fungal Model

Osmotic Adjustment
The unicellular eukaryotic Saccharomyces cerevisiae, baker’s yeast, is an ideal model
for studying cellular and molecular mechanisms of salt tolerance in higher plants. Its small
genome, which has been sequenced,47-49 adds to several other advantages. First, yeast is salt
tolerant and the cells have stress responses similar to halophytic plants. Yeast cells
accumulate compatible solutes, mainly glycerol and some trehalose, to counteract high
external osmolarity during salt stress;36,50-53 this is similar to the reactions of higher plants.
For example, halophytic plants, such as Plantago maritima and Mesembryanthemum
crystallinum, accumulate high concentrations of sorbitol and methylated inositols,
respectively, under salt stress.54,55 Second, both yeast and plants use proton gradients as
the driving force of secondary transport systems which control ion fluxes under stress.56, 57
Many ion flux mechanisms are highly conserved in yeasts and plants. In fact, a number of
plant membrane proteins, such as the potassium, amino acid, and sugar transporters, have
been isolated by functional complementation of yeast mutants.58-61 Third, the accumulation
of glycerol is essential for salt tolerance52,53 and glycerol-deficient mutants are available for
evaluating the functions of other sugar polyols in stress tolerance. Finally, yeast provides
an unsurpassed system for genetic analysis, transformation and functional characterization of
cell-specific functions, especially in light of the recent completion of the yeast genome
sequence.48,62
Yeast cells employ two main mechanisms for adaptation to salt stress: accumulation
of a polyol, glycerol, and maintenance of ion homeostasis. When exposed to NaCl the cells
experience both osmotic stress and ion toxicity. To respond to a low external osmotic
potential, the accumulating glycerol seemingly compensates for the difference between the
extra- and intra-cellular water potential.36 For reducing sodium toxicity, yeast cells have to
maintain low cytosolic Na+ concentrations and this is achieved by several mechanisms: by
restricting Na+ influx, rapidly extruding Na+ and/or efficiently compartmentalizing
sodium into vacuoles. The genetic evidence indicates both mechanisms are essential for
yeast salt tolerance.63-65
Glycerol Accumulation
Yeast cells accumulate glycerol as the major compatible solute when exposed to high
ion concentrations (Fig. 3.1).36 High osmolarity is perceived as a signal by two membrane
osmosensors: the protein products of Sln1 and Sho1. The signal is then transferred via a
MAP-kinase cascade66-68 and finally enhances the expression of the glycerol biosynthetic
pathway. Glycerol is synthesized from dihydroxyacetone phosphate. The first reaction is
catalyzed by glycerol-3-phosphate dehydrogenase which is encoded by two genes, GPD1
and GPD2. The second reaction converts glycerol-3-phosphate to glycerol by glycerol-3-
phosphatase, encoded by GPP1 and GPP2.52,53,69 The osmotic induction of both GPD genes
is mediated by the HOG-MAP kinase signaling pathway. In addition to induced glycerol
production, yeast cells may decrease membrane permeability to glycerol, which leads to an
increased retention of glycerol in the cells under osmotic stress. In fact, the salt-tolerant
yeast Zygosaccharomyces rouxii achieves glycerol accumulation by increased retention of
glycerol within the cell, and probably by active uptake of glycerol rather than by increased
production of glycerol during osmotic stress.36 In contrast, Saccharomyces cerevisiae
appears to increase glycerol production, while it fails to significantly alter membrane
permeability for glycerol retention during osmotic stress. To maintain high glycerol
concentrations in the cell requires a high energy cost, which seems to limit further
increases in salt tolerance in Saccharomyces cerevisiae. A glycerol transport protein (FPS1)
which shows homology with MIP-like water channel proteins has been recently isolated.
The expression of FPS is not regulated by the HOG-MAP kinase signaling pathway.70
Replacing Glycerol in Yeast

Although the correlation between accumulation of glycerol and yeast osmotolerance
has been established, and although the essential role of glycerol in the adaptation to
osmotic stress has been demonstrated by analysis of mutants deficient in glycerol
production,52,53,71 the mechanism(s) by which glycerol can confer such tolerance is not
clear. One obvious possibility is that glycerol is involved in osmotic adjustment to
maintain water flux into the cell. To test whether osmotolerance could be generated by the
presence of polyols other than glycerol, which would support the osmotic adjustment
concept, we introduced the coding regions of genes encoding enzymes for mannitol
and sorbitol production into a glycerol-deficient mutant (Fig. 3.1). However, accumulation of
either sorbitol or mannitol was not able to replace glycerol function (Shen B, Hohmann S,
Bohnert HJ, unpublished). Both foreign polyols accumulated to approximately the same
concentration as glycerol in wild type, and both were retained by the cell better than
glycerol, but both polyols provided only marginal protection. Growth inhibition of 50%
(I50) was 0.6 M NaCl for sorbitol/ mannitol producers in comparison to 0.4 M for the
mutant and 1.2 M for wild type. By reintroducing one of the deleted yeast GPD genes, a
significant increase in tolerance resulted, and the I50 increased from 0.4 M to 0.9 M
NaCl.72 If osmotic adjustment through glycerol is sufficient for salt adaptation, an equal
concentration of sorbitol or mannitol would be expected to confer very similar protection.
The results suggest that the concept underlying the term “osmotic adjustment” may not be
valid, or valid only if the synthesis of a metabolite for osmotic adjustment fullfills species-
specific requirements. The consequence of our results, then, is that glycerol might have specific
protective functions which mannitol and sorbitol cannot replace. Evolutionary adaptations
might have altered yeast proteins such that glycerol, but not other osmolytes, could exert a
protective role. Alternatively, the pathway through which an osmolyte is produced could be
more important than the end-product. Finally, the minimal protection by mannitol or sorbitol
could be caused by a difference in their intracellular distribution compared to glycerol.
Fig. 3.1. Genes involved in yeast osmotic stress signal transduction, and replacement of glycerol
synthesis by foreign osmolytes. The schematic drawing of a yeast cell includes the membrane
osmosensors (Sln1 and Sho1) which transmit signals to a MAP kinase cascade. Specific
transcription is initiated, which leads to the synthesis of several proteins, among them glycerol-
3-phosphate dehydrogenase (GPD) and glycerol phosphatase (GPP). This results in glycerol
synthesis and accumulation. The glycerol facilitator protein, FPS1, a MIP-type channel, is less
permeable under stress than under normal growth conditions. Replacement of both genes
encoding GPD by mannitol-6-P dehydrogenase or sorbitol-6-P dehydrogenase leads to the
accumulation of mannitol or sorbitol to a concentration approximately equal to glycerol
accumulation, but the two foreign polyols only marginally improve salt tolerance of the cells. 72
Whether and how these polyols are compartmentalized in yeast cells is not known, but
glycerol seems to be evenly distributed.73
A major difference exists between glycerol and mannitol/ sorbitol synthesis and
accumulation with respect to energy expenditure. Glycerol biosynthesis, which is also a
requirement for the removal of excess NADH during anaerobiosis,71 is more costly than
mannitol/ sorbitol generation. While NADH oxidation is, in principle, also accomplished
by the mannitol and sorbitol metabolic pathways, which leads to NAD+ increase, the cost
is different. During salt stress, more than 95% of the glycerol produced leaked from the
cells and accumulated in the medium. In contrast, sorbitol and mannitol did not significantly
exit from the cells. We calculated that glycerol biosynthesis under stress conditions
consumed at least 10 times more carbon and NADH than sorbitol/ mannitol biosynthesis.72
Thus, it may be that “burning” of excess reducing power via glycerol biosynthesis is as
important as the increasing osmotic potential provided by the steady-state glycerol
concentration in the cytosol.
Osmosensing and Signaling

The signaling pathways by which yeast cells respond to external osmolarity changes
has been identified (Fig. 3.1). High osmolarity is perceived as a signal by two membrane
osmosensors which are the protein products of Sln1 and Sho1. The Sln1 protein contains an
extracellular sensor domain, a cytoplasmic histidine kinase domain, and a receiver domain.
YPD and Ssk1 receive sensor signals in the cytosol. Sln1 and YPD/Ssk1 function like bacterial
two-component systems.67,74,75 Sho1 is a transmembrane protein containing a cytoplasmic
SH3 domain which can directly activate a MAP kinase kinase, Pbs2, by interaction between
its SH3 domain and a proline-rich motif of Pbs2.68 The signal from Sln1 is transmitted via a
MAP kinase cascade encoded by MAP kinase kinase kinase (Ssk2/22), MAP kinase kinase
(Pbs2), and MAP kinase (Hog1).66-68 The cascade initiates the expression of the glycerol
biosynthetic pathway including the GPD1 and GPP2 (phosphatase) genes.52,69 In addition
to glycerol biosynthesis, genes for other stress responses, such as CCT1 encoding catalase T76
and HSP12 encoding a small heat-shock protein,77,78 are induced by this signaling pathway.
In contrast to hyperosmotic stress, hypoosmotic stress initiates a second MAP kinase
cascade called the protein kinase C1 (PKC1) pathway. The MAP kinase in this PKC1
pathway is phosphorylated when cells are transferred from high osmolarity to low osmolarity.
Protein kinases downstream of PKC1 include BCK1/SLK1, MKK1/MKK2 and MPK1/SLT2.79
PKC1 mutants exhibited a lytic phenotype due to defects in cell wall biosynthesis. The
lytic phenotype can be suppressed by the addition of osmolytes like sorbitol into the
medium.80 How the two osmosensing pathways are coordinated remains to be determined.
Ion Relations
The yeast genome contains approximately 5,800 genes which potentially encode proteins.48
About 250 genes show significant similarity to membrane transport proteins characterized in
yeast and other organisms. Among those, a (partial) functional characterization existed
for only about 60 genes prior to the completion of the sequence, which amply documents
both the value of sequencing projects and our relative ignorance of membrane transport
processes in general.81 Many of these membrane transport proteins are involved in ion
transport and carry out essential functions in salt tolerance.
Potassium Transport
Potassium plays an important role in yeast salinity tolerance. The osmotic potential
generated by high internal potassium concentrations (e.g., in halobacteria) can alleviate
sodium toxicity.36 Three membrane proteins are involved in potassium transport across
the plasma membrane, TRK1, TRK2, and TOK1.82-84 The TRK proteins are involved in K+
influx and TOK1 controls K+ efflux. TRK1 and TRK2 are required for high affinity and
low affinity potassium uptake, respectively. Importantly, TRK proteins can also transport
Na+ but both have a higher affinity for K+. Under high external Na+ concentrations, Na+
can inhibit K+ uptake and enter the cell through the potassium channels. The capacity for
transporting potassium into cells and restricting sodium influx by increased
K+discrimination over Na+ is an essential element for salt tolerance acquisition.85-87
Because the high affinity K+ transport system shows a higher K+/Na+ discrimination than
the low affinity system, under salt stress yeast cells may shift from low to high affinity K+
uptake, allowing the cells to accumulate more K+ than Na+ and to maintain a low Na+/K+
ratio.85,88 Increased K+/Na+ discrimination of a high affinity potassium transporter (HKT1)
from wheat has been shown to increase salt tolerance of yeast strains deficient in potassium
uptake.86
Two halotolerance genes, HAL1 and HAL3, have been isolated by screening for genes
that enhance salt tolerance when overexpressed.89,90 Both have been implicated in the
regulation of K+ concentrations. Overexpression of HAL1 and HAL3 resulted in a stronger

accumulation of K+ under salt stress and increased salt tolerance. The beneficial effects are
specific for NaCl stress, and cannot moderate osmotic stress by sorbitol or excess KCl,
suggesting that high K+ in a physiological range may specifically alleviate sodium toxicity.
Yeast double mutants with deletions of the high affinity K+ transporter (TRK1) and the
Na+-ATPase (ENA1) genes are sensitive to Na+ because of poor Na+/K+ discrimination
and decreased Na+ efflux.85
Similarly, long-term salt-adapted tobacco cells showed increased capacity for K+
uptake compared to wild type cells,91 suggesting better Na+/K+ discrimination by the K+-
uptake system as a significant element for salt tolerance. Likely in the same category,
Arabidopsis sos1 mutants were hypersensitive to salt stress due to a defect in the high affinity K+
uptake system, highlighting the important role of K+ for salt tolerance in plants.92
H+-ATPases
Plasma membrane and vacuolar proton ATPases are essential for generating and
maintaining membrane proton gradients and for pH regulation in yeast and plants.93-96
They must be able to sense and respond to external acidification. The yeast plasma membrane
H+-ATPases (P-ATPase), encoded by the gene PMA1, is predominantly responsible for
proton gradient maintenance, while the product of the PMA2 gene is induced at low pH
when the PMA1 protein cannot function properly.97 Regulation of activity by calcium-
dependent protein kinases, in response to glucose levels, weak organic acids, heat shock
and salt stress, has been shown.98,99 Mutants hypersensitive to the immunosuppressants
cyclosporin A and FK506 were shown to be defective in assembly of the vacuolar H+-ATPase
(V-ATPase). Their characterization indicated involvement of the calcineurin signal
transduction pathway in synthesis, endomembrane transport, assembly and activity
regulation.98,100,101
Sodium Transport Across Membranes

Maintaining low intracellular sodium amounts during salt stress is essential for yeast.
Low sodium concentrations in the cytosol could be achieved by decreased Na+ uptake,
increased Na+ efflux, transport of Na+ into vacuoles or a combination of such activities. In
S. cerevisiae, a Na+-ATPase encoded by a family of 4 or 5 ENA genes has been shown to be
involved in sodium efflux. The expression of the ENA1 gene is induced by salt stress while
the other genes are expressed constitutively and weakly. Mutants defective in the ENA
function are sensitive to sodium and lithium.65,102,103 Also, an Na+/H+ antiport protein,
encoded by Nha1, was found during sequencing of the genome81 and later functionally
characterized.104 NHA1 seems to play a minor role in sodium efflux, but it may be important in
an environment of acidic external pH which would affect the transmembrane proton
gradient.
Based on the results with yeast it was surprising, however, that in Schizosaccharomyces
pombe and Zygosaccharomyces rouxii the major sodium efflux is via such a Na+/H+-
antiporter encoded by the SOD2 gene. SOD2 was initially identified by selection for
increased LiCl tolerance in fission yeast105 and the homologous SOD2 was isolated from
Z. rouxii.106 Functional expression of ENA1 in a sod2 mutant of S. pombe restored Na+
efflux and salt tolerance. Recently, the activity of the SOD2 Na+/H+ antiporter was con-
firmed using microphysiometry, indicating reversible sodium transport, dependent on the
Na+ and H+ gradient across the membrane.107 Based on this property, it should be possible
to utilize SOD2 to transport Na+ into vacuoles by targeting the protein to the tonoplast in
higher plants.
Although mechanisms of Na+ uptake in yeast are still not understood, mutant
analysis has clearly demonstrated an essential role for membrane-located processes.
Disruption of the LIS1/ERG6 gene, encoding a SAM-dependent methyltransferase of the
ergosterol pathway, resulted in increased sodium uptake and decreased salt tolerance. The
mutation seems to affect cation transport indirectly by changing membrane composition.108
Several other uncharacterized mutants showing high internal sodium were salt sensitive
despite normal glycerol accumulation.63,64 Halophytic plants usually sequester Na+ into
vacuoles to lower the concentration of Na+ in the cytoplasm.3 Whether such a mechanism
exists in yeast is unknown, but evidence exists for the vacuole’s important role in salt
tolerance. Mutants defective in vacuole morphology and vacuolar protein targeting are
salt-sensitive.109 A mutant in subunit C of the vacuolar ATPase shows increased sensitivity
to Na+ and Li+.85 The essential function may be associated with both compartmentation
of ions and osmoregulation.
Calcineurin Signaling
The signal transduction pathway regulating ion homeostasis remains unknown in
detail, but it is known to be different from the HOG pathway. Recent studies revealed that
calcineurin and protein phosphatase PPZ seem to be involved in the regulation of ion
fluxes.88,110-112 Calcineurin, a protein phosphatase 2B consisting of a catalytic subunit (CNA)
and a regulatory subunit (CNB), requires Ca2+ and calmodulin for activity.113 Null
mutants of calcineurin fail to recover from G1-arrest in the presence of α-pheromone, but
show normal growth rates under normal growth conditions. Under salt stress, however,
the mutants exhibited a salt-sensitive phenotype,88,110 caused by reduced expression of the
ENA1 gene which is regulated by calcineurin. Also, calcineurin mutants cannot shift from
low- to high-affinity potassium transport under salt stress.88 In contrast, deletions of genes
for the protein phosphatases PPZ1 and PPZ2 increased salt-tolerance due to enhanced
expression of ENA1, suggesting an essential role of these phosphatases in yeast ion
homeostasis.112 At low salt concentrations, the HOG-MAP kinase pathway appears to be
involved in regulation of ion fluxes, while at high salt concentrations ion balance is mainly
controlled by calcineurin.114 Interestingly, calcineurin signaling seems to interact with the
MAP kinase pathway. Disruption of the calcineurin gene (Ppb1) in fission yeast resulted in
sensitivity to chloride. High copy number of the Pmp1 gene, encoding a phosphatase,
suppresses this sensitivity to chloride. The PMP1 phosphatase dephosphorylates PMK1,
the third MAP kinase in fission yeast. As expected, deletion of Pmk1 also suppresses the
chloride sensitivity of calcineurin mutants. 115 Other components in the calcineurin
signaling pathway remain to be identified.
In addition to calcineurin and PPZ, HAL1 and HAL3 are involved in the regulation of
intracellular Na+ and K+ concentrations.90,116 The effect of HAL1 and HAL3 on intracellular
Na+ is mediated by expression of the ENA1 gene. Calcineurin plays a role in the induction
of ENA1 expression by sodium, while HAL1, HAL3 and PPZ determine the basal level of
ENA1. HAL1 and HAL3 are then required for the maximal expression of ENA1 under salt
stress. Overexpression of HAL1 or HAL3 partially suppressed the salt sensitivity of
mutants with a non-functional calcineurin. Increased K+ by overexpression of HAL1 is
independent of the action of TRK1 and TOK1, probably due to decreased export of K+
during salt stress.116 Clearly, multiple regulatory pathways and control circuits govern ion
responses in a complex interaction depending on external signals.
In higher plants, a calcineurin-like protein phosphatase activity has been found in
the regulation of the K+-channel in guard cells of fava bean.117 FK506 and cyclosporin
A, immunosuppressants which bind to cellular receptors, are strong and specific
inhibitors of calcineurin.118 When FK506-receptor complexes were added to guard cells,
the Ca 2+ -induced inactivation of K + channels was inhibited. A Ca 2+ -dependent

phosphatase activity which is sensitive to complexes of FK506 and its binding protein
and to cyclophilin-cyclosporin A was also identified in guard cells. 117 The gene which
encodes cyclophilin has been cloned,119 suggesting an important role of calcineurin in
higher plants. In addition, by complementation of yeast calcineurin mutants, two cDNAs
(STO and STZ) which suppress the calcineurin deficient phenotype in yeast have been
isolated from plants, but the predicted protein sequence did not show significant
homology to phosphatases. 120 Until now, the plant calcineurin gene has not been
identified. The important function of calcineurin in salt tolerance has recently been
demonstrated by overexpression of a truncated yeast calcineurin in transgenic tobacco.
Constitutive expression of this yeast calcineurin increased salt tolerance of transgenic
tobacco plants (Bressan RA, Pardo J, personal communication).121
Molecular Mechanisms of Salt Tolerance in Plants

Metabolite Accumulation
Accumulation of compatible solutes during osmotic stress is a ubiquitous biochemical
mechanism, present in all organisms from bacteria, fungi and algae to vascular plants and
animals.24,36 The accumulating metabolites include amino acids, their derivatives (proline,
glycine betaine, β-alanine betaine, proline betaine), tertiary amines, sulfonium compounds
(choline o-sulfate, dimethylsulfoniopropionate), the raffinose series of sugars, and polyols
(glycerol, mannitol, sorbitol, trehalose, fructans, and methylated inositols).6,14,15, 122,123
Enzymes from halophytes do not show remarkably higher salt resistance than those
from glycophytes, nor do they require sodium for optimal activities. In fact, the activity of
enzymes from both is generally strongly inhibited by high concentrations of either NaCl
or KCl.3 Although halophytes and glycophytes use similar compatible solute strategies to
deal with osmotic stress,124 they use different strategies to cope with ion toxicity. Halophytes
take up sodium and sequester ions into the vacuole. High osmotic potential in vacuoles is
balanced by accumulating compatible solutes in the cytoplasm. Because the cytoplasmic
volume is relatively small compared to the large volume of the vacuole, low concentrations
of compatible solutes suffice to reach the same osmotic potential in the cytoplasm. In
contrast, glycophytes usually attempt to limit sodium uptake or transport sodium to old
leaves as an alternative way to extrude sodium out of plants.3,125,126 The halobacteria deviate
from the general compatible solute strategy, accumulating K+ as the osmolyte rather than
organic solutes to counteract high external osmotic potential. Halobacterial enzymes
require high ion concentration for their optimal activity.36 This adaptation required changes
in protein structure. During evolution, this type of stress adaptation was abandoned,
possibly because it proved inflexible to changing environments, and mechanisms became
favored which utilized organic solutes, likely because they could be synthesized through
pathways attached to basic metabolism.
Several common features characterize the different compatible solutes. First, they can
easily be synthesized from compounds diverted from basic metabolism by novel
enzymatic or regulatory reactions. For example, glycine betaine in higher plants is
synthesized from choline via two reactions catalyzed by choline monooxygenase and
betaine aldehyde dehydrogenase,6 and pinitol is synthesized from myo-inositol in two
reactions catalyzed by inositol o-methyltransferase and ononitol epimerase.4,8,15 Both
choline and myo-inositol are high-flux metabolites and are tightly regulated during growth.
Second, accumulation of compatible solutes under osmotic stress is an active process, rather
than an incidental consequence of other stress-induced metabolic changes. The biosynthetic
pathway for a particular osmolyte is coordinately up-regulated during osmotic stress. For
Table 3.1 Transgenes with Effects on Salt-, Drought- and Low Temperature
Tolerance
ROS Scavenging Enzymes 1991 SOD, catalase, GST/GSX overexpression

leading to enhanced stress tolerance.20,181,232,233,236,238
Mannitol Synthesis 1992 Protection against salt stress.251,252,253
Fructan Accumulation 1995 Enhanced drought tolerance.254
Proline Accumulation 1995 Enhanced salt stress tolerance.135
Glycine betaine Synthesis 1997 Enhanced temperature stress, salt stress tolerance.255
LEA Protein Synthesis 1996 Salinity and drought stress protection.256
Potassium Transporter 1994 Enhanced Na+/K+-discrimination in yeast.86,207
Trehalose Synthesis 1996 Enhanced drought tolerance.141
Glutathione Cycle 1997 Altered redox control, salt and low temperature
Enhancement protection.192
Mannitol as a Hydroxyl 1997 Enhanced salt tolerance, mannitol synthesis in

Radical Scavenger chloroplast.29,30
Inducible Ononitol 1997 Enhanced drought and salt tolerance; inducibility

Accumulation based on changes in substrate amounts.138
Extreme Sorbitol 1998 High accumulation, >600 mM sorbitol, leading to

Accumulation necrotic lesions in sink leaves.227
example, two key genes, Inps1 and Imt1, are transcriptionally enhanced by salt stress, and
higher enzyme amounts lead to increased carbon flux through myo-inositol into pinitol
biosynthesis in stressed Mesembryanthemum.8,127-129 Genes involved in the degradation of
compatible solutes are down-regulated under osmotic stress. This is, for example, the case
for proline oxidase in Arabidopsis thaliana. Stress-dependent lower expression of this
enzyme, at least in part, may explain the increases in proline during salinity and drought
stress.130 Third, many accumulating compounds are end-products of a branch pathway
rather than active intermediates in so far as one enzyme in the pathway catalyzes only the
forward reaction. Examples for this point are DMSP synthesis in marine algae,9,131 pinitol
synthesis in Mesembryanthemum4,55 and glycinebetaine synthesis.6,132-134 Equally, proline
biosynthesis has received much attention, because proline accumulation is a nearly universal
reaction of plants to osmotic stress.135-137 Its true role in stress protection is, however, not
clear—we consider the accumulation of proline a consequence of the necessity for
readjusting carbon nitrogen balance under stress.138 The biosynthesis of ectoine (tetra-
hydropyrimidine and derivatives), an accumulating osmolyte in bacteria, has received
Fig. 3.2. Pathways for the synthesis of selected compatible solutes. Biochemical pathways originating
from glucose-6-P or sorbitol-6-P whose presence in some stress-tolerant species or after gene
transfer into transgenic tobacco is correlated with increased osmotic stress tolerance. Genes/
enzymes used in transgenic experiments are PGM (phosphoglucomutase), INPS (myo-inositol
1-P synthase), IMT (myo-inositol O-methyltransferase), GPDH (sorbitol-6-P dehydrogenase),
MtlDH (mannitol-1-P dehydrogenase), TPS (trehalosephosphate synthase). Pase indicates
unspecific phosphatases. OEP (ononitol epimerase) is found in Mesembryanthemum, but the
gene has not yet been cloned. IMP (myo-inositol monophosphatase) is not regulated in
Mesembryanthemum during stress and has not been included in transgenic plants.
attention recently.130,140 Expression of the three enzymes leading to ectoine in bacteria

confers significant salinity tolerance. Figure 3.2 shows schematically selected pathways that
lead to the synthesis of polyols (mannitol, sorbitol, ononitol and pinitol) and to trehalose
synthesis.141 Apart from the pinitol biosynthetic pathway,8,11 the pathways shown are
engineered pathways (Table 3.1) and may be different from pathways existing in some
plant species naturally. The scheme indicates clearly how the addition of a single gene can
be exploited for metabolic engineering.
Water Channels
Water channels, aquaporins (AQP), are found in all organisms as members of a
super-family of membrane proteins, 26-30 kDa in size, termed MIP (major intrinsic
protein).142,143 The proteins are characterized by six membrane-spanning domains and a
pore-domain with a characteristic sequence signature, NH3-NPAXT-COOH. Aquaporins
enhance membrane permeability to water in both directions depending on osmotic pressure
differences across a membrane, but other members of the gene family in yeast and vertebrates
encode glycerol-facilitators.143 Other MIPs, animal and plant—among them a nodulation-
specific protein, may mediate ion transport and transport of other neutral metabolites, such as
urea.144,145
Complexity of Plant Isoforms

In human DNA, five MIP genes have been characterized among a total of seven MIP-like
genes. They are expressed in different tissues, most highly in erythrocytes, kidney cells and
the brain. In contrast, Arabidopsis contains at least 23 MIP-like coding regions.146 Sequence
signatures of the Arabidopsis MIP indicate two large sub-families of 10 to 12 proteins each
whose members are either plasma membrane-located (PIP) or tonoplast-located (TIP),
and one MIP which diverges from the others has not been characterized.146 While some of
the genes might encode facilitators for diverse small metabolites or ions, eight MIP
proteins have already been identified as aquaporins. Why are there so many plant
aquaporins? We discuss four possibilities which might explain the high number.
1. MIP-intrinsic functional variations might allow AQP to be active at different
membrane osmotic potentials. Yet, all we know is that the intrinsic water permeability
distinguishes four human AQP and one glycerol facilitator by a factor of ~100,147 and
that plant AQP can be either sluggish or effective water transporters when
expressed in Xenopus oocytes.143 There is no report about a functional plant model
that would allow mechanistic studies on AQP. By antisensing with a plasma membrane
AQP coding region148 which supposedly targeted all expressed PIP, the decrease in
AQP amounts led to a decline in water uptake in plants. Such antisense AQP
transgenics increased the root to shoot ratio, suggesting a feedback mechanism
between water uptake and root mass (Kaldenhoff R, personal communication).
Protoplasts from the antisense-expressing plants did not burst as fast as wild type
cells when transferred to hypoosmotic solutions.148
2. Functional differences could have evolved for fine tuning water flux through the
plant—with high conductance AQP located in the root cortex and vascular tissues
which accommodate bulk fluxes and low conductance channels between
mesophyll cells, for example, or even within the cell cytosol and organelles and the
vacuole.
3. Without assuming functional diversification, the number of AQP arising through
gene duplications could have changed gene and protein expression, half-life, and
turnover such that AQP amount shows a gradient that follows the water transport
gradient. In this scenario, the gene number—requiring different promoters, RNA-
stability and translation characteristics and protein half-life regulation—would be
determined by the necessity of cell-specific differences in accommodating water
flux and not by the water transport function per se. This explanation is similar to
the following one, and both find precedence in the presence of, for example, a
large number of genes encoding plasma membrane H+-ATPases, AHA, which are
differentially expressed throughout the plant.95,149,150 Deletion of several AHA genes
did not produce a phenotype under normal growth conditions, but affected growth
significantly under diverse growth and stress conditions, low temperature, salt stress
and external high acidity, for example (Sussman MR, personal communication).
4. Last, AQP/MIP multiplications and diversifications could have been dictated by
the need for a flexible response to environmental changes in water supply or
evaporation, demanding the presence of several sets of AQP. This assumes evolution of
one set of AQP genes for stress responses and that this set is different from others.
It is conceivable that a set of Mip genes exists to take care of the business of cell
expansion following meristematic activity—and this function (missing from animals)
might require regulatory circuits different from those necessary in genes that perform
housekeeping (set 2) and stress-response functions (set 3). Although the data are
not complete with respect to AQP protein expression and cell-specificity, alignments
of sequences indicate that sub-families of two to four closely related sequences
exist146,151 which might represent the three sets of genes. MIP associated with cell
expansion,152 developmental specificity 153,154 and stress functions151,155-158,257 have
been described.
Mechanisms of Regulation
Most important to the topic here is how MIP gene expression, protein amount and
aquaporin activity are controlled during development and under environmental stress.
Regulation is by gene expression and protein amount, and possibly also by post-translational
modification—but we have very little information on mechanistic details in plants.
Weig et al146 used quantitative PCR amplification for the 23 Arabidopsis MIP and
found differences in mRNA amounts spanning several orders of magnitude. Differences
in RNA amounts for each MIP in roots, leaves, bolts and the flowers and siliques were
equally pronounced. No signals were detected for at least three MIP, suggesting that these
might be expressed under conditions not found during normal growth or that they are
expressed in a few cells only or at very low levels. The analysis of such a large gene family,
once all genes are known, can best be done by in situ hybridization, immunocytology with
specific antibodies and DNA microarray analysis through which the amount, location and
regulation of the genes during development and under different environmental conditions
can be monitored. For several MIP in a number of organisms, salt stress altered mRNA
amounts have been reported. AQP expression also responds to drought and low temperature,
hormone treatment (ABA, cytokinine, GA), light, and pathogen infection.143,157,158
Promoter studies have been performed with several MIP, but cell-specificity is
most likely the essential distinguishing factor between AQP and must receive more
attention in order to understand water transport in plants. The promoter for Rb7a159 from
tobacco conveys root-specificity, leads to differential expression in the root in a cell-
specific manner and is induced by nematode feeding.156,159 The Mesembryanthemum MipB
promoter showed highest expression of the gene in roots;151 after transfer into tobacco
and observation of GUS expression, broader specificity was observed, with highest expression
in all meristematic cells and in vascular tissues.160
Even less complete is the information about protein amount, localization and changes
during development and under stress conditions. One essential consideration is that the
large number of genes and high sequence identity among PIP and TIP, respectively,
require excellent controls for avoiding cross-hybridizations between transcripts and
immunological cross-reactivity between antisera. For example, generation of anti-peptide
antibodies against six Mesembryanthemum MIP resulted in distinguishable signals to
different cells.161 However, in the absence of probes for all MIP for this species, it cannot
be excluded that some of the antibodies react to more than one MIP whose sequence is not
yet known, but shares homology with the selected peptide domain.
Regulation has been documented at the level of post-translational modification, mostly
in animal systems. Salt stress conditions in kidney cells lead to changes in protein expres-
sion, which may be controlled by oligomerization, glycosylation, or phosphoryla-
tion.162,163 In addition, the presence in the cell membrane and the half-life of AQP is
determined by the hormone vasopressin in animal cells. Increased vasopressin leads to the
deposition of AQP from internal stores, endosome vesicles, to the outer membrane, and
lower hormone levels lead to cycling of membrane patches through endosomes.164 Clearly,
such traffic and its control would constitute the fastest, most economic way of regulating
water flux. Similar observations remain to be made with plant MIP, but patches of invaginated
plasma membrane regions, termed “plasmalemmasomes” that contain abundant AQP
protein have been found in plants,165 possibly the functional equivalent of animal
endosomes. Our preliminary experiments indicate that PIP from Mesembryanthemum
sediments in different gradient fractions depending on whether salt-stressed or unstressed

cells were used,161 which might indicate that similar membrane shuttle mechanisms exist
in plant cells.
Evidence for plant AQP regulation comes from studies which measured AQP phos-
phorylation.153,166 Regulatory sites for phosphorylation have been mapped in several
MIP/AQP. 143 Also, effects of pharmaceutical agents on water flow in Chara cells, for
example, point towards an association of water flux and the integrity of the cytoskeleton (see
ref. 143). Spinach leaf PIP are reversibly phosphorylated in response to the apoplastic
water potential and calcium.166
The discovery and preliminary characterization of AQP in plants has provided more
questions than answers. Their existence cannot be questioned and they act as water
channels. It is then intuitively obvious that control over their action should be important
under stress conditions. Although there are few data available, it is equally clear that regulation
during stress is complex, involving transcriptional and post-translational controls which
seem to involve synthesis, membrane traffic and reversible insertion into membranes,
complex assembly and MIP protein half-life.
Salt Stress and Radical Scavenging
Reactive Oxygen Species and Radical Scavenging Systems

Production of Reactive oxygen species (ROS) is an unavoidable process in photosynthetic
tissues, but ROS are also produced in mitochondria and cytosol. ROS including singlet
oxygen, superoxide, hydrogen peroxide, and hydroxyl radicals react with and can damage
proteins, membrane lipids, and other cellular components.33,167,168 Some ROS also serve
as signaling molecules, 20 for example, in the initial recognition of attack by
fungal pathogens and the transmission of signals after a primary infection.169,170 Focusing on
chloroplasts, superoxide is abundantly produced from photoreduction of oxygen. Oxygen
concentration as high as 300 mM can be photoreduced to superoxide by photosystem I via
a Mehler reaction.171,172 The production of superoxide has been estimated to be approximately
30 mmol (mg chl)-1 h-1 in intact chloroplasts,173 and the rate of production in isolated
thylakoids was increased 1.5-fold by the addition of ferredoxin and decreased 50% by
addition of NADP+.174 Most of this thylakoid lumen-produced superoxide diffuses to the
stroma.173 H2O2 in chloroplasts is predominantly generated by disproportionation of
superoxide by SODs. In peroxisomes, H2O2 originates directly from glycollate oxidase
activity. Hydroxyl radicals derive from an interaction between hydrogen peroxide and
superoxide or directly from hydrogen peroxide in the presence of transition metals such as
Fe+2 and Cu+ by a Fenton- or Haber-Weiss-reaction. The oxidized metal ions can be
re-reduced by superoxide, glutathione, or ascorbate. Trace amounts, lower than the amount
present in chloroplasts, of metal ions are needed to catalyze the Fenton reaction.168,173 It
has in fact been shown that elevated amounts of iron lead to increased oxidative stress.175
These Reactive oxygen species are scavenged by resident enzyme systems and
nonenzymatic antioxidants. 176 Non-enzymatic detoxification mechanisms include
morphological features such as waxy surfaces and leaf or chloroplast movement, non-
photochemical quenching processes by various compounds, for example, the violaxanthin-
zeaxanthin cycle, and photorespiration. Non-enzymatic antioxidants include flavonones,
anthocyanins, α-tocopherol, ascorbate (at a concentration of ~10 mM in chloroplasts),
glutathione, carotenoids, phenolics and polyols.20,32,168,177 Botanical sources of such
antioxidants not only play important roles in plant stress adaptation, but also retard aging
and diseases related to oxidative damage in animals.178
The enzyme systems involved include SODs which catalyze the reaction from superoxide
to hydrogen peroxide, and ascorbate peroxidases (APX) responsible for the conversion of
hydrogen peroxide to water. Both SOD and APX are represented by isoforms localized to
the stroma and the thylakoid membrane. Ascorbate can be regenerated by the ascorbate-
glutathione cycle. The level of reduced glutathione is maintained by glutathione reductase
using NADPH.168,179,180 In addition, catalase has recently been demonstrated as a sink for
H2O2 in C3 plants.181 In contrast to the detoxification systems for H2O2 and O2-, an
enzyme system that could deal with the short-lived, extremely toxic hydroxyl radical has
not been identified and, in fact, might not have evolved.167,168,179,180 The best way of
detoxifying hydroxyl radicals is to prevent their formation by reducing the concentration
of H2O2 and free metal ions. Once produced, however, protection depends on the presence
of antioxidants in the vicinity of the formation site. Together these systems provide sufficient
protection under normal growth conditions; in fact, the scavenging systems are able to
handle moderate increases of ROS, unless long-term stress exceeds the detoxification
capacity.20,179,182 In chloroplasts, oxidative damage includes first a decline in CO2 fixation,
and then inhibition of photochemical apparatus, loss of pigments, oxidation of proteins,
and lipid peroxidation.183,184
ROS and Environmental Stress

Several lines of evidence support the toxicity of ROS during drought,20 chilling stress184
and salt stress.29,30 First, superoxide production is enhanced, as detected by EPR signals in
drought stressed wheat and sunflower.185,186 Equally, H2O2 content increased about three-fold
during drought and low temperature.187-189 Enhanced production of ROS resulted in an
increase in lipid peroxidation, as documented by a more than 5-fold increase of
malonaldehyde production in wheat.190 Second, the concentration of free transition iron
increased under drought stress,190,191 which stimulated production of hydroxyl radicals in
the presence of high concentrations of H 2O 2 via a Fenton reaction. Compared to
superoxide and H2O2, hydroxyl radicals oxidize a variety of molecules at near diffusion-
controlled rates. Finally, levels of non-enzymatic radical scavengers, such as ascorbate,
carotenoids, flavonoids, sugar polyols, and proline,183 increase and may complement
enzyme protection systems.
Excellent evidence for a protective effect of ROS scavenging systems has recently been
provided by the overexpression of an enzyme with the combined activities of glutathione
S-transferase, GST, and glutathione peroxidase, GPX.192 By doubling the GST/GSX activity in
transgenic tobacco, the seedlings and plants showed significantly faster growth than wild
type during chilling and salt stress episodes. The increased enzyme activities resulted in
higher amounts of oxidized glutathione (GSSG) in the stressed plants, indicating that the
oxidized form could provide an increased sink for reducing power.
Another set of experiments shed light on the relationships between ROS and the
accumulation of polyols. When a bacterial gene (mtlD) encoding mannitol-1-phosphate
dehydrogenase was modified so that the enzyme was expressed in chloroplasts, transgenic
tobacco contained approximately 100 mM mannitol in the plastids. Using transgenic plants,
freshly prepared cells and a thylakoid in vitro system, the protective effect exerted by
mannitol on photosynthesis characteristics could be shown.29, 30 The presence of mannitol
resulted in increased resistance to oxidative stress generated by methylviologen, and cells
exhibited significantly higher CO2 fixation rates than controls during stress. After
impregnation of tissue and cells with dimethyl sulfoxide, a hydroxyl radical generator,
mannitol-containing cells showed a lower rate of methane sulfinic acid production than
wild type, indicating that mannitol acted specifically as a hydroxyl radical scavenger. It
could be shown that the primary damage was to enzymes of the Calvin cycle and not to
components of the light harvesting and electron transfer systems,30 a confirmation of

earlier reports.22 At present, the interpretation which we favor is that mannitol interferes
with either hydroxyl radical production or damage, but it is unknown whether the
protective mechanism is by exclusion of hydroxyl radicals from protein surfaces, a chemical
interaction between mannitol and hydroxyl radicals, or by inhibiting or reducing the
amount of hydroxyl radicals produced in the Fenton reaction.
Plant Ion Uptake and Compartmentation
H+-ATPases and Vacuolar Pyrophosphatase

Plasma membrane and vacuolar proton transporters play essential roles in plant
salinity stress tolerance by maintaining the transmembrane proton gradient that assures
control over ion fluxes and pH regulation (Fig. 3.3).101,193 Three proteins/protein complexes
exist for this purpose: the plasma membrane (H+)-ATPase (P-ATPase) and two vacuolar
transport systems, a (H+)-ATPase (V-ATPase) and a pyrophosphatase (PPiase).
The plant P-ATPase is represented by a gene family of more than 10, encoding
proteins of ~100 kDa, with homology to the yeast PMAs.95,150 As the main proton pump
in the outer cell membrane it is essential for many physiological functions.194 Increased
activity of the proton pump has been shown to accompany salt stress. Halophytic plants
have been shown to increase pump activity under salt stress conditions more drastically
than glycophytes,56,195 but little is know about the regulatory circuits that lead to either
increased protein amount or activity during salt stress.
The V-ATPase, a multi-subunit complex homologous to organellar, yeast (VMA) and
bacterial F0F1-ATPases, has already been shown to be important in plant salinity tolerance.
Electrophysiological studies revealed increased activity of this ATPase when cells or tissues
from stressed plants were analyzed.196,197 Transcripts for several subunits of the V-ATPase
are upregulated following salt shock.198,199 In Mesembryanthemum, V-ATPase activity
increases several-fold following stress.200,201 In a Mesembryanthemum cell culture model it
has now been shown, based on immunological data, that the V-ATPase (and possibly the
P-ATPase) activity does not increase due to more protein being present, but an unknown
mechanism stimulates activity 2 to 3-fold.201 The response is specific for NaCl and could
not be elicited by mannitol-induced osmotic stress.
PPiase genes and tonoplast-located PPiase proteins have been characterized in de-
tail.94 Contrary to previous assumptions, the enzyme has now been authenticated as
also residing in the plasma membrane.202 Its function, if any, under salt stress conditions
is little known. A few reports have indicated that PPase activity declines under salt stress in
some species,203,204 but increases in others.205
Potassium Transporters and Channels

One possible passage for sodium across the plasma membrane is through transport
systems for other monovalent cations. Among those the most significant is the uptake
system for potassium, the most abundant cation in the cytosol, with important roles in
plant nutrition, development and physiological regulation. Many studies have focused on
identifying components involved in K+-transport. Physiological observations indicating a
biphasic uptake of K+ into roots206 gave rise to the assumption that two uptake entities
should be involved, a high-affinity system functioning at µM concentrations of external
K+ and a low-affinity system active in the mM range of potassium. Several plant K+
transporter and K+ channel genes have been isolated by functional complementation of
yeast mutants deficient in K+ uptake58,59,207 or by sequence homology with known K+
transporter or channel genes. 208-210 Electrophysiological studies in heterologous
Fig. 3.3. Transport proteins implicated in plant salinity stress tolerance. The schematic depiction
of a plant cell includes the vacuole, chloroplast (cp), mitochondrion (mt) and cell wall (shaded).
Transmembrane proton gradients established by proton-ATPases and pyrophosphatase are
indicated (+/-). Under NaCl stress, Na+ and Cl- are sequestered to the vacuole, and K+ and
osmolytes are present in high concentrations in the cytosol. Symbols for several membrane-located
transporters and channels are identified by the ion or proton transported and by the direction of
movement. For organelles (mt and cp) no transporters have been characterized through
molecular techniques. A Na+-ATPase, included in the plasma membrane is hypothetical, and a
Na+/H+-antiporter in the plasma membrane has not been detected.
expression systems, such as Xenopus oocytes or yeast cells, indicated that some of them
may function at both affinity ranges.211
Inward-rectifying potassium channels function in the mM range, following the
electrochemical gradient at the plasma membrane and are categorized as low-affinity
systems. 212,231 The AKT1- and KAT1-types of plant channels, similar to the Shaker
channels in animals, contain a pore-forming region conferring ion selectivity. In contrast
to earlier assumptions, these channels are highly selective against Na+,213 and evidence is
lacking for specific regulation under salt stress. We think that the potassium channels play
a minor role in salinity tolerance.
In contrast, K+ transporters which operate at low external potassium may mediate
entry of sodium in saline soil. A high-affinity K+-transporter is known from yeast.214 Some
of the cloned transporters take up potassium with dual-affinity.209,211 A high-affinity K+
transporter from wheat, HKT1, was indicated as a K+/Na+ symporter86 with high-affinity
binding sites for both K+ and Na+. Point mutations, which increased K+ selectivity over
Na+, in one of the 12 transmembrane domains of HKT1 conferred increased salt tolerance
of yeast. Another line of evidence for the involvement of high-affinity K+ uptake system in
salt tolerance came from the study of salt-sensitive mutants. The sos1 mutant of Arabidopsis
thaliana was characterized as hypersensitive to Na+ and Li+ and was unable to grow on low
potassium.92 86Rb uptake experiments showed that sos1 was defective in high-affinity
potassium uptake, and it became deficient in potassium when treated with NaCl. Interestingly
but not surprisingly, expression of the wheat Hkt1 in sos1 mutant plants alleviated the salt-
sensitive phenotype (Schroeder JI, Zhu J-K, personal communication). Further support is
provided by the expression characteristics of a rice homolog of wheat HKT1 in two
varieties that are distinguished by their salinity tolerance. The tolerant variety decreased
expression of the root-specific HKT1 and efficiently excludes sodium, while a salt-sensitive
variety maintained high expression of the HKT1 in the presence of high NaCl.210,215
Irrespective of the indices pointing to the involvement of HKT1-type transporters, or
high-affinity potassium uptake systems in general, in salt tolerance, there are other equally
likely scenarios. First, the presence of sodium is known to interfere with potassium uptake,
as shown for several of the cloned transport proteins, and protective effects exerted by
increased potassium might be based on the nutritional value, and not on a sodium
exclusion mechanism. High sodium sensitivity, as for example shown by the sos1 mutant,
might be due to growth interference when K+ uptake is reduced by the presence of
sodium. In this respect, the improved selectivity of K+ transport systems may increase salt
tolerance, while it is not involved in Na + detoxification or osmotic adjustment. Other
transport systems, finally, might act in sodium uptake. How, for example, the calcium-
regulated outward-rectifying K+-channel KCO1,216 or the regulation of other channels
and transporters, react under sodium stress conditions is unknown. It has been suggested
that sodium might enter through outward-rectifying cation channels.217 Among the many
possibilities, evidence for significant sodium currents through a calcium transporter, LCT1,
exists,218 and hexose and amino acid transporters may also let sodium pass.
Sodium Transport Systems

How sodium enters plant cells, how it enters the plant circulatory system to be
selectively transported over long distances, and how it is partitioned to the vacuole is not
known in detail. Most information is available for the last step in this series: sodium
transport from cytosol to vacuole is accomplished by a sodium/proton antiporter. A
protein of approximately 170 kD219 is a candidate for this tonoplast-located antiporter
based on immunological studies and inhibition of the ameloride-regulated antiport
activity in the presence of the antibody. It will be important to characterize the protein in
detail and to obtain the gene(s), because, when judged by protein size, the putative antiporter
seems to be different from the proteins in bacteria, yeast and vertebrate organisms.
Increased sodium/proton antiport activity during salt stress has been measured in several
model systems, tissues, cells and isolated vacuoles.96,220,221 The increase parallels an
increase in the V-ATPase activity.96,200
Our own data indicate that yet another pathway for sodium uptake may exist. When
analyzing the induction of myo-inositol synthesis in Mesembryanthemum, a surprising
decline of the rate-limiting INPS (myo-inositol-1-phosphate synthase) enzyme in roots
was observed, but the concentration of myo-inositol remained constant in the roots.128,129
This is due to drastically enhanced transport of myo-inositol from the leaves through the
phloem. In addition, myo-inositol is recycled to the leaves through the xylem and the
myo-inositol amount in xylem vessels is correlated with sodium amounts.129 We have cloned
a transcript with homology to vertebrate sodium/myo-inositol and yeast proton/myo-inositol
symporters222 and characterized its activity by complementation of a yeast mutant
defective in myo-inositol uptake.223 It seems possible that such a symporter is responsible
for the excretion of sodium into the xylem, but it is equally possible that sodium/myo-
inositol symport internalizes sodium from the apoplast of the root. The detection of such a
symport mechanism is particularly attractive, considering that the passage of myo-inositol
through the plant circulatory system connects photosynthesis competence with sodium
uptake and transport to mesophyll cells of the leaf.
The Essentiality of Calcium

Increasing calcium improves salinity tolerance of crop plants. Physiological
experiments indicated that the effect is mediated through an increase of intracellular calcium,
changes in vacuolar pH and activation of the vacuolar Na+/H+-antiporter.224, 225 The strict
control over calcium concentrations in the cytosol and calcium storage in a number of
locations (vacuole, mitochondria, endoplasmic reticulum) assign a crucial role to calcium
in plant salinity stress responses.
Recently, an Arabidopsis mutant, sos3, with hypersensitivity to NaCl has been
characterized. The mutant is different from other salt-sensitive mutants92 in that the
phenotype can be masked by the external addition of calcium.226 This phenotype represents
the first mutant with an altered response to calcium in higher plants. The phenotype
reveals the link between calcium and salinity stress tolerance, although the mechanism
through which hypersensitivity and remediation by calcium are connected is not known.
One attractive hypothesis is that a signaling system that responds to calcium spikes at low
calcium concentrations—for example a homolog of the yeast calcineurin-type system—is
defective, and that at higher calcium concentrations a second sensing system can support
the signal and elicit stress defense responses (Zhu JK, personal communication).
Metabolic Engineering of Glycophytic Plants for Increased

Salt Tolerance
In increasing numbers, experiments are reported using transgenic plants for testing
concepts originating from the correlative evidence of physiological analyses. Table 3.1
summarizes some of these reports. The concepts tested target four aspects of tolerance
acquisition:
1. ROS scavenging,
2. Compatible solutes and osmotic adjustment—carbohydrate biosynthesis and
synthesis of charged molecules,
3. Ion balance—potassium uptake vs. sodium uptake, and
4. The synthesis of specific, putatively protective proteins.
A note of caution must be added with respect to the over-expression and accumulation
strategies that have been followed up to now. Too high an accumulation of metabolites, or
too efficient scavenging of H2O2, for example, may not be desirable. When analyzing
transgenic tobacco plants that accumulated sorbitol to extremely high concentrations in
the cytosol, we observed stunted growth and the formation of necrotic lesions that reduced
biomass production, although the plants showed increased salinity and salt stress tolerance.227
The importance of radical oxygen scavenging for preventing oxidative stress in plants
has been demonstrated by genetic engineering of several enzymes into transgenic
plants.179,180,228 Overexpression of superoxide dismutase (Cu/Zn-SOD and Mn-SOD),
ascorbate peroxidase, catalase and glutathione reductase in transgenic plants has already
been shown to lead to increased resistance to oxidative stress.181,229,230,232-238 The most
dramatic protective effect, up until now, was observed after enhancement of the glutathione
cycle.192 In contrast, overexpression of an Fe-SOD in transgenic tobacco neither enhanced
tolerance to chilling-induced photoinhibition in leaf discs nor increased tolerance to salt
stress in whole plants,240 suggesting that isoforms of SOD may have different roles.
Noctor and Foyer20 provided a lucid assessment of the relatively marginal protection
that has been observed in many transgenic plant studies, whether with respect to ROS
scavenging or otherwise. It would certainly be premature to consider the protection
provided by the overexpression of SOD, ASX, or enzymes of the ascorbate/glutathione

cycle as the final word. Protection has typically been observed in strictly controlled
environments, and protective effects have often been marginal. We would like to provide
one consideration as to why this is to be expected. In the case of ASX, at least six different
isoforms exist which are located in mitochondria, in chloroplasts (several, in different
sub-compartments/membranes), soluble in the cytosol, and in the cytoplasmic
endomembrane system.241 A similarly complex distribution has been seen for SOD isoforms
which are found in the cytosol (Cu/Zn-SOD), mitochondria (Mn-SOD) and plastids
(Fe-SOD and Cu/Zn-SOD). Transgenic modifications of single enzymes are likely to have
a minimal effect because of the multitude of compartments that require protection.
Irrespectively, these experiments have clearly shown that—in practically every study—
the engineered expressed transgene elicited some protection. It is now necessary to adopt
multi-gene transfer strategies that alter several components of the stress tolerance system:
1. Targeting, for example, ROS scavenging enzymes to several compartments;
2. Assembling gene constructs that target sodium exclusion and enhanced potassium
uptake;
3. Generating transgenomes in which different pathways are satisfied, for example,
ion homeostasis, carbon allocation, and protein protection simultaneously;
4. Generating transgenomes with strategies that take into account cell-, tissue-,
organ- and developmental specificity.
The last point is particularly important, because little attention has typically been
paid to the “when,” “where,” and “how much” of transgene expression in the presently
concluded transgenic experiments. Significantly more attention needs to be directed to
the promoter elements that drive transgenes. Most attempts have targeted the metabolic
engineering of carbon and nitrogen allocation: ectopic enzyme expression leading to the
synthesis of uncharged carbohydrates—mannitol, sorbitol, trehalose, fructan, and
ononitol—and to glycinebetaine and proline accumulation (Table 3.1). The underlying
mechanism is becoming apparent for some of these strategies, e.g., in the hydroxyl radical
scavenging function of mannitol. 29,30 The mechanisms of protection underlying the
synthesis or presence of chaperones or specific LEA proteins remain to be determined.
Within a very short time, all genes that are essential for the salt tolerance phenotype
shown by some species and all genes that support damage avoidance in sensitive species
will be available. The task remaining, however, is understanding in which metabolic and
signaling pathways the gene products function and in which developmental context stress
protection is necessary. This task will require new approaches. We consider two approaches:
1. Multi-gene transfer into model species—yeast, Arabidopsis, tobacco and rice are
our suggestions; and
2. A focus on metabolic control analysis.
The first strategy utilizes the transfer of all genes, controlled by appropriate promoter
elements, for one or several biochemical pathways to generate protection which can be ana-
lyzed. Through the second approach, a biochemical description of flux in a multitude of
pathways, we will be able to gauge the cost of enzymes/pathways that enhance tolerance in
comparison to the cost and benefits of resident pathways.
Perspectives
High salinity is a major factor responsible for the loss of crop biomass.242 Salinity
caused by irrigation affects many productive agricultural areas. The degeneration of still
productive soils will become a more severe problem in the future. Development of
drought- and salt-tolerant crops has been a major objective of plant breeding programs
for decades in order to maintain crop productivity in semiarid and saline lands. Although
several salt-tolerant varieties have been released, the overall progress of traditional
breeding has been slow and has not been successful.13 The lack of success is mainly due to
the quantitative trait character of salinity tolerance which has to be reconciled with
another multigenic trait, high productivity, which is the ultimate goal of any breeding
program. Marginal progress has equally been grounded in our poor understanding of the
mechanisms of salt tolerance, while the collected body of physiological data has focused
our attention more on details in a large variety of species and less on the principles.
This has changed over the last few years. Biochemical pathways that lead to the
production of compatible solutes such as proline, glycine betaine, DMSP, or pinitol have
been studied and most of the pathway genes have been characterized.6,7,9,14,15 We have the
first glimpses of how the resulting metabolites from such pathways function in protection.
Similarly, the principles of how radical oxygen species act and the principles, genes and
proteins which deter radical damage have emerged. Membrane channels, transporters and
pores are now available through which cells exert control over ion, carbohydrate, amino
acid or water fluxes.58,59,61,146,207,243 We owe most of this recent progress to the power of
the yeast and Arabidopsis thaliana molecular genetic systems. Finding the genes whose
disruptions generate the various mutant phenotypes becomes rapidly easier as additional
mapping data and genomic DNA sequences from Arabidopsis are made available.12
Finally, plant stress perception, and inter- and intracellular signaling of salt stress has
been advanced greatly. Mutants in signal transduction pathways and components of
several signal transduction pathways have been found and are being characterized at
present.244-249 Future studies can follow the blueprint of signaling components isolated
from yeast 10,66,68,88,250 for finding and characterizing homologs of the essential
signaling intermediates in plants. If we accept that a major objective of plant stress
research is application, transgenic crops can be engineered not only for expression of novel
biochemical characters, but also for stress signal transduction that enhances the stress
response inherent to all plants.
Acknowledgments
Because of space constraints a number of references could not be included, and we
apologize. We thank Ms. Pat Adams for help with the manuscript. Different projects have,
off and on, been supported by the US National Science Foundation (Integrative Plant
Biology and International Programs), Department of Energy (Biological Energy), and
Department of Agriculture (NRI). Additional support has been provided by the Arizona
Agricultural Experiment Station, Japan Tobacco Inc., Rockefeller Foundation (New York)
and New Energy Development Organization (Tokyo).
References
1. Cushman JC, Bohnert HJ. Molecular biology of CAM. Plant Physiol 1997; 113:667-676.
2. Shachar-Hill Y. Osmotic coupling and chloride transport in the limonium salt gland.
Proc Roy Soc London-Ser. B: Biol Sci 1990; 241:159-163.
3. Flowers TJ, Troke PF, Yeo AR. The mechanism of salt tolerance in halophytes. Annu Rev
Plant Physiol 1977; 28:89-121.
4. Adams P, Nelson DE, Yamada S et al. Growth and development of Mesembryanthemum
crystallinum (Aizoaceae). New Phytol 1998; 138:171-190.
5. Langdale JA, Nelson T. Spatial regulation of photosynthetic development in C4 plants.
Trends Genet 1991; 7:191-196.
6. McCue KF, Hanson AD. Drought and salt tolerance: Towards understanding and
application. Biotechnol 1990; 8:358-362.
7. Delauney AJ, Verma DPS. Proline biosynthesis and osmoregulation in plants. Plant J
1993; 4:215-223.
8. Ishitani M, Majumder AL, Bornhouser A et al. Coordinate transcriptional induction of

myo-inositol metabolism during environmental stress. Plant J 1996; 9:537-548.
9. Gage DA, Rhodes D, Nolte KD et al. A new route for synthesis of dimethyl-sulfoniopropionate
in marine algae. Nature 1997; 387:891-894.
10. Niu X, Bressan RA, Hasegawa PM et al. Ion homeostasis in NaCl stress environments.
Plant Physiol 1995; 109:735-742.
11. Nelson DE, Shen B, Bohnert HJ. Salinity tolerance-mechanisms, models, and the metabolic
engineering of complex traits, In: Setlow JK, ed. Genetic Engineering, Principles and
Methods. New York: Plenum Press, in press. 1998; 20:153-176.
12. Bevan M, Bancroft I, Bent E et al: Analysis of 1.9 Mb of contiguous sequence from
chromosome 4 of Arabidopsis thaliana. Nature 1998; 391:485-488.
13. Flowers TJ, Yeo AR. Breeding for salinity resistance in crop plants: Where next? Aust J
of Plant Physiol 1995; 22:875-884.
14. Bohnert HJ, Jensen RG. Metabolic engineering for increased salt tolerance-the next step.
Aust J Plant Physiol 1996; 23:661-667.
15. Bohnert HJ, Jensen RG. Strategies for engineering water-stress tolerance in plants. Trends
Biotech 1996; 14:89-97.
16. Bartels D, Nelson DE. Approaches to improve stress tolerance using molecular genetics.
Plant Cell Envir 1994; 17:659-667.
17. Serrano R. Salt tolerance in plants and microorganisms: Toxicity targets and defense
responses. Intl Rev Cytol 1996; 165:1-52.
18. Jain RK, Selvaraj G. Molecular genetic improvement of salt tolerance in plants. Biotech
Annu Rev 1997; 3:245-267.
19. Zhu J-K, Hasegawa PM, Bressan RA. Molecular aspects of osmotic stress in plants. Crit
Rev Plant Sci 1997; 16:253-277.
20. Noctor G, Foyer CH. Ascorbate and glutathione: Keeping active oxygen under control.
Annu Rev Plant Physiol Plant Mol Biol 1998; 49:in press.
21. Levitt J. Responses of Plant to Environmental Stress Chilling, Freezing, and High
Temperature Stresses 2nd ed New York:Academic Press, Inc., 1980.
22. Kaiser WM. Reversible inhibition of the Calvin cycle and activation of oxidative pentose
phosphate cycle in isolated chloroplast by hydrogen peroxide. Planta 1979; 145:377-382.
23. Bieleski RL. Sugar alcohols. In: Loewus FA, Tanner W, eds. Encyclopedia of Plant Physiol
Vol 13A, Plant Carbohydrates I. Berlin:Springer Verlag. 1982:158-192.
24. Yancey PH, Clark ME, Hand SC et al. Living with water stress: Evolution of osmolyte
system. Science 1982; 217:1214-1222.
25. Ford CW. Accumulation of low molecular weight solutes in water stress tropical legumes.
Phytochem 1984; 23:1007-1015.
26. Brown AD, Simpson JR. Water relations of sugar-tolerant yeasts: The role of intracellular
polyols. J Gen Microbiol 1972; 72:589-591.
27. Borowitzka LJ, Brown AD. 1974. The salt relations of marine and halophilic species of
the unicellular green alga, Dunaliella. The role of glycerol as a compatible solute. Arch
Microbiol 1974; 96:37-52.
28. Le Rudulier D, Strom AR, Dandekar et al. Molecular biology of osmoregulation. Science
1984; 224:1064-1068.
29. Shen B, Jensen RG, Bohnert HJ. Increased resistance to oxidative stress in transgenic
plants by targeting mannitol biosynthesis to chloroplasts. Plant Physiol 1997;
113:1177-1183.
30. Shen B, Jensen RG, Bohnert HJ. Mannitol protects against oxidation by hydroxyl radicals.
Plant Physiol 1997; 115:527-532.
31. Schobert B, Tschesche H. Unusual solution properties of proline and its interaction with
proteins. Biochem Biophys Acta 1978; 541:270-277.
32. Smirnoff N, Cumbes QJ. Hydroxyl radical scavenging activity of compatible solutes.
Phytochem 1989; 28:1057-1060.
33. Halliwell B, Gutteridge MC. Role of free radicals and catalytic metal ions in human
disease: An overview. Meth Enzymol 1990; 186:1-85.
34. Orthen B, Popp M, Smirnoff N 1994. Hydroxyl radical scavenging properties of cyclitols.
Proc Royal Soc Edinburgh 1994; 102B:269-272.
35. Pollard A, Wyn Jones RG. Enzyme activities in concentrated solutes of glycine betaine
and other solutes. Planta 1979; 144:291-298.
36. Brown AD. Microbial Water stress Physiology, Principles and Perspectives. New York:
John Wiley & Sons, 1990.
37. Solomon A, Beer S, Waisel Y et al. Effects of NaCl on the carboxylating activity of Rubisco
from Tamarix jordanis in the presence and absence of proline-related compatible solutes.
Plant Physiol 1994;. 90:198-204.
38. Back JF, Oakenfull D, Smith MB. Increased thermal stability of proteins in the presence
of sugars and polyols. Biochem 1979; 18:5191-5196.
39. Paleg LG, Douglas TJ, van Daal A et al. Proline, betaine and other organic solutes protect
enzymes against heat inactivation. Aust J Plant Physiol 1981; 8:107-114.
40. Galinski EA. Compatible solutes of halophilic eubacteria: Molecular principles,
water-solute interaction, stress protection. Experientia 1993; 49:487-496.
41. Papageorgiou G, Murata N. The unusually strong stabilizing effects of glycine betaine on
the structure and function of the oxygen-evolveing photosystem II complex. Photosyn
Res 1995; 44:243-252.
42. Coughlan SJ, Heber U. The role of glycine betaine in the protection of spinach thylakoids
against freezing stress. Planta 1982; 156:62-69.
43. Jolivet Y, Larher F, Hamelin J. Osmoregulation in halophytic higher plants: The protective
effect of glycine betaine against the heat destabilization of membranes. Plant Sci Lett
1982; 25:193-201.
44. Zhao Y, Aspinall D, Paleg LG. Protection of membrane integrity in Medicago sativa L. by
glycine betaine against effects of freezing. J Plant Physiol 1992; 140:541-543.
45. Arakawa T, Timasheff SN. The stabilization of proteins by osmolytes. Biophys J 1985;
47:411-414.
46. Schobert B. Is there an osmotic regulatory mechanism in algae and higher plants? J Theor
Biol 1977; 68:17-26.
47. Dujon B. The yeast genome project: What did we learn? Trends Genet 1996; 12:263-270.
48. Goffeau A, Barrell BG, Bussey H et al. Life with 6000 genes. Science 1996; 274:546-567.
49. Wodicka L, Dong H, Mittman M et al. Genome-wide expression monitoring in Saccha-
romyces cerevisiae. Nature Biotech 1997; 15:1359-1367.
50. Gadd GM, Chalmers K, Reed RH. The role of trehalose in dehydration resistance of
Saccharomyces cerevisiae. FEMS Microbiol Lett 1987; 48:249-254.
51. Mackenzie KF, Singh KK, Brown AD. Water stress hypersensitivity of yeast: Protective
role of trehalose in Saccharomyces cerevisiae. J Gen Microbiol 1988; 134:1661-1666.
52. Albertyn J, Hohmann S, Prior BR. Characterization of the osmotic-stress response in
Saccharomyces cerevisiae: Osmotic stress and glucose repression regulate glycerol-3-
phosphate dehydrogenase independently. Curr. Genet 1994; 25:12-18.
53. Albertyn N, Hohmann S, Thevelein JM et al. GPD1, which encodes glycerol-3-phosphate
dehydrogenase, is essential for growth under osmotic stress in Saccharomyces cerevisiae,
and its expression is regulated by the high-osmolarity osmolarity glycerol response
pathway. Mol Cell Biol 1994; 14:4135-4144.
54. Ahmad I, Larher F, Stewart GR. Sorbitol, a compatible osmotic solute in Plantago
maritima. New Phytol 1979; 82:671-678.
55. Adams P, Thomas JC, Vernon DM et al. Distinct cellular and organismic responses to
salt stress. Plant Cell Physiol 1992; 33:1215-1223.
56. Niu X, Narasimhan ML, Salzman RA et al. NaCl regulation of plasma membrane H+-ATPase
gene expression in a glycophyte and a halophyte. Plant Physiol 1995; 103:713-718.
57. van der Rest ME, Kamminga AH, Nakano A et al. The plasma membrane of Saccharomyces
cerevisiae:Structure, function, and biogenesis. Microbiol Rev 1995; 59:304-322.
58. Anderson JA, Huprikar SS, Kochian LV et al. Functional expression of a probable
Arabidopsis thaliana potassium channel in Saccharomyces cerevisiae. Proc Natl Acad Sci
USA 1992; 89:3736-3740.
59. Sentenac H, Bonneaud N, Minet M et al. Cloning and expression in yeast of a plant
potassium ion transport system. Science 1992; 256:663-665.
60. Frommer WB, Ninnemann O. Heterologous expression of genes in bacterial, fungal,
animal, and plant cells. Annu Rev Plant Physiol Plant Mol Biol 1995; 46:419-444.
61. Rentsch D, Hirner B, Schmelzer E et al. Salt stress-induced proline transporters and salt
stress-repressed broad specificity amino acid permeases identified by suppression of a
yeast amino acid permease-targeting mutant. Plant Cell 1996; 8:1437-1441.
62. Oliver SG. From gene to screen with yeast. Curr Opin Genet Dev 1997; 7(3):405-409.
63. Yagi T, Tada K. Isolation and characterization of salt-sensitive mutants of a salt-tolerant
yeast Zygosaccharomyces rouxii. FEMS Microboil. Lett. 1988; 49:317-321.
64. Ushio K, Ohtsuka H, Nakata Y. Intracellular Na+ and K+ contents of Zygosaccharomyces
rouxii mutants defective in salt tolerance. J Fermen Bioeng 1992; 73:11-15.
65. Garciadeblas B, Rubio F, Quintero FJ et al. Differential expression of two genes encoding
isoforms of the ATPase involved in sodium efflux in Saccharomyces cerevisiae. Mol Gen
Genet 1993; 236:363
66. Brewster JL, De Valoir T, Dwyer ND et al. An osmosensing signal transduction pathway
in yeast. Science 1993; 259:1760-1763.
67. Maeda T, Wurgler-Murphy SM, Saito H. A two-component system that regulates an
osmosensing MAP kinase cascade in yeast. Nature 1994; 369:242-245.
68. Maeda T, Takekawa M, Saito H. Activation of yeast PBS2 MAPKK by MAPKKKs or by
69. Norbeck J, Pahlman A, Akhtar N et al. Purification and characterization of two isoenzymes
of DL-glycerol-3-phosphatase from Saccharomyces cerevisiae. J Biol Chem 1996;
271:13875-13881.
70. Luyten K, Albertyn J, Skibbe WF et al. Fps1, a yeast member of the MIP family of channel
proteins, is a facilitator for glycerol uptake and efflux and is inactive under osmotic stress.
EMBO J 1995; 14:1360-1371.
71. Ansell R, Granath K, Hohmann S, Thevelein JM et al. The two isoenzymes for yeast
NAD-dependent glycerol 3-phosphate dehydrogenase encoded by GPD1 and GPD2 have
distinct roles in osmoadaptation and redox regulation. EMBO J 1997; 16:2179-2187.
72. Shen B, Hohmann S, Jensen RG et al. Roles of sugar alcohols in osmotic stress adaptation-
Replacement of glycerol in yeast by mannitol and sorbitol. To be submitted, 1998.
73. Blomberg A, Adler L. Physiology of osmotolerance in fungi. Adv Microbiol Physiol 1992;
33:145-212.
74. Ota IM, Varshavsky A. A yeast protein similar to bacterial two-component regulators.
Nature 1993; 262:566-569.
75. Posas F, Wurgler-Murphy SM, Maeda T et al. Yeast HOG1 MAP kinase cascade is regulated
by a multistep phosphorelay mechanism in the SLN1-YPD1-SSK1 two-component
osmosensor. Cell 1996; 86:865-875.
76. Schuller C, Brewster JL, Alexander MR et al. The HOG pathway controls osmotic
regulation of transcription via the stress response element (STRE) of the Saccharomyces
cerevisiae CTT1 gene. EMBO J 1994; 13:4382-4389.
77. Hirayama T, Maeda T, Saito H et al. Cloning and characterization of seven cDNAs
for hyperosmolarity-responsive (HOR) genes of Saccharomyces cerevisiae. Mol Gen
Genet 1995; 249:127-138.
78. Varela JCS, Praekelt UM, Meacock PA et al. The Saccharomyces cerevisiae: Hsp12 gene is
activated by the high-osmolarity glycerol pathway and negatively regulated by protein
kinase A. Mol Cell Biol 1995; 15:6232
79. Davenport KR, Sohaskey M, Kamada Y et al. A second osmosensing signal transduction
pathway in yeast: Hypotonic shock activates the PKC1 protein kinase-regulated cell
integrity pathway. J Biol Chem 1995; 270:30157-351
80. Shimizu J, Yoda K, Yamasaki M. The hypo-osmolarity-sensitive phenotype of the
Saccharomyces cerevisiae hop2 mutant is due to a mutation in PCK1, which regulates
expression of β-lucanase. Mol Gen Genet 1994; 242:641-648.
81. Andre B. An overview of membrane transport proteins in Saccharomyces cerevisiae. Yeast

1995; 11:1575-1611.
82. Ko CH, Buckley AM, Gaber RF. TRK2 is required for low affinity K + transport in
Saccharomyces cerevisiae. Genetics 1990; 125:305-312.
83. Ko CH, Gaber RF. TRK1 and TRK2 encode structurally related K + transporters in
Saccharomyces cerevisiae. Mol Cell Biol 1991; 11:4266-4271.
84. Ketchum KA, Joiner WJ, Sellers AJ et al. A new family of outwardly rectifying potassium
channel proteins with two pore domains in tandem. Nature 1995; 376:690-695.
85. Haro R, Banuelos M, Quintero FJ et al. Genetic basis of sodium exclusion and sodium
tolerance in yeast. Plant Physiol 1993; 89:868-874.
86. Rubio F, Gassman W, Schroeder JI. Sodium-driven potassium uptake by the plant
potassium transporter HKT1 and mutations conferring salt tolerance. Science 1995;
270:1660-1663.
87. Gomez MJ, Luyten K, Ramos J. The capacity to transport potassium influences sodium
tolerance in Saccharomyces cerevisiae. FEMS Microbiol Lett 1996; 135:157-160.
88. Mendoza I, Rubio F, Rodriguez-Navarro A et al. The protein phosphatase calcineurin is
essential for NaCl tolerance of Saccharomyces cerevisiae. J Biol Chem 1994; 269:8792-
8796.
89. Gaxiola R, de Larrinoa IF, Villalba JM et al. A novel and conserved salt-induced protein
is an important determinant of salt tolerance in yeast. EMBO J 1992; 11:3157-3164.
90. Ferrando A, Kron SJ, Rios G et al. Regulation of cation transport in Saccharomyces
cerevisiae by the salt tolerance gene HAL3. Mol Cell Biol 1995; 15:5470-5481.
91. Watad AE, Reuveni M, Bressan RA et al. Enhanced net K+ uptake capacity of NaCl-
adapted cells. Plant Physiol 1991; 95:1265-1269.
92. Wu S, Ding L, Zhu J. SOS1, a genetic locus essential for salt tolerance and potassium
acquisition. Plant Cell 1996; 8:617-627.
93. Nelson N. The vacuolar H+-ATPase: One of the most fundamental pumps in nature. J
Exp Biol 1992; 172:19-27.
94. Rea PA, Poole RJ. Vacuolar H+-translocating pyrophosphatase. Annu Rev Plant Physiol
Plant Mol Biol 1993; 44:157-180.
95. Sussman MR. Molecular analysis of proteins in the plant plasma membrane. Annu Rev
96. Barkla BJ, Pantoja O. Physiology of ion transport across the tonoplast of higher plants.
Annu Rev Plant Physiol Plant Mol Biol 1996; 47:127-157.
97. Carmelo V, Bogaerts P, Sa-Correia I. Activity of plasma membrane H+-ATPase and
expression of PMA1 and PMA2 genes in Saccharomyces cerevisiae cells grown at optimal
and low pH. Arch Microbiol 1996; 166:315-320.
98. Estrada E, Agostinis P, Vendenheede JR et al. Phosphorylation of yeast plasma membrane
H+-ATPase by casein kinase I. J Biol Chem 1996; 271:32064-32072.
99. Braley R, Piper PW. The C-terminus of yeast plasma membrane H+-ATPase is essential
for the regulation of this enzyme by heat shock protein HSP30, but not for stress
activastion. FEBS Lett 1997; 418:123-126.
100. Hemenway CS, Dolinski K, Cardenas ME et al. vph6 mutants of Saccharomyces cerevisiae
require calcineurin for growth and are defective in vacuolar H+ -ATPase assembly.
Genetics 1995; 141:833-844.
101. Stevens TH, Forgac M. Structure, function and regulation of the vacuolar (H+)-ATPase.
Annu Rev Cell Dev Biol 1997; 13:779-808.
102. Haro R, Garciadeblas B, Rodriguez-Navarro A. A novel P-type ATPase from yeast
involved in sodium transport. FEBS Lett 1991; 291:189-191.
103. Wieland J, Nitsche AM, Strayle J et al. The PMR2 gene cluster encodes functionally distinct
isoforms of a putative Na + pump in the yeast plama membrane. EMBO J 1995;
14:3870-3882.
104. Prior C, Potier S, Souciet J et al. Characterization of the NHA1 gene encoding a
Na + /H + -antiporter of the yeast Saccharomyces cerevisiae. FEBS Lett 1996; 387:89-93.
105. Jia ZP, McMullough N, Martel R et al. Gene amplification at a locus encoding a putative
Na+/H+ antiport confer sodium and lithium tolerance in fission yeast. EMBO J 1992;
11:1631-1640.
106. Watanabe Y, Miwa S, Tamai Y. Characterization of Na+/H(+)-antiporter gene closely
related to the salt-tolerance of yeast Zygosaccharomyces rouxii. Yeast 1995; 11:839-838.
107. Hahnenberger KM, Jia Z, Young PG. Functional expression of the Schizosaccharomyces
pombe Na+/H+ antiporter gene, sod2, in Saccharomyces cerevisiae. Proc Natl Acad Sci USA
1996; 93:5031-5036.
108. Welihinda AA, Beavis AD, Trumbly RJ. Mutations in LIS1 (ERG6) gene confer increased
sodium and lithium uptake in Saccharomyces cerevisiae. Biochem Biophys Acta 1994;
1193:107-117.
109. Latterich M, Watson MD. Isolation and characterization of osmosensitive vacuolar
mutants of Saccharomyces cerevisiae. Mol Microbiol 1991; 5:2417-2426.
110. Nakamura T, Liu Y, Hirata D et al. Protein phosphatase type 2B (calcineurin)-mediated,
FK506-sensitive regulation of intracellular ions in yeast is an important determinant for
adaptation to high salt stress conditions. EMBO J 1993; 12:4064-4071.
111. Hirata D, Harada S, Namba H et al. Adaptation to high-salt stress in Saccharomyces
cerevisiae is regulated by Ca 2+ /calmodulin-dependent phosphoprotein phosphatase
(calcineurin) and cAMP-dependent protein kinase. Mol Gen Genet 1995; 249:257-264.
112. Posas F, Camps M, Arino J. The PPZ protein phosphatases are important determinants
of salt tolerance in yeast cells. J Biol Chem 1995; 270:13036-13041.
113. Cyert MS, Kunisawa R, Kaim D et al. 1991. Yeast has homologs (CNA1 and CNA2 gene
products) of mammalian calcineurin, a calmodulin-regulated phosphoprotein phosphatase.
Proc Natl Acad Sci USA 1991; 88:7376-7380.
114. Marquez JA, Serrano R. Multiple transduction pathways regulate the sodium-extrusion
gene PMR2/ENA1 during salt stress in yeast. FEBS Lett. 1996; 382:89-92.
115. Sugiura R, Toda T, Shuntoh H et al. Pmp1(+), a suppressor of calcineurin deficiency,
encodes a novel MAP kinase phosphatase in fission yeast. EMBO J 1998; 17:140-148.
116. Rios G, Ferrando A, Serrano R. Mechanisms of salt tolerance conferred by overexpression
of the HAL1 gene in Saccharomyces cerevisiae. Yeast 1997; 13:515-528.
117. Luan S, Li W, Rusnak F, Assmann SM et al. Immunosuppressants implicate protein
phosphatase regulation of K+ channels in guard cells. Proc Natl Acad Sci USA 1993;
90:2202-2206.
118. Liu J. FK506 and cyclosporin, molecular probes for studing intracellular signal
transduction. Immunol Today 1993; 14:290-295.
119. Luan S, Lane WS, Schreiber SL. pCyP B: A chloroplast-localized, heat shock-responsive
cyclophilin from fava bean. Plant Cell. 1994; 6:885-892.
120. Lippuner V, Cyert MS, Gasser CS. Two classes of plant cDNA clones differentially
complement yeast calcineurin mutants and increase salt tolerance of wild type yeast. J
Biol Chem 1996; 271:12859-12866.
121. Mendoza I, Quintero FJ, Bressan RA et al. Activated calcineurin confers high tolerance
to ion stress and alters the budding pattern and cell morphology of yeast cells. J Biol
Chem 1996; 271:23061-23067.
122. Hanson AD, Rathinasabapathi B, Rivoal J et al. Osmoprotective compounds in the
plumbaginaceae: A natural experiment in metabolic engineering of stress tolerance. Proc
Natl Acad Sci USA 1994; 91:306-310.
123. Trossat C, Nolte KD, Hanson AD. Evidence that the pathway of dimethylsulfoniopropionate
biosynthesis begins in the cytosol and ends in the chloroplast. Plant Physiol 1996;
111:965-973.
124. Bray EA. 1997. Plant responses to water deficit. Trends Plant Sci 1997; 2:48-54.
125. Greenway H, Munns R. Mechanisms of salt tolerance in nonhalophytes. Annu Rev Plant
Physiol 1980; 31:149-190.
126. Bohnert HJ, Nelson DE, Jensen RG. Adaptations to environmental stresses. Plant Cell
1995; 7:1099-1111.
127. Vernon DM, Bohnert HJ. A novel methyl transferase induced by osmotic stress in the
facultative halophyte Mesembryanthemum crystallinum. EMBO J 1992; 11:2077-85.
128. Nelson DE, Rammesmayer G, Bohnert HJ. Cell-specific myo-inositol metabolism and
transport in plant salinity tolerance. Plant Cell 1998; 10:753-764.
129. Nelson DE, Koukoumanos M, Bohnert HJ. Myo-inositol is necessary for efficient NaCl
uptake and transport in Mesembryanthemum crystallinum. Plant Physiol 1998; in press.
130. Kiyosue T, Yoshiba Y, Yamaguchi-Shinozaki K et al. A nuclear gene encoding mito-
chondrial proline dehydrogenase, an enzyme involved in proline metabolism, is
upregulated by proline but downregulated by dehydration in Arabidopsis. Plant Cell 1996;
8:1323-1335.
131. Rhodes D, Hanson AD. Quaternary ammonium and tertiary sulfonium compounds in
higher plants. Annu Rev Plant Physiol Plant Mol Biol 1993; 44:357-384.
132. Ishitani M, Nakamura T, Han SY et al. Expression of the betaine aldehyde dehydrogenase
gene in barley in response to osmotic stress and abscisic acid. Plant Mol Biol 1995;
27:307-315.
133. Saneoka H, Nagasaka C, Hahn DT et al. Salt tolerance of glycine betaine-deficient
and -containing maize lines. Plant Physiol 1995; 107:631-638.
134. Russell BL, Rathinasabapathi B, Hanson AD. Osmotic stress induces expression of
choline monooxygenase in sugar beet and amaranth. Plant Physiol 1998; 116:859-865.
135. Kishor PBK, Hong Z, Miao G et al. Overexpresion of pyrroline-5-carboxylate synthetase
increases proline production and confers osmotolerance in transgenic plant. Plant Physiol
1995; 108:1387-1394.
136. Yoshiba Y, Kiyosue T, Katagiri T et al. Correlation between the induction of a gene for
pyrroline-5-carboxylate synthetase and the accumulation of proline in Arabidopsis thaliana
under osmotic stress. J Plant 1995; 7:751-760.
137. Verbruggen N, Hua XJ, May M et al. Environmental and developmental signals modulate
proline homeostasis: Evidence for a negative transcriptional regulators. Proc Natl Acad
Sci USA 1996; 93:8787-8791.
138. Sheveleva E, Chmara W, Bohnert HJ et al. Increased salt and drought tolerance by
D-ononitol production in transgenic Nicotiana tabacum L. Plant Physiol 1997; 115:1211-1219.
139. Malin G, Lapidot A. Induction of synthesis of tetrahydropyrimidine derivatives in
Streptomyces strains and their effect on E. coli in response to osmotic and heat stress. J
Bact 1996; 178:385-395.
140. Louis P, Galinski EA. Characterization of genes for the biosynthesis of the compatible
solute ectoine from Marinococcus halophilus and osmoregulated expression in Escherichia
coli. Microbiol 1997; 143:1141-1149.
141. Holmström KO, Mäntyla E, Welin B et al. Drought tolerance in tobacco. Nature 1996;
379:683-684
142. Chrispeels MJ, Agre P. 1994. Aquaporins: Water channel proteins of plant and animal
cells. Trends Biochem Sci 1994; 19:421-425.
143. Maurel C. Aquaporins and water permeability of plant membranes. Annu Rev Plant
Physiol Plant Mol Biol 1997; 48:399-429.
144. Ishibashi K, Sasaki S, Fushimi K et al. Molecular cloning and expression of a member of
the aquaporin family with permeability to glycerol and urea in addition to water
expressed at the basolateral membrane of kidney collecting duct cells. Proc Natl Acad Sci
USA 1994; 91:6269-6273.
145. Weaver CD, Shomer NH, Louis CF et al. Nodulin 26, a nodule-specific symbiosome
embrane protein from soybean, is an ion channel. J Biol Chem 1994; 269:17858-17862.
146. Weig A, Deswarte C, Chrispeels JJ. The major intrinsic protein family of Arabidopsis
has 23 members form three distinct groups with functional aquaporins in each group.
Plant Physiol 1997; 114:1347-1357.
147. Yang B, Verkman AS. Water and glycerol permeabilities of aquaporins 1-5 and MIP
determined quantitatively by expression of epitope-tagged constructs in Xenopus oocytes.
J Biol Chem 1997; 272:16140-16146.
148. Kaldenhoff R, Källing A, Meyers J et al. The blue light-responsive AthH2 gene of
Arabidopsis thaliana is primarily expressed in expanding as well as in differentiating
cells and encodes a putative putative channel protein of the plasmalemma. Plant J
1995; 7:87-95.
149. DeWitt ND, Sussman MR. Immunocytological localization of an epitope-tagged plasma
membrane proton pump (H+-ATPase) in phloem companion cells. Plant Cell 1995;
7:2053-2067.
150. DeWitt ND, Hong B, Sussman MR et al. Targeting of two Arabidopsis H+-ATPase isoforms
to the plasma membrane. Plant Physiol 1996; 112:833-844.
151. Yamada S, Katsuhara M, Kelly WB et al. A family of transcripts encoding water channel
proteins: Tissue-specific expression in the common ice plant. Plant Cell 1995; 7:1129-1142.
152. Ludevid D, Hoefte H, Himmelblau E et al. The expression pattern of the tonoplast
intrinsic protein γ-TIP in Arabidopsis thaliana is correlated with cell enlargement. Plant
Physiol 1992; 100:1633-1639.
153. Maurel C, Kado RT, Guern J et al. Phosphorylation regulates the water channel activity
of the seed-specific aquaporin γ-TIP. EMBO J 1995; 14:3028-3035.
154. Ikeda S, Nasrallah JB, Dixit R et al. An aquaporin-like gene required for the Brassica
self-incompatibility response. Science 1997; 276:1564-1566.
155. Yamaguchi-Shinozaki K, Koizumi M, Urao S et al. Molecular cloning and characterization
of 9 cDNAs for genes that are responsive to desiccation in Arabidopsis thaliana: Sequence
analysis of one cDNA clone that encodes a putative transmembrane channel protein.
Plant Cell Physiol 1992; 33:217-224.
156. Opperman CH, Taylor CG, Conkling MA. Root-knot nematode-directed expression of a
plant root-specific gene. Science 1994; 263:221-223.
157. Phillips AL, Huttly AK. Cloning of two gibberellin-regulated cDNAs from Arabidopsis
thaliana by subtactive hybridization: Expression of the tonoplast water channel,
gamma-TIP, is increased by GA3. Plant Mol Biol 1994; 24:603-615.
158. Kaldenhoff R, Kaling A, Richter G. Regulation of the Arabidopsis thaliana aquaporin gene
AthH2 (PIP1b). J Photochem Photobiol B 1996; 36:351-354.
159. Yamamoto YT, Cheng CL, Conkling MA. Root-specific genes from tobacco and
Arabidopsis homologous to an evolutionarily conserved gene family of membrane
channel proteins. Nucl Acids Res 1990; 18:7449-7457.
160. Yamada S, Nelson DE, Ley E et al. The expression of an aquaporin promoter from
Mesembryanthemum crystallinum in tobacco. Plant Cell Physiol 1997; 38:1326-1332.
161. Kirch H-H, Michalowski CB, Bohnert HJ. Anti-MIP antibodies recognize distinct cell types
in roots and leaves from Mesembryanthemum crystallinum. Manuscript in preparation, 1998.
162. Verkman AS. Water channels. Austin:RG Landes Company, 1993.
163. King LS, Agre P. Pathophysiology of the aquaporin water channels. Annu Rev Physiol
1996; 58:619-648.
164. Nielsen S, Chou C-L, Marples D et al. Vasoprssin increases water permeability of kidney
collecting duct by inducing translocation of aquaporin-CD water channels to plasma
membrane. Proc Natl Acad Sci USA 1995; 92:1013-1017.
165. Robinson DG, Sieber H, Kammerloher W et al. PIPq aquaproins are concentrated in
plasmalemmasomes of Arabidopsis thaliana mesophyll. Plant Physiol 1996; 111:645-649.
166. Johansson I, Larsson C, Ek B et al. The major integral proteins of spinach leaf plasma
membranes are putative aquaporins and are phosphorylated in response to Ca2+ and
apoplastic water potential. Plant Cell 1996; 8:1181-1191.
167. Buxton GV, Greenstock CL, Helman WP et al. Critical review of rate constants for reactions
of hydrated electrons, hydrogen atoms and hydroxyl radicals in aqueous solution. J Phys
Chem Ref Data 1988; 17:512-579.
168. Asada K. Production and action of active oxygen species in photosynthetic tissues. In:
Foyer CH and Mullineaux PM, eds. Causes of photooxidative stress and amelioration of
defense systems in plants. London: CRC Press, 1994; 77-104.
169. Tenhaken R, Levine A, Brisson LF et al. Function of the oxidative burst in hypersensitive
disease resistance. Proc Natl Acad Sci USA 1995; 92:4158-4163.
170. Lamb C, Dixon RA:The oxidative burst in plant disease resistance. Annu Rev Plant Physiol
Plant Mol Biol 1997; 48:251-275.
171. Mehler AH, Brown AH. Studies on reactions of illuminated chloroplasts. III. Simultaneous
photoproduction and consumption of oxygen studied with oxygen isotopes. Arch Biochem
Biophys 1952; 38:365-370.
172. Robinson JM. Does O2 photoreduction occur within chloroplasts in vivo? Plant Physiol
1988; 72:666-680.
173. Asada K, Takahashi M. Production and scavenging of active oxygen species in
photosynthetic tissues. In: Kyle DJ, Osmond CB, Arntzen CJ, eds. Photoinhibition.
Amsterdam: Elsevier Publ, 1994:227-287.
174. Hodgson RA, Raison JK. Superoxide production by thylakoids during chilling and its
implication in the susceptibility of plants to chilling-induced photoinhibition. Planta 1991;
183:222-228.
175. Kim CS, Jung J. The susceptibility of mung bean chloroplasts to photoinhibition is
increased by an excess supply of iron to plants: A photobiological aspect of iron toxicity
in plant leaves. Photochem Photobiol 1993; 58:120-126.
176. Elstner EF. Metabolism of activated oxygen species. In: Davies DD, ed. Biochemistry of
Metabolism, Vol 11. San Diego: Academic Press Inc, 1987:253-315.
177. Fryer MJ. The antioxidant effects of thylakoid vitamin E (α-tocopheral). Plant Cell Eviron
1992; 15:381-392.
178. Dalton DA. Antioxidant defense of plants and fungi. In: Ahmad S, ed. Oxidant-induced
Stress and Antioxidant Defenses in Biology. New York: Chapman & Hall, 1995:298-355.
179. Foyer CH, Descourvieres P, Kunert KJ. Protection against oxygen radicals: An important
defense mechanism studied in transgenic plants. Plant, Cell Environ 1994; 17:507-523.
180. Foyer CH, Lelandais M, Kunert KJ. Photooxidative stress in plants. Physiol Plant 1994;
92:696-717.
181. Willekens H, Chamnongpol S, Davey M et al. Catalase is a sink for H2 O 2 and is
indispensable for stress defence in C3 plants. EMBO J 1997; 16:4806-4816.
182. Osmond CB, Grace SC. Perspectives on photoinhibition and photorespiration in the field:
Quintessential inefficiencies of the light and dark reactions of photosynthesis? J Exp Bot
1995; 46:1351-1362.
183. Smirnoff N. The role of active oxygen in the response of plants to water deficit and
desiccation. New Phytol 1993; 125:27-58.
184. Wise RR. Chilling-enhanced photooxidation: The production, action and study of
reactive oxygen species produced during chilling in the light. Photosyn Res 1995;
45:79-97.
185. Price AH, Atherton N, Handry GAF. Plants under drought-stress generate activated oxygen.
Free Radical Research Comm 1989; 8:61-66.
186. Quartacci MF, Navari-Izzo F. Water stress and free radical mediated changes in
sunflower seedlings. J Plant Physiol 1992; 139:621-625.
187. Chowdhury SR, Choudhuri MA. Hydrogen peroxide metabolism as an index of water
stress tolerance in jute. Plant Physiol 1985; 65:503-507.
188. Prasad TK, Anderson MD, Martin BA et al. Evidence for chilling-induced oxidative stress
in maize seedling and a regulatory role for hydrogen peroxide. Plant Cell 1994; 6:65-74.
189. Prasad TK. Mechanisms of chilling-induced oxidative stress injury and tolerance in
developing maize seedlings: Changes in antioxidant system, oxidation of proteins and
lipids, and protease activities. J Plant 1996; 10:1017-1026.
190. Price AH, Handry GAF. Iron-catalysed oxygen radical formation and its possible
contribution to drought damage in nine native grasses and three cereals. Plant Cell
Environ 1991; 14:477-484.
191. Moran JF, Becana M, Iturbe-Ormaetxe I et al. Drought induces oxidative stress in pea
plants. Planta 1994; 194:346-352.
192. Roxas VP, Smith RK Jr, Allen ER et al. Overexpression of glutathione S-transferase/glutathione
peroxidase enhances the growth of transgenic tobacco seedlings during stress. Nature
Biotech. 1997; 15:988-991.
193. Guern J, Mathieu Y, Kurkdjian A et al. Regulation of vacuolar pH in plant cells. Plant
Physiol 1989; 89:27-36.
194. Michelet B, Boutry M. The plasma membrane H+-ATPase. Plant Physiol 1995; 108:1-6.
195. Weiss M, Pick U. Primary structure and effect of pH on the expression of the plasma
membrane H+-ATPase from Dunaliella acidophila and Dunaliella salina. Plant Physiol
1996, 112:1693-1702.
196. Reuveni M, Bennett AB, Bressan RA et al. Enhanced H+ transport capacity and ATP
hydrolysis activity of the tonoplast H+-ATPase after NaCl adaptation. Plant Physiol 1990;
94:524-530.
197. Ayala F, O’Leary JW, Schumaker KS. Increased vacuolar and plasma membrane H+-ATPase
activities in Salicornia bigelovii Torr. in response to NaCl. J Exp Bot 1996; 47:25-32.
198. Loew R, Rockel B, Kirsch M et al. Early salt stress effects on the differential expression
of vacuolar H+-ATPase genes in roots and leaves of Mesembryanthemum crystallinum.
Plant Physiol 1996; 110:259-265.
199. Tsiantis M, Bartholomew DM, Smith JAC. Salt regulation of transcript levels for the
c-subunit of a leaf vacuolar H+-ATPase in the halophyte Mesembryanthemum crystallinum.
Plant J 1996; 9:729-736.
200. Barkla BJ, Zingarelli L, Blumwald E, Smith JAC. Tonoplast Na+/H+ antiport activity and
its energization by the vacuolar H+-ATPase in the halophytic plant Mesembryanthemum
crystallinum. Plant Physiol 1995; 109:549-556.
201. Vera-Estrella R, Barkla BJ, Bohnert HJ et al. Salt stress in Mesembryanthemum crystallinum
suspension cells activates adaptive mechanisms identical to those observed in the whole
plant. Planta, in press.
202. Robinson DG. Pyrophosphatase is not (only) a vacuolar marker. Trend Plant Sci 1996;
1:330.
203. Bremberger C, Luettge U. Dynamics of tonoplast proton pumps and other tonoplast
proteins of Mesembryanthemum crystallinum L. during the induction of crassulacean acid
metabolism. Plant 1992; 188:575-580.
204. Matsumoto H, Chung GC. Increase in proton-transport activity of tonoplast vesicles as
an adaptive response of barley roots to NaCl stress. Plant Cell Physiol 1988; 29:1133-1140.
205. Zingarelli L, Anzani P, Lado P. Enhanced K + -stimulated pyrophosphatase activity in
NaCl-adapted cells of Acer pseudoplatanus. Physiol Plant 1994; 91:510-516. 141.
206. Epstein E (1966) Dual pattern of ion absorption by plant cells and by plants. Nature
212:1324-1327.
207. Schachtman DP, Schroeder JI. Structure and transport mechanism of a high-affinity
potassium uptake transporter from higher plants. Nature 1994; 370:655-658.
208. Santa-Maria GE, Rubio F, Dubcovsky J et al. The HAK1 gene of barley is a member of a
large gene family and encodes a high-affinity potassium transporter. Plant Cell 1997;
9:2281-2289.
209. Kim EJ, Kwak JM, Uozumi N et al. AKUP1: An Arabidopsis gene encoding high-affinity
potassium transport activity. Plant Cell 1998; 10:51-62.
210. Golldack D, Su H, Bennett J et al. Differential expression of HKT1-type potassium
transporters in salt-sensitive and salt-tolerant rice lines. Manuscript in preparation, 1998.
211. Fu H-H, Luan S. AtKUP1: A dual affinity K+ transporter from Arabidopsis. Plant Cell
1998; 10:63-73.
212. Maathuis FJM, Verlin D, Smith FA et al. The physiological relevance of Na +-coupled
K +-transport. Plant Physiol 1996; 112:1609-1616.
213. Bertl A, Anderson JA, Slayman CL et al. Use of Saccharomyces cerevisiae for patch-clamp
analysis of heterologous membrane proteins: Characterization of Kat1, an inward-
rectifying K+ channel from Arabidopsis thaliana, and comparison with endogeneous yeast
channels and carriers. Proc Natl Acad Sci USA 1995; 92:2701-2705
214. Gaber RF, Styles CA, Fink GR. TRK1 encodes a plasma membrane protein required for
high-affinity potassium transport in Saccharomyces cerevisiae. Mol Cell Biol 1988;
8:2848-2859.
215. Golldack D, Kamasani U, Quigley F et al. Salt stress-dependent expression of a HKT1-type

high affinity potassium transporter in rice. Plant Physiol 1997; S114:118.
216. Czempinski K, Zimmermann S, Ehrhardt T et al. New structure and function in plant
K+-channels: KCO1, an outward rectifier with a steep Ca2+ dependency. EMBO J 1997;
16:2565-2575.
217. Schachtman DP, Tyerman SD, Terry BR. The K+/Na+ selectivity of a cation channel in
the plasma membrane of root cells does not differ in salt-tolerant and salt-sensitive wheat
species. Plant Physiol 1991; 97:598-605.
218. Schachtman DP, Kumar R, Schroeder JI et al. Molecular and functional characterization
of a novel low-affinity cation transporter (LCT1) in higher plants. Proc Natl Acad Sci
USA 1997; 94:11079-11084.
219. Barkla BJ, Apse MP, Manolson MF et al. The plant vacuolar Na+/H+ antiport. Symp Soc
Exp Biol 1994; 48:141-153.
220. Blumwald E, Poole RJ. Salt tolerance in suspension cultures of sugar beet: Induction of
Na+/H+ antiport activity at the tonoplast by growth in salt. Plant Physiol 1987; 83:884-887.
221. Garbarino J, DuPont FM. NaCl induces Na + /H + antiport in tonoplast vesicles from
barley roots. Plant Physiol 1988; 86:231-236.
222. Cammarata PR, Xu GT, Huang L et al. Inducible expression of Na + /myo-inositol
cotransporter mRNA in anterior epithelium of bovine lens: Affiliation with hypertonicity
and cell proliferation. Exp Eye Res 1997; 64:745-757.
223. Nelson DE, Bohnert HJ. Characterization of the sodium/myo-inositol symporter from
Mesembryanthemum crystallinum. Manuscript in preparation, 1998.
224. Colmer TD, Fan TWM, Higashi RM et al. Interactions of Ca2+ and NaCl stress on the
ion relations and intracellular pH of Sorghum bicolor roots tips: An in vivo 31P-NMR
study. J Exp Bot 1994; 45:1037-1044.
225. Martinez V, Lauchli A. Effects of calcium on the salt-stress response of barley roots as
observed by in vivo phosphorus-31 nuclear magnetic resonance and in vitro analysis.
Plata 1993; 190:519-524.
226. Liu J, Zhu JK. An Arabidopsis mutant that requires increased calcium for potassium
nutrition and salt tolerance. Proc Natl Acad Sci USA 1997; 94:14960-14964.
227. Sheveleva E, Marquez S, Zegeer A et al. Sorbitol dehydrogenase expression in transgenic
tobacco: High sorbitol accumulation leads to necrotic lesions in immature leaves. Plant
Physiol 1998; 117:831-839.
228. Allen RD. Dissection of oxidative stress tolerance using transgenic plants. Plant Physiol
1995; 107:1049-1054.
229. Aono M, Kubo A, Saji H et al. Resistance to active oxygen toxicity of transgenic
Nicotiana tabacum that expresses the gene for glutathione reductase from E. coli. Plant
Cell Physiol 1991; 32:691-697.
230. Aono M, Kubo A, Saji H et al. Enhanced tolerance to photooxidative stress of transgenic
Nicotiana tabacum with high chloroplastic glutathione reductase activity. Plant Cell Physiol
1993; 34:129-135.
231. Maathuis FJM, Ichida AM, Sanders D et al. Roles of higher plant K+ channels. Plant
Physiol 1997; 114:1141-1149.
232. Bowler C, Slooten L, Vandenbranden S et al. Manganese superoxide dismutase can
reduce cellular damage mediated by oxygen radicals in transgenic plants. EMBO J 1991;
10:1723-1732.
233. Bowler C, Van Montagu M, Inze D. Superoxide dismutase and stress tolerance. Annu
Rev Plant Phys Plant Mol Biol 1992; 43:83-116.
234. Gupta AS, Heinen JL, Holaday AS et al. Increased resistance to oxidative stress in
transgenic plants that overexpress chloroplastic Cu/Zn superoxide dismutase. Proc Natl
Acad Sci USA 1993; 90:1629-1633.
235. Van Camp W, Wilekens H, Bowler WH et al. Elevated levels of superoxide dismutase
protect transgenic plants against ozone damage. Biotechnol 1994;.12:165-168.
236. McKersie BD, Chen Y, de Beus M et al. Superoxide dismutase enhances tolerance of
freezing stress in transgenic alfalfa (Medicago sativa L.). Plant Physiol 1993; 103:1155-1163.
237. Foyer CH, Souriau N, Perret S et al. Overexpression of glutathione reductase but not
glytathione synthetase leads to increase in antioxidant capacity and resistance to
photoinhibition in poplar trees. Plant Physiol 1995; 109:1047-1057.
238. McKersie BD, Bowley SR, Harjanto E et al. Water-deficit tolerance and field performance
of transgenic alfalfa overexpressing superoxide dismutase. Plant Physiol 1996;
111:1177-1181.
239. Pitcher LH, Zilinskas BA. Overexpression of copper/zinc superoxide dismutase in the
cytosol of transgenic tobacco confers partial resistance to ozone-induced foliar necrosis.
Plant Physiol 1996; 110:583-588.
240. Van Camp W, Capiau K, Montagu M et al. Enhancement of oxidative stress tolerance in
transgenic tobacco plants overproducing Fe-superoxide dismutase in chloroplasts. Plant
Physiol 1996; 112:1703-1714.
241. Jespersen HM, Kjaersgard IV, Ostergard L et al. From sequence analysis of three novel
ascorbate peroxidases from Arabidopsis thaliana to structure, function and evolution of
seven types of ascorbate peroxidase. Biochem J 1997, 326:305-310.
242. Boyer JS. Plant productivity and environment. Science 1982; 218:443-448.
243. Sauer N, Stolz J. SUC1 and SUC2: Two sucrose transporters from Arabidopsis thaliana;
expression and characrerization in baker’s yeast and identification of the histidine-tagged
protein. Plant J 1994; 6:67-77.
244. Nishihama R, Banno H, Shibata W et al. Plant homologues of components of MAPK
(mitogen-activated protein kinase) signal pathways in yeast and animal cells. Plant Cell
Physiol 1995; 36:749-757.
245. Kakimoto T. CKI1, a histidine kinase homolog implicated in cytokinin signal transduction.
Science 1996; 274:982-985.
246. Ishitani M, Xiong L, Stevenson B et al. Genetic analysis of osmotic and cold stress signal
247. Hirt H: Multiple roles of MAP kinases in plant signal transduction. Trends Plant Sci
1997; 2:11-15.
248. Mizoguchi T, Irie K, Harashida N et al. A gene encoding a mitogen-activated protein
kinase kinase kinase is induced simultaneously with genes for a mitogen-activated
protein kinase and an S6 ribosomal protein kinase by touch, cold, and water stress in
Arabidopsis thaliana. Proc Natl Acad Sci USA 1996; 93:765-769.
249. Mizoguchi T, Ichimura K, Shinozaki K. Environmental stress response in plants: The
role of mitogen-activated protein kinases. Trends in Biotech 1997; 15:15-19.
250. Shinozaki K, Yamaguchi-Shinozaki K. Gene expression and signal transduction in water-
251. Tarczynski MC, Jensen RG, Bohnert HJ. Expression of a bacterial mtlD gene in transgenic
tobacco leads to production and accumulation of mannitol. Proc Natl Acad Sci USA
1992; 89:2600-2604.
252. Tarczynski MC, Jensen RG, Bohnert HJ. Stress protection of transgenic tobacco by
production of the osmolyte, mannitol. Science 1993; 259:508-510.
253. Thomas JC, Sepahi M, Arendall B et al. Enhancement of seed germination in high salinity by
engineering mannitol expression in Arabidopsis thaliana. Plant Cell Environ 1995;
18:801-806.
254. Pilon-Smits EAH, Ebskamp MJM, Paul MJ et al. Improved performance of transgenic
fructan-accumulating tobacco under drought stress. Plant Physiol 1995; 107:125-130.
255. Hayashi H, Alia, Mustardy L et al. Transformation of Arabidopsis thaliana with the codA
gene for choline oxidase; accumulation of glycinebetaine and enhanced tolerance to salt
and cold stress. Plant J 1997; 12:133-142.
256. Xu D, Duan X, Wang B et al. Expression of a late embryogenesis abundant protein gene,
HVA1, from barley confers tolerance to water deficit and salt stress in transgenic rice.
Plant Physiol 1996; 110:249-257.
257. Jones JT, Mullet JE. Developmental expression of a turgor-responsive gene that encodes
an intrinsic membrane protein. Plant Mol Biol 1995; 28:983-996.
CHAPTER 4
Plant Cold Tolerance

Michael F. Thomashow and John Browse
P lants vary greatly in their responses to cold temperatures. At one extreme are many plants
from tropical and subtropical regions which suffer injury when exposed to low nonfreezing
temperatures. These include economically important plants such as cotton, soybean, maize,
rice, and many tropical and subtropical fruits. Such chilling-sensitive plants undergo sharp
reductions in growth rate and development at temperatures between 0˚ and 12˚C.1,2 The
physical and physiological changes in chilling-sensitive plants that are induced by exposure to
low temperatures, together with the subsequent expression of stress symptoms, are termed
chilling injury. The symptoms that are associated with chilling injury include reduced or
retarded germination and seedling emergence, wilting and chlorosis of leaf tissue, electrolyte
leakage and tissue necrosis.
In sharp contrast to plants of tropical origin, those from temperate regions are not
only chilling-tolerant, but many are able to survive freezing. Herbaceous plants from
temperate regions can survive freezing temperatures ranging from -5˚ to -30˚C, depending on
the species, while trees from boreal forests routinely survive winter temperatures below -30˚C.
Significantly, the maximum freezing tolerance of these plants is not constitutive, but is
induced in response to low nonfreezing temperatures (below ~10˚C), a phenomenon known
as “cold acclimation.” For instance, rye plants grown at normal warm temperatures are killed
by freezing below about -5˚C, but after cold acclimation can survive freezing temperatures
down to about -30˚C.
What accounts for the differences in cold tolerance among plant species? Why are
cucumber and rice plants injured at chilling temperatures while cold-acclimated cabbage
and wheat survive freezing below -15˚C? The answers to these questions are of basic scientific
interest and have potential practical applications. Cold temperatures limit the geographical
locations where crop and horticultural plant species can be grown and periodically cause
significant losses in plant productivity. Greater knowledge of the molecular basis of chilling
and freezing tolerance could potentially lead to the development of new strategies to
improve plant cold tolerance, resulting in increased plant productivity and expanded
areas of agricultural production. Here we summarize the current understanding of the
molecular basis for chilling and freezing injury and discuss recent advances in the
identification of genes involved in cold tolerance.
Chilling Tolerance
Role of Membranes in Chilling Injury
The majority of chilling-sensitive plants share a similar threshold for the onset of
low-temperature damage and exhibit a common assembly of symptoms. These observations
have been interpreted by many investigators as indicating that there is a single primary lesion,
or trigger, that initiates cell damage at some critical temperature and leads to a cascade of
secondary events that are the more readily appreciated consequences of chilling damage.2
Several primary lesions have been proposed, but the most widely studied hypotheses
involve temperature-dependent changes in membrane lipid structure.3,4 Early suggestions
envisioned a mechanism in which the lipids in membranes underwent an overall phase
transition from the liquid crystalline (Lα) state to the gel (Lβ) state.1,5 According to this
proposal, the transition from liquid crystalline phase to gel phase would result in alterations in
the metabolism of chilled cells and lead to injury and death of the chilling-sensitive plants. It
was quickly recognized that such a mechanism was an oversimplification6 but it was more
than ten years before a more sophisticated version of the membrane hypothesis was
articulated. Raison and Wright7 observed that small additions of disaturated phospholipids to
preparations of wheat polar lipids could produce entropy changes during differential scanning
calorimetry that were quantitatively similar to those observed for polar lipid extracts from
chilling-sensitive mung bean plants. These experiments suggested that only a portion of
the lipids (4 to 7%) was actually undergoing a phase change in the 0˚ to 12˚C temperature
range.
Meanwhile, Murata and coworkers demonstrated a strong correlation across different
plant species between the degree of chilling sensitivity and the proportion of disaturated
phosphatidylglycerol (PG); molecules that contain only 16:0, 18:0 and 16:1-trans fatty
acids.8 Chloroplast PG is invariably synthesized with 16:0 at the sn-2 position of the glycerol
backbone. Although this 16:0 may be converted to ∆3-16:1-trans, the geometry of this
trans-unsaturated fatty acid is very similar to that of saturated fatty acids. For this reason, the
level of disaturated PG depends on the extent to which the glycerol-3-phosphate
acyltransferase specifically selects 18:1-ACP to the exclusion of 16:0-ACP and 18:0-ACP that
are also available as substrates in the chloroplast stroma.4 Invoking disaturated molecular
species of PG as the cause of chilling sensitivity was attractive because, in contrast to
proposals based on less precise concepts of lipid unsaturation, it provided a mechanism
underpinned by a firm biophysical explanation. Thus, preparations of PG purified from
three chilling-sensitive plants were observed to enter the Lα to Lβ phase transition at 29˚ to
33˚C, whereas PG from chilling resistant plants did not enter the transition until the
temperature was below 15˚C.9 More recently, the molecular-species distribution of PG in
tobacco and Arabidopsis plants has been altered by molecular genetic techniques.10,11 Murata
et al10 transformed tobacco plants with gene constructs encoding glycerol-3-phosphate
acyltransferase from either squash or Arabidopsis. Transgenic plants containing the squash
gene contained elevated levels of disaturated PG (76% of total PG) compared with controls
(36%) and showed more damage after chilling. Conversely, transgenic plants expressing
the Arabidopsis gene contained 28% disaturated PG and showed less chilling damage than
control tobacco plants. One of the measures of chilling injury in this study was the extent
of photoinhibition of photosynthesis. Subsequent studies by Moon et al12 revealed that
there is no difference between the rate at which transgenic and wild type plants undergo
chilling-induced photoinhibition. Rather, the principal effect of the variation in the amount
of disaturated PG seems to be on the rate at which damaged photosystems can be repaired.
The main target for photoinhibition is thought to be the D1 polypeptide of the photosystem II
reaction center.13 When this protein is damaged, presumably by side products from the
photochemical reactions,14 newly synthesized D1 protein is inserted into the photosystem
II complex to restore photochemical activity. Thus, Moon et al12 hypothesize that the
altered amount of disaturated PG has an effect on the rate at which damaged D1 protein is
removed from the photosystem II complex and replaced by synthesis and insertion of new
protein. An important unanswered question is whether the effect of disaturated-PG
Plant Cold Tolerance 63
content on D1 turnover extends to other chloroplast membrane proteins. Conceivably, the

D1 protein is simply an efficient reporter of a general defect in the assembly or removal of
membrane proteins.
Transformation of tobacco plants with a suitably modified version of the des9 gene
from Anacystis nidulans, which encodes a broad-specificity glycerolipid ∆9 desaturase substantially
reduced the level of saturated fatty acids in PG and other membrane lipids.15 Again, the transgenic
plants were more tolerant to chilling treatments than control plants. Interestingly, modest
increases in overall membrane unsaturation also appear to reduce chilling damage in
tobacco. Kodama et al16 used the Arabidopsis FAD7 gene in transgenic tobacco to produce
a modest increase (~10%) in the conversion of dienoic to trienoic fatty acids in membrane
lipids with no apparent change in the level of disaturated PG. Compared with control
plants, the transgenic tobacco exhibited significantly higher leaf expansion rates following
a 7 day exposure to 1˚C. This result is somewhat surprising since the change in lipid
composition would not be expected to influence the tendency of membrane lipids to
undergo a phase transition.
In a similar approach, Wolter et al11 restructured the E. coli plsB gene, which encodes
a membrane-bound glycerol-3-phosphate acyltransferase, so that the peptide was directed
into the chloroplasts of transgenic Arabidopsis that expressed the gene. The bacterial
acyltransferase utilizes both 16:0- and 18:1-ACPs as substrates. The resulting transgenic
Arabidopsis plants contained 48-54% disaturated PG and were damaged by treatment at
4˚C for 7 days. These findings suggest that disaturated PG species can induce low-temperature
sensitivity in a chilling-tolerant plant such as Arabidopsis, although it is not formally
possible to rule out the possibility that accumulation of the plsB protein might have
contributed to the phenotype observed.11
Notwithstanding this accumulated body of evidence, studies on the Arabidopsis fab1
mutant demonstrated that the level of disaturated PG cannot be the sole determinant of
plant chilling sensitivity. fab1 plants contain an increased proportion of 16:0 fatty acids
because of a partial defect in 3-ketoacyl-ACP synthase II, the enzyme responsible for
elongation of 16:0 to 18:0.17 As a consequence, PG from fab1 leaves contains 43%
disaturated molecular species compared with only 9% in PG from the wild type. The
proportion of disaturated PG in fab1 falls close to the middle of the range found for chilling-
sensitive plants and makes the fab1 mutant comparable to species such as castor bean,
cucumber, maize and tobacco.18 However, fab1 plants were able to grow and complete
their life cycle normally at 10˚C. They were also unaffected (as compared with wild type
controls) by more severe chilling treatments that quickly led to the death of cucumber and
other chilling-sensitive plants. These treatments included 4˚C for 7 days in the dark, 2˚C for 7
days in the light and freezing to -2˚C for 24 hours.17 Following each of these treatments,
mutant plants returned to 22˚C remained indistinguishable from wild type controls and
flowered and set seed normally.
fab1 mutant plants do eventually show damage when grown continuously at 2˚C with
reduced photosynthesis, reduced growth and leaf chlorosis developing gradually from 10
to 35 days of low-temperature treatment. At 2˚C, fab1 plants undergo a process of chloroplast
autophagy.19 Therefore, although the fab1 mutant does not exhibit classic chilling sensitivity,
the results of Wu et al19 confirm a deleterious effect of high levels of disaturated PG on
low-temperature fitness and provide a rationale for the relatively low content of disaturated
PG among chilling-tolerant species which may be exposed to low temperatures for
extended periods of the life cycle. Indeed, the chilling-induced chlorosis and slow growth
of the fab1 mutant could be due to a defect in chloroplast membrane protein turnover or
accumulation of the kind described by Moon et al12 However, the side-by-side comparison
of the fab1 mutant with naturally chilling-sensitive species that contain similar levels of
disaturated PG emphasizes the point that factors other than high levels of disaturated PG are
responsible for the injury sustained by these and other chilling-sensitive plants.
In summary, these results make it clear that reducing the proportion of disaturated
PG, and perhaps increasing overall lipid unsaturation, can measurably improve the
low-temperature performance of tobacco plants, possibly through facilitating turnover
of the D1 protein and thereby allowing faster recovery from photoinhibition. At the same
time, disaturated PG cannot be considered the sole cause of chilling sensitivity because
high levels of disaturated PG in Arabidopsis fab1 did not produce a typical chilling-
sensitive phenotype. Nor is the manipulation of membrane lipids the only way to improve
low-temperature performance of tobacco plants. Gupta et al20 produced transgenic
tobacco plants that overexpress a chloroplastic Cu/Zn superoxide dismutase. Leaf disks
from transgenic plants had higher rates of photosynthesis at 10˚C compared with
untransformed controls and also exhibited a greater capacity for recovery at 25˚C after
photoinhibition at 3˚C for 4 hours. These findings imply that protecting tissues from the
effects of oxidative stress may also reduce chilling damage.
The work described here on higher plants is complemented by studies in cyanobacteria.
In these prokaryotes, a high level of saturated fatty acids is correlated with an inability to
grow at low temperatures, either because of reduced processing of D1 protein21 or
reduced activity of nitrate uptake.22 A more complete review of studies in cyanobacteria is
included in Nishida and Murata.23
Additional Mechanisms of Chilling Injury

Although the roles of membrane lipid unsaturation and disaturated PG in chilling
sensitivity have been well established, there are also well-documented examples where other
processes must be responsible. One of these involves the chilling of tomato plants in the
dark, under which condition photoinhibition and considerations of D1 turnover are clearly
not relevant. Tomato plants that have been chilled in the dark show greatly reduced
photosynthesis rates during subsequent illumination. Understanding the possible cause
of this chilling damage started with the observation that many proteins involved in photo-
synthesis are products of genes whose transcriptional activities cycle under control of the
circadian clock. Martino-Catt and Ort24 used genes for the chlorophyll a/b binding
protein of photosystem II (Cab) and for ribulose-1,5-bisphosphate carboxylase/oxygenase
activase (rca) to demonstrate that chilling stops the circadian clock. They discovered
that low temperature has two separate effects on the normal pattern of expression of Cab and
rca proteins:
1. Progression of the timing of the circadian clock controlling gene transcription is
suspended throughout the period of low-temperature exposure; and
2. Normal turnover of the existing transcripts is suspended.
Upon rewarming, the circadian rhythm of transcriptional and translational activity is
reestablished, but is out of phase with the actual time of day by the amount of time that the
tomato plant was at low temperature. In addition, after rewarming, the messages that were
stabilized at low temperature can no longer be translated into protein. From these results, it
is reasonable to suggest that a range of gene-expression and other functions controlled by
the circadian clock may be affected by chilling in tomato. This would clearly be expected to
result in considerable disruption of photosynthesis and other cellular functions.
A few plant species have been demonstrated to have a capacity to increase their chilling
tolerance in response to treatment at modestly cool temperatures. Dark-grown seedlings of the
maize inbred G50 were killed by exposure to 4˚C for 7 days but could be induced to survive
this treatment by prior exposure to 14˚C for 3 days.25 Differential screening techniques
allowed the isolation of cDNAs representing chilling acclimation responsive genes including
Fig.4.1. The effects of chilling

temperature on the growth rates
of Arabidopsis lipid mutants. The
results shown are for (A) fad5 and
fad6 at 5˚C (B); fab1 at 2˚C; and
(C) fad2 at 6˚C. Wild type controls
are included in each experiment.
cat3, which encodes the mitochondrial catalase 3 isozyme. Hydrogen peroxide levels in the
seedlings were increased during acclimation at 14˚C and treatment of seedlings grown at
2˚C with H2O2 induced chilling tolerance and increased both cat3 transcript levels and the
activities of catalase 3 and guaiacol peroxidase. From these results, it appears that peroxide
has dual effects at low temperatures. During acclimation at 14˚C, its early accumulation
signals the production of antioxidant enzymes such as catalase 3 and guaiacol peroxidase.
At 4˚C, in nonacclimated seedlings, it accumulates due to low levels of these, and perhaps
other, antioxidant enzymes and may cause damage through oxidation of lipids and proteins.26
Genes Required for Chilling Tolerance

Much of the discussion of temperature adaptation of plants focuses on finding
defects in chilling-sensitive species that can explain why they are damaged by low tempera-
tures. However, it is probably more useful, especially at the genetic level, to identify the
traits that are responsible for the chilling tolerance observed in temperate plants. Thus, it
is possible to screen mutant populations of a chilling tolerant species such as Arabidopsis
for plants that are no longer fully tolerant to low temperature. Mutants with impaired
chilling tolerance are defined as those which have a wild type appearance at normal growth
temperatures but which show damage when transferred to chilling temperatures. These
mutants each contain a mutation that has no effect at normal temperatures but is disruptive at
chilling temperatures. In such a screen, only mutational defects that sensitize a mutant to
chilling will be identified; mutations associated exclusively with other processes such as
freezing tolerance are excluded. Somerville and coworkers initiated such a mutational
approach.27 They screened a population of Arabidopsis mutated by ethyl methane sulfonate
(EMS). About 20 mutants were isolated that showed damage symptoms in response to a brief
and mild chilling regime; they had a normal, wild type phenotype at 22˚C, but after a week
at 13˚C exhibited visual damage.28,29 An extensive characterization of one mutant, chs1,27-30
reinforces the validity of the mutational approach. The chs1 mutant showed
low-temperature-induced chlorosis indicating a lesion in chloroplast maintenance at low
temperatures. Subsequent investigations revealed a loss of chloroplast integrity30 and
reduced accumulation of proteins localized to the chloroplast.28,29 The detected changes
indicate a sequence of chilling-induced damage caused by disrupted protein accumulation in
the chloroplast. Nevertheless, it has not been possible to identify the precise biochemical
lesion responsible for initiating these changes.
A collection of Arabidopsis mutants with defects in membrane lipid unsaturation31
have offered useful perspectives on the role of membrane unsaturation. Five mutant lines-
fab1 (see above), fad5, fad6, fad2 and the triple mutant—fad3-fad7-fad8—show damage
symptoms when grown at 2˚-6˚C.17,32,33 All of these mutant lines are similar to wild type
when grown at 22˚C, but their growth rates are lower than wild type at chilling temperatures
(Fig. 4.1). However, in all cases (as discussed above for fab1), the low-temperature damage
is distinct from typical chilling sensitivity. For example, symptom development is more
gradual and damage is not exacerbated following a return to 22˚C. These results make it
clear that a suitable membrane lipid composition is required for chilling tolerance, but
that it is unlikely that membrane defects are the sole cause of chilling sensitivity.
A more extreme chilling screen was used by Tokuhisa et al34 who exposed Arabidopsis
plants to 5˚C for up to 42 days and looked for mutants both during the chilling treatment
and after return of the plants to 22˚C. A screen of EMS mutagenized plants using this protocol
identified 3% of the plants as having chilling-induced phenotypes including chlorosis,
reduced growth, necrosis and death. One drawback of EMS mutagenesis is that the
mutations are primarily single base pair substitutions. In many cases, these mutations
destroy the function of the gene product. However, there are many examples where
missense mutations result in an amino acid substitution in the mutated gene such that
the altered polypeptide product functions adequately at normal (permissive) temperatures but
loses function at low (nonpermissive) temperatures. If such a missense mutation is in an
essential gene, the mutation will render a chilling-induced phenotype. Such alleles have
been used extensively in yeast35 and E. coli36 to characterize essential housekeeping genes,
and have been termed cold-sensitive or cs alleles. To circumvent this problem, Tokuhisa et
al34 repeated the screen on a population in which mutations have been generated by
T-DNA insertion.37 Insertion mutagenesis produces a high proportion of null alleles
and will thus facilitate the identification, in our screen, of genes which are unnecessary
at 22˚C but which are essential for proper growth at 5˚C. Just as importantly, the T-DNA
insertion can act as a starting-point to clone and characterize the specific chilling-tolerance
gene. Over 8,000 lines of mutants generated by T-DNA insertional mutagenesis were screened
and about 280 putative mutants were identified. To date, about 200 of these putatives
have been rescreened and 21 mutants have been shown to have heritable, chilling-
impaired phenotypes. Two of these mutants, which exhibited chilling-induced chlorosis
were designated paleface1 (pfc1) and pfc2. A third mutant that was inhibited in leaf
expansion at 5˚C was designated stop1 (sop1). By segregation analysis, each of these
mutants has been shown to have linkage, within 2-3 centiMorgans between the kanamycin
resistance marker in the T-DNA and the chilling-induced phenotype. Therefore, it is
highly probable that the T-DNA in each of these lines is inserted in a gene which is
required for chilling tolerance.
Molecular characterization of the pfc1 mutant has demonstrated a previously
unrecognized requirement for ribosomal RNA processing and modification to provide chilling
tolerance.38 The wild type allele of the mutated gene and a near-full length (>93%) cDNA
clone were isolated by using the T-DNA as a tag. The deduced polypeptide has a 50 amino
acid transit peptide for chloroplast targeting, an S-adenosylmethionine-binding motif and
34% identity with genes from bacteria and yeast encoding ribosomal RNA methylases which
are required for ribosomal RNA processing or translation. The PFC1 transcript was absent
from pfc1 plants and biochemical analyses indicated that the expected methylation of
adenosines 1518 and 1519 in the small subunit rRNA occurred in the wild type but not in
pfc1. Finally, expression of an antisense PFC1 construct in wild type Arabidopsis produced
plants which exhibited the same low-temperature chlorosis seen in the pfc1 mutant. These results
demonstrate that PFC1 function is specifically required for low-temperature tolerance of
Arabidopsis.
Freezing Tolerance
Causes of Freezing Injury
As temperatures drop below freezing, ice forms primarily in the intercellular spaces
(formation of intracellular ice is generally thought to be a fatal event).39 Ice formation is
initiated extracellularly largely due to the apoplastic fluid having a higher freezing point
than the intracellular fluids, but may also involve the relative levels of ice nucleating agents.40
Accumulation of ice in the intercellular spaces can potentially result in physical disruption
of the tissues and cells.41 However, most of the injury is thought to result from the severe
cellular dehydration that occurs with freezing.39,42 The chemical potential of ice is less
than that of unfrozen water at a given temperature. Thus, when ice forms extracellularly,
there is a drop in water potential outside the cell. Consequently, there is movement of
unfrozen water from inside the cell to outside the cell. The net amount of water movement
depends on both the initial solute concentration of the intracellular fluid and the freezing
temperature, which directly determines the chemical potential of the ice. Freezing at -10˚C
results in an osmotic potential of about five osmolar and typically, movement of greater
than 90% of the osmotically active water out of the cell.
Freeze-induced cellular dehydration could have a number of deleterious effects,
resulting in cellular damage such as the denaturation of proteins and precipitation of
solutes. 39,43 However, the best documented injury occurs at the membrane level.42,44
Detailed analyses by Steponkus and colleagues45,46 have demonstrated that multiple forms
of membrane lesions occur in response to freezing. The specific type of membrane damage
depends on the freezing temperature and corresponding severity of cellular dehydration.
At freezing temperatures between about -2˚C and -5˚C, the predominant form of injury in
nonacclimated plants is “expansion-induced lysis”. It results from the cycle of osmotic
contraction and expansion that occurs with freezing and thawing. Specifically, when
protoplasts from leaves of nonacclimated plants are frozen to about -4˚C, they dehydrate, and
as they shrink, endocytotic vesicles bud off from the plasma membrane. When the protoplasts
are thawed and water moves back into the cells, the vesicular material is not reincorporated
into the plasma membranes, resulting in a decrease in membrane surface area. Consequently,
rehydration results in an intolerable osmotic pressure and the cells burst. Freezing of
nonacclimated cells to slightly lower temperatures, approximately -5˚ to -10˚C, results in
another form of membrane damage, lamellar-to-hexagonal-II phase transitions.45,46 In this
case, cells do not burst upon thawing, but instead become osmotically unresponsive due to the
membranes losing their semipermeable characteristics. Freezing cold-acclimated cells to even
lower temperatures, with consequent lower water potentials and more severe dehydration,
results in additional forms of membrane damage, including “fracture jump lesions”.45,46
Mechanisms of Freezing Tolerance

Given the central role of membranes in freezing-injury, it is not surprising that multiple
mechanisms appear to be involved in increasing the cryostability of membranes during cold
acclimation. Steponkus and colleagues45,46 have demonstrated that unlike plasma membranes
from nonacclimated plants, plasma membranes from cold-acclimated plants do not suffer
expansion-induced lysis or formation of hexagonal II phase lipids. The elimination of
these forms of membrane damage involve a number of changes in lipid composition,
including increased levels of fatty acid desaturation in membrane phospholipids.45,46 In
addition, the accumulation of sucrose and other simple sugars that typically occurs with
cold acclimation seems likely to contribute to the stabilization of membranes, as these
molecules can protect membranes against freeze-induced damage in vitro.47,48 Finally, as
discussed below, there is emerging evidence that certain cold-induced hydrophilic
polypeptides help stabilize membranes against freeze-induced injury.
Additional mechanisms could also potentially contribute to freezing tolerance,
including ones that help prevent or reverse freeze-induced denaturation of proteins or
lessen the direct physical damage to cells caused by the accumulation of extracellular ice.
Indeed, molecular chaperones have been shown to accumulate during cold acclimation,
including a spinach Hsp70,49,50 a soybean Hsp7051 and a Brassica napus Hsp90.52 In addition,
there is evidence suggesting that “freeze-inhibitor” sugars might lessen cellular damage by
preventing the formation of adhesions between extracellular ice and the cell walls.53 Also,
recent studies indicate that many plants accumulate antifreeze proteins during cold
acclimation, some of which are present in the apoplastic fluids. 54-56 These proteins
probably do not act by preventing ice formation as they are capable of imparting only a
few tenths of a degree of thermal hysteresis (i.e., lower the freezing temperature by a few
tenths of a degree without affecting the melting point of the solution). However, they
could potentially contribute to freezing tolerance by modifying ice crystal structure and/or
preventing ice recrystallization. In all of these cases, however, further study is required to
clearly establish whether the proteins or sugars contribute significantly to freezing
tolerance.
Role of Cold-Responsive Genes in Freezing Tolerance

In 1985, Guy et al57 established that changes in gene expression occur during cold
acclimation. Since then, a fundamental question in cold acclimation research has been to
determine whether cold-responsive genes have roles in freezing tolerance. To address this
issue, researchers have engaged in the isolation and characterization of genes that are
induced during cold acclimation.58,59 Many of these cold-responsive genes encode proteins
with known activities that could potentially contribute to freezing tolerance. For
instance, the Arabidopsis FAD8 gene60 and barley blt4 genes,61 which encode a fatty acid
desaturase and a putative lipid transfer protein, respectively, are induced in response to low
temperature. These genes might contribute to freezing tolerance by altering lipid composition.
As alluded to above, cold-responsive genes encoding molecular chaperones49-52 might
contribute to freezing tolerance by stabilizing proteins against freeze-induced denaturation.
Cold-responsive genes encoding various signal transduction and regulatory proteins have
also been identified, including MAP kinases,62,63 calcium-dependent protein kinases64,65
and 14-3-3 proteins.66 These proteins might contribute to freezing tolerance by controlling
the expression of cold-responsive genes or by regulating the activity of proteins involved in
freezing tolerance. Whether any of these cold-responsive genes have important roles in freezing
tolerance, however, remains to be determined.
While many of the cold-responsive genes that have been isolated from cold-acclimated
plants encode proteins with known activities, the majority do not. Indeed, most encode
Fig. 4.2. Genomic organization of COR gene families. Coding and intron regions are depicted as
filled and open boxes, respectively. Alternative gene designations are listed. Accession numbers
for the genes are: COR15a, X64138; COR15b, L24070; KIN1, X51474; COR6.6/KIN2, X55053/
X62281; LTI65/RD29b, X67670/D13044; COR78/LTI78/RD29a, L22567/X67071/D13044; LTI29/
ERD10, X90958/D17714; COR47/RD17, X90959/AB004872. The figure was drawn by Kathy
Wilhelm.
extremely hydrophilic polypeptides that are either newly discovered or are homologs of LEA
(late embryogenesis abundant) proteins.59,67 LEA genes are induced late in embryogenesis,
just prior to seed desiccation, and like many of the cold-responsive genes, are induced in
response to dehydration and ABA.68-70 Based largely on these expression characteristics
and the close relationship between freezing and dehydration injury, it has been widely
speculated that the cold-responsive genes encoding the novel hydrophilic and LEA proteins might
contribute to freezing tolerance. Indeed, recent results provide direct evidence that the
Arabidopsis COR (cold-regulated) genes contribute to the increase in freezing tolerance
that occurs with cold acclimation.71,72
Arabidopsis COR Genes

The Arabidopsis COR genes—also designated LTI (low temperature induced), KIN (cold-
inducible), RD (responsive to desiccation) and ERD (early dehydration-inducible)—comprise
four gene families.67 Each family is composed of two genes that are physically linked in the
genome in tandem array (Fig. 4.2). The COR78, COR15, and COR6.6 gene pairs encode newly
discovered hydrophilic polypeptides, while the COR47 gen pair encodes homologs of LEA group
II proteins (also known as dehydrins and LEA D11 proteins).68,69 At least one member of
each gene pair is induced in response to low temperature, dehydration and exogenous
application of ABA.58
To determine whether COR15a might have a role in freezing tolerance, Artus et al.71
constructed transgenic plants that constitutively express the gene and assessed the effects
that this had on freezing tolerance. COR15a encodes a 15 kDa polypeptide that is targeted
to the stromal compartment of chloroplasts.73,74 The mature 9.4 kDa polypeptide,
COR15am, is extremely hydrophilic and, like the other COR polypeptides and many LEA
proteins, has the unusual property of remaining soluble upon boiling in aqueous buffer. In
the initial experiments, Artus et al71 compared the freezing tolerance of chloroplasts in
nonacclimated transgenic and wild type plants. The results indicated that the
COR15am-containing chloroplasts in transgenic plants were 1 to 2˚C more freezing
tolerant than were the chloroplasts in wild type plants that did not contain COR15am
(cold acclimation increased chloroplast freezing tolerance about 6˚C). In additional
experiments, they found that the effects of COR15am were not limited to the chloroplasts.
Protoplasts isolated from leaves of the nonacclimated transgenic plants that constitutively
produced COR15am were about 1˚C more freezing tolerant at freezing temperatures
between -5 and -8˚C than were those isolated from nonacclimated wild type plants.
Significantly, protoplast survival was measured by vital staining with fluorescein diacetate,
a method that reports on retention of the semipermeable characteristic of the plasma
membrane. Thus, it could be concluded from the protoplast survival experiments that
constitutive expression of COR15a resulted in an increase in plasma membrane
cyrostability.
The results of Artus et al71 indicate a role for COR15a in freezing tolerance. However,
unlike cold acclimation which increases protoplast survival over the range of -2˚ to -8˚C,
expression of COR15a only increased survival over the temperature range of -5˚ to -8˚C (if
anything, COR15a expression resulted in a slight decrease in protoplast survival between
-2˚ and -4˚C). A possible explanation for this finding is that COR15a expression might
prevent certain membrane lesions, but not others. As discussed earlier, the predominant
form of membrane injury over the range of -2˚ to -4˚C appears to be expansion-induced
lysis, while over the range of -5˚ to -8˚C, the predominant form of injury is freeze-induced
lamellar-to-hexagonal II phase transitions. Thus, it is possible that constitutive expression
of COR15a might defer the incidence of freeze-induced formation of hexagonal II phase lipids
to a lower temperature, but have little or no effect on the incidence of expansion-induced lysis.
Additional experimentation is required to test this hypothesis.
The mechanism by which COR15a stabilizes membranes against freeze-induced
injury is not yet known. It seems unlikely that the COR15am protein has enzymatic activity,
as it has a very simple amino acid composition and structure: It is rich in alanine (21%),
lysine (18%),glutamic acid (15%) and aspartic acid (10%) residues (which comprise greater
than 60% of the protein); is devoid of proline, methionine, tryptophan, cysteine, arginine
and histidine residues; and is comprised largely of a 13 amino acid sequence that is
repeated (imperfectly) four times. This, however, leaves open many possibilities. COR15am
may act indirectly to stabilize membranes. For example, it could potentially regulate the
activity of proteins that have roles in freezing tolerance, such as enzymes involved in sugar
or lipid metabolism. Alternatively, COR15am might interact directly with the chloroplast
envelope and increase membrane cryostability in some manner. The location of COR15am
within the chloroplast is not necessarily inconsistent with protection of the plasma membrane,
as formation of the hexagonal II phase is an interbilayer event that occurs largely between the
plasmalemma and the chloroplast envelope. Decreasing the propensity of the chloroplast
envelope to fuse with the plasma membrane could result in less damage to the plasma
membrane. Experiments to detect a direct effect of COR15am on the stabilization of
membranes, however, have yielded equivocal results.75,76
Fig. 4.3. CBF1 and COR transcript levels in nonacclimated wild-type Arabidopsis plants (RLD)
and nonacclimated transgenic Arabidopsis plant lines that overexpress either CBF1 (A6 and B16)72
or COR15a (T8).71 Overexpression of CBF1 and COR15a was accomplished by transforming
wild type RLD plants with hybrid gene constructs having the coding sequences for either CBF1
or COR15a under control of the cauliflower mosaic virus 35S promoter.71,72 Total RNA was
prepared from leaves of nonacclimated plants and analyzed for CBF1 and COR transcripts by
RNA blot analysis using 32P-radiolabeled probes.71,72 Overexpression of CBF1 results in the
stimulation of COR gene expression, but does not affect the transcript levels of eIF4A (eukaryotic
initiation factor 4A),92 a constitutively expressed gene that is not responsive to low temperature.
Although constitutive expression of COR15a enhances freezing tolerance at both the

organelle (chloroplast) and cellular (protoplast) level, the effects are modest.71 Moreover,
unlike cold acclimation, COR15a expression alone does not result in a detectable increase
in freezing survival of whole plants.72 These findings are not surprising given the results of
genetic analyses indicating that freezing tolerance is a multigenic trait involving genes with
additive effects.77 Indeed, multiple genes are activated with cold acclimation in Arabidopsis,
including at least one member of each of the four COR gene pairs.58
If multiple COR genes act in concert to increase freezing tolerance, then expression of
the entire COR gene “regulon” would presumably increase freezing tolerance more than
expressing COR15a alone. This hypothesis was recently tested by Jaglo-Ottosen et al.72
Expression of the entire battery of COR genes was accomplished by overexpressing the
Arabidopsis transcriptional activator CBF1 (CRT/DRE binding factor 1).78 CBF1 binds to
a DNA regulatory element, the CRT (C-repeat)/DRE (drought responsive element), that
stimulates transcription in response to both low temperature and water deficit.79 The element
is present in the promoters of COR15a, COR78, COR6.6, COR47 and presumably other yet to
be identified COR genes. Jaglo-Ottosen et al72 found that constitutive overexpression of CBF1
induce expression of the COR genes in nonacclimated Arabidopsis plants (Fig. 4.3) and
increased freezing tolerance at the whole plant level, an effect that was not observed by
expressing COR15a alone. Thus, it appears that additional members of the Arabidopsis
CRT/DRE regulon are freezing tolerance genes that have roles in cold acclimation.
Determining which CRT/DRE-regulated genes have roles in freezing tolerance and
their functions are now important goals. In addition, a critical point to establish is whether
the CRT/DRE-containing COR genes regulate the full array, or only a subset, of the
biochemical changes that occur with cold acclimation (alterations in lipid composition,
accumulation of sugars, synthesis of anthocyanin, etc.).
Other Possible Freezing Tolerance Proteins

More than 20 years ago, Volger and Heber80 reported that cold-acclimated spinach
and cabbage synthesize polypeptides that are highly effective in protecting isolated thylakoid
membranes against freeze-thaw damage in vitro. These putative cryoprotective polypeptides
were detected in cold-acclimated plants, but not nonacclimated plants, suggesting that
they were encoded by cold-regulated genes. Subsequent studies by Hincha and colleagues81,82
indicated that the cryoprotective polypeptides act to protect membranes against freeze-
induced damage by reducing membrane permeability during freezing and increasing
membrane expandability during thawing. A significant limitation in all of these studies,
however, was that only partially purified protein preparations were used. Thus, it was
unclear whether the cryoprotective activity detected was due to a single protein or
multiple polypeptides. Interestingly, however, from the enrichment procedures used, it
was evident that the polypeptides, like the COR polypeptides, were very hydrophilic and
remained soluble upon boiling.
A significant advance in the study of the spinach and cabbage cryoprotective proteins
was recently made by Sieg et al.83 These investigators purified a single cryoprotective protein
from cold-acclimated cabbage that is effective in protecting isolated thylakoids against
freeze-thaw damage in vitro. This protein, which was designated “cryoprotectin,” has a mass of 7
kDa, remains soluble upon boiling and appears to be encoded by a cold-inducible gene
(the protein is present in cold-acclimated plants, but not in nonacclimated plants).
Unfortunately, there is no information on the amino acid sequence of cryoprotectin,
and thus, it is unknown whether it is related to any of the hydrophilic polypeptides
encoded by the cold-responsive genes described above. Additional investigation should
reveal more about the nature of cryoprotectin, its mode of action in vitro, and provide
direct evidence whether it has a role in protecting membranes against freezing-injury
in vivo.
There is evidence accumulating that suggests certain LEA proteins may also contribute to
freezing tolerance. The HVA1 gene of barley, which encodes a LEA group III protein (also
known as LEA D7 proteins), is expressed in aleurone layers late in embryogenesis and in
seedlings in response to low temperature, ABA and water deficit.84 Although there is no
direct evidence that HVA1 expression contributes to increased freezing tolerance, recent
results indicate that the gene is able to confer tolerance to dehydration stress. Xu et al85
have reported that expression of HVA1 in transgenic rice results in increased tolerance to
Table 4.1 Phenotypes associated with sfr mutations in cold-acclimated plants.88,89
Mutant Freezing Anthocyanin Glucose and Fatty Acid

Gene Sensitivity Level (% wt) Sucrose Levels Composition
sfr1 Young leaves 162 Normal No changes detected
sfr2 All leaves (severe) 87 Normal No changes detected
sfr4 All leaves (severe) 8 Reduced amounts of Reduced levels of 16:0,

both glucose (<<10% wt) 18:1 and 18:2 fatty acids
and sucrose (<25% wt)
sfr5 All leaves 115 Normal No changes detected
sfr6 Worst in young leaves 43 Normal No changes detected
both water deficit and high salinity stress. Given the relationship between dehydration
tolerance and freezing tolerance, HVA1 seems to be a strong candidate for having a role in
freezing tolerance. Similarly, Iu et al86 have recently shown that expression of the tomato
Le25 gene, which encodes a LEA D113 protein, increases both the freezing and high salinity
tolerance of yeast cells. Thus, homologs of Le25 would seem to be good candidates for
being freezing tolerance genes. In tomato, which is a chilling sensitive plant that does not
cold acclimate, Le25 is expressed at very low levels (if at all) in response to low temperature.87
Whether it is expressed at high levels at low temperature in plants that cold acclimate
remains to be determined.
Isolation of Mutants Affected in Freezing Tolerance

A powerful approach to identify freezing tolerance genes is to isolate and characterize
mutants that are altered in their ability to survive freezing. Warren and colleagues88,89 and
Xin and Browse90 have recently made exciting progress in this area. The specific approach
taken by Warren and colleagues88,89 has been to isolate Arabidopsis mutants that are less
freezing tolerant than wild type plants after a period of cold acclimation. They screened
M3 seed pools from 1804 chemically mutagenized M2 Arabidopsis plants for mutants with
a decreased capacity to cold acclimate. These efforts resulted in the identification of 13
mutant lines that were defective in freezing tolerance. Seven of these lines displayed chilling
sensitive phenotypes that might have indirectly resulted in a diminished capacity to cold
acclimate. However, six of the lines were able to withstand long periods of cold acclimation
(up to 56 days) without displaying any obvious adverse effects. Thus, in these cases, it would
appear that the mutations have direct effects on the cold acclimation process itself. Genetic
complementation analysis indicated that these lines had suffered mutations in five
SFR (sensitivity to freezing) genes—SFR1, 2, 4, 5 and 6 (two sfr5 mutant alleles were isolated).
Wild type Arabidopsis plants that are cold-acclimated for 2 weeks at 4˚C suffer no
obvious damage upon being frozen at -6˚C for 24 hours followed by transfer to normal
growth temperature. In contrast, Arabidopsis plants carrying the sfr1, 2, 4, 5-1, 5-2 and 6
mutations do suffer injury. The injury observed varies with the different mutations (Table
4.1). The sfr1 mutation affects the freezing tolerance of only young leaves; the sfr6 mutation has
its most severe effects on young leaves, but affects all leaves to some extent; and the sfr2, 4,
sfr5-1 and sfr5-2 mutations affect all leaves equally. Significantly, all of the mutations affect
the cryostability of the plasma membrane, as indicated by the electrolyte leakage test. In
this test, detached leaves are frozen to various temperatures below zero and after thawing,
cellular damage is assessed by measuring electrolyte leakage. Leakage of ions from the cells
is an indication that the semipermeable nature of the plasma membrane has been lost, at
least transiently, in response to freezing. With the sfr1, 4, 5 and 6 mutations, the severity of
the freezing damage observed in the whole plant freezing tests corresponded with the
results of the electrolyte leakage test. Thus, the freezing sensitivity caused by these mutations
appears to result largely from a decrease in membrane cryostability. In contrast, the sfr2
mutation resulted in severe injury in the whole plant freeze test, but only minor damage in
the electrolyte leakage test. Thus, it was suggested89 that the freezing-sensitive lesion caused
by this mutation might not have a primary effect on cellular membranes.
The identity of the sfr genes are not yet known, nor is the molecular basis for how
mutations in the genes affect freezing tolerance. The sfr2, sfr5-1 and sfr5-2 mutations do
not have any obvious effects on the alterations in fatty acid composition and increases in
sucrose and anthocyanin levels that normally occur with cold acclimation (Table 4.1). The
sfr4 mutation, however, results in reduced accumulation of sucrose, glucose and anthocyanin
and lowered levels of 18:1 and 18:2 fatty acids (Table 4.1). Given the likely role of sugars as
cryoprotectants and demonstrated roles of fatty acid composition in membrane
cryostability, it is reasonable to speculate that the effects that the sfr4 mutation has on
sugar and fatty acid composition account, at least in part, for the freezing sensitive phenotype
of these mutants. A role for anthocyanins in freezing tolerance is less clear. Interestingly,
however, the sfr1 and sfr6 freezing sensitive mutations also affect anthocyanin production;
they cause increased and decreased accumulation, respectively (Table 4.1). This suggests
some link between anthocyanin biosynthesis and freezing tolerance, but the nature of that
link is obscure. Finally, both the sfr4 and sfr6 mutations have phenotypes that are not
directly associated with low temperature, namely slow growth of plants at the seedling
stage and racemes having a “bottle brush” appearance, respectively.88 The isolation and
identification of the products encoded by the sfr genes should provide significant new
insight into our understanding of the cold acclimation response.
Another gene that has a major effect on freezing tolerance, eskimo1 (esk1), has
recently been identified by Xin and Browse.90 These investigators screened 800,000 chemically
mutangenized M2 seedlings of Arabidopsis for mutants that displayed “constitutive” freezing
tolerance, i.e., mutants that were more freezing tolerant than wild type plants without cold
acclimation. A number of such mutants were isolated, the best characterized being esk1.
Whereas nonacclimated wild type plants were found to have an LT50 of -2.8˚C in a whole
plant freeze test, the esk1 mutant plants had an LT50 of -10.6˚C. Moreover, the esk1 mutation
increased the freezing tolerance of cold-acclimated plants. Wild type plants that had been
cold-acclimated had an LT50 of -12.6˚C, while cold-acclimated esk1 plants had an LT50
of -14.8˚C.
The molecular basis for the increase in freezing tolerance displayed by the esk1
mutation is not yet certain. However, the esk1 mutation has been shown to have a dramatic
effect on proline concentration; the proline levels in the esk1 mutant are about 30-fold
higher than they are in wild type plants.90 It seems likely that this contributes to the
increased freezing tolerance of the esk1 plants, as proline has been shown to be an effective
cryoprotectant in vitro.91 In addition, total sugars are elevated in the esk1 mutant about
two-fold and expression of the RAB18 cold-responsive LEA group II gene is elevated about
three-fold. These alterations may also contribute to the increase in freezing tolerance.
Perhaps the most important observation made in regard to the esk1 mutation, however, is
that it does not affect the expression of any of the COR genes; expression of COR15a,
COR6.6, COR47 and COR78 remains at low levels under normal growth conditions and is
greatly induced in response to low temperature. These results suggest that there may be
multiple signaling pathways involved in activating different aspects of the cold acclimation
response and that activation of one pathway can result in considerable freezing tolerance
without activation of the other pathways. As discussed above, overexpression of the CBF1
transcription factor induces expression of the CRT/DRE gene regulon and results in a
significant increase in freezing tolerance. The “CBF1 pathway” might therefore control one
set of cold acclimation responses. Similarly, the ESK1 gene may participate in the control of
another set of freezing tolerance responses that includes synthesis of proline, and to a
lesser degree, synthesis of sugars and expression of RAB18. The mechanism of ESK1
action is not known. However, the fact that the two available esk1 alleles are recessive suggests
that ESK1 might act as a negative regulator. Further insight into the nature and function of
ESK1 will come from isolation of the ESK1 gene and characterization of the encoded gene
product.
Conclusions and Perspectives

In a short review such as this, it is impossible to do justice to all of the significant
investigations that have been made regarding plant cold tolerance. We, therefore, chose to
focus primarily on recent advances in our understanding of chilling sensitivity and freezing
tolerance at the molecular and genetic levels. From the studies presented on chilling sensitivity,
we suggest the following hypotheses and framework for considering chilling responses in
plants. For plants that evolved in consistently warm habitats, there has been no selection
against traits that compromise low-temperature growth. Thus, many such traits may have
evolved in tropical and subtropical species, especially if they confer even a small selective
advantage at higher temperatures. Some of these acquired characteristics might affect plant
performance only after extended cold treatment, whereas others might result in damage on
the much shorter time scale normally associated with chilling sensitivity. The progressive
dispersal of angiosperms to temperate climates would have required the elimination of all
traits that affected plant performance in the new, periodically cooler environment. An
important corollary of this proposal is that any chilling-sensitive species is likely to possess
multiple traits that restrict its geographical range. This suggestion is supported by the
examples we have described and by the results of plant breeding experiments indicating
that improvements in chilling tolerance are typically multigenic and may be specific
to particular stages in the plant life cycle. A second corollary is that any particular chilling-
associated trait may not be found in all chilling-sensitive species because each trait may
have arisen more than once during evolution.
In regard to freezing tolerance, it had been widely speculated for more than a decade
that the changes in gene expression that occur with cold acclimation might contribute to
the increased freezing tolerance displayed by cold-acclimated plants. The recent results of Artus
et al71 and Jaglo-Ottosen et al72 indicate that this is indeed the case. Thus, the fundamental
issue of whether cold-responsive genes have roles in freezing tolerance now shifts to identifying
which cold-responsive genes have key roles in cold acclimation and determining their specific
modes of action. In Arabidopsis, it will be important to determine which members of the
CRT/DRE regulon contribute to freezing tolerance and to establish their functions. It will also
be of fundamental interest to determine whether the CRT/DRE regulon is highly conserved
among plants and if so, whether it is regulated by CBF1-related transcription factors.
Finally, it will be important to determine the extent to which the cold acclimation
response is conditioned by the CRT/DRE regulon, i.e., determine how much of the
increase in freezing tolerance that occurs with cold acclimation is due to action of
CRT/DRE-regulated genes. That the CRT/DRE regulon may account for only a portion of
the increase in freezing tolerance is suggested by the study of Xin and Browse;90 analysis of
the esk1 mutants suggests the possibility that cold acclimation involves multiple cold-
activated processes that are controlled by different signal transduction pathways and
that each process/pathway contributes to freezing tolerance in an independent
fashion.The investigations now underway detailing the function and regulation of cold-
responsive genes and the isolation and characterization of mutants altered in freezing tolerance
should provide, in the near future, fundamental insight into the number and nature of
“freezing tolerance regulons” and the signaling pathways that control their expression.
References
1. Lyons JM. Chilling injury in plants. Annu Rev Plant Physiol 1973; 24:445-466.
2. Wang CY. Chilling injury of Horticultural Crops. Boca Raton:CRC Press, 1990.
3. Lyons JM, Graham D, and Raison JK. Low Temperature Stress in Plants. New York:Academic
Press, 1979.
4. Murata N, Nishida I. Lipids in relation to chilling sensitivity of plants. In: Wang CY, ed.
Chilling injury of Horticultural Crops. Boca Raton:CRC Press, 1990:181-199.
5. Raison JK. The influence of temperature-induced phase changes on the kinetics of respira-
tory and other membrane associated enzyme systems. Bioenergetics 1973; 4:285-309.
6. Raison JK, Orr GR. Proposals for a better understanding of the molecular basis of chilling
injury. In: Wang CY, ed. Chilling injury of Horticultural Crops. Boca Raton:CRC Press,
1990:145-164.
7. Raison, JK, Wright LC. Thermal phase transitions in the polar lipids of plant membranes:
Their induction by disaturated phospholipids and their possible relation to chilling injury.
Biochim Biophys Acta 1983; 731:69-74.
8. Murata N, Sato N, Takahashi N et al. Compositions and positional distributions of fatty
acids in phospholipids from leaves of chilling-sensitive and chilling-resistant plants. Plant
Cell Physiol 1982; 23:1071-1079.
9. Murata N, Yamaya J. Temperature-dependent phase behavior of phosphatidylglycerols from
chilling-sensitive and chilling-resistant plants. Plant Physiol 1984; 74:1016-1024.
10. Murata N, Ishizaki-Nishizawa O, Higashi, S. et al. Genetically engineered alteration in the
chilling sensitivity of plants. Nature 1992; 356:313-326.
11. Wolter FP, Schmidt R, Heinz E. Chilling sensitivity of Arabidopsis thaliana with genetically
engineered membrane lipids. EMBO J 1992; 11:4685-4692.
12. Moon BY, Higashi S-I, Gombos Z, et al. Unsaturation of the membrane lipids of chloro-
plasts stabilizes the photosynthetic machinery against low-temperature photoinhibition
in transgenic tobacco plants. Proc Natl Acad Sci USA 1995; 92:6219-6223.
13. Aro E-M, McCaffery S, Anderson JM. Recovery from photoinhibition in peas (Pisum
sativum l.) acclimated to varying growth irradiances. Role of D1 protein turnover. Plant
Physiol 1994; 104:1033-1041.
14. Aro E-M, Virgin I, Andersson B. Photoinhibition of photosystem II-inactivation, protein
damage and turnover. Biochim Biophys Acta 1993; 1143:113-134.
15. Ishizaki-Nishizawa O, Fujii T, Azuma M, et al. Low-temperature resistance of higher plants
is significantly enhanced by a nonspecific cyanobacterial desaturase. Nature Biotech 1996;
14:1003-1006.
16. Kodama H, Hamada T, Horiguchi G et al. Genetic enhancement of cold tolerance by expression
of a gene for chloroplast (∆−3 fatty acid desaturase in transgenic tobacco. Plant Physiol
1994; 105:601-605.
17. Wu J, Browse J. Elevated levels of high-melting-point phosphatidylglycerols do not induce
chilling sensitivity in a mutant of Arabidopsis. The Plant Cell 1995; 7:17-27.
18. Murata N. Molecular species composition of phosphatidylglycerols from chilling-sensitive
and chilling-resistant plants. Plant Cell Physiol 1983; 24:81-86.
19. Wu JW, Lightner J, Warwick N et al. Low-temperature damage and subsequent recovery
of fab1 mutant Arabidopsis exposed to 2˚C. Plant Physiol 1997; 113:347-356.
20. Gupta AS, Heinen JL, Holaday AS et al. Increased resistance to oxidative stress in transgenic
plants that overexpress chloroplastic Cu/Zn superoxide dismutase. Proc Natl Acad Sci USA
1993; 90:1629-1633.
21. Kanervo E, Aro E-M, Murata N. Low unsaturation level of thylakoid membrane lipids
limits turnover of the D1 protein of photosystem II at high irradiance. FEBS Lett 1995;
364:239-242.
22. Sakamoto T, Bryant DA. Growth at low temperature causes nitrogen limitation in the
cyanobacterium synechococcus sp. PCC 7002. Arch Microbiol 1998; 169:10-19.
23. Nishida I, Murata N. Chilling sensitivity in plants and cyanobacteria. The crucial contri-
bution of membrane lipids. Annu Rev Plant Physiol Plant Mol Biol 1996; 47:541-568.
24. Martino-Catt S, Ort DR. Low temperature interrupts circadian regulation of transcriptional
activity in chilling-sensitive plants. Proc Natl Acad Sci USA 1992; 89:3731-3735.
25. Prasad TK, Anderson MD, Martin BA et al. Evidence for chilling-induced oxidative stress
in maize seedlings and a regulatory role for hydrogen peroxide. Plant Cell 1994; 6:65-74.
26. Prasad TK. Mechanisms of chilling-induced oxidative stress injury and tolerance in devel-
oping maize seedlings: Changes in antioxidant system, oxidation of proteins and lipids,
and protease activities. Plant J 1996; 10:1017-1026.
27. Hugly S, McCourt P, Browse J et al. A chilling sensitive mutant of Arabidopsis with altered
steryl-ester metabolism. Plant Physiol 1990; 93:1053-1062.
28. Schneider JC, Hugly S, Somerville CR. Chilling sensitive mutants of Arabidopsis. Plant
Mol Biol Rep 1995; 13:11-17.
29. Schneider JC, Nielsen E, Somerville C. A chilling-sensitive mutant of Arabidopsis is deficient
in chloroplast protein accumulation at low temperature. Plant Cell Environ 1995; 18:23-31.
30. Patterson GW, Hugly S, Harrison D. Sterols and phytyl esters of Arabidopsis thaliana under
normal and chilling temperatures. Phytochemistry 1993; 33:1381-1383.
31. Browse J, Somerville C. Glycerolipids. In: Meyerowitz E, Somerville C, eds. Arabidopsis.
New York: Cold Spring Harbor Press, 1994:881-912.
32. Miquel M, James D, Dooner H et al. Arabidopsis requires polyunsaturated lipids for low
temperature survival. Proc Natl Acad Sci USA 1993; 90:6208-6212.
33. Hugly S, Somerville C. A role for membrane lipid polyunsaturation in chloroplast biogenesis
at low temperature. Plant Physiol 1992; 99:197-202.
34. Tokuhisa JG, Feldmann KA, LaBrie ST et al. Mutational analysis of chilling tolerance in
plants. Plant Cell and Environment 1997; 20:1391-1400.
35. Moir D, Botstein D. Determination of the order of gene function in the yeast nuclear
division pathway using cs and ts mutants. Genetics 1982; 100:565-577.
36. Botstein D, Maurer R. Genetic approaches to the analysis of microbial development. Ann
Rev Genet 1982; 16:61-83.
37. Feldmann KA. T-DNA insertion mutagenesis in Arabidopsis: Mutational spectrum. Plant J
1991; 1:71-82.
38. Tokuhisa JG, Vijayan P, Feldmann KA et al. A homolog of DIM1, a yeast gene encoding a
ribosomal RNA dimethylase is essential for chloroplast development at low temperatures.
Plant Cell 1998; in press.
39. Levitt J. Responses of Plants to Environmental Stress. Chilling, Freezing, and High Tem-
perature Stresses. 2nd ed. New York:Academic Press, 1980.
40. Brush AR, Griffith M, Mlynarz A. Characterization and quantification of intrinsic ice
nucleators in winter rye (Secale cereale) leaves. Plant Physiol 1994; 104:725-735.
41. Olien CR. Analysis of midwinter freezing stress. In: Olien CR, Smith MN eds. Analysis and
Improvement of Plant Cold Hardiness: Boca Raton:CRC Press Inc, 1981:61-87.
42. Steponkus PL. Role of the plasma membrane in freezing injury and cold acclimation. Annu
Rev Plant Physiol 1984; 35:543-584.
43. Guy CL. Cold acclimation and freezing stress tolerance: role of protein metabolism. Annu
Rev Plant Physiol Plant Mol Biol 1990; 41:187-223.
45. Steponkus PL, Uemura M, Webb MS. A contrast of the cryostability of the plasma
membrane of winter rye and spring oat. Two species that widely differ in their freezing
tolerance and plasma membrane lipid composition. In: Steponkus, PL ed. Advances in
Low-Temperature Biology. Volume 2. London:JAI Press Ltd, 1993:211-312.
46. Uemura M, Steponkus PL. Effect of cold acclimation on membrane lipid composition
and freezing-induced membrane destabilization. In: Li PH, Chen THH, eds. Plant Cold
Hardiness. New York:Plenum Press, 1997:171-179.
47. Strauss G and Hauser H. Stabilization of lipid bilayer vesicles by sucrose during freezing.
48. Anchordoguy TJ, Rudolph AS, Carpenter JF et al. Modes of interaction of cryoprotectants
with membrane phospholipids during freezing. Cryobiology 1987; 24:324-331.
49. Neven, LG, Haskell DW, Guy CL et al. Association of 70-kilodalton heat-shock cognate
proteins with acclimation to cold. Plant Physiol 1992; 99:1362-1369.
50. Anderson JV, Li Q-B, Haskell DW et al. Structural organization of the spinach endoplasmic
reticulum-luminal 70-kilodalton heat-shock cognate gene and expression of 70-kilodalton
heat-shock genes during cold acclimation. Plant Physiol 1994; 104:1359-1370.
51. Caban ÈM, Calver P, Vincens P et al. Characterization of chilling-acclimation-related
proteins in soybean and identification of one as a member of the heat shock protein
(HSP70) family. Planta 1993; 190:346-353.
52. Krishna P, Sacco M, Cherutti JF, Hill S. Cold-induced accumulation of hsp90 transcripts
in Brassica napus. Plant Physiol 1995; 107:915-923.
53. Olien CR, Clark JL. Freeze-induced changes in carbohydrates associated with hardiness of
barley and rye. Crop Sci 1995; 35:496-502.
54. Duman JG. Purification and characterization of a thermal hysteresis protein from a plant,
the bittersweet nightshade Solanum dulcamara. Biochim Biophys Acta 1994; 1206:129-135.
55. Griffith M, Antikainen M, Hon W-C et al. Antifreeze proteins in winter rye. Physiol
Plant 1997; 100:327-332.
56. Antikainen M, Griffith M. Antifreeze protein accumulation in freezing-tolerant cereals.
Physiol Plant 1997; 99:423-432.
57. Guy CL, Niemi KJ, Brambl R. Altered gene expression during cold acclimation of spinach.
58. Thomashow MF. Arabidopsis thaliana as a model for studying mechanisms of plant cold
tolerance. In: Meyerowitz E, Somerville C, eds Arabidopsis. New York:Cold Spring Harbor
Laboratory Press, 1994:807-834.
59. Hughes MA, Dunn MA. The molecular biology of plant acclimation to low temperature. J
Expt Bot 1996; 47:291-305.
60. Gibson S, Arondel V, Iba K et al. Cloning of a temperature-regulated gene encoding
a chloroplast omega-3 desaturase from Arabidopsis thaliana. Plant Physiol 1994;
106:1615-1621.
61. White AJ, Dunn MA, Brown K et al. Comparative analysis of genomic sequence and
expression of a lipid transfer protein gene family in winter barley. J Experimental Botany
1994; 45:1885-1892.
62. Jonak C, Kiegerl S, Ligterink W et al. Stress signaling in plants: A mitogen-activated
protein kinase pathway is activated by cold and drought. Proc Natl Acad Sci USA 1996;
93:11274-11279.
63. Mizoguchi T, Irie K, Hirayama T et al. A gene encoding a mitogen-activated protein kinase
is induced simultaneously with genes for a mitogen-activated protein kinase and an S6
ribosomal protein kinase by touch, cold, and water stress in Arabidopsis thaliana. Proc
Natl Acad Sci USA 1996; 93:765-769.
64. Monroy AF, Dhindsa RS. Low-temperature signal transduction: Induction of cold accli-
mation-specific genes of alfalfa by calcium at 25˚C. The Plant Cell 1995; 7:321-331.
65. Tahtiharju S, Sangwan V, Monroy AF et al. The induction of kin genes in cold-acclimating
Arabidopsis thaliana. Evidence of a role for calcium. Planta 1997; 203:442-447.
66. Jarillo JA, Capel J, Leyva A et al. Two related low-temperature-inducible genes of Arabidopsis
encode proteins showing high homology to 14-4-3 proteins, a family of putative kinase regu-
lators. Plant Mol Biol 1994; 25:693-704.
67. Thomashow MF. Role of cold-responsive genes in plant freezing tolerance. Plant Physiol
1998; in press.
68. Dure III, L. Structural motifs in Lea proteins. In: Close TL, Bray EA, eds. Plant Responses
to Cellular Dehydration During Environmental Stress. Rockville:American Society of Plant
Physiologists, 1993:91-103.
69. Close, TJ. Dehydrins: A commonality in the response of plants to dehydration and low
temperature. Physiologia Plantarum 1997; 100:291-296.
70. Ingram J, Bartels D. The molecular basis of dehydration tolerance in plants. Annu Rev
71. Artus NN, Uemura M, Steponkus PL et al. Constitutive expression of the cold-regulated
Arabidopsis thaliana COR15a gene affects both chloroplast and protoplast freezing tolerance.
72. Jaglo-Ottosen KR, Gilmour SJ, Zarka DG et al. Arabidopsis CBF1 overexpression induces
COR genes and enhances freezing tolerance. Science 1998; 280:104-106.
73. Lin C, Thomashow MF. DNA sequence analysis of a cDNA for cold-regulated Arabidopsis
gene cor15 and characterization of the COR15 polypeptide. Plant Physiol 1992; 99:519-525.
74. Gilmour and Thomashow, unpublished data.
75. Webb MS, Gilmour SJ, Thomashow MF et al. Effects of COR6.6 and COR15am polypep-
tides encoded by COR (cold-regulated) genes of Arabidopsis thaliana on dehydration-
induced phase transitions of phospholipid membranes. Plant Physiol 1996; 111:301-312.
76. Uemura M, Gilmour SJ, Thomashow MF et al. Effects of COR6.6 and COR15am polypep-
tides encoded by COR (cold-regulated) genes of Arabidopsis thaliana on the freeze-
induced fusion and leakage of liposomes. Plant Physiol 1996; 111:313-327.
77. Thomashow MF. Molecular genetics of cold acclimation in higher plants. Adv Genet 1990;
28:99-131.
78. Stockinger EJ, Gilmour SJ, Thomashow MF. Arabidopsis thaliana CBF1 encodes an AP2
domain-containing transcriptional activator that binds to the C-repeat/DRE, a cis-acting
DNA regulatory element that stimulates transcription in response to low temperature and
water deficit. Proc Natl Acad Sci USA 1997; 94:1035-1040.
79. Yamaguchi-Shinozaki K, Shinozaki K. A novel cis-acting element in an Arabidopsis gene is
involved in responsiveness to drought, low-temperature, or high-salt stress. Plant Cell 1994;
6:251-264.
80. Volger HG, Heber U. Cryoprotective leaf proteins. Biochim Biophys Acta 1975; 412:335-349.
81. Hincha DK, Heber U, Schmitt JM. Proteins from frost-hardy leaves protect thylakoids
against mechanical freeze-thaw damage in vitro. Planta 1990; 180:416-419.
82. Hincha DK, Schmitt JM. Freeze-thaw injury and cryoprotection of thylakoid membranes. In:
Somero G, Osmond B eds. Water and life: Comparative Analysis of Water Relationships at
the Organismic, Cellular and Molecular Level. Berlin:Springer Verlag, 1992:316-337.
83. Sieg F, Schroder W, Schmidt JM et al. Purification and characterization of a cryoprotective
protein (cryoprotectin) from the leaves of cold-acclimated cabbage. Plant Physiol 1996;
111:215-221.
84. Hong B, Uknes SJ, Ho T-HD. Cloning and characterization of a cDNA encoding a mRNA
rapidly induced by ABA in barley aleurone layers. Plant Mol Biol 1988; 11:495-506
85. Xu D, Duan X, Wang B et al. Expression of a late embryogenesis abundant protein gene,
HVA1, from barley confers tolerance to water deficit and salt stress in transgenic rice. Plant
Physiol 1996; 110:249-257.
86. Imai R, Chang L, Ohta A et al. A lea-class gene of tomato confers salt and freezing tolerance
when expressed in Saccharomyces cerevisiae. Gene 1996; 170:243-248.
87. Cohen A, Plant AL, Moses MS et al. Organ-specific and environmentally regulated
expression of two abscisic acid-induced genes of tomato. Plant Physiol 1991; 97:1367-1374.
88. Warren G, McKown R, Marin A et al. Isolation of mutations affecting the development of
freezing tolerance in Arabidopsis thaliana (L.) Heynh. Plant Physiol 1996; 111:1011-1019.
89. McKown R, Kuroki G, Warren G. Cold responses of Arabidopsis mutants impaired in freezing
tolerance. J Expt Bot 1996; 47:1919-1925.
90. Xin Z, Browse, J eskimo1 mutants of Arabidopsis are constitutively freezing-tolerant. Proc Natl
Acad Sci USA 1998; 95:7799-7804.
91. Rudolph AS, and Crowe JH. Membrane stabilization during freezing: The role of two natural
cryoprotectants, trehalose and proline. Cryobiology 1985; 22:367-377.
92. Metz AM, Timmer RT, Allen ML, Browning KS. Sequence of a cDNA encoding the
alpha-subunit of wheat translation elongation factor 1. Gene 1992; 120:315-316.
CHAPTER 5
Molecular Responses to Heat Stress

Fritz Schöffl, Ralf Prändl and Andreas Reindl
T he heat shock response, defined as a transient reprogramming of gene expression, is a

conserved reaction of cells and organisms to elevated temperatures (heat stress). The
features of this response are the induction of heat shock protein (HSP) synthesis and,
physiologically, the acquisition of a higher level of thermotolerance. In plant species
including soybean the synthesis of HSP is transient and accompanied by a cessation of normal
protein synthesis.1 During heat stress HSP seem to accumulate in a dosage-dependent manner,
sufficient to protect cells from severe damage and to allow resumption of normal cellular
and physiological activites. The transient synthesis of HSP suggests that the signal which is
triggering the reponse is either lost, inactivated or no longer recognized under conditions of
long term heat treatment.
The temperature for the induction of the heat shock response (hyperthermic treatment)
in plants is related to the optimum growth temperature of species and varieties and is usually
5˚-10˚C above normothermic conditions. Plants evolved in low temperature environments
respond to much lower heat shock temperatures than, for example, desert plants.2,3 The
differences in temperature set points for the heat shock response indicate that the cells are
not sensing the absolute temperature. Instead, an internal stress signal, generated at different
temperatures in different species and varieties seems to trigger the response. As exemplified
for the soybean, the heat shock response can be induced by different temperature regimes:2,4
1. Treatment at the maximum heat shock temperature (endured by the unconditioned
cell for several hours),
2. Short heat pulse (applied for several minutes only) at an otherwise lethal temperature, and
3. Also, by slowly, stepwise increasing temperatures.
Severe heat stress leads to cellular damage and cell death. The thermal death point or
killing temperature of plants is highly dependent on the time of exposure. At very high
temperatures, death occurs within minutes and can be attributed to a catastrophic collapse of
cellular organization. Attempts have been made to discriminate between rapid direct (protein
denaturation and aggregation, increased fluidity of lipids in membranes) and slower indirect
types of heat injuries (inactivation of enzymes in chloroplasts and mitochondria, block in
protein synthesis, degradation of proteins and membranes) that eventually lead to starvation,
inhibition of growth, generation of toxic compounds, Reactive oxygen species, ion efflux,
etc.5 However, the crucial point always remaining is the sensitivity of proteins and
enzymes to heat inactivation and denaturation. Hence, adaptive mechanisms that protect
cells from the proteotoxic effects of heat stress should be the key in acquisition of
thermotolerance.
The term tolerance is defined as the ability of cells and organisms to endure an
internal stress induced by an externally applied stressor.6 Thermotolerance refers to the
ability to cope with extreme temperatures; however, it is almost exclusively used for
tolerance to heat, not for cold stress tolerance. Basal tolerance refers to the ability to
survive a certain dose of heat stress without being conditioned by prior heat treatments.
Acquired thermotolerance refers to an enhanced level of tolerance (significantly higher
than basal tolerance) induced by the application of a relatively mild or gradually
increasing heat treatment. All plants tested to date are capable of acclimating or hardening
towards heat stress. It has been shown that heat hardening of plant cells is effective only
under conditions that also induce the synthesis of HSP.7 This acquisition of a higher
level of thermotolerance protects the cells and organisms from a subsequent, otherwise
lethal heat stress.8-10
The heat shock response is of great interest for studying the molecular mechanisms of
stress tolerance and regulation of gene expression. In this paper we summarize the work
on the characterization of the heat shock response with respect to the molecular function
of HSP in thermotolerance and general stress tolerance, the expression of heat shock genes,
and the regulation of the heat shock transcription factor HSF.
Heat Shock Proteins and Thermotolerance

Several classes of HSP have been defined in eukaryotes, including plants. Nomenclature
refers to the approximate molecular mass which is characteristic for each group (HSP100,
HSP90, HSP70, HSP60, HSP20). Synonymous designations are according to the designations
of homologues in prokaryotes (see Table 5.1).
HSP are functionally linked to the large and diverse group of molecular chaperones
which are defined by their capacity to recognize and to bind substrate proteins that are in
an unstable, inactive state. It becomes evident that probably all cellular proteins have to
interact with molecular chaperones at least once in their lifetime, either during synthesis,
subcellular targeting, or degradation. Owing to heat denaturation, the fraction of potential
targets for molecular chaperones seems to dramatically increase upon heat stress and
consequently the cellular chaperone pool has to be replenished. It is not surprising that
except for HSP20 and HSP100 each class of HSP is matched by one or several other
constitutive or cognate proteins (HSC) expressed at normal temperatures or respectively
non-stress conditions. Different HSP may have different functional properties, but common to
almost all of them is their capacity to interact with other proteins and to act as molecular
chaperones (for review see ref. 11). General and plant-specific characteristics of the different
classes of HSP are briefly described; a broader overview is given by Boston et al.12
HSP100
HSP100 belongs to the larger class of Clp proteins and is most similar to the ClpB
subgroup, a family of chaperones with diverged structures. The proteins of the HSP100/
Clp family are crucial for protein disassembly of aggregated and higher order protein
structures. The chaperone activity of certain members of this group assists in proteolytic
pathways, and the ATPase activity of HSP100 is essential for this function. Heat-inducible
plant HSP100 genes cloned from soybean and Arabidopsis were able to functionally
replace the homologous yeast HSP100 in the development of thermotolerance.13,14 In yeast,
HSP100 was shown to be important also for the restoration of heat-inactivated splicing
complexes.15 Of great interest and perhaps also important to plants is the implementation
of a regulatory role of HSP100 in controlling the extrachromosomal inheritance of the
prion-like protein PSI in yeast.15
Molecular Responses to Heat Stress 83
Table 5.1 General Functional Properties of HSP Families
HSP familya Properties
HSP100 Homologous to ClpB , ATP-dependent disassembly of higher

order protein structures, important for thermotolerance in yeast
HSP90 Chaperone with ATP binding site, association with proteins involved
in signal transduction
HSP70 Homologous to DnaK, ATP-dependent chaperone, negative regulator

of the heat shock response, isoforms involved in protein transport
HSP60 GroEL/GroES homologs, ATP-dependent chaperonin, not typically

hs-induced, isoforms in mitochondria and chloroplasts
(Cpn60/Cpn10), cytoplasmic TCP-1
HSP20b Prevalent in plants, aggregation into homooligomeric complexes in

the cytoplasm, formation of heat shock granules, ATP-independent
molecular chaperone in vitro
aclassified by molar mass in kDa, different sizes in many organisms, structural and functional
conservation within families
bHSP-families in plants, only one HSP in yeast and man, structural relationship with the lens
crystallines of the vertebrate eye
HSP90
The members of the HSP90 family, like HSP70 proteins, are highly conserved.
Constitutive and heat-regulated expression has been reported for HSP90 proteins. In yeast
at least one HSP90 is required for viability, and biochemical evidence supports a role of
HSP90 in preventing thermal denaturation and aggregation of protein substrates in vitro.16
Except for studies about the expression of Hsp90 genes, nothing is known about biochemical,
functional, or genetic properties of proteins of this group in plants. The assembly of
glucocorticoid receptor/HSP90 complexes in wheat germ extracts17 suggests that plant
HSP90 may be also involved in signal transduction as shown for animal and human cells,
by forming complexes with, for example, steroid receptors and kinases.18
HSP70
The HSP70 family is probably the most highly conserved protein/gene family in
nature, with about 50% identical amino acid residues between the E. coli DnaK and the
homologous HSP70 proteins in eukaryotes.19 There is genetic evidence for an essential
function of HSC70 in yeast. Members of the chaperone 70 family are involved in translocation
competence of precursor proteins, uncoating of coated vesicles, and association with
ribosomes in the cytosol; others are the driving force for protein uptake in mitochondria
and the ER (for review see Rassow et al 20) and in translocation of proteins into chloroplasts
(for review see Boston et al12). There is structural/biochemical evidence that plant HSP70/
HSC70 genes share the conserved domains for an N-terminally located ATPase activity
and a C-terminally located peptide-binding site. Current models propose that HSP70/
HSC70 bind to nascent or partially denatured polypeptides, thereby preventing improper
folding and keeping proteins in a translocation competent and/or refoldable conformation.
ATP binding and hydrolysis seem to modulate substrate binding. Cochaperones homologous
to DnaJ and GrpE of E. coli accelerate ATP-hydrolysis and stimulate ATP-exchange,
respectively. Numerous genes of the HSP70/HSC70 family have been identified in plants and
it is typical that single species encode multiple genes for cytosolic and organelle-targeted
HSP70-chaperones.12 The multiplicity of HSC70 genes/proteins is partially explained by the
multiple localizations of HSP70-like proteins in different compartments (cytoplasm, chloro-
plasts, mitochondria, ER). The genetic redundancy of cytosolic variants of HSC70 is not
yet understood; multiple forms of this conserved chaperone may reflect the requirement
for differences in biochemical function of these proteins. Identifying the in vivo targets of
HSP70/HSC70 chaperones and a mutant analysis will be a prerequisite to the elucidation
of functional properties.
HSP60
Chaperonins which are homologous to the bacterial GroEL/GroES proteins are not
typically heat-induced stress proteins. The chloroplast-localized Cpn60/Cpn10 are nuclear
encoded proteins and are imported into chloroplasts (for review see Boston et al12). Both
subunit proteins are expressed in the absence of stress and moderately increased upon
heat shock. Ch-Cpn60 forms a cylindrically shaped double-stacked ring of two heptamers
with intrinsic ATPase activity. In contrast to the bacterial GroEL complex, ch-Cpn60
appears to be a hetero-oligomer consisting of stoichiometric amounts of α- and β-subunits.
The chloroplast co-chaperonin Cpn10, compared to its bacterial counterpart GroES, is an
obviously head-to-tail duplicate form of approximately 21 kDa. GroEL/GroES complexes
facilitate folding, assembly, and translocation of numerous other proteins, as exemplified
by the Rubisco complex in chloroplasts. However, in plants the exact mechanism is still
obscure. It is known that GroEL oscillates between states of high and low affinity for
non-native proteins, release requires binding of adenine nucleotides, and that the
GroEL-assisted folding reaction is strictly dependent on ATP-hydrolysis and participation
of co-chaperone GroES. GroEL and GroES-like proteins are also found in plant mitochondria.
The TCP-1 protein of the cytosol is only distantly related to GroEL but probably performs
analogous functions.
HSP20
HSP20 proteins, also termed small (sHSP) or low molecular weight (lmw) HSP, are a
complex group of prevalent HSP in plants. Characteristic to plants is also the occurrence
of members of the HSP20 group in chloroplasts, mitochondria, and in the ER (for review
see Boston et al12). Recently a tomato chloroplast sHSP was found to be imported into
chromoplasts, indicating a function for this HSP in fruit ripening.21 Meanwhile, sHSP
comprise 6 gene families for strictly heat-inducible HSP ranging between 17 and 30 kDa. HSP
of two families (class I and II) are targeted to the cytoplasm. It should be noted that to date
only in higher plants were sHSP also found in the endomembrane system. The conservation of
amino acid residues within classes is 80-90%, but only 30% between different classes.
However, all sHSP share a common structural domain, a feature for all sHSP and the
α-crystalline, in the C-terminal part of the molecule. Small HSP tend to form homooligomeric
complexes of 200 to 800 kDa in vitro.22,23 Studies using recombinant proteins in vitro
suggest that sHSP act as chaperones preventing heat-induced aggregation and promote
renaturation of model substrates. 24 Interestingly, this chaperone activity acts in
substoichiometric amounts and, in contrast to HSP70 and HSP60 chaperones, does not
require stimulation by nucleotides. Although the in vivo targets of sHSP chaperone

complexes are not known, a model suggests that sHSP capture unfolded polypeptides by
hydrophobic interactions and maintain them in a state competent for refolding the native
state by other chaperones, or likewise for degradation by the proteolytic systems.
At the cellular level, one consequence of the heat shock response is the appearance of
granular structures containing class I-sHSP located throughout the cytoplasm and the
nucleus in soybean.25 Heat shock granules may be formed from different HSP; their exact
composition in vivo is unknown. Dependent on the heat stress regimes applied, granules
may form higher order structures of aggregation.26
Mutants Affecting Expression of HSP and Thermotolerance

There is a striking correlation between the occurrence of HSP and acquisition of
thermotolerance, but there is only little direct evidence proving a causal relationship.
Mutagenesis and genetic engineering strategies have been proposed for the analysis of
regulation and for the manipulation of the heat shock response.27,28 Mutations in regulatory
genes resulting in a coordinate change in expression of HSP would be required for studying:
1. the signal pathway from stress to gene,
2. the mechanism of transcriptional regulation and
3. the role of HSP in thermotolerance.
Using genetic engineering of Arabidopsis as a model for higher plants, dominant
mutations were generated showing a deregulated synthesis of HSP at normal temperature.41,42
Such transgenic plants exhibit a significantly higher level of basic thermotolerance. These
data demonstrate the fundamental role of HSP in thermotolerance, however, it is not yet
clear whether, apart from HSP, other genes are also affected and involved in thermotolerance
in transgenic lines expressing a derepressed HSF.
The effects of mutations in individual heat shock genes have been investigated in
different organisms. Analyses in yeast provided evidence for an important role of
HSP10429 and a minor, accessory role of HSP7030 concerning thermotolerance. Mutations
in Hsp26, the sole gene for sHSP in yeast,31 or overexpression of sHSP and antisense
approaches in transgenic plants32,33 had no obvious effect on the phenotype. It is possible
that the protective effect of HSP is sometimes dependent on the physiological conditions
of the cell, as shown by the disruption of a mitochondrial HSP30 gene in Neurospora
resulting in strains that were less thermotolerant under certain carbohydrate limitations. 34
In other eukaryotes, other groups of HSP seem to play important roles in thermotolerance,
as for example shown by HSP70 overexpression in mammalian cells and in Drosophila.35-40
Other Heat-Induced Proteins

Besides the described groups of major HSP chaperones there is a number of plant
proteins/genes including ubiquitin, 43,44 cytosolic Cu/Zn-superoxide dismutase 45 and
manganese peroxidase,46 whose expression is also stimulated upon heat stress. The function of
these proteins is related to protein degradation pathway and oxidative stress response, two
obviously very important processes to be maintained during heat stress or restored during
recovery.
Links to Other Abiotic Stresses

Expression of HSP is also linked to several other environmental stresses and an
increasing number of studies show cross protection in plants. The results imply that HSP
are important constituents of the molecular mechanism of common stress tolerance. The
most obvious links are between HSP and dehydration/drought, cold/chilling/freezing, heavy
metal, and oxidative stress.
Dehydration/Drought
In sunflower47 and the resurrection plant Craterostigma plantagineum48 certain sHSP
are expressed upon water stress/dehydration in vegetative tissue. Expression of sHSP was
identified in a screen for early dehydration responsive genes in Arabidopis thaliana.49 HSP
are also expressed in the absence of heat stress during late seed maturation/desiccation
period in the Arabidopsis embryo.50
Cold/Chilling
Chilling injury is a physiological disorder that develops in some plants that are indigenous
to tropic and subtropic regions. As shown for tomato fruit, heat stress can protect against
chilling injury and the accumulation of HSP correlated with the persistence of chilling
tolerance.51 A similar correlation was found also for chilling tolerance induced by heat
stress in the mung bean hypocotyl52 and in germinating Cucumis sativus seeds.53 Cold-induced
transcripts of HSP90 in Brassica napus, HSC70 in spinach and soybean, and sHSP in
potato, suggest that HSP seem to play a role in plant responses and acclimation to low
temperatures.54-57
Heavy Metal
In soybean hypocotyl the most efficient non-heat shock inducers of heat shock gene
expression are arsenite and cadmium.58-60 Arsenite-treated cells synthesize HSP and
acquire a certain level of thermotolerance.8 Heat shock protects against metal toxicity in
wheat leaves, and cultured cells of Lycopsersicon peruvianum can be protected by heat stress
from injuries inflicted by cadmium treatment.61,62 On the other hand, expression of HSP
is also induced by heavy metal ions, which is in accordance with the concept of cross protection
via HSP.63
Oxidative Stress
Cell death in plants caused by abiotic stresses is frequently associated with symptoms
of oxidative stress. Mutant analyses of SOD, catalase, and cytochrome C oxidase genes in
yeast provide evidence for an involvement of oxidative stress in heat-induced cell death;
overexpression of these enzymes improved thermotolerance under anaerobic conditions.64
In human cells mitochondria were found to be selective targets for the effect of heat shock
against oxidative stress, probably mediated by HSP70.65 In plants, heat shock protects PSII
from photoinhibition and it is speculated that a chloroplast-targeted sHSP is involved in
the protection of the D1 protein.63
On the Molecular Mechanism of Cross Protection

Heat shock-induced cross tolerance and the occurrence of HSP induced by other abiotic
stresses suggest that HSP are important determinants of a common stress response in plants.
If HSP have a generally protective function in abiotic stress response, what is the common
theme in the cellular damage by abiotic stresses? This is most likely the denaturation of
cellular proteins and/or the generation of Reactive oxygen species. Higher levels of
enzymes and enzymatic activities of the antioxidant pathway including superoxide
dismutases (SOD), catalyses, ascobate peroxidase, glutathione reductase and others were
found in response to a number of environmental stresses (for overview see Foyer et al 67).
Overexpression of SOD in tobacco,68,69 alfalfa,70 potato,71 and cotton72 has been found to
enhance tolerance to oxidative stress and to some extent also to freezing, chilling injury
and water deficit.67,73
With the availability of mutants of the heat shock response, showing a synthesis of
HSP at lower non-heat shock temperatures,41,42 it will be possible to test whether under
environmental stress the chaperone function of HSP is involved in stabilization of enzymes

(or enzymatic activities) of the antioxidant pathway.
Transcriptional Regulation
The expression of heat shock genes in plants is, similarly to the situation in other
eukaryotes, primarily regulated at the transcriptional level.74 The thermoinducibility is
attributed to conserved cis-regulatory promoter elements (HSE) which are located in the
TATA box proximal 5'-flanking regions of heat shock genes. The occurrence of multiple
HSEs within a few hundred base pairs is a signature of most eukaryotic heat shock genes.
The importance of HSE for heat-dependent transcriptional regulation in plants has been
verified by promoter deletions of a soybean heat shock gene75,76 and by the capacity of
synthetic HSE sequences, integrated in a truncated CaMV-35S promoter, to stimulate
heat-inducible gene expression in transgenic tobacco.76 The requirement of the TATA box
was demonstrated by deletion analysis of a soybean heat shock gene in sunflower.77 These
data are in accordance with the protection pattern of HSE and TATA box-containing
regions in footprinting experiments using a Drosophila HSP70 promoter.78
HSE-HSF Interaction
The eukaryotic HSE consensus sequence has been ultimately defined by Amin et al79
and Xiao and Lis80 as alternating units of 5'-nGAAn-3'. HSE are the binding sites for the
transactive heat shock transcription factor HSF; efficient binding requires at least three
units resulting in 5'-nGAAnnTTCnnGAAn-3'.81,82 Chemical footprinting and interference
experiments show that the binding of HSF to HSE occurs in the major groove. The nGAAn
box is considered as the fundamental unit of HSF-binding, with each subunit of a HSF
trimer interacting with one nGAAn-unit.81 The stoichiometry for binding was directly
demonstrated by analytical ultracentrifugation.83 There is evidence for trimer formation
of recombinant plant HSF following expression in E. coli and, for binding to both consensus
HSE sequences84,85 and to the HSE-containing regions of an authentic Drosophila HSP70
promoter.85 Transgenic expression provides evidence that HSF-HSE interaction and
transcriptional activation is highly conserved in nature, as demonstrated by the recognition
and proper regulation of a Drosophila HSP70 promoter in plants86 and by the activation of
the same promoter in Drosophila cells via the transiently expressed HSF of Arabidopsis,
AtHSF1.85
The promoter strength seems to depend on several different criteria, including the
degree of conservation and spacing of HSE sequences.87 Mismatches in a naturally occurring
HSE2 consensus sequence of a Drosophila HSP70 gene result in only weak binding of HSF;
however, in the presence of an adjacent full matching HSE1 the binding affinity to HSE2 is
greatly increased, indicating cooperative binding of trimeric HSF.88 The cooperativity in
HSF binding may act as a supplementary mechanism for increasing the range of heat-
inducible binding to DNA.89
Other Sequences Affecting Heat Shock Gene Expression

A number of additional sequence motifs have been identified that have quantitative
effects on expression of certain heat shock genes. In plants, there is evidence for an
involvement of CCAAT box90 and AT-rich sequences.75,91 One AT-rich sequence located in
the flanking region of a soybean heat shock gene exerted an enhancing effect on the
expression of heat shock promoter-reporter gene constructs in tobacco and was able to
bind to the nuclear scaffolds.92 These data suggest that sequences affecting the chromatin
structure may be important for efficient access of transcription factors (e.g., TATA-box
binding protein TBP) and/or the transcriptional activator proteins (e.g., HSF). The following
model integrates the current knowledge about the activation of heat shock gene
expression: The binding of a GAGA sequence binding factor93,94 or, likewise, scaffold
attachment affects chromatin structure in a way that provides TBP access to the TATA box,
which is prerequisite for subsequent assembly of the basal transcription complex. In this
“standby mode” heat shock genes are primed for transcriptional activation upon heat stress,
mediated by the trimerization and binding of HSF to the HSE sequences.
Developmental Expression
In many organisms including plants, the expression of heat shock genes is not only
triggered by a number of environmental stresses but also by developmental cues. In plants
the regulation of developmental expression of HSP, indicated by the occurrence of mRNAs
and HSP in dry seeds, has not yet been investigated in greater detail. The analysis of a
developmentally regulated soybean heat shock promoter in transgenic tobacco provides
evidence for participation of HSE sequences and consequently suggests binding and
involvement of HSF.95 It cannot be excluded that other sequences and transactive factors
are involved in seed specific expression of HSP. The control of this expression by the
developmental program rather than by dehydration is indicated by the negative effect of
the abi3 mutation in Arabidopsis on seed-specific expression of sHSP.96 ABI3, originally
identified as an abscisic acid-insensitive mutant allele in Arabidopsis, appears to have a
dominant regulatory effect on the developmental expression of heat shock genes in the
embryo. Recent models for the action of Vp1,97 the structural/functional homologue of
ABI3 in maize,98 suggest that Vp1 acts in stabilization/activation of regulatory complexes
involved in the transcription of target genes. Further investigation of the activation of heat
shock promoters during seed maturation will be required to test the hypothesis that ABI3,
directly or via the action of secondary factors, is a regulator of HSF activity. It should be
noted that in Drosophila, developmental regulation of certain heat shock genes, such as the
expression of HSP82 and HSP26 in oocytes and early larval stages, seems to be regulated
by steroid hormones and does not involve HSE-HSF interaction. However, the sole HSF of
Drosophila plays an essential role at this stage of development and this function appears to
be not directly related to the expression of HSP.99
The Regulation of HSF

Activation of HSF of higher eukaryotes is a multi-step process. In response to heat
stress, HSF is converted from a monomeric to a trimeric form. Trimeric HSF is localized
predominantly in the nucleus, binding to HSE with high affinity. The acquisition of
transcriptional competence is separately regulated; it can be uncoupled from HSE binding, as
shown for human HSF1 and Drosophila HSF. HSF of Saccharomyces cerevisiae and
Kluyveromyces lactis is bound to heat shock promoters in the absence of stress and is therefore
believed to be regulated primarily at the level of transactivating ability.89,100
HSF Domains
The domains for DNA binding (DBD) and oligomerization (HR-A/B) are located in
the N-terminal region of HSF (Fig. 5.1). Both domains are conserved in primary structure
throughout the HSF protein family. Other regions show significant homology only
between closely related HSF. Nuclear localization signals (NLS), hydrophobic heptad
repeats localized in the C-terminal region (HR-C), and activation domains (AD) have
been identified by functional studies in several HSF.89,100
Similar to vertebrates, all plant species investigated so far contain multiple HSF in
contrast to single HSF genes reported for yeast and Drosophila. To date, three HSF have
been described from Arabidopsis thaliana, six from soybean (Glycine max.), three from
DBD HR-AB NLS HR-C AD
Fig. 5.1. Schematic domain structure of a prototypic HSF. The DNA-binding domain (DBD), the
oligomerization domain (HR-A/B), a nuclear localization signal (NLS), the C-terminal hydro-
phobic heptad repeat (HR-C), and an activation domain (AD) are marked. The sizes of the
individual domains as well as the lengths of the linker sequences are not drawn to scale.
tomato (Lycopersicon peruvianum), and three from maize (Zea maize).42,101-105 Molecular
masses of plant HSF are in the range from 31.2 to 57.5 kD. Based on sequence homology
and domain structure, plant HSF can be subdivided in the two classes A and B.101
DNA-Binding Domain
HSF carry a conserved DBD consisting of an antiparallel four stranded β-sheet packed
against a bundle of three α-helices, as determined for HSF from Kluyveromyces lactis,
Drosophila, and tomato.106-108 The second and the third helix form a typical helix-turn-helix
motif. The third helix is responsible for establishing the specific nucleic acid contacts with
the HSEs. A distinguishing feature between non-plant and plant HSF is an 11 amino acid
deletion between two β-sheets in plant HSF. The significance of the lack of this probably
solvent-exposed loop is unknown.101
Oligomerization Domain
The HR-A/B is separated from the DBD by a linker of variable length and sequence
and comprises the two regions A and B. Region A is based on heptad repeats of hydrophobic
amino acids, whereas region B is composed of two overlapping heptad repeats. In class A
plant HSF, these arrays are separated by additional heptad repeats brought about by the
insertion of 21 amino acids. Plant HSF of class B lack this insertion. It is assumed that the
function of the HR-A/B region is to allow homotrimer formation through a triple-stranded,
α-helical coiled-coil structure.
Obviously, the formation of trimeric HSF in higher eukaryotes requires heat stress,
but how is the suppression of HSF trimerization achieved under non-stress conditions?
The C-terminal heptad repeats of hydrophobic residues (HR-C), which are well conserved
in animal HSF but only poorly in plant and yeast HSF, is involved in the regulation of
trimerization. Mutations in the HR-C region lead to HSF with constitutive trimerization
and DNA-binding competence, as shown for Drosophila HSF, chicken HSF1 and HSF3,
and human HSF1.110-112 A model proposes that intramolecular coiled-coil interactions
between the HR-A/B and HR-C hydrophobic heptad repeats suppress trimer formation
under normal growth conditions. However, deletion mapping of Drosophila HSF revealed
larger portions of HSF involved in the negative control of trimer formation.113
Nuclear Localization Signal

HSF carry two clusters of basic amino acids that have been proposed to function as
NLS. A highly conserved cluster of basic amino acids is located at the C-terminus of the
DBD, and a second cluster resides C-terminal of the HR-A/B.89,100,101 In functional studies
with two class A tomato HSF, the more C-terminal NLS has been found to be exclusively
required for nuclear import.104 Again, this is a discrepancy with vertebrate HSF, which
require either both or solely the more N-terminal NLS for translocation.115,116 By using
fusion proteins between the NLS of the Drosophila HSF and a β-galactosidase reporter, it
has been shown recently that the NLS is sufficient for stress-induced nuclear entry, supporting
the view that nuclear import is one layer of HSF regulation by stress.117
Activation Domain
The activation domains (AD) of HSF of higher eukaryotes are localized C-terminally,
whereas the HSF of Saccharomyces cerevisieae and Kluyveromyces lactis carry AD at the C- and
the N-terminus of the protein.89,100 The AD of human HSF1 and Drosophila HSF show
limited sequence identity and are rich in hydrophobic and acidic amino acids.118,119 Studies
carried out to characterize the AD of tomato HSF indicate the involvement of motifs which
consist of aromatic, hydrophobic, and acidic amino acid residues.120,121
Animal HSF acquire transactivating competence as a final step during the activation
process. Regulatory sequences repress the AD under non-stress conditions and render it
heat-inducible. A central regulatory domain has been mapped between the HR-A/B and
the AD of human HSF1. The regulatory domain confers heat responsiveness even to the
heterologous AD of VP16.118 In contrast to higher eukaryotes, the Saccharomyces cerevisiae
HSF is associated with the high affinity binding sites of heat shock promoters in the
absence of stress.122 HSF in yeast is assumed to be regulated primarily at the level of
transactivating competence. An amino acid in the DBD, the HR-B, and a yeast-specific
control element (CE2) have been shown to be involved in the repression of the AD under
non-stress conditions.123,124
Regulators of HSF Activity
Negative Regulation of HSF by HSP70

There is genetic evidence for an autoregulation of the heat shock response in Escherichia
coli, yeast, and higher eukaryotes.89,100 In Saccharomyces cerevisae mutations in two
constitutively expressed HSC70/HSP70 genes activate a β-galactosidase reporter gene in a
HSE-dependent manner in the absence of heat stress.125 The derived model of HSF regulation
by chaperone titration proposes that the pool of free HSC70/HSP70 is deplenished during
heat shock due to binding of HSC70/HSP70 to unfolded proteins, thereby relieving the
repression of HSC70/HSP70 on HSF. Derepression of the AD domain and increased
trimer formation in higher eukaryotes activate HSF and consequently heat shock gene tran-
scription. In a negative feedback loop, the synthesis of appropriate levels of HSP70 shuts
off HSF. With respect to trimer formation, HSC70/HSP70 may maintain HSF in monomeric
state or may participate in the disassembly of trimeric HSF. Stoichiometric complexes
between non-activated HSF1 and HSP70 have been described, as well as an inhibition of heat
activation of HSF1 in mammalian cells which transiently overexpress HSP70.126
Recently, genetic evidence for a negative regulation of Arabidopsis HSF1 under non-stress
conditions has been obtained. Arabidopsis HSF1 is repressed under non-stress conditions
and trimerizes upon heat shock. A heat stress-independent derepression of Arabidopsis
HSF1 was obtained by constitutive overexpression of HSF1 fusion proteins with a
β-glucuronidase reporter.41 The molecular mechanism of derepression is still unknown
but seems not only restricted to glucuronidase fusions of HSF. The conformation of the
fusion protein may be either inaccessible to a negative regulatory molecule or overexpression
of this protein titrates a transacting negative regulator. Interestingly, overexpression of
another Arabidopsis HSF, HSF3, appears to be sufficient for derepression of the heat shock
response in transgenic Arabidopsis.
Arabidopsis HSF1 shows also a constitutive DNA binding upon heterologous expression
in Drosophila and human cells. Furthermore, a derepressed transactivating competence
has been demonstrated for Arabidopsis HSF1 in Drosophila cells.102 Thus, the negative
control seems to depend on a factor which is absent in the cultured animal cells. Direct
evidence for HSP70 as a negative regulator of HSF in Arabidopsis comes from the analysis
of transgenic Arabidopsis plants carrying an HSP70 antisense gene. As a result of expression of
the antisense RNA, endogenous HSC70/HSP70 levels are reduced, and during the
recovery from heat shock HSF1 trimers are significantly longer present than in control
plants.128
Negative Regulation of HSF by Phosphorylation

Phosphorylation has been proposed to play a role in activation and inactivation of
HSF.89,100 However, recent functional studies suggest that phosphorylation is primarily
involved in repression of HSF. In yeast, phosphorylation of CE2-adjacent serine residues
has been shown to enhance deactivation of HSF after heat shock.129 The major phos-
phorylation sites of human HSF1 in cell culture at control temperature have been localized to
two serines in the regulatory domain, which modulates the AD. Mutational conversion of
these serines to alanine residues leads to constitutively active HSF, whereas conversion to
glutamic acid, mimicking a phosphorylated serine, represses HSF in the absence of stress.
Phosphorylation of the serine residues is increased upon stimulation of the Raf/ERK pathway,
which is a mitogen-activated protein kinase pathway responsive to growth factors.130,131
Recently, HSF phosphorylation has been reported for plants. A kinase activity has been
detected in extracts of Arabidopsis suspension culture cells which phosphorylates HSF1 at
serine residues and consequently decreases the HSE-binding of HSF1. Immunological
characterization has identified the kinase as CDC2a, a cyclin-dependent kinase regulating
the cell cycle.132
Thus, in human cells as well as in Arabidopsis, phosphorylation of HSF through various
kinases may integrate growth signals. As yet it is unknown whether cyclin-dependent
kinases are involved in HSF phosphorylation in animals or whether mitogen-activated
protein kinases play a role in HSF regulation in plants. It is conceivable that in growing
cells, subjected to moderate stress conditions, phosphorylation of HSF may be required
for a repression of the heat shock response, which might otherwise interfere with proliferation.
Such a phosphorylation-dependent repression can be overridden under conditions of heat
stress.129 On the other hand, HSF may have also essential functions during development.
Drosophila HSF is indispensable for oogenesis and early larval development.99 Thus it will
be interesting to investigate whether kinases modulate the developmental activities of HSF
and which pathway is involved in signaling that results in developmental control of HSP
expression in plants.
Conclusions and Perspectives

Some of the plant responses to heat stress show certain characteristics that are unique
to plants, originally discovered in plants or more important to plants than to other organisms.
One such area is the biological role and molecular mechanism of HSP in thermotolerance
and common stress tolerance. Based on the results of genetic engineering of the entire
heat shock response and thermotolerance via HSF,41,42 future research will focus on the
roles of HSP100, HSP90, HSP70 and sHSP as the primary targets for identifying specific
determinants involved in protection from deleterious effects of heat, cold, heavy metal,
desiccation, reactive oxygen, and other stresses. In order to understand the underlying
molecular mechanisms, it will be necessary to identify the cellular targets, enzymes, or
structural proteins to become associated with these HSP in their native form in protein
complexes. The problem of functional specificity of molecular chaperones is exemplified

by the proteins of the chaperone 70 family, comprising the heat-induced cytosolic HSP70
and several constitutive HSC70 proteins localized in different subcellular compartments.
The important questions are: What is the specific target and functional role of HSP70? Is
HSP70 and/or another protein of the 70 kDa stress protein family involved in the negative
regulation of HSF and consequently of the heat shock response?
The regulation of HSF activity and the multiplicity of HSF in plants are other problems of
great scientific interest. The mechanism of repression of HSF activity is still not understood.
HSF1 protein fusions41 and HSF342 of Arabidopsis are active upon transgenic overexpression,
suggesting that negative regulation and/or conformational changes are involved in the
mechanism of activation. Three HSF-like genes were identified in tomato,104 Arabidopsis,42,84
and maize105 and six in soybean.103 The question about the biological role of the genetic
redundancy of HSF has to be addressed and answered by future research. It seems possible
that some putative HSF, classified by the criterion of structural features in the DNA
binding and multimerization domains, may in fact act as DNA-binding-proteins, lacking,
however, the capacity of transcriptional activation. Such proteins could act through DNA
binding, either as repressors or through protein-protein interaction as modulators of HSF
activity.
Is there a signal pathway that senses stress from external sources and triggers the heat
shock response via HSF? Components in the pathway upstream from HSF are not yet
known. It is conceivable that HSF itself or its interaction with HSC70 and other proteins is
the sensor of heat stress and results in an activation of HSF via conformational changes
involving monomer to trimer transition and nuclear targeting. An alternative model for
temperature sensing and regulation of the heat shock response integrates observed membrane
alterations. The lipid perturbation model is based on the effects resulting from changes in
the ratio of saturated: unsaturated fatty acids on the set point of temperature for the heat
shock response in yeast and proposes that these alter HSF activity.133 Developmental signaling
seems to be responsible for the expression of HSP during seed maturation. The involvement of
HSF is indicated by the dependence on HSE promoter sequences; signaling through ABA
pathways is suggested by the negative effect of an abi3 mutation in Arabidopsis.96 However,
neither the responsible HSF nor the level of control by ABI3 have been identified.
Other areas of important research in the field of heat stress responses, not covered in
this review, concern changes in the translation machinery that result in a shut down of
normal protein synthesis and preferential synthesis of HSP during heat stress. In a number
of new approaches, interesting results were gained about translational control for the first
time in this field in plants.134-136
References
1. Key JL, Lin CY, Chen YM. Heat shock proteins of higher plants. Proc Natl Acad Sci USA
1981; 78:3526-3530.
2. Key JL, Kimpel J, Vierling E et al. Physiological and molecular analyses of heat shock
response in plants. In: Atkinson BG, Walden DB, eds. Changes in eukaryotic gene
expression in response to environmental stress. Orlando: Academic Press, 1985:327-348.
3. Kee SC, Nobel PS. Concomitant changes in high temperature tolerance and heat shock
proteins in desert succulents. Plant Physiol 1986; 80:596-598.
4. Key JL Nagao RT, Czarnecka E et al. Heat stress: Expression and structure of heat shock
proteins. In: Von Wettstein D, Chua NH ed. Plant Molecular Biology. New York:Plenum
Publ Corp, 1987:385-397.
5. Levitt J. Responses of plants to envronmental stresses. Vol I: Chilling freezing and high
temperature stresses. 2nd ed. New York:Academic Press, 1980:27-64.
6. Street HE, Oepik H. Physiology of flowering plants: Their growth and development. 3rd
ed., London: Edward Arnold, 1984:110-134
7. Wu MT, Wallner SJ. Heat stress responses in cultured plant cells. Plant Physiol 1984;
75:778-780.
8. Lin CY, Roberts JR, Key JL. Acquisition of thermotolerance in soybean seedlings. Plant
Physiol 1984; 74:152-160.
9. Yarwood CE. Acquired tolerance in leaves to heat. Science 1961; 134:941-942.
10. Yarwood CE. Adaptions of plants and plant pathogens to heat. In: Posser CL, ed.
Molecular mechanisms of temperature adaption, Publ. No. 84 Washington DC: Amer
Assoc Advance Sci 1967:75-89.
11. Vierling E. The roles of heat shock proteins in plants. Ann Rev Plant Physiol Plant Mol
Biol 1991; 42:579-620.
12. Boston RS, Viitanen PV, Vierling E. Molecular chaperones and protein folding in plants.
Plant Mol Biol 1996; 32:191-222.
13. Lee YR, Nagao RT, Key JL. A soybean 101-kD heat shock protein complements a yeast
HSP104 deletion mutant in acquiring thermotolerance. Plant Cell 1994; 6:1889-1897.
14. Schirmer EC, Lindquist S, Vierling E. An Arabidopsis heat shock protein complements a
thermotolerance defect in yeast. Plant Cell 1994; 6:1899-1909.
15. Vogel JL, Parsell DA, Lindquist S. Heat shock proteins HSP104 and HSP70 reactivate
mRNA splicing after heat inactivation. Curr Biol 1995; 306-317.
16. Jakob U, Buchner J. Assisting spontaniety: The role of HSP90 and small HSPS as molecular
chaperones. Trends Biochem Sci 1994; 19:205-211.
17. Stancato LF, Hutchinson KA, Krishna P, Pratt WB. Animal and plant cell lysates share a
conserved chaperone system that assembles the glucocorticoid receptor into a functional
heterocomplex with HSP90. Biochemistry 1996; 35:554-561.
18. Rutherford SL, Zucker CS. Protein folding and the regulation of signaling pathway. Cell
1994; 79:1129-1132.
19. Boorstein WR, Ziegelfoffer T, Craig EA. Molecular evolution of the HSP70 gene family.
J Mol Evol 1994; 38:1-17.
20. Rassow J, von Ahsen O, Böhmer U, Pfanner N. Molecular chaperones: Toward a character-
ization of the heat shock protein 70 family. Trends Cell Biol 1997; 7:129-133.
21. Lawrence SD, Cline K, Moore GA. Chloroplast development on ripening tomato fruit:
Identification of cDNAs for chromoplast targeted proteins and characterization of a cDNA
encoding a plastid localized low molecular weight heat shock protein. Plant Mol Biol
1997; 33:483-492.
22. Waters ER, Lee GJ, Vierling E. Evolution, structure and function of small heat shock
proteins in plants. J Exptl Bot 1995; 47:325-338.
23. Lee GJ, Pokala N, Vierling E. Structure and in vitro molecular chaperone activity of cytosolic
small heat shock proteins from pea. J Biol Chem 1995; 270:19432-10438.
24. Lee GJ, Roseman AM, Saibil HR et al. A small heat shock protein stably binds heat-
denatured model substrates and can maintain a substrate in a folding-competent state.
EMBO J 1997; 16:659-671.
25. Jinn T-L, Chang PFL, Chen YM et al. Tissue-type-specific heat-shock response and
immunolocalization of class I low molecular weight heat-shock proteins in soybean. Plant
Physiol 1997; 114:429-438.
26. Nover L, Scharf KD, Neumann D. Cytoplasmic heat shock granules are formed from
precursor particles and are associated with a specific set of mRNAs. Molec Cell Biol 1989;
9:1298-1308.
27. Schöffl F, Baumann G, Raschke E. The expression of heat shock genes—A model for
environmental stress response. In: Verma DP, Goldberg RB ed. Temporal and spatial
regulation of plant genes, Plant gene research V, New York: Springer Verlag, 1988:253-273.
28. Schöffl F. Molecular genetics of the heat shock response. In: Maybry HT, Nguyen HT,
Dixon RA, Bonness MS ed. Biotechnology for arid plants. Austin: IC2 Institute. University of
Texas at Austin, 1994:83-92.
29. Sanchez Y, Lindquist S. HSP104 is required for induced thermotolerance. Science 1990;
248:1112-1115.
30. Sanchez Y, Taulien J, Borkowich KA et al. HSP104 is required for tolerance to many
forms of stress. EMBO J 1992; 11:2357-2364.
31. Petko L, Lindquist S. HSP26 is not required for growth at high temperatures, nor for
thermotolerance, spore development, or germination. Cell 1986; 45:885-894.
32. Schöffl F, Rieping M, Baumann G. Constitutive transcription of a soybean heat shock
gene by a cauliflower mosaic virus promoter in transgenic tobacco plants. Dev Genetics
1987; 8:365-374.
33. Schöffl F, Diedring V, Kliem M et al. The heat shock response in transgenic plants: The
use of chimeric heat shock genes. In: Wray JL, ed. Inducible plant proteins: Their
biochemistry and molecular biology. Cambridge: Cambridge University Press, 1992:247-265.
34. Plesovsky-Vig N, Brambl R. Disruption of the gene for HSP30, an α-cristallin-related heat
shock protein of Neurosposra crassa, causes defects in thermotolerance. Proc Natl Acad Sci
USA 1995; 92:5032-5036.
35. Morimoto RI, Tissieres A, Georgopoulos C. Stress proteins in biology and medicine. New
York: Cold Spring Harbor Laboratory, 1990.
36. Li GC, Li L, Liu Y-K et al. Thermal response of rat fibroblast stably transfected with the
human 70-Kda heat shock protein-encoding gene. Proc Natl Acad Sci USA 1991;
88:1681-1685.
37. Li GC, Li L, Liu RY et al. Heat shock protein HSP70 protects cells from thermal stress
even after deletion of its ATP-binding domain. Proc Natl Acad Sci USA 1992; 89:2036-2040.
38. Suzuki K, Watanabe M. Modulation of cell growth and mutation induction by introduction
of the expresssion vector of human HSP70 gene. Experimental Cell Res 1994; 215:75-81.
39. Welte MA, Tetrault JM, Dellavalle RP et al. A new method for manipulating transgenes:
Engineering heat tolerance in a complex, multicellular organism. Current Biol 1993;
3:842-853.
40. Kim D, Ouyang H, Li G. Heat shock protein HSP70 accelerates the recovery of heat-shocked
mammalien cells through its modulation of heat shock transcription factor HSF1. Proc
Natl Acad Sci USA 1995; 92:2126-2130.
41. Lee JH, Hübel A, Schöffl F. Derepression of the activity of genetically engineered heat
shock factor causes constitutive synthesis of heat shock proteins and increased
thermotolerance in transgenic Arabidopsis. Plant J 1998; 258:269-278.
42. Prändl R, Hinderhofer K, Eggers-Schumacher G, Schoeffl F. HSF3, a new heat shock
factor from Arabidopsis thaliana, derepresses the heat shock response and confers
thermotolerance when overexpressed in transgenic plants. Mol Gen Genet 1998;
258:269-278.
43. Burke TJ, Callis J, Vierstra RD. Characterization of a polyubiquititn gene from Arabidopsis
thaliana. Mol Gen Genet 1988; 213:435-443.
44. Sun CW, Callis J. Independent modulation of Arabidopsis thaliana polyubiquitin mRNAs
in different organs of and in response to environmental changes. Plant J 1997; 11:1017-1027.
45. Hérouart D, Van Montagu M, Inzé D. Developmental and environmental regulation of
the Nicotiana plumbaginifolia cytosolic Cu/Zn-superoxide dismutase promoter in
transgenic tobacco. Plant Physiol 1994; 104:873-880.
46. Brown JA, Li D, Ic M et al. Heat shock induction of manganese peroxidase gene
transcription in Phanerochaete chryosporium. Appl Environ Microbiol 1993; 59:4295-4299.
47. Almoguera C, Coca MA, Jordano J. Tissue-specific expression of sunflower heat shock
proteins in response to water stress. Plant J 1993; 4:947-958.
48. Alamillo J, Almoguera C, Bartels D, Jordano J. Constitutive expression of small heat shock
proteins in vegetative tissues of the resurrection plant Craterostigma plantagineum. Plant
Mol Biol 1995; 29:1093-1095.
49. Kiyosue T, Yamaguchi-Shinozaki K, Shinozaki K. Cloning of cDNAs for genes that are
early responsive to dehydration stress in Arabidopsis thaliana L.: Identification of three
ERDs as HSP cognate genes. Plant Mol Biol 1994; 25:791-798.
50. Prändl R, Kloske E, Schöffl R. Developmental regulation and tissue specific differences
in the expression of a chimeric heat shock gene in transgenic tobacco and Arabidopsis
plants. Plant Mol Biol 1995; 28:73-82.
51. Sabehat A, Weiss D, Lurie S. The correlation between heat shock protein accumulation
and persistence and chilling tolerance in tomato fruit. Plant Physiol 1996; 110:531-537.
52. Collins GG, Nie XL, Saltveit E. Heat shock proteins and chilling sensitivity in mung bean.
J Exptl Botany 1995; 46:795-802.
53. Jennings P, Salteveit ME. Temperature and chemical shocks chilling tolerance in germinating
Cucumis sativus (cv. Poinsett 76) seeds. Physiol Plant 1994; 91:703-707.
54. Krishna P, Sacco M, Cherutti JF et al. Cold-induced accumulation of HSP90 transcripts
in Brassica napus. Plant Physiol 1995; 107:915-923.
55. Anderson JV, Li QB Haskell DW et al. Structural organization of the spinach endoplasmic
reticulum-luminal 70-kilodalton heat shock cognate gene and expression of 70-kilodalton
heat shock genes during cold acclimation. Plant Physiol 1994; 104:1359-1370.
56. Cabane M, Calvet P, Vincens P, Boudet AM. Characterization of chilling-acclimation-related
proteins in soybean and identification of one as a member of the heat shock protein
(HSP70) family. Planta 1993; 190:346-353.
57. Van Berkel J, Salamini F, Gebhardt C. Transcripts accumulating during cold storage of
potato (Solanum tuberosum L) tubers are sequence related to stress responsive genes.
Palnt Physiol 1994; 104:445-452.
58. Czarnecka E, Edelman L, Schöffl F et al. Comparative analysis of physical stress responses
in soybean seedlings using cloned heat shock cDNAs. Plant Mol Biol 1984; 3:45-58.
59. Edelman L, Czarnecka E, Key JL. Induction and accumulation of heat-shock-specific
poly(A+) RNAs and proteins in soybean seedlings during arsenite and cadmium treatments.
Plant Physiol 1983; 86:1048-1056.
60. Howarth CJ. Heat shock proteins in sorghum and pearl millet; ethanol, sodium arsenite,
sodium malonate and the development of thermotolerance. J Eptl Botany 1990; 41:877-884.
61. Orzech KA, Burke JJ. Heat shock and the protection against metal toxicity in wheat leaves.
Plant Cell Envir 1988; 11:711-714.
62. Neumann D, Lichtenberger O, Günther D et al. Heat-shock proteins induce heavy-metal
tolerance in higher plants. Planta 1994; 194:360-367.
63. Wollgiehn R, Neumann, D. (1995) Stress Response of tomato cell cultures to toxic metals
and heat shock: Differences and similarities. J Plant Physiol 1990; 146:736-742.
64. Davidson JF, Whyte B, Bissinger PH et al. Oxidative stress is involved in heat-induced
cell death in Saccharomyces cerevisiae. Proc Natl Acad Sci USA 1996; 93:5116-5121.
65. Polla BS, Kantengwa S, Francois D et al. Mitochondria are selective targets foe the protective
effects of heat shock against oxidative injury. Proc Natl Acad Sci USA 1996; 93:6458-6463.
66. Schuster G, Even D, Kloppstech K et al. Evidence for protection by heat shock proteins
against photoinhibition during heat shock. EMBO J 1988; 7:1-6.
67. Foyer CH, Descourviers P, Kunert KJ. Protection against oxygen radicals: An important
defense mechanism studied in transgenic plants. Plant Cell Environ 1994; 17:507-523.
68. Bowler C, Slooten L, Vandenbranden S et al. Manganese superoxide dismutase can
reduce cellular damage mediated by oxygen radicals in transgenic plants. EMBO J 1991;
10:1723-1732:
69. Van Camp W, Capiau K, Van Montagu M et al. Enhancement of oxidative stress tolerance
in transgenic tobacco plants overproducing Fe-superoxide dismutase in chloroplasts. Plant
Physiol 1996; 112:1703-1714.
70. McKersie BD, Chen Y, de Beus M et al. Superoxide dismutase enhances tolerance of
freezing stress in transgenic alfalfa (Medicago sativa L.). Plant Physiol 1993; 103:1155-1163.
71. Perl A, Perl-Treves R, Galili S et al. Enhanced oxidative stress defence in transgenic potato
overexpressing tomato CuZn superoxide dismutase. Theor Appl Genet 1993; 85:586-576.
72. Allen RD. Dissection of oxidative stress tolerance using transgenic plants. Plant Physiol
1995; 107:1049-1054.
73. McKersie B, Bowley S, Harjanto E et al.Water-deficit tolerance and field performance of
transgenic alfalfa overexpressing superoxide dismutase. Plant Physiol 1996; 111:1177-1181.
74. Schöffl F, Rossol I, Angermüller S. Regulation of the transcription of heat shock genes in
nuclei from soybean (Glycine max.) seedlings. Plant Cell Environ 1987; 10:113-119.
75. Baumann G, Raschke E, Bevan M et al. Functional analysis of sequences required for
transcriptional activation of a soybean heat shock gene in transgenic tobacco. EMBO J
1987; 6:1161-1166.
76. Schöffl F, Rieping M. Baumann G et al. The function of heat shock promoter elements
in the regulated expression of chimaeric genes in transgenic tobacco. Mol Gen Genet
1989; 217:246-253.
77. Czarnecka E, Key JL, Gurley WB. Regulatory domains of the Gmhsp17.5-E heat shock
promoter of soybean: A mutational ananlysis. Mol Cell Biol 1989; 9:3457-3463.
78. Wu C; Activating protein factors bind in vitro to upstream control sequences in the heat
shock gene chromatin. Nature 1984; 311:81-84.
79. Amin J, Anathan J, Voellmy R. Key features of heat shock regulatory elements. Mol Cell
Biol 1988; 8:3761-3769.
80. Xiao H, Lis JT. Germ line transformation used to define key features of heat shock
response elements. Science 1988; 239:1139-1142.
81. Perisic O, Xiao H, Lis JT. Stable binding of Drosophila heat shock factor to head to head
and tail to tail repeats of a conserved 5 bp recognition unit. Cell 1989; 59:797-806.
82. Sorger PK, Nelson HCM. Trimerization of a yeast transcriptional activator via coiled-coil
motif. Cell 1989; 59:807-813.
83. Kim SJ, Tsukiyama T, Lewis MS, Wu C. Interaction of the DNA binding domain of
Drosophila heat shock factor with its cognate DNA site: A thermodynamic analysis using
ultracentrifugation. Protein Sci 1994; 3:1040-1051.
84. Hübel A, Schöffl F. Arabidopsis heat shock factor: Characterization of the gene and the
recombinant protein. Plant Mol Biol 1994; 26:353-362.
85. Hübel A, Lee JH, Wu C et al. Arabidopsis heat shock factor is constitutively active in
Drosophila and human cells. Mol Gen Genet 1995; 248:136-141.
86. Spena A, Hain R, Ziervogel U et al. Construction of a heat-inducible gene for plants.
Demonstration of heat-inducible activity of the Drosophila HSP70 promoter in plants.
EMBO J 1985; 4:2739-2743.
87. Fernandez M, O’Brien T, Lis JT. Structure and regulation of heat shock gene promoters.
In: Morimoto RI, Tissières A, Gorogopoulus C ed. The biology of heat shock proteins
and chaperones. New Work: Cold Spring Harbor Press, 1994:375-394.
88. Topol J, Ruden DM, Parker CS. Sequences required for in vitro transcriptional activation of
a Drosophila heat shock gene. Cell 1985; 42:527-537.
89. Wu C. Heat schock transcription factors: Structure and regulation. Annu Rev Cell Dev
Biol 1995; 11:441-469.
90. Rieping M, Schöffl F. Synergistic effect of upstream sequences, CCAAT box elements,
and HSE sequences for enhanced expression of chimaeric genes in transgenic tobacco.
Mol Gen Genet 1992; 231,226-232.
91. Czarnecka E, Key JL, Gurley WB. Regulatory domains of the Gmhsp17.5-E heat shock
promoter of soybean: a mutational ananlysis. Mol Cell Biol 1989; 9:3457-3463.
92. Schöffl F, Schröder G, Kliem M. et al. A SAR sequence containing 395 bp fragment
mediates enhanced, gene-dosage-correlated expression of a chimaeric heat shock genen
in transgenic tobacco plants. Transgenic Res 1993; 2:93-100.
93. Giardina C, Perez-Riba M, Lis JT. Promoter melting and TFIID complexes on Drosophila
genes in vivo. Genes Dev 1992; 15:2737-2744.
94. Tsukijama T, Becker PB, Wu C. ATP-dependent nuclesome disruption at a heat shock
promoter mediated by binding of GAGA transcription factor. Nature 1994; 367:525-532.
95. Prändl R, Schöffl F. Heat shock elements are involved in heat shock promoter activation
during tobacco seed maturation. Plant Mol Biol 1996; 31,157-162.
96. Wehmeyer N, Hernandez LD, Finkelstein RR et al. Synthesis of small heat shock proteins is
part of the developmental program of late seed maturation. Plant Physiol 1996;
112:747-757.
97. Quatrano RS, Bartels D, Ho TH et al. New insights into ABA-mediated processes. Plant
Cell 1997; 470-475.
98. Hill A, Nantel A, Rock CD et al. A conserved domain of the viviparous-1 gene product
enhances the DNA binding activity of the bZIP protein EmBP-1 and other transcription
factors. J Biol Chem 1996; 271:3366-3374.
99. Jedlicka P, Mortin MA, Wu C. Multiple functions of Drosophila heat shock transcription
factor in vivo. EMBO J 1997; 16:2452-2462.
100. Mager WH, De Krujiff AJJ. Stress-induced transcriptional activation. Microbiol Rev 1995;
59:506-532.
101. Nover L, Scharf KD, Gagliardi D et al. The HSF world: Classification of plant heat stress
transcription factors. Cell Stress Chaperones 1996; 1:215-223.
102. Hübel A, Lee JH, Wu C et al. Arabidopsis heat shock factor is constitutively active in
Drosophila and human cells. Mol Gen Genet 1995; 248:136-141.
103. Czarnecka-Verner E, Yuan CH, Fox PC et al. Isolation and characterization of six heat
shock transcription factor cDNA clones from soybean. Plant Mol Biol 1995; 29:37-51.
104. Scharf KD, Rose S, Zott W et al. Three tomato genes code for heat stress transcription
factors with a region of remarkable homology to the DNA-binding domain of yeast HSF.
EMBO J 1990; 9:4495-4501.
105. Gagliardi D, Breton C, Chaboud A et al. Expression of heat shock factor and heat shock
protein 70 genes during maize pollen development. Plant Mol Biol 1995; 29:841-856.
106. Harrison CJ, Bohm AA, Nelson HC. Crystal structure of the DNA-binding domain of
the heat shock transcription factor. Science 1994; 263:224-227.
107. Vuister GW, Kim SJ, Orosz A et al. Solution structure of the DNA-binding domain of
Drosophila heat shock transcription factor. Struct Biol 1994; 1:605-614.
108. Schultheiss J, Kunert O, Gase U et al. Solution structure of the DNA-binding domain of
the tomato heat stress transcription factor HSF24. Eur J Biochem 1996; 236:911-921.
109. Peteranderl R, Nelson HCM. Trimerization of the heat shock transcription factor by a
triple-stranded alpha-helical coiled-coil. Biochem 1992; 31:12272-12276.
110. Rabindran SK, Haroun RI, Clos J et al. Regulation of heat shock factor trimer formation:
Role of a conserved leucine zipper. Science 1993; 259:230-234.
111. Nakai A, Morimoto RI. Characterization of a novel chicken heat shock transcription factor,
heat shock factor 3, suggests a new regulatory pathway. Mol Cell Biol 1993; 13:1983-1997.
112. Zuo J, Baler R, Dahl G, Voellmy R. Activation of the DNA-binding ability of human
heat shock transcription factor 1 may involve the transition from an intramolecular to
an intermolecular triple-stranded coiled-coil structure. Mol Cell Biol 1994; 14:7557-7568.
113. Orosz A Wisniewski J, Wu C. Regulation of Drosophila heat shock factor trimerization:
Global sequence requirements and independence of nuclear localization. Mol Cell Biol
1996; 16:7018-7030.
114. Lyck R, Harmening U, Höhfeld I et al. Intracellular distribution and identification of the
nuclear localization signals of two plant heat-stress transcription factors. Planta 1997;
202:117-125.
115. Sheldon LA, Kingston RE. Hydrophobic coiled-coil domains regulate the subcellular
localization of human heat shock factor 2. Genes Dev 1993; 7:1549-1558.
116. Zuo J, Rungger D, Voellmy R. Multiple layers of regulation of human heat shock
transcription factor 1. Mol Cell Biol 1995; 15:4319-4330.
117. Zandi E, Tran TN, Chamberlain W et al. Nuclear entry, oligomerization, and DNA-binding
of the Drosophila heat shock transcription factor are regulated by a unique nuclear
localization sequence. Genes Dev 1997; 11:1299-1314.
118. Newton EM, Knauf U, Green M et al. The regulatory domain of human heat shock factor 1
is sufficient to sense heat stress. Mol Cell Biol 1996; 16:839-846.
119. Wisniewski J, Orosz A, Allada R et al. The C-terminal region of Drosophila heat shock
factor (HSF) contains a constitutively functional transactivation domain. Nucleic Acids
Res 1996; 24:367-374.
120. Treuter E, Nover L, Ohme K et al. Promotor specificity and deletion analysis of three
heat shock transcription factors of tomato. Mol Gen Genet 1993; 240:113-125.
121. Scharf KD, Materna T, Treuter E et al. Heat stress promoters and transcription factors.
In: Nover L, ed. Plant Promoters and Transcription Factors. Springer-Verlag, Berlin, 1994;
125-162.
122. Giardina C, Lis JT. Dynamic protein-DNA architecture of a yeast heat shock promoter.
Mol Cell Biol 1995; 15:2737-2744.
123. Bonner JJ, Heyward S, Fackenthal DL. Temperature-dependent regulation of a heterologous
transcriptional activation domain fused to heat shock transcription factor. Mol Cell Biol
1992; 12:1021-1030.
124. Chen Y, Barlev NA, Westergaard O et al. Identification of the C-terminal activator
domain in yeast heat shock factor: independent control of transient and sustained
transcriptional activity. EMBO J 1993; 12:5007-5018.
125. Boorstein WR, Craig EA. Transcriptional regulation of SSA3, an HSP70 gene from
Saccharomyces cerevisiae. Mol Cell Biol 1990; 10:3262-3267.
126. Baler R, Zou J, Voellmy R. Evidence for a role of HSP70 in the regulation of the heat
shock response in mammalian cells. Cell Stress Chaperones 1996; 1:33-39.
127. Rabindran SK, WisniewskiJ, Li G et al. Interaction between heat shock factor and HSP70
is insufficient to suppress induction of DNA-binding activity in vivo. Mol Cell Biol 1994;
14:6552-6560.
128. Lee JH, Schöffl F. A HSP70 antisense gene affects the expression of HSP70/HSC70, the
regulation of HSF, and the acquisition of thermotolerance in transgenic Arabidopsis
thaliana. Mol Gen Genet 1996; 252:11-19.
129. Høj A, Jakobsen BK. A short element required for turning off heat shock transcription
factor: Evidence that phosphorylation enhances deactivation. EMBO J 1994; 3:2614-2624.
130. Knauf U, Newton EM, Kyriakis J et al. Repression of human heat shock factor 1 activity
at control temperature by phosphorylation. Genes Dev 1996; 10:2782-2793.
131. Chu B, Soncin F, Price BD et al. Sequential phosphorylation by mitogen-activated protein
kinase and glycogen synthase kinase 3 represses transcriptional activation by heat shock
factor 1. J Biol Chem 1996; 271:30847-30857.
132. Reindl A, Schöffl F, Schell J et al. Phosphorylation by a cyclin-dependent kinase modulates
DNA-binding of the Arabidopsis heat shock transcription factor HSF1 in vitro. Plant
Physiol 1997; 115:93-100.
133. Carratu L, Franceschelli S, Pardini CL et al. Membrane lipid perturbation modifies the
set point of the temperature of heat shock response in yeast. Proc Natl Acad Sci USA
1996; 93:3870-3875.
134. Gallie DR, Caldwell C, Pitto L. Heat shock disrupts cap and poly(A) tail function during
translation and increases mRNA stability of introduced reporter mRNA. Plant Physiol
1995; 108:1703-1713.
135. Gallie DR, Le H, Caldwell C et al. The posphorylation state of translation initiation factors is
regulated developmentally and following heat shock in wheat. J Biol Chem 1997;
272:1046-1053.
136. Pitto L, Gallie DR, Walbot V. Role of the leader sequence during thermal repression of
translation in maize, tobacco, and carrot protoplasts. Plant Physiol 1992; 100:1827-1833.
CHAPTER 6
Cellular Responses to Water Stress

Michael R. Blatt, Barbara Leyman and Alexander Grabov
P lant biomass is inextricably tied to water transpiration: increases in crop production

must appear inevitably in an increased water use, regardless of improvements in water
management or farming practice. Thus, on a global scale water use and the physiological
consequences of stress, when water becomes limiting, remain among the most important
factors that influence vegetative plant growth and yield.1 Although water limitation is an
obvious feature of arid environments, conditions of water stress develop in association with
saline soils and, hence, may also become evident where irrigation leads to a buildup of salt.2
Paradoxically flooding of soils, too, can lead to water shortage of the aerial parts of the plant
as the root environment becomes deprived of oxygen.3,4
The primary defense of the plant against water loss and dehydration, especially over
relatively short time scales, is to reduce transpirational water loss. It is now generally
recognized that plants respond to water stress by synthesizing the hormone abscisic acid
(ABA) which is transported via the xylem to the leaf tissues. Most important to this pro-
cess, ABA effects the closure of stomatal pores in the leaf epidermis and dramatically
reduces foliar transpiration, in many cases by two orders of magnitude or more. Foliar
conductance to water vapor of mesophytes and crop plants often lie in the range of 10-20
mm s-1 under conditions in which stomata are largely open, and these figures fall to values
near 0.1 mm s-1 or lower—equivalent to the cuticular conductance—when stomata close.6,7
In xerophytes and many trees, conductances under water stress can fall still lower to values
approaching 0.01 mm s-1.7 Clearly, understanding the factors that control stomatal aperture
will be crucial to future developments toward improving vegetative yields in the face of
increasing pressure on water resources and arable land usage.
At the same time, the guard cells that surround the stomatal pore have become a focus
of attention in fundamental research as well. The ability of these cells to integrate both
environmental and internal signals—and their unique situation within the leaf tissue—
has provided a wealth of experimental access points to signal cascades that link membrane
transport to stomatal control.8 The past 15 years of research has raised the status of the
guard cell to that of an undisputed higher-plant cell “model”, and some of the most
exciting new findings in plant cell signaling have come from work with guard cells, taking
advantage of the “stomatal interface” between molecular genetics and biophysics. In this
chapter we examine some of these recent developments and their background that form
the center of much debate about second messengers in plant cell biology. We also explore
their implications for plant response to water stress.
Molecular Responses to Cold, Drought, Heat and Salt Stess in Higher Plants, edited by Kazuo
The Stomatal Situation

Stomata are pores formed by pairs of specialized cells, the guard cells, and are found
on the epidermis of all aerial parts of most higher plants. Stomata open and close to control gas
exchange between the intercellular spaces within the plant tissue and the surrounding
environment. Thus, stomata have a fundamental role in controlling two of the most
important processes in vegetative plant physiology, photosynthesis and transpiration: they
open to allow sufficient CO2 to enter the leaf for photosynthetic carbon fixation, and they
close to reduce transpiration under conditions of water stress.
Stomata (sing. stoma, Greek for mouth) were some of the first microscopic structures
to be identified in higher plant tissues,9 coinciding with Hooke’s discovery of cells in the
second half of the 17th century, and early on were associated with "plant breathing".10 De
Candolle11 demonstrated that stomatal apertures were variable. However, recognition that
stomatal movements were driven by turgor came only later with von Mohl’s anatomical
and physiological studies.12 The relationship between the accumulation of inorganic K+
and stomatal opening was identified by Immamura13 and Yamashita14 in the first half of
this century, but it was not until Fujino15 published his work in English in 1967 that
scientists outside Japan were made fully aware of there significance. Fischer’s work16,17
independently confirmed these findings in epidermal peels of Vicia, concluding that the
amount of K+ together with an accompanying anion would be sufficient to account for the
increase in osmotic pressure during stomatal opening.
The mechanics of stomatal function is intimately connected with their morphology.
From an anatomical point of view there are two basic types of stomata, although intermediate
forms also exist in gymnosperms and sedges.7 The first type has kidney-shaped guard cells
that surround an elliptical pore in the epidermal surface. These stomata are typical of
dicotyledonous plants including Vicia, Nicotiana and Arabidopsis, but are also found in
Commelina. They remain the best-studied form in terms of guard cell membrane transport
and cell signaling. The second type of stomata are restricted to the monocotyledons and
have dumbell-shaped guard cells and an almost rectangular-shaped pore.18 In each case,
the swelling of the guard cells in opposition to one another leads to an opening of the
pore. However the mechanics differ according to the shape of the guard cells. Swelling
of the monocotyledonous guard cells is restricted by cell wall depositions primarily to the
dumbell-like ends, and opposition between the two cells thus opens the pore slitwise. In
kidney-shaped guard cells the cell-wall thickenings around each guard cell are most
pronounced on the side facing the pore. Cell expansion thus leads to an outward “bowing”
which opens the pore in an oval.
As may be expected, estimates of the change in guard cell volume between the closed
and open states of stomata vary between species. Volume changes of 0.2-0.35 pl change
per µm of pore aperture have been reported for Vicia,19,20 while values for Commelina may
be as high as 0.4 pl/µm.19 These values are significantly larger than our estimates for
Arabidopsis that fall between 0.03 and 0.08 pl/µm. Obviously such estimates depend on
the relative size of the guard cells and provide a rough guide only, because even in one
species guard cell size can vary, dependent on growth conditions and the age of the plant.7
However, the proportional changes in volume they imply are substantial, considering the
small size of guard cells compared with most other plant cell types. Dimensions of Vicia
guard cells, for example, are typically about 40 µm in length and 8-10 µm in diameter at
their widest point, giving volumes near 4 pl when apertures are between 6-8 µm.21 Thus
the volume changes between the closed state and fully-open state (typically 10-14 µm)
entail increases of 2- to 3-fold (200-300%).
Cellular Responses to Water Stress 101
Transport Mechanics
At maturity guard cells lack functional plasmodesmata,22 so all of the osmotic solute
flux that drives cell volume changes for stomatal movement must take place across the
plasma membrane. The transport mechanics that underpin these solute fluxes are now
well-established as a result of the electrophysiological and radiotracer analyses of the past
decade. Until recently, much of this work was carried out on Vicia, primarily because of
their ease of handling. However, comparable data are now to hand for guard cells of
Nicotiana23,24 and of Arabidopsis.25,26 These studies so far indicate only minor differences
between species in transport characteristics, although more subtle differences in their control
have yet to be examined systematically.
Plasma Membrane
The fluxes entailed—notably K+, Cl- and in some instances also organic acids such as
malate—are considerable: between the open and closed states of the stomata, guard cells
of Vicia take up and release, respectively, 2-4 pmol of KCl which on a cell volume basis is
equivalent to changes of approximately 300 mOsM in osmotically active solutes.7 These
events commonly take place over periods of 20 min to 2 h,27 implying a high level of
transport activity by comparison with many other higher-plant cells. In fact, there is
evidence that Vicia guard cells express unusually high levels of H+-ATPase at the plasma
membrane,28-30 consistent with the higher demand expected for membrane energization.
Likewise, measurements of H+-ATPase current under intracellular voltage clamp have
generally yielded values at 0 mV around 3-10 µA cm-2 (= 50-200 pA per cell).31-33 For
comparison a current of 35 pA can be estimated as the minimum required for stomatal
opening. This estimate assumes (probably unrealistically, in view of the energetic needs of
other homeostatic transport processes) that the H+-ATPase current is used entirely to drive
K+ uptake into the cell in exchange for H+ over a 2 h opening period. [It is worth noting
that initial values of 1.5-6 pA, based on patch electrode measurements34,35 were probably
compromised by cytosolic exchange with the patch electrode filling solution and the
consequent loss of regulatory and other cofactors.36 More recent “perforated patch”
techniques have gone some way to overcoming these difficulties and have yielded currents
of 10-20 pA.37,38]
Quantitative differences apart, primary energisation of the guard cell plasma membrane,
like the plasma membranes of most plant and fungal cells, is achieved by ATP-dependent H +
extrusion. At least two distinct H+-ATPase genes are expressed in Vicia guard cells,28
although the physiological characteristics of these isoforms remain to be examined.
Analysis of H+-ATPase function has shown that H+ extrusion is coupled to ATP hydrolysis in a
1:1 ratio33 and shares kinetic features in common with H+ pumps of Neurospora39 and
Chara.40 In theory the H+-ATPase will support stable membrane voltages near -360 mV
with an external pH of 6.33 However, this electrical driving force is utilized to power the
movement of other solutes (notably K+ and Cl- uptake) across the membrane, so the
theoretical limit is never achieved in practice. Membrane voltages near -300 mV have
been observed in low external K+,31 thus ruling out coupling ratios of 2:1 (H+:ATP) or
higher which would limit the H+- ATPase output to a maximum near -200 mV.33
Our knowledge of the mechanisms for K+ across the plasma membrane is dominated
by two classes of K+ channels that give rise to current rectifying inward (IK,in) and outward
(IK,out), respectively. These two pathways are major contributors to K+ flux during stomatal
opening and closing, respectively, and are clearly separable on bases of their biophysical
and pharmacological properties (below; see also ref. 41). However, guard cells probably also
possess secondary coupled transporters for K + 42 similar to those known in fungi 43
and other plant cells.44 Ion flux through channels is inherently passive and thus critically
dependent on the membrane voltage and equilibrium voltage for the permeant ions
(Ex). There are circumstances in which uptake of K+ occurs against the prevailing electro-
chemical driving force for K+ (V>EK), for example in low extracellular [K+] and in the
presence of fusicoccin.42
The major inward-rectifying K+ channel in guard cells of Arabidopsis has been
identified with the KAT1 K+ channel45 and its homologues that show single-channel
conductances of 5-8 pS at near-physiological [K+] as well as similar physiological and
biophysical properties when expressed in Xenopus oocytes46,47 and in yeast.48 Uptake of K+
through IK,in is demonstrable electrically in millimolar external [K+].31,49,50 These channels are
activated normally by membrane voltages more negative than -100 mV,37,51 they show an
apparent gating charge of about 1.4-1.5,37,51 and give rise to an inward-directed (by
convention, negative) current and corresponding K+ flux. The current appears to require
millimolar external [K+] for activity,51 consistent with a putative extracellular K+-binding
site with a Kd near 0.4 mM.52 However, K+ otherwise has no effect on IK,in gating or its
voltage sensitivity.
By contrast, the gating of IK,out is affected by extracellular [K+].53 This channel has a
somewhat higher single-channel conductance (=10-25 pS)54,55 and shows an apparent
gating charge near 2,53,56 implying a steeper dependence on membrane voltage. With 10
mM [K+]outside, activation of IK,out normally occurs at voltages near -70 mV and the
current is fully activated at voltages near 0 mV and more positive. Furthermore unlike
IK,in, the IK,out gate is affected by extracellular [K+]so that its voltage dependence shifts in
parallel with EK.26,53,57 This current has been indicated as the major pathway for K+ efflux
during stomatal closure,42,58 so its dependence on extracellular [K+] makes good “physiological
sense” in an environment in which [K+] can vary over more than an order of magnitude: it
ensures that K+ passage can only occur when the driving force on the ion will lead to its
efflux from the cell.57 Blatt and Gradmann53 carried out a detailed analysis of the current
as a function of [K + ] and found that its activation depended on the
co-operative interaction of 2 K+ ions with the channel, but at a site (or sites) distinct from
the channel pore. They also observed that the apparent K1/2 for interaction was strongly
voltage-dependent, accounting for the equivalence of (negative) membrane voltage and
[K+] in regulating channel activity. Thus, their data indicate that extracellular K+ acts as a
ligand and negative regulator of IK,out, and point to a pair of K+-binding sites associated
with the channel deep within the membrane, but accessible only from the outside of
the cell.
To an extent, the fluxes of K+ carried by these channels—and hence the accompanying
electrical charge—are balanced by parallel transport of Cl. Anion efflux is particularly
important for stomatal closure, as it is essential that the voltage across the guard cell plasma
membrane is brought positive of EK to effect a net loss of K+ through IK,out. Because
cytosolic [Cl-] under most conditions is probably close to 1 mM,59 ECl typically will be
situated close to and positive of 0 mV, that is well positive of EK and sufficient to drive the
membrane voltage positive of EK if the membrane conductance to Cl- is elevated.
Two anion channels have been identified with different macroscopic current
kinetics, although both exhibit roughly equivalent single-channel conductances (≅35
pS). One of these anion currents activates rapidly with halftimes of less than 50 ms on
depolarization to voltages positive of about -100 mV and deactivates with halftimes of
2-5 s on repolarization.60 Once activated, these channels also inactivate with a halftime of
about 10 s, thus disabling the current even when the voltage is held positive of -100 mV.61
The second anion current (hereafter identified as ICl) has been reported to activate and
deactivate slowly (halftimes, 5-30 s), shows no inactivation with time,62,63 and exhibits a
significant residual conductance at voltages negative of -100 mV.23,62,63
Dietrich and Hedrich64 have suggested an interconversion between these two channel
types, proposing a single Cl- channel exists in one of two different gating “modes.”65-67
However, the relationship between the two channel forms remains to be examined in any
detail. Regardless of origin, the fast-activating anion current can be ruled out as a
major pathway for anion efflux during stomatal closure. Its narrow voltage range for acti-
vation and the fact that the current inactivates within seconds cannot be reconciled
with the requirement for sustained channel functioning over periods of 20-30 min or
more during closure.8,68 In fact, recent evidence has confirmed that ICl responds rapidly and
with prolonged activation to stimuli that evoke stomatal closure.23,25,69
Tonoplast
The greater proportion of solutes that pass across the plasma membrane during
stomatal movements must also find their way across the tonoplast. As is the case for
virtually all mature, higher-plant cells the vacuole in guard cells makes up 80-90% of the
total cell volume. Thus most of the volume changes as the guard cells swell and shrink are
taken up in the vacuole7 and the fact implies a degree of transport coordination at the
tonoplast and between the two membranes.70,71 By contrast with events at the plasma
membrane, our understanding of transport across the tonoplast remains very poor
indeed. There has been remarkably little work on the pumps that might energize the
tonoplast in guard cells. All evidence again indicates that significantly higher vacuole-type
ATPase activities are found in these cells compared with leaf mesophyll.72,73 However,
Willmer, et al73 could not find evidence for its control by extrinsic stimuli. Furthermore,
nothing is known of the mechanisms that are responsible for solute accumulation,
although it is likely that the cell must expend energy at least for cation transport into the
vacuole.7
The guard cell tonoplast is known to harbor at least two different cation channels
that are capable of carrying K+ and Ca2+ 74-76 and an anion channel with high selectivity for Cl-
over K+ and dependent on protein phosphorylation for activity.77 Pei, et al77 have suggested
that the anion channel could be a pathway for Cl- accumulation. The cation channels
are more likely to be important to events of solute loss and stomatal closure. Each may
prove important targets for hormonal and/or environmental control, but their respective
contributions still remain to be demonstrated. Equally, the vacuole constitutes an important
source and sink for Ca2+ and H+ generally and, thus, is likely to be a contributor to
signaling events that lead to changes in the free concentration of these ions in the
cytosol. Indeed, the vacuole clearly contributes to hormone-mediated changes in pHi
as well as its homeostatic control78 and it is also thought to comprise a key reservoir
for Ca2+ release and stimulus-evoked changes in [Ca2+]i,79 but much detail is still missing.
Transport Coordination in the Face of Stress

Abscisic acid acts as a universal signal in plants to prevent stomatal opening and
promote stomatal closure. The potency of ABA to inhibit transpiration was first discovered in
the late 1960s 80,81 and these observations were soon linked with early studies on ABA
accumulation under water stress conditions.82,83 It is now recognized that during the
vegetative phase of the plant life cycle ABA mediates generally to signal adverse environ-
mental conditions including cold, drought and high salinity (see especially chapters 4 and
5, this volume). During water stress conditions ABA accumulates in leaf tissues around the
guard cells and promotes stomatal closure to reduce transpirational water loss84 (for
extensive reviews of the literature see refs. 5,7).
Abscisic acid will initiate stomatal closure within minutes of its application to
epidermal strips, to the surface of leaves and via the transpiration stream to intact
leaves. However, stomatal responses can vary even between stomata on one leaf (so-called
“patchy” or mosaic behavior85,86). Stomatal response to ABA is subject to environmental
factors such as external [K+]17 and [Ca2+],87 CO2,18,88 circadian period89 and prior stress
or exposure to ABA.90 There is also a more prolonged effect of ABA that is observed when
plants are first relieved from a period of water stress. Stomata then often remain partially
closed, a characteristic that can extend over periods of hours or even days,91 possibly through
an increased sensitivity to residual ABA levels90 that may reflect an enhanced expression of
signaling components.92 These observations imply a high degree of coordination in
coupling ABA to stomatal movements; they also underscore the integration of ABA-
mediated events with other signal cascades within the cells.
How is such a degree of regulation achieved? To this question a very large body of
data can now be brought to bear. In fact, for guard cells we know more about the
downstream events of evoked transport control than for any other higher-plant cell. In
the first instance, concerted regulation of sets of transporters, including ion channels and
pumps at the plasma membrane and endomembranes, is essential to achieve the requisite
solute flux. For ABA the events leading to stomatal closure requires the modulation of all
of the major ion currents at the plasma membrane to effect K+ and Cl- loss from the guard
cells. Additional effects at the tonoplast as well as other endomembranes are anticipated,
both in relation to bulk K+ salt efflux and to signals dependent on Ca2+ and pH. In the
simplest of schemes, the events that ensue on ABA exposure can be ordered as follows
(Fig. 6.1):
1. ABA evokes the development of an inward-directed current, mediated at least in
part by slow-activating anion channels23,25 to depolarize the membrane and generate a
driving force for K+ efflux;31
2. It inactivates IK,in that normally mediates in K+ uptake;93 and
3. It activates current through IK,out that, together with the anion channels, facilitates
the net loss of salt from the cells.8,93
At least two parallel signaling pathways are known to underpin K+ and anion channel
behavior alone, in addition to alterations in H +-ATPase activity.31,94 Current evidence
supports the idea of [Ca2+]i- and pHi-mediated signaling in guard cells, and the influence
of these signal cascades and their downstream targets overlap significantly. However,
whether both cascades originate with the same receptor-binding event and, more still,
where the site (or sites?) of ABA perception occurs remains speculative. Furthermore, it
has become increasingly clear that signaling pathways evoked by other stimuli such as
CO269 converge with ABA-mediated control of plasma membrane transport independent of
these known second messengers. Thus, as many questions are still unanswered about ABA-
evoked stomatal closure. In the remainder of this chapter we review several of these issues
and their background that form the center of current debate about guard cell signal
transduction and transport control.
The Ca2+ Second Messenger

Evidence of a role for cytosolic-free [Ca2+] ([Ca2+]i) in regulating K+ channels of
guard cells was uncovered early on and has remained a focus of interest, as it has subsequently
for anion channel control in relation to ABA. Elevating [Ca2+]i to micromolar concentrations
greatly reduces I K,in 95 and alters its voltage dependence while promoting both
anion currents (see also refs.60,63). These characteristics mimic the channel responses to
ABA.31,93,96 Because [Ca2+]i has been observed to rise during ABA exposures from rest
near 100 nM to values on some occasions exceeding 1 µM,97-101 the idea that [Ca2+]i
Fig. 6.1. ABA leads to a concerted modulation (+)=increase or activation, (-)=decrease or

inactivation of at least three subsets of plasma membrane ion channels—including the
two dominant K+ currents IK,in and I K,out, and the slow-activating anion current ICl—and,
probably, the H+-ATPase. A scheme for ABA signal perception including hypothical and
known steps entails binding to an integral plasma membrane receptor (1), possibly also
internal ABA binding sites not shown. Binding may activate a G-protein (G) and trigger
activation of phospholipase C (PLC)-mediated hydrolysis of phosphoinositol bisphosphate
(PIP2) to inositol 1,4,5-trisphosphate (IP 3) releasing Ca2+ from intracellular stores (2).
ABA also affects Ca2+ entry across the plasma membrane and its interaction with intracellular Ca2+
release pathways (shading) leads to high-gain Ca2+-induced Ca2+ release (3). The consequent rise
in Ca2+i affects IK,in, ICl and the H+-ATPase. A concurrent, but Ca2+-independent rise in
pHi (4) acts on I K,in , I K,out and I Cl , as well as depleting substrate for the
H+-ATPase. The abi1 protein phosphatase, and by inference also protein kinases (PK/PP),
gate pHi signal transmission (5) to IK,in and IK,out but do not influence ICl. A 2B-type protein
phosphatase may also gate the Ca 2+- sensitivity of IK,in as well as vacuolar Ca 2+ release not
shown. Alterations in phosphorylation state (P*) may mediate in activating ICl directly as
well as modulating the characteristics of other plasma membrane and tonoplast ion channels/
transporters.
increases are crucial to ABA stimulus transduction remains a potent concept. Indeed,
simply raising [Ca2+]i can be sufficient to bias the membrane for solute efflux by controlling
two key ion currents, IK,in and ICl.
Attaching a physiological interpretation to [Ca2+]i changes has not been straightforward
even so. A major problem with the [Ca2+]i hypothesis for guard cell signaling has rested in
the fact that the rise in[Ca2+]i is not universally associated with ABA and related stimuli, or
even stomatal closure.98-101 Fricker, et al99 have pointed out the technical difficulties inherent
to fluorescent measurements of [Ca2+]i in higher-plant cells:
1. That the intrinsic cytosolic volume is small, either sandwiched between plasma
membrane and tonoplast or dominated by perinuclear cytoplasm, making it difficult
to obtain reliable measurements with high spatio-temporal resolution; and
2. That changes in [Ca2+]i relevant to ion channel control almost certainly occur
locally, adjacent to the plasma membrane.
However, environmental factors including growth temperature102 that are "secondary" to
the ABA stimulus can play a critical role in determining the scope and magnitude of the
[Ca2+]i signal. One likely explanation for the influence of secondary factors relates to the
transport “poise” of the cell itself. Recent studies from this laboratory103 have linked changes
of [Ca2+]i in ABA to the voltage across the plasma membrane. These observations suggest
that the [Ca2+]i response may be part of an adaptive feedback loop adjusting the ABA
response of the guard cells to the prevailing requirements for solute flux (see [Ca2+]i
Oscillations... below). So it is all the more remarkable that no further kinetic detail has
been forthcoming relating the activities of either IK,in or ICl to [Ca2+]i. It will be important
to know now whether small increases of [Ca2+]i, for example to values near 0.3-0.5 µM
such are often observed in ABA,97-101 could account for the characteristics of these currents in
the presence of ABA. It is interesting that IK,in was largely unaffected by high [Ca2+]i when
EGTA was used as the Ca2+ buffer in patch-clamp experiments. Kelly, et al104 have
suggested that this dependence on the buffer is related to the dynamics of its binding with
Ca2+ as, unlike BAPTA, EGTA shows a relatively slow Ca2+-binding rate.105-108 Nonetheless, as
these measurements were carried out under steady-state conditions in which [Ca2+]i was
held constant, it is not clear how the relaxation kinetics of Ca2+-buffer binding can
account for such a difference. Whatever the explanation for their observations, the data
make it clear that estimating the Ca2+-sensitivity of IK,in in vivo is not straightforward in
patch clamp experiments.
[Ca2]i Oscillations and the Origins of [Ca2]i Increase

From where does the Ca2+ originate that contributes [Ca2+]i increases? Clearly, the
apoplast is the ultimate source for all cytosolic Ca2+. Entry of Ca2+ across the plasma
membrane also contributes to the [Ca2+]i rise evoked by ABA. In an elegant study of K+ flux
control by ABA, MacRobbie109 demonstrated that the K+ (86Rb+) efflux evoked by ABA in
Commelina guard cells displays two peaks. The initial, rapid stimulation in efflux in the first
1-2 min was probably associated with membrane depolarization and the shift in electro-
chemical driving force for K+ across the plasma membrane. A second, and much slower
stimulation—with a maximum some 10 min after ABA addition—was thought to represent
K+ (86Rb+) release from the vacuole. The slow rise in efflux could also have reflected
longer-term regulatory effects at the plasma membrane. But regardless of its origin, this
second peak in efflux required the presence of extracellular Ca2+. In accord with these
observations, Ca2+ channel blockers have been found to affect ABA-induced stomatal
closure.110 Thus, a simple interpretation is that Ca2+ entry from outside is important to
maintaining osmotic solute flux over the time periods of tens of minutes required for
stomatal closure.
A large body of data also supports a role for Ca2+ release from internal stores during
ABA stimulation—and in guard cell signaling generally—although there remains some
uncertainty about the relative contributions of various pathways. Blatt, et al111 and Gilroy,
et al112 showed that cytosolic inositol-1,4,5-trisphosphate was capable of stimulating a rise
in [Ca2+]i, even in the presence of extracellular Ca2+ channel blocker La3+ to block
Ca2+ entry from outside, and of inactivating IK,in. Subsequent studies have added further
support to the idea of a signaling sequence that passes through inositol-1,4,5-trisphosphate-113-115
and, more recently, also through cADPR- (ryanodine receptor) mediated Ca 2+
release.116,117 These studies have indicated parallels with animal cell models, although the
monophasic dependencies on cADPR reported for Ca2+ release116 are at odds with the
biphasic cADPR characteristics that have been observed in this case.118
Interest in the [Ca2+]i hypothesis in guard cells has been still further heightened by
recent observations that [Ca2+]i may oscillate in response to external stimuli.119,120 It is
plausible that these characteristics constitute “Ca2+ signatures” to encode specific responses,
much as has been argued for animal cells,121,122 and the analogy raises questions about the
origins of such oscillations in guard cells. In many animal cells, [Ca2+]i signals depend on
entry of Ca2+ across the plasma membrane, its release from intracellular stores and its
subsequent elimination from the cell or sequestration within organelles.118,122 Coordination of
these processes gives rise to short-lived and repetitive [Ca2+]i spikes, with durations of a
few seconds, that may serve to frequency-encode cellular responses, including gene
expression.123 However, the oscillations reported in guard cells to date are at least an order
of magnitude slower, with individual cycle durations of minutes and occurring with periods of
5-10 min or more.
It now seems that the situation in guard cells may differ significantly from the animal
models, and that these slow [Ca2+]i fluctuations are driven by oscillations in membrane
voltage. Guard cells commonly show two states of membrane voltage,31,124 one state situated
close to and positive of the K+ equilibrium voltage (EK), and the second characterized by
voltages well negative of EK. Transitions between these two states occur spontaneously and
are potentiated by various stimuli, including ABA;31,49 they are rapid, often exceeding 150
mV in amplitude and can occur in cyclic oscillations similar to cardiac action potentials,
but with periods of tens of seconds to minutes. Our recent work103 has shown that mem-
brane hyperpolarization during such oscillations or under voltage clamp evokes a Ca2+
influx across the plasma membrane and initiates a wave of high [Ca2+]i that propagates
centripetally from the cell periphery. Hyperpolarization beyond about -120 mV is essential to
evoke the response and evidence from pharmacological analyses indicates that the process
depends also on intracellular release from Ca2+ stores.
The observations imply a capacity for Ca2+-induced Ca2+ release (CICR) in the guard
cells that parallels established patterns of CICR coupling in animal cells.122,125 Nonetheless, the
data underscore some important differences to the animal models, notably the dependence on
voltages more negative than -120 mV. By contrast, voltage-evoked CICR in neuromuscular
tissues is normally triggered by membrane depolarization that activates Ca2+ influx through
pharmacologically-distinct, L-type Ca2+ channels.125-127 Both the acute voltage-sensitivity
and [Ca2+]i relaxation kinetics clearly distinguish the present data from one previous
report of [Ca 2+]i transients in Vicia guard cell protoplasts. Schroeder and Hagiwara128
proposed that Ca 2+ release might be triggered after Ca 2+ entry via nonselective,
Ca2+-permeable channels. However these experiments do not convince, because Ca2+-medi-
ated activation of Cl- channels was not ruled out. Furthermore, exchange of the cytosol
with the patch electrode filling solution in these experiments meant that dynamic control
of [Ca2+]i was lost. By contrast, voltage clamp using intracellular microelectrodes leaves
the cytosol largely intact.68,129
Most remarkable, we found that voltage and ABA acted in concert to potentate the
[Ca2+]i signal, with ABA affecting both the voltage threshold for the [Ca2+]i response and
its kinetics. These observations suggest that targets of ABA action may include the voltage-
dependence of gating for the Ca2+ influx pathway itself and activation of intracellular Ca2+
release elements. One explanation may be that Ca2+ channel activation occurs via protein
phosphorylation,79 although these actions of ABA are by no means the only possibilities.
Furthermore, voltage appeared to be a determining factor for [Ca2+]i increases evoked by
ABA. Indeed, with membrane voltage under experimental control, ABA elicited an
appreciable [Ca2+]i increase only once the membrane was driven to voltages near and
negative of -100 mV.
Perhaps it is not surprising, then, that increases in [Ca2+]i in (non-voltage-clamped)
guard cells have never fully correlated with stomatal closure in ABA.98,100,128 Membrane
voltage differs between cells and its distribution between voltage states in a population of
guard cells shows seasonal variability.21,31,38 With 1-10 mM K+ outside the voltage can be
situated positive of -60 mV.31,58 Yet [Ca2+]i increases require that the membrane be
situated at a voltage near or negative of -100 mV. So, the variability in [Ca 2+ ] i
increases evoked by ABA seems very likely to be a direct consequence of the voltage
state of the plasma membrane.
The H+ Second Messenger

Increases in [Ca2+]i do not mimic the effects of ABA in controlling the K+ channels that
mediate IK,out at the guard cell plasma membrane. Thus, additional control elements distinct
from any [Ca2+]i-related events must take part in ABA signal transduction. In fact, this class
of K+ channels has proven singularly insensitive to [Ca2+]i54,95,130 and to factors affecting
upsteam intermediates of [Ca2+]i-related signal cascades, although in each case an influence
on IK,in was reported complementary to a rise in [Ca2+]i. The current has been reported to be
insensitive to G-protein agonists and antagonists such as GTPγS, GDPβS, cholera and
pertussis toxins,131 to mas7, a serpentine receptor mimetic,132 and to inositol-1,4,5,-
trisphosphate.111 Yet, IK,out is activated in a roughly scalar manner in the presence of ABA31,93
which can evoke as much as a 14-fold (1400%) increase in the current at any one voltage.96
Despite early indications of its potential importance in signaling,133-135 a role for
cytosolic-free [H+] (pHi) has surfaced only slowly. At first stimulus-evoked changes in pHi
were linked to control of the H+-ATPase by auxin and fusicoccin136-139 and, hence, to H+ as
a transported substrate rather than as a second messenger. The proposal of pHi as a major
intermediate in ABA signaling rests on two key experimental developments:
1. ABA was found to alkalinize the cytosol in guard cells58,97 as well as in the
mesophyll of Zea and Petroselenium;140 and
2. Experimentally-imposed changes in pHi were found to alter guard cell K+ channels
in a manner distinct from that of changes in extracellular pH.51,58
The first point is important, because the findings implicitly separated the changes in
pHi from the H+- ATPase. Common dogma maintained that ABA reduces H +-ATPase
activity141,142 and it was expected therefore that decreasing H+ extrusion via the pump
should decrease pHi (raise [H+]i). Here was a clue that hormonal control of pHi might
depend on more than just changes in the pool of substrate for primary transport. The
second point, furthermore, ascribed to pHi an action that could not be result directly from
H+ flux across the plasma membrane and, instead, called into question an alternative
function as an internal signal to control ion channels.
Consistent with the idea of a pHi signal intermediate, “pH clamp” experiments demonstrated
that an increase in pHi is both necessary and sufficient to account for the activation of IK,out in
ABA.58 In every case, the current response was fully accommodated as a simple function
of [H+]i without recourse to additional controls, whether guard cells were loaded with pH
buffers or with the weak acid butyrate to suppress changes in pHi with ABA and to raise and
lower pHi in the absence of ABA. Subsequent work also pointed to pHi more generally in
channel control, for example that initiated by auxin,49,143 suggesting a more central role in
cellular stimulus-response coupling in guard cells. Indeed, the evoked changes in pHi
recorded by Thiel, et al143—roughly 0.5 pH units within 20 s, at a conservative estimate—
rivals evoked changes of [Ca2+]i in many plant cells, both in amplitude and kinetics, and
remains difficult to explain in terms of H+ transport across either plasma membrane or
tonoplast.
The action of pHi on IK,out shows an unusual “fingerprint”. Blatt and Armstrong58
found that increasing pHi activated IK,out in a voltage-independent manner, that is the
current rose roughly in proportion at all voltages, and titration of the current as a function
of pHi suggested a cooperative binding of two H+ to effect the response. These results
have since been extended in combined voltage clamp and BCECF fluorescence ratio
measurements of pHi130 to demonstrate that pHi has little effect on the kinetics of IK,out
activation. Instead, it appears to mobilize K+ channels into a pool that are activated on
membrane depolarization. Miedema and Assmann 144 recently confirmed these
observations, showing that the action of pHi results from H+ interaction with binding
sites closely associated with the membrane, possibly with the channel or with closely
associated protein(s).
The action of pHi extends beyond control of IK,out. Along with its sensitivity to [Ca2+]i
(above), IK,in—or one of the major signaling elements controlling these channels—responds
to pHi and, significantly, with a fingerprint that is fundamentally different to that of IK,out.
Initial studies pointed to a pHi dependence opposing that of IK,out in this case. Thus,
reducing pHi was found to activate IK,in and the current was suppressed by alkaline pHi
loads.51,58 Titration of IK,in130 has since demonstrated that the current behaves as a simple
function of [H+]i, again in contrast to the cooperativity between H+ observed for IK,out.
The data indicate a H+ binding at a site with a pKa near 6.4, consistent with the titration of
a single histidine residue and consonant with several recent studies of cloned K+ channels
from yeast145 and mammalian neuromuscular tissues.146
One immediate question is whether pHi action in this latter instance could be
mediated by H+ titration of Ca2+ binding sites, in other words by interference with Ca2+
signal transmission. The answer is negative, but this reply belies the complexity of the
situation (see pHi Interaction with [Ca2+]i below). The Ca2+ and H+ messengers can be
shown to function independently, converging finally on the K+ channels to control their
activity via two kinetically-distinct mechanisms.130 Two lines of evidence support this
conclusion:
1. pHi- mediated control of IK,in is observed in the absence of any measurable changes
in [Ca2+]i; and
2. The effect of pHi is evident predominantly as a voltage-independent change in IK,in
conductance, whereas the action of [Ca2+]i is profoundly voltage-sensitive.
Thus, in a very straightforward sense, the actions of pHi on IK,in can be distinguished
from those of [Ca2+]i. Not surprising then, additional studies — including work with ABA24,58
demonstrating IK,in inactivation concomitant with a rise in pHi when measurements failed
to show a rise in [Ca2+]i — have indicated that changes in pHi can predominate in regulating
IK,in in leu of any changes in [Ca2+]i.
Generality of the pHi Signal

The preceding evidence aside, there occurs at least one circumstance for which pHi
does not appear to play a major role in ion channel control. Stomata normally close in
response to a rise in the partial pressure of CO2 (pCO2), much as they do in the presence of
ABA. Like ABA, elevated pCO2 promotes solute efflux from the guard cells7. Indeed, CO2
and ABA interact closely in control of stomatal aperture.18,88 So, CO2 might be anticipated
to draw on a similar sequence of changes in K+ and anion channel activities to facilitate
these ion fluxes. The expectation has been borne out. Recent voltage clamp studies with
Vicia guard cells69 demonstrated that raising pCO2 from ambient (350 µ11-1) to 1000 µ l-1
evokes a rapid inactivation of IK,in and activation of IK,out as well as enhancing the anion
channel current and altering its voltage-dependent kinetics. Significantly, parallel measure-
ments with the H+- sensitive fluorescent dye BCECF failed to uncover any measurable
change in pHi and, in 10,000 µ11-1 pCO 2 a small 0.1-0.2 unit decrease in pHi was even
observed.
The observations make sense from the physiological standpoint. A rise in pCO2
increases the bicarbonate dissolved in aqueous solution which, as a weak acid, will favor
cytosolic acidification.147 Yet the effect must inevitably suppress K+ salt loss from the guard
cells and stomatal closure, because decreasing pHi inactivates IK,out. So, the existence of an
alternative control pathway circumvents the potential dilemma otherwise inherent when
the response to a weak acid must be to facilitate K+ efflux for stomatal closure. Because
elevating pCO2 actually promoted the activity of IK,out even in the face of mild decreases in
pHi, Brearley, et al69 concluded that the CO2 signal must be transmitted via another, as yet
unidentified, signal cascade. For the present, however, the identify of this third signaling
pathway is a mystery.
Origin of Changes in pHi

What are the possible origins for these pHi changes? Because of the high buffer
capacity of guard cells [≅100 mM/(pH unit)]58,185 a pHi rise of 0.2-0.3 units, such as
occurs in ABA, requires either that approximately 30 mM H+ are eliminated from the
cytosol or that a comparable increase in H+ binding and sequestration occurs within the
cell. Of the possible sinks for H+, transport by the H+-ATPase across the plasma membrane can
be ruled out, because the large flux of H+ would entail substantial currents (approx. 10 µA
cm-2 over 5 min). Such H+ currents have not been observed in response to ABA.23,31,93 Two
other potential sinks for H+ are vacuolar H+ transport, and metabolic H+ consumption
during organic acid breakdown and CO2 release. Neither of these options excludes the
other and both are attractive. The second possibility would tie the pHi signal to the
metabolic turnover of osmolytes that is known to take place during stomatal closure. Guard
cells of many higher plants, including Vicia, use malate as well as Cl- to balance K+ uptake
during stomatal opening.7 Much of this malate is decarboxylated during closure, thereby
consuming cellular H+. Movement of H+ into the vacuole is an equally likely explanation
for the pHi rise, and could contribute to charge balance during K+ efflux from the vacuole.
Roughly 50-70% of the vacuolar K+ salt content passes across the tonoplast before exiting
the guard cell during stomatal closure.7,148,149 A cursory estimate shows that to account for
a 0.3 pHi unit rise, that is approximately 30 mM H+, only 2-4% of the K+ efflux need be
balanced by H+ entry into the vacuole. In fact, vacuolar acidification on this scale is known
to take place during closure.150 Recent studies also lend support to this argument:
Frohnmeyer, et al78 observed that evacuated guard cells and mesophyll protoplasts lose the
ability to dynamically buffer pHi and show no change in pHi in response to hormone
treatments.
Protein Phosphorylation
Channel regulation by phosphorylation/dephosphorylation is indisputably a third
major element coupling ABA, and probably other stimuli, to solute flux during stomatal
movements. Even before a link was established to physiologically-related events,

pharmacological evidence implicated protein kinase and protein phosphatase activities in
maintaining a homeostatic balance of channel function. Regulation of IK,in by a 2B-type
(Ca2+-dependent) protein phosphatase was identified through its sensitivity to FK506,
cyclophilin-cyclosporin A and related compounds in patch electrode studies151 and time-
dependent block of both IK,in and IK,out by low concentrations of okadaic acid implicated a
dependence on protein phosphatase 1/2A activity.152,153 More recent studies have indi-
cated that both the tonoplast SV channel79 and the tonoplast anion channel77 are also
subject to control by phosphorylation. Because ABA, among other stimuli, activates Ca2+-
dependent protein kinases,154,155 it is conceivable that ABA (and Ca2+) action could be
mediated in part through a cascade of protein (de-)phosphorylation.
In the event, direct evidence for protein phosphatase (and by inference, protein
kinase) activity in ABA signaling has come from combined molecular genetic and electro-
physiological studies of plants carrying the mutant abi1 gene of Arabidopsis. The abi1 gene
was originally identified as one of a family of abscisic acid-insensitive mutants that germi-
nated and grew on high concentrations of ABA.156 It is linked to a number of ABA-related
phenotypes, including an aberrant control of stomatal aperture and consequent wilty
characteristic. The ABI1 amino-acid sequence indicated that the protein is type 2C
protein (serine/threonine) phosphatase (PP-2C),157,158 and subsequent work demonstrated
that the ABI1 gene product displays Mg2+-dependent protein phosphatase activity but is
Ca2+-insensitive.159 In vivo, the wild-type gene will partially complement a yeast PP-2C
mutant, and the mutant abi1 (dominant negative) gene confers a subset of phenotypes in
transgenic Nicotiana similar to those of the Arabidopsis mutant, including the strong
tendency to wilt.24 Thus, by association, the abi1 phenotype implicates protein phos-
phorylation in ABA-mediated control of guard-cell ion transport.
Armstrong, et al24 have since linked the mutant phenotype with aberrant control of
the guard cell K+ channels using abi1-transgenic tobacco. These studies showed that the
abi1 gene not only reduced IK,out in the absence of ABA, but eliminated the response of
the current in its presence. The transgene was also found to interfere with ABA-evoked
control of IK,in and to block stomatal closure. Furthermore, the background of IK,out
activity, as well as IK,in and IK,out responses to ABA and stomatal closure could be “rescued” by
treatments with protein kinase antagonists. So, how can abi1 gene action be understood?
The simplest explanation is that the (dominant) mutant protein phosphatase interferes
with a wild type homologue in tobacco preventing protein dephosphorylation. The
kinase antagonists, then, redress the consequent imbalance in phosphorylation of the
target protein(s) to reinstate K + channel sensitivity to ABA. This interpretation is
entirely consistent with the phosphatase activities shown by the wild type and mutant
gene products.159
Protein (de-)phosphorylation also affects anion channel function. Evidence from
pharmacological studies have shown that protein phosphatase antagonists such as okadaic
acid and calyculin A (both protein phosphatase 1/2A antagonists) alter anion channel
characteristics152 and rundown in patch electrode recordings of guard cell protoplasts.160
Recently, ICl was found to be activated by ABA in guard cells of N. benthamiana,23 and
Arabidopsis25 in a manner that depended on the phosphorylation status of the cell. Pei, et
al25 recorded anion currents from protoplasts after obtaining patch seals either in the pres-
ence or absence of ABA and/or with additions of okadaic acid. Statistical comparisons
indicated that the current was activated by ABA and that this activation could be suppressed by
okadaic acid. The effect appeared to be a purely scalar increase in the number of channels
in the active pool. Furthermore, the ABA-mediated increase in I Cl was not
observed in abi1 mutant plants or in Arabidopsis carrying the mutant abi2 gene that
encodes an abi1 homologue,161 although only in the former case was the impairment
rescued in the presence of the kinase antagonist K252A. Grabov et al,23 by contrast, made
use of intracellular microelectrode methods, recording anion currents from individual
guard cells before and throughout treatments with ABA and calyculin A. They found that
ABA reversibly stimulated the anion current and that this stimulation was associated with
alterations in the kinetics of voltage-dependent activation that favored the active state of
the channels. Grabov, et al also observed that calyculin A acted synergistically with ABA in
promoting the anion current, and that the effects in every case were independent of the
abi1 transgene.
How can these two sets of data be reconciled? Pei et al based their assessment on a
mistaken comparison of the Arabidopsis anion current in the presence of ABA with the
N. benthamiana current recorded in the absence of ABA. In fact, resting activities of
the anion channels were roughly equivalent, despite the obvious difference of
species and very different methods used to record these currents. It is likely that subtle
differences in kinase/phosphatase cascades mean that Arabidopsis and N. benthamiana
respond differently to protein phosphatase antagonists (compare also ref. 160), and
this factor might well explain the discrepancy in action of the abi1 gene product. Equally
important, technical differences—including exchange of the cytosol with patch electrode
filling solutions 36 —raises the possibility that different parts of the same kinase/
phosphatase cascade could be favored in each case. Regardless of these complications,
the fact that both sets of experiments demonstrate an action of ABA on the anion
current is salutary.
Interaction of Signaling Elements

Even when evoked by a single stimulus such as ABA, signaling pathways interact in a
manner that can be envisaged to spread its perturbation like a ripple in a pond through a
network of interrelated connections.162 Sufficient evidence is now to hand to demonstrate
such interactions between signaling elements of protein (de-)phosphorylation, [Ca2+]i and
pHi. At present, the precise points of interaction are minimally defined, as is the extent to
which the various signals are interdependent. These issues are currently a major focus of
our attention. Their resolution will be particularly relevant both to determining the
hierarchy of events behind stimulus-response coupling per se, and to establishing the role(s)
of each element, whether primary to signal transmission or secondary in adaptive
conditioning of the response.
pHi Interaction with [Ca2+]i

A case in point can be drawn from work on pHi and [Ca2+]i signals evoked by ABA.
From analysis of the current kinetics and voltage-dependence, it is evident that pHi and
[Ca2+]i controls of IK,in are fundamentally different (see The H+ Second Messenger above).
Nonetheless, these two signaling elements almost certainly interact. Grabov and Blatt130
observed that experimentally lowering pHi to values below about 7.0 resulted in a rise of
[Ca2+]i in roughly 50% of the cells examined. It is significant, too, that these [Ca2+]i
increases did not correlate with the timecourse of pHi loads as might be expected for simple,
bulk titration of Ca2+ binding sites. Instead it appears that pHi may “prime” signaling
elements that mediate in [Ca2+]i control and indirectly trigger a rise in [Ca2+]i 8. The
precise mechanism of this pH-induced [Ca2+]i rise is not currently known, however several
experimentally testable scenarios are now worth examining. One possible explanation
lies in the pH-sensitivity of Ca2+ release mechanisms within the cell, for example the
binding of IP3 to its receptor that leads to cytosolic Ca2+ release,163 although in animals it
is normally alkaline rather than acid pHi that favors IP3 mediated Ca2+ release.164 In the
guard cells pHi is known to affect the gating of vacuolar ion channels that may contribute
to [Ca2+]i changes either directly or indirectly.74,75 The SV channels, especially, are
sensitive to pHi 74 with acid pHi favoring vacuolar Ca2+ release. It is also possible that pHi
could affect Ca2+ influx across the plasma membrane (see Fig. 6.1). The action of pHi in
each of these instances affects events upstream of the [Ca2+]i messenger itself, although
the associated mechanisms are different.
Equally, cytosolic-free Ca2+ may influence pHi. Recent studies (Fricker M, Wood J,
personal communication) have suggested that increasing [Ca2+]i, by promoting Ca2+
influx across the plasma membrane, can trigger increases and fluctuations in pHi that
persist over many minutes. These measurements were carried out using confocal
ratio fluorescence imaging techniques and also indicate a spatial heterogeneity to the
pHi changes. Because local [Ca2+]i “hot spots” are though to underpin Ca2+-induced
Ca2+ release in many cell types,122 we may speculate that complementary pHi domains
also occur within cells, either superimposed on or associated with local changes in
[Ca2+]i. The validity of such a premise certainly deserves attention. But whether or not
correct, it is already evident that the interaction between [Ca2+]i and pHi is not simply
bilateral: whereas decreasing pH i may promote a rise in [Ca 2+] i , the converse of
increasing [Ca2+]i appears to favor pHi increases.
[Ca2+]i, pHi and Protein Phosphorylation

Elevation of [Ca2+]i is probably closely linked to protein phosphorylation in plants as
it is in animals. Principal points of convergence are protein kinases and phosphatases
that dependent on [Ca2+]i. In plant cells there is now considerable evidence for Ca2+-,
Ca2+/calmodulin- and Ca2+/phospholipid- dependent kinases and phophatases.79,151,165-168
Calcium-dependent protein kinase activity is known to activate an ABA-responsive
promoter in Arabidopsis leaf protoplasts169 and a Cl- channel in the guard cell tonoplast.77
By contrast, the ABI1 gene product has proven to be insensitive to [Ca2+]i,159 although it
contains a putative EF-hand Ca2+-binding motif and was originally suspected to be
Ca2+-regulated.157,158
Protein phosphorylation has now been shown to affect pHi signal transmission in
guard cells. Blatt and Grabov8 have reported that the abi1 transgene in N. benthamiana
drastically reduces K+ channel sensitivity to experimentally-imposed changes in pHi. They
found that lowering pHi with 3 and 10 mM butyrate, equivalent to decreases of 0.3-0.5 pHi
units, had only marginal effects on the K+ currents in the transgenic plants compared with
the wild type. The observations are significant, because previous work with these plants24
had demonstrated that the transgene affected K+ channel response to ABA but had no
influence on the rise in pHi evoked by the hormone. So, together, these two sets of data
confirm that events downstream of pHi are dependent on the phosphorylation state of
one or more target proteins.
Initial Events in ABA Stimulus Perception

By contrast with our current knowledge of downstream signaling events and their
targets that mediate solute transport in guard cells, no substantive information exists
relating to the ABA receptor itself. The debate surrounding the localization of the ABA
receptor has itself failed to resolve the most basic question—whether the receptor is cyto-
solic or situated in the plasma membrane and exposed to the external environment—and
has further fueled controversy. It is possible that different ABA receptors situated at the
cell periphery and within the cell all contribute to ABA perception.170 However resolving
this, and related issues will now require that the receptor gene(s) are cloned and
characterized.92
Localization of ABA Receptors

The results of early attempts to identify the nature of the ABA receptor suggested that
the hormone was perceived outside at the cell surface.109 These arguments centered around
the fact that ABA should partition as a weak acid across the plasma membrane in a
pH-dependent manner171-173 and therefore should be more effective at acid than at alkaline
external pH if the receptor were located inside the cell. Hornberg and Weiler174 reported
direct ABA binding to a plasma membrane proteins using photo affinity methods, but the
failure of other groups to duplicate these results thereafter was a setback. Only recently
have more conventional biochemical approaches yielded indications of specific ABA-binding
proteins.175
Other studies exploring the subcellular localization of ABA perception site(s) in guard
cells have used advanced in vivo techniques. Schwartz, et al176 reported a close correlation
between 3H-ABA uptake and stomatal closure in Commelina guard cells. They also found
that microinjecting ABA into individual guard cells promoted stomatal closure while
inclusion of ABA in patch electrodes suppressed IK,in of Vicia guard cell protoplasts. Parallel
studies,102 using a biologically-inert “caged”-ABA, also implicated an internal perception
site for ABA. In this case, microinjection and UV photolysis of the caged compound to
release active ABA promoted stomatal closure even when the epidermal strip was continuously
perfused with ABA-free solution. Other work has indicated that ABA may be perceived at
the cell surface. Anderson, et al177 reported that intracellular microinjections of ABA alone
was insufficient to prevent stomatal opening in Commelina, although stomatal opening
was suppressed when ABA was provided externally. (Note that in a different tissue
preparation, in this case from barley aleurone, Gilroy and Jones178 found that -amylase
secretion that is normally stimulated by gibberelin and antagonized by ABA could not be
suppressed when ABA was first microinjected into the cells.)
The more trivial explanations of differences in plant material aside, the apparent
contradiction between these two data sets might be explained by differences in experimental
design. Evidence of an internal site for ABA action was supported by experiments that
centered on stomatal closure, whereas arguments for an external site of action were
indicated in experiments designed to show the ability of ABA to suppress other activities,
either stomatal opening or amylase secretion. Promotion of stomatal closure and inhibition of
stomatal opening involve partially-separable mechanisms.179 So, it is possible that more
than one site of perception exists, even for short-term events that (presumably) do not
require transcription/translation activities. In fact, this latter idea has found support in
recent radiotracer flux studies. MacRobbie70 observed differences in the efficacy and tim-
ing for ABA flux response in Commelina guard cells. She found that under suboptimal
conditions a delay in vacuolar efflux could be introduced into 86Rb+(K+) efflux measurements
that were separable from the initial flux transient across the plasma membrane. It is
possible that intervening signaling steps—and differences in their kinetics—could
account for such differences in response time and ABA dependence. However, at least one
explanation is that ABA binds with different affinities and at separate sites associated with
the plasma membrane and tonoplast.
Is the ABA Signal G-Protein Coupled?

In lieu of direct evidence for ABA receptor binding, downstream events immediately
following ABA-receptor interaction have been sought that couple to late signaling
intermediates such as [Ca2+]i increases and could yield some information about the
nature of the receptor protein. Widespread interest has focused on the possible role for a
serpentine (7-transmembrane-segment) receptor protein that could couple through the
activation of GTP-binding proteins (G-proteins).180-182 There is certainly some evidence
to support a role for heterotrimeric G-proteins in regulating IK,in. However, none of these
data can be tied to pHi regulation of the current, nor are they wholly consistent with
coupling even through the [Ca2+]i intermediate. In their original studies, Fairley and
Assmann131 showed that IK,in was sensitive to the G-protein antagonists cholera and
pertussis toxins in Vicia guard cell protoplasts; the current was inactivated in the presence
of the non-hydrolyzable GTP analogue GTP-γ-S and this inactivation was overcome by
buffering changes in [Ca2+]i. These studies indicated that control of I K,in could pass
from Gα activation and, plausibly, stimulation of inositol-1,4,5- trisphosphate production
by phospholipase C to a rise in [Ca2+]i. Subsequently, Wu and Assmann183 added evidence
for coupling via a membrane-delimited Gα, and Kelly, et al104 have suggested that [Ca2+]i
may exert a dual effect with G-proteins on IK,in.
G-proteins may also control IK,in activity independent of [Ca2+]i. Armstrong and
Blatt132 found that mas7, that mimics serpentine receptors and promotes G release,
inactivated I K,in. Mas7 action was blocked by GDP-β-S, consistent with Gα control of IK,in.
However, the effect was not linked to [Ca2+]i, as mas7 action could not be overcome by
cytosolic Ca2+ buffering or by neomycin sulfate, an antagonist of phospholipase C and
inositol- 1,4,5-trisphosphate-mediated Ca2+ release. Nor could evidence be found for mas7
action passing through the pHi intermediate. Unlike the effects on auxin- and ABA-evoked
responses of the K+ channels, treatments with weak acid to clamp pHi near 7.0 failed
to rescue IK,in from inactivation by mas7. Furthermore, mas7 had no influence on IK,out
which is profoundly pHi sensitive (see The H+ second messenger above). In fact, there
remains an overwhelming lack of evidence to support a function for G-proteins in
hormonal responses of guard cells per se, whether coupled through [Ca2+]i or pHi.
Thus, it appears ever more probable that relevant signaling events upstream in these
cells could differ from the classic G-protein-receptor model.
Perspectives and Conclusion

The complexity of signaling events that underlie stomatal response to ABA should
come as little surprise as guard cells integrate both environmental and internal signals
to achieve stomatal control. As the primary defense of the plant against water loss and
dehydration, it is imperative that transport functions in guard cells are finely tuned to
these needs along with those of competing demands for CO2. However, our awareness of
the “stomatal situation” arises from a physiological perspective and, most recently, from
the advantages afforded by the interface between molecular genetics and biophysics. It is
likely that many of these control networks, like the transporters that they regulate, are
common among higher-plant cells. Indeed, of the pumps and ion channels that effect
solute movement for stomatal control—and these are, with several notable exceptions,
now well documented at the plasma membrane—counterparts are to be found in a
range of higher-plant cells in every case. Admittedly, our understanding of transport
across the tonoplast remains very poor and much more work is needed here to identify the
major pathways for solute movements and their respective functions. Nonetheless, at present
there is little basis to set vacuolar transport in guard cells apart from other higher-plant
cell types.
Our knowledge of the mechanisms controlling ion channels and, even more still, the
H+-ATPase at the plasma membrane signal integration is unquestionably in the early stages
of development. Still less is known of transport regulation at the tonoplast, although the
bulk of osmotic solutes must pass across this latter membrane. The vacuole is also likely to
comprise an important reservoir and sink for Ca2+ and H+, so separating transport
and signaling functions will be especially important. A role for [Ca2+]i in association
with ABA is certain but, remarkably, major questions still hang over its situation within
the ABA-signaling network, its origin(s), downstream targets and mechanism(s) of

action. The emergence of pHi as a second messenger, and its interaction with the abi1
protein phosphatase, further underscores our relative ignorance about the scope and
nature of events both downstream and upstream within each respective pathway. Significantly,
here is a second messenger for which we may find no direct counterpart in animal cells
and no preconceived model to frame our ideas.
Unquestionably, major advances in resolving each of these issues will now be aided by
a combination of molecular genetic and biophysical tools. Once the identity of a first ABA
receptor has been established, relationships between many signaling elements should fall
in place while still others may only then become evident. But almost certainly some
current perspective on the issues and their possible solutions will appear naive by the
close of this decade. We may echo JBS Haldane’s suspicion “that the universe is not only
queerer than we suppose, but queerer than we can suppose.”184
Acknowledgments
We are grateful to Mark Fricker (Oxford) for sharing unpublished results with us.
This work was supported by the Gatsby Charitable Foundation, Human Frontiers Science
Program grant RG303/95M and EC Biotech grant CT960062. AG is a Senior Research
Associate funded on BBSRC grant 32/C098-1. BL was a Sainsbury Ph.D. Student and is
currently funded on BBSRC grant 32/C08406.
References
1. Loveys BR. How useful is a knowledge of ABA physiology for crop improvement? In:
Davies WJ, Jones HG, eds. Abscisic Acid Physiology and Biochemistry. Oxford, Bios
Scientific. 1991:245-260.
2. Flowers T, Yeo M. Salt Tolerant Plants. In: Baker DA, Hall JL, eds. Solute Transport in
Plant Cells and Tissues. Harlow, Longman Press. 1988:326-355.
3. Bradford KJ, Hsiao TC. Stomatal behavior and water relations of waterlogged tomato
plants. Plant Physiology 1982; 70:1508-1513.
4. Jackson MB. Regulation of water relationships in flooded plants by ABA from leaves,
roots and xylem sap. In: Davies WJ, Jones HG, eds. Abscisic Acid Physiology and
Biochemistry. Oxford, Bios Scientific. 1991:217-226
5. Davies WJ, Jones HG. Abscisic Acid: Physiology and Biochemistry. Bios Scientific, Oxford.
1991:1-266.
6. Nobel PS. Physicochemical and Environmental Plant Physiology. Academic Press, San
Diego. 1991:1-413.
7. Willmer C, Fricker MD. Stomata. Chapman and Long, London. 1996:1-375.
8. Blatt MR, Grabov, A. Signaling gates in abscisic acid-mediated control of guard cell ion
channels. Physiologia Plantarum 1997; 100:481-490.
9. Malpighi M. Anatome Plantarum. Royal Society, London.1675:1-130.
10. Grew N. Anatomy of Plants. London. 1682:1-312.
11. de Candolle AP. Organographie Végétale ou description raisone des organes des plantes.
Deterville, Paris. 1827:1-213.
12. von Mohl H. Welche Ursachen bewirken die Erweiterung und Verengung der
Spaltoeffnungen? Bot Zeit. 1856; 14:697-720.
13. Immamura S. Untersuchungen uber den mechanismus der turgorschwankung der
Spaltoeffnungschlieszellen. Jap J Bot. 1943; 12:251-346.
14. Yamashita T. Influences of potassium supply upon various properties and movement of
guard cell. Sieboldia Acta Biol. 1952; 1:51-70.
15. Fujino M. Role of ATP and ATPase in stomatal movement. Sci Bull Fac Ed Nagasaki
Univ. 1967; 18:1-47.
16. Fischer RA. Stomatal opening: Role of potassium uptake by guard cells. Science 1968;
160:784-785.
17. Fischer R, Hsiao T. Stomatal opening in isolated epidermal strips of Vicia faba: II.
Response to KCl concentration and role of potassium absorption. Plant Physiol. 1968;
43:1953-1958.
18. Raschke K. The stomatal turgor mechanism and its response to CO2 and abscisic acid:
observations and hypothesis. In: Marre E, Cifferrri O, eds. Regulation of cell membrane
activities in plants. Amsterdam, Elsevier. 1977:173-183.
19. Raschke K. Movements of stomata. In: Haupt W, Feinleib ME, eds. Encyclopedia of Plant
Physiology, N.S. Vol. 7. Berlin, Springer. 1979:373-441.
20. Brindley HM. Fluxes of 86Rb+ in “isolated” guard cells of Vicia faba L. Planta 1990;
181:432-439.
21. Blatt MR. Electrical characteristics of stomatal guard cells: The ionic basis of the mem-
brane potential and the consequence of potassium chloride leakage from microelectrodes.
Planta 1987; 170:272-287.
22. Wille A, Lucas W. Ultrastructural and histochemical studies on guard cells. Planta 1984;
160:129-142.
23. Grabov A, Leung J, Giraudat J, Blatt MR. Alteration of anion channel kinetics in wild
type and abi1-1 transgenic Nicotiana benthamiana guard cells by abscisic acid. Plant Journal
1997; 12:203-213.
24. Armstrong F, Leung J, Grabov A, Brearley J, Giraudat J, Blatt MR. Sensitivity to abscisic
acid of guard cell K+ channels is suppressed by abi1-1, a mutant Arabidopsis gene encoding a
putative protein phosphatase. Proc Natl Acad Sci USA. 1995; 92:9520-9524.
25. Pei ZM, Kuchitsu K, Ward JM, Schwarz M, Schroeder JI. Differential abscisic acid
regulation of guard cell slow anion channels in Arabidopsis wild type and abi1 and abi2
mutants. Plant Cell. 1997; 9:409-423.
26. Roelfsema MRG, Prins HBA. Ion channels in guard cells of Arabidopsis thaliana (L.)
Heynh. Planta. 1997; 202:18-27.
27. Raschke K, Firn RD, Pierce M. Stomatal closure in response to xanthoxin and abscisic
acid. Planta. 1975; 125:149-160.
28. Hentzen AE, Smart LB, Wimmers LE, Fang HH, Schroeder JI, Bennett AB. 2 plasma-
membrane H+-ATPase genes expressed in guard-cells of Vicia faba are also expressed
throughout the plant. Plant And Cell Physiology 1996; 37:650-659.
29. Becker D, Zeilinger C, Lohse G, Depta H, Hedrich R. Identification and biochemical
characterization of the plasma membrane H + -ATPase in guard cells of Vicia faba L.
Planta. 1993; 190:44-50.
30. Villalba JM, Lutzelschwab M, Serrano R. Immunocytolocalization of plasma membrane
H+-ATPase in maize coleoptiles and enclosed leaves. Planta. 1991; 185:458-461.
31. Thiel G, MacRobbie EAC, Blatt MR. Membrane transport in stomatal guard cells: The
importance of voltage control. J Membr Biol. 1992; 126:1-18.
32. Blatt MR, Clint GM. Mechanisms of fusicoccin action kinetic modification and inactivation
of potassium channels in guard cells. Planta. 1989; 178:509-23.
33. Blatt MR. Electrical characteristics of stomatal guard cells: The contribution of ATP-
dependent, “electrogenic” transport revealed by current-voltage and difference-cur-
rent-voltage analysis. J Membr Biol. 1987; 98:257-274.
34. Serrano EE, Zeiger E, Hagiwara S. Red light stimulates an electrogenic proton pump in
Vicia guard cell protoplasts. Proc Natl Acad Sci USA 1988; 85:436-440.
35. Assmann SM, Simoncini L, Schroeder JI. Blue light activates electrogenic ion pumping
in guard cell protoplasts of Vicia faba. Nature 1985; 318:285-287.
36. Pusch M, Neher E. Rates of diffusional exchange between small cells and a measuring
patch pipette. Pfluegers Archiv, Euro J Physiol 1988; 411:204-214.
37. Schroeder, J.I. 1988. K+ transport properties of K+ channels in the plasma membrane of
Vicia faba guard cells. J Gen Physiol. 1988; 92:667-683.
38. Lohse G, Hedrich R. Characterization of the plasma-membrane H+- ATPase from Vicia
faba guard cells. Planta. 1992; 188:206-214.
39. Slayman CL, Bertl A, Blatt MR. Partial reaction chemistry and charge displacement by
the fungal plasma-membrane H +-ATPase. In: Hirst BH, ed. Molecular and Cellular
mechanisms of H+ Transport. Berlin, Springer Verlag. 1994:237-244.
40. Blatt MR, Beilby MJ, Tester M. Voltage dependence of the Chara proton pump revealed
by current-voltage measurement during rapid metabolic blockade with cyanide. J Membr
Biol. 1990; 114:205-223.
41. Thiel G, Wolf AH. Operation of K+ channels in stomatal movement. Trends In Plant
Science 1997; 2:339-345.
42. Clint GM, Blatt MR. Mechanisms of fusicoccin action: evidence for concerted modulations
of secondary K+ transport in a higher-plant cell. Planta. 1989; 178:495-508.
43. Rodriguez-Navarro A, Blatt MR, Slayman CL. A potassium-proton symport in Neurospora
crassa. J Gen Physiol. 1986; 87:649-674.
44. Maathuis FJM, Sanders D. Mechanisms of potassium absorption by higher-plant roots.
Physiologia Plantarum 1996; 96:158-168.
45. Nakamura RL, Mckendree WL, Hirsch RE, Sedbrook JC, Gaber RF, Sussman MR.
Expression of an Arabidopsis potassium channel gene in guard cells. Plant Physiology
1995; 109:371-374.
46. Very A-A, Gaymard F, Bosseux C, Sentenac H, Thibaud J-P. Expression of a cloned plant
K+ channel in Xenopus oocytes: analysis of macroscopic currents. Plant J. 1995; 7:321-332.
47. Schroeder JI, Schachtman DP, Lucas WJ, Anderson JA, Gaber RF. The Arabidopsis KAT1
cDNA confers functional expression of an inward-rectifying potassium channel in
Xenopus oocytes. Biophysical Journal 1993; 64:196-196.
48. Bertl A, Anderson JA, Slayman CL, Gaber RF. Use of Saccharomyces-cerevisiae for
patch-clamp analysis of heterologous membrane proteins—characterization of KAT1, an
inward-rectifying K+ channel from Arabidopsis thaliana, and comparison with endogeneous
yeast channels and carriers. Proc Natl Acad Sci USA. 1995; 92:2701-2705.
49. Blatt MR, Thiel G. K+ channels of stomatal guard cells: Bimodal control of the K+
inward-rectifier evoked by auxin. Plant J. 1994; 5:55-68.
50. Schroeder JI, Fang HH. Inward-rectifying K+ channels in guard cells provide a mechanism
for low-affinity K+ uptake. Proc Natl Acad Sci USA 1991; 88:11583-11587.
51. Blatt MR. K+ channels of stomatal guard cells: characteristics of the inward rectifier and
its control by pH. J Gen Physiol. 1992; 99:615-644.
52. Obermeyer G, Armstrong F, Blatt MR. Selective block by α-dendrotoxin of the K +
inward rectifier at the Vicia guard cell plasma membrane. J Membr Biol. 1994;
137:249-259.
53. Blatt MR, Gradmann D. K+-sensitive gating of the K+ outward rectifier in Vicia guard
cells. J Membr Biol 1997; 158:241-256.
54. Hosoi S, Iino M, Shimazaki K. Outward-rectifying K+ channels in stomatal guard cell
protoplasts. Plant Cell Physiol. 1988; 29:907-911.
55. Schroeder JI, Raschke K, Neher E. Voltage dependence of K+ channels in guard-cell
protoplasts. Proc Natl Acad Sci USA 1987; 84:4108-4112.
56. Schroeder JI. Quantitative analysis of outward rectifying K+ channel currents in guard
cell protoplasts from Vicia faba. J Membr Biol. 1989; 107:229-235.
57. Blatt MR. Potassium-dependent bipolar gating of potassium channels in guard cells. J
Membr Biol. 1988; 102:235-246.
58. Blatt MR, Armstrong F. K+ channels of stomatal guard cells: abscisic acid-evoked control
of the outward rectifier mediated by cytoplasmic pH. Planta. 1993; 191:330-341.
59. Allen GJ, Sanders D. Vacuolar ion channels of higher plants. Advances In Botanical
Research Incorporating Advances In Plant Pathology 1997; 25:217-252.
60. Hedrich R, Busch H, Raschke K. Ca2+ and nucleotide-dependent regulation of voltage
dependent anion channels in the plasma membrane of guard cells. EMBO J. 1990;
9:3889-3892.
61. Schulz-Lessdorf B, Lohse G, Hedrich R. GCAC1 recognizes the pH gradient across the
plasma membrane: A pH-sensitive and ATP-dependent anion channel links guard cell
membrane potential to acid and energy metabolism. Plant Journal 1996; 10:993-1004.
62. Linder B, Raschke K. A slow anion channel in guard cells, activating at large hyper-
polarization, may be principal for stomatal closing. FEBS Lett. 1992; 313:27-30.
63. Schroeder JI, Keller BU. Two types of anion channel currents in guard cells with distinct
voltage regulation. Proc Natl Acad Sci USA 1992; 89:5025-5029.
64. Dietrich P, Hedrich R. Interconversion of fast and slow gating modes of GCAC1, a guard
cell anion channel. Planta 1994; 195:301-304.
65. Thomine S, Zimmermann S, Guern J, Barbier-Brygoo H. ATP-dependent regulation of
an anion channel at the plasma membrane of protoplasts from epidermal cells of
Arabidopsis hypocotyls. Plant Cell. 1995; 7:2091-2100.
66. Artalejo CR, Ariano MA, Periman RL, Fox AP. Activation of facilitation calcium channels in
chromaffin cells by D1 dopamine receptors through a cAMP/protein kinase A- dependent
mechanism. Nature. 1990; 348:239-242.
67. Song Y, Huang L-YM. Modulation of glycine receptor chloride channels by cAMP-
dependent protein kinase in spinal trigeminal neurons. Nature 1990; 348:242-245.
68. Blatt MR, Thiel G. Hormonal control of ion channel gating. Ann Rev Plant Physiol Mol
Biol. 1993; 44:543-567.
69. Brearley J, Venis MA, Blatt MR. CO2-mediated control of K+ and anion channels in guard
cells of Vicia. Planta. 1997; 203:145-154.
70. MacRobbie EAC. Effects of ABA on 86 Rb + fluxes at plasmalemma and tonoplast of
stomatal guard cells. Plant Journal. 1995; 7:835-843.
71. MacRobbie EAC. ABA-induced ion efflux in stomatal guard-cells—multiple actions of
aba inside and outside the cell. Plant Journal. 1995; 7:565-576.
72. Fricker MD, Willmer CM. Nitrate-sensitive ATPase activity and proton pumping in guad
cell protoplasts of Commelina. J Exp Botany 1990; 41:193-198.
73. Willmer CM, Grammatikopoulos G, Lasceve , Vavasseur A. Characterization of the
vacuolar-type H+-ATPase from guard cell protoplasts of Commelina. J Exp Botany 1995;
46:383-389.
74. Schulz-Lessdorf B, Hedrich R. Protons and calcium modulate SV-type channels in the
vacuolar lysosomal compartment—channel interaction with calmodulin inhibitors. Planta.
1995; 197:655-671.
75. Ward JM, Schroeder JI. Calcium-activated K+ channels and calcium- induced calcium
release by slow vacuolar ion channels in guard-cell vacuoles implicated in the control of
stomatal closure. Plant Cell. 1994; 6:669-683.
76. Tikhonova LI, Pottosin II, Dietz KJ, Schonknecht G. Fast-activating cation channel in
barley mesophyll vacuoles. Inhibition by calcium. Plant Journal. 1997; 11:1059-1070.
77. Pei ZM, Ward JM, Harper JF, Schroeder JI. A novel chloride channel in Vicia faba guard
cell vacuoles activated by the serine/threonine kinase, CDPK. Embo Journal 1996;
15:6564-6574.
78. Frohnmeyer H, Grabov A, Blatt MR. A role for the vacuole in auxin-mediated control of
cytoplasmic pH by Vicia mesophyll and guard cell protoplasts. Plant J. 1998; 13:1098-
116.
79. Allen GJ, Sanders D. Calcineurin, a type 2B protein phosphatase, modulates the Ca2+-
permeable slow vacuolar ion channel of stomatal guard cells. Plant Cell 1995; 7:1473-1483.
80. Little CHA, Eidt DC Effect of abscisic acid on bud break and transpiration in woody
species. Nature. 1968; 220:498-499.
81. Mittelheuser CJ, van Steveninck RFM. Stomatal closure and inhibition of transpiration
induced by (RS)-abscisic acid. Nature. 1969; 221:281-282.
82. Wright STC. An increase in the “inhibitor-B” content of detached wheat leaves following
a period of wilting. Planta. 1969; 86:10-20.
83. Wright STC, Hiron, RWP. (+)-Abscisic acid, the growth inhibitor induced in wheat leaves
by a period of wilting. Nature. 1969; 224:719-720.
84. Harris MJ, Outlaw WHJ, Mertens R, Weiler E. Water-stress-induced changes in the
abscisic acid content of guard cells and other cells of Vicia faba L. leaves as determined
by enzyme-amplified immunoassay. Proc Natl Acad Sci USA 1988; 85:2584-2588.
85. Cheeseman JM. Patchy: Simulating and visualizing the effects of stomatal patchiness on
photosynthetic CO2 exchange studies. Plant Cell Environ. 1991; 14:593-599.
86. Daley PF, Raschke K, Ball JT, Berry JA. Topography of photosynthetic activity of leaves
obtained from video images of chlorophyll flourescence. Plant Physiol. 1989; 90:1233-1238.
87. Atkinson CJ, Mansfield TA, Kean AM, Davies WJ. Control of stomatal aperture by
calcium in isolated epidermal tissue and whole leaves of Commelina communis L. New
Phytol 1989; 111:9-17.
88. Snaith PJ, Mansfield TA. Control of the CO2 response of stomata by indol-3ylacetic acid
and abscisic acid. J Exptl Bot. 1982; 33:360-365.
89. Gorton HL, Williams WE, Assmann SM. Circadian rhythms in stomatal responsiveness
to red and blue light. Plant Physiol. 1993; 103:399-406.
90. Peng ZY, Weyers JDB. Stomatal sensitivity to abscisic acid following water-deficit stress.
J Exp Botany 1994; 45:835-845.
91. Harris MJ, Outlaw WH, Jr. Rapid adjustment of guard-cell abscisic acid levels to current
leaf-water status. Plant Physiol. 1991; 95:171-173.
92. Leyman B, Blatt MR. Towards cloning of an abscisic acid receptor. Institute Juan March
CRISB 1997; 60:74.
93. Blatt MR. Potassium channel currents in intact stomatal guard cells: Rapid enhancement
by abscisic acid. Planta. 1990; 180:445-455.
94. Goh CH, Oku T, Shimazaki K. Properties of proton-pumping in response to blue-light
and fusicoccin in guard-cell protoplasts isolated from adaxial epidermis of Vicia leaves.
Plant Physiology 1995; 109:187-194.
95. Schroeder JI, Hagiwara S. Cytosolic calcium regulates ion channels in the plasma
membrane of Vicia faba guard cells. Nature. 1989; 338:427-430.
96. Lemtiri-Chlieh F, MacRobbie EAC. Role of calcium in the modulation of Vicia guard
cell potassium channels by abscisic acid: a patch-clamp study. J Membr Biol. 1994;
137:99-107.
97. Irving HR, Gehring CA, Parish RW. Changes in cytosolic pH and calcium of guard cells
precede stomatal movements. Proc Natl Acad Sci USA 1992; 89:1790-1794.
98. McAinsh MR, Brownlee C, Hetherington AM. Visualizing changes in cytosolic-free Ca2+
during the response of stomatal guard cells to abscisic acid. Plant Cell. 1992; 4:1113-1122.
99. Fricker MD, Gilroy S, Read ND, Trewavas AJ. Visualisation and measurement of the
calcium message in guard cells. In: Schuch W, Jenkins G, eds. Molecular Biology of Plant
development. Cambridge, Cambridge Univ. Press. 1991:177-190.
100. Gilroy S, Fricker MD, Read ND, Trewavas AJ. Role of calcium in signal transduction of
Commelina guard cells. Plant Cell. 1991; 3:333-344.
101. McAinsh MR, Brownlee C, Hetherington AM. Abscisic acid-induced elevation of guard
cell cytosolic Ca2+ precedes stomatal closure. Nature 1990; 343:186-188.
102. Allan AC, Fricker MD, Ward JL, Beale MH, Trewavas AJ. Two transduction pathways
mediate rapid effects of abscisic acid in Commelina guard cells. Plant Cell. 1994;
6:1319-1328.
103. Grabov, A. and Blatt, M.R. Ca2+-induced Ca2+ release triggered by membrane voltage
and abscisic acid in guard cells of Vicia faba. Proc Natl Acad Sci USA. 1998; 95:4778-4783.
104. Kelly WB, Esser JE, Schroeder JI. Effects of cytosolic calcium and limited, possible dual,
effects of G-protein modulators on guard cell inward potassium channels. Plant Journal
1995; 8:479-489.
105. Cobbold PH, Rink TJ. Flourescence and bioluminescence measurement of cytoplasmic
free calcium. Biochem J. 1987; 248:313-328.
106. Fabiato A, Fabiato F. Calculator programs for computing the composition of solutions
containing multiple metals and ligands used for experiments in skinned muscle cells. J
Physiol. 1979; 75:463-505.
107. Pethig R, Kuhn M, Payne R, Adler E, Chen T-H, Jaffe LF. On the dissociation constants
of BAPTA-type calcium buffers. Cell Calcium. 1989; 10:491-498.
108. Foehr G, Warchol W, Gratzl G. Calculation and control of free divalent cations in
solutions used for membrane fusion studies. Methods Enzymol. 1993; 221:149-157.
109. MacRobbie EAC. Calcium-dependent and calcium-independent events in the initiation

of stomatal closure by abscisic acid. Proc Roy Soc Lond B Biol Sci. 1990; 241:214-219.
110. McAinsh MR, Brownlee C, Hetherington AM. Partial inhibition of ABA-induced stomatal
closure by calcium—channel blockers. Proc R Soc Lond Ser B Biol Sci. 1991; 243:195-202.
111. Blatt MR, Thiel G, Trentham DR. Reversible inactivation of K+ channels of Vicia sto-
matal guard cells following the photolysis of caged inositol 1,4,5- trisphosphate.
Nature. 1990; 346:766-769.
112. Gilroy S, Read ND, Trewavas AJ. levation of cytoplasmic calcium by caged calcium or
caged inositol trisphosphate initiates stomatal closure. Nature. 1990; 346:769-771.
113. Lee YS, Choi YB, Suh S, Lee J, Assmann SM, Joe CO, Kelleher JF, Crain RC. Abscisic
acid-induced phosphoinositide turnover in guard-cell protoplasts of Vicia faba. Plant
Physiology. 1992; 110:987-996.
114. Parmar PN, Brearley CA. Metabolism of 3-phosphorylated and 4-phosphorylated
phosphatidylinositols in stomatal guard cells of Commelina communis l. Plant Journal.
1995; 8:425-433.
115. Parmar PN, Brearley CA. Identification of 3-phosphorylated and 4-phosphorylated
phosphoinositides and inositol phosphates in stomatal guard cells. Plant Journal. 1993;
4:255-263.
116. Muir SR, Sanders D. Pharmacology of Ca2+ release from red beet microsomes suggests
the presence of ryanodine receptor homologs in higher-plants. Febs Letters. 1996;
395:39-42.
117. Allen GJ, Muir SR, Sanders D. Release of Ca2+ from individual plant vacuoles by both
insp(3) and cyclic ADP-ribose. Science 1995; 268:735-737.
118. Clapham DE. Calcium signaling. Cell. 1995; 80:259-268.
119. Webb AAR, McAinsh MR, Mansfield TA, Hetherington AM. Carbon dioxide induces
increases in guard cell cytosolic free calcium. Plant Journal. 1996; 9:297-304.
120. McAinsh MR, Webb AAR, Taylor JE, Hetherington AM. Stimulus- induced oscillations
in guard cell cytosolic-free calcium. Plant Cell. 1995; 7:1207-1219.
121. Meyer T, Stryer L. Calcium spiking. Ann Rev Biophys Biophys Chem. 1991;20:153-174.
122. Berridge MJ. Microdomains and elemental events in calcium signaling. Cell Calcium. 1996;
20:95-96.
123. Dolmetsch RE, Lewis RS, Goodnow CC, Healy JI. Differential activation of transcription
factors induced by Ca2+ response amplitude and duration. Nature. 1997; 386:855-858.
124. Gradmann D, Blatt MR, Thiel G. Electrocoupling of ion transporters in plants. J Membrane
Biol 1993; 136:327-332.
125. Bootman MD, Berridge MJ. Subcellular Ca2+ signals underlying waves and graded
responses in hela cells. Curr Biol. 1996; 6:855-865.
126. Chavis P, Fagni L, Lansman JB, Bockaert J. Functional coupling between ryanodine
receptors and L-type calcium channels in neurons. Nature. 1996; 382:719-722.
127. Schweitz H, Heurteaux C, Bois P, Moinier D, Romey G, Lazdunski M. Calcicludine, a
venom peptide of the Kunitz-type protease inhibitor family, is a potent blocker of high-
threshold Ca2+ channels with a high-affinity for L-type channels in cerebellar granule
neurons. Proc Natl Acad Sci USA. 1994; 91:878-882.
128. Schroeder JI, Hagiwara S. Repetitive increases in cytosolic calcium of guard cells by
abscisic acid: Activation of nonselective calcium permeable channels. Proc Natl Acad Sci
USA. 1990; 87:9305-9309.
129. Tester M. Plant ion channels: Whole-cell and single-channel studies. New Phytol. 1990;
114:305-340.
130. Grabov A, Blatt MR. Parallel control of the inward-rectifier K+ channel by cytosolic-free
Ca2+ and pH in Vicia guard cells. Planta. 1997; 201:84-95.
131. Fairley GK, Assmann SM. Evidence for G-protein regulation of inward potassium ion
channel current in guard cells of fava bean. Plant Cell. 1991; 3:1037-1044.
132. Armstrong F, Blatt MR. Evidence for K+ channel control in Vicia guard cells coupled by
G-proteins to a 7TMS receptor. Plant J. 1995; 8:187-198.
133. Guern J, Felle H, Mathieu Y, Kurkdjian A. Regulation of intracellular pH in plant-cells.

International Review of Cytology. 1991; 127:111-173.
134. Felle H. pH as a second messenger in plants. In: Boss WF, Morre DJ, eds. Second
messengers in plant growth and development. New York, Alan R. Liss, Inc. 1989:145-166.
135. Morison JIL. Sensitivity of stomata and water use efficiency to high CO2. Plant Cell
Environ. 1985; 8:467-474.
136. Senn AP, Goldsmith MHM. Regulation of electrogenic proton pumping by auxin and
fusicoccin as related to the growth of Avena coleoptiles. Plant Physiol. 1988; 88:131-138.
137. Brummer B, Bertl A, Potrykus I, Felle H, Parish RW. Evidence that fusicoccin and
indole-3-acetic acid induce cytosolic acidifcation of Zea mays cells. FEBS Lett. 1985;
189:109-114.
138. Hager A, Moser I. Acetic acid esters and permeable weak acids induce active proton
extrusion and extension growth of coleoptile segments by lowering the cytoplasmic pH.
Planta. 1985; 163:391-400.
139. Marre E. Fusicoccin: A tool in plant physiology. Ann Rev Plant Physiol Mol Biol. 1979;
30:273-288.
140. Gehring CA, Irving HR, Parish RW. Effects of auxin and abscisic acid on cytosolic
calcium and pH in plant cells. Proc Natl Acad Sci USA 1990; 87:9645-9649.
141. Kasamo K. Effect of abscisic acid on membrane-bound epidermal ATPase from tobacco
leaves. Plant Cell Physiol. 1979; 20:293-300.
142. Lurie S, Hendrix D. Differential ion stimulation of plasmalemma adenosine triphosphatase
from leaf epidermis and mesophyll of Nicotiana rustica L. Plant Physiol. 1979; 63:936-939.
143. Thiel G, Blatt MR, Fricker MD, White IR, Millner PA. Modulation of K+ channels in
Vicia stomatal guard cells by peptide homologs to the auxin-binding protein C-terminus.
144. Miedema H, Assmann SM. A membrane-delimited effect of internal pH on the K +
outward rectifier of Vicia faba guard cells. J Membr Biol 1996; 154:227-237.
145. Lesage F, Guillemare E, Fink M, Duprat F, Lazdunski M, Romey G, Barhanin J. A
pH-sensitive yeast outward rectifier K+ channel with 2 pore domains and novel gating
properties. J Biol Chem 1996; 271:4183-4187.
146. Coulter KL, Perier F, Radeke CM, Vandenberg CA. Identification and molecular localization
of a pH-sensing domain for the inward rectifier potassium channel HIR. Neuron 1995;
15:1157-1168.
147. Bown AW. CO2 and intracellular pH. Plant Cell Environ. 1985; 8:459-465.
148. MacRobbie EAC. Effect of ABA on ion transport and stomatal regulation. In: Davies WJ,
Jones HG, eds. Abscisic Acid Physiology and Biochemistry. Oxford: Bios Scientific.
1991:153-168.
149. MacRobbie EAC. Ionic relations of guard cells. In: Zeiger E, Farquhar GD, Cowan
IR, eds. Stomatal Function. Stanford: Stanford University Press. 1987:125-162.
150. Penny MG, Bowing DJF. Direct determination of pH in the stomatal complex of
Commelina. Planta. 1975; 122:209-212.
151. Luan S, Li W, Rusnak F, Assmann SM, Schreiber SL. Immunosuppressants implicate
protein phosphatase regulation of K+ channels in guard cells. Proc Natl Acad Sci USA
1993; 90:2202-2206.
152. Thiel G, Blatt MR. Phosphate antagonist okadaic acid inhibits steady-state K+ currents
in guard cells of Vicia faba. Plant J. 1994; 5:727-733.
153. Li WW, Luan S, Schreiber SL, Assmann SM. Evidence for protein phosphatase 1 and
phosphatase 2A regulation of K+ channels in 2 types of leaf cells. Plant Physiology. 1994;
106:963-970.
154. Mori IC, Muto S. Abscisic acid activates a 48-kilodalton protein kinase in guard cell
protoplasts. Plant Physiology. 1997; 113:833-839.
155. Li JX, Assmann SM. An abscisic acid-activated and calcium-independent protein kinase
from guard cells of fava bean. Plant Cell. 1996; 8:2359-2368.
156. Koornneef M, Reuling G, Karssen CM. The isolation and characterization of abscisic acid-
insensitive mutants of Arabidopsis thaliana. Physiol.Plant. 1984; 61:377-383.
inArabidopsis thaliana. Science. 1994; 264:1452-1455.
158. Leung J, Bouvier-Durand M, Morris P-C, Guerrier D, Chefdor F, Giraudat J. Arabidopsis
ABA response gene ABI1: Features of a calcium-modulated protein phosphatase. Science.
1994; 264:1448-1452.
159. Bertauche N, Leung J, Giraudat J. Protein phosphatase activiy of the ABI1 (Abscisic acid-
Insensitive 1) protein fromArabidopsis thaliana. European Journal Of Biochemistry. 1996;
241:193-200.
160. Schmidt C, Schelle I, Liao YJ, chroeder JI. Strong regulation of slow anion channels and
abscisic acid signaling in guard cells by phosphorylation and dephosphorylation events.
Proc Natl Acad Sci USA. 1995; 92:9535-9539.
161. Leung J, Merlot S, Giraudat J. The Arabidopsis abscisic acid-insensitive2 (abi2) and abi1
genes encode homologous protein phosphatases 2C involved in abscisic acid signal
transduction. Plant Cell. 1997; 9:759-771.
162. Trewavas AJ. Growth substances in context: a decade of sensitivity. Biochem Soc Trans.
1992; 20:102-108.
163. Busa WB. Mechanisms and consequences of pH-mediated cell regulation. Annual Review
of Physiology. 1986; 48:389-402.
164. Taylor CW, Richardson A. Structure and function of inositol trisphosphate receptors.
Pharmac Ther. 1991; 51:97-137.
165. Stone JM, Walker JC. Plant protein kinase families and signal transduction. Plant Physiology.
1995; 108:451-457.
166. Ma H. Protein phosphorylation in plants—enzymes, substrates and regulators. Trends In
Genetics. 1993; 9:228-230.
167. Roberts DM, Harmon AC. Calcium-modulated proteins—targets of intracellular calcium
signals in higher-plants. Annual Review Of Plant Physiology And Plant Molecular Biology.
1992; 43:375-414.
168. Urao T, Katagiri T, Mizoguchi T, Yamaguchishinozaki K, Hayashida N, Shinozaki K. 2
genes that encode Ca2+-dependent protein-kinases are induced by drought and high-salt
stresses in Arabidopsis thaliana. Molecular & General Genetics. 1994; 244:331-340.
169. Sheen J. Ca2+-dependent protein kinases and stress signal transduction in plants. Science.
1997; 274:1900-1902.
170. MacRobbie EAC. Signaling in guard cells and regulation of ion channel activity. Journal
Of Experimental Botany. 1997; 48:515-528.
171. Hartung W. The site of action of abscisic acid at the guard cell plasmalemma ov
Valerianella locusta. Plant Cell Environ. 1983; 6:427-428.
172. Baier M, Hartung W. Movement of abscisic acid across the plasmalemma and the
tonoplast of guard cells of Valerianella locusta. Bot Acta. 1988; 101:332-337.
173. Paterson NW, Weyers JDB, A'Brook R. The effect of pH on stomatal sensitivity to
abscisic acid. Plant Cell And Environment. 1988; 11:83-89.
174. Hornberg C, Weiler EW. High-affinity binding sites for abscisic acid on the plasmalemma of
Vicia faba guard cells. Nature. 1984; 310:321-325.
175. Wan YS, Hasenstein KH. Purification and identification of ABA-binding proteins and
antibody preparation. Journal Of Molecular Recognition. 1996; 9:722-727.
176. Schwartz A, Wu WH, Tucker EB, Assmann SM. Inhibition of inward K+ channels and
stomatal response by abscisic acid-—an intracellular locus of phytohormone action. Proc
Natl Acad Sci USA. 1994; 91:4019-4023.
177. Anderson BE, Ward JM, Schroder JI. Evidence for an extracellular reception site for
abscisic-acid in commelina guard-cells. Plant Physiology. 1994; 104:1177-1183.
178. Gilroy S. Jones RL. Perception of gibberellin and abscisic-acid at the external face of the
plasma-membrane of barley (Hordeum-vulgare l) aleurone protoplasts. Plant Physiology.
1994; 104:1185-1192.
179. Assmann SM. Ins and outs of guard cell ABA receptors. Plant Cell. 1994; 6:1187-1190.
180. Wickman KD, Clapham DE. G-protein regulation of ion channels. Current Opinion In
Neurobiology. 1995; 5:278-285.
181. Clapham DE. Direct G-protein activation of ion channels. Annual Review Of Neuro-
science. 1995; 17:441-464.
182. Berridge MJ. Inositol trisphosphate and calcium signalling. Natur. 1993; 361:315-325.
183. Wu WH, Assmann SM. A membrane-delimited pathway of G protein regulation of the
guard cell inward K+ channel. Proc Natl Acad Sci USA. 1994; 91:6310-6314.
184. Haldane JBS. Possible Worlds and Other Papers. London, Chatto and Windus. 1927:1-149.
185. Grabov A, Blatt MR. Parallel control of the inward-rectifier K+ channel by cytosolic-free
Ca2+ and pH in Vicia guard cells. Planta. 1997; 201:84-95.
Role of Glycine Betaine and Dimethylsulfoniopropionate in Water-Stress Tolerance 125
CHAPTER 7
Role of Glycine Betaine and

Dimethylsulfoniopropionate
in Water-Stress Tolerance
Douglas A. Gage and Bala Rathinasabapathi
D rought, cold and salinity are three major abiotic conditions limiting the biological
productivity of crops. As a metabolic response to the osmotic stress caused by these
environmental factors, many adapted higher plants, marine algae and bacteria accumulate
organic solutes to equalize external osmotic pressure. The most common osmolytes are
zwitterionic quaternary ammonium compounds such as glycine betaine, the analogous
tertiary sulfonium compound dimethylsulfoniopropionate (DMSP), the amino acid
proline, and polyols like mannitol and glycerol. Together these solutes are known as
“compatible solutes” or “compatible osmolytes” because they are nontoxic compounds that
do not inhibit cellular structure and function even at high concentrations.1 In contrast, high
concentrations of perturbing (incompatible) solutes, such as inorganic ions, can cause
protein denaturation. The exclusion of compatible osmolytes from the hydration sphere of
proteins tends to stabilize the tertiary structure of proteins.2 These compounds also reverse
the disruption of the tertiary structure caused by perturbing solutes, including inorganic
ions.2 In addition, compatible solutes may stabilize proteins during freeze-thaw cycles and
act as cryoprotectants.3-4 To exert a protective metabolic function, it is important that
compatible osmolytes be located primarily in the cytosol and chloroplastic compartments,
but not the vacuole.5
Among the quaternary ammonium compounds, glycine betaine is the most well-known
and widely distributed.6-8 The tertiary sulfonium compound DMSP has an equally effective
osmoprotective role in marine algae and certain higher plants.9-13 Aspects of the occurrence,
compartmentation, synthesis and roles of glycine betaine and DMSP as osmolytes have
been reviewed earlier.5,7,8,14-19 The functional equivalence of these and other quaternary
ammonium and tertiary sulfonium (“onium”) compounds suggests each might be an
appropriate target for metabolic engineering to improve crop stress tolerance.20 However,
the introduction of genes coding for biosynthetic enzymes into new plants requires careful
consideration of the different pathways by which these compounds are formed and the
consequent effects on related primary metabolic processes. Here, we present recent
advances in understanding the synthetic pathways for glycine betaine and DMSP emphasizing
biochemical and molecular genetic results relevant to the development of metabolic
engineering strategies. Comparison of the metabolism of glycine betaine and DMSP is
illustrative of the potential and challenges for engineering stress resistance.
Fig. 7.1. Osmoprotection of Salmonella typhimurium LT2 by glycine betaine and DMSP. Cells
were cultured in minimal medium (MOPS-Glucose) with 0.7 M NaCl (o) or 0.7 M NaCl supplemented
with 1 mM of either glycine betaine (•) or DMSP (∆). (Data replotted from reference 12).
Stress Protection by Glycine Betaine and DMSP In Vivo and In Vitro

Both glycine betaine and DMSP have been shown to afford stress protection for isolated
enzymes, membranes and cultured cells.2,12 For example, glycine betaine has been found
to have unusually strong stabilizing effects on the structure and function of the oxygen-evolving
Photosystem II complex.21 Using bacterial growth bioassays for osmoprotection against
high external salt concentrations,22 glycine betaine and DMSP were found to be equally
effective (Fig. 7.1).12 Many compatible osmolytes also have cryoprotectant activity,3 but
both comparative physiology23-25 and assays on isolated enzymes indicate that DMSP may
be particularly effective in this regard.4,26 Because DMSP does not contain nitrogen, the
accumulation of DMSP could potentially offer advantages over accumulation of glycine
betaine for osmotic adjustment in N-limited environments.
Occurrence of Glycine Betaine in Angiosperms and Other Plants

In certain higher plant taxa, glycine betaine accumulates in response to stress, typically to
50-250 µmol g-1 dry weight. It is known to occur in osmotically significant concentrations
in at least thirteen higher plant families: Amaranthaceae, Asteraceae, Avicenniaceae,
Caryophyllaceae, Chenopodiaceae, Convulvulaceae, Cuscutaceae, Fabaceae, Malvaceae,
Plumbaginaceae, Scrophulariaceae and Solanaceae among the dicots and Poaceae among
the monocots.6,18 A number of other families produce glycine betaine in trace amounts.6,18
Glycine betaine has also been found in a number of marine algae14 and bacteria.27,28 Mass
spectrometry and NMR methods have been most useful for identification and accurate
quantification of glycine betaine. 8,29-31
Occurrence of DMSP
DMSP (also known as dimethyl-β-propiothetin) accumulation to osmotically
significant levels is widespread in marine micro- and macroalgae; it is found in the
evolutionar y diverse classes Chlorophyceae, Rhodophyceae, Dinophyceae,
Chrysophyceae, Bacillariophyceae and Prymnesiophyceae.14,23-25,32 The distribution of
DMSP in marine algae has received considerable attention because this compound is the
primary precursor of biogenic dimethylsulfide (DMS) in the environment. The DMS
formed by specific bacterial or algal enzymatic degradation of DMSP33-35 released into
the water column is a significant component of the global sulfur cycle.36 This gas may
also play a role in climate regulation, because DMS enters the atmosphere from the air-sea
interface (approximately 40 million metric tons/year) and is oxidized to sulfate and other
compounds that form tropospheric aerosols and cloud condensation nuclei.36 Aerosols
and clouds in turn affect the earth's radiative balance and potentially influence climate.
The decomposition of DMSP may also have a defensive function for some phytoplankton
species, because acrylate, the other product in the lyase reaction, deters herbivores.37
Among higher plants, DMSP accumulators are much more restricted, being confirmed
in only two unrelated families, the Asteraceae (Wollastonia biflora13,38) and Poaceae (Spartina
spp39-41 and Saccharum spp- and related taxa12,42). Other members of these two families
contain lower levels of this compound, while some do not produce detectable levels of
DMSP.42 Most of the DMSP-accumulating species are halophytes, although Saccharum
officianale (sugarcane) is a moderately salt-sensitive crop. The marine angiosperms Posidonia
sp7 and Zostera sp43 have also been reported to accumulate DMSP, but these studies did not
exclude potential contributions from marine algal epiphytes. A number of taxa in other
angiosperm families have been shown to make detectable quantities of DMSP, but the
levels are below those likely to contribute to osmotic stress protection. Cyanobacteria may
also produce trace amounts of DMSP.44
Surveys of DMSP in different organisms have often employed assays involving the
detection of dimethylsulfide (DMS) released after the base-catalyzed decomposition of
DMSP44 or thin layer chromatography. While these data must be interpreted cautiously,
many of the reports above have been subsequently confirmed by other methods, primarily
mass spectrometry45 or NMR.14
Genetic Variation for Glycine Betaine Synthesis

Natural variation for glycine betaine content is known in higher plants. Glycine
betaine accumulation has evolved in distantly related families, but non-accumulators do
contain trace amounts. Therefore, it was suggested that glycine betaine accumulation is an
archetypal angiosperm characteristic, strongly expressed by some but weakly by others.8,20
Accumulators and non-accumulators could be found within a family, genus or species. For
example, most members of Poaceae accumulate glycine betaine, but cultivated and the wild
rice species46,47 Eremochloa ophiuroides,48 Echinochloa utilis49 and some cultivars of corn29
and sorghum50 do not. Among dicots, glycine betaine accumulators and non-accumulators
are known within the genus Limonium in the Plumbaginaceae20 and Wollastonia in the
Asteraceae.13,38 In both instances, glycine betaine was replaced with another compatible
solute. 13,17,38
Glycine betaine deficiency in corn has been extensively characterized.8,29 Homozygous
lines deficient in glycine betaine (bet1/bet1) did not oxidize choline to betaine aldehyde,51
suggesting that Bet1 may encode or regulate the choline oxidizing enzyme. To genetically
test the role of glycine betaine in osmotic stress tolerance, near-isogenic F8 pairs of glycine
betaine-containing (Bet1/Bet1) and glycine betaine-deficient (bet1/bet1) lines of corn were
developed.52 When growth parameters were compared between these two genotypes
Table 7.1 Role of Glycine Betaine In Stress Tolerance In Corn (see ref. 53)
Genotype, Treatment LAER LA PH DW
BET/bet1, Control 331 8867 102 68.3
Bet/Bet1, Control 340 9031 107 67.4
bet1/bet1, Salinized 206 5796 53 34.9
Bet/Bet1, Salinized 228a 6364a 61a 40.7a
LSD (0.05) 17 452 6 4.1
Leaf area expansion rate (LAER) (cm2 d-1) for the stress period, total leaf area at harvest (LA)
(cm2), plant height (PH) (cm), and total shoot dry weight (g) at harvest for two F8 families of
corn grown under nonsalinized (control) or salinized conditions (127.5 mM NaCl+ 22.5 mM
CaCl2). aMean (n = 4) of Bet1/Bet1 plants significantly differant at p = 0.05 level from the
mean of sister line bet1/bet1 plants in the same treatment.
under control and salinity stress, both the lines were equal under control conditions, but
the glycine betaine, containing line grew significantly better under salinity stress53 (Table
7.1). Glycine betaine-deficient lines were also more markedly impaired by high temperature
stress as evaluated by membrane integrity and in vivo photochemical activity of PSII.54
Genetic and Environmental Variation in DMSP

There is significant inter- and intraspecific variation in DMSP content among marine
algae.14,25,32,55 Even in the classes, such as the Dinophyceae and Prymnesiophyceae, which
contain a significant number of DMSP accumulating taxa, content can vary widely from
over 2000 µmol/cm3 cell volume (Amphidenium carterae) to below detectable values (some
Prorocentrum accessions). developmental and physiological conditions may account for
some of this variation.55 Environmental conditions can also influence DMSP content in
marine algae. Phytoplankton do not experience significant changes in salinity, but they do
have to maintain high constitutive levels of osmolytes; for some groups DMSP is the dominant
compatible solute, often in concentrations between 50 and 400 mM. In some organisms
(e.g., Tetraselmis subcordiformis) DMSP can occur together with other osmolytes such as
glycine betaine and polyols.4 In contrast to pelagic phytoplankton, estuarine micro- and
macroalgae experience frequent major changes in external salinity. DMSP accumulators
in both constant and variable salinity environments apparently maintain an ability to
adjust to increases in external salt concentrations by increasing DMSP content.55 Reduced
temperatures can also promote increases in DMSP in some algal species.24 Nitrogen nutrition
is another factor that affects DMSP concentrations in a number of algae. Generally, N-limitation
increases DMSP content.56 Internal DMSP concentrations usually increase slowly over a
period of hours in response to environmental conditions, including salinity stress.4 However,
DMSP concentrations can be lowered rapidly by excretion to the outside environment.4
Some algae produce a DMSP lyase to catalyze the breakdown of DMSP to dimethyl sulfide
(DMS) and acrylate, and evidence suggests this enzyme is extracellular.35
DMSP accumulation in higher plants is also variable. In addition to interspecific

variation,12,42 intraspecific variation in a few taxa has been reported. Different strains of
W. biflora contain different concentrations of DMSP13 and some accessions also accumulate
glycine betaine.13,38 Similar variation in DMSP concentrations was found within Saccharum
species and related taxa,12 as well as in Spartina alterniflora41 Increasing salinity causes
DMSP content to increase in W. biflora,13 sugarcane (Saccharum sp) and S. alterniflora.40-41
Other data suggest that DMSP does not function in osmotic adjustment, but rather as a
constitutive osmolyte.57 As in marine algae, DMSP levels were found to be inversely related to
N-nutrition in S. alterniflora.40,57
Glycine Betaine Biosynthesis in Animal and Microbial Systems

In bacteria, animals and higher plants, glycine betaine is synthesized by a two-step
oxidation of choline via betaine aldehyde; however, the enzyme that catalyzes the choline
oxidation step is different in different organisms.
The enteric bacteria Escherichia coli and Salmonella typhimurium, rhizosphere bacteria
such as Rhizobium meliloti, Azospirillum and Pseudomonas, certain marine algae and
cyanobacteria use glycine betaine as a compatible solute. E. coli synthesizes glycine betaine
from exogenous supplies of choline or betaine aldehyde. Choline is oxidized by the action
of a membrane-bound choline dehydrogenase (CDH) to betaine aldehyde, which in turn
is oxidized to glycine betaine by a soluble betaine aldehyde dehydrogenase (BADH). Cells
in which CDH alone was expressed can convert radiolabelled choline into glycine betaine,
leading to the notion that CDH could catalyze both steps of choline oxidation. However,
enzyme activity data suggest that if CDH also catalyzes betaine aldehyde oxidation it should
be at least 40-fold less efficient than its activity with choline.27 A regulatory gene betI controls
the structural genes for choline porter, CDH (betA) and BADH (betB) and choline acts as
an inducer.58 E. coli and S. typhimurium use glycine betaine for osmotic adaptation only,
but R. meliloti demethylates it to glycine through dimethylglycine and sarcosine. The
demethylation of glycine betaine is inhibited in a medium of high osmolality, which thus
permits cells to accumulate glycine betaine as a compatible solute.59 As in E. coli and R.
meliloti, in animals, choline oxidation is catalyzed by CDH and BADH. CDH in liver and
kidney is localized in the inner mitochondrial membrane and the soluble BADH is present
both in the mitochondria and in the cytosol.60-62 Nucleotide sequences coding for CDH
and BADH have been cloned from both animal63-64 and microbial sources.28,65-67 BADH
from bacteria resembles its higher plant counterpart.65
In certain soil bacteria, choline is oxidized by choline oxidase (COX), a soluble flavo
enzyme. In Alcaligenes sp, COX catalyzes the oxidation of both choline and betaine aldehyde,
but the enzyme has about seven-fold lower affinity for betaine aldehyde than for choline.68
Genes encoding COX were cloned from Arthrobacter pascens and A. globiformis.69-70 In Arthrobacter
glycine betaine does not function as an osmoprotectant,70,71 but is utilized as a carbon
source; accordingly COX expression is repressed by the end-product glycine betaine and
regulated by the carbon source in the medium.69
Glycine Betaine Biosynthesis in Higher Plants

Synthesis and accumulation of glycine betaine in higher plants has been studied in
detail in the Chenopodiaceae and the Poaceae. In the leaf cells of halophytic chenopods,
glycine betaine is predominantly localized in the cytoplasm and inorganic ions predominantly
in the vacuole.5,8 Radiotracer studies mainly in chenopods and grasses confirmed that
glycine betaine is synthesized by a two-step oxidation of choline via betaine aldehyde.51,71
Glycine betaine is not catabolized in higher plants tested so far.72-74 In chenopods, the first
step of choline oxidation to betaine aldehyde is catalyzed by choline monooxygenase (CMO)
Fig. 7.2. Glycine betaine biosynthetic pathway, as characterized in spinach. Both of these
enzymatic steps are localized in the chloroplast stroma.77
and the oxidation of betaine aldehyde is catalyzed by betaine aldehyde dehydrogenase

(BADH)75-76 (Fig. 7.2).
Choline Monooxygenase
CMO was purified and partially characterized from salinized spinach.78 Recently,
cDNAs were isolated from spinach79 and sugar beet.80 Unlike CDH and COX from microbial
and animal systems, CMO is a soluble iron sulfur enzyme that requires reduced ferredoxin
for its activity. It is an oligomer of identical subunits of Mr 43,864 with one 2Fe-2S cluster
per subunit. The iron-sulfur cluster is coordinated by two cysteine and two histidine ligands
as in Rieske type iron-sulfur proteins.79 The deduced amino acid sequence for CMO also
has a consensus motif for a mononuclear Fe-binding.80 Upon drought and salinity stress,
CMO mRNA, protein and enzyme activity rose three to seven-fold and returned to
uninduced levels when the stress was removed.80
Betaine Aldehyde Dehydrogenase

In spinach, BADH is a homodimer of nuclear-encoded subunits of Mr 60,000.81-82
Spinach BADH is localized mainly (90%) in the chloroplast stroma.76 Amino acid sequences
deduced from spinach and sugar beet BADH cDNAs83-84 indicates that BADH lacks a typi-
cal stromal targeting peptide. However, expression of these cDNAs in transgenic
tobacco indicated that spinach and sugar beet BADHs were correctly targeted to the chloroplasts in
transgenic tobacco.73 BADH cDNAs or genomic clones were also isolated from spinach,83,85
Atriplex hortensis,86 barley,87 sorghum,88 grain amaranth,89 and rice.90
Several-fold induction of BADH mRNA and enzyme by salinity and drought have
been studied in sugar beet, barley and sorghum.84,87-88 Recent work on purified BADH
from amaranth and spinach BADH expressed in transgenic tobacco indicated that BADH
is not a substrate-specific enzyme. 91,92 BADH efficiently catalyzed dimethyl-
sulfoniopropionaldehyde as well as 4-aminobutyraldehyde and 3-aminopropionaldehyde,
which are intermediates in putrescine and polyamine degradation.92 There were also some
endogenous activities in wild type tobacco with the amino aldehydes tested.92 This suggested
that plants have a family of aldehyde dehydrogenases with distinct but overlapping substrate
specificities. Nakamura et al90 cloned a BADH from rice, a glycine betaine non-accumulator
where it is only weakly expressed. Barley, rice and one of the two sorghum BADHs have a
C-terminal tripeptide, SKL, known to be a signal for targeting preproteins to microbodies;
barley BADH expressed in transgenic tobacco, a dicot, was shown to be localized in the
peroxisomes.90 It was proposed that monocot BADHs function in the peroxisomes.90 This
suggests that glycine betaine synthetic steps are compartmentalized differently between
Fig. 7.3. Choline biosynthesis in spinach. Reversibility of enzyme-catalyzed steps may not be
indicated. Only the major route to phosphatidylcholine is shown. The enzymes are (1) ethanola-
mine kinase (2) to (4) S-Adenosyl methionine-dependent N-methyltransferases, (5) choline
kinase, (6) P-choline phosphatase, (7) CTP:P-choline cytidylyl transferase, (8) CDP-choline
diacylglycerol choline phosphotransferase and (9) phospholipase D.
monocots and dicots. Alternatively, monocot BADHs, all isolated by their homology to
dicot BADHs, may represent one of many aldehyde dehydrogenases with overlapping substrate
specifications and different subcellular location. At present CMO expression is known only
in Amaranthaceae outside of Chenopodiaceae80 and nothing is known about CMO
expression in monocots. Further work on characterization and subcellular localization of
BADH and CMO (or other choline oxidizing enzymes) in a monocot species are required
to confirm any differences in glycine betaine synthesis between monocots and dicots.
Choline Biosynthesis
Choline, the precursor for glycine betaine, is also the precursor for phosphatidylcholine, a
dominant constituent of membrane phospholipids in eukaryotes. In higher plants, choline is
synthesized from serine via ethanolamine. There could be several biosynthetic routes to
free choline from ethanolamine, the routes essentially varying by whether N-methylation
of ethanolamine occurs at the free base, phospho-base or phosphatidyl-base levels (see the
review by Rhodes and Hanson8).
The major steps and enzymes involved in choline synthesis from ethanolamine, mostly
as deciphered in spinach leaf tissue, are shown in Figure 7.3. The metabolic basis for the
increased diversion of choline into glycine betaine was investigated in spinach.93 Ethanolamine
kinase and the three S-adenosylmethionine:phospho-base N-methyl transferases catalyzing
the methylation of phosphoethanolamine are induced by salinity.93 The regulatory step
for choline synthesis appears to be at the enzyme catalyzing the first N-methylation of
phosphoethanolamine. This enzyme's activity is highest at the end of the light period, and
light is required for the salt-responsive upregulation of choline synthesis.94 Comparable
data on regulation of choline synthesis from a glycine betaine non-accumulator or a

monocot are not available. Contrasting the chenopods, where free choline is derived from
phosphocholine (Fig. 7.3), in the Poaceae phosphocholine is incorporated to phosphati-
dylcholine prior to its release as free choline95 and phosphatidylcholine turnover increases
upon salinity treatment.96
Biosynthesis of DMSP
In 1962 Greene97 conducted an initial series of radiotracer experiments with the
chlorophyte alga Ulva lactuca that established that the carbon skeleton, sulfur atom and
methyl groups of DMSP were derived from methionine. It was later demonstrated that
methionine was also the precursor of DMSP in other algal groups98-100 and in higher
plants.37,42,101 The conversion of methionine to DMSP requires four steps: decarboxylation,
deamination, oxidation of the α-carbon and S-methylation. Because intermediates in the
pathway were not characterized in the first biosynthetic studies, the order of these steps
remained unknown until recently, although some evidence from the earlier algal studies97,99
suggested that the S-methylation of methionine to form the sulfonium compound
S-methylmethionine (SMM) was not the first step in the pathway. Maw102 originally
proposed that methionine was converted to DMSP via 4-methylthio-2-oxobutyrate
(MTOB), then to methylthiopropionate (MTP) and finally by methylation to DMSP. However,
there are a number of biochemically plausible alternative routes compatible with the initial
radiolabeling data. Recent work in both higher plants and several algal groups has shed
light on the DMSP pathway and led to the surprising finding that at least three, and perhaps
four, biosynthetic routes are found in nature. The results leading to this conclusion are
reviewed in detail in the following sections. Progress in the isolation of genes coding for
the DMSP biosynthetic enzymes in both marine algae and higher plants lags behind that
for glycine betaine, but some of the enzymes have been characterized, and in a few cases
purified.103-104
DMSP Biosynthesis in Marine Algae

Biosynthetic studies have now been carried out in four classes: the Chlorophyceae,
Bacillariophyceae, Prymnesiophyceae (Haptophyceae) and Dinophyceae. Using radioisotope
feeding, Ushida and coworkers105 showed that methionine was incorporated into DMSP
in the heterotrophic dinoflagellate Crypthecodinium cohnii. As in the chlorophyte alga Ulva
lactuca,97 the C-1 carboxyl group of methionine was lost, confirming that the biosynthesis
must involve decarboxylation, deamination, oxidation of the α-carbon and S-methylation. The
order of these steps in C. cohnii was not completely clear, but several lines of evidence
indicated that the DMSP pathway in this organism did not involve an initial S-methylation
step. First, feeding cold SMM did not inhibit the labeling of DMSP from radiolabeled
methionine, suggesting that SMM was not an intermediate in the pathway.104-105 Alternatively,
the methionine transamination product, methylthio-2-oxobutryate (MTOB) did not block
the incorporation of label from methionine to DMSP. This result could be interpreted to
mean that a pathway, involving transamination of methionine to MTOB, is not the first
step in DMSP biosynthesis. In contrast, methylthiopropionate (MTP) did block uptake of
radiolabel into DMSP, supporting its assignment as a biosynthetic intermediate in this
dinoflagellate and inferring that S-methylation of MTP is the last step to DMSP (Fig. 7.4).104
However, as the authors point out, these data may be explained alternatively by the lack of
uptake of externally-supplied SMM or MTOB to the proper intracellular compartment or
the inhibition of methionine uptake by MTP. It should also be noted that MTOB is somewhat
unstable, so that decomposition may have prevented supplied MTOB from trapping label
in feeding studies.
Fig. 7.4. Biosynthetic pathway to DMSP in the dinoflagellate Crypthecodinium cohnii.104,105 A PLP-
dependent methionine decarboxylase has been isolated and partially characterized.
If MTP is an intermediate, then the formation of MTP from methionine requires C-1
decarboxylation, deamination and oxidation of the C-2 carbon. The isolation of a pyridoxal
5'-phosphate-dependent methionine decarboxylase from C. cohnii104 provided some support
for the notion that decarboxylation was likely the first step in converting methionine to
MTP. However, the labeling patterns of the methionine decarboxylation product,
methylthiopropylamine (MTPA) or the other putative intermediate, MTP, were not investigated, so
the case for this pathway remains circumstantial. The presence of a methionine decarboxylase,
while consistent with the hypothetical pathway, is not conclusive, since similar enzymes
are likely widespread in non-DMSP producers (e.g., the fern, Dryopteris felix-mas and Streptomyces
sp106-107). A different mechanism to form MTP from methionine via the intermediate
methylthiopropylamide by peroxide-dependent oxidative decarboxylation108-109 might also
be possible.
DMSP biosynthesis was recently investigated in several different classes of algae.100 In
the chlorophyte alga Enteromorpha intestinalis, a species not too distantly related to Ulva
lactuca, feeding studies with 35S-labeled methionine established that this amino acid was
Fig. 7.5. Biosynthesis of DMSP in the chlorophyte alga Enteromorpha intestinalis.100 The enzymatic
mechanisms catalyzing the first three steps have been elucidated, and the intermediates MTHB
and DMSHB have been shown to be the D isomers.118
the precursor of DMSP, as in all other organisms investigated to date. In this study the
intermediates in the pathway were identified by following the incorporation of 35S-label
into other metabolites. Two compounds, methylthio-2-hydroxybutyrate (MTHB) and
dimethylsulfonium-2-hydroxybutyrate (DMSHB), were identified as metabolites that rapidly
acquired label and lost it as the 35S-methionine was depleted. Time course experiments and
kinetic analysis showed that the MTHB and DMSHB labeling patterns fit those predicted
for biosynthetic intermediates. With the observed labeling kinetics, the most likely route
from methionine to DMSP via these intermediates would be through the unstable compound
MTOB, which is then reduced to MTHB and subsequently methylated to DMSHB (Fig. 7.5).
Because of its instability, the role of MTOB in the pathway was determined by two methods:
(1) The use of a gentle extraction procedure; and
(2) The conversion of MTOB to a stable product, MTHB, by reduction with sodium
borohydride and correcting for endogenous MTHB levels. This study conclusively
demonstrated that MTOB acquired and lost label consistent with its role as an early
intermediate. Kinetic analysis of the labeling of MTOB, MTHB, DMSHB, confirmed
that flux through the pathway was quantitatively sufficient for these three compounds
to be intermediates in E. intestinalis (Fig. 7.5).
Other compounds monitored in this study, including SMM, dimethylsulfonio-
propionaldehyde, acquired little or no 35S. The putative precursor in the C. cohnii DMSP
pathway, methylthiopropylamine (MTPA), picked up label slowly, as would be expected
for a minor end product, not an intermediate. MTP, another proposed precursor, was
labeled to widely varying extents in different E. intestinalis batches. Gage et al100 concluded
that the labeled MTP detected was formed from oxidative decarboxylation of MTOB, a
catabolic reaction that is known in other organisms that do not accumulate DMSP.98,110-112
Feeding experiments with the synthetically prepared 35S-labeled metabolites SMM,
MTHB and DMSHB supported the proposed route to DMSP (Fig. 7.5). SMM was taken
up, but not metabolized to any appreciable extent, indicating that it was not an intermediate in
the pathway. MTHB was primarily converted back to methionine (and ultimately protein),
which was not an unexpected fate for this metabolite,113,114 but some MTHB was converted to
DMSP. These data do not exclude an indirect route through methionine. However, DMSHB,
the key intermediate in the proposed pathway, was efficiently converted to DMSP, but not
to other products. Stable isotope labeling and mass spectrometry were used to confirm the
identification of DMSHB as an intermediate. These data also shed light on the enzymatic
reactions involved in the pathway in E. intestinalis (see the section below).
The possibility of two different pathways to DMSP in chlorophyte algae and heterotrophic
dinoflagellates (see Figs. 7.4, 7.5), raises questions about the biosynthesis in other classes
of algae that are known to accumulate this osmolyte. The pathway was also investigated in
the prymnesiophyte Emiliania huxleyi, the diatom Melosira nummuloides and Tetraselmis
sp, a prasinophyte.100 Each of these species was found to contain small pools of DMSHB
which accumulated label from 35S-labeled methionine and lost it as the supplied methionine
was depleted. In addition, all three taxa were able to metabolize supplied 35S-labeled DMSHB
to DMSP. These data suggest that at least two other algal classes, the Prymnesiophyceae
and Bacillariophyceae (diatoms), have the same pathway as the Chlorophyceae, while
dinoflagellates may use different means to synthesize this compound. As will be shown
below, DMSP biosynthesis in higher plants proceeds by yet another pathway.
Enzymes in Algal DMSP Biosynthesis

As mentioned above, the first enzyme in the proposed DMSP pathway in the dinoflagellate
C. cohnii was characterized.104,105 A pyridoxal 5'-phosphate-dependent methionine
decarboxylase from this organism was isolated that is a homodimer of two 103 kDa
subunits. This contrasts with the methionine decarboxylases previously isolated from Dryopteris
felix-mas and Streptomyces sp, which are homodimers of 57 and 59 kDa subunits, respectively.
No kinetic data were reported for the dinoflagellate decarboxylase. The enzymes involved in the
subsequent conversions in the C. cohnii pathway (Fig. 7.4) have not been investigated yet.
In E. intestinalis stable isotope labeling studies provided initial information about
several of the enzymatic steps in the DMSP pathway in this chlorophyte. The conversion of
DMSHB to DMSP was shown to be mediated by an oxidase mechanism analogous to that
of lactate oxidase.115 When [U-13C5]methionine was supplied to E. intestinalis in an atmosphere
containing either 16O2 or 18O2, a labeling pattern was observed that was consistent with an
oxygenase-mediated oxidative decarboxylation in the last step in the pathway (DMSHB to
DMSP, Fig. 7.5).100
Stable isotope labeling was also used to investigate the first step in the pathway, the
deamination of methionine to MTOB. There are two common mechanisms that could
effect this conversion: transamination and oxidative deamination. To distinguish the two
reactions, 15N-labeled methionine was supplied to E. intestinalis. Glutamate, alanine and
aspartate all rapidly acquired label, but the amide N of glutamine did not, supporting a
transamination mechanism in the deamination of methionine, rather than oxidative deamination.
In the latter case, free 15NH3 released in the reaction would be predicted to be incorporated in
amide N of glutamine by glutamine synthase.116 Control experiments where 15NH4+ was
supplied showed that the glutamine amide N could be labeled and by inference that
glutamine synthase was operational in E. intestinalis. Additional evidence for the involvement
of a transaminase was provided by the data showing that MTHB was converted back to
methionine; transaminase reactions are reversible.117 By inference, the reductase catalyzing
the conversion of MTOB to MTHB must also be reversible.
Summers et al118 recently provided additional evidence about the enzymes catalyzing
the first three steps in the DMSP pathway in E. intestinalis and other chlorophyte algae.
Employing in vitro assays with partially purified E. intestinalis extracts, these authors were
able to confirm that a transaminase was responsible for the conversion of methionine to
MTOB. Of potential amino group acceptors, 2-oxoglutarate was found to be preferred.
Kinetic analyses showed that the observed enzyme activity was approximately 30-fold higher
than that observed in comparable extracts of three non-DMSP accumulating chlorophyte
algae (Halimeda discoides, Caulerpa ashmedii and Utodea conglutinata). Similar methionine
transaminase activity was found in other DMSP-accumulating chlorophytes (E. fasiculata
and Ulva spp).
The second step, the conversion of MTOB to MTHB (Fig. 7.5) was catalyzed by an
NADPH-dependent reductase. The MTHB produced in this reaction was exclusively the
D-isomer. The reverse reaction, MTHB to MTOB, was NADP-dependent and proceeded
at only 0.4% of the forward rate. Comparisons with other algal extracts demonstrated that
the activity of MTOB reductase was again significantly higher in E. intestinalis and the
other DMSP-producing algae than in three taxa studied that do not accumulate DMSP. The
in vitro results were confirmed in vivo by feeding D-[35S]MTHB to intact E. intestinalis
fronds.
The stereoselectivity for the production of D-MTHB was consistent with the substrate
preference for the next enzyme in the pathway, MTHB S-methyltransferase. E. intestinalis
extracts catalyzed the methylation of D-MTHB, but not L-MTHB, with S-adenosyl
methionine acting as the methyl donor. In accord with the proposed biosynthetic route
(Fig. 7.5) and previous radiolabeling experiments,100 the methyltransferase activity was
selective for MTHB; no activity was detected for MTP, or other thioethers (e.g.,
methylthiopropylamine, D- or L-methionine). As with the previous two enzyme activities,
extracts of the other DMSP-accumulating algae were similar to the preparation from
E. intestinalis, while those of the non-accumulators did not contain significant catalytic
activity. These results suggest that the methylation product, D-DMSHB, might be the preferred
substrate for the oxidase catalyzing the last step in the formation of DMSP. In vivo labeling
experiments with both L- and D-[35S]DMSHB demonstrated that the D isomer was converted
9 times more efficiently to DMSP than the L isomer.
Biosynthesis of DMSP in Higher Plants

The biosynthesis of DMSP in higher plants was first investigated in the Indo-Pacific
strand plant, Wollastonia biflora (synonyms Melanthera biflora and Wedelia biflora), a
member of the Asteraceae.38 DMSP accumulation in this species is variable and some
accessions also accumulate glycine betaine. 13 The DMSP biosynthetic studies were
performed with a genotype rich in DMSP, but lacking glycine betaine. Initial
radiolabeling studies with W. biflora leaf discs confirmed that methionine was the
precursor of DMSP.38
When [U-14C]methionine was fed to leaf discs and then chased by a supply of cold
methionine (pulse-chase labeling), one metabolite, SMM, acquired and lost label in a manner
quantitatively consistent with it being an intermediate. This is in contrast to Greene's 97
results in U. lactuca, where only small amounts of labeled SMM were detected. Another
putative intermediate in the DMSP pathway, MTP, accumulated only a small amount of
label slowly, suggesting that it is a minor end product in methionine metabolism in this
species, and not a precursor to DMSP. Feeding [14C]SMM and [14C]MTP showed that
only SMM was efficiently converted to DMSP. Further, cold SMM, but not MTP, strongly
reduced the incorporation of label into DMSP from [U-14C]methionine.
While these data supported the role of SMM as an intermediate in DMSP biosynthesis in
W. biflora, interpretation of the results is complicated by the facile interconversion of SMM
and methionine via the SMM cycle.119 In this futile cycle that is likely ubiquitous in
angiosperms, methionine is first S-methylated to form SMM by the following reaction:
methionine + S-adenosylmethionine → SMM + S-adenosylhomocysteine. SMM can then
react with homocysteine enzymatically released from S-adenosylhomocysteine by
S-adenosylhomocysteine hydrolase119 to give two molecules of methionine. The function
of this cycle is still unclear, but it has been proposed that SMM may be a way to store
methionine in a less metabolically active form.119 Hanson et al38 showed that SMM was in
fact a direct precursor of DMSP, and not involved in the pathway only by conversion back
to methionine, by labeling one of the methyl groups of SMM with 13C and the other with
2
H3. Both methyl groups of DMSP retained the labels and there was no evidence of methyl
group scrambling. If the SMM was first converted to methionine before going on to DMSP,
then one of the two methyl groups in SMM would be lost and a much more complicated
labeling pattern in DMSP would be expected. Thus, it appears that the widespread
metabolite SMM119,120 has been diverted for DMSP synthesis in this species.
In W. biflora, radiolabeling pulse-chase and trapping experiments demonstrated
that the next intermediate in the pathway from SMM to DMSP was dimethylsulfonio-
propionaldehyde (DMSP-ald).121 Although the conversion of SMM to DMSP-ald requires
two steps, decarboxylation and deamination, no free intermediates such as dimethyl-
sulfoniopropylamine (DMSP-amine) or dimethylsulfonio-2-oxobutanoic acid (DMSOB)
were detected in radiolabeling studies. The latter compound has never been synthesized
and is predicted to be extremely unstable,122 so it is unlikely to be a free intermediate in the
pathway. DMSP-ald itself is rather unstable, and degradation to DMS is the main byproduct
when the radiolabeled compound is supplied.122 Because no precursors of DMSP-ald other
than SMM were detected in feeding studies with a number of labeled, biochemically-plausible
Fig. 7.6. Biosynthesis of DMSP in Wollastonia biflora38,121 and Spartina alterniflora.101 In W. biflora
the 2-step conversion of SMM to DMSP-ald is catalyzed by a transamination/decarboxylase
complex125with no free intermediate (Step1). A different pathway is found in S. alterniflora
involving decarboxylation of SMM to DMSP-amine (Step 2). The mechanism by which
DMSP-amine is converted to DMSP-ald (Step 3) is not clear, but it could be catalyzed by an
amine oxidase, dehydrogenase or transaminase.
intermediates (prepared synthetically), it was proposed that intermediates between SMM

and DMSP-ald might be tightly bound to a multifunctional enzyme complex catalyzing
both the oxidative deamination and decarboxylation steps.122 Evidence that in fact this
enzymatic step proceeds by a combined transamination-decarboxylation reaction123 will be
discussed below. The biosynthetic pathway to DMSP in W. biflora is summarized in Fig. 7.6.
Recent studies with another DMSP accumulator, Spartina alterniflora, has shown an
interesting variation in the higher plant pathway. In contrast to the pathway in W. biflora,
feeding studies with [35S]methionine have shown that there is a free intermediate,
dimethylsulfoniopropylamine (DMSP-amine), between SMM and DMSP-ald in this monocot.101
Both labeling kinetics and direct feeding studies showed that this novel compound was a
free intermediate in the pathway. Modeling of the pathway flux with data from radiolabeling
experiments, and DMSP-amine’s limited ability to act as a cold trap, indicated that there
are metabolically active and “storage” pools of DMSP-amine in S. alterniflora. This analy-
sis also indicated that exogenously-supplied DMSP-amine first enters the storage pool before
being metabolized. In control experiments, non-DMSP accumulating grasses could not

convert [35S]DMSP-amine to DMSP.101 Together these data indicate that DMSP biosynthesis
in the monocot S. alterniflora is distinct, but related to that in the dicot W. biflora. SMM is
a biosynthetic intermediate in several other taxa that are minor producers of DMSP,42 but
subsequent steps in the pathway remain unknown in these groups.
The occurrence of at least three, and likely four, distinct biosynthetic pathways in
different groups of organisms (the dinoflagellate C. cohnii, the chlorophyte E. intestinalis
and other algae, the dicot W. biflora and the monocot S. alterniflora) makes DMSP unique
among natural products. This diversity presents several alternative approaches for the
genetic engineering of this compound, each with possible advantages. The enzymes
involved in higher plant DMSP biosynthesis are only beginning to be isolated and characterized, but
the mechanisms involved in the individual steps are becoming clear.
Enzymes in Higher Plant DMSP Biosynthesis

In both W. biflora and S. alterniflora, the first step in the DMSP pathway is the
methylation of methionine by the enzyme S-adenosylmethionine:methionine
S-methyltransferase (MMT).103 S-adenosylmethionine is the methyl donor. As part of the
ubiquitous SMM cycle,119 this enzyme is likely found in all angiosperms. In DMSP-
producing higher plants, MMT has been recruited as the first committed step in the
pathway. This enzyme has now been purified and characterized from W. biflora. MMT is a
homotetramer of an approximately 115 kDa polypeptide. Antibodies to MMT cross-react
with a similar sized polypeptide in non-DMSP accumulators, such as cabbage, clover and
maize. This structure is significantly different from that of other methyltransferases, which
are usually monomers or dimers of 20-45 kDa polypeptides.124 Because only partial
sequence of MMT has been obtained, and the cloning of the corresponding gene is not yet
complete, the significance of the size of MMT is not yet clear.
The enzyme involved in the direct conversion of SMM to DMSP-ald in W. biflora has
also been investigated.125 As discussed earlier, this conversion requires a two-step reaction
involving removal of the amino group and decarboxylation. Stable isotope labeling experiments with
[15N]SMM showed that a transamination, rather than a deamination, step was likely.125
However, simple transamination of SMM produces the extremely unstable compound
dimethylsulfonio-2-oxobutyrate (DMSOB), which would be expected to rapidly undergo
b,g-elimination to yield DMS and vinylglyoxylate. 122 Because the kinetics of the
decomposition reaction indicates that free DMSOB is not likely a product, a coupled tran-
saminase/decarboxylase enzyme complex was proposed, perhaps involving pyridoxal
5'-phosphate (PLP) as a cofactor or cosubstrate. Given the nature of the substrate and
intermediates in this reaction, this complex would be biochemically unusual, if not
unprecedented.
In S. alterniflora, little is known about the enzymatic steps catalyzing the conversion
of SMM to DMSP-ald via the intermediate DMSP-amine.101 However, an SMM decarboxylase
must be present. Subsequent conversion of DMSP-amine to DMSP-ald could proceed via
one of several alternative mechanisms involving an amine oxidase, dehydrogenase or
transaminase. It is possible that the two novel enzymatic steps in the S. alterniflora DMSP
pathway might have originated from more widespread enzymes. For example, SMM is
structurally analogous to S-adenosylmethionine, and S-adenosylmethionine decarboxylases
are widespread in plants.126 Whether the similarity of DMSP-amine to other diamines
indicates that a diamine oxidase might be involved is intriguing, but speculative at this
point.
The last step in higher plant DMSP biosynthesis is common to both the S. alterniflora
and W. biflora pathways; in monocots and dicots DMSP-ald is oxidized to DMSP. The
Fig. 7.7. Subcellular localization of the DMSP biosynthetic steps in W. biflora. A schematic repre-
sentative of the subcellular localization of the biosynthetic enzymes in the DMSP pathway (after
refs. 125, 127). MMT is cytosolic, but the enzyme(s) involved in the conversion of SMM to DSMP-ald and
DDH are stromal. There is recent evidence for salt stress-activated SMM import into the chloroplast, but
little is known about the export mechanism for DMSP.128 Although it is not shown, the intermediate
identified in S. alterniflora between SMM and DMSP-ald, DMSP-amine,101 is also likely chloroplastic.
enzyme responsible for catalyzing this reaction has not yet been isolated, but some of its
features have been characterized. A pyridine nucleotide-dependent DMSP-ald dehydrogenase
(DDH) activity in W. biflora has been identified.127 This enzyme utilized either NAD or
NADP, though NAD was preferred, and the effect of these cofactors was not additive. The
measured activity and Km was sufficient to account for the in vivo rate of DMSP synthesis.
There is an interesting parallel between this enzymatic step in DMSP synthesis and that of
BADH in glycine betaine. Antisera against BADH neutralized DDH activity in W. biflora
extracts and immunoblot analysis showed a single polypeptide band at 63 kDa,127 similar
to that of BADH subunits in other plants.76,82 Betaine aldehyde was also found to be a
weak competitive inhibitor of DDH.127 These data suggest that DDH and BADH may be
closely related enzymes. BADH isolated from the non-DMSP producer Amaranthus
hypochondriacus efficiently catalyzes the oxidation of DMSP-ald to DMSP.91 In fact, the
Vmax/Km value indicates that DMSP-ald is a better substrate for BADH than betaine aldehyde.91
Further, as was discussed above, tobacco plants engineered to express sugar beet BADH
were able to oxidize DMSP-ald and several other aldehyde substrates.92 Thus, BADH and,
by inference, DDH, are clearly not as substrate-specific as previously supposed and may
have originally evolved for another role, perhaps in polyamine metabolism. In any case,
this enzymatic step in DMSP biosynthesis may have evolved by the recruitment of preexisting
enzymes.
Localization of DMSP Biosynthesis

Little is known of the intracellular localization of the DMSP biosynthesis in marine
algae, but recent studies have provided insight into the compartmentation of DMSP biosynthetic
enzymes in higher plants. Studies with W. biflora mesophyll protoplasts employing subcellular
fractionation and immunological techniques demonstrated that MMT is a cytosolic
enzyme.127 This observation is in accord with earlier investigations of SMM metabolism.119
DDH, in contrast, was localized in the chloroplast stroma (Fig. 7.7).127 The relative instability
of the substrate DSMP-ald suggests that it should be synthesized in close proximity to
DDH and labeling experiments showed this is the case. Feeding [35S]SMM to intact chloroplasts,
incubated in the light with HCO3- and 3-phosphoglyceric acid to promote photosynthesis,
efficiently produced labeled DMSP. (The enhancement of DMSP synthesis under photosynthetic
conditions implies there may be some link to the Calvin cycle, but at present this relationship is
unclear.) The synthesis of DMSP indicates that there is a mechanism to import SMM into
the chloroplast and that the next enzyme in the pathway, the transaminase/decarboxylase
complex that converts SMM to DMSP-ald, is in the same compartment. A similar
compartmentation of DMSP biosynthetic enzymes may be occurring in S. alterniflora,
where DMSP-amine is an intermediate between SMM and DMSP-ald. Although analogous
subcellular fractionation experiments have not been conducted, the limited ability of
exogenously-supplied DMSP-amine to act as a cold trap in feeding studies may reflect the
fact that the metabolically active DMSP-amine is chloroplastic and that the “storage pool”
is cytosolic.101
Follow-up studies have shown that SMM import into the chloroplast may play a key
regulatory role in DMSP biosynthesis in W. biflora.128 When salinized with 30% seawater,
intact protoplasts from W. biflora were found to produce increased levels of DMSP.
Quantitative analysis of SMM and DMSP in the intact protoplasts and in chloroplasts
derived from the protoplasts demonstrated that there was a significant intracellular redis-
tribution of SMM. SMM levels decreased in salinized leaves, with the drop primarily in
extrachloroplastic SMM, while levels inside the chloroplast were similar. Thus, 40% of the
SMM was chloroplastic in unsalinized protoplasts, while 80% was found in the chloro-
plasts following salinization. These results suggest that in W. biflora the SMM transporter
is activated under salt stress in order to increase DMSP biosynthesis. As expected, DMSP
content also increased in the chloroplasts upon salinization.128 Estimates indicate that
stromal concentrations go from approximately 60 mM in control chloroplasts (44% of
the total) to 130 mM in salinized chloroplasts (69% of the total). While there must be a
means to translocate DMSP from the chloroplasts into the cytosol, this compound’s
intracellular distribution in other compartments has not been thoroughly investigated.
Leakage of DMSP from chloroplasts during their isolation was observed, however.128
Whether active or passive transport of DMSP out of the chloroplast occurs in vivo is
unknown.
Metabolic Engineering of Glycine Betaine Synthesis

In recent years glycine betaine has been an active target for genetic engineering into
nonproducing organisms to provide improved osmotic stress resistance. Genes encoding
glycine betaine biosynthetic enzymes from microbes and higher plants have been used to
engineer this pathway in heterologous organisms. For example, the bacterium
S. typhimurium lacks natural ability to oxidize choline; expression of E. coli genes for
choline transport, CDH and BADH in S. typhimurium conferred increased osmotolerance
in the presence of choline.27 Similarly, introduction of a gene for COX from A. pascens into
an E. coli mutant defective in betaine synthesis resulted in glycine betaine synthesis and
osmotolerance upon exogenous choline supply.69 Nomura et al129 transformed the fresh
water cyanobacterium Synechococcus sp. with E. coli bet genes. The transformed cells took
up exogenous choline, accumulated glycine betaine up to a concentration of 45 mM and
tolerated salt stress.129 Similarly, the gene for choline oxidase (codA) from A. globiformis
was introduced to Synechococcus. The transformed cells synthesized glycine betaine from
exogenous choline and were tolerant to salt stress70 and low temperature stress.71 In these
microbial models, the role of glycine betaine in stress protection has been well demonstrated
and in each case the host organism depended on exogenous choline to synthesize glycine
betaine.
Engineering glycine betaine synthesis in higher plants is more complex, especially
with reference to the availability and regulation of choline. Some of the considerations in
this context were presented by McCue and Hanson.5 Many important crops such as rice,
legumes, canola, tomato and cucurbits lack the ability to accumulate glycine betaine. Hence,
engineering its synthesis is an attractive route to improve their stress tolerance. The natural
variation for glycine betaine synthesis in higher plants (see above) suggests that engineering
this pathway should be a biological feasibility. Both glycine betaine non-accumulators and
accumulators have comparable free choline pools and there is evidence that choline is
feedback regulated.8,52 However, at present it is not clear how much plasticity there is in
choline production if significant flux to a new sink is introduced. It is quite possible that
the large methyl group demand required to accumulate osmotically significant quantities
of glycine betaine would exceed the normal capacity of choline synthesis to respond. In
order to engineer glycine betaine (or DMSP) accumulation in higher plants, manipulation
of folate-mediated methyl group metabolism130 may be required.
Some initial experiments have been done to address this question. Coding sequences
for BADH from spinach, sugar beet, barley and E. coli were successfully expressed in
transgenic tobacco under constitutive promoters.73,87,131 When supplied with betaine
aldehyde, the transgenic tobacco expressing spinach or sugar beet BADH accumulated
glycine betaine to levels comparable to glycine betaine accumulators.73 However, endogenous
pools of choline were not diverted to glycine betaine in these transformants because the
choline derived precursor, betaine aldehyde, was provided. Engineering of the choline
oxidation (CMO) step into any of these BADH-expressing transgenic tobacco has not yet
been done.
Lilius et al132 introduced the E. coli betA gene encoding CDH into tobacco. Two
transgenic lines of tobacco were shown to grow better than one wild type control under
salt stress, but glycine betaine synthesis was not confirmed or quantified.132 When betA
was expressed in potato, one transgenic line tested produced glycine betaine up to 108
nanomoles g-1 fwt compared to 44 nanomoles g-1 fwt in the wild type control, though the
plants were supplied with 15 mM choline in the medium.133 Low levels of glycine betaine
accumulation could be due to poor expression of the transgene or limited access by CDH
to supplied choline with resulting poor betaine aldehyde oxidation.
Recently a chimeric construct containing the codA gene (from A. globiformis) for COX
under the control of a constitutive promoter was introduced into Arabidopsis thaliana.134
COX was targeted to the chloroplasts by using the transit peptide of the small subunit of
Rubisco.134 Three lines of transformants were shown to accumulate about 1 µmole g-1 fwt
(equalling about 50 mM internal concentration) and had improved tolerance for salt and
cold stress.134 This is the most unequivocal and direct proof for the role of glycine betaine
in stress tolerance in higher plants. Similar results have been achieved by transforming the
codA gene into rice.134 Yet, the glycine betaine concentrations in these transformants are
still below those found in some glycine betaine producers. Further study is needed to
determine if choline metabolism has reached an upper limit.
When spinach CMO cDNA was expressed under a constitutive promoter in transgenic
tobacco (without spinach or beet BADH), the transgenic lines synthesized about two-fold
more glycine betaine than wild type tobacco or vector alone control (Hanson, personal
communication). Presumably, in this case an endogenous aldehyde dehydrogenase catalyzed
the conversion of the betaine aldehyde product to glycine betaine. The limited increase in
glycine betaine may have been the result of compartmentation and restricted access or
availability of the substrate to the dehydrogenase. Tobacco expressing both CMO and BADH
from heterologous sources should provide valuable insights as to whether the transgenic
plants synthesize osmotically significant quantities of glycine betaine. The results from
these experiments will help to address whether or not choline metabolism will be a limiting
factor for glycine betaine synthesis.
Despite considerable success in engineering glycine betaine synthesis in plants using
genes for choline oxidizing enzymes from microbial sources,134 genes from plants would
appear to have several physiological advantages. The plant enzyme CMO requires reduced
ferredoxin for its activity. This links glycine betaine synthesis with the light reactions of
photosynthesis and could help match the supply of glycine betaine with the demand for
osmotic adjustment and osmoprotection. This demand for osmotic adjustment climbs
rapidly after sunrise as the water potential and water content of salt- or drought-stressed
leaves start falling.135 Similarly, the light control of the N-methyltransferase that catalyzes
phosphoethanolamine methylation in choline synthesis is physiologically linked to the
higher demand for choline flux to glycine betaine under light. Another consideration is
that in higher plants the glycine betaine synthetic pathway has evolved for osmoprotection,
while in microbial systems it may play either a nutritional role alone or a combined role
with osmoprotection. The regulatory controls on CMO, BADH and choline biosynthetic
enzymes from higher plants might therefore be superior to the microbial enzymes. For
example, cis-regulatory elements for these genes should be ideal for engineering higher
plants. Further insight into some of the issues concerning regulation might be provided by
experiments with antisense CMO to downregulate glycine betaine synthesis in plants that
naturally accumulate glycine betaine.
Prospects for Engineering DMSP Synthesis

Because DMSP is a non-nitrogen-containing compatible osmolyte, the genetic
engineering of its biosynthesis into crop plants is attractive. DMSP's effective
cryoprotectant activity would be another potential benefit for crops.4,23-25 However,
because DMSP’s initial precursor is methionine, accumulation of osmotically significant
amounts of DMSP might triple the requirement for reduced sulfur and also increase the
demand for methyl groups.136 While the nitrogen from methionine can be recycled, the
demands on sulfur amino acid and C-1 metabolism may make metabolic compensation diffi-
cult, without additional manipulations of these primary metabolic pathways (see ref. 136 for
further discussion of DMSP in the context of primary metabolism). At present little is known
of the potential metabolic consequences on sulfur and methyl group metabolism of
engineering the DMSP pathway. Investigating primary metabolic adaptations in natural
DMSP accumulators might provide insight into this problem.
Another potential complication is that DMSP accumulation in crops could
introduce a new source of DMS emissions into the environment. At physiological pH,
DMSP is relatively stable. The significant release of DMS from marine systems is primarily
mediated by the enzymatic breakdown of DMSP by specific bacterial and algal lyases.
Therefore, DMS release could be negligible in the absence of these lyases. On the other
hand, in Spartina alterniflora and Wollastonia biflora it has been estimated that
approximately 1% of the DMSP pools turn over by degradation to DMS every day.40,121
However, it is possible that the DMS emissions from these plants do not originate from
endogenous DMSP-lyase activity, but rather from the breakdown of the unstable biosynthetic
intermediate DMSP-ald or degradation of SMM by a specific hydrolase.121 With additional
manipulations, DMS emissions may be avoidable in plants engineered to accumulate DMSP.
The practical opportunities for genetically engineering the DMSP pathway into other
plants are currently limited, compared to those for the glycine betaine pathway, simply
because the first genes for the biosynthetic enzymes are only now being cloned. However,
as these genes become available, the alternative routes to DMSP in different organisms will
provide several options with alternative benefits and disadvantages. While it may be premature
to speculate on particular engineering strategies, it is interesting to consider the alternatives.
Targeting of the gene product or intermediate to the proper compartment (e.g., DDH or
SMM to the chloroplast stroma) is another important issue. Engineering enzymes for
translocation to a specific compartment may be much easier than targeting metabolites.
Specific transporters may be required,128 about which almost nothing is currently known.
There are four potential DMSP biosynthetic pathways to consider (see Figs. 7.4-7.6).
The number of steps in each of these pathways may seem daunting from the perspective of
engineering, especially given the simplicity of the glycine betaine pathway. However, as
was illustrated above, some of the enzymes involved are common to most plants, so that
introduction of new genes may not be required. For example, the dinoflagellate pathway104
is initiated by decarboxylation of methionine by a PLP-dependent decarboxylase. Although
the C. cohnii enzyme appears to be unusual,104-105 functionally equivalent enzymes are
known from other organisms106-107 and may be widespread. The mechanism(s) involved
in the next step in the pathway (MTPA→MTP) is not clear at this time and might require
an amine oxidase, dehydrogenase or transaminase. The last reaction would also require
the engineering of a specific methyltransferase.
The E. intestinalis pathway100 (Fig. 7.5) again appears to have evolved by utilizing
some of the enzymes involved in primary amino acid metabolism.118 Methionine transaminases
(methionine→MTOB) are likely ubiquitous, although the enzyme in the DMSP producer
appears to be 30- to 100-fold more active, with a much lower Km (30 µM vs mM range for
most amino transferases) and is therefore likely a novel enzyme rather than over-expression of
a standard amino transferase.118 To successfully engineer DMSP accumulation, the gene
for the specialized methionine amino transfersase would likely have to be introduced. The
position of a transaminase at the head of the DMSP pathway in E. intestinalis may explain
the increased production of DMSP under N-deficit conditions. Reduction of cellular amino
acid content would increase transamination reactions and thus activate the DMSP pathway.
For some environments this regulatory point in an engineered crop might be valuable. In
contrast, under low N conditions, glycine betaine production is diminished.
Like the first enzyme in the pathway, many plants have a low level of MTOB reductase
activity (MTOB→MTHB).137 The stereochemical configuration of the MTHB produced
has not been determined for plants other than E. intestinalis, however. Because the MTOB
reductase is probably also specialized, introduction of the reductase gene would also be
required to engineer the pathway. Finally, a unique methyltransferase converting D-MTHB
to D-DMSHB is only found in DMSP accumulators and would by necessity be required
for the pathway. However, it is possible that engineering just the first three steps would be
sufficient to impart osmotic stress resistance. The last step in the pathway, conversion of
DMSHB to DMSP, might not be necessary. DMSHB is an effective compatible osmolyte, 118
although it is subject to enzymatic degradation to DMS in vivo in the algae that use this
compound in the DMSP pathway.100 However, it is a chemically stable compound in the
absence of the degradative enzymes, so in principle it would be possible to accumulate
DMSHB without DMS emissions.
The W. biflora and S. alterniflora pathways differ only in the nature of the conversion
of SMM to DMSP-ald (Fig. 7.6). As with the other DMSP pathways, higher plant DMSP
biosynthesis has recruited steps from primary metabolism, in this case the SMM cycle (to
slightly stretch the definition of “primary metabolism”). In W. biflora the first step in the
pathway occurs in the cytosol (methionine→SMM). MMT is abundant in many non-DMSP
accumulators, so that this first step may not have to be engineered. However, as the next
steps in the pathway occur in the chloroplast, cytosolic SMM must be transported into the
chloroplast stroma.127 The transporter(s) that perform this function are unknown. SMM
import to the chloroplast is quite active in the DMSP accumulator W. biflora, particularly
under salinized conditions, while non accumulators have only residual SMM trans-
port function under any conditions.128 Thus, even if the next step in the pathway, the
transamination/decarboxylation of SMM to DMSP-ald,125 could be engineered to be
expressed in the chloroplast of non-DMSP accumultors, it is probable that insufficient
SMM substrate would be available. Further investigation of the SMM transporter in
DMSP and non-DMSP accumulating plants is definitely warranted. If this limitation
is overcome, the last step in the pathway may not require the introduction of DDH.
Both BADH and DDH appear to have less substrate specificity than expected. These
enzymes may both be related to the constitutive aldehyde dehydrogenases involved in
polyamine metabolism.
The relative advantages and disadvantages of engineering a particular DMSP pathway (or
genes from different pathways) will become apparent as we learn more about the biosynthetic
enzymes involved. These studies are only in the early stages.
Conclusion
The engineering of compatible osmolyte biosynthetic pathways into crop plants to
impart improved stress tolerance has long been an objective. There has been recent success
in the introduction of bacterial glycine betaine genes to higher plants to impart improved
stress resistance. Now that the plant genes are available, it will be interesting to compare
glycine betaine sythesis and the stress response of other plants transformed with these
plant genes. The engineering of DMSP accumuation will remain a challenging target for
the future. A great deal of fundamental biochemical investigation must precede any
applications in genetic engineering. The remarkable biosynthetic diverstiy for the production
of this compound is unprecedented and will offer many opportunities for genetic engineering.
That DMSP production has evolved so many times, suggests that it may be a very useful
molecule. Still, attempts to engineer DMSP accumulation will face the same hurdles that
confront glycine betaine engineering. How can substrate limitations be overome without
disrupting primary metabolism? It is clear that the engineering of compatible osmolytes
cannot be viewed in isolation from primary metabolism, particularly that involved in
methyl group metablolism, and in the case of DMSP, sulfur metabolism also. The
introduction of the genes for the biosythesis of compounds like glycine betaine and intro-
duction of the genese for the biosythesis of compounds like glycine betaines and DMSP
may not immediately produce stress-tolerant crops, but they will be useful tools for
understanding how some key areas in primary metabolism interact and are regulated.
Acknowledgments
This publication is Florida Agricultural Experiment Station Journal Series Number
R-06603 and support to B.R. from the College of Agriculture, University of Florida is
gratefully acknowledged.
References
1. Timasheff SN. A physiochemical basis for the selection of osmolytes by nature. In: Somero
GN, Osmond CB, Bolis CL, eds. Water and Life. Berlin:Springer Verlag, 1992:70-84.
2. Yancey PH. Compatible and counteracting solutes. In: Strange K, ed. Cellular and Molecular
Physiology of Cell Volume Regulation. Boca Raton: CRC Press, 1994:81-109.
3. Carpenter JF, Crowe JH. The mechanism of cryoprotection of proteins by solutes. Cryobiol
1988; 25:244-255.
4. Kirst GO. Osmotic adjustment in phytoplankton and macroalgae. The use of dimethyl-
sulfoniopropionate. In:Kiene RP, Visscher PT, Keller MD, Kirst GO, eds. Biological and
Environmental Chemistry of DMSP and Related Sulfonium Compounds. New York: Plenum
Press, 1996:121-129.
5. McCue KF, Hanson AD. Drought and salt tolerance: Towards understanding and applica-
tion. Trends Biotechnol 1990; 8:358-362.
6. Weretilnyk EA, Bednarek S, McCue KF, Rhodes D, Hanson AD. Comparative biochemical
and immunological studies of the glycine betaine synthesis pathway in diverse families of
dicotyledons. Planta 1989; 178:342-352.
7. Wyn Jones RG, Storey R. Betaines. In: Paleg LG, Aspinall D, eds. The Physiology and
Biochemistry of Drought Resistance in Plants. Sydney: Academic Press, 1981:171-204.
higher-plants. Ann Rev of Plant Physiol Mol Biol 1993; 44:357-384.
9. Dickson DJM, Kirst GO. The role of β-dimethylsulfoniopropionate, glycine betaine and
homarine in the osmoacclimation of Platymonas subcordiformis. Planta 1986; 167:536-543.
10. Dickson DJM, Kirst GO. Osmotic adjustment in marine eukaryotic algae: The role of
inorganic ions, quaternary ammonium, tertiary sulfonium and carbohydrate solutes. New
Phytol 1987; 106:657-666.
11. Reed RH. Measurement and osmotic significance of (-dimethylsulfoniopropionate in marine
macroalgae. Mar Biol Lett 1983; 4:173-181.
12. Paquet L, Rathinasabapathi B, Saini H, Zamir, L, Gage DA, Huang Z-H, Hanson, AD.
Accumulation of the compatible solute 3-dimethylsulfoniopropionate in sugarcane and
its relatives, but not other gramineous crops. Aust J Plant Physiol 1994; 21:37-48.
13. Storey R, Gorham J, Pitman MG, Hanson AD, Gage DA. Responses of Melanthera biflora
to salinity and water stress. J Exp bot 1993; 44:1551-1560.
14. Blunden G, Gordon SM. Betaines and their sulphonio analogues in marine algae. Prog
Phycol Res 1986; 4:39-80.
15. Csonka LN, Hanson AD. Prokaryotic osmoregulation: Genetics and physiology. Ann Rev
Microbiol 1991; 45:569-606.
16. Hanson AD. Compatible solute synthesis and compartmentation in higher plants. In: Somero
GN, Osmond CB, Bolis CL, eds. Water and Life. Berlin: Springer-Verlag,1992:52-60.
17. Hanson AD, Rivoal J, Burnet M, Rathinasabapathi B. Biosynthesis of quaternary ammonium
and tertiary sulphonium compounds in response to water deficit. In: Smirnoff N, ed.
Environment & Plant Metabolism. Oxford: Bios Scientific Publishers,1994:89-198.
18. Gorham J. Betaines in higher plants—biosynthesis and role in stress metabolism. In:
Wallsgrove RM, ed. Amino acids and their derivatives in higher plants. Cambridge University Press,
1995:173-203.
19. Miller KJ, Wood JM. Osmoadaptation by rhizosphere bacteria. Ann Rev Microbiol
1996; 50:101-136.
20. Hanson AD, Rathinasabapathi B, Rivoal J, Burnet M, Dillon MO, Gage DA. Osmo-protectant
compounds in the Plumbaginaceae: A natural experiment in metabolic engineering. Proc
Nat Acad Sci USA 1994; 91:306-310.
21. Papageorgiou GC, Murata N. The unusually strong stabilizing effects of glycine betaine
on the structure and function of the oxygen-evolving Photosystem II complex. Photo-
synthesis Res 1995; 44:243-252.
22. Hanson AD, Rathinasabapathi B, Chamberlin B, Gage DA. Comparative physiological
evidence that β-alanine betaine and choline-O-sulfate act as compatible osmolytes in halophytic
Limonium species. Plant Physiol 1991; 97:1199-1205.
23. Kirst GO, Thiel C, Wolff H, Nothnagel J, Wanzec M, Ulmke R. Dimethylsulfonio-propionate

(DMSP) in ice-algae and its possible biological role. Mar Chem 1991; 35:381-388.
24. Karsten U, Wienke C, Kirst GO. The β-dimethylsulfoniopropionate (DMSP) content of
macroalgae from Antarctica and southern Chile. Bot Mar 1990; 33:143-146.
25. Karsten U, Wienke C, Kirst GO. Dimethylsulfoniopropionate (DMSP) accumulation in
green macroalgae from polar to temperate regions: interactive effects of light versus salinity
and light versus temperature. Polar Biol 1992; 12:603-607.
26. Nishiguchi MK, Somero GN. Temperature and concentration dependence of compatibility
of the organic osmolyte dimethylsulfoniopropionate. Cryobiol 1992; 29:118-124.
27. Andersen PA, Kaasen I, Styrvold OB, Boulnois G, Strom AR. Molecular cloning, physical
mapping and expression of the bet genes governing the osmoregulatory choline-glycine
betaine pathway of Escherichia coli. J Gen Microbiol 1988; 134:737-1746.
28. Boch J, Kempf B, Bremer E. Synthesis of the osmoprotectant glycine betaine in Bacillus
subtilis: Characterization of the gbsAB genes. J Bacteriol 1996; 178:5121-5129.
29. Rhodes D, Rich PJ, Myers AC, Reuter CR, Jamieson GC. Determination of betaines by
fast atom bombardment mass spectrometry: Identification of glycine betaine deficient
genotypes of Zea mays. Plant Physiol 1987; 84:781-788.
30. Selvaraj G, Jain RK, Olson DJ, Hirji R, Jana S, Hogge LR. Glycine betaine in oilseed rape
and flax leaves: Detection by liquid chromatography/continuous flow secondary ion-mass
spectrometry. Phytochemistry 1995; 38:1143-1146.
31. Gorham J. Glycine betaine is a major nitrogen-containing solute in the Malvacceae.
Phytochemistry 1996; 43:367-369.
32. Keller MD, Bellows WK, Guillard RRL. Dimethylsulfide production in marine phy-
toplankton. In: Salzman ES, Cooper WJ, eds. Biogenic Sulfur in the Environment.
Washington, DC: American Chemical Society,1989:167-182.
33. Kiene RP. DMS production from DMSP in coastal seawater samples and bacterial cultures.
Appl Environ Microbiol 1990; 56:3292-3297.
34. de Souza MP, Yoch DC. N-terminal amino acid sequences and comparisons of DMSP
lyases from Pseudomonas doudorffii and Alcaligens strain M3A. In: Kiene RP, Visscher
PT, Keller MD, Kirst GO, eds. Biological and Environmental Chemistry of DMSP and
Related Sulfonium Compounds. New York: Plenum Press,1996:293-304.
35. Stefels J, Gieskes WWC, Dijkhuizen L. Intriguing functionality of the production and
conversion of DMSP in Phaeocystis sp. In: Kiene RP, Visscher PT, Keller MD, Kirst, GO,
eds. Biological and Environmental Chemistry of DMSP and Related Sulfonium Compounds.
New York: Plenum Press,1996:305-315.
36. Malin G. The role of DMSP and DMS in the global sulfur cycle and in climate regulation. In:
Kiene RP, Visscher PT, Keller MD, Kirst GO, eds. Biological and Environmental Chemistry of
DMSP and Related Sulfonium Compounds. New York: Plenum Press, 1996:77-189.
37. Wolfe GV, Steinke M, Kirst GO. Grazing-activated chemical defence in a unicellular marine
alga. Nature 1997; 387:894-896.
38. Hanson AD, Rivoal J, Paquet L, Gage DA. Biosynthesis of 3-dimethylsulfoniopropionate
in Wollastonia biflora (L.)DC. Evidence that S-methylmethionine is an intermediate. Plant
Physiol 1994; 105:103-110.
39. Larher F, Hamelin J, Stewart GR. L’acide dimethylsulfonium-3-propanoique de Spartina
angelica. Phytochemistry 1977; 16:2019-2020.
40. Dacey JWH, King GM, Wakeman SG. Factors controlling the emission of dimethylsulfide
from salt marshes. Nature 1987; 330:643-645.
41. Otte ML, Morris JT. Dimethylsulfoniopropionate (DMSP) in Spartina alterniflora Loisel.
Aquatic Bot 1994; 48:239-259.
42. Paquet L, Lafontaine PJ, Saini HS, James F, Hanson AD. Evidence en faveur de la presence
du 3-dimethylsulfoniopropionate chez une large gamme d’Angiospermes. Can J Bot 1995;
73:1889-1896.
43. White H. Analysis of dimethylsulfonium compounds in marine algae. J Marine Res 1982;
40:529-536.
44. Vogt C, Rabenstein A, Rethmeier J, Fischer U. Alkali-labile precursors of dimethyl sulfide in

marine benthic cyanobacteria. Arch Microbiol 1998; 169:263-266.
45. Gage DA, Hanson AD. Characterization of 3-dimethylsulfoniopropionate (DMSP) and
its analogs with mass spectrometry. In: Kiene R, Visscher PT, Keller MD, Kirst GO, ed.
Biological and Environmental Chemistry of DMSP and Related Sulfonium Compounds.
New York: Plenum, 1996:29-44.
46. Hitz WD, Hanson AD. Determination of glycine betaine by pyrolysis-gas chromatography in
cereals and grasses. Phytochemistry 1980; 19:2371-2374.
47. Rathinasabapathi B, Gage DA, Mackill DJ, Hanson AD. Cultivated and wild rices do not
accumulate glycine betaine due to deficiencies in two biosynthetic steps. Crop Sci 1993;
33:534-538.
48. Marcum KB, Murdoch CL. Salinity tolerance mechanisms of six C4 turfgrasses. J Am
Soc Hort Sci 1994; 119:779-784.
49. Ishitani M, Arakawa K, Mizuno K, Kishitani S, Takabe T. Betaine aldehyde dehydrogenase in
the Gramineae: Levels in leaves of both betaine-accumulating and nonaccumulating cereal
plants. Plant Cell Physiol 1993; 34:493-495.
50. Grote EM, Ejeta G, Rhodes D. Inheritance of glycine betaine deficiency in sorghum. Crop
Sci 1994; 34:1217-1220.
51. Lerma C, Rich PJ, Ju GC, Yang W-J, Hanson AD, Rhodes D. Betaine deficiency in maize:
Complementation tests and metabolic basis. Plant Physiol 1991; 95:1113-1119.
52. Yang W, Nadolska-Orczyk A, Wood KV, Hahn DT, Rich PJ, Wood AJ, Saneoka H,
Premachandra GS, Bonham CC, Rhodes JC, Joly RJ, Samaras Y, Goldsbrough PB,
Rhodes D. Near-isogenic lines of maize differing for glycine betaine. Plant Physiol
1995; 107:621-630.
53. Saneoka H, Nagasaka C, Hahn DT, Yang W, Premachandra GS, Joly RJ, Rhodes D. Salt
tolerance of glycine betaine-deficient and -containing maize lines. Plant Physiol 1995; 107:631-638.
54. Yang G, Rhodes D, Joly RJ. Effects of high temperature on membrane stability and chlorophyll
fluorescence in glycine betaine-deficient and glycine betaine-containing maize lines. Aust
J Plant Physiol 1996; 23:437-443.
55. Keller MD, Korjeff-Belows W. Physiological aspects of the production of dimethyl-
sulfoniopropionate (DMSP) by marine phytoplankton. In: Kiene RP, Visscher PT, Keller
MD, Kirst GO, eds. Biological and Environmental Chemistry of DMSP and Related Sulfonium
Compounds. New York: Plenum Press,1996:131-142.
56. Gröne T, Kirst GO. The effect of nitrogen deficiency, methionine, and inhibitors of methionine
metabolism on the DMSP content of Tetraselmis subcordiformis. Mar Biol 1992; 112:97-503.
57. Colmer TD, Fan TW-M, Lauchi A, Higashi RM. Interactive effects of salinity, nitrogen and
sulfur on the organic solutes in Spartina alterniflora leaf blades. J Exp Bot 1996; 47:369-375.
58. Lamark T, Rokenes TP, McDougall J, Strom AR. The complex bet promoters of Escherichia
coli: Regulation by oxygen (ArcA), choline (BetI), and osmotic stress. J Bacteriol 1996;
178:1655-1662.
59. Bernard T, Pocard JA, Perroud B, Le Rudulier D. Variations in the response of salt-
stressed Rhizobium strains to betaines. Arch Microbiol 1986; 143:359-364.
60. Tsuge H, Nakano Y, Onishi H, Futamura Y, Ohashi K. A novel purification and some properties
of rat liver mitochondrial choline dehydrogenase. Biochim Biophys Acta 1980; 614:274-284.
61. Haubrich DR, Gerber GH. Choline dehydrogenase: Assay, properties and inhibitors.
Biochem Pharmacol 1981; 30:2993-3000.
62. Zhang J, Blusztajn JK, Zeisel SH. Measurement of the formation of betaine aldehyde and
betaine in rat liver mitochondria by a high pressure liquid chromatography-radioenzymatic
assay. Biochim Biophys Acta 1992; 1117:333-339.
63. Loeffler M. NCBI sequence Accession number 1154950. 1994.
64. Wilson R. NCBI Accession number Z66494. 1995.
65. Boyd LA, Adam L, Pelcher LE, McHughen A, Hirji R, Selvaraj, G. Characterization of an
Escherichia coli gene encoding betaine aldehyde dehydrogenase (BADH): Structural similarity
to mammalian ALDHs and plant BADH. Gene 1991; 103:45-52.
66. Lamark T, Kaasen I, Eshoo MW, Falkenberg P, McDougall J, Strom AR. DNA sequence
and analysis of the bet genes encoding the osmoregulatory choline-glycine betaine pathway
of Escherichia coli. Mol Microbiol 1991; 5:1049-1064.
67. Pocard JA, Vincent N, Boncompagni E, Smith LT, Poggi MC, Le Rudulier D. Molecular
characterization of the bet genes encoding glycine betaine synthesis in Sinorhizobium
meliloti 102F34. Microbiol 1997; 43:1369-1379.
68. Ohta-Fukuyama M, Miyake Y, Emi S, Yamano T. Identification and properties of the
prosthetic group of choline oxidase from Alcaligenes sp. J Biochem 1980; 88:197-203.
69. Rozwadowski KL, Khachatourians GG, Selvaraj G. Choline oxidase, a catabolic enzyme
in Arthrobacter pascens, facilitates adaptation to osmotic stress in Escherichia coli. J
Bacteriol 1991; 173:472-478.
70. Deshinium P, Los DA, Hayashi H, Mustardy L, Murata N. Transformation of
Synechococcus with a gene for choline oxidase enhances tolerance to salt stress. Plant Mol
Biol 1995; 29:897-907.
71. Deshinium P, Gombos Z. Nishiyama, Y, Murata N. The action in vivo of glycine betaine
in enhancement of tolerance of Synechococcus sp. Strain PCC7942 to low temperature. J
Bacteriol 1997; 179:339-334.
72. Ladyman JAR, Hitz WD, Hanson AD. Translocation and metabolism of glycine betaine.
Planta 1980; 150:191-196.
73. Rathinasabapathi B, McCue KF, Gage DA, Hanson AD. Metabolic engineering of glycine
betaine synthesis: Plant betaine aldehyde dehydrogenases lacking typical transit peptides are
targeted to tobacco chloroplasts where they confer betaine aldehyde resistance. Planta
1994; 193:155-162.
74. Makela P, Sainio P, Jokinen K, Pehu E, Setala H, Hinkkanen R, Somersalo S. Uptake and
translocation of foliar-applied glycine betaine in crop plants. Plant Sci. 1996; 121:221-230.
75. Brouquisse R, Weigel P, Rhodes D, Yocum CF, Hanson AD. Evidence for a ferredoxin-dependent
choline monooxygenase from spinach chloroplast stroma. Plant Physiol 1989; 90:322-329.
76. Weigel P, Weretilnyk EA, Hanson AD. Betaine aldehyde oxidation by spinach chloroplasts.
Plant Physiol 1986; 82:753-759.
77. Hanson AD, May AM, Grumet R, Bode J, Jamieson GC, Rhodes D. Betaine synthesis in
chenopods: Localization in chloroplasts. Proc Nat Acad Sci USA 1985; 82:3678-3682.
78. Burnet M, Lafontaine PJ, Hanson AD. Assay, purification, and partial characterization of
choline monooxygenase from spinach chloroplast stroma. Plant Physiol 1995; 90:581-588.
79. Rathinasabapathi B, Burnet M, Russell BL, Gage DA, Liao P-C, Nye GJ, Scott P, Golbeck
JH, Hanson AD. Choline monooxygenase, an unusual iron-sulfur enzyme catalyzing the
first step of glycine betaine synthesis in plants: Prosthetic group characterization and
cDNA cloning. Proc Natl Acad Sci USA 1997; 94:3454-3458.
80. Russell BL, Rathinasabapathi B, Hanson AD. Osmotic stress induces expression of choline
monooxygenase in sugar beet and amaranth. Plant Physiol 1998; 116:859-865.
81. Arakawa K, Takabe T, Sugiyama T, Akazawa, T. Purification of betaine-aldehyde dehydrogenase
from spinach leaves and preparation of its antibody. J Biochem 1987; 101:1485-1488.
82. Weretilnyk EA, Hanson, AD. Betaine aldehyde dehydrogenase from spinach leaves: Purification, in
vitro translation of the mRNA, and regulation by salinity. Arch Biochem Biophys 1989;
271:56-63.
83. Weretilnyk EA, Hanson AD. Molecular cloning of a plant betaine-aldehyde dehydrogenase,
an enzyme implicated in adaptation to salinity and drought. Proc Natl Acad Sci USA
1990; 87:2745-2749.
84. McCue KF, Hanson AD. Effects of soil salinity on the expression of betaine aldehyde
dehydrogenase in leaves: Investigation on hydraulic, ionic and biochemical signals. Aust
J Plant Physiol 1992; 19:555-564.
85. Shu W, Ai W, Chen S. NCBI Accession number 1813538. 1995.
86. Xiao G, Zhang G, Liu F, Chen S. Study on BADH gene from Atriplex hortensis. Chin Sci
Bull 1995; 40:741-745.
87. Ishitani M, Nakamura T, Han SY, Takabe T. Expression of the betaine aldehyde dehy-
drogenase gene in barley in response to osmotic stress and abscisic acid. Plant Mol Biol
1995; 27:307-315.
88. Wood AJ, Saneoka H, Rhodes D, Joly RJ, Goldsbrough PB. Betaine aldehyde dehydrogenase
in sorghum: Molecular cloning and expression of two related genes. Plant Physiol 1996;
110:1301-1308.
89. Legaria, J, Iturriaga, G. NCBI Accession number 2388710. 1997.
90. Nakamura T, Yokota S, Muramoto Y, Tsutsui K, Oguri Y, Fukui K, Takabe T. Expression of
a betaine aldehyde dehydrogenase gene in rice, a glycine betaine nonaccumulator, and
possible localization of its protein in peroxisomes. Plant J 1997; 11:1115-1120.
91. Vojtechova M, Hanson AD, Munoz-Clares RA. Betaine aldehyde dehydrogenase from amaranth
leaves efficiently catalyzes the NAD-dependent oxidation of dimethyl-sulfoniopropionaldehyde
to dimethylsulfoniopionate. Arch Biochem Biophys 1997; 337:81-88.
92. Trossat C, Rathinasabapathi B, Hanson AD. Transgenically expressed betaine aldehyde
dehydrogenase efficiently catalyzes oxidation of dimethylsulfoniopropionaldehyde and
omega-aminoaldehydes. Plant Physiol 1997; 113:1457-1461.
93. Summers PS, Weretilyn EA. Choline synthesis in spinach in relation to salt stress. Plant
Physiol 1993; 103:1269-1276.
94. Weretilnyk EA, Smith DD, Wilch GA, Summers PS. Enzymes of choline synthesis in spinach.
Response of phospho-base N-methyltransferase activities to light and salinity. Plant Physiol
1995; 109:1085-1091.
95. Hitz WD, Rhodes D, Hanson AD. Radiotracer evidence implicating phosphoryl and phosphatidyl
bases as intermediates in betaine synthesis by water-stressed barley leaves. Plant Physiol
1981; 68:814-822.
96. Giddings TH Jr., Hanson AD. Water stress provokes a generalized increase in phosphatidylcholine
turnover in barley leaves. Planta 1982; 155:493-501.
97. Greene RC. Biosynthesis of dimethyl-β-propiothetin. J Biol Chem 1962; 237:2251-2254.
98. Pokorny M, Marcenko E, Keglevic D. Comparative studies of L- and D-methionine
metabolism in lower and higher plants. Phytochemistry 1970; 9:2175-2188.
99. Chillemi R, Patti A, Morrone R, Piatelli M, Sciutto S. The role of methylsulfonium
compounds in the biosynthesis of N-methylated metabolites in Chondria coerulescens. J
Nat Prod 1990; 53:87-93.
100. Gage DA, Rhodes D, Nolte KD, Hicks WA, Leustek T, Cooper AJL, Hanson, AD. A new
route for synthesis of dimethylsulphoniopropionate in marine algae. Nature 1997; 387:891-894.
101. Kocsis MG, Nolte KD, Rhodes D, Shen T-L, Gage DA, Hanson AD. Dimethyl-
sulfoniopropionate biosynthesis in Spartina alterniflora. Plant Physiol 1998; 117:273-281.
102. Maw GA. The biochemistry of sulfonium salts. In: Stirling CJM, Patai S, eds. The Chemistry
of the Sulfonium Group. Part 2. Chichester, UK: Wiley, 1981:703-770.
103. James F, Nolte KD, Hanson AD. Purification and properties of S-adenosyl-L-
methionine:L-methionine S-methyltransferase from Wollastonia biflora leaves. J Biol
Chem 1995; 270:22344-22350.
104. Uchida A, Ooguri, T, Ishida, T, Kitaguchi, H, Ishida, Y. Biosynthesis of dimethyl-
sulfoniopropionate in Crypthecodinium cohnii (Dinophyceae). In: Kiene RP, Visscher PT,
Keller MD, Kirst GO, eds. Biological and Environmental Chemistry of DMSP and
Related Sulfonium Compounds. New York: Plenum Press, 1996:97-107.
105. Uchida A, Ooguri, T, Ishida, T, Ishida, Y. Incorporation of methionine into dimethyl-
thiopropionic acid in the dinoflagellate Crypthecodinium cohnii. Nippon Suisan Gakkaishi
1993; 59:851-855.
106. Stevenson, DE, Akhtar, M, Gani, D. Streptomyces L-methionine decarboxylase: Purifica-
tion and properties of the enzyme and stereochemical course of substrate decarboxylation.
Biochemistry 1990; 29:7660-7666.
107. Stevenson DE, Akhtar M, Gani D. L-methionine decarboxylase from Dryopteris felix-
mas: Purification, characterization, substrate specificity, abortive transamination of the
coenzyme, and stereochemical courses of substrate decarboxylation and coenzyme
transamination. Biochem 1990; 29:7631-7647.
108. Mazelis M. The pyridoxal phosphate-dependent oxidative decarboxylation of methionine

by peroxide. I. Characteristics and properties of the reaction. J Biol Chem 1962; 237:104-108.
109. Mazelis M, Ingraham LL. The pyridoxal phosphate-dependent oxidative decarboxylation
of methionine by peroxide. II. Identification of 3-methylthiopropylamide as the product.
J Biol Chem 1962; 237:109-112.
110. Scislowski WD, Hokland BM, Davis van-Thienen WIA, Bremer J, Davis EJ. Methionine
metabolism by rat muscle and other organisms. Biochem J 1987; 247:35-40.
111. Walters DS, Steffens JC. Branched chain amino acid metabolism in the biosynthesis of
Lycopersicon pennellii glucose esters. Plant Physiol 1990; 93:1544-1551.
112. Livesy G. Methionine degradation: Anabolic and catabolic. Trends Biochem Sci 1984; 9: 27-29.
113. Dibner JJ, Knight CD. Conversion of 2-hydroxy-4-(methylthio)butanoic acid to L-methionine in
the chick: a stereospecific pathway. J Nutr 1984; 114:1716-1723.
114. Mizayaki HH, Yang SF. Metabolism of 5-methylthioribose to methionine. Plant Physiol
1987; 84:277-281.
115. Walsh C. Enzymatic Reaction Mechanisms. New York:Freeman. 1979.
116. Miflin BJ, Lea PJ. Amino acid metabolism. Ann Rev Plant Physiol 1977; 28:299-329.
117. Christen P, Metzler DE. Transaminases. New York:Wiley. 1985.
118. Summers PS, Nolte KD, Cooper AJL, Borgeas H, Leustek T, Rhodes D, Hanson AD.
Identification and stereospecificty of the first three enzymes of 3-dimethylsulfoniopro-
pionate biosynthesis in a chlorophyte alga. Plant Physiol 1998; 116:369-378.
119. Mudd SH, Datko AH. The S-methylmethionine cycle in Lemna paucicostata. Plant Physiol
1990; 93:623-630.
120. Giovanelli J, Mudd SH, Datko A. Sulfur amino acids in plants. In: Stumpf PK Conn EE,
eds. The Biochemistry of Plants, Vol. 5, Amino Acids and Derivatives. New York: Academic
Press, 1980:453-505.
121. James F, Paquet L, Sparace A, Gage DA, Hanson AD. Evidence implicating dimethyl-
sulfoniopropionaldehyde as an intermediate in dimethylsulfoniopropionate biosynthesis.
Plant Physiol 1995; 108:1439-1448.
122. Cooper ALJ, Hollander MM, Anders MW. Formation of highly reactive vinylglyoxylate
(2-oxo-3-butenoate) from amino acids with good leaving groups in the γ-position.
Toxicological implications and therapeutic potential. Biochem Pharmacol 1989;
38:3895-3901.
123. Toney MD, Hohenester E, Cowan SW, Jansonius JN. Dialkylglycine decarboxylase structure:
Bifunctional active site and alkali metal sites. Science 1993; 261:756-759.
124. Walton NJ, Peerless ACJ, Robins RJ, Rhodes MJC, Boswell HD, Robins DJ. Purification and
properties of putrescine N-methyltransferase from transformed roots of Datura stramonium
L. Planta 1994; 193:9-15.
125. Rhodes D, Gage DA, Cooper AJL, Hanson AD. S-methylmethionine conversion to
dimethylsulfoniopropionate: Evidence for an unusual transamination reaction. Plant
Physiol 1997; 115:1541-1548.
126. Kumar A, Altabella T, Taylor MA, Tiburico, AF. Recent advance in polyamine research.
Trends Plant Sci 1997; 2:124-130.
127. Trossat C, Nolte KD, Hanson AD. Evidence that the pathway of dimethylsulfoniopro-
pionate biosynthesis begins in the cytosol and ends in the chloroplast. Plant Physiol 1996;
111:965-973.
128. Trossat C, Rathinasabapathi B, Weretylnik EA, Shen T-L, Huan Z-H, Gage DA, Hanson
AD. Salinity promotes accumulation of dimethylsulfoniopropionate and its precursor
S-methylmethionine in chloroplasts. Plant Physiol 1998; 116:165-171.
129. Nomura M, Ishitani M, Takabe T, Rai AK, Takabe T. Synechococcus sp. PCC7942 transformed
with Escherichia coli bet genes produces glycine betaine from choline and acquires
resistance to salt stress. Plant Physiol 1995; 107:703-708.
130. Cossins, EA, Chen, L. Folates and one carbon metabolism in plants and fungi. Phytochemistry
1997; 45:437-452.
131. Holmstrom KO, Welin B, Mandal A, Kristiansdottir I, Teeri T, Lamark T, Strom, AR,
Palva ET. Production of the Escherichia coli betaine-aldehyde dehydrogenase, an enzyme
required for the synthesis of the osmoprotectant glycine betaine in transgenic plants.
Plant J 1994; 6:749-758.
132. Lilius G, Holmberg N, Bulow L. Enhanced NaCl stress tolerance in transgenic tobacco
expressing bacterial choline dehydrogenase. Biotechnol 1996; 14:177-180.
133. Holmberg N. Metabolic engineering: Approaches towards improved stress tolerance in
microorganisms and plants. Doctoral Dissertation. Lund, Sweden: University of Lund. 1996:77-92.
134. Hayashi H, Mustardy L, Deshnium P, Ida M, Murata N. Transformation of Arabidopsis
thaliana with the codA gene for choline oxidase; accumulation of glycinebetaine and enhanced
tolerance to salt and cold stress. Plant J 1997; 12:133-142.
135. Hitz WD, Ladyman JAR, Hanson AD. Betaine synthesis and accumulation in barley (Hordeum
vulgare) during field water-stress. Crop Sci 1982; 22:47-54.
136. Hanson AD, Gage DA. 3-Dimethylsulfoniopropionate biosynthesis and use by flowering plants. In:
Kiene RP, Visscher PT, Keller MD, Kirst GO, eds. Biological and Environmental Chemistry of
DMSP and Related Sulfonium Compounds. New York: Plenum Press, 1996:75-86.
137. Pokorny M, Marcenko E, Keglevic D. Comparative studies of L- and D-methionine
metabolism in lower and higher plants. Phytochemistry 1970; 9:2175-2188.
CHAPTER 8
Osmotic Stress Tolerance in Plants:

Role of Proline and Sulfur
Metabolisms
Desh Pal S. Verma
O smotic stress, caused either due to the loss of water or increase in soil salinity, reduces
growth and productivity of plants. The responses of plants to both drought and salinity
have many steps in common and some of them overlap with cold stress. An
increase in external osmolarity results in an efflux of water from the interior, which brings
about a reduction in the turgor pressure in the cell and a reduction in the cytoplasmic
volume. A decrease in cell volume elevates the concentration of various intracellular ions,
which are toxic to the cell. To prevent volume change and loss of water, organisms generally
increase the concentration of compatible solutes in response to osmotic stress. The major
compatible solutes are K+, proline, glutamate, and quaternary ammonium compounds. Some
other molecules that act as osmolytes include trehalose, glycerol, choline, s,s-dimethyl-
sulfoniumacetate, stachydrine (N,N-dimethylproline, proline betaine), β-butyrobetaine,
L-pipecolate, 5-hydroxyl-1-pipecolate, N,N-dimethylglycine, N-methylproline, glutamate
betaine and γ-aminobutyrate.1-3 Different organisms accumulate one or more of these
compounds in response to drought or salinity.
Accumulation of proline in response to osmotic stress is very common in many plants.4
Our recent data suggests that the primary role of proline in osmoprotection may not be
solely as an osmoregulatory osmolyte, but it also helps the cell to overcome oxidative stress.
Other known attributes of proline, such as protecting enzymes from denaturation,5 interacting
with membrane systems,6 regulating cytosolic acidity,7 scavenging free radicals,8 balancing the
ratio of NADH/NAD+,9 and acting as a energy source10 may be more important for the
overall health of the plant under osmotic stress. We have demonstrated that high levels of
endogenous proline help reduce free radicals generated during oxidative stress induced by
the osmotic stress.11 Furthermore, free radicals produced during the oxidative stress cause
serious damage to SH groups, oxidize cystine and methionine, which may impair the function
of proteins.12 We have shown that the regulation of sulfur metabolism can reduce oxidative
stress and thus help alleviate osmotic stress.
Osmoregulation in Microorganisms
Proline as an Osmoprotectant
Although potassium is the most prevalent cation that acts as a major osmolyte in
bacteria,13,14 accumulation of proline has been found to benefit many microorganisms in
sustaining osmotic stress. The role of proline in osmotic stress tolerance was deduced from
the observation that exogenously applied proline could alleviate the growth inhibition of
bacteria imposed by osmotic stress. Proline is taken up from the medium by an active
transport system, and often proline levels are proportional to the osmotic strength of the
medium.15
Csonka demonstrated that proline-overproducing mutants of Salmonella typhimurium
exhibit enhanced tolerance to osmotic stress.16 In bacteria, proline synthesis is controlled
by feedback regulation of the first enzyme of the proline biosynthesis pathway, γ-glutamyl
kinase. Mutations resulting in proline overproduction and enhanced osmotic tolerance
were shown to be located in the proB gene, encoding γ-glutamyl kinase. One of the most
pronounced osmotic stress tolerant mutants (proB74) had a single base pair change which
resulted in a 100-fold loss of sensitivity of g-glutamyl kinase to feedback inhibition by
proline.16,17 When this mutated gene was transferred from S. typhimurium to other
enteric bacteria, it showed an enhanced osmotic stress tolerance.18,19 These studies confirmed
that proline plays a crucial role in osmotolerance in bacteria.
Other Osmolytes
Betaines and glycinebetaine (N,N,N-trimethlglycine) are widely used as osmolytes
by bacteria.20 Most bacteria are unable to synthesize betaines and depend on the transport
of these compounds from their environment.1 The transport of betaines is mediated by
the proP and proU transport system, as used for proline.21- 23 It appears that other systems
for transport also exist1,24 because proP and proU double mutants still grow well on
glycine betaine though with a long lag in growth.
In yeast (Saccharomyces cerevisiae), glycerol seems to be the primary compatible
solute produced under osmotic stress. 25,26 Glycerol is synthesized in cytosol from
dihydroxyacetonephosphate, an intermediate of the glycolytic pathway catalyzed by an
NADH-dependent glycerol-3-phosphate dehydrogenase (GPDH) and a phosphatase,
respectively. The GPDH activity is enhanced several fold under osmotic stress.27 Some
high salt tolerant algae, e.g., Dunaliella, also accumulate glycerol as an osmoprotectant.
Trehalose is accumulated in many organisms and may serve as a non-osmoregulatory
protector in stress conditions.28
Osmosensing and Signal Transduction Machinery

The osmosensing machinery is well studied in E. coli and yeast. In E. coli, two proteins,
OmpF and OmpC, are involved in passive diffusion of small hydrophilic molecules across
the membrane.29 The expression of their genes, ompF and ompC, is affected in a reciprocal
manner by the osmolarity of the medium. As the osmolarity increases, the ompC gene is
preferentially activated, whereas a decrease in osmolarity results in the activation of the
ompF gene. The expression of these two genes is controlled at the transcriptional level by
OmpR and EvnZ proteins. The OmpR is a DNA binding protein and specifically interacts
with the promoter regions of the ompF and ompC genes.30,31 The EnvZ is a membrane
protein containing two membrane-spanning regions at its amino-terminus32 and functions as
an environmental sensor.33 It has been demonstrated that the EnvZ is autophosphorylated
at the histidine residue (His243) in the presence of ATP, and the phosphate group is then
transferred to the OmpR protein.29 Thus, EnvZ acts as a protein kinase. The phosphorylated
OmpR binds with DNA and controls the transcription of the cognate genes. The EnvZ and
OmpR phosphotransfer system has a resemblance to the family of two-component regulatory
systems involved in response to environmental stimuli.34 The osmosensing system must be
coupled with the induction of genes involved in the synthesis of specific osmolyte. The
actual mechanism for this, however, is not yet understood.
Osmotic Stress Tolerance in Plants: Role of Proline and Sulfur Metabolisms 155
In yeast (S. cerevisiae), a more complex signal transduction pathway exists that senses
the osmotic condition of the cell and induces physiological changes that lead to the synthesis of
glycerol, which acts as an osmolyte.35-37 Yeast cells can accumulate glycerol in cytoplasm up to
a concentration of 1 M.35 Several genes of the signal transduction cascade have recently
been isolated from yeast. The mechanism involved in osmosensing is briefly summarized
below.
The primary sensor of osmotic stress, Sln1, contains a classical “input-transmitter-receiver”
structure and shares sequence similarities to both the histidine kinase and the response
regulator protein of the prokaryotic two-component systems.38 The transmitter domain
(histine kinase) of Sln1 is similar to the histidine kinase modules of the bacterial regulators.
Another similarity between Sln1 and bacterial sensor proteins is the presence of two
hydrophobic transmembrane domains in the N-terminal region. This region also contains
a membrane-bound signal-sensing domain which may be exposed on the cell surface. Several
suppressor mutants have been isolated from either sln1∆ or slnts strains. One of the
mutants, ssk1∆, suppresses sln1∆ lethality. The deduced amino acid sequence of Ssk1 shows
homology to the bacterial response regulators. Mutagenesis studies have shown that the
unphosphorylated Ssk1 is functional.36 A new protein, Ypd1p, was recently isolated and
shown to be a constituent of the two component system.37 The mutant of ypd1 is lethal but
it can be suppressed by the overexpression of Ptp2 (tyrosine phosphatase). The ypd1 gene
encodes a protein of 167 amino acids, which has similarity with the chemotactic CheA
protein of bacteria.
Early events of osmosensing start from the autophosphorylation of Sln1 at a His residue
followed by a cascade of phosphorylation steps. The downstream osmotic signal transduction
pathway is composed of three tiers of protein kinases, namely SSK2 and MAPKKKs (MAP
kinase kinase kinases), MAPKK (MAP kinase kinase) and high osmolarity glycerol (HOG)
response MAP (mitogen-activated protein) kinases39,40 (see chapter 2). SSK2 was found to
act as an extragenic suppressor of the sln1D mutant.40 It has high homology with MAPKKK
at the -COOH terminal region. Hog1 is not phosphorylated in pbs2-∆1 cells, suggesting
that phosphorylation of Hog1 requires Pbs2.
Both Ssk2p and Ssk22p interact with Ssk1p and HOG1 as shown by two hybrid analysis.
The phosphorylated Ssk1p is non-functional. HOG1 and HOG4 genes were cloned by
complementing yeast osmoregulation-defective mutants Osms. These mutants grow well
on YEPD medium but not on high-osmolarity medium and show reduced accumulation
of glycerol. The HOG1 sequence contains a single large open reading frame encoding a protein
of 416 amino aids with a molecular weight of 47 kDa. Near the NH2-terminus of the predicted
amino acid sequence of Hog1, a stretch contains the most conserved amino acids found in
this family of protein kinases. Two residues, corresponding to Thr174 and Tyr176, in
comparable positions in the MAP kinases encoded by ERK2 and FUS3 are phosphory-
lated in response to extracellular signals. Pbs2 is also a member of the MAP kinase kinase
gene family. The mutant pbs2 is unable to grow in medium of high osmolarity.
Glycerol synthesis in yeast is limited by the activity of glycerol-3-phosphate dehydroge-
nase (GPD1).27 gpd1∆ mutants produce little glycerol and are sensitive to osmotic stress. As
expected, hog1∆ mutant fails to increase glycerol-3-phosphate dehydrogenase activity. Thus,
the expression of GPD1 appears to be regulated through the HOG pathway. However,
there may be Hog1-independent mechanisms mediating osmotic stress-induced glycerol
accumulation in yeast, since a hog1∆ mutant still shows glycerol accumulation during
osmotic stress, although at a reduced level. The gpd1∆ and hog1∆ double mutants are more
sensitive to osmotic stress than gpd1∆ mutant.
By screening high osmolarity sensitive mutants, an SHO1 gene was identified,37 which
can be rescued by transformation either SSK2 or SSK22. SHO1 encodes a protein of 367
amino acids with four hydrophobic transmembrane domains at the N-terminal region.
The COOH terminal of the SHO1 contains an SH3 domain. The triple mutations of ssk2∆,
ssk22∆, and sho1∆ completely abolish tyrosine phosphorylation of Hog1p and osmotic stress
sensitivity. Sho1p appears to be a component of an alternate pathway that activates Pbs2p
in response to high osmolarity. The HOG pathway thus controls the expression of genes
involved in glycerol biosynthesis.
Two mammalian genes involved in osmoregulation have recently been identified. These
genes, encoding p38 and Jnkl, have high homology with HOG1 and belong to the MAP
kinase gene family. Both genes can complement hog1∆ mutants.41,42 These results suggest
that the signal transduction machinery may be conserved in eukaryotes. We have recently
isolated a plant homolog of the HOG1 gene (unpublished data). However, this gene is
unable to complement the hog1∆ mutant of yeast. The plants have a large MAP kinase gene
family.43 Recently, Jonak et al (ref. 43a). reported that a specific MAP kinase p44MMK4 is
found to be activated by drought and cold stress but not by salt stress. Furthermore,
the transcription of this gene is not induced by ABA. Recent studies from Zhu’s laboratory,44
have demonstrated the presence of ABA-dependent and ABA-independent pathways that
overlap in the transduction of cold and osmotic stress signals in plants.
Osmotic Stress Tolerance in Plants

Plants have evolved a variety of adaptations to water deficit and high salinity. These
adaptations include45 developmental and structural traits, time of flowering, rooting patterns,
leaf waxiness, and physiological mechanisms such as the ability to exclude salt or the
compartmentalization of ions within the cell.46 The biochemical traits, synthesis and
accumulation of compatible osmolytes and changes in patterns of carbon and nitrogen
metabolism are most important. Among the most prevalent osmolytes are proline and
betaines.47 These, however, appear to be non-osmoregulatory osmolytes since the con-
centration of these compounds in most plants is not very high for them to significantly
affect the osmoticum of the cell.
Regulation of Proline Biosynthesis in Plants

The biosynthesis pathways for proline and betaine have been well studied.48,49 In
plants, proline is synthesized from glutamate and ornithine.4 We have demonstrated that
proline biosynthesis from glutamate is accomplished in two major steps and have isolated
both gene encoding ∆1-pyrroline-5-caroxylate (P5C) reductase (P5CR) and P5C synthetase
(P5CS). The latter is a bifunctional enzyme which catalyzes the first two steps in proline
biosynthesis from glutamate; in E. coli these steps are catalyzed by two different enzymes
encoded by two separate genes (proB and proA). In addition, we have demonstrated that
nitrogen flux to proline is tightly regulated.50 Under normal conditions, ornithine is the
primary source of nitrogen for proline synthesis while glutamate takes precedent over
ornithine under stress conditions. Accordingly, the genes encoding ornithine amino
transferase (OAT) and P5CS are reciprocally regulated.50 Furthermore, we have demonstrated
that the accumulation of proline is tightly controlled by the end product of the pathway
when it is synthesized from glutamate and degraded by proline dehydrogenase (PDH).11,51, 52 A
proline cycle is thus established (Fig. 8.1).48
In order to determine the relationship between proline concentration and the levels
of P5CS and PDH gene expression, we first isolated a PDH gene from Arabidopsis.52 The
mRNA levels of P5CS and PDH were measured along with the free proline contents.
Under osmotic stress, the P5CS transcription is significantly induced and results in high
levels of proline synthesis. When osmotic stress is released, P5CS transcript declines to the
normal levels, but at the same time the PDH transcript significantly increases. Consequently,
Fig. 8.1. Proline cycle in plants. Increase in the synthesis of proline during stress conditions and
its degradation to glutamate during recovery from stress maintains a contenuous flux of nitrogen
and energy to the plant. The enzymes involved in the synthesis are P5CS, catalysing step 4; P5CR,
catalysing step 6. For the degradation of proline, the enzymes involved are proline dehydrogenase
for step 1 and P5C dehydrogenase for step 3. Reactions 2 and 5 are the same and are autocatalytic.
the free proline level declines following the induction of the PDH gene. The inhibition of
PDH by osmotic stress, both in bacteria and plants, 53,54 may be a key factor in the
accumulation of proline under stress conditions.
Proline Induces Proline-Degradation, but Under Osmotic Stress Conditions

this Degradation is Inhibited
In yeast and bacteria, the proline oxidase is induced by adding proline to the medium.55
The bacterial PutA protein not only has proline dehydrogenase and P5C dehydrogenase
activities, it also functions as a repressor of the put operon.55 When proline is provided, the
PutA protein binds with proline. The transcription of the put operon is induced and proline
degradation starts.56,57 In plants, our results indicate that PDH transcription is significantly
induced by exogenously applied proline. However, the level of PDH mRNA induced by
proline was reduced during saline stress. During stress, as the free proline concentration
increases, the PDH expression remains at a low level.52 Conversely, the transcription of
PDH is significantly induced within two hours upon removal of the osmotic stress. The
proline concentration has been shown to reach over 125 mM in NaCl-adapted tobacco
cells.58 The above results suggest that a transcriptional mechanism may be involved in
order to prevent proline-induced proline degradation during osmotic stress. This mechanism
may avoid the futile cycling and would precisely control the level of enzymes needed for
degradation of proline in response to osmotic stress to ensure proline accumulation.
Proline Cycle and the Role of Proline Degradation During Recovery of Plants
from Stress
Proline is one of the few amino acids that can be used as a sole source of carbon and
nitrogen, and as shown above, the synthesis and degradation of proline is tightly regulated. In
E. coli, PDH couples proline oxidation to the reduction of FAD cofactor, which is bound to
the PutA protein and thus delivers electrons to the membrane-associated electron transport
chain.56,57 Removal of FAD retains P5C dehydrogenase activity but not the PDH activity
of the PutA protein. Plant PDH is also a flavoprotein located in the inner membrane of
mitochondria,54 suggesting that proline oxidation donates electrons to the respiratory electron
transport chain and thus provides energy during recovery from stress. This was also suggested
from the studies on root nodules.9 In soybean root nodules, the proline concentration is
very high and it has been proposed that the NADP+ generated during proline synthesis
provides cofactor for the synthesis of the purine precursor, ribose 5-phosphate, and thus
regulating the synthesis of purines.9 Proline is oxidized by bacteroids which provide much
required energy for symbiotic nitrogen fixation.
In yeast, it has been reported that the proline degradation is regulated by nitrogen
sources and both PUT1 and PUT2 genes have been shown to be regulated by nitrogen. 55,59,60 In
maize, proline degradation was shown to be inhibited to 25% of the control levels in
water-stressed mitochondria. However, only PDH was inhibited, while P5C dehydrogenase
was not affected.54 Our results indicate that exogenously applied proline significantly
induces PDH expression. Removal of osmotic stress also induced PDH expression, resulting in
the decline of proline levels. These data indicate that the PDH expression level is directly
correlated with the increase in proline degradation, suggesting that proline dehydrogenase
is the rate limiting enzyme in the proline degradation pathway. Since accumulation of
proline is not deleterious to the cell, it acts as a reserve for nitrogen and energy and may
help plant recover rapidly from stress.52 The proline cycle (Fig. 8.1) thus helps maintain an
optimum flow of nitrogen and energy in the cell.
Accumulation of Other Osmolytes

Accumulation of Sugar Alcohol
Sugar alcohols, also called polyols or polyalcohol, have many hydroxyl groups, and
could possibly take the place of water in biopolymers of cell cytoplasm and help maintain
the function of enzymes and membranes at a time when water is limited due to osmotic
stress.61,62 Furthermore, sugar alcohols are inert and harmless to enzymes even at very
high concentrations.63 In bacteria and animals, the function of sugar alcohols in osmoregulation is
well studied. In plants, sugar alcohols are not as common as proline, glycine betaine, and
choline-0-sulfate. Mannitol-1- phosphate deyhdrogenase is an enzyme that regulates mannitol
synthesis in bacteria but plants do not normally produce and accumulate mannitol. The
expression of the bacterial mt1D gene results in the synthesis and accumulation of mannitol in
transgenic plants.64 The mannitol concentration in leaves and roots of the transformed
plants was estimated to be at a level of 100 mM. When the control and transgenic plants
were subjected to salinity stress, the plants producing mannitol were found to have
an increased ability to tolerate high salinity. These results clearly demonstrated that
overproducing sugar alcohols also helps plants deal with high salinity stress.
Betaines
Many organic molecules such as β-alanine betaine, proline betaine, and hydroxyproline
betaine, act as effective non-osmoregulatory osmolytes in plants.3,65 Different osmolytes
appear to have different selective advantages in a particular stress environment and in a
given species. Proline betaine and hydroxyproline betaine are found in plants adapted to
very dry or saline environments such as species of Citrus.66 It has been demonstrated that
proline betaine and hydroxyproline betaine are much better osmoprotectants than proline
(our unpublished data), particularly in chronically dry environments.3 The pathway
involved in the synthesis of proline betaine has, however, not yet been worked out and no
gene involved in this pathway has so far been identified. Once proline methyltransferases
are isolated and characterized, their genes can be used for plant transformation. Expression of
these genes in plants producing high levels of proline may convert some of this excess
proline into dimethyl proline and thus provide excellent protection from osmotic stress.
Sulfur Osmolytes
Choline-0-sulfate is another important osmolyte in plants.65,66 Choline-0-sulfate is
formed from choline. The choline sulfotransferase has not yet been identified in plants,
but this enzyme has been identified in both fungi and bacteria. The sulfate salinity leads to
higher choline-0-sulfate accumulation.3 With an increase in choline-0-sulfate, a proportional
decrease in glycine betaine level is observed.2 In some green alga, red algae and brown
algae and in the higher plants, e.g.,Wedelia biflora,65 β-dimethyl sulfoniopropioate (DMSP),
a sulfur osmolyte, accumulates. DMSP originates from methionine via methylation. The
enzyme S-adenosylmethionine:methionone S-methyltransferrase, catalyzing the first step
of DMSP synthesis, has recently been purified. 67 It appears that the regulation of
nitrogen and sulfur metabolisms may influence the type of osmolytes which a given
species accumulates.
Accumulation of Proline in Transgenic Plants Expressing Elevated

Levels of P5CS
We have earlier demonstrated that P5CS is the rate limiting enzyme in the proline
biosynthesis pathway and the level of P5CR has no effect on proline synthesis.68,69 We
introduced a Vigna P5CS cDNA under the control of 35S promoter in tobacco plants and
assessed the level of proline produced. Transgenic lines that produced high levels of Vigna
P5CS mRNA and proteins were analyzed and the effect of proline accumulation on plant
tolerance to water stress was assessed. Plants expressing high levels of P5CS accumulated
high levels of proline.69 Transgenic lines that did not produce Vigna P5CS mRNA nor
P5CS protein were found to have as low levels of proline as the control plants. Proline
levels in leaves of 10 transgenic lines ranged from 830 to 1590 µg/g fresh weight of leaves
(on average:1100 µg/g), compared to 80 to 89 µg/g in control plants. Corresponding to the
level of expression of P5CS, proline contents were increased in transgenic lines. A direct
correlation between P5CS expression level and proline accumulation in the transgenic
plants confirms that P5C synthetase is rate limiting in proline biosynthesis.
Data on amino acid analysis showed that accumulation of proline occurs at the
expense of glutamate.69 This indicates that availability of glutamate may act as a factor for
proline overproduction under water or salt stress. This conclusion is consistent with our
observation that transgenic plants expressing high levels of P5CS and supplied with
20 m M NH4NO3 produced nearly 25 times more proline than did the pBI121 controls.69
A recent study from B. Hirel’s group showed (unpublished data) that antisense of GS in
phloem cells reduces the level of proline, rendering plants sensitive to osmotic stress. Thus,
proline synthesis is tightly coupled with nitrogen assimilation. Furthermore, the source of
nitrogen for proline is also regulated, as was observed by our studies on OAT expression.4
Fig. 8.2. Effect of free radicals-induced damage due to osmotic stress as measured by malondialdehyde
(MDA) production, a measure of lipid peroxidation reaction. (A) Effect of externally added proline
on the level of MDA produced in transgenic tobacco cell lines after treatment with 250 mM NaCl for
8 h. (B) Effect of endogenously accumulated proline on MDA content in 14 day old seedlings of wild
type, P5CS and P5CS129A seedlings.
Feedback Inhibition of Proline Biosynthesis

Earlier experiments suggested that proline accumulation in plants under stress may
involve the loss of feedback regulation due to a conformational change in the P5CS
protein.70,71 In bacteria, proline biosynthesis has been shown to be regulated by the end
product inhibition of γ-GK activity.72 A Salmonella typhimurium mutant resistant to a toxic
proline analog accumulated proline and showed enhanced tolerance to osmotic stress.73
The mutation was due to the change of an aspartate (at position 107) to asparagine in the
γ-GK, resulting in a mutant γ-GK which was much less sensitive to proline inhibition.16,74
Our experiments revealed that the conserved aspartate residue (at position 128) in the
Vigna P5CS is not involved in the feedback inhibition. Using site-directed mutagenesis, a
replacement of phenylalanine at position 129 by alanine in Vigna P5CS (P5CSF129A) was
created. This mutant enzyme was shown to retain similar kinetic characteristics as the
wild type P5CS, but its feedback inhibition was virtually eliminated.51
We compared plants overexpressing a wild type form of Vigna P5CS and the
mutant enzyme P5CSF129A, whose feedback inhibition was eliminated. These two groups
of transgenic plants expressed comparable levels of Vigna P5CS mRNA and proteins as
revealed by Northern and Western blot analyses. Under normal conditions, P5CSF129A
A B C
Fig. 8.3. Overexpression of Vigna P5CS and its mutagenized derivative devoid of feedback
inhibition by proline (F129A), and the effect of overproduction of proline on the ability of
trangennic tobacco seeds to germinate on 200 mM NaCl. Pictures are taken 20 days after
germination. WT, wild type control seeds. For details, see ref. 10.
plants accumulated about two-fold more proline over the plants expressing Vigna wild type
P5CS.10 This difference was further increased in plants under salinity stress,11 demonstrating
that feedback regulation of P5CS does play a role in controlling the level of proline in
plants in both normal and stress conditions.
Proline Accumulation Confers Osmoprotection

Proline levels are increased in both control and transgenic plants after drought treatment.
These values increased from about 80 µg proline/g fresh leaf (before stress) to about 3000
µg/g (after stress) in control (wild type and pBI121) plants, and from 1000 µg/g to an
average of 6500 µg/g in transgenic lines.69 While the proline contents were approximately
14-fold greater in transgenic lines than in control plants before stress, only about 2-fold
greater after stress. Control plants started wilting in 5-6 days after drought treatment, while
wilting was delayed by at least two to three days in transgenic plants and the wilting was
more severe in controls than in transgenic plants. High constitutive levels of proline in
P5CS transgenic lines may be responsible for the observed effect on wilting rather than the
induced level of proline. Shinozaki’s group has recently constructed P5CS antisense
Aribidopsis plants and has demonstrated (personal communication) that the transgenic
plants with reduced P5CS are more sensitive to osmotic stress suggesting a direct role of
proline in osmoprotective machinery in plants.
The elevated levels of proline, either from exogenous addition in the medium or
produced endogenously effectively reduces free radical levels caused by salinity stress
(Fig. 8.2). This was determined by measuring the levels of melondialdehyde (MDA), a
marker for lipid peroxidation and membrane damage due to free radicals. The results
of externally added proline to suspension culture cells and the transgenic seedlings
overproducing proline due to P5CS were very similar. 10 These data clearly show that
proline confers a significantly increased ability for the transgenic seedlings to grow in media
containing NaCl (Fig. 8.3). These findings shed new light on the regulation of proline
biosynthesis in plants and its role in reducing oxidative damage conferred by the osmotic
stress. Figure 8.4 summarizes different roles of proline in plan metabolism and opens the
possibility of improving crops for stress tolerance through genetic engineering for proline
synthesis.
Fig. 8.4. Possible roles of proline in protecting plants from osmotic stress.
The other roles, besides being an osmolyte, appear to be more important
for reducing osmotic stress-induced oxidative stress.
Role of Sulfur Metabolism in Osmotic Stress Tolerance

Proteins, sulfated polysaccharides, sulfolipids, coenzymes and other sulfur-containing
secondary compounds are actively involved in cellular metabolism. Sulfate must be activated
in order to be used in cellular metabolism. This activation is achieved by the enzymes ATP-
sulfurylase forming adenosine 5'-phosphosulfate (APS) and the APS-kinase forming
phosphoadenosine 5'-phosphosulfate (PAPS).75 The PAPS is further reduced to sulfite and
sulfide, which contribute sulfur to cysteine and methionine. Some of the intermediates
and their derivatives in the sulfur assimilation pathway, such as sulfite,75 are toxic to the
cell when accumulated beyond certain levels. Under oxidative stress imposed by osmotic
stress, the SH groups of proteins are likely to be oxidized, which may impair the function
of many proteins.12 To avoid accumulation of the sulfur compounds to toxic levels, a sensitive
flux (rate of flow) of the intermediates in the sulfur assimilation pathway is necessary. A
3'(2'), 5'- diphosphonucleoside 3'(2')-phosphohydrolase (DPNPase) catalyzes the conversion
of PAPS to APS in vitro. This enzyme is widely distributed from algae to plants and suggests the
presence of a futile cycle (substrate cycle). The futile cycle is one of the most sensitive flux
control systems in metabolic pathways.
In yeast, a halotolerance gene (HAL2) was identified by functional assay of supporting the
growth of cells under high salinity stress.76 The yeast HAL2 gene is identical to MET22.
The yeast met22 mutant is a methionine auxotroph and can not use sulfate, sulfite or sulfide as
sulfur sources. However, the mutant exhibits wild type activities of the enzymes necessary to
assimilate sulfate and has a normal sulfur uptake system. HAL2 also shows high homology
with the E. coli cysQ gene.77 The cysQ mutant is a cysteine auxotroph but mutations which
resulted in sulfate transport defects compensated for cysQ mutation. Over expression of
HAL2 in yeast improved salt tolerance.74
The protein encoded by yeast HAL2 gene is shown to have the activity of 3'(2'),5'-bisphosphate
nucleotidase (DPNPase) with both PAPS and PAP as substrates.78 We have isolated a plant
homolog of HAL2 gene from rice (RHL) and enzymatic studies on the expressed protein
Fig. 8.5. The futile sulfur cycle and the role of DPNPase in the control of this cycle, leading to the
regulation of thr flux of active sulfur. For details, see ref. 52.
confirmed that it encodes a DPNPase. The RHL cDNA complemented both cysQ and met22
mutations. Thus, we demonstrated that the proteins encoded by cysQ, HAL2 and RHL genes
have the same function in sulfur assimilation pathways in E. coli, yeast and plants. This
enzyme, together with APS kinase, appears to catalyze a futile cycle in sulfur assimilation
pathway (Fig. 8.5).
Similar to the Chlorella enzyme, the rice DPNPase is inhibited by Ca2+ and has optimal at
9. The optimum Mg2+ concentration of DPNPase is about 2.5 mM. Since the RHL enzyme
and the Chlorella DPNPase have the same substrate specificity, similar kinetics and both
depend on Mg2+ and, inhibited by Ca2+, we believe the two are the same enzyme. DPNPase,
together with APS sulfotransferase, transfers sulfur from PAPS to a thio carrier which is
further reduced. The isolation of the RHL gene and the complementation of yeast HAL2
and E. coli cysQ mutants provided direct evidence that DPNPase is involved in sulfur
reduction in plants.
The assimilation of sulfur starts with sulfur activation. The first enzyme, ATP-
sulfurylase, catalyzes APS synthesis. Since the equilibrium for APS formation is far to the
left, the second reaction catalyzed by APS kinase, to phosphorylate APS at the 3' position,
plays an important role in pulling the first reaction forward. [Consequently, the PAPS accumulates
and an enzyme that controls the PAPS pool and removes unnecessary PAPS]. DPNPase
performs this function. Murguía et al78 suggested that PAPS is the substrate for the yeast
HAL2 encoded enzyme and the function of the HAL2 gene product is to remove PAP. This
hypothesis explains some phenotypes of the HAL2 mutant, met22; however, it does not
explain several facts.52
The phenotypes of cysQ and met22 can be explained if PAPS is the native substrate.
The cysQ gene product converts PAPS to APS and APS kinase catalyzes APS to PAPS. The
two enzymes run a futile cycle in the sulfur activation pathway (Fig. 8.5). In the cysQ mutant,
the PAPS accumulates immediately which is toxic to the cell.76 When cysteine is provided to E coli, the
sulfate uptake system is inhibited by feedback regulation and the cell stops synthesizing the
enzymes involved in the sulfur assimilatory pathway. Therefore, the toxic PAPS is not accumulated
and the cysQ mutant grows normally. When sulfite is provided, it can be easily converted to
cysteine without producing PAPS. The same principles may apply in yeast.
It has been demonstrated that fructose 1,6-biphosphatase (FBPase) and DPNPase
belong to the same structural protein family.80 FBPase is an allosterically regulated
enzyme which controls the flux of glycolysis by forming a “futile cycle” with phosphofructokinase
(PFK). Furthermore, FBPase is also regulated by cellular redox and the oxidized form is
inactive.81 If DPNPase is also regulated by cellular redox, the free radicals generated by
osmotic stress will render the enzyme inactive and thus disturb the balance of the sulfur
assimilation pathway. Consequently, overexpression of the HAL2 gene may help the cell
restore the sulfur flux under stress conditions and confer salt tolerance. In addition, the
cations accumulated during osmotic stress may also inhibit the DPNPase enzyme activity,
and overexpression of the HAL2 gene may overcome this problem.
The product of RHL gene (DPNPase) converts PAPS to APS controlling the sulfur
flux and thus may reduce the accumulation of the toxic compounds. Gläser et al (1993)
reported that supplying methionine improved salt tolerance in yeast. This phenotype could
be due to the fact that availability of methionine inhibits sulfate uptake.82 Sequestering
toxic sulfur compounds may be a common phenomenon under salt stress. In plants,
choline-O-sulfate (an osmolyte) may sequester extra sulfate under stress conditions.3
Although the transcription of the DPNPase gene is not increased by salt, the activity
of the DPNPase enzyme and the yeast HAL2 enzyme are increased by K+. Osmotic stress
increases the K+ uptake.82 This indicates that the RHL enzyme may response to salt stress
at the protein level. Overexpression of RHL in plants produces more glutathione, rendering
plant cells more resistant to heavy metal ions. These cells also produce less free radicals.
The latter may be the primary reason why HAL2 overexpression shows osmotic protection by
lowering oxidative damage, as does the proline.
A Possible Role of DPNPase in Salt Tolerance

Since our results have demonstrated that salt stress causes free oxygen radical production
and oxidative damage, it is likely that salt stress causes some of the HAL2 enzyme to lose
activity due to the oxidation of protein. The decrease in DPNPase activity disturbs the
sulfur metabolism which is required to remove oxidative stress. Because the DPNPase is
sensitive to Li+ and Na+, Murguia et al,78 proposed that Ha12 loses enzyme activity during
salt stress due to the increase of cellular Li+ or Na+ and overexpression of DPNPase overcomes
the problem of sulfur reduction.
Considering that the sulfur-rich compounds play significant roles in antioxidation
stress and that overexpression of the HAL2 gene in yeast conferred osmotolerance, the role
of DPNPase in controlling the “futile cycle” in sulfur assimilation pathways is important.
Overexpression of Plant HAL2 Gene Confers Reduction in Free

Radical Production and in Heavy Metal Toxicity
Plants exposed to various oxidative stresses including heavy metal stress, exhibit an
increase in lipid peroxidation due to excessive free radical generation.83 Interaction of heavy
metals with functional -SH groups has been proposed as the mechanism of inhibition of
several physiological reactions.84,85 A rapid decline in cellular glutathione (GSH) levels
was shown in plant cells exposed to cadmium.86,87 It has also been proposed that glutathione
may be of significance in the protection of plants against the harmful effects of active
oxygen species and free radicals.88 Exposure of plants to excess Cd particularly enhances
the demand for organic sulfur and even sulfide.89 The metal chelating phytochelatins are
synthesized from glutathione by plants exposed to metals such as Cd2+, Cu2+ and Zn2+.90
The studies on the overexpression of P5CS, mutated P5CS and HAL2 have clearly
demonstrated that metabolic engineering to produce specific compounds is now a reality.
For ensuring a continuous supply of nitrogen for proline synthesis a multiple gene cassette
containing constitutive GS, and stress-inducible P5CS may be necessary for producing
optimum levels of proline. Coexpression of HAL2 and superoxide dismutase may be
complementory in reducing drought or salt stress. A significant reduction in free radical
formation may, however, be deleterious from the point of pathogen attack.
References
1. Csonka LN. Physiological and genetic responses of bacteria to osmotic stress. Microbiological
Reviews. 1989; 51:121-147.
2. Hanson AD, Rathinasabapathi B, Chamberlin B, Gage DA. Comparative physiological
evidence that β-alanine betaine and choline-O-sulfate act as compatible osmolytes in
halophytic Limonium species. Plant Physiol. 1991; 97,1199-1205.
3. Hanson AD, Rathinasabapathi B, Rivoal J, Burnet M, Dillon MO, Gage DA.
Osmoprotective compounds in the plumbaginaceae: A natural experiment in metabolic
engineering of stress tolerance. Proc Natl Acad Sci USA 1994; 29,306-310.
4. Delauney AJ, Verma DPS. Proline biosynthesis and osmoregulation in plants. The Plant
Journal. 1993; 4,215-223
5. Rajendrakumar CS, Reddy BV, Reddy AR. Proline-protein interaction: Protection of
structural and functional integrity of M4 lactate dehydrogenase. Biochem Biophys Res
Commun. 1994; 201,957-963.
6. Rudolph AS, Crowe, JH, Crowe LM. Effects of three stablizing agents-proline, betaine,
and trehalose on membrane phospholipids. Arch Bioche Biophys. 1986; 245,134-143.
7. Venekamp JH. Regulation of cytosol acidity in plants under conditions of drought. Physiol
Plantarum. 1989; 76,112-117.
8. Smirnof N, Cumbes QJ. Hydroxyl radical scavenging activity of compatible solutes.
Phytochemistry. 1989; 28,1057-1060.
9. Alia, Saradhi PP. Suppression in mitochondrial electron transport is the primary cause
behind stress induced proline accumulation. Biochem Biophys Res Commun, 1991;
193:54-58.
10. Kohl DH, Schubert KR, Carter MB, Hagedorn CH, Shearer G. Proline metabolism in
N2-fixing root nodules: Energy transfer and regulation of synthesis. Proc Natl Acad Sci
USA 1998; 85,2036-2040.
11. Hong Z, Lakinini K, Zhang Z, Verma DPS. Removal of feedback inhibition of ∆1-pyrroline-
5-carboxylate synthetase enhances proline accumulation and removes oxidative stress
1999; submitted.
12. Smirnof N. The role of active oxygen in the response of plants to water deficit and desiccation.
New Phytol 1993; 125:27-58.
13. Christian JBH, Waltho J. The sodium and potassium content of non-halophilic bacteria
in relation to salt tolerance. J Gen Microbiol. 1961; 25:292-335.
14. Walderhaug MO, Dosch DC, Epstein W. Potassium transport in bacteria. In: Ion transport
in prokaryotes. Rosen BP, Silver S, eds. San Diego: Acadmic Press, 1987; 85-130.
15. Britten RJ, McClure FT. The amino acid pool in Escherichia coli. Bacteriol Rev 1962;
26:292-335.
16. Csonka LN. Regulation of cytopalsmic proline levels in Salmonella typhimurium: Effect
of osmotic stress on synthesis, degradation, and cellular retention of proline. J Bacteriol
1988; 170:2374-2378.
17. Smith LT. Characterization of g-glutamyl kinase from Escerichia coli that confers proline
overproduction and osmotic tolerance. J Bacteriol 1985; 164:1088-1093.
18. Jakowec MW, Smith LT, Dandekar AM. Recombinant plasmid conferring proline overproduction
and osmotic tolerance. Appl Environ Microbiol 1985; 50:441-446.
19. LeRudulier D, Yang SS, Csonka LN. Nitrogen fixation in Klebsiella pneumoniae during
osmotic stress: Effect of exogenous proline or a proline overproducing plasmid. Biophys
Acta 1982; 719:273-283.
20. Hanson AD, Hitz WD. Metabolic responses of mesophytes to plant water deficits. Ann
Rev Plant Physiol. 1982; 33:163-203.
21. Cairney J, Booth IR, Higgins CF. Salmonella typhimurium proP gene encodes a transport
system for the osmoprotectant betaine. J Bacteriol 1985a; 164:1218-1223.
22. Cairney J, Booth IR, Higgins CF. Osmoregulation of gene expression in Salmonella
typhimurium: proU gene encodes an osmotically induced betaine transport system.
J Bacteriol,1985b; 164:1224-1232.
23. May, G, Faatz E, Villarejo M, Bremer E. Binding protein dependent transport of glycine
betaine and its osmotic regulation in Escerichia coli K12. Mol Gen Genet 1983; 205:225-233.
24. LeRudulier D, Bouillard L. Glycine betaine, an osmotic effector in Klebsiella pneumoniae
and other members of the Enterobacteriaceae. Appl Environ Microbiol,1986; 46:152-159.
25. Bloomberg A, Adler, L. Physiology of osmotolerance in fungi. Adv Microbiol Physiol
1992; 33:145-212.
26. Mager WH, Varel JCS. Osmostress response of the yeast Saccharomyces. Mol Microbiol,
1993; 10:253-258.
27. Albertyn JS, Hohman JM, Thevelein, and Prior (1994). GPD1, which encodes glycerol-3-
phosphate dehydrogenase, is essential for growth under osmotic stress in Saccaromyces cerevisiae,
and its expression is regulated by the high-osmolarity glycerol response pathway. Mol
Cell Biol 1994; 14:4135-4144.
28. Streeter J. Accumulation of a,a-trehalose by Rhizobium bacteria and bacteroids. J Bact
1985; 164:78-84.
29. Aiba H, Mizuno T, Mizushima S. Transfer of phosphoryl group between two regulatory
proteins involved in osmoregulatory expression of the ompF and ompC genes in Escherichia coli.
J Biol Chem 1989; 264:8563-8567.
30. JoY-L, Nara F, Ichihara S, Mizuno T, Mizushima S. Purification and characterization of
the OmpR protein, a positive regulator involved in osmoregulatory expression of the
ompF and ompC genes in Escerichia coli. J Biol Chem 1986; 261:15252-15256.
31. Norioka S, Ramakrisshnan G, Ikenaka K, Inouye M. Interaction of a transcriptional activator,
OmpR, with reciprocally osmoregulated genes, ompF and ompC, of Escerichia coli.
J Biol Chem 1986; 261:17113-17119.
32. Forst S, Comeau D, Norioka S, Inouye M. Localization and membrane topology of EnvZ,
a protein involved in osmoregulation of OmpF and OmpC on Escherchia coli. J Biol Chem,
1987; 262:16433-16438.
33. Hall MN, Silhavy TJ. Genetic analysis of the major outer membrane proteins of
Escherchia coli K-12. Annu Rev Genet 1981b; 15:91-142.
34. Albright LM, Huala E, Ausubel FM. Prokaryotic signal transduction mediated by sensor
and regulator protein pairs. Annu Rev Genet 1989; 23:311.
35. Brewster JL, DeValoir T, Dwyer ND, Winter E, Gustin MC. An osmosensing signal transduction
pathway in yeast. Science. 1993; 259:1760-1763.
36. Maeda T, Wurgler-Murphy SM, Saito H. A two-component system that regulates an
osmosensing MAP kinase cascade in yeast. Nature. 1994; 369:242-245.
37. Posas F, Wurgler-Murpy SM, Maeda T, Witten EA, Thai TC, Saito H. Yeast HOG1 MAP
kinase cascade is regulated by a multistep phosphorelay mechanism in the SLN1-YPD1-SSK1
“Two-Component” osmosensor. Cell 1996; 86:865-875.
38. Ota IM, Varshaavsky A. A yeast protein similar to bacterial two-compound regulators.
Science. 1993; 262:566-569.
39. Boguslawski G. PBS2, a yeast gene encoding a putative protein kinase, interacts with the
RAS2 pathway and afects osmotic sensitivity of Saccharomyces cerevisiae. J Gen Microbiol
1992; 138:2425-2432.
40. Maeda T, Takekawa M, Saito H. Activation of yeast PBS2 MAPKK by MAPKKKs or by
41. Han J, Lee JD, Bibbs L, Ulevitch RJ. A MAP kinase targeted by endotoxin and
hyperosmolarity in mammalian cells. 1994; 265:808-811.
42. Galcheva-Gargova Z, Derijard B, Wu I-H, Davis RJ. An osmosensing signal transduction
pathway in mammalian cells. Science 1994; 265:806-808.
43. Mizoguchi T, Hayashida N, Yamaguchi-Shinozaki K, Kamada H, Shinozaki K. ATMPKs:
A gene family of MAP kinases in Arabidopsis thaliana. FEBS Letters. 1993; 336:440-444.
43a. Jonak C, Kiegerl S, Ligterink W, Barker PJ, Huskisson NS. Stress signaling in plants:
A mitogen-activated protein kinasee pathway is activated by cold and drought. Proc Natl
Acad Sci USA 1996; 93:11274-11279.
44. Ishitani M, Xiong L, Stevenson B, Zhu J-K. Genetic Analysis of Osmotic and cold stress
signal transduction in Arabidopsis: Interactions and convergence of abscisic acid-dependent
and abscisic acid-independent pathways. Plant Cell 1997; 9:1935-1949.
45. McCue KF, Hanson AD. Salt-inducible betaine aldehyde dehydrogenase from sugar beet:
cDNA cloning and expression. Plant Mol Biol. 1990; 8:358-362.
46. Binzel ML, Hess FD, Bressan RA, Hasegawa PM. Intracellular compartmentation of ions
in salt-adapted tobacco cells. Plant Physiol. 1988; 86:607-614.
47. Yancey PH, Clark ME, Hand SC, Bowlus RD, Somero G. Living with water stress: Evolution
of osmolyte systemes. Science 1982; 217:1214-1222.
48. Verma DPS, Zhang C-S. Regulation of proline and arginine biosynthesis in plants. In:
Singh BK, ed. Amino acid biosynthesis. Marcell and Decker Pub 1998:249-265.
49. Hanson AD, May AM, Grumet R, Bode J, Jamieson GC, Rhodes D. Betaine synthesis in
chenopods: Localization in chloroplasts. Proc Natl Acad Sci USA 1985; 82:3678-3682.
50. Delauney AJ, Hu CA, Kishore PBK, Verma DPS. Cloning of ornithine δ-aminotransferase
cDNA from Vigna aconitifolia by transcomplementation in Escherichia coli and regulation of
proline biosynthesis. J Biol Chem. 1993; 268:18673-18678.
51. Zhang C, Lu Q, Verma DPS. Removal of feedback inhibition of D1-pyrroline-5-carboxylate
synthetase, a bifunctional enzyme catalysing the first two steps of proline biosynthesis in
plants. J Biol Chem. 1995; 270:20491-20496.
52. Peng Z, Verma DPS. Reciprocal regulation of ∆1-Pyrroline-5-carboxylate synthetase and
proline dehydrogenase genes control probline levels during and after osmotic stress in
plants. Mol Gen Gent. 1996; 253:334-341.
53. Ekena K, Maloy S. Regulation of proline utilization in Salmonella typhimurium: How do
cells avoid futile cycle. Mol Gen Genet. 1990; 220:492-494.
54. Rayapati PJ, Stewart CR. Solublization of proline dehydrogenase from maize mitochondria.
Plant Physiol. 1991; 95:787-791.
55. Wang S, Brandriss MC. Proline utilization in Saccharomyces cervisiae: Analysis of cloned
PUT1 gene. Mol and Cell Biol 1986; 6:2638-2645.
56. Wood JM. Membrane association of proline dehydrogenase in Escherichia coli is redox
dependent. Proc Nat Acad Sci USA 1987; 84:373-377.
57. De Spicer PO, Maloy S. PutA protein, a membrane-associated flavin dehydrogenase, acts
as a redox-dependent transcriptional regulator. Proc Natl Acad Sci, USA 1993; 90:4295-4298.
58. Binzel ML, Hasegawa PM, Rhodes D, Handa S, Handa AK, Bressan RA. Solute
acuumulation in tobacco cells adapted to NaCl. Plant Physiol 1987; 84:1408-1415.
59. Wang S, Brandriss MC. Proline utilization in Saccharomyces cerevisiae: Sequence, regula-
tion, and mitochondrial localization of the PUTI gene product. Mol Cell Biol 1987;
7:4431-4440.
60. Brandris MC. Proline utilization in Saccharomyces cerevisiae: Analysis of cloned PUT2
gene. Mol Cell Biol 1983; 3:1846-1856.
61. Schobert B. Is there an osmotic regulatory mechanism in algae and higher plants? J Theor
Biol 1977; 68:17-26.
62. Ahmad I, Lahrher F, Stewart GR. Sorbitol, a compatible osmotic solute in Plantago maritima,
New Phytol 1979; 82:671-678.
63. Bieleski RL, Pedgewell RJ. Sorbitol metabolism in nectarines from flowers of Rosaeae.
Aust J Plant Physiol 1980; 7:15-25.
64. Tarczynski MM, Jensen RG, Bohnert HJ. Stress protection of transgenic tobacco by production of
the osmolyte mannitol. Science 1993; 259:508-510.
higher plants. Annu Rev Plant Physiol Plant Mol Biol 1993; 44:357-384.
66. Hanson AD, Gage DA. Identification and determination by fast atom bombardment mass
spectometry of the compatible solute choline-o-sulfate in Limonium species and other
halophytes. Aust J Plant Physiol 1991; 18:317-327.
67. James F, Nolte KD, Hanson AD. Purification and properties of S-adenosyl-l-methionine:
L-methionine s-methyltransferase from Wollastonia biflora leaves. J Biol Chem 1995; 270:
22344-22350.
68. Szoke A, Miao GH, Hong Z, Verma DPS. Subcellular location of D1-pyrroline-5-carboxylate
reductase in root/nodules and leaf of soybean. Plant Physiol 1992; 99:1642-1649.
69. Kavi Kishore PB, Hong Z, Miao G, Hu C, Verma DPS. Overexpression of ∆1-pyrroline-5-
carboxylate synthetase increases proline production and confers osmotol-erance in transgenic
plants. Plant Physiol 1995; 108:1387-1394.
70. Boggess SF, Aspinall D, Paleg L. Stress metabolism IX. The significance of end-product
inhibtion of proline biosynthesis and of compartmentation in relation to stress induced
proline accumulation. Aust J Plant Physiol 1976; 3:513-525
71. Boggess SF, Stewart CR, Aspinall D, Paleg L. Effect of water stress on proline synthesis
from radioactive precursors. Plant Physiol 1976; 58:398-401.
72. SmithJC, Deutch AH, Rushlow KE. Purification and characteristics of a γ-glutamyl kinase
involved in E. coli proline biosynthesis. J Bacteriol 1984; 157:545-551.
73. Csonka LN. Proline overproduction results in enhanced osmoregulation in Salmonella
typhimurium. Mol Gen Genet 1981; 182:82-86.
74. Dandekar AM, Uratsu SL. A single base pair change in proline biosynthesis genes causes
osotic stress tolerance. J Bacteriol 1988; 170:5943-5945.
75. Schmidt A, Jager K. Open questions about sulfur metabolism in plants. Ann Rev Plant
Physiol Plant Mol Biol 1992; 43:325-349.
76. Glaser HU, Thomas D, Gaxiola R, Montrichard F, Surdin-Kerjan Y, Serrano R. Salt tolerance
and methionine biosynthesis in Saccharomyces cerevisiae involve a putative phosphatase
gene. EMBOJ 1993; 3105-3110.
77. Neuwald AF, Krishnan BR, Brikun I, Kulakauskas S, Suziedelis K, Tomcsanyi T, Leyh
TS, Berg De. cysQ: A gene needed for cysteine synthesis in Escherichia coli K-12 only
during aeobic growth. J Bacteriol 1992; 174:415-425.
78. Murguia JR, Belles JM, Serrano R. A salt-sensitive 3'(2'), 5'-bisphosphate nucleotidase
involved in sulfate activation. Science 1995; 267:232-243.
79. York JD, Ponder JW, Majerus PW. Definition of a metal-dependant/Li + -inhibited
phosphomonoesterase protein family based upon a conserved three-dimensional core
structure. Proc Natl Acad Sci USA 1995; 92:5149-5153.
80. Chardot T, Meunier JC. Properties of oxidized and reduced spinach (Spinacia oleracea) chlo-
roplast fructros-1,6-bisphosphatase activated by various agents. J Biochem 1991; 278:787-791.
81. Markovich D, Forgo J, Strange G, Biber J, Murer H. Expression cloning of rat renal
Na+/So4(2-) cotransport. Proc Natl Acad Sci USA 1993; 90:8073-8077.
82. Watad A-E, A, Reuveni M, Bressan RA, Hasegawa PM. Enchanced net K+ uptake capacity of
NaCl-adapted cells. Plant Physiol 1991; 95:1265-1269.
83. Chaoui A, Mazhoudi S, Ghorbal MH, Ferjani EE. Cadmium and zince induction of lipid
peroxidation and effects on antioxidant enzyme activities in bean (Phaseolus vulgaris L.). Plant
Sci 1997; 127:139-147.
84. Shio Y, Tamai H, Sasa T. Effects of copper on photosynthetic electron transport systems
in spinach chloroplasts. Plant Cell Physiol 1978; 19:203-209.
85. Sandmann G, Boger P. Copper mediated lipid peroxidation processes in photosynthetic
membranes. Plant Physiol 1980; 66:797-800.
86. Goldsbrough PB, Mendum ML, Gupta SC. Effect of glutathione on phytochelation synthesis
and cadmium tolerance in tomato cells. In: Rennenber et al, eds. Sulfur and Sulfur Assimilation in
Higher Plants. The Hague, The Netherlands: SPB Academic Publishing 1990:185-189.
87. Scheller HV, Huang B, Hatch E, Goldsbrough PB. Phytochelation synthesis and glutathione
levels in response to heavy metals in tomato cells. Plant Physiol 1987; 85:1031-1035.
88. De Kok LJ, Stulen I. Role of glutathione in plants under oxidative stress. In: De Kok LJ
et al, eds. Sulfur Nutrition and Assimilation in Higher Plants. The Netherlands: SPB
Academic Publishing, 1993; 125-138.
89. Rauser WE. Metal binding peptides in plants. In: De Kok LJ et al, eds. Sulfur Nutrition and
Assimilation in Higher Plants. The Hague, The Netherlands: SPB Academic Publishing 1993;
239-251.
90. Meuwly P, Thibault P, Schwan AL, Rauser WE. Three families of thiol peptides are induced
by cadmium in maize. Plant J 1995; 7:391-400.
Index
A G
ABA 3-7, 11, 14-20, 22-25, 41, 71, 74, 101, Guard cell 5, 24, 36, 37, 101-106, 108-118
105-118, 158
ABA signaling 7, 19, 24, 110, 113 H
ABA-dependent pathway 14, 15, 17, 18
Abiotic stress 1-4, 7, 11, 87, 88 H+-ATPase 103, 106, 107, 110, 112, 117
ABRE 7, 15-18 Heat shock protein 83, 84, 94
Arabidopsis 102-104, 113-115, 144, 158 HSF phosphorylation 93
Arabidopsis thaliana 29, 38, 45, 49, 90, 144 HSF regulation 92, 93
B I
Betaines 1, 4, 147, 156, 158, 160 Ion homeostasis 31, 36, 48
Biosynthesis 3, 5, 6, 12, 14, 15, 17-20, 24, 25,
33, 34, 38, 47, 49, 76, 131, 133-137, K
139-143, 145, 147, 156, 158, 161-163
BZIP transcription factor 17 K+ channel 103, 104, 106, 110, 111, 113, 115,
117
C
M
Ca2+ 105-111, 113-117, 165
CBF1 16, 17, 73, 74, 77, 81 MAP kinase cascade 19, 21, 22, 33, 34
Chilling injury 63, 64, 66, 78, 88 Marine algae 38, 127-131, 134, 143
Chilling tolerance 63, 66-68, 77, 79, 88 Metabolic engineering 39, 47, 48, 127, 143,
Cl- channel 105, 109, 115 166
Cold acclimation 63, 70-78 Metabolite accumulation 29, 37
Compatible solute 29-32, 37-39, 47, 49, 127, MYB 13, 18, 19, 27
129-131, 155, 156 MYC 13, 18, 19
COR genes 71, 73, 74, 77
Cross protection 87, 88 N
CRT/DRE 74
Nicotiana 102, 103, 113
D
O
Development 1, 3, 5, 6, 14, 23, 24, 29, 30, 41,
44, 48, 49, 63, 68, 84, 90, 93, 94, 106, 117, Osmolytes 12, 30, 32-34, 45, 51, 112, 127,
158 128, 130, 147, 155, 156, 158, 160, 161
DRE/CRT 15, 16 Osmoprotectants 1, 14, 30, 161
Drought 1, 2, 6, 7, 11-25, 29, 38, 41, 43, 48, Osmoregulation 36, 49, 155, 157, 158, 160
49, 63, 74, 83, 87, 88, 101, 105, 127, 132, Osmotic adjustment 7, 30-32, 46, 47, 128,
145, 155, 158, 163, 166 131, 145
Osmotic stress 2, 5, 11, 18-22, 29, 31-35,
F 37-39, 44, 127, 129, 143, 146, 155-164,
166
Freezing tolerance 4, 63, 67, 69-78
Frost sensitive mutants 5
Oxidative stress 42, 43, 47, 66, 87, 88, 155,

164, 166
P
Plasma membrane 13, 31, 34, 35, 40, 41, 44,
45, 69, 70, 72, 76, 103-112, 115-117
Proline 1, 12-14, 29, 34, 37, 38, 43, 48, 49, 72,
76, 77, 127, 155, 156, 158-164, 166
Protein dephosphorylation 19, 113
Protein phosphorylation 24, 105, 110, 112,
113, 115
Q
QTLs 3, 7
Quantitative trait loci 3
Quaternary ammonium compounds 1, 127,
155
R
Reactive oxygen species (ROS) 2, 3, 42, 43, 48,
83, 88
Reverse genetics 3, 25
S
Salinity 11, 14, 17, 21, 22, 29, 30, 34, 38, 39,
44-49, 75, 105, 127, 130-134, 155, 158,
160, 161, 163, 164
Salt tolerance 4, 31-38, 45, 46, 48, 49, 164,
166
Seed dormancy 4, 6
SFR 75
Sulfur metabolism 147, 155, 161, 164, 166
T
Thermotolerance 83-85, 87, 88, 93
Tonoplast 13, 35, 40, 44, 46, 105-108,
111-113, 115-117
Two component system 157
V
Vicia 102, 103, 109, 112, 116, 117
W
Water channels 39, 42
Water stress 5, 20, 88, 101, 102, 105, 106, 161

Molecular Responses

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Molecular Responses

Uploaded by

Copyright:

Available Formats

BIOTECHNOLOGY I N T E L L I G E N C E U N I T 1

Kazuo Shinozaki and

Kazuo Shinozaki, Ph.D.

Kazuko Yamaguchi-Shinozaki, Ph.D.

R.G. LANDES COMPANY

Library of Congress Cataloging-in-Publication Data

2. Molecular Responses to Drought Stress ................................................ 11

4. Plant Cold Tolerance ............................................................................... 63

6. Cellular Responses to Water Stress ...................................................... 101

Bo Shen, Ph.D. Hua Su, Ph.D.

Hans J. Bohnert, Ph.D. M. Koornneef, Ph.D.

Douglas A. Gage, Ph.D. A.J.M. Peeters, Ph.D.

Bala Rathinasabapathi, Ph.D.

Andreas Reindl, Ph.D.

Fritz Schöffl, Ph.D.

Michael F. Thomashow, Ph.D.

Desh Pal S. Verma, Ph.D.

The Genetic Approach in Stress Physiology—General Principals

Genetic Differences in Salt Tolerance

Genetic Differences in the Response to Low Temperatures

ABA Related Mutants

The Molecular Function of ABA, ABI and ERA Genes

The Isolation of Mutants Affected in the Response to Stress

Genetic Variation for Morphological Traits Related to Stress Tolerance

Molecular Responses to Drought

signal transduction in drought stress response. Promoter analysis of several drought

A Variety of Functions of Drought-Inducible Genes

Existence of a variety of drought-inducible genes suggests complex responses of plants

Improvement of Stress Tolerance Using Gene Transfer

Regulation of Gene Expression by Drought

Major ABA-Independent Regulatory System of Gene Expression during

Drought-Specific ABA-Independent Regulatory System (Pathway III)

Major ABA-Dependent Regulatory System (Pathway II): Important Roles of

Roles of MYC and MYB Homologs in ABA-Dependent Gene Expression

Signal Perception and Signal Transduction in Drought Stress

homologues, which suggests the involvement of similar osmosensing mechanisms in higher

Fig. 2.6. Second messengers and factors

Sensing of Osmotic stress: The Two-Component System

Fig. 2.7. Possible roles of the two-component

Roles of MAP Kinase Cascades in Stress Response

Roles of Phospholipids Metabolism and Calcium in Drought Stress Signaling

ABA Signal Transduction

Regulation of ABA Biosynthesis

Conclusion and Perspectives

3. Shinozaki K, Yamaguchi-Shinozaki K. Molecular responses to drought and cold stress.

23. Nakashima K, Kiyosue T, Yamaguchi-Shinozaki K, Shinozaki K. A nuclear gene, erd1,

42. Meskiene I, Boerge L, Glaser W, Balog J, Brandstoetter M, Zwerger K, Ammerer G, Hirt

Molecular Mechanisms of Salinity

Osmolytes, Osmoprotectants, Compatible Solutes, Osmotic

Cellular Mechanisms of Salt Tolerance—the Fungal Model

Replacing Glycerol in Yeast

Osmosensing and Signaling

regulation of K+ concentrations. Overexpression of HAL1 and HAL3 resulted in a stronger

Sodium Transport Across Membranes

the Ca 2+ -induced inactivation of K + channels was inhibited. A Ca 2+ -dependent

Molecular Mechanisms of Salt Tolerance in Plants

ROS Scavenging Enzymes 1991 SOD, catalase, GST/GSX overexpression

Mannitol Synthesis 1992 Protection against salt stress.251,252,253

Fructan Accumulation 1995 Enhanced drought tolerance.254

Proline Accumulation 1995 Enhanced salt stress tolerance.135

LEA Protein Synthesis 1996 Salinity and drought stress protection.256

Potassium Transporter 1994 Enhanced Na+/K+-discrimination in yeast.86,207

Trehalose Synthesis 1996 Enhanced drought tolerance.141