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. Laboratory of Plant Physiology and Biophysics University of London, Wye College Wye, England, U.K. Chapter 6

Alexander Grabov, Ph.D. Laboratory of Plant Physiology and Biophysics University of London, Wye College Wye, England, U.K. Chapter 6

Bo Shen, Ph.D. Departments of Plant Sciences The University of Arizona Tucson, Arizona, U.S.A. Chapter 3

Hua Su, Ph.D. Departments of Plant Sciences The University of Arizona Tucson, Arizona, U.S.A. Chapter 3

Hans J. Bohnert, Ph.D. Departments of Biochemistry, Molecular and Cellular Sciences The University of Arizona Tucson, Arizona, U.S.A. Chapter 3

M. Koornneef, Ph.D. Laboratory of Genetics Wageningen Agricultural University Wageningen, The Netherlands Chapter 1

John Browse, Ph.D. Institute of Biological Chemistry Washington State University Pullman, Washington, U.S.A. Chapter 4

Barbara Leyman, Ph.D. Laboratory of Plant Physiology and Biophysics University of London, Wye College Wye, England, U.K. Chapter 6

Douglas A. Gage, Ph.D. Department of Biology Michigan State University East Lansing, Michigan, U.S.A. Chapter 7

A.J.M. Peeters, Ph.D. Laboratory of Genetics Wageningen Agricultural University Wageningen, The Netherlands 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. Lehrstuhl Allgemeine Genetik Universität Tubingen Tubingen, Germany Chapter 5 Fritz Schöffl, Ph.D. Lehrstuhl Allgemeine Genetik Universität Tubingen 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 osmoadjustment 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

4

Molecular Responses to Cold, Drought, Heat and Salt Stress in Higher Plants

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 genotypes. 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 highaffinity 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.

Genetic Approaches to Abiotic Stress Responses

5

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.

6

Molecular Responses to Cold, Drought, Heat and Salt Stress in Higher Plants

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.,

Genetic Approaches to Abiotic Stress Responses

7

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.

Molecular Responses to Cold, Drought, Heat and Salt Stress in Higher Plants

8

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 contribution 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.

Genetic Approaches to Abiotic Stress Responses

9

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 waterstress 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; 61:385393. 37. Koornneef M, Reuling G, Karssen CM. The isolation and characterization of abscisic acidinsensitive 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 abscisicacid-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 positional 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.

10

Molecular Responses to Cold, Drought, Heat and Salt Stress in Higher Plants

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 survive 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 ABAdependent 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. Therefore, 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 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.

12

Molecular Responses to Cold, Drought, Heat and Salt Stress in Higher Plants

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 embryogenesis 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 macromolecules 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 phosphatases are reported.8 These protein kinases and phosphatases may be involved in modification of functional proteins and regulatory proteins involved in stress signal transduction 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.

14

Molecular Responses to Cold, Drought, Heat and Salt Stress in Higher Plants

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

Molecular Responses to Drought Stress

15

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 ABAindependent 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

16

Molecular Responses to Cold, Drought, Heat and Salt Stress in Higher Plants

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 DRErelated 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

Molecular Responses to Drought Stress

17

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-boxlike sequences. These results suggest that ABA-inducible bZIP proteins are also involved in

18

Molecular Responses to Cold, Drought, Heat and Salt Stress in Higher Plants

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 DNAbinding 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 DNAbinding 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 droughtresponsive expression of the rd22 gene.29 A cDNA for a transcription factor MYC homologue, 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

Molecular Responses to Drought Stress

19

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 droughtstress 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 signaling, 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 function 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.

20

Molecular Responses to Cold, Drought, Heat and Salt Stress in Higher Plants 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

Molecular Responses to Drought Stress

21 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, YamaguchiShinozaki, 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 phosphorylating 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).

22

Molecular Responses to Cold, Drought, Heat and Salt Stress in Higher Plants

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

Molecular Responses to Drought Stress

23

Fig. 2.8. Possible roles of phospholipids metabolism in drought stress and ABA signal transduction pathways. Many genes involved in phospholipids metabolism (PI turnover and PC turnover), 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

24

Molecular Responses to Cold, Drought, Heat and Salt Stress in Higher Plants

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 calciumdependent protein kinase functions as a positive regulator. These results indicate that a protein phosphorylation and dephosphorylation process might be involved in ABAresponsive 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 biochemical 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 biosynthesis 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

Molecular Responses to Drought Stress

25

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-responsive 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.

26

Molecular Responses to Cold, Drought, Heat and Salt Stress in Higher Plants

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 stress response. Plant Physiol 1997; 115:327-334. 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 ∆1pyrroline-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 temperatureresponsive 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.

Molecular Responses to Drought Stress

27

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 cisacting 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 acidVIVIPAROUS1 (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 transduction. 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.

28

Molecular Responses to Cold, Drought, Heat and Salt Stress in Higher Plants

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 mitogenactivated 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 phosphatidylinositolspecific 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. 52. Meyer K, Leube MP, Grill E. A protein phosphatase 2C involved in ABA signal transduction 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 biosynthesis 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 transduction in Arabidopsis: Interactions and convergence of abscisic acid-dependent and abscisic acid-independent pathways. Plant Cell 1997; 9:1-16. 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 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.

30

Molecular Responses to Cold, Drought, Heat and Salt Stress in Higher Plants

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 environmental 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 solutes 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

32

Molecular Responses to Cold, Drought, Heat and Salt Stress in Higher Plants

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-3phosphatase, 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 speciesspecific 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.

Molecular Mechanisms of Salinity Tolerance

33

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 glycerol3-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.

34

Molecular Responses to Cold, Drought, Heat and Salt Stress in Higher Plants

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

Molecular Mechanisms of Salinity Tolerance

35

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 calciumdependent 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 confirmed 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.

36

Molecular Responses to Cold, Drought, Heat and Salt Stress 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,

Molecular Mechanisms of Salinity Tolerance

37

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

38

Molecular Responses to Cold, Drought, Heat and Salt Stress in Higher Plants

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 Enhancement

1997 Altered redox control, salt and low temperature protection.192

Mannitol as a Hydroxyl Radical Scavenger

1997 Enhanced salt tolerance, mannitol synthesis in chloroplast.29,30

Inducible Ononitol Accumulation

1997 Enhanced drought and salt tolerance; inducibility based on changes in substrate amounts.138

Extreme Sorbitol Accumulation

1998 High accumulation, >600 mM sorbitol, leading to 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 (tetrahydropyrimidine and derivatives), an accumulating osmolyte in bacteria, has received

Molecular Mechanisms of Salinity Tolerance

39

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 nodulationspecific protein, may mediate ion transport and transport of other neutral metabolites, such as urea.144,145

40

Molecular Responses to Cold, Drought, Heat and Salt Stress in Higher Plants

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, RNAstability 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

Molecular Mechanisms of Salinity Tolerance

41

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 cellspecific 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 expression, which may be controlled by oligomerization, glycosylation, or phosphorylation.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

42

Molecular Responses to Cold, Drought, Heat and Salt Stress in Higher Plants

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 phosphorylation.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, nonphotochemical quenching processes by various compounds, for example, the violaxanthinzeaxanthin 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

Molecular Mechanisms of Salinity Tolerance

43

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 ascorbateglutathione 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 diffusioncontrolled 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

44

Molecular Responses to Cold, Drought, Heat and Salt Stress in Higher Plants

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 detail.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

Molecular Mechanisms of Salinity Tolerance

45

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

46

Molecular Responses to Cold, Drought, Heat and Salt Stress in Higher Plants

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 saltsensitive 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 calciumregulated 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/myoinositol 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

Molecular Mechanisms of Salinity Tolerance

47

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

48

Molecular Responses to Cold, Drought, Heat and Salt Stress in Higher Plants

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 analyzed. 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

Molecular Mechanisms of Salinity Tolerance

49

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.

50

Molecular Responses to Cold, Drought, Heat and Salt Stress in Higher Plants

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.

Molecular Mechanisms of Salinity Tolerance

51

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 Saccharomyces 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-3phosphate 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.

52

Molecular Responses to Cold, Drought, Heat and Salt Stress in Higher Plants

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 binding of an SH3-containing osmosensor. Science 1995; 269:554-558. 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 adaptationReplacement 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.

Molecular Mechanisms of Salinity Tolerance

53

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; 269:87928796. 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 NaCladapted 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 Plant Physiol Plant Mol Biol 1994; 45:211-234. 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.

54

Molecular Responses to Cold, Drought, Heat and Salt Stress in Higher Plants

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.

Molecular Mechanisms of Salinity Tolerance

55

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 mitochondrial 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.

56

Molecular Responses to Cold, Drought, Heat and Salt Stress in Higher Plants

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.

Molecular Mechanisms of Salinity Tolerance

57

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.

58

Molecular Responses to Cold, Drought, Heat and Salt Stress in Higher Plants

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 inwardrectifying 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.

Molecular Mechanisms of Salinity Tolerance

59

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.

60

Molecular Responses to Cold, Drought, Heat and Salt Stress in Higher Plants

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 transduction in Arabidopsis: Interactions and convergence of abscisic acid-dependent and abscisic acid-independent pathways. Plant Cell 1997; 9:1935-1949. 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 waterstress response. Plant Physiol 1997; 115:327-334. 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 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.

62

Molecular Responses to Cold, Drought, Heat and Salt Stress in Higher Plants

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 chillingsensitive 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

64

Molecular Responses to Cold, Drought, Heat and Salt Stress in Higher Plants

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 chillingsensitive 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 photosynthesis 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

Plant Cold Tolerance

65 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 temperatures. 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

66

Molecular Responses to Cold, Drought, Heat and Salt Stress in Higher Plants

(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 linesfab1 (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, chillingimpaired 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

Plant Cold Tolerance

67

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

68

Molecular Responses to Cold, Drought, Heat and Salt Stress in Higher Plants

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

Plant Cold Tolerance

69

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 (coldinducible), 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

70

Molecular Responses to Cold, Drought, Heat and Salt Stress in Higher Plants

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

Plant Cold Tolerance

71

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

72

Molecular Responses to Cold, Drought, Heat and Salt Stress in Higher Plants

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 freezeinduced 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

Plant Cold Tolerance

73

Table 4.1 Phenotypes associated with sfr mutations in cold-acclimated plants.88,89 Mutant Gene

Freezing Sensitivity

Anthocyanin Level (% wt)

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 both glucose (<<10% wt) and sucrose (<25% wt)

Reduced levels of 16:0, 18:1 and 18:2 fatty acids

sfr5

All leaves

115

Normal

No changes detected

43

Normal

No changes detected

sfr6 Worst in young leaves

Glucose and Sucrose Levels

Fatty Acid Composition

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

74

Molecular Responses to Cold, Drought, Heat and Salt Stress in Higher Plants

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.

Plant Cold Tolerance

75

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 chillingassociated 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

76

Molecular Responses to Cold, Drought, Heat and Salt Stress in Higher Plants

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 coldactivated 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 coldresponsive 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 respiratory 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 chloroplasts 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.

Plant Cold Tolerance

77

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 contribution 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 developing 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 Temperature 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.

78

Molecular Responses to Cold, Drought, Heat and Salt Stress in Higher Plants

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. Proc Natl Acad Sci USA 1986; 83:2422-2426. 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. Proc Natl Acad Sci USA 1985; 82:3673-3677. 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 acclimation-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.

Plant Cold Tolerance

79

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 regulators. 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 Plant Physiol Plant Mol Biol 1996; 47:377-403. 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. Proc Natl Acad Sci USA 1996; 93:13404-13409. 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 polypeptides encoded by COR (cold-regulated) genes of Arabidopsis thaliana on dehydrationinduced 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 polypeptides encoded by COR (cold-regulated) genes of Arabidopsis thaliana on the freezeinduced 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.

80

Molecular Responses to Cold, Drought, Heat and Salt Stress in Higher Plants

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 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.

82

Molecular Responses to Cold, Drought, Heat and Salt Stress in Higher Plants

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

84

Molecular Responses to Cold, Drought, Heat and Salt Stress in Higher Plants

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, chloroplasts, 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

Molecular Responses to Heat Stress

85

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.

86

Molecular Responses to Cold, Drought, Heat and Salt Stress in Higher Plants

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

Molecular Responses to Heat Stress

87

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 heatinducible 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

88

Molecular Responses to Cold, Drought, Heat and Salt Stress in Higher Plants

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

Molecular Responses to Heat Stress

DBD

HR-AB

89

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 hydrophobic 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

90

Molecular Responses to Cold, Drought, Heat and Salt Stress in Higher Plants

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 transcription. 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.

Molecular Responses to Heat Stress

91

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 phosphorylation 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

Molecular Responses to Cold, Drought, Heat and Salt Stress in Higher Plants

92

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.

Molecular Responses to Heat Stress

93

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 characterization 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 heatdenatured 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.

94

Molecular Responses to Cold, Drought, Heat and Salt Stress in Higher Plants

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.

Molecular Responses to Heat Stress

95

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.

96

Molecular Responses to Cold, Drought, Heat and Salt Stress in Higher Plants

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.

Molecular Responses to Heat Stress

97

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.

98

Molecular Responses to Cold, Drought, Heat and Salt Stress in Higher Plants

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 process, 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 Shinozaki and Kazuko Yamaguchi-Shinozaki. ©1999 R.G. Landes Company.

100

Molecular Responses to Cold, Drought, Heat and Salt Stress in Higher Plants

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

102

Molecular Responses to Cold, Drought, Heat and Salt Stress in Higher Plants

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 electrochemical 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

Cellular Responses to Water Stress

103

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 activation 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 Clover 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 environmental 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).

104

Molecular Responses to Cold, Drought, Heat and Salt Stress in Higher Plants

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 ABAmediated 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 ABAevoked 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

Cellular Responses to Water Stress

105

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.

106

Molecular Responses to Cold, Drought, Heat and Salt Stress in Higher Plants

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 electrochemical 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.

Cellular Responses to Water Stress

107

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 membrane 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+-mediated 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

108

Molecular Responses to Cold, Drought, Heat and Salt Stress in Higher Plants

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 voltagedependence 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

Cellular Responses to Water Stress

109

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

110

Molecular Responses to Cold, Drought, Heat and Salt Stress in Higher Plants

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 measurements 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

Cellular Responses to Water Stress

111

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 timedependent 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 indicated 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 electrophysiological 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 germinated 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 phosphorylation 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 presence 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

112

Molecular Responses to Cold, Drought, Heat and Salt Stress in Higher Plants

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

Cellular Responses to Water Stress

113

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 cytosolic 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

114

Molecular Responses to Cold, Drought, Heat and Salt Stress in Higher Plants

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 timing 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

Cellular Responses to Water Stress

115

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

116

Molecular Responses to Cold, Drought, Heat and Salt Stress in Higher Plants

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.

Cellular Responses to Water Stress

117

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 membrane 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 plasmamembrane 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 ATPdependent, “electrogenic” transport revealed by current-voltage and difference-current-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.

118

Molecular Responses to Cold, Drought, Heat and Salt Stress in Higher Plants

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.

Cellular Responses to Water Stress

119

62. Linder B, Raschke K. A slow anion channel in guard cells, activating at large hyperpolarization, 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 cAMPdependent 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; 13:1098116. 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.

120

Molecular Responses to Cold, Drought, Heat and Salt Stress in Higher Plants

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.

Cellular Responses to Water Stress

121

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 stomatal 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 highthreshold 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.

122

Molecular Responses to Cold, Drought, Heat and Salt Stress in Higher Plants

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. Proc Natl Acad Sci USA 1993; 90:11493-11497. 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 acidinsensitive mutants of Arabidopsis thaliana. Physiol.Plant. 1984; 61:377-383.

Cellular Responses to Water Stress

123

157. Meyer K, Leube MP, Grill E. A protein phosphatase 2C involved in ABA signal transduction 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 acidInsensitive 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.

124

Molecular Responses to Cold, Drought, Heat and Salt Stress in Higher Plants

181. Clapham DE. Direct G-protein activation of ion channels. Annual Review Of Neuroscience. 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.

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.

126

Molecular Responses to Cold, Drought, Heat and Salt Stress in Higher Plants

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

Role of Glycine Betaine and Dimethylsulfoniopropionate in Water-Stress Tolerance

127

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

128

Molecular Responses to Cold, Drought, Heat and Salt Stress in Higher Plants

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

17

452

6

4.1

LSD (0.05)

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

Role of Glycine Betaine and Dimethylsulfoniopropionate in Water-Stress Tolerance

129

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)

130

Molecular Responses to Cold, Drought, Heat and Salt Stress in Higher Plants

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 typical 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 dimethylsulfoniopropionaldehyde 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

Role of Glycine Betaine and Dimethylsulfoniopropionate in Water-Stress Tolerance

131

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) ethanolamine 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

132

Molecular Responses to Cold, Drought, Heat and Salt Stress in Higher Plants

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 phosphatidylcholine 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.

Role of Glycine Betaine and Dimethylsulfoniopropionate in Water-Stress Tolerance

133

Fig. 7.4. Biosynthetic pathway to DMSP in the dinoflagellate Crypthecodinium cohnii.104,105 A PLPdependent 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

134

Molecular Responses to Cold, Drought, Heat and Salt Stress in Higher Plants

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

Role of Glycine Betaine and Dimethylsulfoniopropionate in Water-Stress Tolerance

135

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, dimethylsulfoniopropionaldehyde, 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

136

Molecular Responses to Cold, Drought, Heat and Salt Stress in Higher Plants

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,

Role of Glycine Betaine and Dimethylsulfoniopropionate in Water-Stress Tolerance

137

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 dimethylsulfoniopropionaldehyde (DMSP-ald).121 Although the conversion of SMM to DMSP-ald requires two steps, decarboxylation and deamination, no free intermediates such as dimethylsulfoniopropylamine (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

138

Molecular Responses to Cold, Drought, Heat and Salt Stress in Higher Plants

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 analysis also indicated that exogenously-supplied DMSP-amine first enters the storage pool before

Role of Glycine Betaine and Dimethylsulfoniopropionate in Water-Stress Tolerance

139

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 DMSPproducing 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 transaminase/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

140

Molecular Responses to Cold, Drought, Heat and Salt Stress in Higher Plants

Fig. 7.7. Subcellular localization of the DMSP biosynthetic steps in W. biflora. A schematic representative 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.

Role of Glycine Betaine and Dimethylsulfoniopropionate in Water-Stress Tolerance

141

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 redistribution 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 chloroplasts 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

142

Molecular Responses to Cold, Drought, Heat and Salt Stress in Higher Plants

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.

Role of Glycine Betaine and Dimethylsulfoniopropionate in Water-Stress Tolerance

143

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 difficult, 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

144

Molecular Responses to Cold, Drought, Heat and Salt Stress in Higher Plants

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.

Role of Glycine Betaine and Dimethylsulfoniopropionate in Water-Stress Tolerance

145

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 transport 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 introduction 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.

146

Molecular Responses to Cold, Drought, Heat and Salt Stress in Higher Plants

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 dimethylsulfoniopropionate. 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 application. 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. 8. Rhodes D, Hanson AD. Quaternary ammonium and tertiary sulfonium compounds in 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. Photosynthesis 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.

Role of Glycine Betaine and Dimethylsulfoniopropionate in Water-Stress Tolerance

147

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 phytoplankton. 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.

148

Molecular Responses to Cold, Drought, Heat and Salt Stress in Higher Plants

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 dimethylsulfoniopropionate (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 saltstressed 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.

Role of Glycine Betaine and Dimethylsulfoniopropionate in Water-Stress Tolerance

149

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.

150

Molecular Responses to Cold, Drought, Heat and Salt Stress in Higher Plants

87. Ishitani M, Nakamura T, Han SY, Takabe T. Expression of the betaine aldehyde dehydrogenase 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. Dimethylsulfoniopropionate 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-Lmethionine: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 dimethylsulfoniopropionate 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 dimethylthiopropionic acid in the dinoflagellate Crypthecodinium cohnii. Nippon Suisan Gakkaishi 1993; 59:851-855. 106. Stevenson, DE, Akhtar, M, Gani, D. Streptomyces L-methionine decarboxylase: Purification 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 felixmas: Purification, characterization, substrate specificity, abortive transamination of the coenzyme, and stereochemical courses of substrate decarboxylation and coenzyme transamination. Biochem 1990; 29:7631-7647.

Role of Glycine Betaine and Dimethylsulfoniopropionate in Water-Stress Tolerance

151

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-dimethylsulfoniopropionate 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 dimethylsulfoniopropionaldehyde 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 dimethylsulfoniopropionate 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.

152

Molecular Responses to Cold, Drought, Heat and Salt Stress in Higher Plants

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-dimethylsulfoniumacetate, 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 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.

154

Molecular Responses to Cold, Drought, Heat and Salt Stress in Higher Plants

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 phosphorylated 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 dehydrogenase (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

156

Molecular Responses to Cold, Drought, Heat and Salt Stress in Higher Plants

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 concentration 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,

Osmotic Stress Tolerance in Plants: Role of Proline and Sulfur Metabolisms

157

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.

158

Molecular Responses to Cold, Drought, Heat and Salt Stress in Higher Plants

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

Osmotic Stress Tolerance in Plants: Role of Proline and Sulfur Metabolisms

159

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

160

Molecular Responses to Cold, Drought, Heat and Salt Stress in Higher Plants

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

Osmotic Stress Tolerance in Plants: Role of Proline and Sulfur Metabolisms

A

B

161

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.

162

Molecular Responses to Cold, Drought, Heat and Salt Stress in Higher Plants

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 ATPsulfurylase 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

Osmotic Stress Tolerance in Plants: Role of Proline and Sulfur Metabolisms

163

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, ATPsulfurylase, 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

164

Molecular Responses to Cold, Drought, Heat and Salt Stress in Higher Plants

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.

Osmotic Stress Tolerance in Plants: Role of Proline and Sulfur Metabolisms

165

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-pyrroline5-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.

166

Molecular Responses to Cold, Drought, Heat and Salt Stress in Higher Plants

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-3phosphate 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 binding of an SH3-containing osmosensor. Science 1995; 269:554-558. 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.

Osmotic Stress Tolerance in Plants: Role of Proline and Sulfur Metabolisms

167

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, regulation, 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. 65. Rhodes D, Hanson AD. Quaternary ammonium and tertiary sulfonium compounds in 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.

168

Molecular Responses to Cold, Drought, Heat and Salt Stress in Higher Plants

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-5carboxylate 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) chloroplast 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, 105-118, 158 ABA signaling 7, 19, 24, 110, 113 ABA-dependent pathway 14, 15, 17, 18 Abiotic stress 1-4, 7, 11, 87, 88 ABRE 7, 15-18 Arabidopsis 102-104, 113-115, 144, 158 Arabidopsis thaliana 29, 38, 45, 49, 90, 144

Guard cell 5, 24, 36, 37, 101-106, 108-118

B

I

Betaines 1, 4, 147, 156, 158, 160 Biosynthesis 3, 5, 6, 12, 14, 15, 17-20, 24, 25, 33, 34, 38, 47, 49, 76, 131, 133-137, 139-143, 145, 147, 156, 158, 161-163 BZIP transcription factor 17

Ion homeostasis 31, 36, 48

C Ca2+ 105-111, 113-117, 165 CBF1 16, 17, 73, 74, 77, 81 Chilling injury 63, 64, 66, 78, 88 Chilling tolerance 63, 66-68, 77, 79, 88 Cl- channel 105, 109, 115 Cold acclimation 63, 70-78 Compatible solute 29-32, 37-39, 47, 49, 127, 129-131, 155, 156 COR genes 71, 73, 74, 77 Cross protection 87, 88 CRT/DRE 74

D Development 1, 3, 5, 6, 14, 23, 24, 29, 30, 41, 44, 48, 49, 63, 68, 84, 90, 93, 94, 106, 117, 158 DRE/CRT 15, 16 Drought 1, 2, 6, 7, 11-25, 29, 38, 41, 43, 48, 49, 63, 74, 83, 87, 88, 101, 105, 127, 132, 145, 155, 158, 163, 166

F Freezing tolerance 4, 63, 67, 69-78 Frost sensitive mutants 5

H H+-ATPase 103, 106, 107, 110, 112, 117 Heat shock protein 83, 84, 94 HSF phosphorylation 93 HSF regulation 92, 93

K K+ channel 103, 104, 106, 110, 111, 113, 115, 117

M MAP kinase cascade 19, 21, 22, 33, 34 Marine algae 38, 127-131, 134, 143 Metabolic engineering 39, 47, 48, 127, 143, 166 Metabolite accumulation 29, 37 MYB 13, 18, 19, 27 MYC 13, 18, 19

N Nicotiana 102, 103, 113

O Osmolytes 12, 30, 32-34, 45, 51, 112, 127, 128, 130, 147, 155, 156, 158, 160, 161 Osmoprotectants 1, 14, 30, 161 Osmoregulation 36, 49, 155, 157, 158, 160 Osmotic adjustment 7, 30-32, 46, 47, 128, 131, 145 Osmotic stress 2, 5, 11, 18-22, 29, 31-35, 37-39, 44, 127, 129, 143, 146, 155-164, 166

172

Molecular Responses to Cold, Drought, Heat and Salt Stress in Higher Plants

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