Gene 388 (2007) 1 – 13 www.elsevier.com/locate/gene

Review

Deciphering the regulatory mechanisms of abiotic stress tolerance in plants by genomic approaches N. Sreenivasulu a,⁎, S.K. Sopory b , P.B. Kavi Kishor c a

Leibniz-Institute of Plant Genetics and Crop Plant Research (IPK), Corrensstrasse 3, 06466, Gatersleben, Germany International Centre for Genetic Engineering and Biotechnology, Aruna Asaf Ali Marg, New Delhi 110 067, India c Department of Genetics, Osmania University, Hyderabad 500 007, India

b

Received 19 July 2006; received in revised form 8 October 2006; accepted 12 October 2006 Available online 24 October 2006 Received by A.J. van Wijnen

Abstract Environmental constraints that include abiotic stress factors such as salt, drought, cold and extreme temperatures severely limit crop productivity. Improvement of crop plants with traits that confer tolerance to these stresses was practiced using traditional and modern breeding methods. Molecular breeding and genetic engineering contributed substantially to our understanding of the complexity of stress response. Mechanisms that operate signal perception, transduction and downstream regulatory factors are now being examined and an understanding of cellular pathways involved in abiotic stress responses provide valuable information on such responses. This review presents genomic-assisted methods which have helped to reveal complex regulatory networks controlling abiotic stress tolerance mechanisms by high-throughput expression profiling and gene inactivation techniques. Further, an account of stress-inducible regulatory genes which have been transferred into crop plants to enhance stress tolerance is discussed as possible modes of integrating information gained from functional genomics into knowledge-based breeding programs. In addition, we envision an integrative genomic and breeding approach to reveal developmental programs that enhance yield stability and improve grain quality under unfavorable environmental conditions of abiotic stresses. © 2006 Elsevier B.V. All rights reserved. Keywords: Abiotic stress tolerance; Osmoregulation; Regulators; Genomic approaches; Transgenics

1. Introduction Abbreviations: ROS, reactive oxygen species; cDNA, complementary DNA; EST, expressed sequence tags; BLAST, basic local alignment search tool; SAGE, serial analysis of gene expression; MPSS, massively parallel signature sequencing; qRT-PCR, quantitative real time polymerase chain reaction; ABA, abscisic acid; ABF, abscisic acid responsive element binding factor; DREB, drought-responsive element binding factor; DRE, drought responsive element; LEA, late embryonic abundant; CCA1, circadian clock associated 1; bZIP, basic-leucine zipper; AP2, apetala 2; EREBP/ERF, ethylene-responsive element binding factor; NAC, acronym of no apical meristem; CBF, C-repeat binding factor; COR, cold-regulated; HSP, heat shock protein; PP2C, protein phosphatase type 2C; HXK, hexokinase; MAPK, mitogen-activated protein kinase; MKK, MAPK kinase; AtHK1, Arabidopsis thaliana histidine kinase 1; CDPK, calcium-dependent protein kinase; TILLING, targeted-induced local lesions in genomes; SNF1, sucrose non-fermenting 1; SOS, salt overly sensitive; los, low expression of osmotically responsive; hos, high expression of osmotically responsive; ABI, abscisic acid insensitive; eQTL, expression quantitative trait loci. ⁎ Corresponding author. Tel.: +49 39482 5172; fax: +49 39482 5595. E-mail address: [email protected] (N. Sreenivasulu). 0378-1119/$ - see front matter © 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.gene.2006.10.009

Plant productivity is severely affected by abiotic stress factors which include salinity, drought, high and low temperature, and heavy metals. As a sequel to it, physiological and biochemical responses in plants vary and cellular aqueous and ionic equilibriums are disrupted. Also, hundreds of genes and their products respond to these stresses at transcriptional and translational level (for reviews see Cushman and Bohnert, 2000; Sreenivasulu et al., 2004a; Yamaguchi-Shinozaki and Shinozaki, 2005; Umezawa et al., 2006a). Understanding the functions of these stress-inducible genes helps to unravel the possible mechanisms of stress tolerance. Of late, functional genomic approaches triggered a major paradigm from single gene discovery to thousands of genes by using multi-parallel highthroughput techniques. Generation of expressed sequence tags (ESTs) from cDNA libraries prepared from abiotic stress-treated seedlings of various crops, complete genome sequence

2

N. Sreenivasulu et al. / Gene 388 (2007) 1–13

information of rice and Arabidopsis provided a valuable resource for gene discovery. Furthermore, employment of multi-parallel techniques such as expression profiling by microarrays, random and targeted mutagenesis, complementation and promoter-trapping strategies allow the identification of the key stress-responsive gene pools and in turn provide important clues for functional characterization of stressresponsive genes and stress tolerance mechanisms. Recent genomic studies show considerable overlap of plant responses to osmotic stresses such as drought, and salinity (Chen et al., 2002; Kreps et al., 2002; Buchanan et al., 2005). Dehydration, salinity, low as well as high-temperature stresses lead to metabolic toxicity, membrane disorganization, generation of reactive oxygen species (ROS), inhibition of photosynthesis and altered nutrient acquisition (Hasegawa et al., 2000). At the molecular level, abiotic stress tolerance can be achieved through gene transfer by altering the accumulation of osmoprotectants, production of chaperones, superoxide radical scavenging mechanisms, exclusion or compartmentation of ions by efficient transporter and symporter systems (see reviews by Ingram and Bartels, 1996; Hasegawa et al., 2000; Apse and Blumwald, 2002; Zhu, 2002; Viswanathan and Zhu, 2004; Sangam et al., 2005; Valliyodan and Nguyen, 2006). In this article, we provide an account of the recent developments in the functional genomic approaches of abiotic stress tolerance including recently developed genetical genomic approaches. The stellar contributions of high-throughput functional genomics in revealing signaling cascades as well as regulatory mechanisms will be discussed. Further, we decipher components of tolerance mechanisms, their regulation and control of gene expression by signaling networks and transcriptional regulators. 2. Functional genomic approaches role in elucidating abiotic stress tolerance mechanisms In response to osmotic stress such as drought and salinity, plants respond and adapt to environmental stresses by altering thousands of genes, as a result cellular, physiological and biochemical processes will be modified. Transcriptome analysis

using microarray technology offers a powerful platform to find out the expression of genes during osmotic stress at the global level. This led to a cascade of challenges that occur at molecular, metabolic, physiological, cellular and morphological levels. This perhaps would not have been possible without technical advancement in the realms of (a) gene discovery, (b) highthroughput gene expression, (c) altering gene expression by transformation technologies, (d) functional characterization of genes of interest via high-throughput gene inactivation techniques, and (e) genetical genomic approaches. The role of different disciplines of functional genomic approaches in crop improvement for stress tolerance is shown in Fig. 1 and details are discussed in the following sections. 2.1. Gene discovery An important genomic approach to identify abiotic stressrelated genes is based on ESTs generated from different cDNA libraries representing abiotic stress treated tissues collected at various stages of development. Detailed information on the type of libraries and number of ESTs generated from each library of various abiotic stress-tolerant species is indexed at the National Center for Biotechnology Information (NCBI) dbEST (http:// www.ncbi.nlm.nih.gov/dbEST/dbEST_summary.html). In order to enrich plant EST datasets with stress-responsive genes, specific sequencing programs based on cDNA libraries from stress-treated plant tissues and organs of diverse species at many developmental time points are necessary. Since EST data sets generated from control as well as stress-treated tissues are derived mostly from non-normalized cDNA libraries, counting the abundance of a particular gene provide information on relative expression levels of stress-responsive genes. In addition, the clustering of EST sequences generated from abiotic stress-treated cDNA libraries provides information on gene number, gene content and possible number of gene families involved in stress responses. Putative functions are assigned to such stress-responsive genes by BLASTX comparison to the Swissprot database. This type of analysis provides a valuable resource of information regarding a gene index

Fig. 1. Schematic representation of multidisciplinary genomic approaches for abiotic stress tolerance.

N. Sreenivasulu et al. / Gene 388 (2007) 1–13

associated with stress-responsive genes among various species. Further, the clustering data yields consensus sequences that provide a much cleaner data set than typical EST data. Outcome of such studies indicates that unknown genes still represent a very high percentage (20–30%) in all cDNA libraries of stresstreated plants. They need to be annotated in order to find possible functions and to get a comprehensive picture of the tolerance mechanisms. Recently, an attempt was made to identify abundantly expressed ESTs in libraries of a salt-treated halophyte Thellungiella halophila (Wang et al., 2004a,b) as well as from monocots like barley, wheat, maize and rice (Sreenivasulu et al., 2004a). Analyzing the various EST collections enabled us to find stress-regulated genes. Further, these data should also assist in unraveling the underlying regulatory and metabolic networks. 2.2. Identification of a stress-responsive gene by transcript profiling In contrast to digital in silico quantification of expression levels based on EST counts, approaches such as SAGE, MPSS, array-based transcript profiling technologies and quantitative real time PCR (qRT-PCR) allow us to perform an assessment of high-throughput expression of thousands of genes in control and stress-treated tissues at various developmental stages. Insights into gene expression patterns and functions coupled with stress tolerance can be explored by EST-based cDNA arrays. Gene expression profiling using cDNA macroarrays (Sreenivasulu et al., 2006) or microarrays (Chen et al., 2002) are novel approaches to identify higher number of transcripts and pathways related to stress tolerance mechanisms than before. There are several studies reported related to abiotic stress transcriptome profiling in model species such as Arabidopsis and rice that have revealed several new stress-related pathways in addition to the previously well described stress-related genes (Desikan et al., 2001; Chen et al., 2002; Kreps et al., 2002; Seki et al., 2002a; Oh et al., 2005). 2.2.1. Comparative transcriptome response of acquired desiccation tolerance in seeds and dehydration response in vegetative tissues Most of the recent emphasis has been made on dissecting the mechanisms of dehydration responses in vegetative tissues triggering gene expression associated with desiccation tolerance in an ABA-dependent manner via ABA-responsive element binding factors (ABF), MYC and MYB transcription factors and in an ABA-independent manner via drought-responsive element binding factors (DREB) (Shinozaki and Yamaguchi-Shinozaki, 2000). As a recent outcome of transcriptome studies it is apparent that ABA is not only involved in drought-specific responses but also there is a cross-talk in cold and salinity stress responses (Seki et al., 2002a,b). Here we are making a particular attempt to summarize results from one of our laboratories regarding the identification of regulatory genes involved in desiccation tolerance program during natural development as a part of maturation program in seed embryos which become desiccation tolerant. Signaling to switch on the desiccation

3

tolerance program in developing seed embryos occur via abscisic acid by activating genes involved in ABA biosynthetic and signaling pathway (Sreenivasulu et al., 2006). The transcriptome analysis and prediction of cis elements in tightly co-expressed gene set indicates that embryo-specific gene expression patterns show peak of expression during late maturation delineate a regulatory network for the acquisition of desiccation tolerance, which are controlled by ABAmediated signal transduction via ABF and in an ABAindependent manner via DREB 2A transcription factor. Further, the same coupling regulatory factors ABF and DREB triggers oleosin and lipid biosynthesis genes which are co-expressed with desiccation-related LEA genes. Such comparative studies confirm the participation of well known regulators described in vegetative tissues during dehydration responses also in the natural process of desiccation tolerance in seeds. Transcriptome data covering natural acquisition of desiccation tolerance during embryo maturation and transcriptional changes occurring in 3 mm long radicles of Medicago truncatula seeds leading to drought tolerance were compared recently. The protective mechanisms had clear overlap of ABA-dependent and ABAindependent regulatory pathways involved in both drought and desiccation tolerance (Buitink et al., 2006). 2.2.2. Genotype specific transciptome responses from salt stress treatments Genes responding to a particular stress vary between species and even genotypes due to the fact that certain genotypes have efficient signal perception and transcriptional changes that lead to successful adaptations and eventually tolerance. Kawasaki et al. (2001) reported large-scale gene expression profiling in the salt-tolerant rice variety Pokkali as well as in the saltsensitive variety IR29 at 15-min to 7-day time intervals under control and high salinity conditions. These authors concluded that the tolerant cultivar responded at the level of transcription after 15 min of salt stress and displayed up-regulation of many genes encoding glycine-rich proteins, ABA and stress-induced proteins, metallothionein-like proteins, glutathione S-transferase, ascorbate peroxidase, water channel protein isoforms, subtilisin inhibitor, tyrosine inhibitor and others. Many transcripts that were up-regulated in the tolerant cultivar responded more slowly in the sensitive cultivar during salt treatment. In another study, where cDNA subtraction between sensitive and tolerant rice cultivars was done, about 1266 clones were obtained. And out of the 86 clones that were sequenced, 22 clones were similar to earlier reported stress-responsive genes, and 36 clones were novel (Sahi et al., 2003). Similarly, when the gene expression patterns in tolerant and sensitive seedlings of foxtail millet (Setaria italica L.) exposed to 250 mM NaCl were monitored, it was found that 14 unique ESTs up-regulated in the salt-tolerant foxtail millet line were identified under prolonged salt stress and found to be similar to that identified in rice (Sreenivasulu et al., 2004b). In conclusion, among cereal plants during long-term abiotic stress treatments, protease inhibitors, stress proteins, aquaporins and antioxidant components were induced and expected to impart tolerance during various abiotic stress treatments.

4

N. Sreenivasulu et al. / Gene 388 (2007) 1–13

2.2.3. Transciptome studies of abiotic stress responses Differences in gene expression during abiotic stress responses such as drought, salinity, cold and high temperature varies to the type and extent of stress. In the model plant Arabidopsis deeper insights were gained into functional genomic aspects of multiple stress interactions. Using 1300 full-length clones (Seki et al., 2001) and 7000 full-length clone inserts (Seki et al., 2002a,b) multi-stress interactions of abiotic stress treatments were studied to overlapping responses as well to identify genes of potential interest to salt, drought and cold responses. By using 1300 full-length clones, Seki et al. (2001) identified a set of only 44 and 19 genes, which were induced either by drought or cold stress response, respectively. By using 7000 full-length inserts, 299 drought-inducible genes, 213 high salinity-stress-inducible genes, 54 cold-inducible genes and 245 ABA-inducible genes were identified (Seki et al., 2002a,b). Multi-stress interactions of abiotic stress treatments were studied by Kreps et al. (2002) using a larger array containing oligonucleotides for about 8100 Arabidopsis genes, to identify genes of potential role in salt, drought and cold responses. They identified changes in gene expression (more than 2-folds over control) for 2409 out of 8100 genes as part of cold, drought and salt responses. The results obtained from plant abiotic stressrelated transcriptome studies are difficult to compare even among related species due to the fact that stress treatments were performed in different tissue types, during different time course, type and design of array. 2.2.4. Genome-wide transcriptome studies reveal molecular cross-talk of gene regulatory networks among abiotic stress treatments An important aspect of the current research interest using functional genomic tools is the identification of key regulators based on gene expression patterns related to multi-stress interactions. Genome-wide transcriptome analysis has identified hundreds of genes encoding transcription factors that are induced or repressed by many environmental stresses (Chen and Zhu, 2004). The expression patterns of these transcription factors are highly complex and they suggest that stress tolerance and resistance are controlled at the transcriptional level by an extremely intricate gene regulatory network. Chen et al. (2002) identified groups of transcription factors regulated, (a) specifically by abiotic stress (class I) and (b) both by biotic and abiotic stresses (class II) in Arabidopsis. Among the class I group, approximately 20 genes were preferentially induced by abiotic stresses such as salinity, osmotic, cold and jasmonic acid treatments. These transcription factors include DRE/CRT binding factors activated by cold stress, CCA1 and Athb8 (regulated by hormones, Baima et al., 2001), Myb proteins as well as bZIP/HD-ZIPs and AP2/EREBP domain proteins (Kizis et al., 2001). Further, Seki et al. (2002a) employed a full-length cDNA microarray containing 7000 independent Arabidopsis cDNAs to identify cold, drought and salinity-induced target genes and stress-related transcription factor family members such as DREB, ERF, WRKY, MYB, bZIP, helix-loop-helix and NAC. These results indicate that there is a greater cross-talk between salt and drought stress signaling processes in com-

parison to salt and cold stresses. Fowler and Thomashow (2002) revealed rapid expression of CBF1 (DREB1b), CBF2 (DREB1c) and CBF3 transcripts (DREB1a) during short-term cold acclimation transcriptome studies of Arabidopsis. Six additional long-term up-regulated genes encode transcription factors: putative zinc finger protein (At4g38960), R2R3-Myb transcription factor AtMYB73, H-protein promoter binding factor 2a (AF079503), the HD-Zip protein AthB-12, and two AP2 domain proteins, RAP2.7 and RAP2.1. Similarly, transcriptome response to dehydration, salinity and ABA has been monitored in sorghum seedlings and identified approximately 22 transcription factors (Buchanan et al., 2005). These regulators include ABF from bZIP factors, DREB from AP2/EREBP family, HD-ZIP and MYB factors, which are also known to be stress-responsive in other model species such as Arabidopsis and rice. Also, there is a greater need to verify the roles that these transcription factors play in the networks for better designing plants that can tolerate a variety of environmental stresses. 2.2.5. Complex transcriptional responses in transgenic plants conferring abiotic stress tolerance Revealing complex transcriptional responses in transgenic plants or knockout mutants conferring abiotic tolerance is a useful method to identify gene interactions and downstream elements. Transcript profiling (1300 genes) of Arabidopsis plants overexpressing DREB1a identified 12 genes as cold and drought-inducible target genes of the DREB1 transcription factor family (Seki et al., 2001). In an extensive analysis, Fowler and Thomashow (2002) analyzed CBF1 (DREB1b), CBF2 (DREB1c) and CBF3-expressing transgenic plants (DREB1a) and identified 41 downstream genes as CBF targets, among them AP2 transcription factor (RAP2.6) is subregulon of CBF regulon. In a recent attempt from Thomashow's group, a total of 514 CBF2 target genes were identified as cold-responsive gene set using 24 K Arabidopsis Affymetrix Genechip array (Vogel et al., 2005). The transcription factors co-regulated along with CBF2 includes ZAT12, ZAT10, RAV1, MYB73, cold induced zinc finger proteins (CZF1 and CZF2), whose transcription peaks within 1 h of low temperature treatment. Among them CBF2 and ZAT12 regulons turned out to control transcription of approximately 92% of the most highly induced genes to cold stress response and in turn both regulators shown to increase freezing tolerance in Arabidopsis (Vogel et al., 2005). Similarly, Maruyama et al. (2004) performed transcriptome analysis in transgenic plants overexpressing DREB1a/CBF3 conferring strong tolerance to freezing stress. They identified 38 upregulated genes encoding cor 15a, cor 15b, cor 47 from LEA/ Dehydrin genes, known to participate in freezing tolerance, and C2H2 Zinc finger and AP2 transcription factors as CBF3 targets. This was done using 7000 full-length RIKEN microarray and 8000 affymetrix Arabidopsis gene chip which further verified their regulation under cold stress. Further, they found enriched presence of DRE/C-repeat cis elements in the promoters of target genes of CBF3. Recently, Zhu and his group revealed interesting regulatory interconnections between early and late cold responses by revealing statistically significant differences

N. Sreenivasulu et al. / Gene 388 (2007) 1–13

between transcriptome of inducer of CBF expression1 (ice1) and corresponding wild type plants (Lee et al., 2005). As expected, CBF3 transcript levels were significantly lower in ice1 mutant until 24 h of cold treatment due to the fact that bHLH ICE1 transcription factor known to bind CBF3. In addition, 204 of the 939 cold-regulated genes were affected in ice1. Among them, key regulators belonged to AP2 transcription factor family followed by bZIP, MYB and bHLH (Lee et al., 2005). Although DREB1/CBF is characterized to function in cold stressresponsive gene expression and DREB2 transcription factor in drought-responsive gene expression, until recently it was not clear how DREB1 and DREB2 activate cold and drought specific set of genes by binding to the same dehydrationresponsive element (DRE/C-repeat element TACCGACAT). Recently, Sakuma et al. (2006) addressed this interesting issue by performing transcriptome studies in overexpression lines of DREB2 and DREB1 transgenic plants and identified only 8 genes in common. The remaining 14 genes are probable targets of DREB2a which consists of at least 9 LEA members thought to confer dehydration tolerance. Further, by performing promoter analysis and gel mobility shift assay of the DREB2a and DREB1a up-regulated genes authors concluded that the DREB2a and DREB1a proteins have different binding specificities, therefore genes downstream of DREB2a and DREB1a trigger different set of genes conferring drought and freezing tolerance, respectively (Sakuma et al., 2006). 2.2.6. Outlook from transcriptome studies of abiotic stress tolerance Comparative analysis of the response to abiotic stresses among diverse tolerant species can lead to the identification of evolutionarily conserved and unique stress defence mechanisms. As large-scale gene expression data sets become available from the analyses of stress responses of various plant systems, it is important to set up effective clustering algorithms and bioinformatic tools for the identification of the stress regulons (for technical details see the review Sreenivasulu et al., 2002). By applying clustering algorithms to large-scale gene expression data of abiotic stress responses, stress regulons, i.e. sets of genes regulated in a similar fashion, can be identified. Genes up-regulated in Arabidopsis to saline, drought and cold stress conditions are enriched with DRE-related core motif and abscisic acid-response elements (Seki et al., 2001, 2002a). In general, promoters of several stress-responsive genes from various species such as rice, maize and Arabidopsis harbour DRE, ABF, GCC-box and w-box cis elements (Chen et al., 2002; Kizis and Pages, 2002; Seki et al., 2002a; Dubouzet et al., 2003). This approach also enables the identification of new promoter elements/transcription factor binding sites in coexpressed gene sets among cereals and other tolerant model systems, which in turn helps to explore regulatory networks controlling abiotic stress responses. In summary, genome-wide transcriptome analysis resulted in identifying transcription factors that are induced or repressed to a range of abiotic stresses. However, mere expression patterns do not allow verifying the roles of these co-expressed transcription factors, which acts in certain networks. In order to obtain a compre-

5

hensive picture of complex regulatory networks, along with gene expression profiling data of transcription factors, yeast two-hybrid data, genome-wide location data in combination with computational methods are necessary to identify the interacting partners. Such studies are expected to yield more valuable insights into the regulation mechanisms of abiotic stress tolerance. Identifying transcription factors and their target genes responding to abiotic stress responses is one paradigm among multiple level interactions. Other such levels of regulations of gene expression include the role of microRNA and epigenetic-based chromatin remodelling and modification, which is not the focus of this review. To achieve a more holistic picture of stress tolerance mechanisms, it is important to combine recently developed genetical genomic approaches along with functional genomic approaches such as transcriptome, proteome, and metabolome studies. A final test of the validity of the developed concepts is either testing the putative function of the genes by forward genetics or by the transfer of several of the stress-responsive genes into crops to enhance the tolerance levels. 2.3. Functional aspects of stress tolerance mechanisms identified through transgenic approaches Based on their functions, we divided the genes that are associated with abiotic stress tolerance into 4 groups such as osmolyte biosynthesis, antioxidant protectants, protection of cell integrity and ion homeostasis (Sangam et al., 2005; Valliyodan and Nguyen, 2006). Research in genetic engineering for stress tolerance was limited in the pre-genomics era by the inadequate availability of genes and specific promoters (Zhu et al., 1997). It is now possible to study many genes simultaneously on a genome-wide scale with respect to their structure and function. Thus, the current trend in stress-biology is to use large-scale genomics data to scrutinize and revalidate the protective mechanisms which were described based on transgenic approaches (sugar alcohol biosynthesis, osmolyte biosynthesis, antioxidants, LEA proteins, molecular chaperones, cell membrane proteins, aquaporins, ion homeostasis and transcription factors). Thus, there is a good correlation among the gene sets identified by traditional and genomic studies. Moreover, many important components identified for abiotic stress tolerance were tested by transgenic technologies and confirmed the roles of these genes associated with the biosynthesis of osmolytes, stress proteins, antioxidants, aquaporins and ion homeostasis (Kishor et al., 2005; Sangam et al., 2005). Based on summarized results from one of our laboratories, it is apparent that genetic engineering of glyoxalase pathway (glyoxalase I and II) contributing to glutathione-based detoxification resulted in limited yield penalty under extreme salinity (Singla-Pareek et al., 2003) and heavy metal stress conditions (Singla-Pareek et al., 2006). Other proteins which have influence on DNA metabolism like helicases have shown to overcome salinityinduced reduction in plant productivity and yield in transgenic tobacco plants (Sanan-Mishra et al., 2005). With a special focus, we discuss here the success of implying regulatory genes in transgenic approaches.

6

N. Sreenivasulu et al. / Gene 388 (2007) 1–13

2.3.1. Engineering transgenic abiotic stress-tolerant plants with transcription factors Based on molecular and genomic studies several key transcription factors were identified to be induced under abiotic stress treatments. Among them DREB and ABF are well characterized transcription factors known to play an important role in regulating gene expression in response to abiotic stresses via ABA-independent and ABA-dependent manner. Kasuga et al. (1999) overexpressed the cDNA encoding DREB1a under the control of 35S promoter in transgenic Arabidopsis plants. As a result, many of stress tolerance genes such as cor, erd, P5CS and rd29 are expressed under normal growing conditions and resulted in improved tolerance to drought, salinity, and freezing stress. However, constitutive expression of DREB1a resulted in severe growth retardation under normal growing conditions. In contrary, expression of DREB1a gene under the control of a stress-inducible promoter rd29A gave rise to minimal effects on plant growth under normal growing conditions and provided even greater tolerance to abiotic stress treatments. Oh et al. (2005) ectopically expressed Arabidopsis DREB1a (CBF3) in transgenic rice plants under the influence of constitutive promoter. The authors report that DREB1a transgenic rice plants show enhanced tolerance to drought and salinity but to a little extent for low temperature stress with no visible stunted phenotype despite its constitutive expression. These results show controversies to Arabidopsis results in relation to freezing tolerance as reported earlier from Kasuga et al. (1999). This could be partly explained due to finding expression of different sets of stress-related genes such as Lip5, Dip1, PSLS, HSP70, PP2Ca, Rab21 in DREB1a transgenic rice plants (Oh et al., 2005), which are known to enhance osmotic stress tolerance. In another attempt, Ito et al. (2006) analyzed OsDREB1 transgenic rice plants conferring improved tolerance to drought, salt and low temperatures and identified large portion of stress-inducible genes. These results confirm that DREB1/CBF cold-responsive pathway is conserved in rice and Arabidopsis. Overexpression of CBF1 (DREB1b) in Arabidopsis and Brassica induces cor genes in both species (Jaglo et al., 2001; Thomashow, 2001) but not in tomato (Hsieh et al., 2002). Hsieh et al. (2002) reported improved drought, chilling and oxidative stress tolerance of tomato plants expressing Arabidopsis CBF1. ABA-responsive element binding factor (ABF) members belong to bZIP transcription factor family, which show distinct roles in sugar, ABA and stress responses (Uno et al., 2000). To reveal the involvement of ABFs in stress tolerance, Kang et al. (2002) generated ABF3 and ABF4 transgenic Arabidopsis lines by overexpressing them constitutively. ABF3 and ABF4 overexpression lines exhibited an altered transpiration rate in response to water deficit conditions, eventually all transgenic plants survived a 12-day drought treatment and set seed in contrast to wild type plant with 33% survival rate. Both ABF3 and ABF4 overexpression lines showed induction of ABAsignaling ABI1, ABI2 phosphatase and other stress-responsive genes including desiccation-related LEA genes via ABAdependent pathway (Kang et al., 2002). In contrast, both ABF3 and ABF4 transgenic lines were hypersensitive to salinity

treatments during germination. It is interesting to note that although both ABF4 and ABF3 play an essential role in drought tolerance, yet their constitutive overexpression resulted in stunted phenotype in ABF4 line and no visible growth inhibitions in ABF3 overexpression lines in comparison to wild type plants (Kang et al., 2002). Similarly, constitutive overexpression of ABF3 in rice also did not show visible phenotypic growth alterations and overexpression line showed drought tolerance by activating specific groups of stressregulated genes Wsi18 and Rab21 (Oh et al., 2005). To investigate the function of ABF2, Kim et al. (2004) generated ABF2 transgenic Arabidopsis plants under the control of 35S promoter. ABF2 overexpression lines exhibit seedling growth inhibition due to hexokinase HXK1 and HXK2 expression inhibition, which was eventually relieved by the addition of sucrose suggesting the involvement of ABF2 in sucrose response. In contrast to ABF3 and ABF4, the two-week old ABF2 transgenic plants showed only 10–15% survival during drought stress and exhibited higher salt tolerance. However, when 3-week old ABF2 transgenic plants were tested for drought tolerance, nearly 97% of the transgenic plants survived in comparison to only 22% survival rate seen in wild type plants (Kim et al., 2004). So far, ABF1 overexpression or knockout phenotypes were not described in Arabidopsis. Recently, Furihata et al. (2006) demonstrated that overexpression of Arabidopsis phosphorylated active form of AREB1 (ABF2) resulted not only induction of stress-responsive RD29B gene but also seed-specific genes without ABA treatment. The germination efficiency of 35S:AREB1pa plants are reduced, probably due to interference of ABA in seed germination events. Authors also demonstrated that SnRK2 protein kinase (42 kDa) phosphorylates AREB1 protein at R–X–X–S/T site. Among the bZIP family, a pathogen-induced pepper bZIP transcription factor (CAbZIP1) was constitutively overexpressed in transgenic plants, which exhibited resistance to Pseudomonas syringae, drought and salt tolerance during all stages of plant growth (Lee et al., 2006). One important way of achieving tolerance to multiple stress conditions is to overexpress transcription factor genes that control multiple genes from various pathways. Although important role of well characterized DREB and ABF stressresponsive transcription factors was revealed recently in conferring abiotic stress tolerance by transgenic approaches (see above), very little is known about the function of other major families of transcription factors that acts under the influence of ethylene, jasmonic acid, salicylic acid and other phytohormones conferring abiotic stress tolerance. Ethyleneresponsive element binding proteins (EREBP) transcription factor belongs to AP2 type family, which are known to act under the influence of ethylene that mediates plant responses to biotic and abiotic stresses. Park et al. (2001) identified stress-induced EREBP transcription factor (Tsi1), known to bind GCC and DRE/CRT boxes, in tobacco, which was induced by ethylene, methyl jasmonate, salicylic acid, salt treatments and wounding. Overexpressed transgenic lines of tobacco showed enhanced salt tolerance and resistance to pathogens (Park et al., 2001). In another attempt, Ethylene Response Factor (ERF), a member of

N. Sreenivasulu et al. / Gene 388 (2007) 1–13

EREBP transcription family from tomato known to be responsive to jasmonate and ethylene was constitutively overexpressed in transgenic tobacco, and found to confer enhanced tolerance to salt and pathogens by activating expression of pathogen-related genes (Wang et al., 2004a). Based on these results, it is apparent that some of the stress-responsive EREBP family members are potential linkers in ethylene and osmotic stress signaling. 2.3.2. Engineering transgenic abiotic stress-tolerant plants with signaling pathway genes Stress perception and signaling pathways are critical components of adaptive response that is vital for the plant species to survive under extreme environmental constraints (drought, salt and extreme temperatures). Osmotic stress and associated oxidative stress are a common consequence of such a stress exposure and share one or more intermediates/components or outputs as part of their signaling (Viswanathan and Zhu, 2004). It is also highly desirable to have stress signaling sensors that can transduce the signal to the target cells. Although signal perception and transduction pathway genes are attractive targets for genetic engineering, a detailed knowledge of cascades of signal perception and transduction activated under abiotic stress response is lacking. Nevertheless we made an attempt to summarize few initial reports made in connection with signaling. The two-component system consists of sensory histidine kinase and response regulator functions as stress sensors in bacteria and yeast (Aguilar et al., 2001). Although direct evidence has not been given to SLN1 homologue of A. thaliana histidine kinase (AtHK1), it is thought to be a candidate sensor, as it is upregulated during salt and low temperature stresses (Urao et al., 1999). By transferring A. thaliana histidine kinase (AtHK1) in yeast double mutant sln1/sho1 lacking both yeast sensors, double mutant has been successfully recovered (Urao et al., 1999). In addition involvement of receptor-like kinases in osmosensing has been suggested. Receptor-like protein kinase (NtC7) induced under abiotic stress response has been shown to confer osmotic stress tolerance in overexpressing transgenic tobacco plants (Tamura et al., 2003). Mitogen-activated protein kinase (MAPK) cascades seem to be the convergent points for crosstalk and it has been shown that signals perceived by 60 MAPKKKs have to be transduced through 10 MAPKKs to 20 MAPKs, providing scope for cross-talk (Ligterink and Hirt, 2000). Constitutive expression of MAPKKK/Nicotiana protein kinase 1 (MAPKKK/NPK1) in maize activated an oxidative signal cascade and the transgenics showed tolerance to cold, heat, salinity and also higher photosynthetic rates (Shou et al., 2004). These results clearly demonstrate that NPK1 somehow protects the photosynthetic machinery. In addition, activation of MAPK cascade in response to cold and salt stress has been identified in Arabidopsis by yeast two-hybrid interactions, complementation of osmosensitive yeast mutants and by overexpression studies in plants (Teige et al., 2004). Overexpressing Arabidopsis MAPK kinase 2 (MKK2) plants resulted in the constitutive upregulation of downstream MPK4 and MPK6 activity and other 152 genes involved in stress metabolism, signaling and transcriptional regulation. While

7

overexpression of MKK2 resulted in salt and freezing tolerance, on the contrary mkk2 null mutant plants show hypersensitivity to salt and cold stress (Teige et al., 2004). These results again suggest that MAPK cascade is an important convergent point for cross-talk between different abiotic stress responses. Interestingly, calcium-dependent protein kinase (CDPK) has been identified as an important component of osmotic signaling pathways. Rice CDPK7 gene has been overexpressed and shown to be a positive regulator in triggering stress-responsive genes in response to salt/drought, however transgenic plants confer tolerance to cold, drought and salinity stress (Saijo et al., 2000). These results again suggest that CDPK signaling pathway follows at least two distinct pathways one specific to osmotic responses and other to cold stress response. As indicated above, until now we have only fragmentary knowledge about the abiotic stress signaling pathways, which could be dissected in a systematic manner by forward and reverse genetics approach. 2.4. High-throughput gene inactivation techniques to identify putative functions of genes involved in abiotic stress responses Even though a growing number of genomic-assisted programs continue to be developed, the identification of abiotic stress-responsive candidate genes is still at large. Interestingly, more than 40–50% of identified stress-responsive gene functions remain to be characterized. In order to reveal their putative functions involved in abiotic stress tolerance, various highthroughput methods were developed for the confirmation and validation of gene function by gene inactivation. There are two main complementary approaches developed for identifying mutations in target genes, namely TILLING (Targeted Induced Local Lesions In Genomes) and T-DNA insertion mutant lines. Using these techniques various attempts were made in order to show the putative functions of abiotic stress-responsive genes in Arabidopsis, rice, maize and barley. In Arabidopsis (ecotype C24) 43,000 T-DNA insertion lines were generated (Weigel et al., 2000; http://stress-genomics.org/stress.fls/tools/mutants. html), of which about 30,000 lines were screened for stressrelated gene regulation mutants (Xiong et al., 1999). By conventional genetic screen, Arabidopsis sos mutants showing hypersensitiveness to growth inhibition by NaCl stress were identified (Zhu, 2001). The sos mutant growth is impaired on media that are deficient in K+ and particularly hypersensitive to Na+ and Li+ ions (Zhu et al., 1998). Based on recent biochemical and physiological data, the role of sos1 in K+ acquisition is indirect and SOS1 gene is identified as Na+/H+ antiporter, which maintains low concentration of Na+ in cytoplasm by pumping Na+ ions into acidic vacuoles (Shi et al., 2000). Subsequently, another locus, SOS2, and a third salt-tolerant locus, SOS3, were identified in Arabidopsis by Zhu et al. (1998). SOS2 locus was also necessary for K+ nutrition since sos2 mutants could not be grown in a medium with low potassium and recently been identified as a serine/threonine type protein kinase with an Nterminal catalytic domain similar to that of the yeast SNF1 kinase (Liu et al., 2000). High concentration of Na+ elicits a cytoplasmic calcium signal, which will be perceived by SOS3, a calcium binding protein. SOS3 interacts with SOS2, a protein

8

N. Sreenivasulu et al. / Gene 388 (2007) 1–13

kinase, and the regulatory domain is located in SOS2 gene. The calcium-dependent kinase pathway of SOS3–SOS2 activates the Na+/H+ antiporter gene under excess of Na+. It, therefore, appears that SOS1, SOS2 and SOS3 encode regulatory components controlling K+ nutrition essential for ion homeostasis in plants under stress. Recently, using 22 K Agilent oligoarrays sos2 and sos3 mutant transcriptome studies were performed (Kamei et al., 2005). The genome-wide transcription response of sos2 differed significantly from sos3. In addition, major transcription factor families such as AP2, MYB and WRKY were induced in the absence of SOS2. Interestingly stress-inducible genes such as RD29A, RD17, RD22 and KIN1 did not show any affect in its transcriptional levels in both sos2 and sos3 mutants. In addition, Shi et al. (2002) isolated sos4 mutant from Arabidopsis (hypersensitive to Na+, K+, and Li+ ions) and demonstrated that SOS4 encodes a pyridoxal kinase that is involved in the biosynthesis of pyridoxal-5-phosphate, an active form of vitamin B6. Genetic screening for identifying loci associated with abiotic stress responses in signaling did not yield major success due to the limitations of not finding major visible phenotypes and better screening systems. To tackle these problems Zhu et al. developed an elegant screening approach first by generating transgenic plants by expressing the luciferase coding sequence under the control of the stress-responsive RD29A promoter and eventually treated the seeds of these plants with ethyl methanesulfonate mutagen to generate a large number of mutants. By looking for the alteration in luciferase expression pattern in mutants under various abiotic stress conditions three major groups of mutants (los—low expression of osmotically responsive genes, cos—constitutive expression of osmotically responsive genes and hos—high expression of osmotically responsive genes) were identified and presented genetic analysis of osmotic and cold stress signal transduction pathways mediated by ABA-dependent and ABA-independent pathways (Ishitani et al., 1997). From these mutants remarkable insights were gained in ABA signal transduction network during drought and cold responses. The los5-1 and los5-2 mutants are identical to aba3 identified locus and exhibit reduced tolerance to drought, salt stress and freezing with reduced expression of RD29A-luciferase expression and other stress genes COR15, COR47, P5CS and RD22 (Xiong et al., 2001a). The LOS5 gene was cloned and found to encode molybdoprotein cofactor sulfurylase, an enzyme required by aldehyde oxidase for ABA biosynthesis (Xiong et al., 2001a). In fiery1 mutant, superinduction of RD29A-luciferase expression in response to exogenous ABA, low temperature and osmotic stress shown to encode inositol polyphosphate 1-phosphatase involved in phosphoinositol metabolism, which in turn suggests a positive role for phosphoinositols in ABA and stress signaling (Xiong et al., 2001b). Among the hos class of mutants hos5 showed osmotic stress specific response (Xiong et al., 1999) and hos1 and hos2 mutations showed affect only in cold stress response (Lee et al., 2001). With an intention to identify mutants show disrupt in the regulatory genes involved in drought response, another elegant imaging screening system has been developed to screen the

mutagenized population of Arabidopsis to track the altered stomatal movements in mutants under drought response in comparison to normal wild type plants (Merlot et al., 2002). As an outcome of genetic screening at least 4 mutants affected in ABA biosynthesis (aba1, aba2, aba3 new allele of aao3) and 3 others involved in ABA signaling (abi1-1, ost1, ost2) have been identified (Merlot et al., 2002). One interesting outcome of these studies is the identification of OST1 as ABA-activated SnRK2 kinase known to act downstream of ABI1 required for dehydration stress signaling (Mustilli et al., 2002; Yoshida et al., 2002). Conducting genome-wide transcription profiling studies from such mutants provides clues for the identification of target genes of ABA signaling. In an attempt to look for new targets of ABA signaling, genes were identified by SAGE analysis using abi1-1 mutant and identified ABI2, cascade of other transcription factors, ABA-responsive ribosomal proteins and ubiquitin– proteosome complex as potential targets of abi1-1 (Hoth et al., 2002). Besides these, quite a number of T-DNA insertion-based mutants have been identified from ABA biosynthesis and signaling cascades. In a reverse genetics approach, Umezawa et al. (2006b) screened T-DNA pools for higher ABA levels and identified cyp707a3-1 mutant, which showed reduced transpiration rate and exhibited efficient ABA-inducible gene expression and enhanced drought tolerance. These results confirm the role of cytochrome P450 CYP707A family gene in participating ABA catabolic pathway. Mutant carrying T-DNA tag in protein phosphatase 2C (AtPP2CA) gene, which is characterized as negative regulator of ABA-signaling pathway showed ABA hypersensitivity during early growth of seed germination (Yoshida et al., 2005). 2.5. Genomics-assisted breeding — a way forward to develop stable abiotic stress-tolerant lines The conventional breeding programs are slow but in some cases have been successful in developing abiotic stress-tolerant lines; however in most of these cases the tolerant plant shows an inverse relationship with yield (Ceccarelli, 1987). With the advent of newly developed genomic resources, two major approaches could be used in exploiting the gene pool for imparting abiotic stress tolerance: first, identification of stresstolerant genes via functional genomic approaches and introduction of stress-tolerant genes into crops of interest (summarized above), and secondly, identification of QTLs/genes conferring tolerance to stress in germplasm collections, development of respective molecular markers and use in marker-assisted breeding programs. Although by using functional genomic approaches regulatory pathways involved in abiotic stress response has been dissected and shown to enhance abiotic stress tolerance in laboratory conditions by activating stress-responsive signal transduction and downstream transcription factor genes in transgenic plants, its success in field conditions are rather poor. Hence it is equally important to integrate developed knowledge as an outcome of functional genomics into a real knowledge-based breeding programs via genomics-assisted breeding to develop stable populations conferring both stress tolerance and yield. In this respect, we

N. Sreenivasulu et al. / Gene 388 (2007) 1–13

need to explore variation between species and accessions for abiotic stress tolerance (summarized for cereals by Nevo, 1992) and the identified tolerant accessions/wild species should be introgressed into elite genome showing higher yield under abiotic stress conditions. Such breeding material could be used in -omics technology to explore regulatory mechanisms controlling yield under stress conditions (Fig. 1). Wild species and landraces have unique resistance alleles that are not found in the cultivated gene pool and are known to be especially potent sources of abiotic stress tolerance traits (Ellis et al., 2000). Here we focus on cereal plants. Accordingly, wild barley possesses considerably more variation than the cultivated species and many alleles are associated with adaptation to specific environments. For instance, some of the wild species such as Hordeum jubatum, H. chilense and H. spontaneum were identified as tolerant to abiotic stress especially to salinity (Forster et al., 1997). Gorham et al. (1990) suggested that the D genome of hexaploid bread wheat provides a source of salt tolerance. In other species of the tribe Triticeae, Gorham et al. (1990) demonstrated that discrimination for K+/Na+ uptake is present in the U-genome of Aegilops spp., the R-genome of rye and the T-genome of Triticale. Likewise, the international maize and wheat improvement center (CIMMYT) has made considerable effort during the past 25 years to scrutinize and select tropical germplasm to improve drought tolerance in maize (Bruce et al., 2002). Thus, wild species offer a potentially rich source of resistance gene alleles for crop improvement. Therefore the genetic loci involved in the control of abiotic stresses in cultivated species can now be targeted by investigating in wild gene pool (Ellis et al., 2000). Keeping in view the importance of abiotic stress genes in wild species, progress was made in direction of producing salt-tolerant wheat/Thinopyrum bessarbicum amphiploids after transferring salt tolerance genes from T. bessarbicum to wheat (King et al., 1997). Currently, the following approaches can be used in breeding programs to utilize wild species for crop improvement in respect to abiotic stress tolerance. First, wild species may be backcrossed to an elite cultivar. At the end of three years, the best performing plants can be assessed for characters inherited from the wild donor species. In this approach, the discovery and transfer of desirable QTLs from unadapted to elite germplasm is simultaneous. The scheme was devised by Tanksley and Nelson (1996) and generally termed ‘Advanced Backcross QTL Analysis’. Molecular genetic analysis of such traits has usually been based on observable phenotypes, without knowledge of the trait-determining genetic architecture, and on anonymous polymorphisms linked to quantitative trait loci (QTL). Also in a very recent review, Tuberosa and Salvi (2006) summarized QTL studies related to drought in cereals and application of marker-assisted selection to introgress QTL alleles for drought tolerance. Dramatically, genomic approaches increased our ability to study genes and regulatory networks involved in abiotic stress responses (see above) and even QTL effects were tracked to a single gene level in response to salt tolerance (Ren et al., 2005) and submergence tolerance in rice (Xu et al., 2006). Currently, an approach called ‘Genetical Genomics’ has been developed based on gene expression profiling and marker-based

9

fingerprinting of related lines or individuals in a segregating population to analyze cis- and trans-acting factors and to delineate a trait-related genetic network (Jansen and Nap, 2001; Jansen, 2003). We have adopted and modified the genetical genomics concept of Jansen and Nap to identify sets of traitrelated genes and pathways controlling storage events in developing seed and presently extending this approach to decipher molecular regulatory networks underlying both tolerance and yield. Instead of a segregating population, we are using a set of introgression lines where an alien genome (in our case a wild barley line, Hordeum spontaneum HS213 conferring higher abiotic stress tolerance) is introgressed in the genetic background of an elite line (in our case Hordeum vulgare cv. Brenda, showing higher yield) [Li et al., 2005]. Such lines offer specific advantages to track characteristics for both yield and tolerance for crop improvement. Extensive expression profiling from these introgression lines are underway to detect expression QTLs (eQTLs) as originally proposed by Jansen and Nap (2001). Such data can be used to develop trait-linked molecular markers and direct transgenic approaches, both important components of molecular breeding strategies (Fig. 2). Although the ‘genetical genomics’ approach is still in its infancy, efforts are underway in this direction in some plant species. We believe that integration of information obtained from genomic approaches with genetic-based breeding will accelerate the success for different agronomic traits including abiotic stress tolerance. 2.6. Enhancing yield stability of crops that are challenged by abiotic stresses The impact of environmental stresses on crop yields is multidimensional and hence numerous studies were carried out in the model plant Arabidopsis and to some extent in other crop plants but mainly these studies were performed at the vegetative stage. As summarized above, partial tolerance was shown to be achieved in the vegetative phase through gene transfer by altering the ABA-dependent as well ABA-independent signaling cascades and in addition by manipulating accumulation of osmoprotectants, production of chaperones, protection of cell integrity by expression of LEA proteins, improved superoxide radical scavenging mechanisms and efficient exclusion or

Fig. 2. Proposed scheme of bridging the gap between breeding and genomic approaches for abiotic stress tolerance.

10

N. Sreenivasulu et al. / Gene 388 (2007) 1–13

compartmentation of toxic ions from cytoplasm (for recent reviews see Umezawa et al., 2006a; Valliyodan and Nguyen, 2006). Although such genetically modified plants may survive abiotic stress conditions, in majority of the cases their performance was not tested for yield stability under field conditions. Recently, Bahieldin et al. (2005) tested six independent wheat transgenic lines overexpressing barley HVA1 gene under field conditions for improved biomass productivity and grain yield under water deficit conditions. Wang et al. (2005) generated antisense transgenic Brassica napus plants with suppression of beta-subunit of Arabidopsis farnesyltransferase gene under drought-inducible rd29A promoter leads to significant reduction in water transpiration under drought stress. In addition, the three consecutive years of field evaluations of these transgenic plants confirm minimized yield loss by conferring tolerance to water deficit-induced seed abortion during flowering. These results confirm that genetic manipulation of the stress responses to ABA is successful. Similarly, overexpression of stress-induced NAC1 transcription factor (SNAC1) enhanced drought tolerance at the reproductive stage with 22–34% higher seed setting than control plants under field conditions (Hu et al., 2006). These transgenic plants seemed to loose water more slowly under water deficit conditions by closing stomatal pores, which hints that SNAC1 might participate in ABA responses. These reports are few successful field screening of plants and confirmation of transgenics for improved grain yield in comparison to wild types. Besides these initial success stories, there is a large need to understand the basic molecular mechanisms influencing grain yield under stress conditions. The genetic engineering of such key regulatory genes governing improved grain yield under severe abiotic stress environments appears to be one of the most promising approaches. Unlike salt stress, drought and temperature stresses might occur in combination at any stage of plant development, which is a critical factor influencing reduced grain weight and yield loss due to a decrease in photosynthetic efficiency and altered remobilization processes in source–sink relationships. Shortage of assimilates and sometimes nitrogen availability is a major cause of arrested grain development, and as a result duration of grain filling is reduced. If the rate of grain filling is not adjusted upward, final grain weight is reduced. An alternative source of assimilates are pre-anthesis stem reserves in the form of sugars, starch or fructans, which constitute a buffer in case source capacities are reduced as a result of drought-induced senescence. These reserves are readily utilized for grain filling and their availability may become a critical factor in sustaining grain filling and grain yield under drought stress (Yang and Zhang, 2006). Zinselmeier et al. (1999) demonstrated that lowering of water potential by drought stress during pollination resulted in a drastic decrease in kernel number. Base on in vitro experiments, infusion of sucrose into the stem results in improved grain set of maize under drought stress, leading to the conclusion that sugar starvation could be an important factor affecting grain set and ovary abortion under stress. To understand maize kernel response to drought during early seed development, tissuespecific transcription profiling was performed using 2500

unigene array of maize and identified a large number of differentially expressed genes in placenta under water deficit conditions (Yu and Setter, 2003). In addition, recent physiological evidences reported from cereals such as wheat and rice provide hints that the presence of a higher ABA content in superior grains is correlated with efficient seed filling by optimizing faster remobilization events (Yang et al., 2001; Yang et al., 2006). However, no detailed molecular-physiological data have been reported in this context. To this end, we propose 3 main perspective points for yield improvement under challenging environments of abiotic stress. 1. To address the basic question of assessing yield under abiotic stresses, new insights need to be gained at the molecular mechanisms of source–sink relationships by an integrated approach of genomics and breeding. We propose the following integrative scheme: (i) exploiting the untapped genetic reserves of crop plants (introgression lines of BC3 DH lines, near-isogenic lines, recombinant-inbred lines or pre-selected gene bank material imparting abiotic stress tolerance) for identifying genetic networks controlling agronomically relevant seed traits under drought/temperature stress by OMICS technologies, (ii) validating the deduced genetic networks by genetical genomic approaches and characterizing pre-selected lines for future molecular breeding programs, (iii) using the molecular knowledge to generate transgenic plants with improved seed yield under extreme stress treatments and (iv) as a proof of concept assessing yield as a selection criterion for abiotic stress resistance (refer to Figs. 1 and 2). 2. In depth understanding about plant characters such as root architecture and plasticity under drought conditions especially in agronomically superior genotypes is of paramount importance to the molecular biologists to improve the yield parameters. Similarly, attention should be drawn that developmental traits such as flowering and fruiting are not affected in the field by any one of the abiotic stresses or a combination of them. 3. It is known that there is always a co-occurrence of several abiotic stresses rather than a single stress in the field condition. But the response of the plants to simultaneous occurrence of more than one stress is different compared to the plants that are exposed to one particular abiotic stress. While engineering the plants for drought stress, perhaps one needs to keep in mind about temperature stress also and try to modify the plant. Tolerance to a combination of drought and high temperature for instance or cold coupled with ozone or photooxidative stress conditions should be the focus of future research that is aimed at designing and developing transgenic crops with enhanced tolerance in the field conditions. Achieving tolerance for such combined stresses might lead to an important gateway for yield improvement. 3. Outlook In recent years, large amount of data have accumulated on plant's responses to stress, both at the level of signaling and

N. Sreenivasulu et al. / Gene 388 (2007) 1–13

perception. Certainly a plethora of regulatory proteins like transcription and signaling factors associated with salt, drought and temperature tolerance have helped our understanding of the molecular mechanisms associated with them and also the plant survival and crop yields to some extent under these conditions per se. But there is an intriguing amount of cross-talk and interconnections that are involved in stress signaling, such as trans- and cis-acting factors. Further, understanding about the post-translational modifications of proteins, degradation of proteins and also non-coding microRNA interactions will allow us the modulation of the target proteins. It has been shown that some siRNAs are stress-inducible and they affect transcriptional and translational processes including alternative splicing. Only then, we can better fine tune salt, drought and temperature stresses so as to suit the climatic needs. As pointed out in this review, failure to gain the necessary knowledge of exploring natural mechanisms of achieving optimum yield by stress adaptations will hamper the programs of developing agriculturally sustainable crop. A large number of functionally characterized genes were introduced into crop plants to build up tolerance against various abiotic stress conditions. Most of the studies recorded an increased stress tolerance of the transgenic lines in different crop plants when compared to the controls in the laboratory conditions. Since genetically engineered plants exhibited higher accumulation of osmolytes and stress proteins, an increased tolerance for water and cold stresses, and gain in terms of agricultural productivity in such plants is expected to be even more dramatic. However, adequate scientific data on extensive field tests were not available. Therefore, we are yet to show which of the above discussed approaches are promising for further development of products (varieties) that can overcome the environmental constraints and perform better. It is evident from the published research data that extensive work is imperative in several research fronts before we release some of these transgenic lines as varieties. In most cases constitutive expression of stress-tolerant genes is likely to cause unwanted effects. Therefore, it is highly desirable to achieve organ-specific and stress-responsive expression of the introduced genes by introducing specific promoters. In this regard, genomic tools identified a set of stress-inducible genes, and their promoters can now be isolated and thoroughly tested for specificity. Such promoters could be used to prevent gene silencing when gene pyramiding is sought as a feasible strategy to obtain higher tolerance levels. Genomic approaches have resulted in identifying whole pathways involved in abiotic stress response and their relationships to the whole metabolic network. At the same time general pathway regulators can be identified. However, unwanted side effects of manipulating such regulatory genes are even more likely than with single gene members of pathways and, therefore, intensive research and breeding efforts are needed to harness their potential. Besides the genetic engineering approach, other more classical approaches related to breeding programs are of no less importance. There is an air of optimism that the promise ushered by the recent approaches by functional genomics and genomics-assisted breeding of abiotic stress tolerance could

11

generate valuable information for engineering stress-tolerant plants for their ultimate use in sustainable agriculture. Acknowledgements This work has been supported by IPK internal funds and funds of the Federal Ministry of Education and Research (BMBF; GABI SEED II grant). We gratefully acknowledge Prof. U. Wobus, IPK for expert comments on the manuscript.

References Aguilar, P.S., Hernandez-Arriaga, A.M., Cybulski, L.E., Erazo, A.C., de Mendoza, D., 2001. Molecular basis of thermo-sensing: a two-component signal transduction thermometer in Bacillus subtilis. EMBO J. 20, 1681–1691. Apse, M.P., Blumwald, E., 2002. Engineering salt tolerance in plants. Curr. Opin. Biotechnol. 13, 146–150. Bahieldin, A., et al., 2005. Field evaluation of transgenic wheat plants stably expressing the HVA1 gene for drought tolerance. Physiol. Plant. 123, 421–427. Baima, S., et al., 2001. The Arabidopsis ATHB-8 HD-zip protein acts as a differentiation-promoting transcription factor of the vascular meristems. Plant Physiol. 12, 643–655. Bruce, W.B., Edmeades, G.O., Barker, T.C., 2002. Molecular and physiological approaches to maize improvement for drought tolerance. J. Exp. Bot. 53, 13–25. Buchanan, C.D., et al., 2005. Sorghum bicolor's transcriptome response to dehydration, high salinity and ABA. Plant Mol. Biol. 58, 699–720. Buitink, J., et al., 2006. Transcriptome profiling uncovers metabolic and regulatory processes occurring during the transition from desiccationsensitive to desiccation-tolerant stages in Medicago truncatula seeds. Plant J. 47, 735–750. Ceccarelli, S., 1987. Yield potential and drought tolerance of segregating populations of barley in contrasting environments. Euphytica 36, 265–273. Chen, W.Q.J., Zhu, T., 2004. Networks of transcription factors with roles in environmental stress response. Trends Plant Sci. 9, 591–596. Chen, et al., 2002. Expression profile matrix of Arabidopsis transcription factor genes suggests their putative functions in response to environmental stresses. Plant Cell 14, 559–574. Cushman, J.C., Bohnert, H.J., 2000. Genomic approaches to plant stress tolerance. Curr. Opin. Plant Biol. 3, 117–124. Desikan, R., Mackerness, S.A.H., Hancock, J.T., Neill, S.J., 2001. Regulation of the Arabidopsis transcriptome by oxidative stress. Plant Physiol. 127, 159–172. Dubouzet, J.G., et al., 2003. OsDREB genes in rice, Oryza sativa L., encode transcription activators that function in drought-, high-salt- and coldresponsive gene expression. Plant J. 33, 751–763. Ellis, R.P., et al., 2000. Wild barley: a source of genes for crop improvement in the 21st century. J. Exp. Bot. 51, 9–17. Forster, B.P., et al., 1997. Locating genotypes and genes for abiotic stress tolerance in barley: a strategy using maps, markers and the wild species. New Phytol. 137, 141–147. Fowler, S., Thomashow, M.F., 2002. Arabidopsis transcriptome profiling indicates that multiple regulatory pathways are activated during cold acclimation in addition to the CBF cold response pathway. Plant Cell 14, 1675–1690. Furihata, T., et al., 2006. Abscisic acid-dependent multisite phosphorylation regulates the activity of a transcription activator AREB1. Proc. Natl. Acad. Sci. U. S. A. 103, 1988–1993. Gorham, J., Wyn-Jones, R.G., Bristol, A., Jones, R.G.W., 1990. Partial characterization of the trait for enhanced K+–Na+ discrimination in the D genome of wheat. Planta 180, 590–597. Hasegawa, P.M., Bressan, R.A., Zhu, J.K., Bohnert, H.J., 2000. Plant cellular and molecular responses to high salinity. Ann. Rev. Plant Physiol. Plant Mol. Biol. 51, 463–499.

12

N. Sreenivasulu et al. / Gene 388 (2007) 1–13

Hoth, S., Morgante, M., Sanchez, J.P., Hanafey, M.K., Tingey, S.V., Chua, N.H., 2002. Genome-wide gene expression profiling in Arabidopsis thaliana reveals new targets of abscisic acid and largely impaired gene regulation in the abi1-1 mutant. J. Cell Sci. 115, 4891–4900. Hsieh, T.H., Lee, J.T., Charng, Y.Y., Chan, M.T., 2002. Tomato plants ectopically expressing Arabidopsis CBF1 show enhanced resistance to water deficit stress. Plant Physiol. 130, 618–626. Hu, H.H., et al., 2006. Overexpressing a NAM, ATAF, and CUC (NAC) transcription factor enhances drought resistance and salt tolerance in rice. Proc. Natl. Acad. Sci. U. S. A. 103, 12987–12992. Ingram, J., Bartels, D., 1996. The molecular basis of dehydration tolerance in plants. Ann. Rev. Plant Physiol. Plant Mol. Biol. 47, 377–403. Ishitani, M., Xiong, L.M., Stevenson, B., Zhu, J.K., 1997. 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 9, 1935–1949. Ito, Y., et al., 2006. Functional analysis of rice DREB1/CBF-type transcription factors involved in cold-responsive gene expression in transgenic rice. Plant Cell Physiol. 47, 141–153. Jaglo, K.R., et al., 2001. Components of the Arabidopsis C-repeat/dehydrationresponsive element binding factor cold-response pathway are conserved in Brassica napus and other plant species. Plant Physiol. 127, 910–917. Jansen, R.C., 2003. Studying complex biological systems using multifactorial perturbation. Nature Rev. Genet. 4, 145–151. Jansen, R.C., Nap, J.P., 2001. Genetical genomics: the added value from segregation. Trends Genet. 17, 88–391. Kamei, A., et al., 2005. Analysis of gene expression profiles in Arabidopsis salt overly sensitive mutants sos2-1 and sos3-1. Plant Cell Environ. 28, 1267–1275. Kang, J.Y., Choi, H.I., Im, M.Y., Kim, S.Y., 2002. Arabidopsis basic leucine zipper proteins that mediate stress-responsive abscisic acid signaling. Plant Cell 14, 343–357. Kasuga, M., Liu, Q., Miura, S., Yamaguchi-Shinozaki, K., Shinozaki, K., 1999. Improving plant drought, salt and freezing tolerance by gene transfer of a single stress-inducible transcription factor. Nature Biotechnol. 17, 287–291. Kawasaki, S., et al., 2001. Gene expression profiles during the initial phase of salt stress in rice. Plant Cell 13, 889–905. Kishor, P.B.K., et al., 2005. Regulation of proline biosynthesis, degradation, uptake and transport in higher plants: its implications in plant growth and abiotic stress tolerance. Curr. Sci. 88, 424–438. Kim, S., Kang, J.Y., Cho, D.I., Park, J.H., Kim, S.Y., 2004. ABF2, an ABREbinding bZIP factor, is an essential component of glucose signaling and its overexpression affects multiple stress tolerance. Plant J. 40, 75–87. King, I.P., et al., 1997. Introgression of salt tolerance genes from Thinopyrum besarabicum into wheat. New Phytol. 137, 75–81. Kizis, D., Pages, M., 2002. Maize DRE-binding proteins DBF1 and DBF2 are involved in rab17 regulation through the drought-responsive element in an ABA-dependent pathway. Plant J. 30, 679–689. Kizis, D., Lumbreras, V., Pages, M., 2001. Role of AP2/EREBP transcription factors in gene regulation during abiotic stress. FEBS Lett. 498, 187–189. Kreps, J.A., Wu, Y.J., Chang, H.S., Zhu, T., Wang, X., Harper, J.F., 2002. Transcriptome changes for Arabidopsis in response to salt, osmotic, and cold stress. Plant Physiol. 130, 2129–2141. Lee, H.J., Xiong, L.M., Gong, Z.Z., Ishitani, M., Stevenson, B., Zhu, J.K., 2001. The Arabidopsis HOS1 gene negatively regulates cold signal transduction and encodes a RING finger protein that displays cold-regulated nucleocytoplasmic partitioning. Genes Dev. 15, 912–924. Lee, B.H., Henderson, D.A., Zhu, J.K., 2005. The Arabidopsis cold-responsive transcriptome and its regulation by ICE1. Plant Cell 17, 3155–3175. Lee, B.J., Park, C.J., Kim, S.K., Kim, K.J., Paek, K.H., 2006. In vivo binding of hot pepper bZIP transcription factor CabZIP1 to the G-box region of pathogenesis-related protein 1 promoter. Biochem. Biophys. Res. Comm. 344, 55–62. Li, J., Huang, X.Q., Heinrichs, F., Ganal, M.W., Roder, M.S., 2005. Analysis of QTLs for yield, yield components, and malting quality in a BC3-DH population of spring barley. Theor. Appl. Genet. 110, 356–363. Ligterink, W., Hirt, H., 2000. MAP kinase pathways in plants: versatile signaling tools. Int. Rev. Cyt. 201, 209–258.

Liu, J.P., Ishitani, M., Halfter, U., Kim, C.S., Zhu, J.K., 2000. The Arabidopsis thaliana SOS2 gene encodes a protein kinase that is required for salt tolerance. Proc. Natl. Acad. Sci. U. S. A. 97, 3730–3734. Maruyama, K., et al., 2004. Identification of cold-inducible downstream genes of the Arabidopsis DREB1A/CBF3 transcriptional factor using two microarray systems. Plant J. 38, 982–993. Merlot, S., et al., 2002. Use of infrared thermal imaging to isolate Arabidopsis mutants defective in stomatal regulation. Plant J. 30, 601–609. Mustilli, A.C., Merlot, S., Vavasseur, A., Fenzi, F., Giraudat, J., 2002. Arabidopsis OST1 protein kinase mediates the regulation of stomatal aperture by abscisic acid and acts upstream of reactive oxygen species production. Plant Cell 14, 3089–3099. Nevo, E., 1992. Origin, evolution, population genetics and resources for breeding of wild barley Hordeum spontaneum in Fertile Crescent. In: Sherwy, P.R. (Ed.), Barley Genetics Biochemistry, Molecular Biology and Biotechnology. CAB International, Wallingford UK, pp. 19–43. Oh, S.J., et al., 2005. Arabidopsis CBF3/DREB1A and ABF3 in transgenic rice increased tolerance to abiotic stress without stunting growth. Plant Physiol. 138, 341–351. Park, J.M., Park, C.J., Lee, S.B., Ham, B.K., Shin, R., Paek, K.H., 2001. Overexpression of the tobacco Tsi1 gene encoding an EREBP/AP2-type transcription factor enhances resistance against pathogen attack and osmotic stress in tobacco. Plant Cell 13, 1035–1046. Ren, Z.H., et al., 2005. A rice quantitative trait locus for salt tolerance encodes a sodium transporter. Nat. Genet. 37, 1141–1146. Sahi, C., Agarwal, M., Reddy, M.K., Sopory, S.K., Grover, A., 2003. Isolation and expression analysis of salt stress associated ESTs from contrasting rice cultivars using a PCR-based subtraction method. Theor. Appl. Genet. 106, 620–628. Saijo, Y., Hata, S., Kyozuka, J., Shimamoto, K., Izui, K., 2000. Over-expression of a single Ca2+-dependent protein kinase confers both cold and salt/drought tolerance on rice plants. Plant J. 23, 319–327. Sakuma, Y., et al., 2006. Functional analysis of an Arabidopsis transcription factor, DREB2A, involved in drought-responsive gene expression. Plant Cell 18, 1292–1309. Sanan-Mishra, N., Pham, X.H., Sopory, S.K., Tuteja, N., 2005. Pea DNA helicase 45 overexpression in tobacco confers high salinity tolerance without affecting yield. Proc. Natl. Acad. Sci. U. S. A. 102, 509–514. Sangam, S., Jayasree, D., Reddy, K.J., Chari, P.V.B., Sreenivasulu, N., Kavi Kishor, P.B., 2005. Salt tolerance in plants — transgenic approaches. J. Plant Biotechnol. 7, 1–15. Seki, M., et al., 2001. Monitoring the expression pattern of 1300 Arabidopsis genes under drought and cold stresses by using a full-length cDNA microarray. Plant Cell 13, 61–72. Seki, et al., 2002a. Monitoring the expression profiles of 7000 Arabidopsis genes under drought, cold and high-salinity stresses using a full-length cDNA microarray. Plant J. 31, 279–292. Seki, et al., 2002b. Monitoring the expression pattern of around 7000 Arabidopsis genes under ABA treatments using a full-length cDNA microarray. Funct. Integr. Genomics 2, 282–291. Shi, H.Z., Ishitani, M., Kim, C.S., Zhu, J.K., 2000. The Arabidopsis thaliana salt tolerance gene SOS1 encodes a putative Na+/H+ antiporter. Proc. Natl. Acad. Sci. U. S. A. 97, 6896–6901. Shi, H.Z., Xiong, L.M., Stevenson, B., Lu, T.G., Zhu, J.K., 2002. The Arabidopsis salt overly sensitive 4 mutants uncover a critical role for vitamin B6 in plant salt tolerance. Plant Cell 14, 575–588. Shinozaki, K., Yamaguchi-Shinozaki, K., 2000. Molecular responses to dehydration and low temperature: differences and cross-talk between two stress signaling pathways. Curr. Opin. Plant Biol. 3, 217–223. Shou, H.X., Bordallo, P., Wang, K., 2004. Expression of the Nicotiana protein kinase (NPK1) enhanced drought tolerance in transgenic maize. J. Exp. Bot. 55, 1013–1019. Singla-Pareek, S.L., Reddy, M.K., Sopory, S.K., 2003. Genetic engineering of the glyoxalase pathway in tobacco leads to enhanced salinity tolerance. Proc. Natl. Acad. Sci. U. S. A. 100, 14672–14677. Singla-Pareek, S.L., Yadav, S.K., Pareek, A., Reddy, M.K., Sopory, S.K., 2006. Transgenic tobacco overexpressing glyoxalase pathway enzymes grow and set viable seeds in zinc-spiked soils. Plant Physiol. 140, 613–623.

N. Sreenivasulu et al. / Gene 388 (2007) 1–13 Sreenivasulu, N., Kavi Kishor, P.B., Varshney, R.K., Altschmied, L., 2002. Mining functional information from cereal genomes — the utility of expressed sequence tags (ESTs). Curr. Sci. 83, 1–9. Sreenivasulu, N., Varshney, R.K., Kavi Kishor, P.B., Weschke, W., 2004a. Tolerance to abiotic stress in cereals: a functional genomics approach. In: Gupta, P.K., Varshney, R.K. (Eds.), Cereal Genomic, pp. 483–514. Sreenivasulu, N., Miranda, M., Prakash, H.S., Wobus, U., Weschke, W., 2004b. Transcriptome changes in foxtail millet genotypes at high salinity: identification and characterization of a PHGPX gene specifically upregulated by NaCl in a salt-tolerant line. J. Plant Physiol. 161, 467–477. Sreenivasulu, N., Radchuk, V., Strickert, M., Miersch, O., Weschke, W., Wobus, U., 2006. Gene expression patterns reveal tissue-specific signaling networks controlling programmed cell death and ABA-regulated maturation in developing barley seeds. Plant J. 47, 310–327. Tamura, T., Hara, K., Yamaguchi, Y., Koizumi, N., Sano, H., 2003. Osmotic stress tolerance of transgenic tobacco expressing a gene encoding a membrane-located receptor-like protein from tobacco plants. Plant Physiol. 131, 454–462. Tanksley, S.D., Nelson, J.C., 1996. Advanced backcross QTL analysis: a method for the simultaneous discovery and transfer of valuable QTLs from unadapted germplasm into elite breeding lines. Theor. Appl. Genet. 92, 191–203. Teige, M., et al., 2004. The MKK2 pathway mediates cold and salt stress signaling in Arabidopsis. Mol. Cell 15, 141–152. Thomashow, M.F., 2001. So what's new in the field of plant cold acclimation? Lots! Plant Physiol. 125, 89–93. Tuberosa, R., Salvi, S., 2006. Genomics-based approaches to improve drought tolerance of crops. Trends Plant Sci. 11, 405–412. Umezawa, T., Fujita, M., Fujita, Y., Yamaguchi-Shinozaki, K., Shinozaki, K., 2006a. Engineering drought tolerance in plants: discovering and tailoring genes to unlock the future. Curr. Opin. Plant Biotech. 17, 113–122. Umezawa, T., et al., 2006b. CYP707A3, a major ABA 8′-hydroxylase involved in dehydration and rehydration response in Arabidopsis thaliana. Plant J. 46, 171–182. Uno, Y., Furihata, T., Abe, H., Yoshida, R., Shinozaki, K., YamaguchiShinozaki, K., 2000. Arabidopsis basic leucine zipper transcription factors involved in an abscisic acid-dependent signal transduction pathway under drought and high-salinity conditions. Proc. Natl. Acad. Sci. U. S. A. 97, 11632–11637. Urao, T., et al., 1999. A transmembrane hybrid-type histidine kinase in Arabidopsis functions as an osmosensor. Plant Cell 11, 1743–1754. Valliyodan, B., Nguyen, H.T., 2006. Understanding regulatory networks and engineering for enhanced drought tolerance in plants. Curr. Opin. Plant Biotech. 9, 189–195. Viswanathan, C., Zhu, J.K., 2004. Molecular perspectives on cross-talk and specificity in abiotic stress signaling in plants. J. Exp. Bot. 55, 225–236. Vogel, J.T., Zarka, D.G., Van Buskirk, H.A., Fowler, S.G., Thomashow, M.F., 2005. Roles of the CBF2 and ZAT12 transcription factors in configuring the low temperature transcriptome of Arabidopsis. Plant J. 41, 195–211. Wang, H., et al., 2004a. Ectopic overexpression of tomato JERF3 in tobacco activates downstream gene expression and enhances salt tolerance. Plant Mol. Biol. 55, 183–192.

13

Wang, Z.I., et al., 2004b. Expressed sequence tags from Thellungiella halophila, a new model to study plant salt-tolerance. Plant Sci. 166, 609–616. Wang, Y., et al., 2005. Molecular tailoring of farnesylation for plant drought tolerance and yield protection. Plant J. 43, 413–424. Weigel, et al., 2000. Activation tagging in Arabidopsis. Plant Physiol. 122, 1003–1013. Xiong, L.M., Ishitani, M., Zhu, J.K., 1999. Interaction of osmotic stress, temperature, and abscisic acid in the regulation of gene expression in Arabidopsis. Plant Physiol. 119, 205–211. Xiong, L.M., Ishitani, M., Lee, H., Zhu, J.K., 2001a. The Arabidopsis LOS5/ ABA3 locus encodes a molybdenum cofactor sulfurase and modulates cold stress- and osmotic stress-responsive gene expression. Plant Cell 13, 2063–2083. Xiong, L.M., Lee, B.H., Ishitani, M., Lee, H., Zhang, C.Q., Zhu, J.K., 2001b. FIERY1 encoding an inositol polyphosphate 1-phosphatase is a negative regulator of abscisic acid and stress signaling in Arabidopsis. Genes Dev. 15, 1971–1984. Xu, K., et al., 2006. Sub1A is an ethylene-response-factor-like gene that confers submergence tolerance to rice. Nature 442, 705–708. Yamaguchi-Shinozaki, K., Shinozaki, K., 2005. Organization of cis-acting regulatory elements in osmotic-and cold-stress-responsive promoters. Trends Plant Sci. 10, 88–94. Yang, J.C., Zhang, J.H., 2006. Grain filling of cereals under soil drying. New Phytol. 169, 223–236. Yang, J.C., Zhang, J.H., Wang, Z.Q., Zhu, Q.S., Wang, W., 2001. Hormonal changes in the grains of rice subjected to water stress during grain filling. Plant Physiol. 127, 315–323. Yang, J.C., Zhang, J.H., Liu, K., Wang, Z.Q., Liu, L.J., 2006. Abscisic acid and ethylene interact in wheat grains in response to soil drying during grain filling. New Phytol. 171, 293–303. Yoshida, R., et al., 2002. ABA-activated SnRK2 protein kinase is required for dehydration stress signaling in Arabidopsis. Plant Cell Physiol. 43, 1473–1483. Yoshida, T., Nishimura, N., Asami, T., Shinozaki, K., Hirayama, T., 2005. Characterization of Arabidopsis ABA hypersensitive mutant, ahg3. Plant Cell Physiol. 46 (S51-S51, Suppl. S.). Yu, L.X., Setter, T.L., 2003. Comparative transcriptional profiling of placenta and endosperm in developing maize kernels in response to water deficit. Plant Physiol. 131, 1921–1922. Zhu, J.K., 2001. Cell signaling under salt, water and cold stresses. Curr. Opin. Plant Biol. 4, 401–406. Zhu, J.K., 2002. Salt and drought stress signal transduction in plants. Ann. Rev. Plant Biol. 53, 247–273. Zhu, J.K., Hasegawa, P.M., Bressan, R.A., 1997. Molecular aspects of osmotic stress in plants. Crit. Rev. Plant Sci. 16, 253–277. Zhu, J.K., Liu, J.P., Xiong, L.M., 1998. Genetic analysis of salt tolerance in Arabidopsis: evidence for a critical role of potassium nutrition. Plant Cell 10, 1181–1191. Zinselmeier, C., Jeong, B.R., Boyer, J.S., 1999. Starch and the control of kernel number in maize at low water potentials. Plant Physiol. 121, 25–35.