Plants cope with drought stress by manipulating key physiological processes like photosynthesis, respiration, water relations, antioxidant and hormonal metabolism. There exist multiple and often redundant stress sensors, which transduce the stress signal through secondary signalling molecules to the nucleus, where the expression of stress-response genes is regulated. Transcription factors play an important role in regulating the expression of the stress-response genes. Another level of regulation of gene expression is at the epigenetic level and involves modifications either at the chromatin level or at the mRNA level. Crop plants show various adaptive and acclimatization strategies to drought stress, which range from seemingly simple morphological or physiological traits that serve as important stress tolerance markers to major upheavals in gene expression in which a large number of transcription factors are induced. Studies on contrasting crop genotypes or genetic engineering of crops help in differentiating responses to drought from those leading to drought tolerance. Of specific importance to crop plants is not whether they survive stress, but whether they show good yields under stress conditions.
Global climate changes are leading to increases in temperature and atmospheric CO2 levels as well as alterations in rainfall patterns. Periods of inadequate rainfall leading to drought are predicted to arise more frequently under such conditions. Terminal drought conditions bring about a progressive decrease in soil water availability to plants and cause premature plant death, while intermittent drought conditions affect the plant growth and development but are not usually lethal. The ability to survive longer and maintain function under intermittent or terminal drought conditions leads to subsistence yields, which are much lower than those observed under hydrated conditions. Drought tolerance enables plants to grow and maintain relatively high yields in spite of drought conditions and is an outcome of the plant's efforts to withstand or recover from stress. If the tolerance is restricted to that particular generation, the plant is said to be acclimated to drought. If it persists over generations, the plant genotype is said to be adapted to drought conditions.
A large number of molecular, biochemical and physiological processes at the cellular or whole plant level are altered in response to drought and play an important role in mitigating stress. What is crucial but difficult is to distinguish between the responses that lead to tolerance from those that arise due to stress-induced damage. The molecular machinery involved in drought stress perception, signalling and regulation of gene expression has been fairly well understood. However, there are lacunae in our understanding of how it correlates with phenotypic alterations in the plant (Blum 2011). On the other hand, several phenotypic markers have been identified in crop plants that correlate with drought tolerance, but we know little of either the gene expression involved in these phenotypic traits or how they correlate with the yield parameters.
The review attempts at going through the breadth of processes involved in giving rise to a drought-response phenotype. Some of these processes have been compared in contrasting genotypes of crops, with the objective of understanding those that correlate with better yields under drought conditions. Genetic engineering of crop plants has also emerged as an important technique to validate the role of specific genes in giving rise to the drought phenotype.
Drought Responses of Plants
Growth and water relations
A primary response of plants subjected to drought stress is growth arrest. Shoot growth inhibition under drought reduces metabolic demands of the plant and mobilizes metabolites for the synthesis of protective compounds required for osmotic adjustment. Root growth arrest enables the root meristem to remain functional and gives rise to rapid root growth when the stress is relieved (Hsaio and Xu 2000). Lateral root inhibition has also been seen to be an adaptive response, which leads to growth promotion of the primary root, enabling extraction of water from the lower layers of soil (Xiong et al. 2006). Growth inhibition can arise due to the loss of cell turgor arising from the lack of water availability to the growing cells. Water availability to cells is low because of poor hydraulic conductance from roots to leaves caused by stomatal closure. Although a decrease in hydraulic conductance decreases the supply of nutrients to the shoot, it also prevents embolism in xylem and could constitute an adaptive response. Osmotic adjustment is another way by which plants cope with drought stress. Synthesis of compatible solutes like polyols and proline under stress prevents the water loss from cells and plays an important role in turgor maintenance (Blum 2005, DaCosta and Huang 2006). Modification of growth priorities as well as reduction in the performance of photosynthetic organs due to stress exposure leads to alterations in carbon partitioning between the source and sink tissues (Roitsch 1999). Hence, carbohydrates that contribute to growth under normal growth conditions are now available for selective growth of roots or for the synthesis of solutes for osmotic adjustment (Lei et al. 2006, Xue et al. 2008).
Water deficit–induced ABA synthesis brings about stomatal closure, which causes a decrease in intercellular carbon dioxide concentration and inhibits photosynthesis. This inhibition is reversible and photosynthesis can resume if stomata open upon stress removal (Chaves et al. 2009). On the other hand, open stomata and high hydraulic conductance under drought enable photosynthesis and nutrient supply to the shoot at the cost of risking turgor loss (Sade et al. 2012). Some plants appear to adopt the latter strategy to enable the synthesis of osmotic metabolites from photoassimilates, which help in preventing turgor loss.
Carbon dioxide limitation due to prolonged stomatal closure in the face of continued photosynthetic light reactions leads to the accumulation of reduced photosynthetic electron transport components, which can reduce molecular oxygen and give rise to reactive oxygen species (ROS), thus causing indiscriminate damage to the photosynthetic apparatus. This metabolic inhibition of photosynthesis is irreversible and leads to injury (Lawlor and Cornic 2002). Hence, photophosphorylation and ATP generation is reduced, which inhibits Rubisco activity. Adaptive responses to prevent drought-induced damage to photosynthetic apparatus include thermal dissipation of light energy, photodestruction of D1 protein of PSII, the xanthophyll cycle, water–water cycle and dissociation of the light-harvesting complexes from photosynthetic reaction centres (Niyogi 1999, Demmig-Adams and Adams 2006) (Table 1).
Table 1. Physiological responses contributing to drought tolerance in plants
ROS, reactive oxygen species.
Adjustment of chlorophyll antenna size. Photodestruction of D1 protein of PSII
Plant growth is determined by the ratio between photosynthetic CO2 assimilation and respiratory CO2 release. The rate of respiration is regulated by processes that use the respiratory products – ATP (water and solute uptake by roots, translocation of assimilates to sink tissues), NADH and TCA cycle intermediates (biosynthetic processes in growing parts of a plant), which together contribute to plant growth. Under drought stress, these processes are affected and lead to a decrease in respiration rate. On the other hand, increased respiratory rates have also been observed under water scarcity and these lead to an increase in the intercellular CO2 levels in leaves (Lawlor and Tezara 2009). Higher respiration may arise due to uncoupling of respiratory oxygen evolution from oxidative phosphorylation, which prevents the accumulation of reductants and reduces the generation of ROS. Increased respiratory rates are also observed due to the activation of energy-intensive processes like osmolyte synthesis and antioxidant metabolism that occur under drought conditions.
Interdependence of metabolic processes in chloroplasts and mitochondria has been reported (Raghavendra and Padmasree 2003). For example, mitochondria are involved in processing the glycolate produced in chloroplasts during photorespiration (Taira et al. 2004). Mitochondrial respiration also plays an important role in dissipating the NADPH generated during photosynthetic light reactions through type II NADPH dehydrogenases situated on the matrix side (Plaxton and Podesta 2006). Hence, leaf mitochondria act as a safety engine that enables the plant to cope with variations in chloroplast metabolism under water stress (Atkin and Macherel 2009). Plant mitochondria also prevent ROS generation within themselves by employing the alternative oxidase (AOX) pathway, in which the complexes III and IV of the respiratory electron transport system are bypassed and electrons are directly transferred to oxygen, with the generation of thermal energy instead of ATP (Siedow and Umbach 2000). The AOX pathway as well as the photorespiratory pathway is operational when a plant is exposed to stress and serves a role in maintaining cell function by preventing the accumulation of ROS (Lambers et al. 2005, Florez-Sarasa et al. 2007).
In addition, the TCA cycle is modified to prevent the generation of excess reductants. One of the modifications is GABA synthesis, in which two steps in the TCA cycle related to the generation of reducing power are bypassed. GABA accumulation occurs during stress conditions and may constitute a stress adaptive response (Fait et al. 2007).
Prohibitins are large protein complexes that localize to the inner mitochondrial membrane, where they appear to play a role in maintaining the superstructure of the inner mitochondrial membrane and the protein complexes associated with it (Van Aken et al. 2010). They have been implied in stress tolerance not only because of their role in protecting mitochondrial structure, but also in triggering retrograde signalling between mitochondria and the nucleus in response to stress, thus altering the expression of several stress-responsive transcripts, including AOX, heat-shock proteins (HSP) and genes involved in hormone homoeostasis.
Reactive oxygen species are generated due to metabolic perturbation of cells, and these cause cell damage and death. While mechanisms to prevent the generation of ROS have been mentioned earlier, an important adaptive mechanism consists of their effective scavenging if and when these harmful species do arise. Antioxidant substrates like ascorbate, α-tocopherol and carotenoids and antioxidant enzymes like superoxide dismutase, catalase, ascorbate peroxidase and glutathione reductase exist in cell organelles and the cytoplasm and play an important role in detoxifying these reactive species (Shao et al. 2008). Methionine sulfoxide reductases are another class of antioxidant enzymes that play a role in preventing damage to proteins due to ROS generation in plastids (Rouhier et al. 2006). These enzymes use thioredoxin to reduce the methionine sulfoxide residues generated in proteins due to oxidative stress.
Plant hormones regulate diverse processes in plants, which enable acclimation to stress. On exposure to water deficits, ABA synthesized in roots is known to be translocated to leaves, where it brings about stomatal closure and inhibits plant growth, thus enabling the plant to adapt to stress conditions (Wilkinson and Davies 2010). In barley, fivefold increase in endogenous ABA levels was observed in drought-tolerant varieties as compared to susceptible ones, indicating its role in improving stress tolerance (Thameur et al. 2011). The role of ABA in regulating aquaporin activity, which contributes to the maintenance of a favourable plant water status, has also been reported (Parent et al. 2009). Improvement of shoot growth under drought was observed when 9-cis-epoxycarotenoid dioxygenase (NCED3), a key enzyme in abscisic acid biosynthesis, was overexpressed in Arabidopsis (Iuchi et al. 2001). ABA accumulation during the expression of drought tolerance is known to bring about a reduction in ethylene production and an inhibition of ethylene-induced senescence and abscission. ABA-deficient maize seedlings showed drought susceptibility as well as an increase in ethylene production (Sharp 2002). Auxins have been identified as negative regulators of drought tolerance. In wheat leaves, drought stress tolerance was accompanied by a decrease in ndole-3-acetic acid (IAA) content (Xie et al. 2003). Downregulation of IAA was seen to facilitate the accumulation of late embryogenesis-abundant (LEA) mRNA, leading to drought stress adaptation in rice (Zhang et al. 2009). However, there are evidences of a transient increase in IAA content in maize leaves during the initial stages of exposure to water stress, which later drops sharply as the plant acclimates to water stress (Wang et al. 2008). A rapid decline in endogenous zeatin and gibberellin (GA3) levels was also observed in maize leaves subjected to water stress, which correlated with higher levels of cell damage and plant growth inhibition. Reduced cytokinin content and activity caused by either reduced biosynthesis or enhanced degradation was observed in drought-stressed plants. (Pospisilova et al. 2000). In alfalfa, decreased cytokinin content during drought led to accelerated senescence (Goicoechea et al. 1995). Cytokinins are known to delay senescence, and an increase in the endogenous levels of cytokinins through the overexpression of the ipt gene involved in cytokinin biosynthesis led to stress adaptation by delaying drought-induced senescence (Peleg and Blumwald 2011). Cytokinins are also negative regulators of root growth and branching, and root-specific degradation of cytokinin contributed to primary root growth and branching induced by drought stress, hence increasing drought tolerance in Arabidopsis (Werner et al. 2010).
Brassinosteroids (BRs) have also been reported to protect plants against various abiotic stresses (Kagale et al. 2007). Application of BR was seen to increase water uptake and membrane stability, as well as to reduce ion leakage arising from membrane damage in wheat plants subjected to drought stress (Sairam 1994). However, it was shown that changes in endogenous BR levels did not occur during the exposure of pea plants to water stress (Jager et al. 2008).
Stress Perception and Signalling
Acclimation to stress involves processes starting from perception of stress to the expression of large number of genes involved in the manifestation of a morphological or physiological response that increases the chances of survival under the stress condition (Fig. 1).
Molecular mechanisms that sense stress consist of a number of classes of cell surface receptors like serine/threonine-like receptor kinases called receptor-like kinases (RLKs), ion channel–linked receptors, G-protein-coupled receptors (GPCRs) and two-component histidine kinase receptors. RLKs are major contributors to the processing of a vast array of plant developmental and environmental cues. Their activity is regulated by receptor oligomerization and phosphorylation, receptor internalization and dephosphorylation or regulation at the transcriptional level (Chae et al. 2009). Brassinosteroid receptor BR1 belongs to the RLK family, which in response to BR or stress is internalized by the responding cells and the stress signal transduced. Cre 1 (cytokinin response 1) is a two-component histidine kinase receptor that transduces signal via a phosphorelay pathway. This receptor kinase, besides binding cytokinins, is also thought to act as a sensor of osmotic stress (Bartels and Sunkar 2005). Ca2+ channels are responsible for the influx of Ca2+ into the cytoplasm when activated by various stress situations (Xiong et al. 2002). These channels therefore act as ion channel–linked receptors of stress. GPCRs are another group of membrane receptors, which on sensing stress activate enzymes like phospholipase C or D which in turn release second messengers and transduce the stress signal (Tuteja and Sopory 2008).
An intracellular receptor for ABA, PYR/RCAR, has been shown to signal for drought stress through the activation of a serine/threonine kinase SnRK2, in response to ABA binding (Sheard and Zheng 2009). Because ABA synthesis is known to be induced in response to stress, the ABA receptor can be considered to be a stress sensor.
Sugar signalling has emerged as an important component of stress responses. Hexokinases were identified as glucose sensors in plants, which played a role in repressing photosynthetic gene expression when the hexose levels in leaf cells were high (Kim et al. 2000, Hanson and Smeekens 2009). The trehalose biosynthesis pathway, in which trehalose 6 phosphate (T6P) acts as an indicator of G6P and UDPG pool size, is known to link growth and development to metabolite content, because both sucrose synthesis and trehalose synthesis pathways feed into the same metabolite pool (Vogel et al. 2001, Paul et al. 2008). Trehalose phosphate phosphatases are upregulated under stress conditions and in turn regulate the T6P levels. Hence, multiple facets of drought stress appear to be simultaneously perceived by a cell through various receptors that respond to osmotic pressure, membrane rigidity, metabolic status, Ca2+-level perturbations, respectively, thereby ensuring plant response and improving the chances of survival on drought exposure.
Reactive oxygen species, which are toxic by-products of stress metabolism, also serve as important signalling molecules (Miller et al. 2010) and the oxidative signal is transduced via secondary signalling intermediates like Ca2+ or phosphatidic acid (PA)–activated serine/threonine protein kinases and mitogen-activated protein (MAP) kinases to bring about transcription of genes that play a role in acclimation (Cheeseman 2007). Due to the short half-life of ROS, redox signalling is likely to occur through the redox status of ascorbate/dehydroascorbate and reduced glutathione/oxidized glutathione couples (Foyer and Noctor 2000). Nitric oxide radical (NO) is synthesized in plants, probably either from arginine via a nitric oxide synthase or by nitrite reduction, and has been shown to be a component of secondary messenger cascades (Mazid et al. 2011), involving cyclic GMP and Ca2+.
Signal perception is followed by the generation of secondary signalling molecules such as protein kinases and phosphatases (serine/threonine phosphatases), phospholipids like phosphoinositides (Bartels and Sunkar 2005), ROS, Ca2+, nitric oxide, cAMP and sugars, which play an important role in signal transduction (Tuteja and Sopory 2008). Many of these secondary messengers are common to diverse stress situations, indicating that cross-talk between different stress-response pathways may occur through these common signal transducers.
Mitogen-activated protein kinases bring about protein phosphorylation and constitute one of the major mechanisms for signal transduction. They are located in the cytoplasm and consist of three classes of enzymes (MAPK, MAPKK and MAPKKK) that form a signalling cascade from the stress sensor located on the plasma membrane to the regulation of gene expression in the nucleus. Translocation of the MAPK into the nucleus brings about the activation of transcription factors through phosphorylation (Tena et al. 2001).
Calcium levels in the cytoplasm have been shown to increase transiently on stress exposure. The source of this stress-induced cytoplasmic Ca2+ is either from the apoplast or from the cellular reserves. Several Ca2+ sensors like calmodulin (CaM) or CaM-binding proteins have been identified in the cells, which transduce the stress signal to the nucleus through other messengers like phospholipase D or Ca2+-dependent protein kinases (Tuteja and Sopory 2008).
Phospholipids like phosphoinositides that are located in the plasma membranes are a source of several secondary signalling molecules like phosphotidylinositol phosphates, which are phosphorylated by kinases (e.g. PI3Kase) (Drobak and Watkins 2000). Phospholipases act on these phospholipids to generate signalling molecules like inositol 1,4,5-trisphosphate (IP3), diacylglycerol (DAG) and PA, which play a role in the transmission of the signal across plasma membrane and in intracellular signalling.
Transcriptional regulation of gene expression
A large number of genes are seen to be involved in the expression of the stress phenotype (Xiong et al. 2002, Shinozaki and Yamaguchi-Shinozaki 2003). The transcriptional response initially is composed of a core set of multistress-responsive genes and becomes increasingly stress specific as time progresses (Ma and Bohnert 2007). DNA microarrays provide a high-throughput means of analysing gene expression at the whole-genome level and have been used to study the patterns of gene expression in response to drought or high-salinity stresses in several plant species (Seki et al. 2002, Guo et al. 2009, Hayano-Kanashiro et al. 2009).
Some of the genes seen to be upregulated under drought stress conditions include the genes involved in osmolyte synthesis, genes coding for LEA proteins, aquaporins, signalling molecules and transcription factors (TFs). Of these, the genes coding for TFs were particularly interesting because TFs act as master switches and trigger the simultaneous expression of a large number of stress-response genes that contribute to the stress phenotype (Bartels and Souer 2004). About 104 TFs, whose expression was increased on exposure to dehydration stress, have been identified by transcriptome analysis in Arabidopsis plants exposed to drought stress (Rhizsky et al. 2004). While most of the transcription factors were upregulated under stress, a few transcription factors that played a role in primary growth processes were downregulated. Drought stress–induced gene expression was seen to be regulated by TFs belonging to bZIP, AP2/ERF, HD-ZIP, MYB, bHLH, NAC, NF-Y, EAR and ZPT2 families (Yang et al. 2010). These TFs are activated at the transcriptional or protein level by the transduced drought signal. Because drought stress is accompanied by an increase in ABA levels, some TFs are activated specifically by ABA. The ABA-responsive TFs (ABFs) predominantly belong to the bZIP family of TFs and bind to ABA-response elements (ABRE) present in the promoters of stress-response genes (Jakoby et al. 2002, Yoshida et al. 2010). TFs belonging to the AP2/ERF family bind to the drought-response element (DRE) present in their promoters of a large number of drought-response genes (Yamaguchi-Shinozaki and Shinozaki 2005, Maruyama et al. 2009). The HD-ZIP TFs are plant specific and show the presence of a homeodomain adjacent to leucine zipper. Among several functions attributed to this family of transcription factors, one function is the regulation of ABA-dependent genes under dehydration stress (Deng et al. 2002). Most of the plant MYBs consist of two repeats R2R3 (Jin and Martin 1999) and play a role in regulating the expression of dehydration-responsive genes (Abe et al. 2003). The ZPT2 TFs are characterized by the presence of two zinc finger motifs separated by a single long linker. These act as transcriptional repressors by downregulating the activity of other transcription factors (Sakamoto et al. 2004) and are induced during dehydration stress as well as with ABA treatment. Transcription factors belonging to NAC family bind to promoters of not only dehydration-response genes (Tran et al. 2004), but also auxin-response genes (Hegedus et al. 2003).
The promoters of stress-response genes are known to have several types of cis elements to which TFs of the same family or different families can bind (Narusaka et al. 2003, Srivastav et al. 2010). Hence, gene expression under different stress situations can be combinatorially regulated by employing suitable TFs, which often form homo- or heterodimers in bringing about transcriptional activation under specific stress situations. However, manipulations of transcription factors in engineering complex traits such as abiotic stress tolerance are known to produce unintended pleiotropic effects which may have adverse effects on the growth and development of plants (Abdeen et al. 2010).
Post-transcriptional regulation of gene expression
Besides stress-induced regulation of gene expression at the transcription level, stress conditions also bring about epigenetic regulation of gene expression (Table 2). Stress-induced changes in histone variants, histone N-tail modifications and DNA methylation have been shown to regulate stress-responsive gene expression and plant development under stress. Drought stress induced the expression of a variant of histone H1 called H1-S, which appeared to play a role in stomatal closure (Scippa et al. 2004). ABA downregulated the expression of a histone deacetylase AtHD2C, while overexpression of this enzyme brought about enhanced expression of ABA-responsive genes and greater salt and drought tolerance than the wild-type plants (Sridha and Wu 2006). Drought-induced expression of stress-responsive genes was also seen to be associated with modifications in histones H3 and H4. Histone H3K4 trimethylation, H3K9 acetylation, H3 Ser-10 phosphorylation, H3 phosphoacetylation and H4 acetylation were observed, which correlated with the expression of stress-induced genes (Sokol et al. 2007). Histone acetyltransferases (HATs), which interact with transcription factors, were also seen to be involved in activating stress-responsive genes. Stresses can induce changes in gene expression through hypomethylation or hypermethylation of DNA. In tobacco, stress-induced DNA demethylation was observed in the coding sequence of a glycerophosphodiesterase-like protein gene, while DNA hypermethylation was induced by drought stress in pea (Chinnusamy and Zhu 2009).
Table 2. Epigenetic regulation and RNA-related processes in response to drought stress
HATs, histone acetyltransferases; SUMO, small ubiquitin-like modifier.
MicroRNAs (miRNAs) are ~20- to 22-nt non-coding RNAs that specifically base pair to target mRNAs and induce the cleavage of target mRNAs or repress their translation. Hence, they constitute a gene-silencing mechanism that regulates the expression of target genes post-transcriptionally. Regulation of stress-response genes by miRNAs has been demonstrated recently (Shukla et al. 2008). For example, abiotic stress brought about downregulation of miR398 that targets stress-inducible Cu-Zn SOD genes that play a role in scavenging superoxide radicals generated in plants on exposure to stress (Sunkar et al. 2006). MiRNA159 was seen to be upregulated in response to ABA, and this miRNA silenced several MYB transcription factors that are known to positively regulate ABA responses. MiR169 regulated target genes for carbohydrate metabolism, leading to stem sugar accumulation in sweet sorghum (Calvino et al. 2011). MiRNAs miR172 and miR395 were reported to target genes related to time of flowering and permitted greater biomass build-up.
The mRNA transcribed is processed to give rise to the mature mRNA, and RNA-binding proteins are involved in post-transcriptional RNA modifications through processes like splicing and regulation of its stability and turnover. Under stress conditions, alternative splicing of some mRNAs coding for transcription factors has been reported in wheat (Egawa et al. 2006). There are reports indicating the occurrence of alternative splicing in at least 42% of genes in Arabidopsis during abiotic stress conditions (Filichkin et al. 2010, Nakaminami et al. 2012). Degradation or stabilization of mRNA levels under stress conditions is brought about by processing bodies (PBs) and stress granules (SGs), respectively (Weber et al. 2008, Xu and Chua 2011). P-bodies are RNP complexes known to play a role in translational repression and mRNA decapping. Removal of 5′ m7GDP by decapping proteins (DCP1, DCP2) from the mRNA cap takes place in P-bodies, which leads to further degradation of the mRNA by exonucleases (XRN4). SGs have been shown to contain nuclear proteins (UBP1 and RBP47), polyA+ mRNA and translation initiation factors, which under stress conditions are observed as distinct complexes in the cytoplasm (Weber et al. 2008).
Post-translational modification of proteins also plays an important role in the drought stress response. The importance of phosphorylation cascades in signal transduction has already been mentioned earlier. Protein modifications are also known to affect the conformation, activity, localization and stability of transcription factors (Kline et al. 2010). Ubiquitin-dependent protein degradation is another post-translational protein modification, which was shown to play an important role in hormonal signalling (Santner and Estelle 2009). Upregulation of an E3 ubiquitin ligase XERICO in Arabidopsis enhanced the expression of an ABA biosynthesis gene, AtNCED3, thereby increasing the cellular ABA levels and hence drought tolerance (Ko et al. 2006). In addition to ubiquitin, plants use a variety of other polypeptide tags to post-translationally modify and regulate various intracellular proteins. Small ubiquitin-like modifier (SUMO) is one such peptide that brings about sumoylation. In Arabidopsis, the amount of AtSUMO1 and AtSUMO2 conjugates increased in response to various stress treatments, and when these were overexpressed, the increased sumoylation levels induced ABA-/stress-responsive genes by masking ubiquitin sites on regulatory proteins (Kurepa et al. 2003). Hence, the post-translational modifications like sumoylation and ubiquitination modulate plants response to stress.
Drought Adaptation Strategies in Crop Plants
Drought-tolerant plants like xerophytes, halophytes, resurrection plants show morphological and physiological adaptations to cope with poor water availability either through growth arrest till favourable conditions return, or through shortened growth cycles comprising limited vegetative growth followed by flowering and seed set during the short periods of water availability. Such adaptations are not desirable traits in crop species, which develop large yields over long growth periods. Genotypes that differ in drought tolerance serve as important systems for studying adaptive responses to drought in crop species, and exploitation of natural variation for drought-related traits has resulted in an improvement of crop performance (Ribaut et al. 2004, Reynolds and Tuberosa 2008).
Physiological studies on contrasting genotypes provide information on the mechanisms involved in drought tolerance and provide a useful screening strategy for drought tolerance, albeit at a smaller scale and often in an ‘unnatural’ drought exposure (Fig. 2). For example, drought tolerance in durum wheat was attributed to alterations in mitochondrial metabolism. The mitochondria showed an active AOX pathway and an uncoupling protein, both of which played a role in the dissipation of energy and prevented the accumulation of ROS (Pastore et al. 2007). In addition, the cytosolic NADH produced was oxidized by an active malate/oxaloacetate shuttle in the mitochondria. On comparing drought responses of wheat genotypes with the related Aegilops biuncialis genotypes, a higher photosynthetic activity was observed in Aegilops, which are adapted to drier habitats. Higher CO2 fixation was attributed to better stomatal conductance and more efficient non-radiative energy dissipation in Aegilops (Molnar et al. 2002). In comparisons made between drought-tolerant and drought-susceptible sorghum genotypes, it was observed that the genotypes differed in stress thresholds at which transition from stomatal to metabolic inhibition of photosynthesis occurred (Bhargava and Paranjpe 2004). This has important implications because stomatal inhibition of photosynthesis is reversible and an ability to delay metabolic inhibition of photosynthesis would facilitate the recovery from stress. Tolerant genotypes of sorghum were also seen to have higher levels of Rubisco under drought stress than susceptible genotypes, and this correlated with higher transcript levels of the chloroplast chaperone HSP60, which probably protected the Rubisco protein from drought-induced damage (Jagtap et al. 1998). Source–sink relationships also play an important role in drought tolerance of crop plants because carbohydrate reserves are utilized for grain filling and their availability is a critical factor in sustaining grain filling and grain yield under drought stress (Yang and Zhang 2006). Although osmotic adjustment is another mechanism for coping with drought stress, it is seen to be of relevance mainly in root development into deeper soils, which can give plants access to water. This was seen in wheat lines showing better osmotic adjustment as compared to those showing low osmotic adjustment (Morgan 1995). However, in drought-tolerant genotypes of prairie junegrass, genes involved in proline and fructan biosynthesis were seen to play an important role in drought tolerance (Jiang et al. 2010). Efficiency of antioxidant metabolism in protecting plants against oxidative damage has been reported in drought-tolerant crop genotypes as compared to drought-susceptible ones. Drought-tolerant genotypes of sorghum showed higher activities of antioxidant enzymes on exposure to stress, but not under non-stress conditions (Jagtap and Bhargava 1995). An increase in activities of specific isozymes of antioxidant enzymes has also been reported in drought-tolerant rapeseed genotypes subjected to drought stress (Abedi and Pakniyat 2010). However, a drought-tolerant genotype Oryza longistaminata of rice accumulated smaller amounts of ROS as well as antioxidant substrates, indicating that it had other acclimation mechanisms that prevented oxidative stress (Kumar et al. 2011). The role of ABA in drought tolerance has been studied in barley genotypes differing in their ability to survive water-limiting conditions (Thameur et al. 2011). Drought tolerance correlated with an increase in ABA accumulation, and the genotype showing highest tolerance had fivefold more ABA levels as compared to the susceptible genotype.
At the molecular level, differences in gene expression in drought-susceptible and drought-tolerant genotypes have been observed. Generally, the genes involved in protecting plants from drought stress through stress perception, signal transduction, transcriptional regulatory networks in cellular responses or tolerance to dehydration were seen to be upregulated in drought-tolerant barley genotypes, while those concerned with primary metabolic processes like photosynthesis were downregulated (Guo et al. 2009). In tolerant land races of maize, genes encoding hormones, aquaporins, HSPs, LEAs and detoxification enzymes were induced to a greater extent than in the susceptible land races (Hayano-Kanashiro et al. 2009).
Many of the drought-related traits have been tagged using molecular markers, and the loci associated with these traits [quantitative trait loci, (QTLs)] have been used to select genotypes that are able to yield better under field drought conditions. For example, the ‘anthesis-silking interval’ typically increased under water deficit and negatively correlated with yield in maize (Duvick 2005). Screening genotypes for QTLs associated with lower anthesis-silking interval enabled the identification of genotypes showing better yields under water-limiting conditions. In sorghum, genotypes resistant to post-flowering drought stress, referred to as the stay-green phenotypes, have been shown to have a positive impact on yield under terminal drought. Four major QTLs designated as Stg2, Stg3 and Stg4 and additional minor QTLs were identified in sorghum, which modulate the expression of the stay-green trait (Harris et al. 2007). In rice, a QTL with a large effect on grain yield in upland rice growing under drought stress was associated with improved root architecture (Bernier et al. 2007). In maize, QTLs like root-ABA and root-yield-1.06 were identified, which were associated with root traits, ABA concentration as well as agronomic traits, especially grain yield across water regimes. These QTLs have been used to improve yield stability in maize under water-limiting conditions by marker-assisted selection (Landi et al. 2005, 2010). In cotton, QTLs for a physiological trait like low osmotic potential showed a strong association with plant height as well as with productivity in water-limiting conditions. Eleven QTLs associated with low osmotic potential were seen to be associated with thirteen QTLs associated with seed cotton yield (Saranga et al. 2004). Similarly, two significant QTLs affecting osmotic potential (qtlOP-2) and plant height (qtlPH-1) under drought conditions were also identified (Saeed et al. 2011). Such QTLs have been used for developing high-yielding cotton cultivars under water-stress conditions using marker-assisted selection. In wheat, two QTLs were found to be associated with plant height, kernel weight and yield under varying water availability (Maccaferri et al. 2008). However, contribution of QTLs to a trait is often low and QTLs associated with adaptive responses to drought differ across environments, while those that are constitutive are stable across environments (Collins et al. 2008). Dissecting the phenotypic traits into smaller and simpler traits, which show high heritability in genotypes exhibiting drought tolerance, has led to the identification of stable QTLs associated with these traits across diverse environments (Tardieu and Tuberosa 2010).
Identification of stable QTLs enables gene discovery through map-based cloning, and this serves as an important input in breeding for drought tolerance using transgenic approaches. Two approaches have been mainly used for the molecular dissection of a QTL: positional cloning and association mapping. Positional cloning enables the identification of the genetic and physical interval cosegregating with the QTL, while association mapping establishes a statistical association between allelic variation at a locus and the phenotypic value of a trait across a large number of unrelated accessions. Identification of the candidate genes associated with a QTL is difficult because a QTL is known to span a large genomic region. For example, a QTL was shown to span a region of over 12 Mb and 310 genes in maize (Salvi and Tuberosa 2005). A few genes identified from the QTL regions include the CRY2 gene that is involved in cryptochrome synthesis from the rice QTL for flowering time ED1, or a gene coding for a transcription factor from the plant architecture QTL Tb1 in maize (Salvi and Tuberosa 2005). The candidate genes or sequences that cosegregate with the QTL are then functionally tested with reverse genetics tools based on gene tagging, TILLING or RNAi and validated for function by producing transgenic plants (Tuberosa and Salvi 2006).
Transgenic Technology for Improved Drought Tolerance in Crops
Drought tolerance has been achieved using genetic engineering strategies to improve (i) water-use efficiency of plants, (ii) cell protection mechanisms against ROS, (iii) hormonal balance to alter the growth and development in order to avoid drought and (iv) alter the expression of drought-induced transcription factors that act as master switches in regulating a large number of downstream drought-response genes.
Late embryogenesis-abundant proteins are known to accumulate during seed desiccation and in vegetative tissues when plants experience water deficit. Transgenic expression of a group 3 LEA protein from barley (HVA1) showed improved drought and salt tolerance in rice and wheat plants (Xu et al. 1996, Sivamani et al. 2000). Overexpression of trehalose or polyamines was also seen to confer tolerance to abiotic stress in rice (Garg et al. 2002, Capell et al. 2004). Transgenic alfalfa plants overexpressing the antioxidant enzyme superoxide dismutase showed improved tolerance to drought stress (McKersie et al. 1996). Transgenic rice plants overexpressing the isopentenyl transferase (IPT) gene, which plays a role in cytokinin biosynthesis, showed increased expression of brassinosteroid-related genes and repression of jasmonate-related genes (Peleg et al. 2011). Besides alterations in hormone homoeostasis, the transgenic rice plants also showed a change in source–sink relationships and a stronger sink capacity when subjected to water limitation.
Attempts at overexpressing TFs that show higher expression under drought stress in tolerant as compared to susceptible genotypes (Hayano-Kanashiro et al. 2009) have led to an improvement of drought tolerance in several crops. Wheat transgenics expressing the DREB1 gene from Arabidopsis showed better tolerance to drought under glasshouse conditions (Pellegrineschi et al. 2004). Rice transgenics overexpressing ABA-inducible TF (ABF3) or drought-inducible TF (DREB2) showed improved survivability and significantly higher number of panicles, respectively, in response to drought stress, as compared to wild-type plants (Oh et al. 2005, Bihani et al. 2011). Overexpression of OsbZIP23 in rice exhibited significantly improved tolerance to drought and high salinity and sensitivity to ABA (Hadiarto and Tran 2011).
Although transgenic technologies provide a targeted approach for improving drought tolerance, the transgenic plants are often tested under ‘unnatural’ stress conditions and it is not clear whether they would also give rise to better yields under field stress conditions. However, such studies are important as they give an indication of genes that could serve as potential candidates for improving stress tolerance in crops, because the slow progression of dehydration that is seen in the field does not lead to drastic changes in gene expression that are observed in potted plants (Barker et al. 2005).
Climate Change and Crop Adaptation
Drought stress, especially in the tropics, is accompanied by high temperature stress, and the responses of crops to a combination of these two stress factors appear to differ from the responses to either of the stresses applied singly (Sreenivasulu et al. 2007). Hence, yield responses of crop plants when exposed to abiotic stress combinations may differ from individual stress exposures. Besides, climate change–induced higher temperatures are predicted to increase the water requirements of crops (Nelson et al. 2009). Exploiting the genetic variability available in crop species in adjusting to climate change may be a useful strategy for identifying traits contributing to improved tolerance to a combination of stresses expected to occur due to climate change. For example, pearl millet varieties have shown adaptation to persistent drought as well as high temperatures in Sahel region (Niger) of Africa. Changes in morphological and phenological characteristics (flowering time, plant height and spike length) in varieties sampled in 2003, during which drought and high temperatures prevailed as compared to the same varieties sampled in 1976, when such stress situations did not occur (Bezancon et al. 2009), showed a significant shift in adaptive traits. The varieties flowered slightly earlier and had shorter spikes in 2003 than in 1976, suggesting that selection for these traits occurred in the face of environmental change over this time period. Two genes, PHY and PgMADS11, that play a role in flowering time regulation were found to show polymorphism, which could also have arisen in response to selection. In the context of climate change, a shorter life cycle may mitigate the effect of climate change by allowing flowering and seed production in stressed environments. Similarly, there would be a large number of genes involved in different adaptive processes occurring in response to unpredictable stresses arising due to climate change, which could be mined by comparative studies on genotypes adapted to different environments.
A number of advancements have been made in our understanding of how a plant responds to drought stress. Adaptation to drought is seen to involve metabolic and morphological alterations that prevent injury to plants. Underlying these physiological and morphological alterations are molecular mechanisms that regulate the expression of genes involved in the various adaptive processes. Although much is known now about the different type of stress sensors, the secondary signalling molecules involved and entire stress-specific signalling pathways have not been deciphered, largely due to cross-talk between different stress-signalling pathways.
Stress-response gene expression is regulated largely by transcription factors, which in turn are subjected to very intricate regulation at the chromatin level, RNA level and protein level. Stress-induced chromatin remodelling may mediate acclimation responses and help a plant to cope better with subsequent stress situations. Micro-RNA-mediated gene silencing of stress-response TFs under non-stress conditions and their activation by downregulation of miRNA expression have emerged as another important means of regulating downstream stress-response gene expression.
Information on the stress adaptive mechanisms shown by drought-tolerant genotypes of crop species has been fragmentary. Gene expression studies in response to drought provide information on processes involved in stress tolerance, but the sheer magnitude of information generated in such studies makes it a daunting task to distinguish the adaptive responses from those that arise secondarily as an outcome of growth arrest or cell damage. Phenotypic traits associated with drought-tolerant crops serve as important breeding tools in identifying stress-tolerant genotypes and in introgressing the tolerance traits into cultivated genotypes. Dissecting these complex phenotypic traits into simpler, heritable traits has led to the identification of genes associated with some QTLs for drought tolerance. Understanding stress-tolerant strategies using model plants and testing these in crop genotypes that show adaptation to stress appear to be a useful approach in improving drought tolerance of crops. However for studies on adaptation of crop plants to complex stress situations arising due to climate change, there is a need to exploit the available biodiversity in crop genotypes growing in diverse environments to understand the mechanisms involved in coping with different stress combinations.
KS acknowledges University Grants Commission, Government of India, for financial assistance through award of a research fellowship.