Improving plant trade-off between assimilated carbon and water by using controlled irrigation
The understanding of the factors that regulate the trade-off
between carbon assimilation and water loss, and those that drive
partitioning of assimilates between reproductive and non-reproductive
structures in relation to water availability are essential to
identify the technologies for matching water input with plant
requirements. Irrigation strategies that exploit the knowledge
of a plant's long-distance signalling system are increasingly
being used to get improved crop water use efficiency under sustained
or improved quality of the product (Davies et al.
and Ping, 2002
). Indeed, it was demonstrated that large unregulated
fluxes of water are not essential to plant functioning and that
water can be saved by manipulating stomatal functioning (Loveys
and Davies, 2004
). A measure of successful regulation of carbon
assimilation under variable water availability is the plant
ability to maintain an equilibrium among the intervening processes,
diffusion, light harvesting, photochemistry, and
biochemistry (Geiger and Servaites, 1994
), so that the flux
through each component of the process is in balance with the
others, except for brief periods of transition. When water deficits
start to build up, leaf stomatal conductance usually decreases
faster than carbon assimilation, leading to increased water
use efficiency, WUE (Chaves et al.
). It is also well known
that when irrigation is above the optimum, an excessive shoot
growth can occur at the expense of roots and fruits (Zhang,
). Manipulation of pre- and post-flowering water use in
crops can be used to increase harvest index (HI) and by using
methods of controlled irrigation the optimized water use by
stomata can lead to an increase in WUE, without a significant
decrease in production and eventually with beneficial effects
Closure of stomata under dehydrating conditions is the resulteither from a feedback response to the generation of water deficitsin the leaf itself that is transmitted to the guard cells, orfrom a feed-forward control before any alteration in leaf tissuewater status takes place (Schulze, 1986). These feed-forwardresponses of guard cells comprise the responses to high vapourpressure deficit, whose mechanisms area still under debate (Franksand Farquhar, 1999) and dehydration taking place elsewhere inthe plant, namely in the roots (Davies and Zhang, 1991). Inaddition to stomatal closure, shoot growth is slowed down ata very early stage of water stress (Hsiao, 1973; Kramer, 1983).As discussed in the previous section, strong evidence has accumulatedsuggesting that this kind of response to decreasing soil watermay be mediated by long-distance signals produced in dryingroots, namely of chemical origin (such as the hormone ABA orcytokinins) and transported to the shoot in the transpirationstream (Wilkinson and Davies, 2002). They will provide to theshoot a measure of the water available in the soil. However,ABA signalling is a complex process which involves not onlythe up-regulation of ABA biosynthesis and transport via thexylem to the leaf, but ultimately depends on homeostasis ofxylem sap along the length of the transport system and on thevariable role of anion trapping (Wilkinson and Davies, 2002).In fact, a large proportion of ABA transported from the rootsis catabolized in the cells of the leaf in a process termedABA filtration (Wilkinson, 2004). The pH of the xylem sap andof the leaf apoplast was shown to prevent ABA from enteringthe apoplast via the xylem. This is based on the ‘aniontrap’ concept (Wilkinson and Davies, 2002), which establishesthat ABA accumulates in the most alkaline compartments of thecells. The arrival of these signals at the guard cells (Alvimet al., 2001) or the growing tissues (Wilkinson, 2004) is thereforeultimately governed by the apoplastic pH,. Environmental factors(such as PPFD, temperature or VPD) that influence shoot physiologicalprocesses will interact with factors that affect the rhizosphere,determining the final apoplastic pH. As a consequence, plantWUE will reflect the multiple environmental stimuli perceivedand the ability of the particular genotype to sense the onsetof changes in moisture availability and therefore fine-tuneits water status in response to the environment (Wilkinson,2004).
This knowledge has inspired a special kind of deficit irrigation,the so-called partial root-zone drying (PRD), where each sideof the root system is irrigated during alternate periods. InPRD the maintenance of the plant water status is insured bythe wet part of the root system, whereas the decrease in wateruse derives from the closure of stomata promoted by dehydratingroots (Davies et al., 2000). Large-scale implementation of PRDirrigation in vineyards has already taken place in Australia(Loveys and Ping, 2002). This irrigation type has been furtherstudied in grapevines (Souza et al., 2003; Santos et al., 2003)and in other crops, such as tomato (Davies et al., 2000; Mingoet al., 2003), raspberries (Grant et al., 2004), orange trees(Loveys and Davies, 2004) or olive trees (Mentritto et al.,unpublished data). Although the nature of the signals is nottotally clear, it is recognized that stomatal closure and growthinhibition are likely to be responding simultaneously to differentstimuli, some of which may operate through common signal transductionsystems (Webb and Hetherington, 1997; Shinozaki and Yamaguchi-Shinozaki,2000). Physiological data that are being accumulated (e.g. ingrapevines under PRD) point to subtle differences between PRDand the deficit irrigation (DI), where the same amount of wateris distributed by the two sides of the root system (Souza etal., 2003; Santos et al., 2003). These differences include somereduction of stomatal aperture in PRD (more apparent when measurementsof stomatal conductance are done under constant light and temperature,rather than under the fluctuating conditions prevailing in thefield), a depression of vegetative growth, and an increase incluster exposure to solar radiation, with some potential toimprove fruit quality (Table 1). An interesting finding is thelink found between the intensity of the PRD stomatal responseand VPD, high VPD intensifying PRD stomatal closure comparedwith the controls (Loveys and Davies, 2004). These authors suggestthat the enhanced response of stomata to VPD in PRD irrigationcould be related to an increased ability of the xylem to supplyABA.
Table 1. Effect of controlled irrigation on physiological responses of field-grown grapevines
There is also evidence that PRD can increase fruit quality intomato, presumably as a result of differential effects on vegetativeand reproductive production (Davies et al., 2000). The rootsystem also seems to be significantly altered in response topartial dehydration, not only in respect to total extensionand biomass but also in architecture (Dry et al., 2000; TPdSantos et al., unpublished results; MA Bacon and WJ Davies,personal communication). It is likely that this alteration inthe root characteristics and in the source/sink balance playsan important role in plant performance under PRD.
In some crops, such as cereals (Blum et al., 1994; Gent, 1994)and some legumes (Chaves et al., 2002), reserves accumulatedin the stem before anthesis can be utilized for grain fillingin addition to current assimilates, therefore contributing toimportant gains in HI. Under stress conditions (Blum et al.,1994) or high respiration rates (for example, high temperatures)stem reserves are essential to complete grain filling (Gent,1994). The potential for storing reserves in the stem is dependenton stem length and weight density, although these characteristicsper se are not sufficient to ensure that those reserves wouldbe translocated to the fruit. Mobilization of reserves is dependenton sink strength, which varies with the genotype and is affectedby the environment (e.g. water availability). On the other hand,the stem (in particular, the stem stele, which is associatedwith the vascular tissue) is especially well protected againstenvironmental stress. In fact, studies in lupin subjected todrought indicated that the stem stele never dropped its relativewater content (RWC) below 83%, whereas the other organs in theplant exhibited values below 60%, namely the leaves 57%, theroots 58%, and the stem cortex 58% (Pinheiro et al., 2004).It can be speculated that this response is associated with theprotection given by the accumulation of assimilates, mainlyglucose, fructose, and sucrose whose concentration in the stemstele doubles under water deficits (Pinheiro et al., 2001).These sugars could also act as signals for the observed inductionof protective proteins such as late embryogenesis abundant (LEA)proteins, much more pronounced in the stele than in the cortex(C Pinheiro et al., unpublished data).
Controlled soil drying was shown to promote the remobilizationof carbon reserves during late grain filling in wheat and improveHI, especially when the crop is grown under high nitrogen (Yanget al., 2000, 2001). In fact, under such conditions, a mildsoil drying counteracts the delay in senescence of vegetativetissues that usually accompanies the heavy use of N, and improvesremobilization of stem reserves to the grains. Stay-green fortoo long results in the non-remobilization of pre-anthesis reservesin leaves, glumes, and stems, which may account for 30–47%of the carbon in protein and 8–27% of the carbon in carbohydratesdeposited in the grain (Gebbing and Schnyder, 1999). In China,if crop maturation is delayed, dry winds at the end of the growingseason can dehydrate wheat very rapidly and reduce grain yield.Yang et al. (2001) showed that, by applying a moderate soildrying and thus inducing an earlier senescence, they could accelerategrain filling and therefore improve yield.
However, in regions without the constraints described aboveextending the grain filling period, and therefore delaying leafsenescence, could benefit yield by allowing more time for thetranslocation of assimilates to the grain (Richards et al.,2001). This can be achieved either by controlling irrigationand/or by selecting genotypes for stay-green capability.
Genetic engineering for improved plant response to water deficit: recent advances
In the past decade most of the genetic engineering work thathas been successful in agricultural terms was directed towardscrop resistance to biotic stresses or to technological properties(see the review by Sonnewald, 2003). The studies addressingplant resistance to abiotic stress, namely in relation to drought,have been confined so far to experimental laboratory work andto single gene approaches, which has led to marginal stressimprovement (Ramanjulu and Bartels, 2002). However, recent advancessuggest that rapid progress will be possible in the near future,with large economical impact in many areas of the globe (Dunwell,2000; Garg et al., 2002; Wang et al., 2003) (Table 2). In fact,even modest improvements in crop resistance to water deficitsand in water use efficiency will increase yield and save water.One of the major challenges of this technology is to developplants not only able to survive stress, but also able to growunder adverse conditions with reasonable biomass production,overcoming the negative correlation between drought resistanttraits and productivity, which was often present in past breedingprogrammes (Mitra, 2001). Such a compromise requires improvedefficiency in maintaining homeostasis, detoxifying cells fromharmful elements (like ROS), and recovering growth that is arrestedupon acute osmotic stress (Xiong and Zhu, 2002). This also meansthat there is the need to introduce sets of genes that governquantitative traits, a technological approach that has alreadyproved to be successful, for example, in the case of transgenicrice with introduced provitamin A (Ye et al., 2000). The progressivecloning of many stress-related genes and responsive elements,and the proof of their association to stress-tolerant QTLs (QuantitativeTrait Loci), suggests that these genes may represent the molecularbasis of stress tolerance (Cattivell et al., 2002). On the otherhand, the identification of QTLs associated with drought toleranceis also an important tool for marker-assisted selection (MAS)of tolerant plants. These studies have been conducted on a broadvariety of species (see for instance Casasoli et al., 2004;Lanceras et al., 2004; Tuberosa et al., 2002). A lot of workhas been done on this topic and will not be covered here; however,it is clear that the combination of traditional and molecularbreeding (MAS and genetic engineering) will allow a more rapidway to improve abiotic stress tolerance in agricultural crops.
Table 2. Recent achievements in improving drought tolerance in crops through genetic engineering
The increasing knowledge of stress adaptation processes andthe identification of key pathways and interactions involvedin the plant response to the stress conditions is being exploitedto engineer plants with higher tissue tolerance to dehydrationor with drought avoidance characteristics (Laporte et al., 2002).The latter is, of course, more difficult to achieve, becauseit is linked to whole-plant morphological and physiologicalcharacteristics (Altman, 2003).
The recent progress in gene discovery and knowledge of signaltransduction pathways is raising the possibility of engineeringimportant traits by manipulation of one single gene, downstreamof signalling cascades, with putative impact on more than onestress type. Moreover, in genetic engineering, it is importantto mimic nature and activate, at the correct time, only thegenes that are necessary to protect the plants against stresseffects. This may be achieved by using appropriate stress-induciblepromoters and will minimize effects on growth under non-stressingconditions, which is essential for agricultural crops. It isalso desirable to target the desired tissue/cellular location,to control the intensity and time of expression, and to ensurethat all the metabolic intermediates are available, so thatno negative effects will arise (Holmberg and Bulow, 1998). Finally,to be able to prove that a transgenic plant is more resistantto water stress than the wild type, one needs a rigorous evaluationof the physiological performance as well as the water statusof transformed plants. This will avoid ambiguous interpretationsof the gene effects on plant drought resistance, such as thoseoften appearing in the literature (see, for example, the commentby Blum in www.plantstress.com/admin/Files/Hsieh_PlantPhysiol_130.htmand Hsieh et al., 2002). In other words, the impact of the introducedgenes must be separated in their direct versus indirect effects(for example, increased resistance of the photosynthetic apparatusversus effects on plant or leaf size, phenology etc.).
Among the genes that are known to respond to drought stressand which are being manipulated by genetic engineering, someencode enzymes involved in metabolism (for example, linked todetoxification or osmotic response), others are active in signalling,or in the transport of metabolites (for example, the prolinetransporter) or in regulating plant energy status. Some genesdo not have a well-established function, such as those encodingthe LEA proteins, but result in protection of the cellular machineryagainst various stresses (Bray, 1997; Xu et al., 1996).
Engineering for osmotic adjustment and/or protection of macromolecules:
Engineering for increasing osmolytes, such as mannitol, fructans,trehalose, ononitol, proline, or glycinebetaine, among othersmay increase resistance to drought, although the protectionmechanisms are still not fully understood (Ramanjulu and Bartels,2002). If the osmolyte accumulation is sufficient to decreasecell osmotic potential thereby enabling the maintenance of waterabsorption and cell turgor at lower water potentials (Morgan,1984), one can talk of osmotic adjustment. When the accumulationis low, it is reasonable to ascribe osmolytes a function inprotecting macromolecules (such as, for example, enzymes) eitherby stabilizing proteins or by scavenging reactive oxygen speciesproduced under drought (Shen et al., 1997a; Zhu, 2001b). Althoughthe benefits of osmolyte accumulation for crop yield are thesubject of some controversy (Serraj and Sinclair, 2002), someresults of genetic transformation point to advantages for plantperformance under drought, which may open avenues for the future.Still, transgenic plants that have been engineered to overproduceosmolytes often exhibit impaired growth in the absence of stress.This is probably due to the involvement of osmolytes in signalling/regulatingplant responses to multiple stresses, including reduced growththat may be part of the plant adaptation strategy against stress,as suggested by Maggio et al. (2002).
The raffinose family oligosaccharides, such as raffinose andgalactinol, are among the sugars involved in desiccation tolerance.Taji et al. (2002) engineered Arabidopsis plants for over-expressionof AtGolS 1, 2, or 3, all genes coding for galactinol synthasefrom A. thaliana. The overexpression of AtGolS2 did increaseendogenous galactinol and raffinose in transgenic plants andwas found to reduce transpiration from leaves and to improvedrought tolerance. These compounds seem to act as osmoprotectants,rather than by providing osmotic adjustment (Taji et al., 2002).
Expression of bacterial fructan in tobacco and sugar beet ledto an improved growth under water deficits in transgenic plantsthan in the wild type (Pilon-Smits et al., 1995, 1998).
Mannitol, the most widely distributed sugar alcohol in nature(Stoop et al., 1996), was demonstrated to scavenge hydroxylradicals and stabilize macromolecular structures, such as phosphoribulokinase(a thiol-regulated enzyme), thioredoxin, ferredoxin, and glutathione(see for example, Shen et al., 1997a, b). The protective effectseems to result from the formation of hydrogen bonds betweenmacromolecules and osmolytes under limited water availability,thus preventing the formation of intramolecular H-bonds thatcould irreversibly modify the three-dimensional molecular structures.Recently, Abebe et al. (2003) achieved a significant improvementof wheat tolerance to water and salt stress through the ectopicexpression of the mtlD gene (mannitol-1-phosphate dehydrogenase)from E. coli. The authors found that the amount of mannitolaccumulated (0.6–2.0 µmol g–1 FW) was toolow to ensure protection through osmotic adjustment, but waseffective in improving stress tolerance. Lines containing over0.7 µmol g–1 FW in the flag leaf, started showingside effects of mannitol accumulation and lines with over 1.6µmol g–1 FW in the flag leaf showed severe abnormalities,including sterility. This was accompanied by an exceptionallylow sucrose content. The plants with the lower mannitol contents(up to 0.7 µmol g–1 FW), however, did not sufferfrom the adverse effects of excess mannitol, which would depletethe sucrose pool and negatively impact the growth of wheat plants.Because mannitol is a naturally occurring sugar-alcohol andis used as an additive in many processed foods, its overexpressionmay prove to be a useful tool to enhance crop resistance todrought and salt. The overexpression of IMTI1 (inositol methyltransferase) gene, from the ice plant Mesembryanthemum crystallinuminto tobacco, led to the accumulation of another sugar-alcohol,the methylated form of inositol, D-ononitol, leading to an increasedtolerance to drought and salt stress (Sheveleva et al., 1997).
Trehalose, a non-reducing disaccharide of glucose, has beenshown to stabilize biological structures and macromolecules(proteins, membrane lipids) in different organisms during dehydration(Crowe et al., 1992). Through the regulated over-expressionof a fusion gene containing the coding regions of both otsAand otsB (trehalose-6-P synthase and trehalose 6-P-phosphatase)of E. coli, Garg et al. (2002) showed that trehalose has a primarypositive effect in transformed plants under abiotic stress conditions.This effect was linked to the maintenance of an elevated capacityfor photosynthesis under stress. The positive effect of trehaloseaccumulation (an increase in 3–9-fold compared with thewild type) was observed under salt, drought, and low-temperatureconditions. Under drought, trehalose accumulation accountedfor an increased protection of Photosystem II against photo-oxidativedamage, as assessed by in vivo chlorophyll fluorescence (PSIIand Fv/Fm). These effects were observed both when the fusiongene was directed to the chloroplast (with a transit peptideand under the control of the promoter of the small subunit ofrbcS) or to the cytosol (under the control of an ABA-induciblepromoter). The reason why photosynthetic capacity was preservedin drought-stressed transgenic rice is, however, not clear;is it because shoot water status was improved, or is it simplybecause, under a dehydration intensity similar to that affectingthe wild-type plants, the photosynthetic apparatus is protectedagainst oxidative stress? It may be speculated that becausethe transgenic lines with gene expression in the chloroplastshowed protection against drought at lower trehalose concentrationsthan those with cytosolic expression, the second hypothesisis the most likely.
Garg et al. (2002) also found an increase in other soluble carbohydratesafter exposure to abiotic stress (20% higher concentrationsin transformed than in wild-type plants). These results areconsistent with the hypothesis raised by Paul et al. (2001),working with tobacco plants expressing E. coli trehalose biosyntheticgenes, that trehalose may play a role in the modulation of carbonmetabolism in response to external factors, through sugar-sensingmechanisms. The work by Garg et al. (2002) confirmed some beneficialeffects observed in earlier transformation work done by Pilon-Smitset al. (1998) in tobacco. However, very significant progresswas achieved by comparison with previous studies, where undesirablepleiotropic effects, including stunted growth and the formationof abnormal leaves, occurred in plants where the two enzymesinvolved in the trehalose biosynthesis were overexpressed (Goddijnet al., 1997; Holmstrom et al., 1996). If these studies areconfirmed by field trials, they increase the possibility forcultivating rice, a major staple crop worldwide, in rainfedconditions or in saline soils (Penna, 2003).
Betaines, ectoine, and proline are among the compatible solutesthat also accumulate in plants as a widespread response againstenvironmental stress (Chen and Murata, 2002; Rontein et al.,2002). Some crop plants have low levels of these compounds,and engineering their biosynthetic pathways is a potential wayto improve stress tolerance. For instance, in wheat, the accumulationand mobilization of proline was found to correlate with thelevel of tolerance towards water stress (Nayyar and Walia, 2003),the tolerant genotype being more responsive to ABA. Overexpressingthe gene P5CS from Vigna aconitifolia in tobacco led to a 2-foldincrease in proline and a better growth under water and saltstress (Kavi Kishor et al., 1995). A number of genes involvedin the biosynthetic pathways of such compounds, such as choline-oxidaseor sorbitol-6-phosphate dehydrogenase, have been tested in transgenicplants with positive results in increasing stress tolerance(Chen and Murata, 2002). In some cases, the accumulation ofthese solutes is marginal, implying that they were not actingthrough an effect of osmotic adjustment (Holmstrom et al., 1996).
A group of proteins commonly involved in the enhancement ofstress tolerance are the LEA proteins. The role of LEA proteinswas suggested as chaperones, in binding water, in protein ormembrane stabilization, and in ion sequestration (Cushman andBohnert, 2000). Rice and wheat plants expressing the barleygroup 3 LEA gene HVA1 in leaves and roots showed improved osmoticstress tolerance and improved recovery after drought and salinitystress (Xu et al., 1996; Sivamani et al., 2000). Group 2 ofthe LEA proteins, the dehydrins (also known as the Lea D11 family)has been commonly observed accumulating in response to dehydrationor low temperature (Close, 1997). With one or more copies ofa putative amphipathic -helix-forming domain (the K-segment),dehydrins are the best-studied LEA proteins. They have beenconsidered as having a role as surfactants, preventing the coagulationof numerous macromolecules (Close, 1997).
Other proteins may also play a role in protection against drought.This is the case of some heat shock (HS) proteins, includingsmall HS (smHS) such as the At-HSP17.6A class from Arabidopsisthaliana, which, upon over-expression, could increase salt anddrought tolerance, presumably due to its chaperone activitydemonstrated in vitro (Sun et al., 2001). Their action includespreventing protein degradation and assisting the refolding ofproteins denaturated during stress. In transgenic tobacco plants,the enhanced accumulation of the chaperone-binding protein BiP,of the endoplasmic reticulum (shown to be induced by a varietyof environmental stresses), conferred tolerance to water stress(Alvim et al., 2001). Under progressive drought, leaf BiPs concentrationwas correlated with shoot water content and photosynthetic rateswere maintained in stressed transgenic plants to values similarto those measured in wild-type well-watered plants.
NtC7, a gene encoding a membrane-located receptor-like protein,with transmembrane domains, was also found to induce, in transgenictobacco plants, a marked increase in tolerance to mannitol-inducedosmotic stress, with rapid recovery from severe wilting, whereaswild-type plants showed leaf necrosis (Tamura et al., 2003).The authors suggested that the NtC7 gene is involved in thesignalling pathway that activates genes responsive to osmoticstress (independently of ion homeostasis), presumably as partof the osmosensor system. Osmotic adaptation may occur throughmechanic-sensitive signalling, in which alterations in turgorcould be the starting point for a signalling cascade, by generatinga signal eventually triggering conformational changes in membraneproteins. In potato, mechanical stress has an early cellularresponse of the significant and rapid synthesis of superoxideradicals (Johnson et al., 2003).
Protection against excessive accumulation of ROS has been achievedby overexpressing a stress-inducible aldehyde dehydrogenasegene, already present in Arabidopsis thaliana (Sunkar et al.,2003). The function of this enzyme is to catalyse the oxidationof various toxic aldehydes, accumulated as a result of sidereactions of ROS with lipids and proteins. Transgenic linesshowed improved tolerance when exposed to dehydration, as wellas to other types of stress (salt, heavy metals, H2O2) and thiswas accompanied by a decreased accumulation of lipid peroxidation-derivedtoxic aldehydes. Transgenics also survived for longer periodsof drought than wild-type plants. The authors claim that thesefindings may lead to applications in crop plants, such as maize,wheat or soybean. In addition, the ectopic expression, in tobacco,of the alfalfa aldose/aldehyde reductase MsALR, provided toleranceto multiple stresses, including drought stress, with reducedamounts of reactive aldehydes generated from lipid peroxidation(Oberschall et al., 2000). Manipulation of ROS scavenging enzymes,yielding the effective reduction of ROS concentration, however,may lead to increased susceptibility to biotic stress, sincecell wall fortification, as a barrier to pathogen penetration,is increased by ROS (Xiong et al., 2002). On the other hand,manipulation of ROS scavenging enzymes aiming to reduce oxidativedamage is limited by the high number of isoforms and by theirlocation in different sub-compartments and membranes (Bohnertand Sheveleva, 1998).
Engineering for water transporters:
Water transport in plants uses both the apoplastic and the symplasticroutes. This means that a high number of water molecules haveto cross numerous cell membranes. This process is facilitatedby aquaporins, membrane-intrinsic proteins found in all livingorganisms and forming water-permeable complexes (Uehlein etal., 2003). The apoplastic water potential influences the phosphorylationstatus of aquaporins, so that its ability to transport waterincreases when phosphorylated. Therefore, aquaporins are likelyto play an important role in the control of cellular water statusin response to water deficits (Assmann and Haubrick, 1996; Bray,1997). Differential expression of genes that encode differentaquaporin isoforms during plant development were shown to beassociated with different physiological processes, includingstomatal opening (Chrispeels and Agre, 1994). However, the relationshipbetween the role of aquaporins in the regulation of plant waterstatus and the regulation of aquaporin gene expression is stillunclear (Aharon et al., 2003). For example, the over-expressionin tobacco of the Arabidopsis aquaporin AthH2, which encodesPIP1b aquaporin, improved growth performance under non-stressconditions, but it was not effective under drought or salt stress(Aharon et al., 2003).
Aquaporins may also transport other small molecules such asglycerol, solutes and ions (Tyerman et al., 2002) and they showcytosolic pH-dependent gating (changes in the conductance ofindividual water channels), a feature providing a mechanismof co-ordinated inhibition of plasma membrane aquaporins uponcytosol acidosis (Tournaire-Roux et al., 2003). This behaviourjustifies the reduced ability of roots to absorb water underflooding conditions, as a consequence of anoxia.
Recently it was found that the tobacco aquaporin NtAQP1 actsas a CO2 membrane-transport-facilitating protein, playing asignificant role in photosynthesis and in stomatal opening (Uehleinet al., 2003). The overexpression of NtAQP1 in tobacco raisedmembrane permeability for CO2 and water, and increased leafgrowth (Uehlein et al., 2003), a feature that may have an impactin plant performance under drought. Photosynthesis increasedin these transgenic plants by 36% under ambient CO2 (380 ppm)and by 81% at elevated CO2 (810 ppm). This was accompanied byan increase in stomatal conductance in both situations. Therefore,the increase in photosynthesis may result from a combinationof more open stomata and a higher mesophyll conductance, resultingfrom the decreased membrane resistance to CO2. Both effectsled to an increase in CO2 availability to the cells.
Engineering for C4 traits:
The ability to optimize net carbon gain and therefore increaseWUE under reduced water availability is critical for plant survival(Chaves et al., 2004). In species with C4 photosynthesis highphotosynthetic rates can be associated with low stomatal conductance,leading to high WUE (Cowan and Farquhar, 1977; Schulze and Hall,1982). Manipulating WUE is a highly complex desideratum, becauseit implies co-ordinated changes relating to stomatal apertureand photosynthesis. Following various attempts to use conventionalhybridization to get C3–C4 hybrids, several groups havesuccessfully transformed C3 plants to acquire C4 characteristics(see the review by Matsuoka et al., 2001). Ku et al. (1999),for example, introduced in rice the phosphoenolpyruvate carboxylase(PEPC) from maize, achieving a high-level expression of thePEPC protein (1–3-fold that of maize leaves). Althoughno significant effects were observed in the rates of photosynthesis,the transformed rice plants exhibited a reduction in the O2inhibition of photosynthesis characteristic of C3 plants thatmay attain 40% of potential photosynthesis. These transgenicplants may theoretically have some advantage over the wild type,especially under low CO2 conditions, prevalent for example underwater deficits, when carbon loss associated with photorespirationbecomes maximal. Some beneficial effects of the introductionof PEPC were observed under supra-optimal temperatures in transgenictobacco and potato (see Matsuoka et al., 2001). The hypothesisunderlying this response is that PEPC participates in the initialCO2 fixation or it increases CO2 in the vicinity of Rubisco.
A recent paper by von Caemmerer (2003) suggests, based on amodelling exercise, that C4 photosynthesis in a single C3 cell,although theoretically inefficient due to the absence of appropriatestructural features of C4 plants (see the review by Leegood,2002), may ameliorate the CO2-diffusion limitations of C3 leaves.Again, this could be beneficial under water-limited conditions,when stomata close and intercellular CO2 decreases drastically.
An alternative strategy to improve WUE would be to enhance photosyntheticcapacity in C3 crop plants by expressing improved forms of Rubisco,exhibiting higher relative specificity for CO2 compared withO2, such as those encountered in rodophyte algae, or to increasethe catalytic rate of Rubisco (Spreitzer and Salvucci, 2002;Parry et al., 2003). There is also scope for over-expressingRubisco activase, which seems to be more susceptible to extremeenvironments, namely high temperatures (Feller et al., 1998;Rokka et al., 2001).
Engineering via signal components and transcription factors:
In spite of the complex nature of the physiological adaptationof plants to the stress conditions and the difficulty of understandingthe regulatory mechanisms behind adaptation, there are alreadya number of genes that have been found to be involved in thesignal transduction pathways. They play important roles downstreamof signalling cascades, which could be used to engineer a higherability for plant protection from abiotic stress (Iba, 2002;Zhu, 2002). The modulation of these genes has been reportedto improve abiotic stress tolerance in a number of plant specieswith positive effects, sometimes regarding more than one stresstype (Dubouzet et al., 2003).
Multiple stress stimuli lead to Ca2+ influx in the cell andto its increased concentration in the cytoplasm. A number oftransport proteins such as the aquaporins, H+-ATPases and ionchannels, responsible for cytosolic osmoregulation and involvedin stress adaptation, are regulated by calcium-dependent proteinkinases (CDPKs). Saijo et al. (2000) investigated the functionof the rice cold- and salt-inducible OsCDPK7, and found thatits over-expression in transgenic rice plants conferred saltand drought-tolerance, apparently through the induced expressionof LEA proteins, namely rab16A (group 2 LEA protein), salT (aglycine-rich protein) and wsi18 (group 3 LEA protein). Thiseffect, however, was only observed in the rice cells after stressstimuli, pointing to a strong post-translational control andOsCDPK7 activation after the stress-induced calcium influx.The over-expression of OsCDPK7 did not significantly affectplant development and fertility.
The transfer of individual genes to plants, for acquiring higherstress tolerance, has so far only had a limited impact; however,the simultaneous transcriptional activation of a subset of thosegenes, by transferring transcription factors, has been revealedas a promising strategy (Jaglo-Ottosen et al., 1998; Liu etal., 1998).
There are several classes of transcription factors (TFs) playingmajor roles in dehydration and desiccation (Ramanjulu and Bartels,2002). In Arabidopsis, the TFs DREBs/CBFs specifically interactwith the dehydration responsive element/C repeat (DRE/CRT) cis-activeelement, controlling the expression of many stress-induciblegenes. DREB/CBF proteins are encoded by AP2/EREBP multigenefamilies and mediate the transcription of a number of genes,such as rd29A, rd17, cor6.6, cor15a, erd10, kin1, kin2, andothers, in response to cold and water stress (Ingram and Bartels,1996; Liu et al., 1998; Seki et al., 2001; Thomashow et al.,2001). A novel transcriptional regulator of the DRE/CRT classof genes, FIERY2 (FRY2), acts by repressing stress inductionof the upstream DREBs/CBFs TFs (Xiong et al., 2002). Recessivemutations in FRY2 result in super-induction of the DRE/CRT classof stress-responsive genes. Because FRY2/CPL1 contains dsRNA-bindingdomains, Xiong and Zhu (2002) speculated that dsRNA could bea regulator of the phosphatase enzymatic activity of FRY2/CPL1.RNA could then regulate hormone and stress responses in plants,as it does in animals. As cited by Xiong and Zhu (2002), somecomponents in mRNA processing (such as the cap-binding proteinABH1 and Sm-like snRNP protein SADI) are specifically involvedin ABA and stress responses.
The over-expression in Arabidopsis of DREB1 and DREB2 improvedtolerance to dehydration (Liu et al., 1998). Under the controlof a constitutive promoter, DREB1A was, however, detrimentalwhen stress was not applied, although it had a positive effectfor plants under stress. The use of the stress-inducible promoterrd29A, instead of the CaMV 35S promoter, to over-express DREB1Aminimized the negative effects on plant growth (Kasuga et al.,1999). DREB genes under the control of rd29A are presently beingtested on tropical rice (Datta, 2002).
The Arabidopsis CBF1 (DREB1B) ectopically expressed in tomato,resulted in enhanced resistance to water-deficit, although growthretardation was observed as well as reduced fruit and seed numberswhen under the control of the 35S promoter (Hsieh et al., 2002).An ABA-inducible promoter did not affect plant morphology orgrowth, but was less effective under stress conditions. In transgenicCBF1 tomato under water-deficit, stomata closed faster thanin wild-type plants and proline concentration was higher, whilecatalase activity increased and H2O2 decreased compared withwild plants. Another gene, CBF4, found to be up-regulated onlyby drought (and not cold) when over-expressed in transgenicArabidopsis was able to activate genes involved in both droughtadaptation and cold acclimation (Haake et al., 2002). The authorsproposed that plant responses to cold and drought evolved froma common CBF-like transcription factor, first through gene duplicationand then through promoter evolution.
An homologous gene isolated from rice, OsDREB1A, and testedin Arabidopsis indicated a functional similarity to the ArabidopsisDREB1A, although in microarray and RNA blot analyses some differenceswere observed regarding the induced target genes (Dubouzet etal., 2003). The authors suggested that OsDREB1A is potentiallyuseful for producing transgenic monocots tolerant to drought,high-salt, and/or cold stresses.
Improved osmotic stress tolerance was achieved by 35S:AtMYC2/AtMYB2in transgenic plants, as assessed by electrolyte-leakage tests(Abe et al., 2003). Constitutive expression of TFs, however,usually leads to growth retardation (Abe et al., 2003; Hsiehet al., 2002; Kasuga et al., 1999). The Arabidopsis MYB TF proteinsAtMYC2 and AtMYB2 were found to function as transcriptionalactivators in ABA-inducible gene expression (Abe et al., 2003).This role points to a novel regulatory system for gene expressionin response to ABA, other than the ABRE (abscisic acid responsiveelement)-ZIP regulatory system (Wang et al., 2003).
The over-expression of bZIP (basic region leucine zipper) TFs,
binding to ABRE cis
-elements (e.g. ABF3
found to cause ABA hypersensitivity, reduced transpiration rate,
and enhanced drought tolerance in transgenic plants (Kang etal.