Plant adaptation to O2-deficient waterlogged soils and flood-prone environments involves a suit of morphological, anatomical, and metabolic traits, as outlined in following sections. Plants need to cope with tissue anoxia, or avoid this adverse condition via a well-developed system of inter-connected gas-filled channels (aerenchyma) for internal O2 transport to supply submerged parts. Even species with large volumes of aerenchyma can experience anoxia in parts of their body, if only transiently. For these anoxic cells and tissues to survive, acclimative metabolic responses are essential. Furthermore, in addition to O2 deficits, plants must also cope with increased free radicals and reduced uptake of nutrients as additional components of flooding stress.
Brian Atwell, Macquarie University, Australia
Flooding is an environmental hazard writ large across the agricultural regions of the planet. Major inundations constrain food production because few of the world’s crop species are hydrophytes (wetland plants). It is essential to discover mechanisms of tolerance that can be used as a basis for breeding and engineering flood tolerance in rice and also other crops.
Rice (Oryza spp.) is notable among the staple crops of the world for its extreme flood tolerance. Thus it has been a natural choice for discovery of flood-tolerance mechanisms. More specifically, the coleoptile of rice is subject in nature to extreme hypoxia or even anoxia and over thousands of years these organs have concentrated gene expression patterns that confer tolerance to O2 deficits.
The coleoptile, a small sheath that emerges from the seed (caryopsis) as a ‘proto-shoot’, is especially critical in this scenario because it grows preferentially while roots and leaves remain suppressed without O2 present. Coleoptiles also have special characteristics to make their growth energetically ‘cheap’. They consist of pre-formed cells and can increase in length by 30% each hour (Figure 1 of this Case Study) while the organ as a whole can elongate by 3 cm in a day under water. This illustrates that coleoptiles have evolved to be a perfect emergent organ, providing a conduit from anoxic soil/hypoxic floodwater to the water surface above. Reaching O2 is essential before energy deficits overwhelm the embryonic meristems. Indeed, this specific role of coleoptiles was called a ‘snorkel’ almost 50 years ago (Kordan, 1974) because it was realised that on reaching a source of O2, other organs like leaves and roots could grow.
Figure 1. A time series of rice seedlings growing in hypoxic stagnant solution for 6 days. Note the lack of roots or true leaf. The seedling on the far right has commenced a new phase of development, with the production of a mesocotyl at the base of the coleoptile. This coleoptilar node is the first true shoot meristem. (Photograph courtesy of R. Oldfield and B.J. Atwell).
Fast coleoptile growth under water is stimulated by ethylene. The build up of this phytohormone causes coleoptiles of submergence-tolerant genotypes to ‘stretch’ towards to the water surface. Hence, rice coleoptiles constitute an example of the ‘escape’ strategy. As described in Case Study 1, a failure to perceive ethylene in shoots ensures survival in mature plants under long-term floods; the presence of the Sub1A gene induces dormancy and conserves carbohydrates. By contrast with that quiescent strategy, in areas such as river deltas with prolonged deep floods, internode elongation is beneficial for escape and with the discovery of two snorkel genes, has also been shown to be an ethylene-mediated phenomenon.
Mechanisms of anoxia tolerance in plants have been elucidated through studies of coleoptiles. Ethanolic fermentation can accelerate in anoxia because the Pasteur Effect speeds up glycolysis, using carbohydrates from the seed reserves. This provides a modestly better ATP supply than a non-hydrophyte could generate. Indeed, while ethanol was long been thought to be the dominant player in fermentation, other pathways involving haemoglobin and nitric oxide are now being invoked in energy production under anoxia. The realisation that maintenance and even growth in anoxia could be achieved by preserving critical energetically demanding reactions at the cost of non-essential reactions has opened new avenues of research. For example, we now know that protein synthesis becomes a very dominant use of ATP in anoxia while many other energised reactions succumb to energy shortages. Modification of the hierarchy of energy use occurs in animals and could be an evolutionary step unique to hydrophytic plants.
Ultimately, membrane potentials are vital for all living cells. The cost of maintaining potentials has been studied in depth, including under anoxia in coleoptiles. Proton pumping between compartments is vital and in anoxia, ATP-driven proton pumps become disabled. This induces a fall in the vital cytosolic pH but this fall is modest in rice whereas acidification can be lethal in intolerant species. Most remarkably, pyrophosphate is an alternative energy source to ATP for pumping protons in plants. This relatively ‘cheap’ energy source appears to energise proton pumping into the cell vacuole by a unique protein that is expressed most strongly in anoxia-tolerant cultivars of rice and is inducible within < 2 h of imposition of anoxia. Genetically knocking out the gene for this tonoplastic proton pump increases sensitivity to anoxia. Thus, genes encoding a pyrophosphate-driven pump at the tonoplast have been identified as targets for improving survival during flooding and have also recently been implicated in regulation of Na+ compartmentation crucial also to salinity tolerance.
This short account of some of the applications of rice coleoptiles to improve our understanding of stress tolerance illustrates the power of an ideal model. That one simple, undifferentiated organ could instruct us about hormone physiology, energy metabolism and membrane integrity is truly remarkable. This does not preclude the need for studies in more organisms, including wild plants or animals, but does reinforce the importance of model systems
Kordan HA (1974) The rice shoot in relation to oxygen supply and root growth in seedlings germinating under water. New Phytol 73: 695–697
Edwards JM, Roberts TH, Atwell BJ (2012) Quantifying ATP turnover in anoxic coleoptiles of rice (Oryza sativa) demonstrates preferential allocation of energy to protein synthesis. J Exp Bot 63: 4389-4402
When plants are completely submerged and in darkness, so that no photosynthesis occurs, O2 can become exhausted by respiration resulting in tissue anoxia, especially in tissues/organs buried in anoxic soil. Prolonged anoxia is tolerated by rhizomes, tubers and some shoot organs of wetland species, and by germinating seeds of rice and some paddy weeds (e.g. barnyard grass). More commonly, anoxia can occur in portions of the plant body (e.g. roots), or parts of tissues within roots. ‘Anoxic cores’, coexistence of an anoxic stele and aerobic cortex, were demonstrated for maize roots in hypoxic conditions, using O2-microelectrodes (Figure 18.5) and biochemical indicators of fermentative metabolism (Thomson and Greenway 1991).
Figure 18.5. O2 concentration (mM) measured across a maize seedling root by using an O2-microelectrode. The profile was taken radially through differentiated tissues 75 mm behind the apex of a 135 mm long root. O2 concentration in the bathing medium was about 0.05 mM (hypoxia), so that the cortex received O2 whereas the stele had an ‘anoxic core’. The abrupt gradient in O2 status results from the lower porosity and higher metabolic demand in steal tissue as compared with the cortex. (Profile reproduced from Gibbs et al., 1988 with courtesy of W. Armstrong).
Roots in drained soil respire by catabolising carbohydrates in the tricarboxylic acid (TCA) cycle, with the ‘reducing power’ produced used in the electron transport chain (ETC) with O2 as the terminal electron acceptor. Energy in the form of ATP is generated, predominantly through oxidative phosphorylation in mitochondria (Chapter 2.4). However, in waterlogged soils O2 is scarce or even absent and therefore respiration is inhibited. Carbohydrates are then broken down via fermentative pathways to yield at least some ATP, produced during substrate-level phosphorylation in glycolysis (Figure 18.6). Conversion of the pyruvate to an end-product, such as ethanol, is essential to remove this metabolite as well as to recycle the NADH to NAD+, so that the pathway can continue to flow. Breakdown of carbohydrates to ethanol and CO2 is the principal fermentative pathway in plants. Some lactate and alanine are also produced, but in contrast to fermentation leading to lactate and alanine, ethanolic fermentation can be sustained over days in anoxic tissues, end-product feedback or even toxicity being minimised by leakage of ethanol and CO2 to the environment.
Figure 18.6. Scheme denoting the important metabolic reactions during anaerobic carbohydrate catabolism. Anoxia prevents pyruvate from entering the TCA cycle because O2 is unavailable as a terminal electron acceptor. Carbon is diverted to fermentative end-products, allowing oxidation of NADH and sustained catabolism of carbohydrates. Key enzymes are: 1. ATP-dependent phosphofructokinase; 2, PPi-dependent phosphofructokinase (PFK); 3, lactate dehydrogenase; 4, pyruvate decarboxylase (PDC); 5, alcohol dehydrogenase (ADH); 6, glutamate-pyruvate transaminase; 7, pyruvate dehydrogenase. The enzyme that catalyses oxidation of NADH as pyruvate is converted to alanine has not been identified. Note that some reactions are reversible (two-way arrows).
Figure 18.7. Curves showing pH optima of enzymes at the branch point for carbon flow to aerobic and anaerobic pathways. These in vitro determinations from extracts of rice coleoptiles indicate how cytoplasmic pH controls carbon flow. In aerobic conditions, pyruvate dehydrogenase (PDH) catalyses entry of pyruvate to the TCA cycle when pH is above 7. Anoxia results in a decline in cytoplasmic pH to below 7, causing PDH activity to give way to pyruvate decarboxylase (PDC) and fermentation to commence (i.e. PDC becomes engaged at pH below 7, whereas PDH ceases to function). In addition, PDC extracted from coleoptiles of rice seedlings previously exposed to anoxia is in a more active state, enhancing pyruvate consumption for ethanol production. Based on Morrell et al. (1989).
Carbon flow from pyruvate to ethanol (with CO2 also produced) occurs via the fermentative enzymes pyruvate decarboxylase (PDC) and alcohol dehydrogenase (ADH) (Figure 18.6). This flow is probably regulated by the activity of PDC which catalyses the first step of ethanolic fermentation. In wheat roots, for example, the PDC in vitro activity approximates the measured in vivo rate of ethanol production. Increases in the amounts of PDC and ADH proteins have been observed in a range of plant genotypes and tissues in response to O2 deprivation. Indeed, these enzymes form part of a suite of ‘anaerobic proteins’, enzymes synthesised during anoxia. In addition to increased protein abundance, post-translational regulation of PDC activity is also exerted by changes in cytoplasmic pH, which decreases from around 7.5 in aerobic cells to around 6.8–7.2 in anoxic cells. Below pH 7.2, the activity of PDC reaches its optimum. For example, PDC extracted from anoxic rice coleoptiles becomes very active as pH drops below 7 according to the broad pH response curve in Figure 18.7. Following a return to aerobic conditions, cytoplasmic pH increases back to its normal level, the activity of PDC decreases, and carbon then flows again via pyruvate dehydrogenase (PDH) to the TCA cycle, rather than via PDC for fermentation to ethanol.
During anoxia, normal protein synthesis is replaced by the selective transcription and translation of a set of proteins called ‘anaerobic proteins’. In maize roots, there are 20–22 of these proteins which include fermentative enzymes (e.g. PDC and ADH), enzymes involved in anaerobic carbohydrate catabolism (e.g. sucrose synthase and enzymes responsible for the reversible breakdown of sucrose) and several glycolytic enzymes (e.g. aldolase). Other ‘anaerobic proteins’ of maize include superoxide dismutase (SOD), responsible for scavenging O2-free radicals. ‘Anaerobic proteins’ are also formed in rice embryos, with a suit of proteins similar to those described for maize but also others - one very interesting additional ‘anaerobic protein’ in rice is the tonoplast H+-pyrophosphatase (Carystinos et al. 1995). By maintaining an ‘energised’ tonoplast capable of ion and solute transport, this enzyme might help stabilise cytoplasmic pH. Use of pyrophosphate (PPi) as an energy source reduces the dependence of tonoplast ion transport on ATP regeneration to drive the H+-ATPase. Studies using knockout mutants in rice have further demonstrated the importance of the H+-pyrophosphatase for anoxia tolerance (see Case Study 2).
Root tissues can acclimate to low O2 with improved anoxia tolerance, if exposed to hypoxia (low, but not zero O2) prior to the onset of anoxia. As examples, roots of maize and wheat survive anoxia more than three times longer if exposed first to hypoxia rather than abrupt transfer from aerated solution into anoxia (Table 18.4). The elimination of ADH activity reduced the survival of maize Adh-
Mutants to almost zero following anoxic shock but allowed recovery following hypoxic pre-treatment (Table 18.4)
Metabolic acclimation set in train by hypoxia included changes in gene expression and therefore the protein complement (‘proteome’) in cells. Hypoxic pretreatment raised activities of the fermentative enzymes PDC and ADH, and resulted in a faster rate of ethanolic fermentation during the subsequent anoxia. How plants sense and initiate signal cascades to invoke these metabolic acclimations is a current topic of debate and could involve sensing of changes in cellular energy charge, cytosolic pH, and/or possibly oxygen (see Case Study 18.3) or other possibilities.
The changes in fermentative enzymes, together with observations that exogenous sugars prolong tissue survival during anoxia, point to carbohydrate catabolism as an important factor in tolerance to anoxia. Even with fermentation operational, however, the anoxic root cells still face an ‘energy crisis’, as the ATP generated via fermentation is often insufficient even for cell maintenance in some species. Compared with respiration, fermentation produces 85–95% less ATP per hexose unit consumed. So, although such anaerobic energy generation is vital, a rapid rate of fermentation alone does not endow anoxia tolerance. Pea root tips, for example, ferment 45% faster than maize root tips, but survive less than half as long in anoxia. Greenway and Gibbs (2003) have highlighted that in addition to fermentation, other more subtle aspects of energy consumption must also be involved in anoxia tolerance, such as a reduction of energy requirements for cell maintenance and the redirection of energy flow to essential cellular processes, including maintenance of membrane integrity, regulation of cytoplasmic pH, and synthesis of appropriate ‘anaerobic proteins’.
The key to anoxia tolerance therefore lies in integration of energy production via anaerobic carbohydrate catabolism and energy consumption in reactions essential for survival. Accumulating evidence suggests two modes of tolerance based on slow and rapid rates of fermentation (Greenway and Gibbs, 2003). As one example of the ‘slow fermentation mode’, lettuce seeds appear to survive anoxia by slowing carbohydrate catabolism in anoxia to less than 35% of the rate in air. After 14 d without O2, lettuce seeds germinate normally (Raymond and Pradet, 1980). Other plant tissues which survive but do not grow in anoxia, produce an initial burst of fermentative activity over 6–24 h before settling to slower fermentation rates. This two-phase pattern presumably provides the higher ATP required as cells acclimate to anoxia, but then the lower rates of fermentation would conserve carbohydrates for long-term survival. To be of adaptive value, this conservation of substrates through slower catabolism must be compatible with the smaller ATP yield available for cell maintenance. Calculations show that, for example, non-growing beetroot tissue in anoxia used 10- to 25-fold less ATP for cell maintenance than aerobic tissues (Zhang and Greenway, 1994).
The coleoptile of rice provides an example of the ‘fast fermentation mode’, this organ grows in anoxia (a second example is the stem of Potamogeton spp.). Fast fermentation is sustained by accelerated glycolysis, a phenomenon known as the ‘Pasteur Effect’. However, even in rice, glycolytic rate is only about twice as fast in anoxia as in air (Table 18.5).
The glycolytic enzyme ATP-dependent phosphofructokinase (PFK), might in addition to PDC, contribute to control of glycolysis (Figure 18.6) and thus fermentation in the coleoptile of rice. Starch breakdown and sugar transport from the endosperm to coleoptile of rice seedlings in anoxia fuels the ethanolic fermentation. For plants without such starch reserves, however, low carbohydrate levels would limit the rate of anaerobic carbohydrate catabolism in tissues that experience anoxia.
LACJ (Rens) Voesenek, Utrecht University, The Netherlands
Plant life relies on light-energy driven fixation of CO2 into carbohydrates through photosynthesis. These carbohydrates are subsequently used to construct various plant structures and fuel energy production through respiration in non-photosynthetic tissues and even in photosynthetic cells during dark periods. Respiration requires a sufficient supply of O2. Cells in various organs of submerged terrestrial plants typically suffer from severe shortage of O2 and thus energy deficits. The extremely slow diffusion rate of O2 through water to plant organs and cells means that O2 supply becomes limiting for respiration. Low O2 conditions during flooding most frequently occur in root tissues as these are surrounded by a soil environment characterized by very low O2 levels and because of the absence of photosynthetic capacity to generate O2. Upon low O2 conditions metabolism shifts from efficient mitochondrial ATP production (O2 dependent) to inefficient anaerobic substrate-level production of ATP (glycolysis linked to fermentation), as long as sugar substrates are available. Further adjustments involve restrictions in ATP consumption and translational activities so that general protein turnover slows to save energy.
Reliable sensing of O2 levels would allow rapid acclimation to declining O2 in flooded plants. It was shown recently that O2 sensing is achieved by a mechanism in which the N-end rule pathway of protein degradation serves to regulate the low O2 response in plants (Gibbs et al. 2011; Licausi et al. 2011). Up to now the unraveling of the O2 sensing machinery was one of the biggest challenges in flooding research. The identification of such a mechanism sheds light on the earliest step in the signaling pathway leading to low O2 acclimation (Figure 1).
Many genes that are typically associated with low O2 conditions in plants are regulated by transcription factors belonging to the so-called group VII Ethylene Response Factors (ERFs). Typical for these ERFs in Arabidopsis (5 members) is that they possess a specific N-terminal motif (N-degron). Due to this motif these proteins are post-translationally modified in an O2-dependent manner via the N-end rule pathway of protein degradation. The N-terminal of Arabidopsis group VII ERFs is composed of a methionine followed by a cysteine as the second residue. The constitutive activity of methionine amino peptidase cleaves these ERF proteins between the methionine and the cysteine, yielding an N-terminal exposed cysteine. In this conformation the cysteine can be oxidized in an O2-dependent manner. Under normoxic conditions, cysteine is oxidized and an arginine residue is added to the cysteine, catalyzed by an arginyl tRNA transferase (ATE). In this form the ERF protein can be recognized by the E3 ligase PROTEOLYSIS 6 (PRT6), leading to ubiquitination and 26S proteosome-mediated degradation. However, when O2 is limited (hypoxia or anoxia) as in many organs of flooded plants, degradation of ERFs is inhibited as a consequence of a lack of cysteine oxidation. Under these conditions stable ERFs can function as transcription factors and drive transcription of genes needed in plants to survive in low-O2 environments. As soon as the plant is re-oxygenated (e.g. upon withdrawal of flood water) the ERFs are again destabilized and the transcription of hypoxia-induced genes is halted.
At least one Arabidopsis ERF, RAP2.12, is sequestered at the plasma membrane, mediated by an interaction with the membrane-bound Acyl CoA binding proteins 1 and 2 (ACBP1/2). This sequestration of RAP2.12 is functional to prevent degradation by the N-end rule pathway under normoxic conditions. Via docking to ACBP1/2, high levels of RAP2.12 can be maintained even under normoxic conditions, without the risk of being degraded. Upon hypoxia RAP2.12 translocates rapidly to the nucleus to switch on acclimative pathways for low O2 conditions (Figure 1).
Figure 1. A model describing O2-sensing in plants. Under normoxic conditions the ERF protein RAP2.12 has a protected plasma membrane localization due to its interaction with Acyl CoA binding proteins (ACBPs). Upon hypoxia RAP2.12 and ACBP dissociate and RAP2.12 moves to the nucleus where it induces transcription of adaptive hypoxia-response genes. Upon re-oxygenation RAP2.12 is rapidly degraded via the N-end rule pathway to down regulate the hypoxia response (Licausi et al. 2011).
Bailey-Serres J, Fukao T, Gibbs DJ et al. (2012) Making sense of low oxygen sensing. Trends Plant Sci 17: 129-138
Gibbs DJ, Lee SC, Isa NM et al. (2011) Homeostatic response to hypoxia is regulated by the N-end rule pathway in plants. Nature 479: 415-418
Licausi F, Kosmacz M, Weits DA et al. (2011) Oxygen sensing in plants is mediated by an N-end rule pathway for protein destabilization. Nature 479: 419-422
Hypoxic conditions can favour generation of reactive oxygen species (ROS). The major sources of ROS generation are ETC in mitochondria and oxidase activities (Blokhina et al., 2003). Various ROS species may be produced; among the major ones are superoxide radical (O2•-), hydroxyl radical (OH•), and hydrogen peroxide (H2O2). Although H2O2 is less reactive than the two other ROS, in the presence of reduced transition metals such as Fe2+ (abundant in waterlogged soils), the formation of OH• can occur in the Fenton reaction. These ROS can damage plant cells by causing lipid peroxidation in membranes, DNA damage, protein denaturation, carbohydrate oxidation, pigment breakdown and an impairment of enzymatic activity (Noctor and Foyer, 1998).
The extent of the ROS-induced damage to cells depends on duration and severity of stress. Short-term O2 deprivation results in a limited accumulation of ROS and lipid peroxidation. In the short-term, the rate of ROS formation and the degree of lipid peroxidation can be regulated by constitutive endogenous antioxidants (Blokhina et al. 2003). In addition, hypoxia induces increased activities of antioxidant systems. Prolonged deprivation of O2, however, can diminish or even abolish synthesis, transport and turnover of antioxidants. As a consequence of the depleted antioxidants and associated enzymes, cells are unable to cope with the ROS and lipid peroxidation can become severe, particularly during re-oxygenation (see also section 18.6). In addition to causing non-specific increases in membrane permeability resulting from lipid peroxidation, both H2O2 and OH• have also been shown to directly control activity of Ca2+- and K+-permeable plasma membrane ion channels (Demidchik et al. 2007, 2010). Perturbations in intracellular ionic homeostasis may initiate programmed cell death (Demidchik et al. 2010).
Plant nutrient acquisition is dramatically reduced in sensitive species when in waterlogged soil. Upon waterlogging, root growth can be immediately arrested whereas shoots can continue to grow. The resulting increased shoot:root ratio causes an imbalance between shoot nutrient demands and supply by roots. Nutrient ion uptake be roots is also greatly reduced on a per root weight basis (Elzenga and van Veen 2010; Colmer and Greenway, 2011), primarily as a result of reduced O2 availability inhibiting respiration. Ion uptake by roots consumes energy. The plasma membrane proton pump (H+-ATPase) requires ATP and the proton motive force generated is used to drive symporter-mediated ion uptake. Indeed, all anions (e.g. NO3-) enter root cells via H+-anion symporters. Furthermore, the H+-ATPase maintains the negative membrane potential, essential to creating electrochemical gradients allowing channel-mediated uptake of cations (e.g. K+ uptake). Absence of O2 inhibited respiration and lowered H+-ATPase pumping, causing a substantial membrane depolarization, making such cation uptake via channels thermodynamically impossible (Pang and Shabala 2010). Not only is K+ uptake significantly reduced, but roots can also loose substantial amounts of K+ through depolarization-activated channels. It is not surprising, therefore, that waterlogged plants often exhibit acute K+ deficiency. The organic acids present in waterlogged soils, from anaerobic microbial metabolism, can also lead to membrane depolarisation of root cells and reduced ion uptake.
The diminished capacity for ion transport, together with initial ‘dilution’ of shoot nutrient concentrations by continued shoot growth relative to roots, explains a range of nutrient deficiencies observed in leaves of intolerant plants under waterlogged conditions. Waterlogging tolerant species with adequate O2 supply to roots via large volumes of aerenchyma, can sustain root respiration and therefore plasmamembrane H+-ATPase functioning for nutrient uptake, as well as having adequate O2 and energy for deeper root penetration. Efficient internal aeration of roots, together with a barrier to ROL in basal zones, also enables an aerobic rhizosphere at the root tips and regions of dense laterals, altering the rhizosphere (e.g. diminished soil toxins), presumably also with benefits for nutrient uptake by the roots.
