Ca2+ can be released from cell compartments to the cytosol during stress conditions. We discuss here the causes of Ca2+ release under conditions of ATP concentration decline that result in the suppression of ATPases and activation of calcium ion channels. The main signaling and metabolic consequences of Ca2+ release are considered for stressed plant cells. The signaling function includes generation and spreading of calcium waves, while the metabolic function results in the activation of particular enzymes and genes. Ca2+ is involved in the activation of glutamate decarboxylase, initiating the γ-aminobutyric acid shunt and triggering the formation of alanine, processes which play a role, in particular, in pH regulation. Ca2+ activates the transcription of several genes, e.g. of plant hemoglobin (phytoglobin, Pgb) which scavenges nitric oxide and regulates redox and energy balance through the Pgb–nitric oxide cycle. This cycle involves NADH and NADPH oxidation from the cytosolic side of mitochondria, in which Ca2+- and low pH-activated external NADH and NADPH dehydrogenases participate. Ca2+ can also activate the genes of alcohol dehydrogenase and pyruvate decarboxylase stimulating hypoxic fermentation. It is concluded that calcium is a primary factor that causes the metabolic shift under conditions of oxygen deficiency.
Introduction: calcium in the plant cell
Calcium is a widespread element in plant cells, but most of it is in inactive pools within the vacuole, cell wall, endoplasmic reticulum, mitochondria, and possibly in Golgi vesicles. The stored calcium is usually present in submillimolar concentrations, e.g. the mean vacuolar Ca2+ level is in the range of 0.2 mM , while the concentration of free calcium (‘resting calcium’) in the cytoplasm is commonly in the range of 100–200 nM . It can increase to submicromolar and micromolar concentrations in stress conditions, triggering signal transduction and activation of several enzymes. Calcium, binding to proteins in the cytoplasm and membrane surfaces, can reach submillimolar concentrations , an estimation based on the concentration of cytoplasmic calcium-binding protein sites reaching 1 mM in human neutrophils . In the course of biological evolution, living cells acquired efficient mechanisms to pump Ca2+ as its concentrations increased in the Proterozoic ocean in parallel with the increase of photosynthetic oxygen .
Ca2+-dependent ATPases actively pump Ca2+ out of the cytosol and into the vacuole, vesicles, and cell wall, a process which is suppressed upon a decrease in energy charge . There are two distinct families of Ca2+-ATPases: auto-inhibited Ca2+-ATPases (ACA), present on the plasma membrane (PM) and tonoplast; and endoplasmic reticulum Ca2+-ATPases (ECA), found also in membranes of the tonoplast, Golgi, and other vesicles . ACAs are active at micromolar concentrations of free Ca2+, transport one calcium ion per ATP, and can also use ITP and GTP. They are activated by calmodulin and auto-inhibited at high Ca2+ concentrations. Their inhibition results in the activation of a salicylic acid-dependent programmed cell death pathway . ECAs function at submicromolar [Ca2+], exclusively use ATP to transport two ions per ATP consumed, and can also transport Mn2+ and Zn2+. In addition to these pumps, Ca2+ channels, regulated by membrane potential, are also found in membranes. They are classified into three groups: hyperpolarization-activated (HACC), depolarization-activated (DACC), and voltage-independent (VICC) [7,9].
Calcium is not freely mobile: when injected, it remains in the proximity of the injection site . As a result, calcium waves generated during signaling relate not to its direct diffusion, but to more complex mechanisms that may involve, in particular, the elevation of inositol 1,4,5-trisphosphate (IP3), i.e. calcium waves do not implicate the forward movement of calcium ions, but they involve the forward movement of Ca2+ release from vesicle stores and uptake back into cell compartments . This initiates downstream processes of Ca2+-dependent gene expression, opening of K+ and chloride channels and triggering release of second messengers such as ADP-ribose which further mobilize calcium from intracellular stores .
Calcium signaling is the main event in all stress responses . Its pattern differs in hypoxia-tolerant and -intolerant plants. Behera et al.  detected marked differences between Arabidopsis and rice in the spatiotemporal parameters of the observed Ca2+ signatures in Arabidopsis thaliana and Oryza sativa plants. The hypoxia-tolerant plant rice responded with a lower maximal signal amplitude, but greatly increased signal duration when compared with the hypoxia-sensitive Arabidopsis. This indicates that adaptation to hypoxia is mediated by the amplitude and duration of the calcium wave. Plants experience low oxygen stress frequently throughout development, e.g. due to restricted diffusion through compact tissues or during flooding. The conditions of oxygen deficiency could be either hypoxic or anoxic, which generally reflects the atmosphere to which the tissue has been exposed, while the actual conditions in the tissue are always difficult to attest . In this review, we will discuss the role of Ca2+ release to the cytosol as a primary factor that causes metabolic shift under the conditions of oxygen deficiency.
Ca2+ release into the cytosol under hypoxia initiates calcium signaling events
Ca2+ release into the cytosol can be viewed as a consequence of the decrease in ATP resulting from the suppression of operation of the mitochondrial electron transport chain due to hypoxia. There may be secondary events leading to the elevation of Ca2+ concentration under stress, like the changes in redox potential and modulation of phytohormone balance, but the primary cause remains the decrease in ATP. Lower ATP production in stressed cells is associated with depolarization of cell membranes , which activates depolarization-activated Ca2+ channels (DACC) . In Arabidopsis, two mechanosensitive Ca2+ channels respond to hypo-osmotic stress caused by flooding . Hypoxically induced ethylene formation [18,19] can also activate weak voltage-dependent Ca2+ channels . Modulations of total glutathione content can control Ca2+ release via the changes in glutathione redox potential .
Elevation of the Ca2+ concentration leads to activation of calmodulin and Rho GTPases  as well as of numerous enzymes that are regulated by Ca2+. Ca2+ can directly activate PM NADPH oxidase (Respiratory Burst Oxidase Homolog, RBOH) , which triggers further elevation of Ca2+ and initiates the signaling cascade associated with reactive oxygen species (ROS) generation (Figure 1). The RBOH-dependent calcium signaling cascade involves hydrogen peroxide (H2O2) production being controlled by [Ca2+]cyt in a positive feedback manner . H2O2 production by hypoxic cells at PM  is attributed to the increased redox level causing one-electron reduction of oxygen at the site of NADPH oxidase, even at low oxygen concentrations, followed by dismutation of superoxide anion. ROS signaling events connected with H2O2 generation are directly linked to the stress-induced Ca2+ release. The activation of NADPH oxidase is achieved via Ca2+-dependent protein kinases that specifically activate this enzyme in response to stress [26,27]. ROS production by RBOH stimulates PM Ca2+-permeable channels, which, in turn, further contribute to the elevation of cytosolic free Ca2+ triggering signal transduction events [9,28]. ROS production, in particular, was shown to be a major event resulting in the induction of alcohol dehydrogenase .
Cross-talk between ROS generated by PM NADPH oxidase (RBOH) and Ca2+ influx in the cytosol.
Direct support for the role of NADPH oxidase in the modulation of cytosolic Ca2+ signatures in response to hypoxia comes from the recent experiments with RBOHD mutants . Superoxide anion, generated by RBOH, is converted by superoxide dismutase (SOD) to H2O2, which is transported to the cytosol via aquaporin  and activates HACC from both sides of the membrane, followed by Ca2+ influx that depolarizes the PM, activating DACC. Owing to a positive feedback loop between NADPH oxidation by RBOH and cytosolic Ca2+ concentration, the majority of Ca2+ accumulated in the cytosol comes from the extracellular space. This PM calcium channel loop constitutes the major source of the hypoxia-induced cytosolic Ca2+ changes, complemented by the release of Ca2+ from mitochondria and other cell compartments. Another calcium signaling cascade involves Ca2+/calmodulin (CaM) complexes that have been shown to control anaerobic protein degradation and solute release in different cell compartments .
Mitochondria play an important role in Ca2+ release to the cytosol. Schwarzländer et al.  observed stress-induced pulsing (via transient partial depolarization of the inner membrane) of membrane potential in Arabidopsis mitochondria resulting from Ca2+ flux and suggested that it represents a transient uncoupling mechanism to prevent ROS production. Wagner et al.  have demonstrated that At-MICU, an EF-hand protein of Arabidopsis with homology to the mitochondrial Ca2+ uniporter in mammals, mediates mitochondrial Ca2+ uptake and choreographs mitochondrial Ca2+ dynamics. The concentration of Ca2+, similar to Mg2+, can be under control of the equilibrium governed by adenylate kinase [35–37], since Ca2+ is chelated by adenine nucleotides to nearly the same extent as Mg2+ . The total concentration of free calcium is always a few orders of magnitude lower than that of magnesium. As a consequence, the ratio of [Ca2+]/[Catotal] is roughly equal to that of [Mg2+]/[Mgtotal] , which means that a tenfold increase of [Mg2+] will lead to a tenfold increase of [Ca2+] (for details, see ref. ). The release of Mg2+ upon energy charge lowering can cause correspondent changes in internal Ca2+ . Thus, the equilibrium of adenylate kinase can control not only Mg2+ but also Ca2+ homeostasis (Figure 2), with consequences for signaling and regulation of metabolic events.
Adenylate kinase (AK) reaction and Ca2+ balance.
The increase of Ca2+ in the intermembrane space of mitochondria under intensive respiration, which follows the increase of Mg2+ , will subsequently lead to the activation of Ca2+-dependent enzymes, including NADPH and NADH dehydrogenases in the inner mitochondrial membrane facing the intermebrane space  and NAD kinase in this space . It also leads to the activation of glutamate decarboxylase, phytoglobin (Pgb) synthesis, external NADH and NADPH oxidation, internal NADPH oxidation in mitochondria, cysteine proteases (calpain), Ca/CaM-dependent NAD kinase, Ca/phospholipid-dependent protein kinases, and several other processes. It has been also shown that Mg2+ is an allosteric activator of Ca2+ binding to calmodulin , with the latter activating target enzymes in response to submicromolar increases in Ca2+.
A significant contribution to Ca2+ signaling is linked to the tonoplast because the central vacuole is the largest store of calcium in a mature plant cell and Ca2+ release from it regulates many physiological processes . The transient receptor potential Ca2+-permeable channels (TPC channels) located on the tonoplast may play a crucial role in low oxygen sensing, such as the voltage-dependent slow vacuolar (TPC1/SV) channel in Arabidopsis identified as the product of the TPC1 gene. The homological TCP1 channels in other species can be also localized in the PM. Other vacuolar Ca2+ channels, which are activated by IP3 or cADP ribose, are less well characterized . Bioinformatic studies  have led to the prediction of potential candidate genes in Arabidopsis for oxygen sensing and calcium signaling based on homology with known oxygen-sensing domains in mammalian systems, which include TPC channels. It becomes evident that the co-ordinated Ca2+ transport through the PM, tonoplast, and mitochondrial membranes after the onset of hypoxia resulting in Ca2+ elevation in the cytosol initiates signaling events playing a key role in the adaptive response of plants to the conditions of oxygen deficiency. Ca2+ elevation can significantly influence metabolic fluxes and substrate oxidation which makes Ca2+ a major signaling ion in the hypoxic response .
Using the transgenic AEQUORIN system to follow Ca2+ changes, the cotyledons and leaves of Arabidopsis seedlings demonstrated a biphasic and tissue-specific change in cytosolic Ca2+ concentration in response to anoxia, a fast peak occurring in the first minutes, followed by a second response that lasted 1.5–4 h . Ca2+ channel blockers inhibited the first response and promoted the second response. It is possible that the first response was caused by the efflux of free Ca2+ from the apoplast and organelles, while the second response was due to the release of bound, including extracellular, calcium upon the decrease in the energy charge. Blocking calcium channels could prevent the outflow of calcium released at this stage to organelles and the apoplast. The addition of the PM Ca2+-channel antagonists, verapamil and La3+, reduced anoxia-induced Ca2+ accumulation in rice , and blockage of Ca2+ uptake by Gd3+ reduced anoxic seed survival in Arabidopsis, suggesting an extracellular source for the increase in [Ca2+]cyt in the second response . Comparing hypoxia-induced changes in [Ca2+]cyt in rice and wheat roots, Yemelyanov et al.  suggested that both external and internal Ca2+ stores are important for stress-induced [Ca2+]cyt elevation in rice, whereas the hypoxia-intolerant wheat does not necessarily require an external source for the rise in Ca2+ concentration.
The earlier work on the effects of hypoxia on Ca2+ release [49–52] has been extended by other research groups. Virolainen et al.  demonstrated that the onset of anoxia caused the release of previously accumulated Ca2+ from mitochondria. Wagner et al.  also consider the significance of mitochondrial Ca2+ control and release as a key event in anoxic plant signaling and metabolism. It has been suggested that the mitochondrial release of Ca2+ can trigger the release of Ca2+ from other organelles and from the apoplast [49,55]; however, further studies are needed to test whether the release of Ca2+ from mitochondria precedes its release from other compartments.
Changes in [Ca2+]cyt appear to be an important component of the low oxygen-sensing mechanism. Plants are known to possess two classes of GTP-binding proteins (G-proteins), which regulate numerous signal transduction pathways . They also possess RHO-like GTPases (ROP) that act as molecular switches between active GTP- and inactive GDP-bound forms . It is suggested that an intracellular release of Ca2+ is needed before ROP–GTP can initiate signaling, owing to the dependence of both PM and mitochondrial NAD(P)H oxidases on Ca2+ for activity (reviewed in ref. ).
The calmodulin-like protein CML36 in Arabidopsis functions as a Ca2+ sensor by interacting with the PM Ca2+-ATPase isoform ACA8 and stimulating its activity . This interaction may explain previously reported effects of anoxia on Ca2+-ATPase. The form of Ca2+-ATPase having CaM-binding properties and phosphorylated during the reaction is induced by Ca2+ during early anoxia in maize, playing an important role in calcium transport in O2- deprived maize cells [60,61]. Lokdarshi et al.  have identified, in Arabidopsis, another hypoxia-responsive Ca2+ sensor, CML38, a calmodulin-like protein involved in the regulation of gene silencing (rgsCaMs). CML38 transcripts are up-regulated more than 300-fold in roots within 6 h of hypoxic stress. The sensor CML38 is localized in hypoxia-induced cytosolic granule structures accumulating and storing messenger RNA under stress. Joyce et al.  revealed two ATP-dependent Ca2+ transport activities in tomato differing in sensitivity to nitrate and affinity for Ca2+. A low-affinity Ca2+ uptake system (Km >200 µM) was inhibited by nitrate and represented a tonoplast H+/Ca2+ antiport. A high-affinity Ca2+ uptake system (Km = 6 µM) associated with the endoplasmic reticulum was not inhibited by nitrate. Wang et al.  studied tissue-specific Ca2+ changes using several ACA (Ca2+-ATPase) and CAX (Ca2+/proton exchanger) knockout Arabidopsis mutants. Hypoxia decreased the expression of ACA8, CAX4, and CAX11 resulting in 11-fold tissue-dependent increase in Ca2+ accumulation in root tissues. The increase was tissue-specific being much higher in the cytosol of stelar cells in the mature zone. In addition, a significantly increased Ca2+ concentration was found in the cytosol of stelar cells in the mature zone after hypoxic treatment. The authors suggested that the transporter CAX11 plays a key role in maintaining cytosolic Ca2+ homeostasis and signaling under the hypoxic conditions.
Glutamate decarboxylase and the γ-aminobutyric acid shunt
An important element of anaerobic signal transduction is CaM-dependent glutamic acid decarboxylase (GAD; EC 220.127.116.11), which converts glutamic acid to γ-aminobutyric acid (GABA) . Plant GAD is characterized by the presence of a CaM-binding domain at its C-terminus. Activation of GAD results in the formation of GABA, which is significantly accumulated under hypoxic conditions. Within a few hours after the onset of hypoxia, the GABA concentration in plants increases three to four orders of magnitude , reaching millimolar levels. These concentrations of GABA are efficient in the regulation of the PM Ca2+ and K+ transporters under hypoxic stress conditions . This regulation is achieved via direct sensitization/desensitization of the PM Ca2+ and K+ transporters  or through the negative regulation by GABA of the anion flux through plant aluminum-activated malate transporter (ALMT) . GABA significantly increased the ability of roots to actively extrude Ca2+ especially when they were exposed to H2O2 indicating the link between ROS production, GABA, and Ca2+ signaling . The peak Ca2+ flux values were three times higher in GABA-treated roots, and the total amount of Ca2+ extruded over the first 15 min after low oxygen stress onset differed almost 10-fold. Thus, the activation of GAD by Ca2+ results, via GABA formation, in the stimulation of Ca2+ extrusion leading to the establishment of calcium homeostasis (Figure 3).
The cross-talk between Ca2+ and GABA in hypoxic cells.
The GABA-exerted control of ROS production via the modulation of anion flux leading to the altered cycling of tricarboxylic acid cycle intermediates and the regulation of RBOH is an important consequence of Ca2+-dependent GABA formation [69,70]. This mechanism results in GABA-dependent restoration of the basal H2O2 and Ca2+ levels in the cytosol and their maintenance at the optimal levels required for signaling, preventing the formation of HO· and transition to PCD. The mechanism of GABA regulation of Ca2+ transport probably does not include binding to the glutamate receptors, as in animal cells, but rather the established GABA-dependent down-regulation of the expression of 14-3-3 gene family members , which affects the activity of the PM H+-ATPase, with downstream implications to Ca2+ transport (see ref.  for details).
Glutamate decarboxylase consumes protons (as a co-substrate) and releases CO2 in the reaction [72,73]. Thus, Ca2+-dependent activation of GAD under anoxia becomes an important tool for counterbalancing acidification of the cytosol which can reach significant values during lactate fermentation [74,75]. Acidification promotes association of enzyme dimers into a stable functional hexameric enzyme . The transcript level of GAD can also increase during hypoxia; however, its activation at the enzyme level is always more pronounced .
Although GABA synthesis may also occur via polyamine (putrescine and spermidine) degradation [73,77] and possibly by a non-enzymatic reaction from proline under oxidative stress , its main source is the GAD reaction. GABA is metabolized mainly in mitochondria. GABA transaminases are also present in the cytosol and chloroplasts with lower activity . Transamination of GABA occurs preferably with glyoxylate and pyruvate, but not with 2-oxoglutarate . As a result, GABA can be an efficient amino donor for alanine formation, which is also accumulated under hypoxia, and it can also stimulate glycine/glyoxylate interconversion that may take place under stress .
GABA accumulation causes a shift from lactate to alanine formation under hypoxia, preventing further acidification of the cell [82,83]. Thus, Ca-dependent GAD activation leads to pH stabilization and to a shift toward alanine and succinate formation. Bao et al.  suggested that the GABA shunt, by controlling the balance of succinate and hydroxybutyrate, prevents ROS accumulation, which may be important in stress conditions when the redox level is elevated.
Ca2+ and up-regulation of the Pgb–nitric oxide cycle
Under hypoxic conditions, Ca2+, besides its known participation in the activation of the GABA shunt via activation of glutamate decarboxylase , can participate in the regulation of vacuolar H+-PPase , in activation of external mitochondrial NAD(P)H dehydrogenases, which are also more active at lower pH values (using NADH and NADPH) , and in lower pH-stimulated activation of NAD kinase . Another important protein, which is significantly up-regulated under hypoxia, is class 1 Pgb1 that can oxygenate substrates such as nitric oxide (NO) at nanomolar concentrations of oxygen, which is two orders of magnitude lower than the oxygen affinity of cytochrome c oxidase and thus fully anaerobic for mitochondria [88,89]. The term Pgb has been suggested for hemoglobin-like proteins of plants  and is now generally accepted by the scientific community.
The expression of Pgb under hypoxic conditions is not directly affected by oxygen availability, but linked to energy charge, i.e. availability of ATP . There is also the possibility that this hypoxic induction, at least in barley aleurone layers, is induced by Ca2+ in a calmodulin-independent process . Expression of the class 1 Pgb, depending on the availability of Ca2+, was also shown for rice [93,94]. The importance of Pgb expression is related not only to NO scavenging, which is crucial by itself for prevention of nitrosylation of proteins and of other effects connected with NO accumulation , but also for initiation of the Pgb–nitric oxide (Pgb–NO) cycle, which involves three components: nitrate reductase-catalyzed nitrite synthesis, nitrite-dependent NO formation occurring in mitochondria and other cell compartments, and oxygenation of NO to nitrate by oxygenated Pgb and its associated reductase . This process regulates redox balance of the hypoxic cell by oxidation of NADH and contributes to a limited rate of ATP synthesis due to proton pumping at the sites of complexes III and IV . The cycle helps to regulate the cellular NO level and to maintain mitochondrial electron transport by using nitrite as an alternative to oxygen electron acceptor. Thus, cytosolic NADH can be recycled and some ATP produced as long as there is stored nitrate.
In the reaction of NO oxygenation by Pgb (Figure 4), one O2 molecule is used for each NO molecule cycled. This means that the turnover of Pgb-NO cycle is limited by the amount of oxygen available. The extremely high affinity of Pgb, which is in the nanomolar range differing from cytochrome c oxidase by two orders of magnitude , ensures that almost every last oxygen molecule is used; however, the production of NO when oxygen concentration drops remains higher than the availability of oxygen . This results in NO release from the hypoxic plants, causing the plant to lose stored nitrogen. Because of the limited availability of oxygen, the use of nitrite as an electron acceptor can be viewed as a stop gap measure but not as a long-term solution to hypoxia .
Operation of the Pgb–NO cycle between mitochondria and cytosol.
Other genes that are affected by Ca2+ increase are those encoding alcohol dehydrogenase 1 and sucrose synthase 1 [50,51,100,101]. Therefore, both the glycolytic pathway (initiated with sucrose synthase forming UDP-glucose and fructose and completed with alcohol dehydrogenase) and the Pgb–NO cycle can be up-regulated under hypoxic stress via Ca2+ elevation (Figure 4). Other enzymes up-regulated by NO include Ca2+-dependent protein kinases involved in different signaling events including adventitious root formation [102,103].
Nitrate reductase is up-regulated under hypoxia. Its activity is regulated (inhibited) by phosphorylation, in a process mediated by Mg2+ and 14-3-3 proteins . Application of calcium significantly increases nitrate reductase activity and promotes uptake . In aerated roots, the cytosolic NR is highly phosphorylated (inactive). It becomes dephosphorylated (activated) under hypoxia and these changes in the NR phosphorylation state are triggered mainly by a decrease in cytosolic pH . Thus, although NR is probably not directly affected by Ca2+ (indirect activation of expression is possible, see, e.g. Chandok and Sopory ), Ca2+ stimulates nitrate transporters located in the plasmalemma. In Arabidopsis roots, the PM nitrate influx transporter NRT1.1 is phosphorylated. Low-affinity calcium transport into the vacuole is inhibited by nitrate, which may be a factor in calcium accumulation in the cytosol . Douglas et al.  demonstrated that nitrate reductase kinase from Spinacea oleracea leaves is a calmodulin-domain protein kinase.
Ca2+ activates the external NADH and NADPH dehydrogenases in mitochondria
There are two or more externally facing rotenone-insensitive dehydrogenases on the inner membrane of plant mitochondria (abbreviated as NDB), encoded by four genes in Arabidopsis . Two detectable Ca2+-dependent activities, one specific for NADH and another for NADPH, can be distinguished by sensitivity to diphenyleneiodonium . The expression of both activities is affected by nitrogen supply but not by light, as in the case of internal rotenone-insensitive dehydrogenases NDA1 and NDC [111,112]. Mitochondrial oxidation of cytosolic NADH and NADPH occurring via these dehydrogenases does not result in proton pumping at the site of electron transport from NAD(P)H to ubiquinone . Low pH activates external dehydrogenases by protonating the NAD(P)H-binding site , while Ca2+ has a specific role in the catalysis . Recent research has expanded our knowledge of plant external mitochondrial NADH and NADPH dehydrogenases , suggesting that Ca2+-dependent external NADPH oxidation is an ancient process having fundamental importance for eukaryotic cellular redox metabolism, while the external Ca2+-dependent NADH oxidation is a result of more recent divergence of NAD(P)H-oxidizing enzymes, mirroring the evolutionary development of the alternative oxidase genes.
The internal NADPH dehydrogenase (NDC) is also activated by Ca2+ suggesting it, too, may be involved under hypoxic conditions. The gene coding for the protein has similarities to the cyanobacterial type II NADH dehydrogenase genes, suggesting that this gene may have evolved from a chloroplast progenitor, becoming nuclear encoded and acquiring a signal peptide for mitochondrial targeting of the protein product . Xu et al.  showed that NDB proteins could also be targeted to peroxisomes where they can oxidize peroxisomal NADH. Geisler et al.  found that binding of Ca2+ by NDB proteins is mediated by an EF-hand motif, and a single amino acid substitution in this motif of NDB1 abolished the Ca2+ binding. Suppression of NDB1in Arabidopsis resulted in reduced plant growth and decreased tissue levels of sugars, citric acid cycle intermediates, and amino acids . Similar changes reflecting a shift to fermentation were reported for plants down-regulating NDA proteins , but no direct association between photosynthesis and NDA proteins was found. Ca2+ activation of NADPH, but not NADH, oxidation by NDB-type dehydrogenases was dependent on pH, being pronounced at a pH close to 6 and having half-maximum stimulation at low micromolar concentrations of Ca2+ . NADPH oxidation responded to changes in Ca2+ concentrations more rapidly than NADH oxidation. Cytosolic acidification together with Ca2+ elevation is more important for externally facing NADPH oxidation than for NADH oxidation . The internally facing NADPH dehydrogenase, NDC, is also dependent on Ca2+ with half-activation at 3 µM .
The link between operation of NAD(P)H dehydrogenases and prevention of fermentation suggests a possible involvement (especially of the external dehydrogenases, NDB) in hypoxic utilization of cytosolic NADH and NADPH. NADH oxidation by external NADH dehydrogenases in maize mitochondria may play an important role in imbibed seeds, which remain hypoxic before radicle protrusion . NADH-dependent O2 consumption was insensitive to rotenone and inhibited by dicumerol indicating the involvement of the externally facing NADH dehydrogenases, but not that of complex I. The alternative oxidase was not involved either. This has similarities to the use of nitrite as a terminal electron acceptor by anoxic rice and barley mitochondria . In this study, mitochondria were insensitive to rotenone, but NADH and NADPH oxidation were inhibited by diphenyleneiodonium, which inhibits external rotenone-insensitive dehydrogenases to a higher degree than NADPH dehydrogenase . The nitrite-dependent NADH and NADPH oxidation were also sensitive to KCN and myxothiazol, indicating the involvement of complexes III and IV .
The role of NDB dehydrogenases in ROS and NO formation under hypoxia may represent an important signaling event in adaptation to hypoxic stress . It is likely that oxidation of the cytosolic NADH and NADPH by externally facing dehydrogenases of mitochondria represents an important mechanism for redox regulation in hypoxic cells. Although a part of NADH is utilized in the terminal reactions of glycolysis, not all pyruvate can stoichiometrically participate in this process since alanine can be formed from pyruvate, via transamination, without oxidizing NADH. Consequently, mitochondrial oxidation of NAD(P)H can stimulate amino acid metabolism  with the operation of the mitochondrial ETC using nitrite as an alternative electron acceptor. Miyashita and Good  have demonstrated the involvement of GABA transamination in the formation of alanine from pyruvate under hypoxia. When pyruvate is converted to alanine, the glycolytic NADH can be oxidized by the NDB-type dehydrogenases, and the coordination of this process with nitrite reduction by the complexes III and IV of mitochondria initiates the Pgb–NO cycle.
Ca2+ can also act as a physiological transducer for the expression of the mitochondrial alternative oxidase 1a (AOX1a) gene . AOX expression under the hypoxic stress controls NO formation and shifts metabolism toward the formation of amino acids . Although AOX induction is triggered by accumulation of citrate, Ca2+ can mediate this process, in particular, via promotion of ethylene synthesis .
Tissue nitrogen status and calcium
Calcium appears to be implicated in several ways in the N status of plants . The calcium sensor, CBL7, was induced under nitrate starvation in young Arabidopsis plants and roots of cbl7 mutants respond poorly to starvation relative to wild type . The response in the mutants was only observed at low nitrate concentrations and pH values below 6.8. Two high-affinity nitrate transporter genes, NRT2.4 and NRT2.5, were down-regulated in the mutant during nitrate starvation. An Arabidopsis calcium antiporter gene, CAX1, was up-regulated at low nitrate concentrations, but the high-affinity nitrate transporter genes, NRT1 and NRT2, did not show any significant changes in expression with added nitrate , although the authors suggested that this may have been due to the presence of NH4+ in the medium. There is an interesting proposal that amino acid-gated Ca2+ channels may be a mechanism by which plants sense their N status . Glutamate-like receptor (GLR) proteins are numerous in Arabidopsis and are gated by a wide range of amino acids to affect Ca2+ transport . This amino acid binding might allosterically regulate the GLRs. GLRs are ubiquitous in plants and have been found in the PM, in the inner and outer chloroplast membranes and in the inner membrane of mitochondria.
Ca2+ and cell death
Ca2+ is involved in the later events of hypoxic stress resulting in necrosis in roots. This occurs via regulation of cysteine proteinases . After 48 h of anoxia, maize root tips develop signs of cell death. Several pre-existing proteases become repressed and Ca2+- and anoxia-induced cysteine protease (AIP) activity is induced, becoming the most abundant protease activity in the root axis and tip. De-tipping of roots before anoxia reduced root damage, and this was associated with a decreased AIP activity .
Another mechanism of Ca-associated necrosis is linked to sucrose synthase (SuSy). Phosphorylation of sucrose synthase (SuSy) in a Ca2+-dependent manner was observed during the first 2 h of anoxia . During prolonged anoxia (48 h), however, SuSy became unphosphorylated. The phosphorylated protein was associated with the microsomal fraction in root tips, which preceded callose induction, indicative of cell death. The membrane association of SuSy can be seen as an important early event in anoxic root tips which represents a potent regulatory mechanism of sugar metabolism in response to oxygen deprivation. The mutants that showed constitutive SuSy phosphorylation lacked the anoxia-induced relocation of SuSy, which resulted in lower callose deposition and greater tolerance to prolonged anoxia. Both the transcriptional control of genes encoding SuSy and its reversible phosphorylation provide a potent regulatory mechanism of PCD in anoxic plants.
Aerenchyma formation in hypoxic roots is triggered by ethylene, an action that initiates a signal transduction pathway involving phosphoinositides and Ca2+ [11,133]. The process can be also ROS-mediated and involves PM NADPH oxidase [134,135]. It is initiated by ethylene activation of PM Ca2+ channels with weak voltage dependence . Differential expression of the class 1 (and possibly class 2) Pgb in certain cells, which is also stimulated by Ca2+ , results in the prevention of PCD in these cells via scavenging of NO and stabilization of redox level, while in other cells where Pgb is down-regulated, Ca2+-stimulated events leading to PCD may take place and aerenchyma is formed supporting viable root cells via transport of oxygen from the shoot .
It was suggested that the influx of Ca2+ into the cytoplasm is a necessary step in the process leading to cell death  and hence to aerenchyma formation under oxygen deficiency. The supporting evidence comes from experiments with thapsigargin, a known inhibitor of Ca2+-ATPases. Thapsigargin treatment resulted in increased aerenchyma formation in normoxic roots and in a faster progress of PCD in oxygen-deficient roots . Oxygen deprivation and Ca2+ ions together cause cytochrome c release from mitochondria initiating programmed cell death .
Virolainen et al.  reported that the onset of anoxia caused rapid dissipation of the inner membrane potential, shrinkage of the mitochondrial matrix, and the release of previously accumulated Ca2+. The release of cytochrome c which is a key event in PCD took place when oxygen deprivation and Ca2+ ions occurred together in anoxic wheat roots. In mammalian systems, cell death in oxygen-deprived neurons is mediated by Ca2+ entry through over-activation of N-methyl-d-aspartate receptors , and silencing of these receptors involves the mitochondrial permeability transition pore through release of mitochondrial Ca2+. A similar mechanism for plants remains to be elucidated.
Elevation of the Ca2+ concentration in the cytosol under hypoxia, when ATP levels are limiting, occurs both as a direct consequence of lower Ca2+-binding capacity and indirectly due to Ca2+ release from mitochondria and other organelles when Ca-ATPases are suppressed and calcium channels activated. This results in the chain of signaling and metabolic events that include the generation and spreading of calcium waves and the activation of particular enzymes and genes. The important metabolic events triggered by Ca2+ release are the induction of Pgb–nitric oxide cycle, the up-regulation of glycolysis, and the activation of GABA shunt. These processes mold the bioenergetics of the hypoxic cell and contribute to redox balance, energy production, and pH regulation. They help plants to survive the conditions when the mitochondrial ATP production becomes limited. Ca2+ release in the cytosol, therefore, can be considered as a primary event that causes metabolic shift under hypoxia.
anoxia-induced cysteine protease
depolarization-activated calcium channel
hyperpolarization-activated calcium channel
plasma membrane NADPH oxidase
reactive oxygen species
- TPC channels
transient receptor potential Ca2+-permeable channels
The authors listed, have made substantial, direct and intellectual contribution to the work, and approved it for publication.
The research leading to these results has received funding from the Natural Sciences and Engineering Research Council of Canada [grant no. RGPIN 355753/2013 (to A.U.I.)].
The authors thank Dr Claudio Stasolla (University of Manitoba, Canada) for his valuable suggestions and comments.
The Authors declare that there are no competing interests associated with the manuscript.