Sugar signalling and antioxidant network connections in plant cells


W. Van den Ende, Laboratorium voor Moleculaire Plantenfysiologie, Katholieke Universiteit Leuven, Kasteelpark Arenberg 31, bus 2434, B-3001 Leuven, Belgium
Fax: +32 161967
Tel: +32 16321952


Sugars play important roles as both nutrients and regulatory molecules throughout plant life. Sugar metabolism and signalling function in an intricate network with numerous hormones and reactive oxygen species (ROS) production, signalling and scavenging systems. Although hexokinase is well known to fulfil a crucial role in glucose sensing processes, a scenario is emerging in which the catalytic activity of mitochondria-associated hexokinase regulates glucose-6-phosphate and ROS levels, stimulating antioxidant defence mechanisms and the synthesis of phenolic compounds. As a new concept, it can be hypothesized that the synergistic interaction of sugars (or sugar-like compounds) and phenolic compounds forms part of an integrated redox system, quenching ROS and contributing to stress tolerance, especially in tissues or organelles with high soluble sugar concentrations.


abscisic acid


ascorbic acid


Arabidopsis thaliana hexokinase 1






nucleoside diphosphate-sugar


methyl jasmonate


mitochondrial hexokinase


programmed cell death


raffinose family oligosaccharides


reactive oxygen species


one of three isoforms of the B subunit of the peripheral V1 complex in vacuolar-type H-ATPase


The ability of cells to perceive and correctly respond to their microenvironment forms the basis of cellular signalling. Glucose is an important cellular nutrient that also acts as a regulatory metabolite modulating gene expression in yeast, animals and plants. In plants, sugars play important roles as both nutrients and signal molecules. Both glucose and sucrose are recognized as pivotal integrating regulatory molecules that control gene expression related to plant metabolism, stress resistance, growth and development [1–3]. There is great interest in dissecting the processes that are involved in the sugar sensing and response pathways allowing plants to adapt to the constantly changing environment, but the sugars’ dual function as nutrients and signalling molecules significantly complicates analyses of the mechanisms involved [4]. Although our knowledge of the detailed molecular mechanisms of sugar perception and signalling in plants is far from complete, hexokinase (HXK) and Snf1-related kinase 1 have already been identified as conserved sugar signalling components controlling eukaryotic energy homeostasis, stress resistance, survival and longevity. Furthermore, invertase-related sugar signals seem to be very important during plant defence reactions [5,6]. In conclusion, sugar signals integrate numerous environmental and endogenous developmental and metabolic cues and therefore operate in a complex network with plant-specific hormone signalling and stress-related pathways.

The field of antioxidants has received much attention lately, both in fundamental and applied research, but especially in the field of medicine (food supplements). Because the toxic oxygen gas appeared in the Earth’s atmosphere ∼2.2 billion years ago, aerobic organisms such as animals and plants only survived because they evolved antioxidant defence processes [7]. The need for oxygen for the efficient production of energy (ATP) in mitochondria is in balance with the necessity of controlling the level of reactive oxygen species (ROS), such as the hydroxyl radical OH; the superoxide radical O2•− and hydrogen peroxide H2O2, which are always produced in the presence of oxygen and particularly under stress. In general, metabolic rates appear to inversely correlate with stress resistance and lifespan in a variety of organisms and this has been attributed to oxidative stress [8], a deleterious process caused by the overproduction of ROS that can be important as a mediator of cell damage leading to various disease states, senescence and aging in humans, animals [9] and plants [7]. In plants, high photosynthetic activity in source leaves can induce the accumulation of both soluble sugars and ROS. Intriguingly, however, sugar starvation can lead to ROS accumulation as well [10]. Enzymic and nonenzymic antioxidants as endogenous protective mechanisms work in a complex co-operative network to reduce the cytotoxic effects of ROS in both plant and animal cells [11–13].

Increased cellular oxidation is also a key feature of leaf senescence [14], a process that seems to be associated with temporal high levels of soluble sugars. Although an involvement of metabolic signals in the regulation of plant senescence has been demonstrated in a range of studies [15, and references therein], the exact mechanisms involved in the stimulation of the ROS detoxification systems remain largely unresolved. So far, the protective effects of soluble sugars against oxidative stress have been mostly attributed to (indirect) signalling effects, triggering the production of specific ROS scavengers [10,16]. However, it was recently proposed that soluble sugars, especially when they are present at higher concentrations, might act as ROS scavengers themselves [17]. Here, the putative roles of sugars and sugar metabolizing enzymes in both sugar and ROS signalling pathways and their co-operation with antioxidant networks in plant cells are presented. A picture is emerging in which the activity of HXK may control ROS production in plant organelles such as mitochondria and chloroplasts. Also, the effect of these pathways on plant senescence is discussed.

Sugar signalling, a multifunctional network

Sugars not only fuel growth and development as carbon and energy sources, but in addition have acquired important regulatory roles as signalling molecules [2,3]. This is now particularly well established for glucose, the prime carbon and energy source in eukaryotic cellular metabolism. HXK, the first enzyme in glucose catabolism, was identified as a genuine glucose sensor, with separable catalytic and signalling activities [18,19] (Fig. 1). In addition, HXK-independent glucose and sucrose signalling pathways appear to be active in plants [20–24], but a sucrose sensor remains to be identified (Fig. 1). Although glucose signalling has been associated primarily with active cell division, respiration, cell wall biosynthesis and sugar-mediated feedback regulation of photosynthesis (Fig. 1), sucrose signalling appears to be specifically associated with anthocyanin production and with the regulation of storage- and differentiation-related processes [24–26] (Fig. 1). In addition to the important post-translational effects on metabolic enzyme activity, genome-wide expression profiling has recently uncovered the huge impact of carbon status on gene expression [27–32]. Diverse environmental stress conditions can cause cellular energy starvation, triggering metabolic reprogramming through the Snf1-related kinase 1 pathway [33 and references therein].

Figure 1.

 A hypothetical sugar–antioxidant network in plant cells. Glucose (Glc) and the activity of (organellar) HXK take a central position in this scheme, as they emerge as important regulators of cytosolic ROS. The HXK generated glucose-6-phosphate (Glc6P) and NDP-glucose boost the glycosylation of phenolic compounds, the synthesis of ASC and contribute to hormone homeostasis. The released ROS is used as a signal to stimulate the antioxidant defence system. In parallel with HXK activity, HXK acts as a glucose sensor, controlling cell division and cell expansion, in concert with hormone and ROS signalling. Next to glucose, sucrose can also be sensed and metabolized. Invertases (INVs) can influence plant growth and development either directly (e.g. by influencing source–sink relationships or by acting as a true regulatory protein) or indirectly via sugar signalling (altering the hexose/sucrose ratio). Sucrose is sensed by an unknown sensor stimulating the accumulation of reserve compounds (e.g. vacuolar fructan biosynthesis) and anthocyanins. Both anthocyanins and phenolic compounds might be involved in the scavenging of cytosolic ROS, creating phenolic compound radicals (phenolic compounds), which are reduced by ASC. It is postulated that sugars at high concentrations (sucrose, fructans, sugar-like compounds) could directly scavenge ROS derived from excess H2O2 entering the vacuole, producing sugar radicals (sugar). Vacuolar glycosylated phenolic compounds might assist in recycling the sugars from sugar radicals.

Excess photosynthate is generally transiently stored as starch in the chloroplast during the day, in part through sugar-mediated induction of gene expression and redox activation of ADP-glucose pyrophosphorylase, a key enzyme in starch biosynthesis [34–36]. Apart from the breakdown of sucrose by invertases (see below), the degradation of chloroplastic starch in leaf cells during the night (mainly via maltose and glucose export) and from plastids (amyloplasts) in starch-storing organs is also a putative source for glucose signals [25,37]. Also, trehalose metabolism appears to play an important regulatory role in co-ordinating metabolism with plant growth and development [38]. Apart from the breakdown of sucrose and reserve carbohydrates, the hydrolysis of cell wall polysaccharides might also generate sugar signals. Several cell wall glycoside hydrolases are upregulated under stress conditions such as darkness, sugar depletion, senescence and infection [33,39,40].

The exact molecular mechanisms involved in sugar signalling are still largely unknown, but mutant screens have revealed a tight interaction with various hormone signalling pathways [2]. Glucose controls abscisic acid (ABA) signalling and biosynthesis gene expression [41] and HXK-dependent signalling interacts positively and negatively with auxin and cytokinin signalling, respectively [19]. Glucose and ethylene signalling converge at the level of EIN3 protein stability [42].

Role of sucrose splitting enzymes

Sucrose is one of the most widespread disaccharides in nature. In higher plants it represents the major transport compound bringing carbon skeletons from source (photosynthetically active leaves) to sink tissues (roots, young leaves, flowers, seeds, etc.). Invertases mediate the hydrolytic cleavage of sucrose into glucose and fructose (Fig. 1). These enzymes are very well studied in plants, fungi and bacteria. For a long time, invertases were believed to be absent in the animal kingdom. Recently, however, at least two invertases have been discovered in the genome of Bombyx mori, but their exact functions remain unclear [43].

Plants possess three types of invertase: acid cell wall invertases (glycoside hydrolase family 32) located in the apoplast (i.e. the continuum of cell walls of adjacent plant cells), acid vacuolar invertases (glycoside hydrolase family 32) located in the vacuole and neutral/alkaline invertases (glycoside hydrolase family 100) located in the cytoplasm, chloroplasts and mitochondria [44–47]. Neutral invertases can be inhibited by glucose and fructose [48,49]. Knock outs of two cytoplasmic neutral invertases in Arabidopsis [50] and the model legume Lotus japonicus [51] resulted in severely reduced growth rates. It remains to be further explored whether the invertase exerts a direct effect on plant growth and development or an indirect effect via glucose signalling (Fig. 1). Intriguingly, the nuclear localization of AtCINV1, a well-characterized neutral invertase of Arabidopsis, and its interaction with a phosphatidylinositol monophosphate 5-kinase [52] strongly suggest that (at least some) neutral invertases can fulfil regulatory functions apart from their catalytic function [53]. It is becoming increasingly clear that invertases can do much more than simply hydrolyse sucrose. For instance, the LIN6 cell wall invertase of tomato is considered to be a pivotal enzyme for the integration of metabolic, hormonal and stress signals, regulated by a diurnal rhythm [54]. Expression of a yeast-derived invertase in the apoplast under a meristem-specific promoter caused accelerated flowering and enhanced branching of the inflorescence and seed yield, whereas a cytoplasmic localization of this invertase resulted in delayed flowering and both reduced seed yield and branching in Arabidopsis [55]. In potato, cytoplasmic localization of yeast invertase was detrimental to tuber yield, whereas the opposite was observed for an apoplastic yeast invertase construct [56]. Targeting of this yeast invertase to plastids resulted in early leaf senescence, consistent with reduced sucrose and increased hexose levels in the leaves [57]. These results emphasize the importance of the exact source, nature and location of the sugar signals [55] and the importance of invertases as modulators of the sucrose/glucose ratio. The sucrose/glucose ratio might be more important than the absolute sucrose and glucose concentrations, as suggested by different groups [58–63], but this requires further investigation.

Biotic or abiotic stresses and hormonal signals can also induce cell wall invertase expression and sink formation in leaf tissues [45]. Sucrose synthase is also able to split sucrose, but it generates fructose and UDP-glucose instead of glucose. It cannot be excluded that fructose-specific signalling pathways exist in plants [64]. Intriguingly, some of the so-called cell wall invertases and sucrose synthases have lost their catalytic activities, and they may function as regulatory proteins [65–68], but this requires further investigation. They are possibly involved in sucrose sensing processes.

Soluble sugars as a part of an antioxidant system

ROS are continuously produced during mitochondrial respiration [69,70] and photosynthesis [71]. The mitochondrial source of ROS production is as important in nonphotosynthesizing plant cells as it is in mammalian cells [69]. Small soluble sugars and the enzymes associated with their metabolic pathways are widely believed to be connected to oxidative stress and ROS signalling [10,72–74]. On the one hand, endogenous sugar availability can feed the oxidative pentose phosphate pathway [10,75], creating reducing power for glutathione (GSH) production, contributing to H2O2 scavenging. On the other hand, excess sugar production in source leaves by increased photosynthetic activities may result in the generation of excess cytosolic H2O2, especially when the export of sugars from these leaves is hampered due to decreased sink strength under stress. Therefore, it was proposed that sugars themselves, especially the longer water-soluble oligo- and polysaccharides, such as fructans, might be effective candidates for capturing ROS in tissues exposed to a wide range of environmental stresses [17]. Fructans are known to protrude deep between the headgroups of the membranes (Fig. 2) to stabilize them [76]. Under stress, excess cytoplasmic H2O2 can diffuse through the tonoplast (aquaporins can assist in this process), where tonoplast-bound class III peroxidases catalyse the reduction of H2O2 by taking electrons to various donor molecules, such as phenolic compounds, lignin precursors, auxin or secondary metabolites. However, the functioning of these peroxidases is also accompanied by the production of OH and OOH through the so-called hydroxylic cycle of these enzymes [77,78]. Vacuolar fructans are ideally positioned to stabilize the tonoplast, but also to temporarily scavenge the aggressive OH and OOH radicals that are produced in the vicinity of these membranes (Fig. 2). In this process, the fructans (or other sugars/sugar-like compounds) are converted into (less harmful) fructan radicals. It has been proposed that such sugar radicals could be recycled back into sugars with the help of phenolic compounds or anthocyanins [17,79]. Interestingly, the synthesis of phenolic compounds/anthocyanins can also be stimulated by sugar-mediated signalling and metabolic pathways (Fig. 1). Synthetic oligosaccharides also show strong antioxidant activity in vitro and counteract lipid peroxidation processes in mice [80]. In conclusion, it can be postulated that vacuolar sugars or sugar-like compounds, present in the vicinity of the tonoplast and interacting with this membrane (Fig. 2), might fulfil crucial roles in scavenging radicals and thus preventing lipid peroxidation by excess H2O2 produced under stress conditions.

Figure 2.

 A possible dual role for vacuolar fructans in the vicinity of the tonoplast under stress. Abiotic and biotic stresses can lead to increased concentrations of cytosolic H2O2, which can enter the vacuole via diffusion and/or through aquaporins. Vacuolar fructans (blue) can insert deeply between the headgroups of the tonoplastic membranes, stabilizing them under stress. Type III peroxidases (green) associate intimately with the inner side of the tonoplast. Peroxidases produce OH radicals. Fructans are well positioned to scavenge these radicals, a process in which fructan radicals and water are formed. Fructan radicals might be generated back into fructans with the help of phenolic compounds (see also Fig. 1).

Among a range of small sugars tested, sucrose showed the strongest antioxidant capacity in vitro [81–83], strongly suggesting that similar antioxidant reactions with sucrose can also occur in planta. At low concentrations, sucrose might serve as a substrate or signal for stress-induced modifications, whereas at higher concentrations it may function directly as a protective agent (e.g. in vacuoles of sugar beet and sugar cane plants, co-operating with the classic, cytoplasmic antioxidant systems) (Fig. 1) [17].

Other important water-soluble carbohydrates derived from sucrose (sucrosyl oligosaccharides) include the raffinose family oligosaccharides (RFOs: α-galactosyl extensions of sucrose), next to the fructans (β-fructosyl extensions of sucrose). Sucrosyl oligosaccharides and the enzymes associated with their metabolism might interact indirectly with ROS signalling pathways. Recently, RFOs as well as galactinol have been proposed to fulfil important roles in oxidative stress protection in plants [81,84] and seeds [85–87]. Previously, raffinose was shown to protect photophosphorylation and electron transport of chloroplast membranes against freezing, desiccation and high temperature stress [88], strongly suggesting that chloroplastic RFOs might be operating as ROS scavengers. The oxidized RFO radicals might be regenerated by ascorbic acid (ASC) or other reducing antioxidants, such as flavonoids [89]. The overexpression of galactinol synthase (GolS1, GolS2, GolS4) and raffinose synthase in transgenic Arabidopsis plants increased the galactinol and raffinose concentrations and resulted in effective ROS scavenging capacity and oxidative stress tolerance [81]. Moreover, lipid peroxidation was significantly lower than in wild-type plants. Furthermore, these transgenic plants exhibited higher photosystem II (PSII) activities compared with wild-type plants, and appeared to more tolerant under high light and chilling conditions [81].

Fructans might protect plants against freezing/drought stresses by stabilizing membranes [90,91]. Recent studies on transgenic plants carrying fructan biosynthetic genes [92–94] suggest that the enhanced tolerance of these plants is associated with the presence of fructans. Their reduced lipid peroxidation levels indicate that fructans, similar to RFOs, might also act directly as ROS scavengers. Alternatively, fructans might work indirectly by stimulating other specific antioxidative defence mechanisms [17]. Intriguingly, changes in fructan concentrations showed a close correlation with changes in ASC and GSH concentrations in immature wheat kernels, strongly suggesting a connection with the well-known cytoplasmic antioxidant systems. Therefore, fructans may form an integral part of a more complex ROS scavenging system in fructan-accumulating plants. It was proposed that glucose, produced by the vacuolar fructan initiator enzyme sucrose : sucrose 1-fructosyl transferase, after retranslocation to the cytoplasm could directly fuel biosynthesis of classical antioxidants [95,96], establishing a direct connection between vacuolar and cytoplasmic antioxidation mechanisms.

Similarly, sugar alcohols (mannitol, inositol, sorbitol) also possess ROS scavenging capacities. In tobacco, mannitol is believed to protect thioredoxin, ferredoxin, GSH and the thiol-regulated enzyme phosphoribulokinase against OH radicals. The targeting of mannitol biosynthesis to chloroplasts in transgenic tobacco plants resulted in an increased resistance to methyl viologen-induced oxidative stress. It became clear that mannitol in the chloroplast does not reduce ˙OH radical production, but that it increases the capacity to scavenge these radicals and protects cells against oxidative damages [97]. Contrary to glucose, fructose and sucrose, mannitol, even at high concentrations, does not repress photosynthesis and its presence has no obvious harmful effects on plants [97,98].

A role for trehalose in protection against ROS has also been demonstrated [99]. This is particularly important in micro-organisms, as only a few plants are known to accumulate trehalose to a great extent [100]. Trehalose is capable of reducing oxidant-induced modifications of proteins during exposure of yeast cells to H2O2 [101]. The ability of trehalose to reduce intracellular oxidation during dehydration has been demonstrated, especially when yeast cells were deficient in superoxide dismutase [102]. Moreover, as observed for fructans, trehalose reduced the levels of lipid peroxidation, suggesting an additional property of this sugar for improving tolerance to water loss. Also it was shown that, in vitro, trehalose significantly reduces oxidation of unsaturated fatty acids through a weak interaction with the double bonds [103]. The regulated overexpression of trehalose biosynthetic genes in transgenic rice plants produced increased amounts of trehalose in the shoot and conferred high levels of tolerance to salt, drought and low-temperature stresses. Compared with nontransformed rice, several independent transgenic lines exhibited sustained plant growth, less photo-oxidative damage and a more favourable mineral balance under stress conditions [104].

Implication of HXK in sugar signalling and antioxidant activity

HXKs are ubiquitous proteins in all living beings, catalysing the phosphorylation of glucose and fructose [105]. HXKs are not only essential in the first step of glycolysis. There is also evidence that some HXK forms are involved in NDP-sugar synthesis [106]. HXKs appear to occur in two groups: one with plastid signal peptides (type A) and one with N-terminal membrane anchors (type B) [107], and have been found in the cytosol, the stroma of plastids, the Golgi complex [108–113] or associated with the mitochondrial membrane [114]. The major glucose phosphorylating enzyme in the moss Physcomitrella patens is a chloroplast stromal HXK [108].

In addition to its familiar role as a metabolic enzyme, Arabidopsis thaliana HXK1 (AtHXK1) serves as an intracellular glucose sensor. This sensor is predominantly associated with mitochondria (mtHXK), but was also reported to be found in the nucleus [115,116]. The nuclear AtHXK1 appears to be part of a glucose signalling complex (Fig. 3) that suppresses the expression of photosynthetic genes. The signalling activity of AtHXK1 requires two unexpected partners: VHA-B1 and RPT5B (Fig. 3). VHA-B1 is one of the three isoforms of the B subunit of the peripheral V1 complex in vacuolar-type H-ATPase and is responsible for noncatalytic ATP binding. RPT5B is a subunit of the 19S regulatory particle, which binds either end of the 20S proteasome to provide ATP dependence and the specificity for ubiquitinated proteins. At a low glucose concentration, the target photosynthetic genes are expressed by a specific transcription factor. At a high glucose concentration, glucose diffuses into the nucleus and binds to nuclear HXK1. This process probably triggers a conformational change in HXK1, which then acts as a transcriptional repressor together with VHA-B1 and RPT5B (Fig. 3). In animals, the MondoA/Mlx complex is believed to monitor intracellular glucose-6-phosphate concentrations, translocating the complex to the nucleus when levels of this key metabolite increase. However, nuclear localization of the MondoA/Max-like X (Mlx) complex depends on the enzymatic activity of HXK [117]. In yeast, glucose-dependent changes in gene expression utilize at least three mechanisms, including carbon catabolite repression in which Saccharomyces cerevisiae hexokinase 2 (ScHXK2) has a nonmetabolic role in modulating specific transcriptional regulators [118]. In maize, the inhibition of mtHXK by ADP, mannoheptulose and glucosamine as compared with the insensitive cytosolic HXK, has been considered as evidence for a putative role of mtHXK in hexose sensing [109]. However, this interesting hypothesis awaits more direct demonstration. Such investigations may also clarify whether (and under which conditions) HXK is translocated from the mitochondrion to the nucleus, or mediates distinct signalling events depending on its subcellular distribution [119].

Figure 3.

 The AtHXK1 signalling complex driving the expression of photosynthetic genes in Arabidopsis, depending on the glucose concentration. (A) When the glucose concentration (white dots) is low, the target photosynthetic genes are turned on by a specific transcription factor (white). This activation may involve an unidentified coactivator complex. (B) When glucose is in excess (for instance, under nitrate-deficient conditions or stress condition when invertases are activated, and glucose is not utilized in cells), glucose diffuses freely into the nucleus, where it binds to nuclear AtHXK1. Glucose binding triggers a conformational change in AtHXK1, which then acts together with VHA-B1 (blue) and RPT5B (yellow) as a transcriptional repressor. Both VHA-B1 and RPT5B interact with the transcription factor, thereby linking AtHXK1 to DNA binding and suppressing gene expression. The mechanism of the nuclear translocation of AtHXK1, VHA-B1 and RPT5B is still unknown. This figure has been reproduced from [116] and reprinted with permission from AAAS.

The rice OsHXK5 and OsHXK6 proteins appear to have a similar dual localization and sensor function [120]. In addition to HXKs, plants also contain several fructokinases, some of which might also be involved in sugar sensing [121]. Surprisingly, the AtHXK1 mutant glucose insensitive 2 is insensitive to glucose, but is still sensitive to fructose and sucrose [2].

The mitochondrial localization of the HXKs is thought to improve access to the ATP produced in respiration for consumption by active metabolite fluxes through sucrose cycling, glycolysis and sugar nucleotide synthesis. Consistently, an entire functional glycolytic metabolon appears to be associated with the outer mitochondrial membrane, allowing pyruvate to be provided directly to the mitochondrion, where it is used as a respiratory substrate and glycolytic enzymes appear to associate dynamically with mitochondria in response to respiratory demand [122,123].

These observations suggest that mtHXK activity could be involved in the regulation of both mitochondrial respiration and ROS production in plants, similar to the key preventive antioxidant role of mtHXK through a steady-state ADP recycling mechanism in rat brain neurons [124]. Similar to mammalian HXKs [125,126], plant mtHXKs have also been reported to associate with porin/voltage-dependent anion channels [115], but further research is needed to unravel the precise interactions between plant mtHXKs, voltage-dependent anion channels and the outer mitochondrial membrane. The presence of different hydrophobic N-terminal sequences in mammalian and plant mtHXKs suggests that these interactions might substantially differ between mammals and plants. It was demonstrated that HXKI and HXKII reduce the intracellular levels of ROS and inhibit the mitochondrial permeability transition pore in mammalian cells [127]. An authentic mtHXK activity, which is subject to inhibition by ADP, was detected on potato tuber (Solanum tuberosum) outer mitochondrial membranes [128]. A mtHXK activity with similar kinetics has been described in pea leaves [129]. The potato mtHXK is much more sensitive to ADP inhibition in the micromolar range with glucose as a substrate than with fructose, suggesting a different affinity for these hexoses [128]. Moreover, it was demonstrated that this mtHXK can contribute to a steady-state ADP recycling (ADP production by mtHXK, bound to the outer mitochondrial membrane; ADP consumption through oxidative phosphorylation) that regulates H2O2 formation in the electron transport chain on the inner mitochondrial membrane. Importantly, this mitochondrial ADP recycling mechanism led to a decrease in the mitochondrial membrane potential, whereas an inhibition of mtHXK led to an increase in H2O2 production. Thus, mtHXK bound to the outer mitochondrial membrane can guide the ADP delivery to the F0F1ATP synthase enzyme via the adenine nucleotide transporter in an efficient channelling to the mitochondrial matrix. In conclusion, mtHXK activity plays a specific role in generating ADP to support oxidative phosphorylation, thereby avoiding an ATP synthesis-related limitation of respiration and subsequent H2O2 release in plants [128]. In maize roots, mtHXK activity is believed to be directly linked to NDP-sugar synthesis (Fig. 1), needed for cellulose, phenylpropanoid and flavonoid biosynthesis. The attachment of mtHXK to the mitochondrial membrane is absolutely necessary to perform its preventive antioxidant activity, in both animal and plant mitochondria [124,127,128]. Interestingly, both methyl jasmonate (MeJa) and glucose-6-phosphate are known to induce the detachment of mammalian mtHXK from the outer mitochondrial membrane [124,130,131]. Glucose-6-phosphate had no such effect on plant mitochondria [128]. The effect of MeJa on the detachment of mtHXK from plant mitochondria has not yet been reported. However, MeJa addition to plants leads to ROS accumulation, alterations in mitochondrial movements and morphology, and cell death [132], similar to that observed in animal cells [131], suggesting that similar mechanisms might be operating in plant and animal cells, but this remains to be demonstrated.

Next to the important signalling function of HXKs, as described above for AtHXK1, these new data indicate that mtHXK activities might play a key role as a regulator of ROS levels (Fig. 1). mtHXKs could respond rapidly to changes in the cellular demand for glucose-6-phosphate, which is known to be a key intermediate in several metabolic pathways sensitive to ADP/ATP ratios, including glycolysis, sucrose synthesis, the pentose phosphate pathway, cellulose biosynthesis and phenylpropanoid and flavonoid biosynthesis (Fig. 1). However, the links between glucose, HXK and ROS control extend the processes described in mitochondria. Indeed, the HXK-derived glucose-6-phosphate (Fig. 1) can feed the l-galactose route in the so-called Smirnoff–Wheeler pathway, leading to biosynthesis of ASC [133], which has major implications in the well-known cytoplasmic ROS detoxification processes, cell elongation and, possibly, cell division [134]. ASC works in close co-operation with GSH to remove H2O2 via the Halliwell–Asada pathway. Moreover, ASC takes part in the regeneration of α-tocopherol, providing extra protection of the membranes [135]. The NDP-sugars, produced by the activity of sucrose synthases (and other enzymes), can serve as donor substrates for glycosyltransferases that catalyse the glycosylation of most plant hormones (except ethylene), contributing to hormone homeostasis [136]. Moreover, NDP-sugars are also important as donor substrates for the glycosylation and stability of many secondary metabolites, such as phenolic compounds (Fig. 1), contributing to increasing antioxidant abilities [17] and assisting in recycling sugars from sugar radicals (Fig. 1). Cellular NDP-sugar concentrations are often very low. This includes that they are a limiting substrate for product synthesis, as was shown for (U/A)DP-glucose-dependent starch biosynthesis in potato [137] and UDP-glucose-dependent based cellulose synthesis in Avena coleoptiles [138]. Therefore, it can be postulated that an accumulation of phenolic compounds would also greatly depend on the cellular NDP-sugar concentrations, but this requires further investigation. It is widely recognized that phenolic compounds are involved in the H2O2 scavenging cascade in plant cells [139], but the links with HXK activities and sugar recycling processes have so far received little attention.

Apart from glucose, fructose and sucrose, sugar alcohols, such as sorbitol, mannitol and myo-inositol, may also be transported into vacuoles [140,141]. Anthocyanins, flavonoids and a wide array of conjugated endogenously synthesized toxic or xenobiotic compounds typically accumulate in the vacuole and several of these substances cross the tonoplast via ATP-binding cassette (ABC)-type carriers [142,143]. Phenolic compounds and fructans (or other vacuolar, sugar-like compounds) might operate in a synergistic way to scavenge excess H2O2 (Figs 1 and 2).

Taken together, glucose and HXK take a central position and a dual role in both sugar signalling and antioxidant networks, with important links to ASC biosynthesis and, via NDP- glucose production, to the synthesis of glycosylated phenolic compounds (Fig. 1). High concentrations of vacuolar compounds (sugars or sugar-like compounds) could also form an integral part of this antioxidant network (Fig. 1).

Senescence – a link between sugar signalling and ROS production pathways

In addition to aging, plants are characterized by a highly specific process termed leaf senescence. After a period of active photosynthesis, the leaf’s contribution to the plant diminishes and the leaf then enters its final stage of development: senescence. This highly regulated process is characterized by the loss of chlorophyll, the breakdown of macromolecules and the massive remobilization of nutrients to other parts of the plant [15]. Senescence can be triggered by multiple developmental and environmental signals. Drought, darkness, leaf detachment and the hormones ABA and ethylene induce leaf yellowing [144]. Cytokinins, on the other hand, can delay plant senescence, and studies with the AtHXK1 mutant glucose insensitive 2 show that sugars and cytokinins work antagonistically [19]. Interestingly, cytokinin-induced cell wall invertase expression is an essential downstream component of cytokinin-mediated local delay (green islands) of leaf senescence [58]. Leaf senescence is accompanied by considerable changes in cellular metabolism and gene expression [145]. Several of these senescence-associated genes encode transporters of sugar, peptides, amino acids and transporters that could participate in the substrate and nutrient mobilization that occurs as part of the senescence programme [146]. However, micro-array analysis has revealed significant differences in gene expression between dark/starvation-induced and developmental senescence [147].

ROS play an important role in the response of plants to biotic and abiotic stress, plant cell growth, regulation of gene expression, stomatal opening, hormone signalling and programmed cell death (PCD) [148–153]. This multifaceted role for ROS indicates a tight control of their production and accumulation levels. The observation that ROS may mediate both ABA signalling and ABA biosynthesis [154] suggests that the feedback regulation of ABA biosynthetic genes by ABA may be mediated in part by ROS through a protein phosphorylation cascade [155]. Senescence-associated genes are expressed in response to increases in the tissue contents of ROS [14]. Increased cellular oxidation is a key feature of leaf senescence [14]. PCD may be triggered by enhanced ROS levels and cellular oxidation [156–158]. Genetic evidence suggests that ROS do not trigger PCD or senescence by causing physicochemical damage to the cell, but rather act as signals that activate pathways of gene expression that lead to regulated cell suicide events [152,153]. An imbalance between ROS production and antioxidant defence can lead to an oxidative stress condition. Increased levels of ROS may be a consequence of the action of plant hormones, environmental stress, pathogens, altered sugar levels and fatty acids [159–162] and ROS may in turn induce ROS scavengers and other protective mechanisms (Fig. 1) [163].

For leaves, evidence supports a role for sugar accumulation in the initiation and/or acceleration of senescence. However, the regulation of senescence or aging may respond to different metabolic signals in heterotrophic plant organs and heterotrophic organisms [164]. van Doorn [165] questioned whether sugar increases or decreases cause leaf senescence, and concluded that there is not enough hard evidence to demonstrate a causative relationship. The long-lasting debate on this matter probably reflects the complexity of senescence regulation, with sugars being only one factor. The initial trigger of processes leading to cell death probably needs to be searched for in mitochondria [132]. Apart from the actual concentrations of glucose and sucrose within mitochondria and their surroundings, it becomes increasingly clear that the activity of mtHXK is essential to regulate ROS levels in mitochondria (Fig. 1). Although a temporal ROS overshoot, via ROS signalling, leads to an increased production of antioxidants contributing to ROS scavenging and stress tolerance, it can be speculated that a more drastic ROS overshoot could serve as an initial trigger of senescence. Consistent with this hypothesis and the central importance of mtHXK activity, it was recently demonstrated that MeJa addition, perhaps acting by releasing mtHXK from the outer membrane, results in four sequential processes: (a) ROS accumulation, (b) alterations in mitochondrial movements and morphology, (c) photosynthetic dysfunction and (d) cell death [132].

The Arabidopsis hypersenescing mutant hys1 provides the clearest link between sugar, senescence and stress signalling. The repressive effect of sugars on photosynthetic gene expression and activity and the correlation between HXK expression and the rate of leaf senescence [19,21] are indicative of an important role for HXK-dependent sugar signalling in leaf senescence. Similar to the effects of targeting the yeast invertase to chloroplasts, an accelerated senescence was observed during overexpression of HXK [166]. HXK overexpression in tomato plants impaired growth and photosynthesis, and induced rapid senescence in photosynthetic tissues [167]. It was also demonstrated that mtHXKs play a role in the control of PCD in Nicotiana benthamiana [168]. HXK dislocation from mitochondria leads to the closure of voltage-dependent anion channels and subsequent mitochondrial swelling and cell death in mammalian cells [169]. HXK binding to the mitochondria inhibits apoptosis [169]. Increasing glucose phosphorylation activity by mtHXK may reduce apoptosis through both the inhibition of mitochondrial permeability transition and more efficient glucose metabolism due to their better access to ATP [127].

In conclusion, plant mtHXKs might act as a glucose sensor (regulatory function) and, through its catalytic activity, as a crucial regulator of ROS levels in mitochondria (Fig. 1). Similar functions could be postulated for chloroplastic HXKs.

All abiotic stresses generate ROS, potentially leading to oxidative damage affecting crop yield and quality. Next to the well-known classical antioxidant mechanisms, sugars and sugar metabolizing enzymes come into the picture as important new players in the defence against oxidative stress. Therefore, sugar metabolizing enzymes, such as organellar HXKs, form promising targets to improve crop yield, stress tolerance and longevity in plants.


Sugars are finally being recognized as important regulatory molecules with both signalling and putative ROS scavenging functions in plants [2] and other organisms [170–172]. The exact source, nature and location of the sugar and nonsugar signals, such as ROS and hormones, are important to provide an integrative regulatory mechanism controlling various functions of a plant cell. The responses to sugars and oxidative stress are not only linked (Fig. 1), but sugars also affect scores of stress-responsive genes [27]. Apart from acting as signals, we hypothesize that vacuolar sugars or sugar-like compounds, possibly in combination with phenolic compounds, form a so far unrecognized vacuolar redox system acting in concert with the well-established cytoplastic antioxidant mechanisms.


WVdE, KLR, LX and FR are supported by grants from FWO Vlaanderen.