Plant Growth Regulation

, Volume 62, Issue 3, pp 241–255

Desiccation tolerant plants as model systems to study redox regulation of protein thiols

SI Plant Desiccation Stress

DOI: 10.1007/s10725-010-9482-9

Cite this article as:
Colville, L. & Kranner, I. Plant Growth Regul (2010) 62: 241. doi:10.1007/s10725-010-9482-9

Abstract

While the majority of plants and animals succumb to water loss, desiccation tolerant organisms can lose almost all of their intracellular water and revive upon rehydration. Only about 300 ‘resurrection’ angiosperms and very few animals are desiccation tolerant. By contrast, many bryophytes and most lichens are desiccation tolerant and so are the seeds and pollen grains of most flowering plants. The current literature reveals that the extreme fluctuations in water content experienced by desiccation tolerant organisms are accompanied by equally extreme changes in cellular redox state. Strongly oxidizing conditions upon desiccation can cause irreversible oxidation of free cysteine residues of proteins, which can change protein structure and function, and contribute to protein denaturation. It appears likely that reversible formation of disulphide bonds, in particular through protein glutathionylation, contributes to the set of protection mechanisms that confer desiccation tolerance. Upon rehydration, de-glutathionylation can be catalyzed by glutaredoxins (GRXs) and protein disulphide bonds can be reduced through NADPH-dependent thioredoxins (TRXs). Due to their ability to survive severe oxidative stress, desiccation tolerant plants and seeds are excellent models to study protein redox regulation, which may provide tools for enhancing tolerance to drought and more generally, to oxidative stress, in crops.

Keywords

Desiccation Glutaredoxin Glutathione Glutathionylation Thiol Thioredoxin 

Introduction

Desiccation tolerance

Desiccation tolerance is the ability to revive from the air-dried state. Desiccation tolerant (DT) life forms can lose more than 90% of their water, and resume metabolism when water is available again. Desiccation tolerance is a complex trait with a wide taxonomic range. Species belonging to five animal phyla and four plant divisions demonstrate tolerance to desiccation at some point in their life cycle. Among fungi, desiccation tolerance is widespread in lichens, and some yeasts are DT. Examples of DT organisms from the animal kingdom include certain species of nematodes, rotifers, tardigrades, some crustacean embryos and the larva of the fly Polypedilum vanderplanki. In the plant kingdom desiccation tolerance is common amongst mosses and liverworts, but is a rare trait of adult angiosperms, with only around 300 species able to survive desiccation as mature plants, often referred to as ‘resurrection plants’. However, the spores of pteridophytes and the majority of the seeds and pollen of the flowering plants are DT (Alpert 2006). DT seeds are termed ‘orthodox’ and dry to a low water content (WC) during maturation on the mother plant. After shedding dry orthodox seeds can remain viable for long periods of time. By contrast, ‘recalcitrant’ seeds are desiccation sensitive (DS). They do not undergo maturation drying and remain metabolically active throughout development and after shedding from the mother plant. Recalcitrant seeds are adapted to germinate immediately, and if conditions are unsuitable the seeds will start to desiccate, resulting in loss of germination capacity (Berjak and Pammenter 2008). A notable example of orthodox seed longevity was demonstrated by the successful germination of sacred lotus seeds that, according to radiocarbon dating, were over 1000 years old (Shen-Miller et al. 1995). Some species of mosses, liverworts and resurrection plants have been reported to survive several years in the desiccated state (Alpert 2005).

Desiccation tolerance is thought to be an ancient trait that was lost during the evolution of higher plants but retained in their seeds and pollen. The observation that some protective mechanisms are shared in seeds and vegetative tissues of resurrection plants led to the hypothesis that desiccation tolerance in vegetative tissue had re-evolved from seeds (Illing et al. 2005). It is conceivable that desiccation tolerance is an ancestral trait of land-dwelling plants, because the first organisms that made the transition from life in the sea to life on land will have had to rely on constitutive protection mechanisms before they had evolved specialized desiccation-avoidance mechanisms, including cuticles, bark or skin (Kranner et al. 2008).

Sugars, LEA proteins and vitrification

Desiccation tolerance appears to be based upon the induction of protective mechanisms during dehydration and repair during rehydration (Black and Pritchard 2002; Hoekstra et al. 2001; Jenks and Wood 2007; Kranner and Birtic 2005; Kranner et al. 2008; Moore et al. 2009). Whether DT life forms have a common set of structural, physiological and molecular mechanisms that constitute desiccation tolerance based on a shared ‘desiccome’ still needs to be elucidated (Potts et al. 2005). However, there is strong evidence for the presence of specific protection mechanisms. Desiccation tolerance correlates with the presence of non-reducing di- and oligosaccharides and other compatible solutes, e.g. polyols, and amino acids, e.g. proline. Late embryogenesis abundant proteins (LEAs), dehydrins and heat shock proteins (HSPs) also appear to confer desiccation tolerance (Wise and Tunnacliffe 2004). These proteins may act as molecular shields that help to prevent molecular interactions, membrane fusion and protein aggregation (Hoekstra et al. 2001). Drying to low WCs (<0.3 g H2O g−1 dry mass) leads to saturation of cytosolic components and a decrease in the molecular mobility, known as intracellular glass formation or ‘vitrification’. Intracellular glasses decelerate changes in ionic strength and pH, and chemical reactions generally, so that the cells of dry DT organisms are, to some extent, preserved in space and time, i.e. they still deteriorate, but at lower rates. Sugars and LEA proteins facilitate glass formation, and the composition of non-reducing sugars influences the temperature at which glass formation occurs, and thereby affects desiccation tolerance. For example, the soluble sugars in DT Zea mays embryos consist of 85% sucrose and 15% raffinose, a mixture which can form glasses at temperatures above 0°C, whilst DS axes typically contain 75% glucose and 25% sucrose, and can form glasses only at sub-zero temperatures (Koster 1991; Buitink and Leprince 2004). Accumulation of raffinose and stachyose also coincided with the acquisition of desiccation tolerance in the embryonic axes of Phaseolus vulgaris seeds, and may facilitate glass formation (Bailly et al. 2001; Koster and Leopold 1988). Intracellular glass formation does not provide protection during drying since glasses are formed at WC below those at which DS tissue can survive. However, intracellular glasses enable prolonged survival in the dry state (Buitink and Leprince 2004).

Protection from oxidative damage

Some of the deleterious effects of desiccation have been related to reactive oxygen species (ROS) formation as a result of unbalanced metabolism and impairment of respiratory and photosynthetic electron transport. In particular, photosynthesis is very sensitive to water deficit. Electron leakage during photosynthetic electron transport and the formation of singlet oxygen are increased when cells of photosynthetic tissue are desiccated in the light (Kranner et al. 2002; Seel et al. 1992; Smirnoff 1993). ROS react with cellular macromolecules, e.g. proteins and lipids, causing damage and disruption of cell function. Orthodox seeds undergo a concerted downregulation of metabolism during the early stages of desiccation, which may limit ROS formation (Leprince et al. 2000). The embryos of recalcitrant seeds can be flash-dried and cryopreserved, which avoids metabolism-linked damage associated with slow drying by rapidly passing through intermediate WCs (Berjak and Pammenter 2008). Likewise, bryophytes and lichens are poikilohydric, which means that the tissue WC rapidly equilibrates to the water potential of the surrounding environment. Therefore, most bryophyte cells exist either in a fully turgid or a desiccated state for most of the time, with transition between the two occurring relatively fast, which means that metabolism under reduced water potential occurs only transiently (Proctor et al. 2007). Whilst DT organisms such as lichens and bryophytes may partly prevent ROS formation by being able to desiccate rapidly, others, such as orthodox seeds, may use concerted downregulation of their metabolism. Hence, increased ROS formation due to disturbances in electron transport chains may be confined to certain time periods upon cellular desiccation, but cannot be entirely avoided.

In the dry state, metabolic activity ceases, and ROS production becomes less likely. However, ROS can still be generated through auto-oxidation processes, for example, of lipids, and damage to cellular macromolecules can occur as a result of Maillard reactions (Bailly 2004; Murthy et al. 2003; Wettlaufer and Leopold 1991). Rehydration of desiccated tissue is also associated with increased intracellular ROS formation due to the restarting of metabolism. In addition, ROS may be generated extracellularly by NADPH oxidase or peroxidases as part of an oxidative burst during rehydration, where they may play roles in pathogen defence, growth and differentiation. Extracellular ROS production was demonstrated for desiccation-stressed thalli of the liverwort Dumortiera hirsuta (Beckett et al. 2004), for lichens and bryophytes (Minibayeva and Beckett 2001) and during seed imbibition (Kranner et al. 2010; Schopfer et al. 2001).

Considering the repeated exposure of DT organisms to desiccation and rehydration, both of which cause increased ROS formation, a potent antioxidant machinery is essential to scavenge excess ROS that are inevitably formed upon desiccation. The ability of DT organisms to induce repair mechanisms upon rehydration includes rapid regeneration of antioxidant systems (Kranner 2002; Kranner et al. 2002). In contrast to DS organisms, DT ones can apparently maintain function of their antioxidant machinery in the desiccated state or quickly re-synthesize antioxidants upon rehydration (Kranner and Birtic 2005; Pukacka and Ratajczak 2007). In summary, the extreme changes in water content in DT organisms give rise to severe changes in cellular redox state through repeated cycles of ROS formation and scavenging.

Thiol-disulphide conversions

Such changes in cellular redox state can affect structure and function of redox-sensitive proteins. Free cysteine (Cys) residues of proteins can form intra-molecular disulphide bonds, changing protein tertiary structure and thus, function, or be irreversibly oxidized to sulphonic acids, also affecting protein function. Kranner and Grill (1996, 1997) reviewed the significance of thiol-disulphide exchange in desiccation tolerance, and proposed that the accumulation of protein-bound glutathione (PSSG) during dehydration serves to protect protein thiol groups from desiccation-induced oxidative damage. At the time, the mechanisms of thiol-disulphide exchange and the enzymes involved in reducing mixed disulphides during rehydration were unknown. Meanwhile, the general importance of glutathionylation in protein redox regulation has been recognized and enzymes involved in thiol-disulphide exchange have been uncovered. This review provides an update to reflect recent findings and discuss the role of thiol-disulphide exchange in desiccation and rehydration. Thiol-disulphide conversions have wide significance outside the field of desiccation tolerance, but due to their ability to survive extreme changes in water status and cellular redox state, we believe that desiccation tolerant organisms are excellent models to study redox regulation of proteins. We first summarize findings on ROS formation and antioxidant response in DT plants and lichenized fungi that affect cellular redox state, followed by more detailed discussion of glutathione (γ-glutamyl-cysteinyl-glycine; GSH), protein glutathionylation during desiccation and enzymes that re-reduce protein thiol groups (PSH), namely GRXs and TRXs, and finally present an updated thiol-disulphide cycle with a proposed role in protein protection.

Redox changes during desiccation and rehydration

Antioxidants

Antioxidant systems are essential to maintain redox homeostasis by scavenging excess ROS and have been abundantly studied in DT organisms. The major low-molecular-weight (LMW) antioxidants are ascorbate (AsA), GSH and tocopherols. AsA and GSH are water soluble and found throughout the cell whereas tocopherols are lipid soluble and are located in membranes. Both AsA and GSH can scavenge ROS directly and also act as electron donors for ROS-detoxifying enzymes e.g. AsA peroxidase (APX) and glutathione-S-transferase (GST) (Noctor and Foyer 1998). AsA and GSH participate in the AsA-GSH cycle, in which AsA is oxidised to dehydroascorbate (DHA) through reaction with ROS, and the DHA is subsequently reduced by GSH-dependent DHA reductase (DHAR). The resulting glutathione disulphide (GSSG) is reduced by NADPH-dependent glutathione reductase (GR). AsA is also required for the regeneration of α-tocopherol from the α-tocopheroxyl radical (Noctor and Foyer 1998; Smirnoff and Wheeler 2000). NADPH for GSSG reduction can be delivered from photosynthesis, and from the oxidative pentose phosphate pathway in non-photosynthetic tissues such as seeds and before photosynthesis commences in DT photosynthetic organisms at the early stages of rehydration. Glucose-6-phosphate dehydrogenase (G6PDH) is a key enzyme of the oxidative pentose phosphate pathway and is often used as an indirect indicator of NADPH availability (Kranner and Birtic 2005).

Another class of antioxidant enzymes with putative roles in dehydration response are peroxiredoxins (PRXs), which are ubiquitous thiol-based peroxidases. There are four PRX subfamilies: 1-Cys PRX, 2-Cys PRX, PRX Q and PRX II (Dietz 2003). In addition, GPXs have recently been reclassified as a fifth class of PRXs (Navrot et al. 2006). 1-Cys PRXs contain one conserved catalytic Cys residue and PRXs of the other subfamilies have two. 1-Cys PRXs are located in the nucleus, and are thought to have a DNA-protective role (Dietz et al. 2006).

Ascorbate and glutathione during seed desiccation and imbibition

In the dry state the decreased molecular mobility dramatically slows enzymatic catalysis, and AsA and GSH become gradually oxidised as they react with ROS, but are not regenerated. In the case of AsA this leads to an overall depletion of the AsA/DHA pool, because DHA is relatively unstable and can be broken down non-enzymatically to diketogulonic acid (De Tullio and Arrigoni 2003). In orthodox seeds, AsA declines during maturation drying and is completely absent from dry seeds, and only low levels of DHA may be detected. Besides its role as an antioxidant, AsA is involved in cell cycle activity, which will be unwanted when a maturing seeds enters into the quiescent state, so AsA synthesis is down-regulated which also contributes to the depletion of the AsA pool in dry orthodox seeds. Synthesis of AsA begins immediately upon imbibition, and galactonolactone dehydrogenase, the last enzyme in AsA biosynthesis, is present in dry seeds. Dry seeds also contain active DHAR to reduce DHA and provide AsA whilst biosynthesis is upregulated during the intitial stages of rehydration (De Tullio and Arrigoni 2003; Tommasi et al. 1999).

GSSG on the other hand accumulates and persists in the dry state. The ratio of GSH/GSSG decreases as seeds desiccate. Increasing time in the desiccated state leads to progressive GSSG formation with or without loss of total glutathione (i.e., GSH + GSSG) (reviewed by Kranner and Birtic 2005). For example, during seed maturation of Triticum durum the WC of the kernels decreased from 77 to 52%, and the total ascorbate (AsA + DHA) and total glutathione contents, and DHAR and catalase (CAT) activities increased. A further drop in WC to 9% correlated with a decrease in APX and CAT, DHAR, ascorbate free radical reductase (AFRR) and GR activities. Total AsA content drastically decreased and only DHA remained in mature kernels. The total glutathione content also declined and the GSH/GSSG ratio increased (de Gara et al. 2003). Upon rehydration GSH is restored through reduction of GSSG. GSSG accounted for 24% of the total glutathione pool in dry Pisum sativum seeds, but was reduced to 3% within 14 h of imbibition (Kranner and Grill 1993). Similarly, in dry Pinus pinea seeds GSSG represented 50% of the total glutathione pool in embryos. This was reduced to less than 4% after 24 h of germination as a result of GSSG reduction by GR (Tommasi et al. 2001).

A study comparing AsA and GSH metabolism in DT Acer platanoides (Norway maple) seeds and recalcitrant Acer pseudoplatanus (sycamore) seeds showed that an increase in GSH content coincided with the acquisition of desiccation tolerance in Acer platanoides seeds and throughout desiccation the GSH/GSSG ratio was much higher in Acer platanoides seeds than in Acer pseudoplatanus (Pukacka and Ratajczak 2007). ROS production was higher in Acer pseudoplatanus seeds during desiccation, which coincided with lower activity of antioxidant enzymes [APX; monodehydroascorbate reductase (MDHAR); DHAR; GR and glutathione peroxidase, GPX] and likely contributed to desiccation sensitivity and viability loss. ROS production also increased during desiccation of recalcitrant seeds of Acer saccharinum and this correlated with a loss of viability (Pukacka and Ratajczak 2006).

Ascorbate and glutathione during desiccation and rehydration of resurrection plants and lichens

Oxidation of antioxidants also occurs during desiccation of resurrection plants and their antioxidant status correlates with their ability to revive from the desiccated state. Kranner et al. (2002) reported that the resurrection plant, Myrothamnus flabellifolia, could revive after four months in a desiccated state with no visual signs of damage. Similar as in seeds, desiccation caused losses of leaf GSH and AsA in conjunction with formation of GSSG and DHA, which was rapidly reversed upon desiccation of viable plants. After remaining eight months in the desiccated state, antioxidant defences completely broke down; the plants failed to re-establish AsA, GSH and α-tocopherol and abscised their leaves upon rehydration. G6PDH activity was restored during rehydration, but GR activity was not, so GSSG reduction appeared to be limited by insufficient GR activity rather than NADPH availability.

Dhindsa (1987) demonstrated a dependency of GSH oxidation and re-reduction during desiccation and rehydration, respectively, on the speed of water loss in the DT moss, Tortula ruralis. GSSG formation was only observed when the moss was slowly dehydrated. GSH levels were maintained in rapidly dehydrated moss. However, it appears that the rapidly dehydrated moss had received a signal to oxidize GSH, because GSSG formation was found upon rehydration of the rapidly dehydrated moss, while the slowly dehydrated moss reduced GSSG immediately upon water uptake. A later study showed that the activities of GR, GPX and GST increased during slow drying but were unchanged by rapid drying (Dhindsa 1991). During subsequent rehydration the activities of GR, GPX and GST gradually declined to control levels within 24 h in the slowly dried moss. Conversely, in rapidly dried moss the activities of all three enzymes increased between two and eight hours of rehydration. When transcription and translation of these enzymes were limited by treatment with inhibitors, GSSG accumulated and lipid peroxidation and solute leakage increased. Others found that the DT Tortula ruraliformis maintained larger pools of GSH, α- and γ-tocopherol and greater superoxide dismutase (SOD) and GR activities following desiccation compared to the DS moss, Dicranella palustris, especially when desiccated in the light (Seel et al. 1992).

Furthermore, lichens also accumulate GSSG during desiccation and reduce it to GSH upon rehydration. In three lichens with different degrees of desiccation tolerance, the most DT species, Pseudevernia furfuracea, showed the highest rate of GSH oxidation during desiccation, indicating that this may be an active, controlled process (Kranner 2002). Rehydration was associated with rapid regeneration of GSH which correlated with the re-establishment of G6PDH activity during rehydration. The least DT species, Peltigera polydactyla, accumulated less GSSG during desiccation and GSSG reduction upon rehydration was delayed, correlating with diminished delivery of NADPH through the oxidative pentose phosphate pathway.

Protein expression in resurrection plants

A proteomic study of protein expression during dehydration and rehydration of the leaves of a resurrection plant, Boea hygrometrica, found that of 223 proteins, 35% were induced during dehydration and 5% were upregulated by rehydration. During dehydration the WC of the detached leaves dropped by ~97% within 24 h. Within 30 min of dehydration 30 proteins were induced, including a putative ABC transporter ATPase, GPX and polyphenol oxidase. These may arise due to de novo synthesis or as a result of desiccation-induced post-translational processing of existing proteins. Proteins that increased at later stages of dehydration included the putative Rubisco large subunit (rbcL), oxygen evolving complex (OEC) of photosystem II, vacuolar H+-ATPase A subunit and GST. The increase in rbcL may be due to programmed proteolysis associated with the shut down of photosynthesis which is reported to occur during dehydration of ‘poikilochlorophyllous’ resurrection plants as a means of restricting ROS formation. The induction of GPX and GST indicates the involvement of GSH in desiccation response, and in agreement with this GSH levels were found to increase during the early stages of dehydration (Jiang et al. 2007). Others found that expression of two Boea hygrometrica group 4 LEA genes in transgenic tobacco conferred increased drought tolerance compared to wild-type tobacco plants. This was associated with increased SOD and peroxidase (POX) activity and enhanced stability of photosynthesis-related proteins and membranes in the transgenic plants following drought stress (Liu et al. 2009).

Peroxiredoxins in seeds and resurrection plants

Studies of 1-Cys PRX expression in barley found that expression only occurred in the embryo and aleurone layer, and not in the rest of the endosperm tissue that undergoes programmed cell death during maturation drying. 1-Cys PRX expression increased during late embryogenesis and desiccation, and then protein levels declined following imbibition of non-dormant seeds but were maintained in dormant seeds, perhaps to strengthen antioxidant activity in the dormant state (Stacy et al. 1999). 1-Cys PRX expression also increased during seed maturation drying in Arabidopsis, leading to the assumption that 1-Cys PRXs protect against oxidative damage during desiccation (Aalen 1999; Dietz 2003). In support of this hypothesis, a 1-Cys PRX homologue (XvPer1) was identified in the nuclei of leaf cells of the resurrection plant, Xerophyta viscosa. XvPer1 was not present in unstressed plants, but expression was induced in response to stresses such as dehydration, heat, cold, high light and ABA (Mowla et al. 2002). Another role proposed by Haslekas et al. (2003) based on studies using Arabidopsis plants altered in expression of 1-Cys PRX was that 1-Cys PRX acts as a sensor for unfavourable environmental conditions and prevents seed germination occurring under such conditions. Hence, 1-Cys PRXs may play a role in desiccation tolerance in protection against oxidative damage during desiccation.

In summary, many studies have documented an increase in antioxidant capacity in the early stages of desiccation in DT organisms followed by a subsequent decline in the activities of antioxidant enzymes and progressive oxidation of the AsA and GSH pools the longer the tissues were kept in the dry state. In the dry state ROS formation and enzymatic activity may slow down and deterioration continue at lower rates, progressively impairing recovery upon rehydration the longer the tissues remain in the desiccated state. Overall, desiccation tolerance appears to correlate with maintenance of higher antioxidant capacity in the dry state in conjunction with an ability to resume antioxidant function rapidly upon rehydration.

Glutathione and protein redox state

The redox state of glutathione

Glutathione is the most abundant low molecular weight thiol in plants and is considered to be the major intracellular redox buffer (Schafer and Buettner 2001; Szalai et al. 2009), in particular in the above-mentioned absence of AsA in dry DT organisms. As discussed above, changes in the size and redox state of the total glutathione pool are common responses to desiccation both in tolerant and sensitive organisms, with a tendency of DT plants being able to maintain higher GSH levels. Importantly, DT plants and lichens seem to be particularly well equipped to reduce GSSG upon rehydration while DS ones may fail to reduce GSSG quickly (Kranner and Birtic 2005). However, GSH/GSSG ratios in DT and DS organisms may only differ subtly or not at all. The reason for this is that GSH/GSSG ratios only imprecisely reflect the reducing capacity of the GSSG/2GSH redox couple. Concentration-dependent redox couples such as GSSG/2GSH are better defined by their half cell reduction potentials, which can be calculated using the Nernst equation.
$$ GSSG + 2H^{ + } + 2e^{ - } \to 2GSH\quad {\text{Half-cell reaction}} $$
$$ E_{GSSG/2GSH} = \Updelta E^\circ - \frac{RT}{nF}\ln {\frac{{\left[ {GSH} \right]}}{{\left[ {GSSG} \right]}}}^{2} \quad {\text{Nernst equation}} $$
Where R is the gas constant (8.314 J K−1 mol−1); T, temperature in K; n, number of transferred electrons; F, Faraday constant (9.6485 × 104 C mol−1); E0′, standard half-cell reduction potential at pH 7 \( \left[ {E^{{0^{\prime } }}_{{{\text{GSSG}}/ 2 {\text{GSH}}}} = - 2 40 {\text{mV}}} \right] \); [GSH] and [GSSG] are molar concentrations of GSH and GSSG. Because two molecules of GSH are converted to one molecule of GSSG upon oxidation, [GSH] is entered as a squared term in the Nernst equation. Hence, the glutathione half-cell reduction potential (EGSSG/2GSH) is dependent on the ratio of GSH to GSSG and also on the concentration of GSH. Therefore, the relationship between the GSSG/2GSH redox couple and desiccation tolerance should be revisited. Ideally, the entire ‘intracellular redox environment’ would be assessed, which has been defined as the sum of the products of the reduction potentials and reducing capacities of all of the redox couples present. However, EGSSG/2GSH may be a good indicator and be representative of the intracellular redox environment (Schafer and Buettner 2001). Kranner et al. (2006) proposed that EGSSG/2GSH is a universal marker of plant viability, with increases in the GSSG/GSH ratio, sometimes in conjunction with GSH loss, causing EGSSG/2GSH to shift towards more positive values. A shift of EGSSG/2GSH to the zone between −180 and −160 mV corresponded to the initiation of pathways leading to cell death. This was demonstrated for orthodox Pisum sativum seeds subjected to accelerated ageing and for recalcitrant Quercus robur seeds following desiccation, where both species suffered a 50% loss of viability within this zone. Moreover, a meta-analysis of published data from a wide range of species across 13 plant and fungal orders exposed to various abiotic stresses confirmed that cell viability is lost in this zone (Kranner et al. 2006). Embryonic axes excised from recalcitrant Castanea sativa seeds showed a shift in EGSSG/2GSH of +70 mV as a consequence of declining levels of total glutathione and a decrease in the GSH/GSSG ratio that correlated with viability loss in response to desiccation (Roach et al. 2010). A re-calculation of the data in the above-mentioned study by Kranner et al. (2002) shows that desiccation of the leaves of the resurrection plant, Myrothamnus flabellifolia (Fig. 1), resulted in EGSSG/2GSH becoming more positive as the period of desiccation increased (Fig. 2a). In plants that fully recovered after desiccation for four months, EGSSG/2GSH remained more negative than −180 mV (Fig. 2c). By contrast, a shift of the EGSSG/2GSH to the zone between −180 mV and −160 mV occurred after eight months of desiccation, after which period EGSSG/2GSH rose to values more positive than −160 mV and the leaves were abscised upon rehydration. The DHA/AsA pool was also affected in Myrothamnus flabellifolia leaves following desiccation as demonstrated by the positive shift in EDHA/AsA, which was reversed during rehydration of plants that had been desiccated for up to four months, but not in leaves desiccated for eight months (Fig. 2b, d).
Fig. 1

The resurrection plant Myrothamnus flabellifolia. Topleft: Desiccated plant in the natural habitat growing on a granite outcrop. Top right: When water is available again the seemingly dead plant revives, unfolds its leaves and photosynthesises within a few hours. Bottom row: Detailed views of desiccated (left) and rehydrated (right) leaves. Images are a courtesy of Wolfgang Stuppy

Fig. 2

Redox changes during desiccation and rehydration, shown here for the leaves of the resurrection plant, Myrothamnus flabellifolia. Desiccation causes oxidation of the glutathione and ascorbate pools resulting in shifts of (a) the glutathione half cell reduction potential (EGSSG/2GSH) and (b) the ascorbate half cell reduction potential (EDHA/AsA) towards more positive values. For EGSSG/2GSH a zone between −180 to −160 mV has been defined in which cell death processes are initiated, depicted as the grey region in (a) and (c). A corresponding zone for EDHA/AsA has yet to be determined. In contrast to EGSSG/2GSH, EDHA/AsA is not concentration dependent and DHA can be irreversibly oxidized easily and broken down as opposed to the more stable GSSG. For example, the last data point in (b) suggests that the DHA/AsA system had recovered whereas the high AsA/DHA ratio was due to complete breakdown of DHA while very small amounts of AsA were still detected (original data in Kranner et al. 2002). Therefore, EDHA/AsA may be a less suitable marker of the intracellular redox environment than EGSSG/2GSH. Nonetheless, a clear trend of AsA oxidation during desiccation can be recognized in (b). The open circles represent leaves desiccated for up to four months and closed circles represent leaves desiccated for eight months. (c) Upon rehydration, leaves desiccated for eight months (solid symbols) showed a further increase in EGSSG/2GSH and (d) EDHA/AsA towards more positive values, indicating that the desiccation-induced damage was irreversible, in agreement with the fact that the plants abscised their leaves upon rehydration. By contrast, leaves desiccated for four months were able to recover (open symbols), and this coincided with a decline in EDHA/AsA and maintenance of the EGSSG/2GSH below the threshold at which cell death is predicted to occur. Data for EGSSG/2GSH and EDHA/AsA have been recalculated using data in Kranner et al. (2002); (a) and (c) are modified after Kranner and Birtic (2005). EDHA/AsA was calculated using the Nernst equation (note that DHA is not a squared term) using an \( {\text{E}}^{{0^{\prime } }}_{{{\text{DHA}}/{\text{AsA}}}} \) of +60 mV

Changes in EGSSG/2GSH together with changes of other redox couples in response to desiccation will have an impact on the intracellular redox environment. This will have a downstream effect on redox-sensitive molecules. Under normal non-stressed conditions the intracellular redox environment of the cytosol is reducing and most proteins will contain free Cys residues. In contrast, the oxidising environment of the endoplasmic reticulum (ER) promotes disulphide bond formation and protein folding. When exposed to oxidative stress, such as during desiccation, the redox environment of the cytosol becomes more oxidising due to shifts in the half-cell reduction potentials of redox couples such as GSSG/2GSH towards more positive values. This can result in the formation of disulphide bonds between redox-sensitive protein thiols, thereby altering the activity and conformation of proteins. Under strongly oxidizing conditions, protein thiols can also be irreversibly oxidized, contributing to protein denaturation. Such irreversible oxidative damage can be partly prevented by targeted disulphide formation in proteins, for example by the formation of PSSG and through TRXs. Such alteration of activity and conformation of proteins through the oxidation of Cys residues can provide a means of translating a desiccation signal into a physiological response.

Irreversible oxidation of protein thiol groups

As discussed above, protein Cys residues are highly vulnerable to oxidation and can be irreversibly oxidized. The first step of oxidation leads to formation of sulphenic acid (Cys-SOH). This step is reversible, but subsequent oxidation to sulphinic (Cys-SO2H) and sulphonic acid (Cys-SO3H) is not, which ultimately leads to protein degradation (Meyer and Rüdiger 2005). The one known exception to this is the 2-Cys Prx specific reduction of sulphinic acid catalysed by sulphiredoxin (Woo et al. 2005). Cys residues are present in the active sites of many proteins and in protein motifs that function in protein regulation and trafficking, cellular signalling and control of gene expression. Cys residues can form intra- and intermolecular disulphide bridges, which alter protein conformation and regulate activity (Dalle-Donne et al. 2007). The pKa of most Cys thiols is greater than 8, so they are completely protonated at physiological pH, which means that they are unlikely to react with ROS or reactive nitrogen species (RNS). However, specific Cys residues in redox-sensitive proteins exist as thiolate anions due to lowering of pKa values through charge interactions with nearby polar or basic amino acids. Thiolate anions are more readily oxidised to sulphenic acid, which facilitates formation of disulphide bonds with other Cys residues or with GSH (Dalle-Donne et al. 2007; Dalle-Donne et al. 2009; Paulsen and Carroll 2010).

In some organisms redox-active Cys-containing proteins have evolved into selenoproteins, where the catalytic Cys residue is replaced by selenocysteine. Most selenoproteins have close Cys-containing homologues, a feature used by Fomenko et al. (2007) to identify catalytic redox-active Cys residues through screening of nucleotide sequence databases for sequences that align with selenoproteins. This method specifically identified the active Cys residues since these aligned with selenocysteine. Over 10,000 unique sequences containing redox-active Cys were detected, and included all known families of oxidoreductases including TRXs, GRXs, PRXs and GPX.

S-Glutathionylation of protein thiol groups

The post-translational modification of protein Cys residues by the addition of glutathione to form mixed disulphides (PSSGs) is termed S-glutathionylation and protects vulnerable thiol groups of proteins from undergoing irreversible oxidation. In addition to GSH (and its homologues), mixed disulphides can form between protein thiols and cysteine, cystine and homocysteine (hCys). Cysteine/cystine is the most abundant extracellular redox buffer, where it is predominantly present in the disulphide form. Extracellular proteins therefore tend to be S-cysteinylated whereas intracellular proteins are S-glutathionylated because GSH/GSSG is the major intracellular redox buffer (Dalle-Donne et al. 2007). As discussed above, only certain Cys residues are redox-active and susceptible to oxidation or S-glutathionylation. In addition, the accessibility of the thiol group within the protein structure determines the susceptibility of the Cys residue to S-glutathionylation. Together, thiol accessibility and reactivity confer specificity to protein S-glutathionylation (Dalle-Donne et al. 2007; Dalle-Donne et al. 2009; Paulsen and Carroll 2010).

Mechanisms of S-glutathionylation

Several mechanisms of S-glutathionylation have been proposed: (i) direct interaction between GSH and a partially oxidised thiol group e.g. thiyl radical, sulphenic acid or protein S-nitrosothiol; (ii) reaction between PSH and oxidised forms of GSH e.g. S-nitrosoglutathione (GSNO), glutathione sulphenic acid (GSOH) and glutathione disulphide S-monoxide [GS(O)SG]; (iii) thiol-disulphide exchange between PSH and GSSG (Kranner and Grill 1996; Schafer and Buettner 2001; Dalle-Donne et al. 2009). Thiol-disulphide exchange requires extreme changes in the GSH to GSSG ratio under conditions of severe oxidative stress such as occurs during desiccation. The majority of PSSGs are transient intermediates in protein disulphide bond formation, but some persist, and the number of constitutive PSSGs can increase by 20 to 50% under conditions of oxidative stress. This may represent a means of storing GSH which would otherwise be exported from the cell as GSSG, as well as a protective mechanism against irreversible oxidation of protein thiols (Kranner and Grill 1996; Dalle-Donne et al. 2007). S-glutathionylation fulfils several of the criteria for consideration as a mechanism of redox regulation e.g. it is reversible, specific for certain Cys residues of particular proteins, it occurs in response to physiological stimuli to elicit an appropriate response, and protein targets of S-glutathionylation are modified in activity and cellular function (Dalle-Donne et al. 2009).

S-glutathionylation in stressed plants

Proteomic studies have identified potential targets of S-glutathionylation. For example, Dixon et al. (2005) treated Arabidopsis cell cultures with biotinylated GSSG prior to exposure to tert-butylhydroperoxide to induce oxidative stress, labelled proteins were then purified using streptavidin-agarose, separated by 2D-PAGE and identified through peptide mass fingerprinting. This approach identified 22 proteins that underwent S-glutathionylation in vivo, including sucrose synthase, α- and β-tubulin, acetyl-CoA carboxylase, actin, glyceraldehyde-3-phosphate dehydrogenase, transducin and a heat shock protein. In vitro studies found a greater number of S-glutathionylation targets, many of which were metabolic enzymes; others included GST, 2-Cys PRX and GRX.

The involvement of decreases in the GSH/GSSG ratio in promoting disulphide bond formation was demonstrated in a study by Cumming et al. (2004) who also showed that H2O2 affected disulphide bonding in a dose-dependent manner. Many of the proteins identified as undergoing disulphide bonding in the cytosol had been shown in other studies to be targets of S-glutathionylation. Therefore, it was proposed that S-glutathionylation protects redox-sensitive proteins during low levels of oxidative stress but more severe stress leads to disulphide bond formation (Cumming et al. 2004).

Examples for protein S-glutathionylation in seeds and desiccation tolerant plants

Butt and Ohlrogge (1991) reported on S-glutathionylation of an acyl carrier protein in seeds of Spinacia oleracea during seed maturation in parallel with a decrease in the GSH/GSSG ratio and a decline in GR activity. In mature dry seeds half of the total acyl carrier protein was S-glutathionylated, and this was rapidly reversed during imbibition so that within 24 h only the reduced form of acyl carrier protein was detected. Mixed disulphide formation was also observed during desiccation of wheat grains, along with a decrease in the GSH/GSSG ratio and GR activity (Rhazi et al. 2003). Rehydration of wheat embryos resulted in rapid reduction of PSSG within the first 10 mins of imbibition (Fahey et al. 1980).

In agreement with the above studies, seed maturation in Triticum durum also coincided with a decline in monobromobimane-labelled protein thiol groups, which is indicative of disulphide bridge formation and S-glutathionylation involving GSSG reduction and incorporation of GSSG into mixed disulphides, respectively (de Gara et al. 2003). Dehydroascorbate reductase activity was found to increase during seed maturation. In addition to DHAR, DHA-reducing activity is displayed by a number of enzymes including protein disulphide isomerase (PDI), which is involved in storage protein maturation. Therefore it was suggested that the high DHA reducing capacity could have been due to PDI catalysis of disulphide bridge formation during protein-folding. During seed development storage proteins are synthesised on the rough ER and transported into the ER lumen where PDI-catalysed protein folding takes place. The formation of disulphide bridges stabilises the proteins before they are transported into protein storage vacuoles. The redox environment in the ER lumen is considerably more oxidising than in the cytosol, which facilitates disulphide bridge formation. In addition, protein disulphide bond formation may lead to H2O2 formation within the ER and contribute to programmed cell death of the endosperm during seed desiccation (Onda et al. 2009).

Protein thiol content also declined in orthodox Acerplatanoides seeds during desiccation, but was unchanged in recalcitrant Acer pseudoplatanus seeds (Pukacka and Ratajczak 2007). This suggests that in orthodox seeds protein thiol-disulphide conversion was a targeted response to desiccation. Likewise, accumulation of PSSG during desiccation and reduction upon rehydration was also found in the resurrection plant Boea hydroscopica (Navari-Izzo et al. 2000), and earlier in the DT fungus Neurospora crassa (Fahey et al. 1975), a type of red bread mould. Taken together, these studies suggest that protein glutathionylation accompanies the desiccation of DT seeds, plants and fungi.

Reduction of mixed disulphides and protein disulphide bonds by glutaredoxins and thioredoxins

Thioredoxins

The reversibility of protein thiol modifications relies on TRXs and GRXs. Thioredoxins are small (12–14 kDa) oxidoreductases that catalyse the reduction of disulphide bridges in proteins. They represent a large multigene family with over 40 genes of TRX and TRX-like sequences identified in Populus trichocarpa and Arabidopsis thaliana. Six types of TRX have been characterised, f-, m-, x-, y-type TRXs are found in chloroplasts; h-type TRXs are present in the cytosol and mitochondria; and o-type TRXs are found in mitochondria (Jacquot et al. 2009). Proteins of the TRX superfamily contain a highly conserved CysXXCys motif in their active site, where XX is usually Gly or Pro, and one of the Cys residues can be substituted by Thr or Ser.

There is increasing evidence for the involvement of TRXhs in oxidative stress responses, with control at the transcriptional level due to the presence of antioxidant responsive cis-elements in several TRXh gene promoters (Dos Santos and Rey 2006). In rice a cis-element involved in response to methyl viologen was identified in the promoters of TRXh, GRX and SOD, which indicates that these genes may be induced in a co-ordinated response to high ROS levels. Sequence analysis revealed that eleven other rice genes contained the cis-element in their promoter region, but the motif was not detected in the Arabidopsis genome, so does not appear to be conserved in all plant species (Tsukamoto et al. 2005). The expression of TRXh8 and TRXh5 is induced by biotic and abiotic stress, and expression of chloroplastic TRXs is light-responsive. A number of TRX targets have been identified, which include TRX-dependent reductases and TRX-regulated enzymes, many of which are involved in oxidative stress response e.g. APX, CAT, DHAR, GPX, GST, methionine sulphoxide reductase A and B (MSRA and MSRB), 1-Cys PRX, 2-Cys PRX, PRX Q, PRX II, SOD, aldehyde dehydrogenase, germin and secretory heme-peroxidase (Dos Santos and Rey 2006).

The type-h TRXs are abundant in cereal seeds during development and germination and play a role in the mobilisation of seed storage reserves by reducing storage proteins, α-amylase inhibitors and activating a seed-specific serine protease, thiocalsin. A study into the expression of TRXh1, TRXh2 and TRXh3 in wheat seeds showed that expression of all three was low during the early stages of seed development, but increased during maturation when the seed WC decreased to below 40%. TRXh expression in response to salt and oxidative stress was only affected in the aleurone layer, where induction may help to protect against oxidative stress associated with lipid catabolism in the oleosomes stored in aleurone cells (Cazalis et al. 2006). Another study of TRXh proteins in developing wheat seeds showed that TRXh was present only in the maternal tissues of the nucellar projection during the early stages of seed development. However, as development progressed TRXh accumulated in the scutellum and aleurone layer and during maturation drying was localised exclusively in these tissues. In contrast to the endosperm, the embryo and aleurone layer remain viable. ROS were generated in the embryo and aleurone cells during maturation drying, and at the same time the proportion of reduced TRXh increased, indicating that desiccation triggers TRXh reduction that may play a role in protecting against oxidative stress (Serrato and Cejudo 2003).

Alkhalfioui et al. (2007) examined the redox state of proteins during germination of Medicago truncatula seeds, and showed that in dry seeds few proteins were reduced. Proteins were progressively reduced as germination proceeded, especially during the later stages. Many of these proteins were also shown to be reduced by TRX. The TRX targets were proteins involved in a range of functions, demonstrating that redox regulation is central to a wide diversity of processes. This study also highlighted the scale and importance of thiol-disulphide conversion in the dry state for maintaining proteins in a stable form that can be rapidly recovered upon rehydration. TRX reduction occurs via TRX reductases (TRs) which differ in their mode of action depending on their subcellular location. In the chloroplast there are two TRs; the first is ferredoxin:thioredoxin reductase (FTR), which is a 20–25 kDa heterodimer with an Fe–S centre and a redox-active disulphide which is reduced by ferredoxin (FDX). The second is NADPH-dependent TRX reductase (NTRC), a fusion between a TR module and a TRX module with a flavin adenine dinucleotide (FAD) cofactor and a redox-active disulphide at the active site. NTRC maintains 2-Cys PRX in the reduced state. Reduction of TRXs in the cytosol and mitochondria is also NADPH-dependent and relies on NTRs that are similar to NTRC. NADPH reduces FAD which in turn activates the catalytic disulphide to enable TRX reduction via a disulphide-dithiol exchange reaction (Jacquot et al. 2009).

Thioredoxins play key roles in reducing protein disulphides during seed germination (Shahpiri et al. 2008; Wong et al. 2003, 2004) and their roles in development and germination appear to be linked to the desiccation and rehydration associated with maturation and imbibition, respectively. The association of the TRX system with response to water deficit has been demonstrated in several studies. For example, many wheat proteins that were altered in expression following growth under drought conditions were identified as TRX targets (Hajheidari et al. 2007). In addition, an Arabidopsis NTRC knockout mutant was hypersensitive to drought stress as well as salt and methyl viologen treatment, suggesting that NTRC provides protection against oxidative stress (Serrato et al. 2004). Likewise, increased sensitivity to drought and methyl viologen treatment was observed in transgenic Solanum tuberosum plants lacking a TRX known as chloroplastic drought-induced stress protein (CDSP32). This correlated with over-oxidation of a 2-Cys PRX that was previously identified as a CDSP32 target. Therefore it was proposed that the increased sensitivity to oxidative stress in the CDSP32-lacking plants was due to impaired H2O2 detoxification as a result of decreased capacity to regenerate 2-Cys PRX (Broin and Rey 2003). PRXs are regulated by thiol-disulphide exchange using TRX as an electron donor, and have been associated with desiccation tolerance. For example, 1-Cys PRXs are reportedly only found in DT tissue of seeds and in some resurrection plants, and proteomic studies have revealed that PRXs are desiccation-responsive proteins in the DT moss Physcomitrella patens (Wang et al. 2009).

Glutaredoxins

Glutaredoxins also belong to the thiol-disulphide oxidoreductase superfamily. There are estimated to be around 30 GRX isoforms in higher plants, which are divided into three groups, the CPYC-type, the CGFS-type and the CC-type. Some GRXs catalyse de-glutathionylation via a dithiol mechanism whilst others use a monothiol mechanism. GRXs may also catalyse protein glutathionylation especially in the presence of the glutathiyl radical, but it is likely that de-glutathionylation is the major role in vivo. GRXs are reduced by GSH in the presence of GR and NADPH, although some GRXs of the CGFS-type may be reduced by TRs (Rouhier et al. 2008). GRX activity is significantly decreased under conditions of oxidative stress due to its requirement for GSH for reduction. In addition, under oxidising conditions the Cys residues of GRX themselves undergo S-glutathionylation or form a stable intramolecular disulphide bond, which protects the enzyme from oxidative damage (Peltoniemi et al. 2006). In plants, GRX targets identified so far include GSH-dependent DHAR and a type II PRX, both of which are involved in ROS detoxification (Rouhier et al. 2008). Loss of activity of a chloroplastic GRX in an Arabidopsis knockout mutant led to increased carbonylation of proteins in the chloroplasts. Carbonylation is an irreversible oxidative protein modification, so GRX appears to play a role in protecting proteins against oxidation (Cheng et al. 2006).

Protein modification and redox signalling

ROS were once considered to be toxic by-products of photosynthesis and aerobic metabolism, and the only ‘positive’ role ascribed to them was that of host defence, especially in mammalian systems. More recently, ROS have gained acceptance as key players in cellular signalling pathways, participating in processes including stomatal closure, root growth, programmed cell death and the hypersensitive response (Apel and Hirt 2004; Mittler 2004). In plants ROS accumulation has been observed in response to biotic and abiotic stresses, as a result of production catalysed by enzymes such as NADPH oxidases and POX or through leakage of electrons from respiratory and photosynthetic electron transport chains. Signalling cascades rely on proteins such as kinases and phosphatases, hence a ROS signal must be perceived and transmitted by proteins. The most likely way in which ROS can interact with proteins is through thiol modification of Cys residues, which can be oxidised to varying degrees to cause changes in protein conformation and activity (Hancock et al. 2006). Schafer and Buettner (2001) proposed that EGSSG/2GSH could play an important role in activating nano-switches, defined as switches that operate on a nanometre scale such as the distance between thiol groups of Cys residues separated by two intervening amino acids, as occurs in TRXs. These nano-switches or ‘sulphur switches’ act as redox sensors in response to changes in the intracellular redox environment to control and integrate metabolic pathways. In mammalian cells three major redox control nodes comprising TRXs, GSSG/2GSH and Cystine/2Cys have been proposed to regulate sulphur switches, and the redox control nodes are maintained independently in different subcellular compartments (Kemp et al. 2008). Using redox-sensitive probes expressed in Arabidopsis as artificial target proteins, Meyer et al. (2007) showed that changes in EGSSG/2GSH as a result of inhibition of GSH synthesis led to disulphide bridge formation, and that GRX was specifically required for reduction of the disulphide bridges. This provided evidence that EGSSG/2GSH may trigger biological responses via GRX as the mediator.

Proposed model for thiol-disulphide exchange in protecting proteins from stress-induced oxidative damage

In summary, many studies have shown that desiccation causes accumulation of GSSG at the expense of GSH. This will cause a shift of the intracellular redox environment to more oxidising conditions, favouring oxidation of thiols to sulphenic acid or generation of thiyl radicals which facilitates disulphide bond formation and S-glutathionylation. Both can modify protein activity to activate or inactivate enzymes and thereby regulate cellular processes. In addition, formation of disulphide bonds and S-glutathionylation can protect protein Cys residues from further oxidation and irreversible damage. Building upon the previously published thiol-disulphide cycle of Kranner and Grill (1996) we propose that disulphide bond formation of oxidation-prone protein Cys residues through glutathionylation and disulphide bonding is part of the response of DT organisms to desiccation (Fig. 3). In addition, during maturation drying of orthodox seeds and desiccation of DT plants, proteins undergo a thiol-disulphide transition, so that in dry tissues most proteins are in the disulphide state. This process is reversed during imbibition and TRXs may play a major role in protein reduction, which results in activation of enzymes required for germination, for example pullulanase that facilitates starch breakdown and thiocalsin, a seed-specific Ser protease, and inactivation of trypsin and α-amylase inhibitors. In addition, reduction of disulphide bridges results in the unfolding of storage proteins which increases their accessibility to proteases. On rehydration de-glutathionylation also occurs, likely catalyzed by GRXs.
Fig. 3

Simplified scheme of thiol-disulphide conversions during desiccation and rehydration. Red (bold lines) arrows denote processes that may occur predominantly during desiccation and blue (thin lines) arrows show those that primarily occur during rehydration. Desiccation is associated with an increase in ROS levels. The thiol (-SH) groups of redox-active Cys residues in proteins (P; green) and glutathione (G; yellow) are susceptible to oxidation, and in the dry state a shift towards the disulphide forms occurs as the probability for enzymatic reduction decreases with progressive water loss. The formation of disulphide bridges and mixed disulphides with glutathione protects Cys residues from further oxidation to sulphenic (PSOH), sulphinic (PSO2H) and sulphonic (PSO3H) acids. Sulphenic acid can be reduced by glutathione whereas sulphonic and sulphinic acid formation are irreversible. S-glutathionylation can help in avoiding irreversible oxidation of free Cys residues and may occur through several mechanisms, shown here is the reaction between a protein thiol (PSH) and GSSG, resulting in the formation of protein-bound glutathione (PSSG). This will occur under conditions of severe oxidative stress when GSSG accumulates such as during desiccation. Other mechanisms include reaction of protein thiyl radicals or sulphenic acid intermediates with GSH, and glutaredoxins (GRX) can also catalyse the reaction between GS and PSH to yield PSSG (not shown), although GRXs normally act as reductants in de-glutathionylation reactions. Upon rehydration the reduction of protein disulphides can be catalysed by thioredoxin (TRX), which is subsequently reduced by TRX reductases (see ‘NADPH-dependent thioredoxin system’). In the cytosol and mitochondria the NADPH-dependent thioredoxin system operates, in which TRX reductases utilise electrons supplied by NADPH, which are transferred to TRX disulphide via flavoproteins (FAD). In chloroplasts, TRX reductases may use ferredoxin as the electron donor. Moreover, during rehydration GRXs reduce PSSG leading to formation of a mixed disulphide between GSSG and GRX. This is reduced by GSH and the resulting GSSG is reduced by glutathione reductase using NADPH as an electron donor (see ‘NADPH-dependent glutaredoxin system’). The active site of GRXs can either have one (monothiol) or two (dithiol) catalytic Cys residues; only the monothiol mechanism for PSSG reduction is shown here, for simplicity

Outlook and perspectives

There is much evidence for increased thiol-disulphide conversion during desiccation and many studies have linked TRXs and GRXs with the reduction of protein disulphides and de-glutathionylation. However, direct evidence that these processes are key to conferring desiccation tolerance is yet to be provided through studies of knockout or knockdown mutants. The use of mutants affected in GSH synthesis, and TRX and GRX expression may prove useful in dissecting the role of each of these components in the desiccation response. Moreover, comparative proteomic studies between DT and DS species will likely help unravel the importance of protein redox regulation to surviving desiccation. Recently, great advances in the field of redox proteomics have been made (Dietz 2008). Several techniques were successfully employed to analyse oxidation and S-glutathionylation of protein Cys residues. Methods for quantifying free redox-active thiols often employ labelling reagents such as 5, 5-dithio-bis-(2-nitrobenzoic acid) (Pukacka and Ratajczak 2007), monobromobimane (De Gara et al. 2003), 4-dithiopyridine (Hansen et al. 2009) or iodoacetamide. These react with protein thiols and may be detected directly using spectrophotometric analysis or through the use of radiolabelled or fluorescent tags to purify labelled proteins (Ying et al. 2007). Labelled protein targets of thiol-disulphide exchange may be separated using 2D-PAGE and identified by mass spectrometry (Alkhalfioui et al. 2007; Rinalducci et al. 2008; Ying et al. 2007). Diagonal gel electrophoresis has been used to study proteins with inter- and intramolecular disulphide bonds and has seen a recent revival due to development of high-throughput mass spectrometry to enable protein identification (Cumming et al. 2004; Rinalducci et al. 2008). A similar principle has been applied to chromatographic separation of thiol-containing proteins, in a method known as combined fractional diagonal chromatography. Labelling of free thiols with isotope-coded affinity tags enables labelled proteins to be separated by affinity chromatography prior to LC–MS analysis. Analysis of S-glutathionylated proteins is more difficult; radiolabelled or biotinylated GSH has been used in combination with 2D PAGE, however drawbacks include disturbance of the intracellular GSH pool by incubation with labelled GSH and the possibility of not detecting constitutively S-glutathionylated proteins since they may not incorporate labelled GSH (Dixon et al. 2005; Giustarini et al. 2004; Rinalducci et al. 2008; Ying et al. 2007). Taken together, forward and reverse genetics in conjunction with advanced hyphenated techniques and proteomics may enable us to understand precisely how thiol-disulphide interactions contribute to the ability to revive from the desiccated state. As discussed earlier, DT plants are excellent models to study stress response, and the knowledge gained from using them may well be transferable to DS plants, enabling the breeding of more drought tolerant, or generally more stress tolerant crops.

Acknowledgments

The Royal Botanic Gardens, Kew receive grant-in-aid from Defra.

Copyright information

© Springer Science+Business Media B.V. 2010

Authors and Affiliations

  1. 1.Seed Conservation DepartmentRoyal Botanic Gardens KewArdinglyUK

Personalised recommendations