Biochemical Journal

Review article

A nuclear glutathione cycle within the cell cycle

Pedro Diaz Vivancos, Tonja Wolff, Jelena Markovic, Federico V. Pallardó, Christine H. Foyer

Abstract

The complex antioxidant network of plant and animal cells has the thiol tripeptide GSH at its centre to buffer ROS (reactive oxygen species) and facilitate cellular redox signalling which controls growth, development and defence. GSH is found in nearly every compartment of the cell, including the nucleus. Transport between the different intracellular compartments is pivotal to the regulation of cell proliferation. GSH co-localizes with nuclear DNA at the early stages of proliferation in plant and animal cells. Moreover, GSH recruitment and sequestration in the nucleus during the G1- and S-phases of the cell cycle has a profound impact on cellular redox homoeostasis and on gene expression. For example, the abundance of transcripts encoding stress and defence proteins is decreased when GSH is sequestered in the nucleus. The functions of GSHn (nuclear GSH) are considered in the present review in the context of whole-cell redox homoeostasis and signalling, as well as potential mechanisms for GSH transport into the nucleus. We also discuss the possible role of GSHn as a regulator of nuclear proteins such as histones and PARP [poly(ADP-ribose) polymerase] that control genetic and epigenetic events. In this way, a high level of GSH in the nucleus may not only have an immediate effect on gene expression patterns, but also contribute to how cells retain a memory of the cellular redox environment that is transferred through generations.

  • antioxidant
  • cell cycle
  • epigenome
  • gene expression
  • glutathione
  • nucleus
  • oxidative signalling
  • poly(ADP-ribose) polymerase (PARP)
  • redox state

INTRODUCTION

The paradigm concerning ROS (reactive oxygen species) that traditionally viewed these metabolites as potentially lethal and thus presenting an ever-present danger due to their physicochemical toxicity, has existed in redox biology for many years. More recent data challenge this view, indicating that these compounds are key determinants of redox status whose increased availability in the cell is favoured by programmed ROS production or withdrawal of antioxidant capacity [1,2]. Enhanced ROS accumulation influences many pathways of redox signalling and control [3]. ROS signalling is made possible by homoeostatic regulation that is dependent on cellular antioxidant status. Antioxidants continuously process ROS and they are crucial components of the cellular redox signalling network. It has been known for many years that the cellular redox state is a crucial regulator of the cell cycle [46]. Low levels of cellular oxidation triggered by superoxide and hydrogen peroxide activate cell signalling pathways leading to proliferation, and are required for the correct mitogenic signalling [5,7,8]. An oxidation event that occurs early in the G1-phase of the cell cycle is a critical regulatory step in the progression to S-phases [9]. Moreover, in the G1-phase cellular GSH levels are low and an increase in total GSH is subsequently necessary for the cells to progress from the G1- to the S-phase [10]. Thus the current model of cell-cycle regulation incorporates an intrinsic redox cycle, in which transient oxidations, perhaps related to bursts of ROS, serve to regulate key proteins by processes such as thiol–disulfide exchange reactions at critical cysteine residues and in this way regulate cell-cycle progression or cause an arrest in the proliferation cycle [11]. However, conditions that cause excessive or prolonged cellular oxidation arrest the cell cycle and trigger cell death. Uncontrolled oxidation which can damage DNA, leading to altered bases and damaged sugar residues, resulting in single- and double-stranded DNA breaks [12,13], is considered to be important in this response. Strand breaks trigger a DDR (DNA damage response) by inducing the expression of molecular markers associated with DNA damage repair, such as PARP [poly(ADP-ribose) polymerase], RAD51 and BRCA (breast cancer early-onset) family members [1417]. An inherent part of the DDR is the activation of checkpoints with conserved key regulators such as the ATM (ataxia telangiectasia mutated) and Rad3-related protein kinases that arrest cell-cycle progression and stimulate repair processes in order to preserve genome integrity [1820]. It should be noted, however, that while the initiation of a DDR by oxidative DNA base damage is well characterized in mammalian and yeast cells [21,22], there is little evidence that oxidative stress leads to DNA damage or DDR induction in plants [12,23,24]. Oxidative DNA modifications display a negative linear correlation with nuclear GSH [25].

Until recently, much uncertainty has remained concerning the enzymes and antioxidants that afford protection to the nucleus and prevent DNA damage. However, the characterization of the recruitment of one of the cell's major antioxidants, the thiol tripeptide GSH (Figure 1), into the nucleus in the G1- and S-phases of the cell cycle [26] provides a realistic and powerful mechanism for strategic deployment of antioxidant defence mechanisms during the cell cycle in animals and plants.

Figure 1 A simple scheme showing the pathway of glutathione synthesis and cycling between reduced (GSH) and oxidized (GSSG) forms

(1) γ-Glutamyl cysteine synthetase (glutamate–cysteine ligase); (2) glutathione synthetase; (3) glutathione reductase.

GLUTATHIONE SYNTHESIS AND INTRACELLULAR COMPARTMENTATION

Glutathione is synthesized in a two-step ATP-dependent pathway. The first reaction, which is catalysed by the enzyme γ-glutamyl cysteine synthetase (encoded by the GSH1 gene in plants) also called GCL (glutamate-cysteine ligase), produces γ-EC (γ-glutamyl cysteine). In the second step, catalysed by GSH-S (glutathione synthetase; encoded by the GSH2 gene in plants), glycine is added to form glutathione (Figure 1). Much of our understanding of the functions of GSH in plants has come from studies on mutants altered in these enzymes or from transgenic plants where the activities of GCL or GSH-S have been manipulated [27,28]. For example, in the model plant Arabidopsis thaliana knockout mutations of GSH1 that completely block glutathione synthesis are lethal [29]. This demonstrates that glutathione is essential for the plant survival, as is the case in mammalian and yeast cells.

Less severe mutations in the GSH1 gene, which result in decreased glutathione contents, have been extremely useful in elucidating the functions of glutathione in plants. For example, the rml1 (rootmeristemless1) mutant, which has less than 5% of wild-type glutathione contents, fails to develop a root apical meristem because all the cells arrest at the G1-phase of the cell cycle [28]. Intriguingly, this mutant shows a much less severe shoot than root phenotype because of redundancy between glutathione and TRXs (thioredoxins) in the control of shoot apical meristem function. The redundancy in TRX/glutathione functions was demonstrated by combining the rml1 mutation with mutations in the two genes encoding cytosolic/mitochondrial NADPH-TRX reductase (NTRA, NTRB) [30]. Other mutations in GSH1, resulting in decreases in tissue glutathione contents of between 25 and 50% that of the wild-type plants, do not lead to an altered growth phenotype. However, lower glutathione levels result in a decreased ability to withstand biotic and abiotic stresses. For example, the cad2 mutant was identified by an enhanced sensitivity to cadmium, the rax1 mutant by altered expression of a cytosolic ascorbate peroxidase gene and the pad2 mutant by decreased camalexin contents and enhanced sensitivity to pathogens [31].

In Arabidopsis leaves, the first step of GSH synthesis takes place exclusively in the chloroplasts [32]. In other plant species the situation is less clear and γ-EC production may also take place in the cytosol, as well as the chloroplasts [33]. While the second enzyme, GSH-S, is encoded by a single gene (GSH2), alternative splicing results in localization of the gene product in the chloroplast and cytosol [33,34]. While GSH1 knockouts are embryonic lethal, GSH2-knockout mutants are seedling-lethal, probably reflecting partial replacement of GSH functions by γ-EC, which accumulates to very high levels in these plants [35]. The wild-type phenotype can be restored in gsh2 mutants by complementing the mutant with targeted expression of the enzyme to the cytosol alone [35]. All of the other compartments of the cell are dependent on the import of glutathione from the cytosol.

GSH synthesis is subject to multiple levels of control, but the most important of these are considered to be cysteine availability and GCL activity [2]. It has long been recognized that cysteine availability can impose an important limitation on GSH synthesis. In general, however, up-regulation of the cysteine synthesis pathway occurs concomitantly with that of GSH synthesis [27]. The expression of the GSH1 or GSH2 genes is not modulated by many known environmental or endogenous triggers, and marked increases in the abundance of GSH1 or GSH2 transcripts have only been reported in response to jasmonic acid and in certain stress situations such as exposure to heavy metals or certain pathogens. Crucially, increases in hydrogen peroxide and other oxidants, which are known to cause accumulation of glutathione, do not alter the expression of the GSH1 or GSH2 genes. However, GCL activity has long been known to be regulated by GSH [31,32,36] (Figure 1). The GCL homodimer is linked in plants by two disulfide bonds, one of which is involved in redox regulation [36]. Although the exact mechanistic details remain to be resolved, activation through disulfide formation is considered to be important in contributing to up-regulation of GSH synthesis in response to oxidative and other stresses [36]. While it remains unclear precisely how redox regulation interacts mechanistically with the classic model of GSH feedback inhibition of GCL activity, these two regulatory processes would appear to have distinct functions from a physiological perspective. For example, feedback inhibition acts as a homoeostatic control mechanism to restrict the extent of cellular GSH accumulation and accelerate synthesis in response to depletion and, conversely, covalent thiol–disulfide regulation allows enhanced synthesis, specifically in response to increased cellular oxidation.

There is strong interplay between GSH concentrations and GSH/GSSG ratio, where an accumulation of GSSG often causes subsequent increases in total glutathione [2]. Oxidative activation of GSH synthesis can be rationalized in terms of the homoeostatic maintenance of glutathione redox potential, as this is related to [GSH]2/[GSSG] rather than to [GSH]/[GSSG]. At a constant GSH/GSSG ratio, increases in glutathione pool size entail subtle, but potentially important, decreases in redox potential.

GLUTATHIONE AND CELLULAR REDOX STATE

The redox potential of the cytosol, which is established by the major redox buffers such as ascorbate, glutathione and NADP(H), is considered to be in the order of −300 mV. If we assume that all three of these redox buffers are in equilibrium in the cytosol, then at NADP+/NADPH=1 (representing a redox potential of −320 mV), there should be very little oxidized glutathione. Thus the overall redox state of the glutathione [GSH/(GSH+2GSSG)] pool in a organ such as the leaf is considered to be between 0.9 and 0.95 assuming that GR (glutathione reductase) activity allows the glutathione and NADP(H) couples to be in redox equilibrium (i.e. at the same redox potential). However, a rather different picture emerges from the application of the in vivo probes based on GFPs (green flourescent proteins) containing oxidizable thiol groups that are able to monitor glutathione status. Such measurements suggest that cytosolic GSH/GSSG ratios are underestimated in whole-tissue measurements, and that GSH/GSSG ratios can be considerably higher than 0.95 in certain intracellular compartments that therefore have a redox potential of less than −320 mV [37]. This would confer a high sensitivity on the signalling function of glutathione redox potential mediated through GRX (glutaredoxin)-dependent changes in protein thiol–disulfide status.

Increases in GSSG relative to GSH are a useful indicator of oxidative stress or ‘disulfide stress’. However, it should be noted that an important property of cellular glutathione homoeostasis is that increased oxidation is generally accompanied or rapidly followed by increases in the total pool size.

The concentration and/or redox state of the glutathione pool in each cellular compartment are important for cellular redox homoeostasis and redox signalling [38,39]. While it remains unclear how GSH/GSSG transporters function in inter-compartmental redox regulation, the transport of GSH between the different cellular compartments is fundamental for the maintenance of cellular GSH levels and redox-based signalling pathways. Recently, a screen for Arabidopsis mutants that are insensitive to the GCL-inhibitor BSO (buthionine sulfoximine), revealed the identity of the chloroplast γ-EC and GSH transporter, called chloroquinone-like transporter [40]. However, whereas GSH transport mechanisms have been identified at the plastid membrane [40] and at the plasma membrane and tonoplast membranes [4144], no information is available on the mechanism that regulate GSH transport into the nucleus. Until recent observations suggested that much of the cellular GSH pool could be restricted to the nucleus [26], transport of GSH into the nucleus had not been thought to be important because nuclear pores were not considered to restrict diffusion of low-molecular mass solutes such as GSH.

THE FUNCTIONS OF GLUTATHIONE

Glutathione is the major non-protein cellular thiol and is present in many cellular compartments at millimolar concentrations. This high abundance imposes a low cellular thiol–disulfide redox potential on the cell and allows GSH to function as a thiol buffer, maintaining cytoplasmic thiols in the reduced state. Like other thiols, glutathione can undergo numerous redox reactions and almost all GSH functions are linked to oxidation of the cysteine group. GSH is oxidized by ROS at high rates, but it is able to function as an efficient antioxidant scavenger and ‘sacrificial’ nucleophile because it is present in the cell at relatively high concentrations. Oxidized forms notably include disulfides, either with another glutathione molecule to form GSSG or with a different thiol to form ‘mixed disulfides’, as well as more oxidized forms in which the thiol group is converted into sulfenic, sulfinic or sulfonic acids [1]. However, glutathione functions are not restricted to these compounds because of the enormous potential array of glutathione conjugates that can be formed with electrophilic species.

The cellular glutathione pool is an effective buffer or barrier against excessive oxidation protecting redox-sensitive molecules in the cell. The glutathione pool is maintained predominantly in a reduced state because of the action of GR. GR is found in many cellular compartments and has a very high affinity for the substrates GSSG and NADPH. In addition to chemical oxidation, GSH oxidation can also occur as a result of a number of enzyme-catalysed reactions that use GSH to reduce hydrogen peroxide or other peroxides to water or the corresponding alcohol. It should be noted, however, that in plants enzymes that have been annotated as glutathione peroxidases use TRXs more efficiently than glutathione. However, certain plant GSTs (glutathione transferases) can catalyse glutathione peroxidation and GSH is also involved in peroxide metabolism through GRX-dependent peroxiredoxins. In addition, GSH is required for the regeneration of reduced ascorbate through the action of DHARs [DHA (dehydroascorbate) reductases]. GSH-dependent reduction of DHA allows NADPH oxidation to be coupled to ROS removal via ascorbate and glutathione pools.

The nucleophilic nature and the chemical reactivity of its γ-cysteine thiol group make GSH particularly suitable for a broad range of functions in metabolism and signalling [1,2]. Glutathione participates in post-transcriptional protein modification through thiol–disulfide exchange and through S-glutathionylation, i.e. formation of a stable mixed disulfide between GSH and a protein thiol, and in this way GSH protects proteins of irreversible modifications induced by oxidation or nitrosylation by reactive nitrogen species. Moreover, GSNO (S-nitrosoglutathione) has a number of physiological functions, particularly as a signalling molecule or as a reservoir of NO. In addition to its antioxidant activities, glutathione is important in the detoxification of xenobiotics through the action of GSTs that catalyse the conjugation to glutathione and phytochelatins, which contribute to heavy metal tolerance. In plants GSH also participates in the regulation of sulfur metabolism with roles in the uptake, assimilation, transport and storage of reduced sulfur. As indicated above, cellular glutathione status has long been implicated in relaying defence signals in plants [1,2,38]. For example, GSH is an important mediator of the salicylate-dependent suppression of jasmonate signalling in plants that occurs upon pathogen or insect attack [45]. In animal cells, the glutathionylation of proteins such as SERCA (sarcoplasmic/endoplasmic reticulum Ca2+-ATPase) pumps and myofibrils in muscle, and transcription factors such as NF-κB (nuclear factor κB). In the latter case, the GSH-generating agent NAC (N-acetyl-L-cysteine) has been shown to enhance hypoxic apoptosis in mouse fibroblasts by blocking the NF-κB survival pathway. Since GRXs, which specifically catalyse reduction of protein-SSG mixed disulfides, reversed the inhibition of p65-NF-κB DNA binding in extracts from hypoxic cells plus NAC and restored NF-κB activity, GRX-dependent S-glutathionylation of p65-NF-κB is considered to be the most probable mediator of NF-κB inactivation and enhanced hypoxic apoptosis [46].

GLUTATHIONE-DEPENDENT REGULATION OF CELL PROLIFERATION

A considerable amount of evidence in the literature obtained from studies on both animal and plant cells support the view that glutathione is a key regulator of cell proliferation. Early studies showed that GSH was an important factor in the control of tumour growth [47], whereas low-molecular-mass thiols were implicated in the regulation of cell proliferation [48]. In a similar manner to mammalian cells, plants cells in the G1-phase of the cell cycle were found to have a very low level of GSH [10]. Moreover, an increase in the amount of total GSH was shown to be necessary for the cells to progress from the G1- to the S-phase of the cell cycle [10]. In contrast with the positive effect of GSH, addition of GSSG caused the cell cycle to arrest at G1 [49]. As discussed above, depletion of GSH blocks the transition from G1- to S-phase in the cell cycle of the root apical meristem, but redundancy between GSH and TRX functions are observed in the development of the shoot meristem [30]. GSH is also important in the formation of nodules, which are specialized organelles where plants cells are in symbiotic association with rhizobial bacteria [50]. Although in Arabidopsis cells the GSH/GSSG ratio is high and constant throughout the cell cycle [51], proliferating human colon epithelial cells were characterized by particularly high GSH/GSSG ratios [52].

Recent studies have shown that the intracellular distribution of GSH is also a crucial factor in proliferation, particularly the distribution of GSH between the nucleus and cytosol [26]. A number of methods are available for the detection of GSH in cells. Pioneering work using monochlorobimane–GSH conjugates demonstrated the compartmentalization of GSH in the nucleus of hepatocytes [53]. Thereafter, microinjection studies revealed that GSH conjugates preferentially localize to nuclei [54]. Immunocytochemistry using specific antibodies and computer-supported transmission electron microscopy has been used in plants to quantify the relative amounts of glutathione in the different compartments of the plant cell [55,56]. Real-time imaging is generally restricted to cytosol and mitochondrial compartments, but it also provides a useful tool with which to study the participation of glutathione in cellular redox homoeostasis [57]. These in vivo probes, which are based on GFPs containing oxidizable thiol groups, were designed to ‘sense’ the glutathione redox potential by equilibration of their thiol–disulfide status with that of glutathione. To date, these redox-sensitive GFPs have largely been used to monitor glutathione status in the cytosol [57]. Confocal microscopy using CellTracker Green [CMFDA (5-chloromethyl fluorescein diacetate)] to detect GSH has proved to be a useful tool in the characterization of GSH distribution between the nucleus and cytoplasm in animal and plant (Figure 2) cells [26,51].

Figure 2 Confocal microscopy images showing GSH localization in the cytoplasm (a–c) and in the nucleus (d–f) of Arabidopsis cells

Double staining using Hoechst 33342 to localize nuclei (blue stain; a, c, d and f) and CMFDA (green stain; b, c, e and f) to track GSH, was performed on living cells at points where GSH was localized throughout the cytoplasm (ac) or in the nucleus (df). Confocal images (a) and (d) show blue Hoechst fluorescence, images (b) and (e) show distributions of green CMFDA fluorescence, and (c) and (f) show the two superimposed stains.

RECRUITMENT INTO THE NUCLEUS DURING THE CELL CYCLE

As discussed above there is a general consensus of opinion that the cell cycle includes a redox cycle involving changes in the abundance of ROS and the intracellular partitioning of GSH. Moreover, the hierarchy of redox-dependent regulatory checkpoint changes during cell proliferation in a manner that not only is crucial for correct cell-cycle progression, but also links cell-cycle progression to cell fate [58]. An oxidation event is required at G1 [9] in order to stimulate mitogenic pathways that control the activities of CDKs (cyclin-dependent kinases) and initiate the phosphorylation cascade including the phosphorylation of the pRB (retinoblastoma protein) in order to allow entry into S-phase and activate DNA replication and cell division [58,59]. Thereafter, a more general reduction of the cellular environment is required in order to enable the cells to progress to the G2/M-phases [60]. GSH recruitment into the nucleus and concomitant regulation of glutathione homoeostasis in plant [51] and mammalian [26] cells will not only have implications for the regulation of the abundance of ROS in the cytosol, but also have a profound impact on the redox state of the nucleus and the cytosol with many repercussions for processes that are subject to redox regulation in each cellular compartment. Combining information obtained from our studies on mammalian [26,61] and plant [51] cells, we propose a model for the glutathione cycle that exists within the cell cycle illustrated in Figure 3 (see also the accompanying animation at http://www.BiochemJ.org/bj/431/0169/bj4310169add.htm). In this model, intracellular GSH partitioning and homoeostasis undergoes the following pattern during the cell cycle: (i) GSH is recruited and sequestered in the nucleus early (G1) in cell proliferation, (ii) as a result of GSH sequestration in the nucleus, the cytoplasm is starved of GSH, (iii) the sharp fall in cytosolic GSH availability and the accompanying change in cytosolic redox state triggers GSH synthesis in the cytoplasm, (iv) the total GSH pool of the cells increases rapidly, (v) the nuclear envelope dissolves at the end of prophase/beginning of metaphase allowing equilibration between the cytosol and nuclear GSH pools during the G2- and M-phases, and (vi) the nuclear envelope re-forms during telophase (before cytokinesis), the cells divide and the cellular GSH pool is re-distributed between the daughter cells.

Figure 3 The glutathione cycle within the eukaryotic cell cycle

This model for intracellular GSH partitioning between nucleus and the cytoplasm and its effects on whole-cell glutathione homoeostasis during the cell cycle has been derived from our studies on mammalian and plant cells ([26,51,61] and P. Diaz Vivancos, J. Markovic, F.V. Pallardo and C.H. Foyer, unpublished work). Prior to initiation of the cell cycle (A), GSH is equally distributed between the cytoplasm and nucleus. An appropriate cell-cycle trigger (cell-cycle induction) causes changes in nuclear envelope transport properties (B) that include orchestration of GSH-transporting proteins that are either rapidly synthesised de novo or activated in order to recruit and sequester GSH in the nucleus. GSH sequestration into the nucleus has immediate repercussions for the cytoplasm, particularly organelles such as the mitochondria and chloroplasts, because the cytosol is depleted in GSH. GSH synthesis in the chloroplasts and cytosol is stimulated as a result of GSH depletion until the GSH pools in the nucleus and cytoplasm reach similar levels (C). The barrier between the nuclear and cytosolic GSH pools disappears when the total cellular GSH level is highest because the nuclear envelope dissolves at the end of prophase and the beginning of metaphase (D). During telophase the nuclear envelope reforms and the total cellular GSH pool is divided between the two newly formed daughter cells (re-distribution; E). The GSH-transporting proteins in the nuclear envelope are then either inactivated or degraded so that the GSH pools in the nucleus and cytoplasm are again in equilibrium (A). An animated version of this Figure is available at http://www.BiochemJ.org/bj/431/0169/bj4310169add.htm.

THE IMPACT OF NUCLEAR GSH SEQUESTRATION ON THE REDOX PROCESSES IN THE CYTOPLASM

A depletion of cytoplasmic GSH upon GSH recruitment into the nucleus is entirely consistent with previous observations of an oxidation event occurring early in the G1-phase of the cell cycle [9,58]. The recruitment of GSH into the nucleus early in cell proliferation will have a profound effect on the cytosolic GSH concentration without necessarily affecting the GSH/GSSG ratios. The glutathione redox potential depends not only on the GSH/GSSG ratio of the cytosol, but also on the overall GSH concentration. GSH depletion will greatly change the cytosolic glutathione redox potential of plant cells, which will rise from values that are generally accepted to be well below −300 mV. This will greatly alter signalling functions associated with glutathione redox potential-mediated events such as GRX-dependent changes in protein thiol–disulfide status, where a change of 50 mV significantly alters the relationship between oxidized and reduced forms. Two considerations are important in analysing the effects of changes in the cytosolic and nuclear GSH pools on thiol-based ROS sensors [1]. These are the thermodynamics (redox potential of oxidizable thiols) of the system and kinetics (ability to compete with the antioxidative system). In oxyR, for example, the hydrogen-peroxide-reactive thiols have a midpoint redox potential of approx. −0.18 V and a rate constant for the reaction with hydrogen peroxide that is comparable with peroxidases. These properties mean that under optimal conditions where the redox potential of the glutathione pool is approx. −0.24 V, oxyR will generally be in its reduced inactive form. However, any increase in hydrogen peroxide availability or change in glutathione redox potential (or both) can readily cause oxidative activation of the sensor [1]. Redox-dependent shifts in oxyR–DNA contacts suggest that such a mechanism might operate in cells [62]. Sensor oxidation may also be facilitated by programmed withdrawal of GSH, or it could be catalysed by specific peroxidases, as shown for the yAP-1 system in yeast [63].

A depletion of cytoplasmic GSH upon GSH recruitment into the nucleus is entirely consistent with previous observations of an oxidation event occurring early in the G1-phase of the cell cycle [9,58]. Comparisons of the plant cell transcriptome when GSH is localized in the nucleus relative to cytoplasmic GSH show that the abundance of a large number of proteins involved in oxidative defence pathways is decreased when GSH is localized within the nucleus (P. Diaz Vivancos, J. Markovic, F.V. Pallardo and C.H. Foyer, unpublished work). Thus the oxidative defence shield is lowered when GSH is localized within the nucleus. The GSH cycle within the cell cycle, illustrated in Figure 3, thus fits well with the intrinsic redox cycle model, which incorporates transient oxidations that regulate cell-cycle progression [11].

A cytoplasmic oxidation event associated with GSH recruitment into the nucleus will also have a profound effect on GCL activity, which is subject to redox regulation as well as feedback inhibition by GSH, as illustrated in Figures 1 and 4. These two regulatory processes could function together in this situation, redox regulation allowing enhanced synthesis in response to increased cellular oxidation, while feedback inhibition provides a homoeostatic control mechanism limiting the extent of cellular GSH accumulation. Thus the observed large increases in the total cellular pool of GSH [26,51] that rapidly follow GSH recruitment into the nucleus are entirely consistent with a large depletion of the cytoplasmic GSH pool. The rapid increase in total cellular GSH occurs before the nuclear envelope dissolves at the end of prophase (prometaphase) providing evidence for severe depletion of the cytoplasmic GSH pools when GSH is recruited into the nucleus. The cytoplasmic and nuclear GSH pools become in equilibrium once more once the nuclear envelope has dissolved at a point where the total cellular GSH content has greatly increased.

Figure 4 A simple model for changes in cellular GSH homoeostasis

Prior to initiation of the cell cycle (A), GSH is equally distributed between the cytoplasm and nucleus and the rate of GSH synthesis is regulated by feedback inhibition and redox regulation of γ-GCL. In plants the first step of GSH synthesis is localized in the chloroplasts. GSH accumulation in the nucleus depletes the chloroplasts and mitochondria of GSH and leads to activation of γ-GCL, leading to de novo GSH synthesis and accumulation (B). The stimulation of de novo GSH synthesis would continue until the GSH levels in the cytoplasm are sufficiently high to limit GSH synthesis by feedback inhibition and redox regulation of γ-GCL activity.

The nuclear glutathione pool may be able to resist depletion by mechanisms such as inhibition of GCL activity by BSO [64]. The addition of BSO failed to decrease the nuclear GSH pool or impair cell proliferation [64]. In contrast, non-specific agents such as DEM (diethylmaleate), which is a weak electrophile and forms DEM–GSH adducts through a reaction catalysed by GST, prevented nuclear GSH recruitment and strongly inhibited cell proliferation. These results suggest that the nuclear GSH pool can be depleted by GST-catalysed reactions. [65].

THE CELLULAR DEFENCE SHIELD IS LOW WHEN GSH IS LOCALIZED IN THE NUCLEUS

The depletion of the cytoplasmic GSH pool also has important consequences for the cellular defence capabilities. When GSH is localized in the nucleus, the oxidative defence capacity of the cell is low, as indicated first by the transcriptome profile, which shows a decreased abundance of antioxidant and defence transcripts, and secondly by the increased abundance of oxidants in the cells (P. Diaz Vivancos, J. Markovic, F.V. Pallardo and C.H. Foyer, unpublished work). The presence of a GSH cycle within the cell cycle, as illustrated in Figure 3, thus fits well with the intrinsic redox cycle model, which incorporates transient oxidations that regulate cell-cycle progression [11].

A second major consequence of a depletion of the cytoplasmic GSH pool for plant cells could be an impairment of the signalling pathways that up-regulate pathogen defences in response to triggers, such as those that induce changes in salicylic acid contents, which have a direct requirement for cytosolic GSH. Arabidopsis pad2-1 mutants that are deficient in GSH are more susceptible to insect and pathogen attack [31], whereas those lacking the chloroplast CLT transporters are heavy metal-sensitive and hypersensitive to fungal infection [39]. In the latter case, the absence of the transporters traps much of the cellular GSH pool in the plastids and, despite having total GSH levels similar to the wild-type, the mutants exhibit an altered systemic acquired resistance response with the significant decrease in the ability to induce PRs (pathogenesis resistance proteins), such as PR1 in leaves [40].

GLUTATHIONE-MEDIATED REGULATION OF NUCLEAR PROTEINS

GSH can interact with protein cysteine groups in a large number of ways from simple thiol–disulfide exchange reactions to glutathionylation that alter their properties and functions [1]. For example, GSH sequestration in the nucleus induces changes in the pattern of glutathionylation of nuclear proteins [26,64]. In addition, many proteins associated with nuclear functions are regulated by thiol–disulfide exchange mechanisms [66]. These include transcription, translation, chromatin stability, nuclear protein import and export, as well as DNA replication and repair. GSH acts as the driving force for ribonucleotide reductase activity through the action of nuclear GRXs [67]. Much remains to be understood regarding thiol-regulated redox pathways in nuclei, but it is generally accepted that GSH and TRX have non-redundant functions in the regulation of transcription, nuclear protein trafficking and DNA repair [66]. One the simplest effects resulting from the sequestration of large amounts of GSH in the nucleus in G1 is that a much more reduced nuclear environment will be created. Moreover, there will be a profound change in the redox state of the nucleus and this could serve to protect DNA and redox-sensitive nuclear proteins from oxidation, as well driving GRX-related processes, which could have a pronounced influence on the binding of transcription factors and play an essential role in the regulation of gene expression.

As discussed above, the activity of a number of key transcription factors in animals such as NF-κB, AP-1 (activator protein-1) and p53 are subject to regulation by the cellular redox environment. In plants, perhaps the best characterized redox-modulated protein that has a profound effect on gene expression is called NPR1. This protein, which is necessary for salicylic acid signalling, interacts with TGA transcription factors to elicit pathogen defence responses [68,69]. Given that it is considered that alterations as small as ±15 mV in the redox potential can result in transcription factor translocation and activation or deactivation [70,71], the pronounced change in the redox potentials of the cytoplasm and of the nucleus brought about as a result of recruitment of GSH into the nucleus will certainly have a profound effect on gene expression patterns. Nuclear GSH has been considered to act as a transcriptional regulator of NF-κB, AP-1 and p53 [72] because the process of binding of these transcription factors to DNA is activated by stimuli known to result in cytoplasmic ROS production [73]. In the case of NF-κB, cysteine residues within the DNA-binding domain must be reduced. Disruption of Nrf2 causes oxidative stress and compromises cell-cycle progression with cell arrest at the G2/M-phase [74].

GSH recruitment into the nucleus has the potential to modify a large number of key proteins in the nucleus, such as histones, telomerase and PARPs. The presence of high levels of GSH in the nucleus has already been shown to cause alterations in telomerase activities in co-ordination with changes in critical cell-cycle proteins, particularly Id2 and E2F4 [61]. The accumulation of GSH in the nucleus has the potential to control the structure of chromatin and the dynamics of chromatin condensation [75]. Various residues in the histone tails are subject to post-translational modification by processes such as methylation (lysine and arginine), acetylation (lysine), phosphoryaltion (serine and threonine) and ubiquitination (lysine), thereby changing the histone–DNA interaction which creates or blocks protein-binding sites. Different combinations of histone modifications alter the recruitment of proteins to chromatin in a manner that either represses or activates gene expression [76]. The effect of a particular histone modification on gene expression depends on its spatial distribution across a gene region termed ‘the histone-modification landscape’ and on the presence of other nearby modifications [77]. Thus post-translational modifications can be predictive of the transcriptional ability of a given region of chromatin. Moreover, the combinational nature of histone N-terminal modifications provides a ‘histone code’ that considerably extends the information of the genetic code [78]. There are many ways in which a high nuclear GSH pool could influence ‘the histone-modification landscape’, from effects on DNA or histone methylation to other histone modifications via modulation of enzymes such as the histone acetyltransferases.

Although little information is available to date on how nuclear GSH influences ‘the histone-modification landscape’ there is some evidence that PARP activity is influenced by a high level of GSH [51]. Poly(ADP-ribosyl)ation is a unique post-translational protein modification catalysed by PARPs, which tag long, linear or multiply branched poly(ADP-ribose) polymers on target proteins, using NAD+ as a substrate. Most of the known physiological acceptors are nuclear proteins, which are almost exclusively involved in the metabolism of nucleic acids, the maintenance of chromatin architecture or the modulation of cell-cycle activities [79,80]. PARP is also closely linked with cell-cycle regulation in mammals and with the control of stress tolerance in plants. Transgenic plants with down-regulated PARP activity display enhanced stress tolerance [81,82]. Of the different antioxidants present in plant cells, only GSH showed a good correlation with changes in PARP activity during cell proliferation [51]. The changes in PARP activity and associated decreases in the cellular pyridine nucleotide pools that showed correlations with increases in total glutathione preceded changes in the expression of the PARP1 and PARP2 genes [51]. Thus it is possible that GSHn (nuclear GSH) has a direct effect on PARP activity, as well as the abundance of PARP-mRNAs as illustrated in Figure 5.

Figure 5 A simple model for the effects of GSH on nuclear proteins such as PARP and on gene expression

One possible sequence of events would involve GSH transport into the nucleus as a result of changes to nuclear pore proteins such as Bcl-2 (A). Redox modification of histones, for example by glutathionylation of (B). Changes in the expression of genes such as PARP (C). Transport of PARP mRNAs (blue lines) to the cytoplasm; translation of PARP peptides (blue chain) and transport of peptides to the nucleus (D). Direct effects of GSH on nuclear proteins such as histones and PARP (E), i.e. activity, protein folding or activation etc.

MECHANISMS OF GSH MOVEMENT TO THE NUCLEUS

Molecular trafficking across the nuclear envelope is controlled by the nuclear pore complex [83]. While ions and small hydrophilic molecules, such as glutathione, are considered to move rapidly by diffusion, the nuclear envelope can maintain ion gradients and support ATP-dependent membrane potentials. Moreover, the permeability of the nuclear envelope membrane and associated transport were known to be greatly increased in proliferating cells because of alterations in the characteristics of the pores. One pore-forming protein that has been linked to changes in nuclear GSH content is Bcl-2, which is localized at the nuclear envelope [83]. It has long been known that Bcl-2 is an intracellular regulator of apoptosis and there are numerous reports concerning its functions in programmed cell death. However, the potential role of Bcl-2 in cellular proliferation emerged only recently. GSH binds to Bcl-2 in mitochondria [84] and overexpression of Bcl-2 results in GSH recruitment into the nucleus [85]. Overexpression of Bcl-2 in HeLa cells not only increased the total cellular glutathione level, but it also caused a re-distribution of cellular GSH, with accumulation in the nucleus leading to the hypothesis that the nuclear compartmentalization of GSH was facilitated by Bcl-2 [85]. Similarly, high nuclear Bcl-2 expression correlates with higher nuclear GSH levels in rat CC531 colorectal cancer cells [86]. We have found that Bcl-2 levels are higher in the nucleus of proliferating than in quiescent 3T3 fibroblasts, and this coincides with the high level of glutathione in the nucleus, as well as with the intense nuclear transport regulated by the nuclear pores (P. Diaz Vivancos, J. Markovic, F.V. Pallardo and C.H. Foyer, unpublished work). It is now generally accepted that Bcl-2 is a pore-forming protein and that it is a candidate GSH transporter [84]. The incorporation or activation of GSH transport proteins such as Bcl-2 allows the specific translocation of selected substances, particularly GSH, into the nucleus. In accordance with this view, we suggest that Bcl-2 becomes associated with the nuclear envelope membrane prior to cell division and that this binding facilitates the translocation of glutathione to the cell nucleus, as illustrated in Figure 5. Bcl-2 family proteins could thus act as ‘gate-keepers’ or as docking proteins, capable of pulling other proteins and peptides out from the cytosol.

While plant cells do not contain Bcl-2 proteins, a number of BAG (Bcl-2-associated athanogene) proteins have been described [87,88]. Of the seven BAG genes in the Arabidopsis genome, only the nuclear-localized AtBAG6 gene showed a differential expression pattern when GSH was localized in the nucleus relative to the cytoplasm (P. Diaz Vivancos, J. Markovic, F.V. Pallardo and C.H. Foyer, unpublished work).

It is also possible that GSTs are involved in nuclear compartmentalization of GSH. Although the nuclear localization of GSTs remains controversial, it is possible that at least two GSTs are present in the nucleus of mammalian cells [89,90]. For example, approx. 10% of the cytosolic GST pool that consisted mainly of the Alpha GST class, particularly GSTA1-1, was associated with the outer nuclear membrane and approx. 10% was compartmentalized in the nucleus hepatocytes [90]. To date, relatively few examples of nuclear GSTs have been identified in plants [91]. However, GSTU12 was localized in the nucleus in A. thaliana [92] and a member of the Phi class of GSTs, which act as transferases and peroxidases, was localized in the nucleus of vine leaves [93].

CONCLUSIONS AND FUTURE DIRECTIONS

Since its discovery over a century ago, GSH has been shown to fulfil many crucial functions in animal and plant cells. The finding that this essential redox metabolite is recruited into the nucleus during the cell cycle in both animal and plant cells [26,51] is perhaps one of the most important findings related to central GSH functions to date. Many aspects of the process of GSH recruitment into the nucleus remain to be discovered particularly regarding the mechanisms that facilitate GSH transport and sequestration in the nucleus, the effects of high nuclear GSH on the activities and functions of nuclear proteins, and how GSH depletion alters redox processes in the cytoplasm.

The field of redox biology has recently witnessed a dramatic reappraisal of the significance of ROS and cellular oxidation. For many years considered as only damaging agents to be suppressed or policed by the antioxidant system, ROS are now known to be important signalling molecules. The concept persists that ROS exert their principal effects through chemical toxicity that causes damage, but enhanced or even irreversible oxidation can also be considered in terms of cell signalling and marking molecules for turnover. While the signalling compared with damage opposition is largely irrelevant to the description of the basic biochemical mechanisms by which ROS oxidize cellular components, the choice between ‘damage’ and ‘signalling’ remains crucial in the evaluation of the physiological significance of redox-regulated mechanisms. Our view is that extensive evidence now supports the view that oxidative signalling has an important role in the regulation of cell proliferation, with glutathione acting as an essential regulator of the nuclear redox potentials and having multiple signalling functions. In this regard, it is important to quantify the changes in the nuclear and cytoplasmic redox potential causes by nuclear GSH accumulation, particularly in relation to protein targets. The Nernst equation could be used to calculate the magnitude of changes in redox potential or redox-sensitive GFPs could be targeted to the nucleus as well as the cytosol to analyse effects of GSH movement on the redox potential in each cellular compartment.

One of the most important future research directions in our view will be to determine how the large GSH-mediated increases in the redox state of the nucleus influence genetic and epigenetic processes. The redox regulation of nuclear proteins by GSH-dependent pathways will greatly improve our understanding of nuclear processes and also provide new insights into the control of cell fate, the inverse relationships between growth and defence in relation to stress, and also how to treat aging and diseases associated with uncontrolled oxidation.

FUNDING

Work in our laboratories was funded by the Biotechnology and Biological Sciences Research Council [grant number BB/C51508X/1A.P] and the European Union [grant number PITN-GA-2008-215174:Chloroplast Signals]. P.D.V. thanks the Fundación Séneca (Spain) for a postdoctoral research fellowship. J.M. and F.V.P. thank the Ministry of Science and Innovation-Spain [project number SAF2008-01338].

Acknowledgments

We thank Dr Eva Serna and Ms Sonia Priego (UCIM-Central Research Unit, Faculty of Medicine, University of Valencia, Spain) for their technical assistance in our laboratory.

Abbreviations: AP-1, activator protein-1; BAG, Bcl-2-associated athanogene; BSO, buthionine sulfoximine; CMFDA, 5-chloromethyl fluorescein diacetate; DDR, DNA damage response; DEM, diethylmaleate; DHA, dehydroascorbate; γ-EC, γ-glutamyl cysteine; GCL, glutamate-cysteine ligase; GFP, green fluorescent protein; GR, glutathione reductase; GRX, glutaredoxin; GSH-S, glutathione synthetase; GST, glutathione transferase; NAC, N-acetyl-L-cysteine; NF-κB, nuclear factor κB; PARP, poly(ADP-ribose) polymerase; rml1, rootmeristemless1; ROS, reactive oxygen species; TRX, thioredoxin

References

View Abstract