Journal of Experimental Botany, Vol. 53, No. 372, pp. 1283-1304,
May 15, 2002
© 2002 Oxford University Press
Original Papers |
Interactions between biosynthesis, compartmentation and transport in the control of glutathione homeostasis and signalling
1 Institut de Biotechnologie des Plantes, Bât 630 Université Paris VII, 91405 Orsay Cedex, France
2 Instituto de Fitopatologia y Fisiologia Vegetal, CNIA-INTA, Cno. 60 Cuadras Km 51/2, X5020ICA Cordova, Argentina
3 Crop Performance and Improvement, IACR-Rothamsted, Harpenden, Herts AL5 2JQ, UK
Received 10 July 2001; Accepted 10 December 2001
 |
Abstract |
|---|
 Top Abstract IntroductionBiosynthesis of glutathioneIntracellular compartmentation...Glutathione homeostasis: the...Glutathione and signallingConclusionsReferences |
|---|
Glutathione has numerous roles in cellular defence and in sulphur metabolism. These functions depend or impact on the concentration and/or redox state of leaf glutathione pools. Effective function requires homeostatic control of concentration and redox state, with departures from homeostasis acting as signals that trigger adaptive responses. Intercellular and intracellular glutathione pools are linked by transport across membranes. It is shown that glutathione can cross the chloroplast envelope at rates similar to the speed of biosynthesis. Control of glutathione concentration and redox state is therefore due to a complex interplay between biosynthesis, utilization, degradation, oxidation/reduction, and transport. All these factors must be considered in order to evaluate the significance of glutathione as a signalling component during developement, abiotic stress, or pathogen attack.
Key words:
Chloroplasts, 
-glutamylcysteine synthetase, glutathione, maize, plant–pathogen interaction, transporter, Triticum aestivum, wheat, Zea mays.
|
Introduction |
|---|
|
TopAbstractIntroductionBiosynthesis of glutathioneIntracellular compartmentation...Glutathione homeostasis: the...Glutathione and signallingConclusionsReferences |
|---|
Glutathione (
-Glu-Cys-Gly) is a multifunctional metabolite in plants (Fig. 1
). It is a major reservoir of non-protein reduced sulphur, and has crucial functions in cellular defence and protection. Glutathione reacts chemically with a range of active oxygen species (AOS), while enzyme-catalysed reactions link GSH to the detoxification of H2O2 in the ascorbate–glutathione cycle. Importantly, GSH protects proteins against the denaturation that is caused by oxidation of protein thiol groups during stress. All these functions involve the oxidation of the thiol group, principally to form glutathione disulphide (GSSG). Cellular GSH:GSSG ratios are maintained by glutathione reductase (GR), a homodimeric flavoprotein that uses NADPH to reduce GSSG to two GSH. Like all other aerobic organisms, plants maintain cytoplasmic thiols in the reduced (–SH) state because of the low thiol-disulphide redox potential imposed by millimolar amounts of glutathione, which acts, therefore, as a thiol buffer. Although transient disulphide bonds do occur during the catalytic cycle of some enzymes, stable protein disulphide bonds are relatively rare except in quiescent tissues such as seeds. The multiple roles of GSH within the cell, together with the stability of GSSG, may make this redox couple ideally suited to information transduction. The GSH/GSSG ratio is likely to be far more influential in the control of gene expression and protein function than the absolute size of the glutathione pool. This review specifically addresses the key factors involved in glutathione homeostasis (synthesis, sinks, partitioning, and transport) that are central to the putative roles of this compound in signalling.
|
In addition to its antioxidant functions, glutathione is a precursor of phytochelatins and a substrate for glutathione-S-transferase (GST: Fig. 1
). Over and above its function in the ascorbate–glutathione cycle, glutathione acts as a direct electron donor to peroxides in reactions catalysed by glutathione peroxidase (GPX). The animal GPX is a selenoprotein that detoxifies H2O2 at high rates. By contrast, plant GPXs are not constitutive but are induced in response to stress. They do not contain selenium and only catalyse GSH-dependent reduction of H2O2 at rates which are very low compared with the high rates of H2O2 generation in plants (Foyer and Noctor, 2000). In plants, ascorbate peroxidase (APX) and catalase (CAT) are predominant in the detoxification of H2O2 while GPXs are more important in other areas of oxidant metabolism, including the removal of lipid and alkyl peroxides (Eshdat et al., 1997). The only clear demonstration of GPX targeting thus far has shown direction to the chloroplast (Mullineaux et al., 1998). In addition to their role in conjugation, GSTs can use GSH to reduce peroxides (Cummins et al., 1999). Transgenic tobacco lines overexpressing plant GST/GPX were reported to show enhanced antioxidant capacity and substantial improvement in seed germination and seedling growth under stress (Roxas et al., 1997). GSTs form a large, heterogeneous family of proteins that share the defining characteristic of catalysing the nucleophilic attack of the sulphur atom of GSH (or homologue) on the electrophilic centre of their substrates. They are therefore responsible for the removal of compounds that are potentially genotoxic or cytotoxic by virtue of their reaction with electrophilic sites in DNA, RNA and proteins. It has become clear, however, that the function of GSTs is not limited to these reactions: GSTs also seem to be involved in a ‘ligandin’ function important, for example, in the anthocyanin synthesis pathway (Marrs, 1996). It may be that certain GSTs function as flavonoid-binding proteins as suggested recently for AN9, a GST required for efficient anthocyanin export from the cytosol in petunia (Mueller et al. 2000). Consistent with this is the observation that the anthocyanin content of Arabidopsis leaves correlated with GSH content in plants with modified capacity for GSH biosynthesis (Xiang et al., 2001).
While it has long been accepted that glutathione is essential for vigour, it has only recently been recognized that this tripeptide cannot be functionally replaced, except perhaps by one of its homologues. The rml1 mutant of Arabidopsis, which is deficient in -ECS and contains no detectable glutathione, has a marked phenotype with an absence of root development and a small shoot system, and can survive only in tissue culture supplied with GSH (May et al., 1998a). Similarly, transgenic Arabidopsis with less than 5% wild-type leaf glutathione contents were shown to be significantly decreased in size and biomass, and were more sensitive to environmental stress (Xiang et al., 2001). A strong correlation has also been demonstrated beween root GSH content and the capacity of the cells in the root apical meristem to proliferate (May et al., 1998a). However, although a 96% reduction of shoot glutathione contents was associated with shorter roots in transformed Arabidopsis, the decrease in root length was only of the order of 40% (Xiang et al., 2001).
All these recent developments underline the importance of the control of glutathione concentration and redox state in plant cells. This review discusses the many advances made over the last decade in the understanding of glutathione biosynthesis, and also begins to address the key issue of compartmentation and related transport processes. These processes, complex and poorly defined, are discussed here in the light of new data obtained in the authors’ laboratory.
|
Biosynthesis of glutathione |
|---|
|
TopAbstractIntroductionBiosynthesis of glutathioneIntracellular compartmentation...Glutathione homeostasis: the...Glutathione and signallingConclusionsReferences |
|---|
Enzymes and genes
The pathway of glutathione biosynthesis is well established and is similar in plants, animals and micro-organisms. In two ATP-dependent steps, catalysed by
-glutamylcysteine synthetase (-ECS) and glutathione synthetase (GSH-S), the constituent amino acids are linked to form the complete tripeptide (Fig. 1). The N-terminal peptide bond linking glutamic acid to cysteine in GSH is unusual in that glutamic acid is linked via the - rather than the 
-carboxyl group. The low activities of -ECS and GSH-S in plants, and the complexities of the procedures for enzyme extraction and assay, have precluded extensive purification and kinetic characterization. Consequently, much of the current knowledge of their structure, regulation and function has been gleaned from molecular techniques and plant transformation. It is clear that the two-step reaction sequence occurs in both chloroplastic and non-chloroplastic compartments and is found in photosynthetic and non-photosynthetic tissues (Foyer and Noctor, 2001). A gene encoding -ECS, here denoted as gsh1, was originally cloned from Arabidopsis thaliana by complementation of an E. coli mutant deficient in this enzyme (May and Leaver, 1994). Heterologous expression of the Arabidopsis -ECS in a yeast mutant recovered only 10% of the GSH measured in the wild-type yeast (May and Leaver, 1994). This discrepancy provoked much speculation concerning the identity of the cloned gene, but further complementation studies have now confirmed that this gene does indeed encode a protein with true -ECS activity (May et al., 1998a).
Functional complementation of an E. coli mutant deficient in GSH-S activity was also used to clone the Arabidopsis thaliana gene for this enzyme, which is denoted here as gsh2 (Rawlins et al., 1995). The ability of several plant species to make homologues of glutathione depends on the specificity of the synthetases involved. Specific legume GSH-Ss use either glycine to form GSH or ß-alanine to form homoglutathione. Recent work in Medicago truncatula suggests that separate genes encode GSH-S and homoglutathione synthetase (hGSH-S) and that the divergence in specificity has arisen by gene duplication after the evolutionary divergence of the Leguminosae (Frendo et al., 1999). The two genes are very homologous and are found on the same fragment of genomic DNA. In a consideration of the distribution of the biosynthetic enzymes in legume nodules, Becana et al. have suggested that -ECS is plastidic, hGSH-S is cytosolic and GSH-S isoforms exist in both the cytosol and mitochondria in several legume species (Becana et al., 2000).
Regulation of biosynthesis
Glutathione synthesis is controlled primarily by -ECS activity and cysteine availability: As in animals, the activity of -ECS limits the rate of glutathione synthesis in plants under most conditions. Consistent with this notion is the observation that the cad-2 Arabidopsis mutant, which has a mutation in the gsh1 gene, has only one-third of the tissue glutathione contents of the wild-type (Cobbett et al., 1998). Antisense suppression of -ECS in Arabidopsis also causes substantial decreases in leaf glutathione (Xiang et al., 2001). A key role for -ECS in controlling the rate of glutathione synthesis is supported by the increases in extractable enzyme activity in tissues treated with cadmium (Rüegsegger and Brunold, 1992). Most tellingly, overexpression of an E. coli -ECS but not GSH-S, in poplar or tobacco substantially increases leaf glutathione contents (Strohm et al., 1995; Noctor et al., 1996, 1998; Creissen et al., 1999), as does homologous overexpression of the Arabidopsis -ECS (Xiang et al., 2001).
Bacterial genes encoding -ECS and GSH-S have been introduced into poplar, mustard and tobacco (Strohm et al., 1995; Foyer et al., 1995; Noctor et al., 1996, 1998; Zhu et al., 1999; Pilon-Smits et al., 1999; Creissen et al., 1999). Over-expression, with targeting of the bacterial enzyme protein to either the chloroplast or cytosol, led to marked increases in enzyme activity. Increases in -ECS, but not GSH-S, not only led to constitutive increases in leaf glutathione (up to 400%: Noctor et al., 1996, 1998; Creissen et al., 1999) but glutathione was also increased in xylem sap, phloem exudates and roots (Herschbach et al., 2000). Of particular note is the observation that the cysteine pool was not depleted by the increased demand for thiols, but was even slightly enhanced in response to increased -ECS activities, pointing to co-ordinate regulation of cysteine synthesis and glutathione synthesis. This observation is supported by leaf thiol contents in plants homologously overexpressing the plant -ECS (Xiang et al., 2001). Despite enhanced GSH contents in the phloem of poplar, sulphur uptake by the roots was markedly enhanced to meet the requirements of increased demand for sulphur (Herschbach et al., 2000). Nevertheless, incubation of leaf discs with cysteine increased glutathione contents substantially in untransformed and transformed poplars, particularly in the light, suggesting that cysteine supply remains a key limitation (Strohm et al., 1995; Noctor et al., 1996, 1997).
The dipeptide produced by the -ECS reaction is present at very low levels in most untransformed plants. In the poplars overexpressing -ECS, however, -EC was greatly increased. In some conditions, this was attributable to insufficient availability of Gly (Noctor et al., 1997). Even when Gly was abundant, however, -EC was still much higher than in untransformed plants, reflecting a shift in control from -ECS to GSH-S, whether the bacterial -ECS was present in the cytosol or chloroplast (Noctor et al., 1998). This suggested that overexpression of both enzymes together would increase the potential for constitutive enhancement of tissue glutathione contents even further than that achieved by -ECS overexpression alone. This effect was observed when tobacco lines overexpressing each of the biosynthetic enzymes from E. coli were crossed to produce hybrids over-producing both enzymes although, suprisingly, -EC contents were found to be higher in the hybrid lines than in those lines overexpressing -ECS alone (Creissen et al., 1999). The marked phenotype produced by chloroplastic -ECS overexpression in tobacco complicates the interpretation of these results. By contrast, a phenotype linked to chloroplastic -ECS overexpression was not observed in poplar (Noctor et al., 1998) or Brassica juncea (Zhu et al., 1999; Pilon-Smits et al., 1999), except that these transformed plants were more, rather than less, stress tolerant. Similar results have been reported in transformed Arabidopsis overexpressing the endogenous -ECS (Xiang et al., 2001). The above evidence demonstrates that the most important factors controlling plant glutathione are the activity of -ECS and the availability of cysteine, and recent work suggests that these two factors may be co-ordinated (H Rennenberg, personal communication). The in vivo activity of -ECS is determined by control at multiple levels, and these are discussed in the following sections.
Control of transcription and translation of -ECS and GSH-S:
Studies in animals, particularly on cancer cells challenged with chemotherapeutic agents, have shown that transcription of the -ECS gene is regulated by protein factors and by conserved antioxidant response elements upstream of the coding sequence (Foyer and Noctor, 2001). Relatively little is known about the co-ordinate regulation of expression of gsh1 and gsh2 in plants, but it is clear that GSH and GSSG per se exercise little or no control over transcription (Xiang and Oliver, 1998). Similarly, H2O2 did not affect transcript abundance. The abundance of gsh1 and gsh2 transcripts was increased by cadmium in Brassica juncea (Schäfer et al., 1998) and by both cadmium and copper in Arabidopsis (Xiang and Oliver, 1998). Jasmonic acid (JA) also increased gsh1 and gsh2 transcripts and a common signal transduction pathway may be involved (Xiang and Oliver, 1998). Interestingly, JA is involved in the control of glucosinolate synthesis, activation of which can represent a significant increase in sulphur demand in the Brassiceae (Doughty et al., 1995). Although transcript abundance was increased by heavy metals and JA, oxidative stress was required for the translation of the transcripts, implicating regulation at the post-transcriptional level and a possible role for factors such as H2O2 or modified GSH/GSSG ratios in de-repressing translation of the existing mRNA (Xiang and Oliver, 1998). It is interesting to note that among the most spectacular increases in glutathione are those observed when plants deficient in CAT are placed in conditions favouring photorespiration (Smith et al., 1984; Willekens et al., 1997), where the accumulation in total glutathione is accompanied by a marked decrease in the reduction state of the pool. A similar response was elicited by exposing poplar leaves to ozone (Sen Gupta et al., 1991). The 5'untranslated region (5'UTR) of the gsh1 gene was found to interact with a repressor-binding protein that was released upon addition of H2O2 or changes in the GSH/GSSG ratio (Xiang and Bertrand, 2000). A redox-sensitive 5'UTR-binding complex is thus suggested to control -ECS mRNA translation in A. thaliana (Xiang and Bertrand, 2000).
Post-translational control of -ECS:
Post-translational regulation of -ECS through end-product inhibition by GSH is a crucial factor in controlling GSH concentration in animals and plants. Covalent modification may also be influential. There is some evidence to suggest that rat -ECS is regulated by protein phosphorylation (Sun et al., 1996) but this has not yet been found in studies on -ECS from plants. May et al. concluded that protein factors are involved in post-translational control of -ECS and are required for full activity (May et al., 1998b). The failure of the plant enzyme to operate ectopically was explained by the absence of such endogenous plant factors (May et al., 1998a, b). In the animal enzyme system, a smaller regulatory subunit acts to increase the catalytic potential of the larger catalytic subunit by increasing its Ki value for GSH and decreasing the Km for glutamate, thereby alleviating feedback control and allowing the enzyme to operate under in vivo conditions (Huang et al., 1993). It is nevertheless clear that even in the absence of the smaller subunit, the large catalytic subunit is capable of effective catalysis, since overexpression of this polypeptide alone yielded increased glutathione levels in transfected human cells (Mulcahy et al., 1995). The highest glutathione contents were, however, obtained by dual overexpression of both subunits (Mulcahy et al., 1995).
While protein factors have not been identified in plants, and there is as yet no evidence for control of -ECS by phosphorylation, several enzymes in plants are controlled by interactions between phosphorylation status and factors such as 14-3-3 proteins, regulatory components found in several compartments of the plant cell (DeLille et al., 2001). This interaction inactivates enzymes such as nitrate reductase, but also confers stability against proteolytic attack. It is as yet unclear whether this type of regulation might be important in glutathione synthesis in plants, but it is perhaps worth considering a potential role in the post-translational control of glutathione synthesis. Non-linearity of the specific activity of -ECS with protein concentration has been alluded to previously (Hell and Bergmann, 1990). As part of a study on the intracellular distribution of GSH metabolism in wheat (see below), -ECS activity was measured in unfractionated leaf extracts and in purified chloroplasts. In both cases, specific activity increased substantially as protein was increased (Fig. 2). At the highest amounts of protein in the assay, the activity in wheat leaf extracts was very similar to that reported previously (approximately 0.5 nmol mg-1 protein min-1: cf. Fig. 2 and Hell and Bergmann, 1990). In chloroplast extracts, the activity showed an almost linear increase with protein and, at the highest protein concentration, attained values that were more than 4-fold higher than those found in whole leaf extracts (Fig. 2). On a chlorophyll basis, maximal -ECS activity was approximately twice as high in chloroplast extracts as in whole leaf extracts. These effects are unlikely to be due to low-molecular-weight effectors since similar results were obtained whether or not extracts were desalted prior to assay.
|
-ECS from whole tobacco and parsley cells has been shown to be subject to inhibition by GSH, in a manner that is competitive with respect to glutamate (Hell and Bergmann, 1990; Schneider and Bergmann, 1995). The KiGSH was 0.27–0.45 mM at a glutamate concentration of 10–20 mM (Hell and Bergmann, 1990). Figure 3 shows that the enzyme extracted from purified wheat chloroplasts, assayed at a constant protein concentration, was also sensitive to GSH. -ECS from wheat chloroplast and whole leaf material showed a similar sensitivity to inhibition by GSH and by buthionine sulphoximine (Table 1). It would appear, therefore, that chloroplastic and extrachloroplastic isoforms possess similar regulatory properties, although this notion can only be definitely confirmed by biochemical studies of the extrachloroplastic enzyme(s). Given the likely difficulties of isolating cytosolic -ECS through classical techniques, purification through heterologous expression of a cloned cDNA offers the best approach to this question.
|
|
|
Intercellular compartmentation of glutathione synthesis
The implications of heterogeneity of cells with regard to glutathione metabolism have only recently become the focus of research effort. Apart from the pioneering studies of Rennenberg (see below), little information on this question had appeared in the literature. Given that the sensitivity of certain cells may be explained by their lack of adequate glutathione production or recycling, this question is key.
Compartmentation in C3 leaves:
Glutathione is not produced at equivalent rates by all tissues or, indeed, by all cells within a tissue. Of particular note is the high capacity for GSH biosynthesis in some types of trichomes. These specialized unicellular or multicellular structures on the epidermis can have a protective function in excreting toxic compounds such as cadmium. The trichomes found on the stem and leaf surface of various A. thaliana ecotypes, show much higher expression of enzymes involved in the synthesis of cysteine and GSH and have GSH contents 2–3 times higher than the surrounding basal and epidermal cells (Gutierrez-Alcala et al., 2000). The evidence for GSH, GSSG and GS-conjugate transport systems on the plasma membrane associated with systemic transport has recently been discussed (Foyer et al., 2001) and hence these systems will not be elaborated upon here further.
Compartmentation in maize leaves:
An extreme example of the differential intercellular partitioning of glutathione metabolism is observed in maize. In common with other plants that show ‘Kranz anatomy’ and C4 photosynthesis, maize leaves have two photosynthetic cell types whose functions are very different. The enzymes of the Benson–Calvin cycle, which are very sensitive to redox regulation, are localized in bundle sheath chloroplasts. In contrast to the mesophyll, the bundle sheath cells have very low amounts of photosystem II, ferredoxin and ferredoxin-NADP+ reductase and, therefore, a lower capacity for the photochemical production of reducing power. H2O2 was found only in the mesophyll compartment in optimal growth conditions (Doulis et al., 1997) but accumulated in the bundle sheath cells at low temperatures (Pastori et al., 2000a). Following exposure to cold stress, oxidative damage was found almost exclusively in the bundle sheath (Kingston-Smith and Foyer, 2000).
Recent evidence suggests that the sensitivity of maize leaves to chilling-induced precocious senescence is related to the ability to synthesize and regenerate GSH (Kocsy et al., 2000) and to the absence of antioxidant generation and recycling capacity in the bundle sheath cells (Kingston-Smith and Foyer, 2000). Like DHAR, GR is localized only in the leaf mesophyll cells. By contrast, other antioxidant enzymes are either restricted to the bundle sheath cells (APX and superoxide dismutase) or are found to be approximately equally distributed between the two cell types (CAT and MDHAR; Doulis et al., 1997; Pastori et al., 2000a). The exclusive localization of GR activity in the mesophyll cells may be explained by the comparative lack of reductant in the bundle sheath cells (Doulis et al., 1997). Because of their low water-splitting capacity, bundle sheath cells may not generate sufficient NADPH to support the reduction of GSSG and DHA. GSSG and DHA produced in the bundle sheath tissues must, therefore, be transported to the mesophyll tissues to be reduced.
The absence of GR from the maize bundle sheath is due to post-transcriptional regulation, since GR transcripts are found in both cell types (Pastori et al., 2000b). Cysteine is synthesized in the bundle sheath whereas GSH-S activity is located predominantly in the mesophyll cells (Burgener et al., 1998). It appears, therefore, that glutathione is synthesized in the cells where GR is present and that the bundle sheath relies on the mesophyll for both the synthesis of glutathione and the reduction of GSSG. cDNAs corresponding to maize -ECS and GSH-S mRNA have recently been isolated. The 1664 bp nucleotide sequence obtained for -ECS mRNA (EMBL Nucleotide Sequence Database Acc. No. AJ302783) consists of a 38 nt 5' untranslated region that precedes the first ATG, a 1317 nt open reading frame encoding 437 amino acids and a 309 nt 3' untranslated region. No transit peptide could be identified from the amino acid sequence and analysis with the PSORT program showed high probability of a cytosolic localization. The 1608 bp nucleotide sequence of the mRNA isolated for GSH-S (EMBL Nucleotide Sequence Database Acc. No. AJ302784) presents an open reading frame between nucleotides 278 and 1510 that encodes a peptide of 409 amino acids, with no transit peptide. Southern blot analysis indicates that the gene encoding GSH-S is present as a single copy. Screening of a BAC library from the flint inbred F2 line with probes obtained in this work confirmed this result. Although the analysis of several clones of the -ECS cDNA showed only one sequence of mRNA, the screening of the BAC library indicated that this gene is present in two copies. Northern blot analysis showed that the expression of both enzymes is higher in leaves than in roots but there are no results to date on the intracellular regulation of the expression of these genes.
|
Intracellular compartmentation of glutathione metabolism |
|---|
|
TopAbstractIntroductionBiosynthesis of glutathioneIntracellular compartmentation...Glutathione homeostasis: the...Glutathione and signallingConclusionsReferences |
|---|
Intercompartmental variations (e.g. chloroplast versus cytosol, apoplast versus cytosol) in glutathione concentration and redox state may be crucial in signalling. By the end of the 1970s, it was known that spinach chloroplasts contained high concentrations of glutathione (Foyer and Halliwell, 1976
) and that photoautotrophic tobacco cells exported glutathione into the culture medium much faster than did hetrotrophically grown cells (Bergmann and Rennenberg, 1978). It was subsequently shown that GSH is translocated from source leaves in the phloem (see references in Herschbach et al., 2000). As discussed above, -ECS and GSH-S activities were shown to be located both inside and outside the chloroplast (Klapheck et al., 1987; Hell and Bergmann, 1988, 1990) and the ability of photosynthetic cells to synthesize GSH in chloroplastic and cytosolic compartments was confirmed by overexpression studies (Noctor et al., 1996, 1998; Creissen et al., 1999). Together, these observations indicate that photosynthetic cells are likely to be able to export chloroplastically-produced glutathione from the cell. Studies have been conducted of 35S-GSH uptake into tobacco cells and bean leaf protoplasts (Schneider et al., 1992; Jamaï et al., 1996), and a GS-conjugate transporter is known to operate at the tonoplast (Foyer et al., 2001), but nothing is known about glutathione transport across the chloroplast envelope. The following sections present results that have recently been obtained on compartmentation and transport, and these findings are discussed within the context of published literature data in order to evaluate the role of different processes in the control of the intracellular distribution of glutathione metabolism.
Distribution of glutathione metabolism between the chloroplast and the rest of the cell
Because of the agronomic importance of wheat, the compartmentation of glutathione metabolism was undertaken by isolating intact chloroplasts from wheat leaves. Although highly intact wheat chloroplasts can be prepared by lysis of protoplasts prepared via enzymatic digestion (Edwards et al., 1978), this method is of questionable suitability for the study of the distribution of stress-linked components because the isolation procedure itself affects AOS production and the antioxidant system (Ishii, 1987; Papadakis et al., 2001). Therefore, the development of a more direct and rapid method for the isolation of intact wheat chloroplasts was sought, using the mechanical homogenization of leaf tissue. While mechanical homogenization is easily applied to species with soft leaf tissue (notably spinach and pea), the literature contains very few reports of the successful use of this technique in the isolation of wheat chloroplasts, probably because fragments of fibrous bundle-sheath strands are released during tissue homogenization and cause rupture of the fragile chloroplasts. Nevertheless, by using young leaves, it was possible to develop a method producing an adequate yield (0.5–1 mg chlorophyll) of predominantly intact chloroplasts (Table 2). This method allowed chloroplasts to be liberated and separated from the homogenate in under 5 min and to be purified in less than 30 min. Given that all leaf antioxidative enzymes, including chloroplastic isoforms, have thus far been shown to be nuclear-encoded, this rapid isolation technique should avoid possible artefactual changes in enzyme distribution. The chloroplasts were photosynthetically competent, as judged by their high rates of electron transport (Table 2). Although unable to catalyse CO2-dependent O2 evolution at appreciable rates, the chloroplasts were competent in O2 evolution when supplied with 3-phosphoglyceric acid (Table 2). This indicates activity of the phosphate translocator and at least two enzymes of the Calvin cycle, as well as retention of nucleotides (NADP(H), adenylates). The chloroplast volume (Table 2) was within the range of that measured in other species (Heldt, 1980).
Expressed with respect to chlorophyll, NADP-GAPDH activity in the wheat chloroplasts was similar to that measurable in the unfractionated homogenate (Table 3). The latency of this activity in the chloroplast preparation was greater than 80%. While a large part of GR activity was located in the chloroplasts, the proportion of DHAR associated with the organelle was rather less (Table 3). The fraction of GSH-S found in the chloroplasts was low (Table 3). The use of chlorophyll to calculate the distribution of enzymes within the photosynthetic cell is complicated by the existence of significant numbers of non-photosynthetic cells within leaves. In an analysis of wheat of the same age as that used here, only 50% of the leaf cells were found to be mesophyll cells, with only slightly higher values found in other monocotyledonous and dicotyledonous species, including spinach and pea (Jellings and Leech, 1982). Values for the chloroplastic enzyme fraction based on chlorophyll (Table 3) must therefore be treated as lower limits for the chloroplast allocation within photosynthetic cells: true values would be higher if a significant fraction of enzyme is also present in non-photosynthetic cells.
|
Four thiols were detected in wheat leaves and chloroplasts (Table 4
). The glutathione homologue -Glu-Cys-Ser (hmGSH) was first described in 1992 and reported to exist in numerous grass species (Klapheck et al., 1992). It was not possible to detect the formation of this compound in GSH-S assays where Gly was replaced by Ser, consistent with this thiol being produced by hydroxymethylation of the Gly residue of GSH. Contents of all thiols on a chlorophyll basis were much lower in chloroplasts, only 8% of the total leaf glutathione being recovered in the purified chloroplasts (Table 4). Assuming that the vacuole occupies 80–90% of the cell volume, and that glutathione concentration is low in this compartment (Wolf et al., 1996), a leaf content of 199 nmol mg-1 chl (Table 4) yields a global concentration outside the chloroplast of 1–2 mM (chlorophyll content of the wheat leaves was around 1 mg g-1 FW). This value is in agreement with estimations of root cytosolic concentration using in situ imaging techniques (1.8–4 mM: Fricker and Meyer, 2001). The glutathione content of the isolated chloroplasts (Table 4) can be converted directly, using the measured chloroplast volume (Table 2), to a concentration of around 0.5 mM. This would suggest that the glutathione concentration in young wheat leaves is lower in the chloroplast than in other compartments such as the cytosol. Several observations speak against this notion. First, a higher chloroplast glutathione concentration (63–81 nmol mg-1 chl) was reported in spinach, representing a mean concentration of 3.5 mM (Foyer and Halliwell, 1976). Second, a higher proportion of glutathione was also found in chloroplasts isolated non-aqueously from barley (Smith et al., 1985). Here, more than 50% of leaf glutathione was recovered in the chloroplasts, although this method may lead to overestimation due to adhesion of extrachloroplastic material to the isolated chloroplasts. Using marker enzymes to correct for this artefact, a value of 35% leaf glutathione in pea chloroplasts isolated in non-aqueous media was calculated (Klapheck et al., 1987). By contrast, the same article reported that chloroplasts prepared in aqueous media from pea protoplasts retained only 5% of glutathione present in the parent protoplasts (Klapheck et al. 1987). When isolated in aqueous media, pea chloroplasts had glutathione contents of 7–22 nmol mg-1 chl, values very similar to those we found in wheat (cf. Table 3). Moreover, another study in pea found only 10% of the leaf glutathione in percoll-purified chloroplasts (Bielawski and Joy, 1986). It was concluded that pea chloroplasts lose glutathione during extraction in aqueous media, probably due to leakiness to small molecules during the isolation (Klapheck et al., 1987). Another possibility is the operation of transporters (see below). Whatever the processes responsible for low contents following aqueous isolation, it seems clear that chloroplasts from wheat, like those from pea, lose glutathione and other thiols during this method of extraction while retention of glutathione is higher in spinach chloroplasts. Interestingly, it was reported that, expressed on a chlorophyll basis, the glutathione content of pea protoplasts was similar to that of pea leaves (Klapheck et al., 1987). By contrast, mesophyll protoplasts have been purified from wheat (authors’ unpublished results) that have high photosynthetic activity (100–170 µmol O2 mg-1 chl h-1 in the presence of NaHCO3) but that contain only approximately 30% of the initial leaf glutathione content (c. 60 and 200 nmol mg-1 chl in protoplasts and leaves, respectively).
|
Table 5
presents a summary of the principal literature data on the chloroplast complement of glutathione, together with values for the four enzymes that were measured in wheat. In the leaves of dicotyledonous species, -ECS and GSH-S have been found to be more or less equally divided betwen chloroplastic and non-chloroplastic compartments, with a somewhat higher proportion of -ECS found in the chloroplast (Table 5). The effects of protein concentration on -ECS activity (Fig. 2) prevented estimating the distribution of this enzyme in wheat leaves, but it seems clear that, whatever the precise distribution, a higher proportion of this enzyme is found in the wheat chloroplast than GSH-S. In nmol chloroplast protein-1 min-1, maximal activities were 2.3 (-ECS: Fig. 2) and 0.23 (GSH-S). This is surprising considering that overexpression studies strongly suggest that the first enzyme exercizes the major limitation on synthesis of GSH in both cytosolic and chloroplastic compartments (Strohm et al., 1995; Noctor et al., 1996, 1998). However, studies of spinach chloroplasts, and of barley and pea chloroplasts isolated non-aqueously, indicate a chloroplastic GSH concentration close to 5 mM (Foyer and Halliwell, 1976; Smith et al., 1984; Klapheck et al., 1987) which would be sufficient to inhibit the chloroplastic -ECS by about 90% (Fig. 3) at 10–20 mM Glu, a likely chloroplastic concentration of this amino acid (Winter et al., 1994). It is perhaps worth noting that analysis of maize roots also found only a small fraction of GSH-S in the plastid (Table 5). The distribution of these enzymes could be dependent on developmental stage. Even though the wheat leaves used here were photosynthetically competent, they were nevertheless young. Literature data suggest, perhaps, that in more mature leaves, GSH-S is more strongly associated with the chloroplast (Table 5).
|
The data presented in Table 2
for GR are in agreement with literature studies (Table 5). In leaves, the bulk of GR activity is found in the chloroplast whereas root plastids may contain a lower proportion of the total cellular activity (Table 5). About 20% of the pea leaf activity was associated with the cytosol (Edwards et al., 1990). Lower GR activities have been reported in isolated mitochondria and peroxisomes (Edwards et al., 1990; Jimenez et al., 1997). Of the four enzymes discussed here, GR is by far the best characterized at the gene level. The first gene sequence encoding plant GR was isolated from pea, shown to encode a product with an N-terminal sequence characteristic of chloroplast-targeting sequences (Creissen et al., 1992), and later the product was found to be targeted to both chloroplasts and mitochondria (Creissen et al., 1995). Subsequently, cDNAs for other isoforms have been isolated from various species. Multiple activity bands have been identified in protein extracts from pea and spinach (Edwards et al., 1990; Foyer et al., 1991).
Less is known about the compartmentation of DHAR. The data in wheat (Table 3) give a value intermediate between those found in spinach and pea (Table 5). Because various enzymes can catalyse GSH-dependent reduction of dehydroascorbate as a ‘secondary’ reaction, the presence of a specific DHAR activity in chloroplasts has been the subject of controversy (Foyer and Mullineaux, 1998). Very recently in spinach, however, a chloroplast DHAR has been purified and a corresponding gene cloned and sequenced (Shimaoka et al., 2000). The authors reported a stromal DHAR activity of 34 µmol mg-1 chl h-1 (Shimaoka et al., 2000). Since DHAR activities from leaf tissue are typically 100–400 µmol mg-1 chl h-1, this would also represent a fairly small proportion associated with the chloroplast, although part of the high extrachloroplastic activity may be attributable to proteins other than a specific DHAR. It is interesting that GR/DHAR ratios appear to be markedly higher in the chloroplast than outside this organelle: on the basis of differential effects on the ascorbate and glutathione redox states observed in transgenic and mutant plants, it was suggested that DHAR activity in the chloroplast may be too low to couple the ascorbate and glutathione pools effectively (Noctor et al., 2000). In a recent theoretical article, the first attempt to model flux through the chloroplast Mehler-peroxidase and ascorbate–glutathione cycles is reported (Polle, 2001). It was concluded that chloroplastic DHAR is likely to be unimportant in chloroplast redox cycling, but that the chemical reduction of DHA by GSH would be sufficiently fast to allow an effective ascorbate–glutathione cycle (Polle, 2001). It is nevertheless clear that the redox states of the leaf ascorbate and glutathione pools can vary independently: decreased expression of the chloroplastic protein 2-cys peroxiredoxin led to a more oxidized ascorbate pool without effect on the highly reduced glutathione pool (Baier et al., 2000). Both thermodynamic and kinetic considerations predict that glutathione should be oxidized before ascorbate (Noctor et al., 2000; Polle, 2001), and this is indeed observed in plants deficient in CAT (see below). The factors that may allow DHA to accumulate in the presence of abundant GSH have been discussed, and include microcompartmentation within the chloroplast (Noctor et al., 2000; Polle, 2001). Polle's model is an excellent first step towards the comprehensive evaluation of oxidant processing and antioxidant cycling within the chloroplast, and presents or reinforces several important conclusions concerning the importance of non-enzymatic reduction and the likely location of ‘rate-limitations’ (Polle, 2001). Such approaches are likely to be very useful in understanding the response of chloroplast-derived stress to leaf physiology. Nevertheless it is noted that the common assumption that the chloroplast is the predominant site of AOS production in leaf cells may not be valid under many conditions (Noctor et al., 2002), and that it cannot be excluded that the chloroplast antioxidative system may sometimes be subject to oxidative loads of both chloroplastic and extra-chloroplastic origin.
|
|
Transport of glutathione across the chloroplast envelope
It has previously been reported that incubation of wheat chloroplasts with 35S-labelled GSH at 1 and 100 µM resulted in time-dependent uptake that was linear for at least 15 and 8–10 min, respectively (Noctor et al., 2000
). To characterize the uptake process, rates were measured over the first 5 min following addition of 35S-GSH, within which time uptake was linear at all concentrations between 1 µM and 1 mM (Fig. 4). Regression analysis was used to calculate rates: fitted lines gave a positive intercept on the y-axis (Fig. 4), probably indicating a rapid binding to the external surface of the chloroplast followed by a constant rate of uptake. At low concentrations (1–50 µM) the time-dependent uptake results in an accumulation of glutathione to a calculated chloroplast concentration of up to 5-fold the external concentration. Given that these external concentrations represent values at least 10-fold lower than the internal chloroplast concentration before addition of labelled GSH (c. 0.5 mM: see above), the data suggest active uptake of external glutathione. This conclusion is supported by the concentration dependence of GSH uptake (Fig. 5). Uptake was linear up to 20–30 µM GSH then showed saturation at around 100–200 µM GSH followed by a further increase in rate up to a concentration of 1 mM (Fig. 5). The results suggest that at least two systems are able to take up GSH across the chloroplast envelope, one showing half-saturation at around 30–50 µM GSH with a maximal capacity of approximately 0.6–0.8 nmol mg-1 chl min-1 and a second with lower affinity and higher capacity. Transport at 50 µM GSH was not affected by added ATP or by light (data not shown). No difference in uptake was observed if the possible oxidation of GSH was countered by the presence of NADPH and yeast GR in the external medium. Although direct uptake of the disulphide form has not been examined, the presence of GSSG at 10 and 500 µM significantly inhibited uptake of GSH at 10, 50 and 500 µM. At a constant GSH concentration of 50 µM, inhibition by GSSG was half-maximal at approximately 0.4 mM (Fig. 6). These data suggest that the two forms of glutathione can be transported by common systems, although GSH appears to be preferred. GSSG was also reported to inhibit GSH uptake by bean protoplasts (Jamaï et al., 1996), but not by tobacco cells (Schneider et al., 1992). It is conceivable that the weak effect of GSSG in wheat chloroplasts (Fig. 6) could be due to an effect on 35S-GSH concentration rather than an effect on the transport process itself. However, mixtures of GSSG and GSH are relatively stable at neutral pH so an effect of GSSG on the chemical GSH concentration is unlikely given the brevity of each incubation. It is possible, nonetheless, that thiol-disulphide exchange reactions result in conversion of 35S-GSH to 35S-GSSG, particularly when GSSG is in excess of GSH. Here, it is worth noting that Schneider reported no effect of 0.25 mM GSSG on the uptake of 35S-GSH (50 µM) into tobacco cells during a 2 h incubation at pH 6.0 (Schneider et al., 1992). Although this study's experiments were conducted at pH 7.6, which is likely to be more conducive to thiol-disulphide exchange, all solutions were prepared immediately before assay and the incubation period used (5 min) was much shorter than those employed previously in studies of glutathione uptake across the plasmalemma. Nevertheless, it remains unclear whether the effect of GSSG is mediated at the level of transport.
|
|
The capacity of the high-affinity wheat chloroplast glutathione uptake system was 40–60 nmol mg-1 chl h-1 while, at physiological GSH concentrations (1 mM), the uptake rate was about 200 nmol mg-1 chl h-1 (Fig. 5
). By applying a modified but similar method, it was not possible to detect significant uptake into wheat mesophyll protoplasts within 60 min, at either pH 7.6 or pH 5.5. This contrasts with the literature report of significant GSH uptake into broad bean protoplasts, where transport was shown to be due to a single saturable phase with Km 0.4 mM and Vmax of 2.1 nmol 107 protoplasts min-1 (Jamaï et al. 1996). This difference may be attributable to differences in methodology or to inactivation of the wheat plasmalemma translocator during protoplast isolation. It is worth comparing maximum GSH transport rates across the plasmalemma with those measured in wheat chloroplasts. A mean chl content per wheat protoplast of 11.8±2.0 pg (n=five independent protoplast preparations) was measured. Assuming a similar chl content, a rate of 2.1 nmolx107 broad bean protoplasts min-1 converts to around 1 µmol mg-1 chl h-1, though this capacity is unlikely to be reached at physiological extracellular GSH concentrations. In photoheterotrophic tobacco cells, as in wheat chloroplasts, two phases were identified in kinetic characterization of GSH uptake (Schneider et al., 1992). A high-affinity system displayed a Km of 18 µM and a capacity of 19–20 nmol GSH g-1 dry weight min-1, while at higher concentrations a second phase was observed with Km 780 µM and a capacity of c. 170 nmol GSH g-1 dry weight min-1 (Schneider et al., 1992). The physiological significance of the second system is unclear. Given a chl content in the region of