Function of phytochelatin synthase in catabolism of glutathione-conjugates

Authors


(fax +49 8161 715432; e-mail erwin.grill@wzw.tum.de).

Summary

Detoxification of xenobiotic compounds and heavy metals is a pivotal capacity of organisms, in which glutathione (GSH) plays an important role. In plants, electrophilic herbicides are conjugated to the thiol group of GSH, and heavy metal ions form complexes as thiolates with GSH-derived phytochelatins (PCs). In both detoxification processes of plants, phytochelatin synthase (PCS) emerges as a key player. The enzyme is activated by heavy metal ions and catalyzes PC formation from GSH by transferring glutamylcysteinyl residues (γ-EC) onto GSH. In this study with Arabidopsis, we show that PCS plays a role in the plant-specific catabolism of glutathione conjugates (GS-conjugates). In contrast to animals, breakdown of GS-conjugates in plants can be initiated by cleavage of the carboxyterminal glycine residue that leads to the generation of the corresponding γ-EC-conjugate. We used the xenobiotic bimane in order to follow GS-conjugate turnover. Functional knockout of the two PCS of Arabidopsis, AtPCS1 and AtPCS2, revealed that AtPCS1 provides a major activity responsible for conversion of the fluorescent bimane-GS-conjugate (GS-bimane) into γ-EC-bimane. AtPCS1 deficiency resulted in a γ-EC-bimane deficiency. Transfection of PCS-deficient cells with AtPCS1 recovered γ-EC-bimane levels. The level of the γ-EC-bimane conjugate was enhanced several-fold in the presence of Cd2+ ions in the wild type, but not in the PCS-deficient double mutant, consistent with a PCS-catalyzed GS-conjugate turnover. Thus AtPCS1 has two cellular functions: mediating both heavy metal tolerance and GS-conjugate degradation.

Introduction

The tripeptide glutathione (GSH) is ubiquitous in eukaryotes and generally represents the major thiol compound of the water-soluble non-protein thiol fraction in plants (Stulen and De Kok, 1993). It is involved in many basic physiological processes, including metabolism, transport and storage of reduced sulphur, redox regulation, signal transduction, conjugation of metabolites, and detoxification of xenobiotics (Meister, 1995; Xiang et al., 2001). GSH metabolism, redox regulation and the expression of stress defence genes are interlinked (Ball et al., 2004; Mullineaux and Rausch, 2005). Along with these functions, GSH serves as a precursor of heavy metal-chelating oligopeptides, the phytochelatins (PCs), which are synthesized by the action of phytochelatin synthase (PCS). PCS is a specific γ-glutamylcysteinyl (γ-EC) dipeptidyl transpeptidase (EC 2.3.2.15; Grill et al., 1989; Vatamaniuk et al., 2004) that belongs to the papain-type proteases (Rea et al., 2004; Vivares et al., 2005), and is activated by heavy metal ions (Beck et al., 2003; Oven et al., 2002). Phytochelatins have the general structure (γ-EC)n-glycine (n = 2–11) and are the major scavengers of heavy metal ions in plants. On exposure to heavy metals, plants synthesizing GSH isoforms (γ-EC)-Xaa, where Xaa presents a β-alanine, serine, glutamic acid or glutamine residue, form the corresponding PC-isoforms, (γ-EC)n-Xaa, via PCS (Loscos et al., 2006; Oven et al., 2002; Zenk, 1996). PC peptides accumulate in vacuoles and can be transported over long distances from root to shoot and vice versa (Chen et al., 2006; Gong et al., 2003).

Analysis of PCS and PCS-like proteins from plants, selected fungi, and invertebrates revealed an aminoterminal catalytic domain and a cysteine-rich carboxyterminal domain unique to eukaryotic PCS (Clemens et al., 1999; Cobbett and Goldsbrough, 2002; Ha et al., 1999; Ruotolo et al., 2004; Tsuji et al., 2005; Vatamaniuk et al., 1999). The carboxyterminal part of eukaryotic PCS is proposed to ensure protein stability, allosteric activity and broader responses to metal ions. A PCS-like protein corresponding to the aminoterminal moiety of eukaryotic PCS was found in the cyanobacterium Nostoc sp. (Harada et al., 2004; Tsuji et al., 2004). Crystal structure analysis of NsPCS from Nostoc showed a γ-EC acyl-enzyme intermediate that is stabilized in a cavity of the protein adjacent to a second putative GSH binding site (Vivares et al., 2005). However, NsPCS primarily hydrolyzes GSH into γ-EC and glycine (Harada et al., 2004; Tsuji et al., 2004). The inefficiency of NsPCS in PC biosynthesis, and the constitutive expression of plant PCS, have raised the question of another cellular function of PCS different from metal detoxification (Clemens, 2006).

In vitro analyses revealed that PCS also accepts S-acylated GSH derivatives as substrates (Beck et al., 2003; Grill et al., 1989; Oven et al., 2002; Vatamaniuk et al., 2000). Substrates with hexyl and methyl groups linked to the thiol group of GSH led to the generation of correspondingly modified PC analogues (Oven et al., 2002; Vatamaniuk et al., 2000), while substrates such as GS-bimane and -nitrophenole conjugates were converted quantitatively into the corresponding dipeptidyl-derivatives (Beck et al., 2003). The PCS-catalyzed reactions reflect the transfer of γ-EC, S-modified or not, onto different acceptors, either a peptide or a water molecule. The mode of transfer is possibly controlled by the bulkiness of the S-conjugated group (Beck et al., 2003). Despite the mechanistic insights into PCS-catalyzed reactions, a cellular function of PCS in mediating GS-conjugate turnover has so far not been elucidated.

GS-conjugates in plants comprise adducts of GSH and secondary metabolites (Alfenito et al., 1998; Marrs et al., 1995) as well as a diverse range of xenobiotics, including widely used herbicides, the chloro-S-triazine herbicides such as atrazine, and the sulphonylurea herbicide chlorimuron ethyl (Lamoureux and Rusness, 1993; Lamoureux et al., 1991, 1998). Conjugation of electrophilic xenobiotics to GSH is catalyzed by glutathione S-transferases (GSTs) and some peroxidases of plants (Edwards and Dixon, 2005). GS-conjugates are formed in the cytoplasm, then transported to the vacuole via the action of tonoplast ABC-type transporters (Martinoia et al., 2000; Rea et al., 1998). In plants, the catabolism of GS-xenobiotics is frequently initiated by removal of the glycine residue yielding the corresponding γ-EC conjugate (Lamoureux and Rusness, 1986, 1993). In animal systems, GS-conjugates are catabolized by the removal of the γ-glutamyl residue generating the corresponding cysteinylglycine conjugate (Meister, 1995). The key step in degradation is the first cleavage reaction. Similarly, the cleavage of the glycine residue from plant GS-conjugates is considered to be the key catabolic reaction (Lamoureux and Rusness, 1986, 1993). Based on in vitro analyses, a vacuolar carboxypeptidase (Steinkamp and Rennenberg, 1985; Wolf et al., 1996) as well as PCS (Beck et al., 2003) have been postulated to catalyze the initial step of the pathway.

Here we used Arabidopsis thaliana mutants to analyse the role of PCS in the metabolism of GS-conjugates. The Arabidopsis genome encodes two PCS genes. AtPCS1 has been identified as the major player in PC biosynthesis, and AtPCS1 deficiency of the cad1 mutant results in impaired heavy metal tolerance (Ha et al., 1999). The AtPCS2 gene encodes a functional PCS that is, however, expressed at a much lower level than AtPCS1. The enzyme activity is not sufficient to confer substantial metal tolerance in Arabidopsis (Cazale and Clemens, 2001; Lee and Kang, 2005). Our analysis with AtPCS1 and AtPCS2 knockouts revealed that AtPCS1 controls γ-EC-bimane levels. The abundance of the γ-EC-bimane conjugate was strongly reduced in AtPCS1-deficient, but not in AtPCS2-deficient plants and cells. Degradation of GS-bimane to γ-EC-bimane was rescued by transfection of protoplasts deficient in both PCS with AtPCS1. We conclude that AtPCS1 of Arabidopsis has a function in the plant-specific pathway of GS-conjugate degradation.

Results

γ-glutamylcysteine is the major intermediate in glutathione-S-bimane catabolism in Arabidopsis

We used monochlorobimane (MCB) as a model xenobiotic. The conjugation of bimane to GSH via GST provides a fluorescent label that is highly selective for GSH, and allows for a convenient chromatographic analysis of GS-bimane catabolites (Newton and Fahey, 1995). Due to the low non-enzymatic rate constant for the reaction of MCB with thiol groups, in vivo conjugation of the label is specific for GSH and dependent on GST activity (Coleman et al., 1997; Meyer et al., 2001). Arabidopsis seedlings were exposed to varying concentrations of MCB, and growth inhibition was determined in order to identify the threshold level of toxicity for the reagent (Figure 1a). Inhibition of root growth required concentrations higher than 30 μm of MCB applied to seedlings exogenously for 3 days. Based on these results, a sub-toxic level of 5 μm MCB was chosen for all further analyses. In seedlings, MCB was conjugated efficiently to GSH (data not shown). The GS-bimane was catabolized to γ-EC-bimane, cysteinylglycine-bimane (CG-bimane), and the cysteine (C-bimane) derivatives (Figure 1b). The identities of these compounds were confirmed by mass spectrometric analysis and all derivatives were absent in non-labelled plants, which served as controls (data not shown). After an exposure time of 4 h, the γ-EC-bimane fraction represented about 20% of total GS-bimane catabolites. With further onset of incubation, the GS-bimane fraction decreased considerably, and simultaneously C-bimane accumulated consistent with a subsequent conversion of γ-EC-bimane into C-bimane, such that C-bimane levels constituted the major catabolite (86%) 16 h later.

Figure 1.

 Metabolism of bimane-GS-conjugate (GS-bimane) in Arabidopsis.
(a) Effect of varying monochlorobimane (MCB) concentrations on root growth of Arabidopsis seedlings. Four-day-old seedlings were transferred to a medium containing MCB. Root growth was determined after 3 days. Mean values and SD include 20 seedlings for each data point.
(b) Turnover of GS-bimane. Arabidopsis seedlings were exposed to 5 μm MCB for 4 h, were washed and transferred into medium without MCB. At this time point, bimane derivatives of seedlings were analysed from cell-free extracts (t = 0 h).
(c,d) Growth of a cell-suspension culture in the presence and absence of 5 μm MCB. (c) Total amount of bimane derivatives within the cells; (d) levels of catabolites from GS-bimane. The first sample was analysed 10 min after administration of MCB. Values represent means of duplicates (±SD).

At that time point, approximately 40% of the GS-bimane fraction had been catabolized. Low levels of CG-bimane were detected, corresponding to approximately one-fifth to one-eighth of the γ-EC-bimane level. A similar efficient conversion of the GS-bimane to its cysteinyl derivative occurred in cell-suspension cultures (Figure 1c,d). Growth of the cell-suspension culture was not affected by the presence of MCB (5 μm), and the sum of labelled GSH and its catabolites remained fairly constant for 6 days, indicating that no other major catabolite was generated (Figure 1c). The γ-EC-bimane level peaked 1 day after MCB challenge, while CG-bimane was only a minor fraction but increased steadily over time and equalled γ-EC-bimane after 6 days. C-bimane accumulated very quickly, and >80% of the GS-conjugate was converted into C-bimane within 3 days.

Phytochelatin synthase-deficient mutants are impaired in degradation of GS-bimane to γ-glutamylcysteine-bimane

The previous analyses confirmed the catabolism of the GS-conjugate to the cysteinyl derivate via the corresponding γ-EC intermediate. However, during the initial phases of GS-conjugate catabolism a small amount of the CG-bimane was also detectable, which increased during the onset of the experiment, indicating a second pathway of catabolism operating in Arabidopsis. In order to determine the contribution of PCS activity to the generation of γ-EC-bimane, mutants with functional knockouts of both PCS genes were isolated. The cad1 mutant expresses a truncated AtPCS1 due to a premature stop codon in the gene that results in PC deficiency (Ha et al., 1999). However, it is uncertain whether it is a complete null mutant. Arabidopsis with an insertional inactivation of PCS were identified by screening a mutant collection (Rios et al., 2002), and two lines with T-DNA integration sites at exons encoding the catalytic domain were selected for further analysis (Figure 2a). Both T-DNA integrations disrupt the integrity of the PCS catalytic domain at amino acid residues 40 and 178 of AtPCS1 and AtPCS2, respectively. The double-knockout ΔPCS was generated by crossing the single-knockout lines and selecting siblings homozygous for insertions at the two loci. RT-PCR analysis revealed the presence of AtPCS1 transcripts in wild type and ΔPCS2 but not in ΔPCS1 and ΔPCS, indicating the disruption of gene function. Similarly, AtPCS2 transcripts could be detected only in wild type and ΔPCS1 (Figure 2b).

Figure 2.

 Characterization of phytochelatin synthase knockout lines.
(a) Location of the T-DNA insertions in AtPCS1 and AtPCS2. Arrows mark positions of primers used for RT-PCR analysis.
(b) RT-PCR analysis for AtPCS1 and AtPCS2 transcripts of seedlings from wild type, ΔPCS1, ΔPCS2, and the double-knockout line ΔPCS with primers as indicated in (a).
(c) Phytochelatin formation in seedlings of wild type and the double-knockout line ΔPCS. Four-day-old seedlings were exposed to 50 μm Cd2+ for 2 days before PC analysis.

PCS activity was analysed in the knockout plants by examining their capacity to synthesize PCs. Exposure of seedlings to 50 μm Cd2+ for 2 days induced PC biosynthesis in wild-type and indistinguishably in ΔPCS2 plants, but not in ΔPCS1 (not shown) and ΔPCS seedlings (Figure 2c), supporting the functional knockout of AtPCS1.

AtPCS1-catalyzed PC formation and GS-conjugate breakdown is activated by Cd2+ (Beck et al., 2003; Oven et al., 2002). We reasoned, therefore, that the presence of Cd2+ might affect the turnover of GS-bimane to γ-EC-bimane depending on a functional PCS. Seedlings of the knockout lines and the wild-type control were challenged with MCB by vacuum infiltration for 15 min and subsequent incubation for 4 h in the presence and absence of 10 μm Cd2+. Wild type and ΔPCS2 revealed comparable γ-EC-bimane levels (7.5 and 6.3 pmol per seedling, respectively; Figure 3a). ΔPCS1 and the double knockout also contained the dipeptidyl-conjugate, albeit at markedly reduced levels of 2.2 and 2.5 pmol per seedling, corresponding to 29% and 33% of wild type, respectively. Concomitant exposure of the seedlings to MCB and Cd2+ led to a more than fourfold elevation of γ-EC-bimane levels in wild-type and ΔPCS2 plants. The induction was not observed in lines deficient in AtPCS1, which revealed a >10-fold lower γ-EC-bimane abundance compared with the wild type in the presence of Cd2+.

Figure 3.

 AtPCS1-mediated γ-EC-bimane formation.
(a) Activation of phytochelatin synthase (PCS)-catalyzed GS-bimane turnover by Cd2+. Seedlings of Arabidopsis were exposed to 5 μm monochlorobimane (MCB) in the presence (filled bars) and absence (open bars) of 10 μm Cd2+. After an incubation period of 4 h, seedlings were harvested and analysed for γ-EC-bimane levels. Values represent means ± SD from triplicates.
(b) Biosynthesis of γ-EC-bimane from MCB in Arabidopsis protoplasts depending on a functional PCS. Protoplasts from wild type (bsl00001), ΔPCS1 (○), ΔPCS2 (bsl00000), and ΔPCS (•) were incubated for 2 h in 5 μm MCB, and bimane conjugates were analysed at the intervals indicated. The total amount of bimane derivatives was set to 100% and the percentage of γ-EC-bimane is given (±SD).

The results were corroborated by a time-course experiment analysing γ-EC-bimane in protoplasts of these plants. The protoplast system provides the advantage of direct and homogeneous access of the xenobiotic to the cells, and a more synchronized catabolism can be expected. The results shown in Figure 3(b) emphasize the role of AtPCS1 in GS-bimane breakdown. At 2 h after MCB administration, there was an approximately sevenfold difference in γ-EC-bimane abundance between the AtPCS1-functional and -deficient cells. Over time, the γ-EC-bimane level of the AtPCS1-deficient lines increased while that of the AtPCS1 functional lines decreased, which resulted in only a twofold difference 32 h after the start of the experiment. The accumulation of γ-EC-bimane also occurred in protoplasts of the double mutant. The identity of the γ-EC-bimane detected in ΔPCS was confirmed by electrospray-ionization mass spectrometry (ESI–MS) analysis. The results indicate that, while AtPCS1 is the major activity responsible for γ-EC-bimane levels in Arabidopsis, an additional activity distinct from AtPCS2 also catalyzes the formation of γ-EC-bimane from GS-bimane.

Analysis of protoplasts revealed not only a strongly reduced γ-EC-bimane level, but also a still functioning breakdown of GS-bimane to C-bimane in the PCS double-knockout (Figure 4a). The C-bimane formation, however, was markedly reduced in the PCS-deficient cells (31% reduction), whereas the CG-bimane content was increased by 52% compared with the wild type. The protoplasts were challenged to bimane for 4 h to exclude impaired viability of the cells by prolonged incubation. The formation of CG-bimane is presumably dependent on the activity of γ-glutamyl transpeptidases (γ-GTs), which are selectively inactivated by the inhibitor acivicin (Benlloch et al., 2005; Griffith and Meister, 1980; Nakano et al., 2006). Acivicin is a quite potent growth inhibitor, and a 35% decrease in biomass accumulation was observed in the presence of 0.1 μm levels (Figure 4b). We used increasing levels of acivicin in an attempt to block a γ-GT-dependent pathway. In the presence of 1 μm acivicin, overall turnover of GS-bimane to the cysteinyl adduct was reduced compared with the control, while CG-bimane levels were unchanged (Figure 4a). It required concentrations above 10 μm acivicin to observe a reduction in CG-bimane content that was paralleled by a strong inhibition in cell growth and overall GS-bimane turnover. The analyses do not provide a conclusive picture of the contribution of the PCS-dependent versus γ-GT-dependent pathway, but it reveals that loss of PCS function impairs conjugate breakdown, and leads to an accumulation of the CG-catabolite being part of an alternative pathway.

Figure 4.

 GS-bimane catabolism in the presence of acivicin, an inhibitor of γ-glutamyl transpeptidase.
(a) Levels of the bimane derivatives γ-EC-bimane, CG-bimane and C-bimane are presented as fractions of the sum including non-metabolized GS-bimane. Protoplasts from phytochelatin synthase (PCS)-deficient (ΔPCS) and wild-type (WT) plants were pre-incubated for 30 min (±acivicin) before monochlorobimane (MCB) administration (5 μm). Protoplasts (1 × 105) were analysed for GS-bimane breakdown 4 h after MCB exposure. Levels of total bimane derivatives were comparable (0.15 nmol μg−1 chlorophyll a±20%).
(b) Growth inhibition of Arabidopsis cell-suspension cultures in the presence of acivicin. The gain in biomass (FW) was determined after 2 days’ incubation (±acivicin). All values represent means with SD from triplicates.

Generation of the γ-glutamylcysteine conjugate is restored by expression of AtPCS1 in phytochelatin synthase-deficient cells

In order to demonstrate unequivocally the in vivo role of AtPCS1 in the breakdown of the GS-bimane conjugate, we attempted to complement the deficiency of dysfunctional AtPCS1 lines by transient AtPCS1 expression. The previous analysis clearly revealed a lower level of γ-EC-conjugate in the mutant compared with wild-type cells in the first hours after MCB challenge. Hence we chose a short period for phenotypic expression of the AtPCS1 cDNA under the control of a strong constitutive promoter. Protoplasts of knockout lines and wild type were analysed for γ-EC-bimane formation 2 h after exposure to MCB (Figure 5). The analysis of GS-conjugate breakdown revealed a strong increase of γ-EC-bimane levels in ΔPCS knockout protoplasts transfected with the AtPCS1 construct. The γ-EC-bimane levels increased by a factor of >10 in the ΔPCS1 and the double mutant (Figure 5b), while in wild type and ΔPCS2 there was a minor increase by overexpression of AtPCS1 (factor of 1.5 and 1.3, respectively). The data clearly indicate the functional complementation of the knockout line deficient in γ-EC-conjugate formation by AtPCS1.

Figure 5.

 Complementation analysis of phytochelatin synthase (PCS)-deficient cells.
(a) Determination of the γ-EC-bimane conjugate in transiently transformed protoplasts. Protoplasts derived from wild-type (WT) and ΔPCS plants were transfected with vector DNA that provides constitutive expression of AtPCS1 (WT + PCS1, ΔPCS + PCS1). Protoplasts transformed with an empty expression cassette (WT, ΔPCS) served as controls. After 2 h phenotypic expression, protoplasts were exposed to MCB (5 μm) for 2 h and subsequently the γ-EC-bimane levels were analysed. The γ-EC-bimane fractions of the HPLC chromatograms are indicated by arrows.
(b) Increase of the γ-EC-bimane level in protoplasts of WT, ΔPCS1, ΔPCS2, and the double-knockout line ΔPCS transformed with AtPCS1. The increase is expressed relative to the γ-EC-bimane level of protoplasts transformed with the empty expression cassette. Amounts of γ-EC-bimane were normalized either to the chlorophyll a content (filled bars) or to the total level of bimane-derivatives (open bars). Means ± SD of three independent experiments are given.

Discussion

Glutathione is considered to be essential for normal plant development. Among multiple functions of glutathione, the conjugation to endogenous and exogenous electrophilic organic compounds is central to a wide range of metabolic pathways (Marrs, 1996). GS-conjugates of endogenous metabolites are involved in biosynthesis, transport and storage. Anthocyanins, for example, are among the known endogenous substrates of glutathionation in higher plants, which are conjugated to GSH within the cytoplasm and are subsequently transported to the vacuole via the action of a tonoplast-bound GS-X pump requiring Mg2+-ATP (Rea et al., 1998). Many herbicide families, including sulphonylureas, aryloxyphenoxypropionates, triazinone sulphoxides and thiocarbamates, are susceptible to GSH conjugation (van Eerd et al., 2003). Conjugation of xenobiotics, such as the widely used herbicides atrazine and the thiocarbamate S-ethyl dipropylthiocarbamate (EPTC), initiates a detoxification pathway that generally leads to intracellular transport, secretion or compartmentalization of the biotransformed compounds. The xenobiotic conjugates with GSH are frequently metabolized via its γ-EC intermediate to the corresponding S-cysteine conjugate, which can either persist or be subjected to further catabolism (Lamoureux and Rusness, 1993).

In this study, we used MCB as a substrate for glutathionation, to assess the role of PCS in the catabolism of this xenobiotic. Previous biochemical characterization of Arabidopsis AtPCS1 has revealed the capacity of the dipeptidyl transferase to break down a number of GS-conjugates, including GS-bimane, to the corresponding γ-EC adduct (Beck et al., 2003). The bimane-conjugates offer the advantage of high fluorescence in contrast to the reagent, and of convenient separation of the various catabolites (Fahey et al., 1980; Fricker et al., 2000). MCB is readily taken up by plant cells and is conjugated to GSH (Fricker and Meyer, 2001; Hartmann et al., 2003; Meyer and Fricker, 2002). The fluorescent signal is present primarily in the cytosol and the nuclear compartment before it accumulates in the vacuole. In Arabidopsis, we show that the dominant catabolite of GS-bimane is the cysteinyl adduct. Turnover of GS-bimane was most efficient in cell-suspension cultures in which more than half of the GS-conjugate was transformed into C-bimane within 24 h (Figure 1). C-bimane represented about 90% of all breakdown products identified. We cannot exclude the existence of minor fractions of additional catabolites. However, the sum of identified GS-bimane and catabolites remained fairly constant at approximately 60% of the amount of administered reagent throughout the experiment with cell cultures (7 days), indicating that there is no major additional sink for GS-catabolites. The recovery rate is well in agreement with other extraction efficiencies from such cultures (Grill et al., 1991), supporting the notion of a quantitative conjugation of MCB to GSH and its analysed catabolites. In addition to C-bimane, the catabolites γ-EC- and CG-bimane were present. Immediately after MCB administration, the levels of both derivatives in cell cultures were low and increased over time, while GS-bimane declined concomitantly and C-bimane accumulated. The occurrence of both catabolites is consistent with their generation from GS-bimane by hydrolysis of either the carboxyterminal or aminoterminal amino acid residue, respectively. The γ-EC derivative is by far the more prominent intermediate shortly after administration of the xenobiotic, which is in agreement with other analyses of GS-herbicide catabolism in plants (Riechers et al., 1996).

The first and rate-limiting step in GS-conjugate metabolism in mammals is removal of the aminoterminal γ-glutamic acid residue by the action of specific glutamyltransferases (Meister, 1995), while in plants the breakdown of GS-conjugated herbicides is frequently initiated by removal of the carboxyterminal glycine residue (Lamoureux and Rusness, 1993). The latter pathway was also evident in our MCB feeding experiments, and we now show that in Arabidopsis AtPCS1 catalyzes the plant-specific formation of the γ-EC-derivative (Figure 6).

Figure 6.

 Pathways of GS-conjugate catabolism.
Electrophilic compounds (R) are conjugated to GSH (γ-Glu-Cys-Gly) by glutathione-S-transferases (GST). There are two routes of GS-conjugate (GS-R) degradation. The plant-specific pathway yields the corresponding glutamylcysteinyl-conjugate (γ-Glu-Cys-R) and involves PCS, while a second pathway of plants generates the cysteinylglycine adduct (R-Cys-Gly) and possibly involves γ-GT. Both routes lead to the cysteinyl conjugate (Cys-R) via subsequent dipeptide hydrolysis. Inset, bulky S-residues of glutathione (GSH) converted to γ-Glu-Cys-conjugates by phytochelatin synthase (PCS; Beck et al., 2003): from left top to bottom: -benzyl, -nitrophenyl, -phenylbenzyl, -uracil, -bimane, -acetamido-fluorescein.

Analysis of knockout lines of AtPCS1 and AtPCS2, as well as the corresponding double mutant, revealed a clear deficiency both in PC biosynthesis and in γ-EC-bimane generation in the absence of a functional AtPCS1 gene. The AtPCS2 mutant was indistinguishable from the wild type, consistent with previous reports that expression of this gene is negligible compared with that of AtPCS1 (Cazale and Clemens, 2001; Lee and Kang, 2005). Analysis of the double mutant did not show any contribution of AtPCS2 to either PC biosynthesis (Chen et al., 2006) or GS-bimane turnover. Functional inactivation of PCS in our analysis was indicated by the T-DNA insertion in exons encoding the catalytic domain of the AtPCS1 and AtPCS2, respectively, and by the failure to detect functional cDNA transcripts of these genes in the corresponding homozygous mutant. The complementation of γ-EC-bimane deficiency in the double knockout by transient expression of AtPCS1 further supported the role of PCS in GS-conjugate catabolism.

The in vitro analysis of PCS (Beck et al., 2003) demonstrated unequivocally that the enzyme catalyzes the breakdown of GS-bimane and other GS-conjugates such as GS-benzyl, GS-nitrobenzyl, GS-phenylbenzyl, GS-uracil and GS-acetamido-fluorescein into the corresponding γ-EC-conjugates. All these different conjugates contain a bulky S-substitution. The analysis now clearly demonstrates a cellular function of AtPCS1 in GS-bimane turnover, leading to the γ-EC-conjugate generation.

While PCS provides the major activity responsible for the conversion of GS-bimane to γ-EC-bimane, the double-knockout lines still had residual levels of γ-EC-bimane. This finding, and the existence of an alternative pathway, may explain why we have thus far failed to detect an enhanced sensitivity to xenobiotics in the PCS double mutant. In addition to PCS, other catabolic activities are capable of hydrolyzing the glycine residue from the GS-conjugate. In vitro studies implicated a vacuolar carboxypeptidase of barley in catalyzing such a reaction (Steinkamp and Rennenberg, 1985; Wolf et al., 1996). Due to the subsequent conversion of γ-EC conjugates to the cysteinyl adduct and the existence of an alternative, CG-conjugate-dependent pathway (Figure 6), it is difficult to assess the contribution of AtPCS1 and the possible carboxypeptidase activity to overall fluxes in GS-bimane turnover. The AtPCS1-dependent accumulation of γ-EC-bimane in the initial phase of xenobiotic metabolism points to a prominent role of PCS in that pathway. Furthermore, loss of AtPCS1 function impaired conjugate breakdown and led to an enhanced level of CG-bimane catabolite (Figure 4a). The increased formation of CG-bimane is consistent with a redirection of GS-conjugates into a parallel pathway that probably involves γ-GTs (Meister, 1995; Nakano et al., 2006). There are four annotated γ-GTs in the Arabidopsis genome (Orsomando et al., 2005) that might act in the pathway. Unfortunately, the general toxicity of the γ-GT inhibitor acivicin prevented a more conclusive picture of the fluxes of GS-bimane catabolites in the different turnover pathways.

In conclusion, AtPCS1 is involved in two different cellular functions, mediating heavy metal tolerance and GS-conjugate degradation. Both functions are probably executed in the cytosolic compartment. Blocking the vacuolar transport by azide led to an accumulation of γ-EC-bimane consistent with a cytoplasmic action of PCS (unpublished results). The catabolic function reflects the hydrolysis of the γ-EC-moiety, rather than transpeptidation as required for PC formation. It is interesting to note that NsPCS, the PCS-like protein of the cyanobacterium Nostoc, hydrolyzes GSH to γ-EC rather than catalyzing the biosynthesis of PC peptides (Harada et al., 2004; Tsuji et al., 2004). NsPCS lacks the carboxyterminal domain of plant PCS. This finding indicates that γ-EC formation is probably the more ancient function of PCS in phylogenetic terms, and that the carboxyterminal domain was acquired later in evolution in plants, selected fungi and a few invertebrates (such as Caenorhabditis) to provide regulatory capacity and to contribute to heavy metal tolerance. This domain contains a putative second cysteine acylation site that may bind the acceptor for dipeptidyl transpeptidation required for PC formation (Clemens, 2006). Alternatively, plant PCS may act as a dimer, and γ-EC transpeptidation for PC biosynthesis involves transfer from one monomer to the other. The crystallographic analysis of NsPCS revealed a dimeric structure comprising a cleft that allows for accommodating a bulky side chain at the cysteinyl residue (Vivares et al., 2005). The PC formation from GS-conjugates with small side chains (Oven et al., 2002; Vatamaniuk et al., 2000), but without bulky S-residues (Beck et al., 2003), may reflect steric interference with the transpeptidation step, resulting only in GS-conjugate catabolism that is transfer onto a water molecule, rather than transfer onto an acceptor peptide.

The breakdown of GS-conjugates appears not to be just a side reaction catalyzed by PCS, but possibly evolutionarily more ancient than PC biosynthesis. GS-conjugates are formed not only from xenobiotics, including widely used herbicides and pesticides, but also from endogenous compounds. Future analyses need to elucidate further the role of PCS in the metabolism of GS-conjugates, including endogenous adducts of, for example, cinnamic acid and anthocyanins (Alfenito et al., 1998).

Experimental procedures

Plant material and chemicals

Plants of Arabidopsis thaliana Heynh. ecotype Columbia Col-0 were grown in pots on a perlite/soil mixture at 23°C under long-day conditions with 16 h light (250 μE m−2 sec−1). The plants were used for crossing, protoplast preparation and DNA/RNA extraction. Seedlings were raised under sterile conditions on agar containing nutrient B5 solution (Gamborg et al., 1968). Cell-suspension cultures of Arabidopsis were maintained as reported for cultures of Silene cucubalus (Beck et al., 2003). All chemicals and standards used were of analytical grade or the highest purity available (Fluka; Sigma-Aldrich http://www.sigmaaldrich.com/; Merck http://www.merck.com). Acivicin (LKT Laboratories http://www.lktlabs.com) was dissolved in sterile de-ionized water to yield a 10 mm stock solution. Monochlorobimane (Calbiochem http://www.emdbiosciences.com) was dissolved in 100% acetonitrile to give a 100 mm stock solution. No inhibitory effect of acetonitrile was observed on growth of roots and cell-suspension cultures at levels used for the experiments (up to 0.2%).

Exposure to monochlorobimane, cadmium nitrate and acivicin

Arabidopsis seedlings were transferred onto agar-solidified nutrient B5 solution containing various concentrations of MCB, 4 days after germination. Root growth was determined within 3 days after transfer. For analysis of bimane conjugates, 5-day-old seedlings were transferred to B5 medium supplemented with 5 μm MCB and vacuum-infiltrated for 15 min. Subsequently the seedlings were incubated for 4 h on a gyratory shaker (100 rpm). After incubation, seedlings were washed in B5 medium, frozen in liquid nitrogen, and ground in Eppendorf tubes using a glass pestle. Bimane derivatives were extracted from 50 mg plant tissue by adding 0.5 ml extraction buffer (8 mm glycine, 10 mm EDTA, pH 3). After ultrasonic homogenization (Sonopuls; Bandelin http://www.bandelin.com) with two pulses of 30 sec (90% intensity) and subsequent centrifugation (16 000 g, 5 min), the supernatant was transferred into a new reaction tube and incubated at 70°C (10 min). The samples were stored in ice for 10 min, and denatured proteins were removed by centrifugation (16 000 g, 10 min). Aliquots of the supernatant were analysed for bimane derivatives.

Seven-day-old seedlings were exposed to 50 μm Cd(NO3)2 for 2 days and analysed for phytochelatins (Grill et al., 1987). Cell-suspension cultures of Arabidopsis were exposed to MCB (5 μm) 1 day after transfer into fresh LS medium (Linsmaier and Skoog, 1965) supplemented with 0.1 mg l−1 kinetin and 0.5 mg l−1 naphthaleneacetic acid. After various exposure periods, cells were harvested by filtration. Aliquots of plant material (0.1 g FW) were used for determination of dry weight or snap-frozen in liquid nitrogen for analysis of bimane derivatives within 2 days.

Acivicin toxicity was assessed using 1-week-old Arabidopsis cell-suspension cultures diluted 1:10 with fresh LS medium and allowed to recover for 24 h. Aliquots (5 ml) were filled in 25-ml Erlenmeyer flasks and the biomass (FW) determined at the start and after 2 days’ incubation with or without acivicin. For investigation of GS-bimane turnover in the presence of acivicin, 1 × 105 protoplasts (0.5 ml) were incubated with various concentrations of acivicin for 30 min before MCB administration (5 μm) for 4 h on a gyratory shaker (50 rpm, 23°C). The protoplasts were sedimented for 3 min at 500 g and bimane derivatives were subsequently extracted for HPLC analysis as described above.

Analysis of glutathione-conjugate catabolites and phytochelatin biosynthesis

The GS-bimane and catabolites thereof were separated by HPLC analysis (Summit ASI-100; Dionex, Idstein, Germany), essentially as reported (Beck et al., 2003; Newton and Fahey, 1995). The GSH and PC levels were analysed by online derivatization of sulphhydryl groups with Ellman reagent (Grill et al., 1987). To confirm the identity of the bimane conjugates, the fractions were purified by HPLC, dried over a stream of nitrogen, and subjected to ESI–MS analysis (Beck et al., 2003).

Arabidopsis T-DNA insertion lines

Knockout plants of Arabidopsis (accession Col-0) were obtained by screening a collection of T-DNA lines with the T-DNA-specific primers FISH1 and FISH2 (Rios et al., 2002) and the PCS-specific primers AtPCS1-for 5′-TTTATATCGGCGATCTCTTCCTTCTCCTCC-3′; AtPCS1-rev 5′-GATTCATCAAACCACCTCCAAGGCC-3′; AtPCS2-for 5′-AATCTTCAATGAAGCGCTTCAGAAAG-3′; AtPCS2-rev 5′-TGATTTACATCCTCTGTTCTTCGAATCTC-3′.

Five T-DNA insertion lines were recovered: a single ΔPCS1 and four ΔPCS2 (ΔPCS2-1 to ΔPCS2-4). The T-DNA insertion was localized by DNA sequence analysis of the amplified DNA products. ΔPCS2-1 had an insertion in an exon within the catalytic domain of PCS2, and was chosen for further analysis (subsequently named ΔPCS2). Crossing ΔPCS1 and ΔPCS2 and isolating the siblings homozygous for ΔPCS1 and ΔPCS2 generated the ΔPCS double-knockout line. The single-knockout lines were back-crossed three times to wild type (Col-0) before generation of the double mutant. Insertional inactivation of AtPCS1 (At5g44070) and AtPCS2 (At1g03980) was confirmed by RT-PCR analysis. Total RNA was extracted from 4-week-old leaves using the Aurum Total RNA Mini Kit (Bio-Rad http://www.bio-rad.com/). Total RNA (600 ng) was used for synthesis of cDNA according to the supplier's instructions (First Strand cDNA Synthesis Kit; Fermentas http://www.fermentas.com). Aliquots of the cDNA were subsequently tested for the presence of AtPCS1 and AtPCS2 transcripts as well as actin transcripts, using the primers AtPCS1a-for 5′-GGAAGCCATGGACAGTATTG-3′; AtPCS1a-rev 5′-TTCTCCTCTGCGCTGAGATT-3′; AtPCS2a-for 5′-AATGTTGCGAGCCGCTTGAA-3′; AtPCS2-rev (see above); Actin-for 5′-TGGGATGACATGGAGAAGAT-3′; Actin-rev 5′-ATACCAATCATAGATGGCTGG-3′.

Protoplast analysis

Isolation of protoplasts from rosette leaves of 3-week-old Arabidopsis plants and polyethyleneglycol-mediated protoplast transfection were performed as described previously (Himmelbach et al., 2002). For transfection experiments, approximately 5 × 104 protoplasts (0.1 ml) were transfected with 20 μg DNA of the expression plasmid 35S::AtPCS1. 20 μg of the empty vector pBI221 (Jefferson et al., 1987) served as a control, and 10 μg of a constitutive reporter 35S::LUC (Himmelbach et al., 2002) was used to normalize for transfection efficiency.

The plasmid 35S::AtPCS1 is a derivative of the pBI221 vector in which the AtPCS1 gene replaces the glucuronidase gene. AtPCS1 was amplified using the primers 5′-TATCGGATCCATGGCTATGGCGAGTTTA-3′ (for) and 5′-TATCGAGCTCCTAATAGGCAGGAGCAGC-3′ (rev) and cloned under control of the cauliflower mosaic virus 35S promoter via the unique BamHI and SacI restriction sites. After transfection, the samples were kept for 2 h on a gyratory shaker (50 rpm, 23°C) for phenotypic expression prior to MCB (5 μm) exposure or mock treatment. After further 2 h incubation, protoplasts were processed for HPLC analysis as described above.

Acknowledgements

The financial support of the Deutsche Forschungsgemeinschaft and the Fonds der Chemischen Industrie is gratefully acknowledged. We thank Dr Csaba Konzc, MPI Cologne, for providing the T-DNA collection and the Arabidopsis cell-suspension culture as well as for helping in identifying the knockout lines. We are grateful to Johanna Berger and Sylvia Große for technical assistance and to Dr Farhah Assaad for critical reading of the manuscript. T. Letzel would like to thank Agilent (Waldbronn, Germany) for providing the LC-quadrupole-MS.

Ancillary