Introduction

The tripeptide glutathione (γ-Glu-Cys-Gly; GSH) is the major non-protein thiol compound in eukaryotic cells and participates in the regulation of cellular redox status. Essential for plant growth and development, this antioxidant is a key cellular redox component that functions in the regulation of gene expression and the cell cycle (Noctor et al. 2002). This is particularly important in plants during exposure to oxidative stress, when reactive oxygen species (ROS) are formed and GSH operates as a central component of the ascorbic acid–GSH cycle (Noctor and Foyer 1998). Other functions of GSH include the storage and long-distance transport of reduced sulfur (Brunold and Rennenberg 1997), conjugation with secondary plant metabolites and xenobiotics via glutathione S-transferases (Marrs 1996; Alfenito et al. 1998), and the formation of phytochelatins in response to heavy metal exposure (Cobbett 2000). Important developmental processes such as root meristem activity (Vernoux et al. 2000) and flowering (Ogawa et al. 2001) are also regulated by GSH levels.

GSH must be in its reduced form to fulfill many of its roles. This is particularly true when it acts as an antioxidant against ROS such as hydrogen peroxide and super oxide (Noctor and Foyer 1998). Thus, the enzyme glutathione reductase (GR) is essential for the functioning of GSH and is a prevalent member of the flavoprotein oxidoreductase family in both eukaryotes and prokaryotes. GR catalyzes the reaction that converts oxidized glutathione (GSSG) to reduced glutathione (GSH) using NAD(P)H as an electron donor (Meister 1988). In plants, GR plays a key role in the response to oxidative stress by maintaining the intracellular pool of GSH. Expression of GR is upregulated under stresses such as changes in salinity, drought, high light intensity, mechanical wounding, chilling, and exposure to heavy metals and herbicides (Foyer et al. 1991; Mullineaux and Creissen 1997; Apel and Hirt 2004; Romero-Puertas et al. 2006).

Iron (Fe) is an essential element in plants, and is required for cellular events such as respiration, chlorophyll biosynthesis, and photosynthetic electron transport. Fe is required for the synthesis of heme and chlorophyll. Indeed, low chlorophyll content (chlorosis) of young leaves is the most obvious visible symptom of Fe-deficiency. A number of heme and non-heme proteins involved in redox systems also require Fe. The most well-known heme proteins are cytochromes, which contain Fe as a prosthetic group. Other heme enzymes such as catalase and peroxidase also depend on Fe for their activity. In non-heme proteins such as those containing a Fe–S domain, Fe associates with the thiol group of cysteine, with inorganic sulfur in clusters, or with both, such as in ferredoxin. Isozymes of superoxide dismutases (SOD), which detoxify superoxide anions, also require Fe as a prosthetic group. The activities of catalase, peroxidase (Iturbe-Ormaetxe et al. 1995; Tewari et al. 2005), and ascorbate peroxidase (Iturbe-Ormaetxe et al. 1995; Ishikawa et al. 2003; Zaharieva et al. 2004; Tewari et al. 2005) decrease under Fe-deficient conditions, whereas the activities of different isozymes of SOD increase (Tewari et al. 2005). These enzymes play a role in scavenging ROS; thus Fe-deficiency, like Zn-deficiency (Cakmak and Marschner 1988), can trigger secondary oxidative stress.

Although Fe is abundant in soil, it is mainly present in oxidized Fe(III) compounds, which are poorly soluble in neutral to alkaline soils. Calcareous soils account for 30% of the world’s cultivated soil (Chen and Barak 1982). As such, Fe-deficiency is a worldwide problem that is responsible for serious reductions in crop yields. The molecular basis of Fe-acquisition from soil has been identified and strategies have been proposed to overcome this problem (For review see: Curie and Briat 2003; Mori 2001). Graminaceous plants secrete a family of small molecules called mugineic acid family phytosiderophores (MAs), to solubilize soil Fe (Takagi 1976). The Fe(III)–MAs complex is readily absorbed by the plant. DMA, synthesized by DMA synthase (DMAS; Bashir et al. 2006), is the first MA synthesized in the MA biosynthetic pathway and the subsequent steps differ depending on the plant species or even the cultivar (Ma et al. 1999; Mori 2001). The production and secretion of MAs markedly increases in response to Fe deficiency, and tolerance to Fe deficiency in graminaceous plants is strongly correlated with MAs secreted. Although excess Fe is known to trigger oxidative stress (Halliwell and Gutteridge 1986), little is known about the role of Fe-deficiency-induced oxidative stress and the role of antioxidants such as GSH in plant responses to Fe-deficiency. Therefore, understanding the mechanisms involved in the Fe-deficiency response is of extreme importance to create Fe-deficiency-tolerant plants and to thereby increase crop production.

The expression of GR is regulated by a variety of biotic and abiotic stresses; however, to date no report has described the expression pattern of GR in response to Fe-deficiency stress in plants. These responses are especially of interest in barley, which is one of the most Fe-deficiency-tolerant species among graminaceous crops. To investigate the regulation of GR in graminaceous plants, cytosolic and chloroplastic isoforms were cloned from barley and their expression patterns as well as enzyme activity were examined under Fe-sufficient and Fe-deficient conditions in graminaceous plants.

Materials and methods

Cloning of HvGR1 and HVGR2

A full-length cDNA library prepared from Fe-deficient barley roots (Higuchi et al. 2001) was screened to clone the full-length cDNA of barley cytosolic GR (HvGR2). The GenBank accession AB063249, having homology with OsGR2 (LOC_Os02g56850) was previously isolated in an attempt to clone DMAS. This clone is shorter than OsGR2 and consists of about 70% of the sequence of OsGR2. ORF of AB063249 was labeled with 32P, and approximately 200,000 colonies of the cDNA library were screened through colony hybridization as described previously (Higuchi et al. 2001).

Partial sequences of barley chloroplastic GR (HvGR1) were identified as unigene numbers 7437 and 37686 in the HarvEST database (Version 1.47; http://www.harvest.ucr.edu/) because of their strong homology to rice chloroplastic GR (OsGR1; LOC_Os03g06740). The HvGR1 ORF was amplified from the cDNA library using the 5′-CACCATGGCGACAACCGCGGCCCTCCC-3′ and 5′-TACTTCTGAGCGACGACCTCATC-3′ as forward and reverse primers, respectively. The amplified cDNA clone was subcloned into pENTR/D-TOPO (Invitrogen, Carlsbad, CA) and sequenced using a Thermo Sequenase Cycle Sequencing Kit (Shimadzu, Kyoto, Japan) and a DSQ-2000L DNA sequencer (Shimadzu).

The rice genome database (http://www.tigr.org/) was searched to determine the number and location of GR genes in the rice genome. The homologs of GR in different crops were identified through BLAST (http://www.ncbi.nlm.nih.gov/BLAST/), and phylogenetic tree was constructed using CLUSTAL W Neighbor-Joining method and the tree was visualized with TreeView.

Plant material

Seeds of barley (Hordeum vulgare L. cv. Ehimehadaka no. 1), wheat (Triticum aestivum L. cv. Chinese spring), maize (Zea mays cv. Alice), and rice (Oryza sativa L. cv. Nipponbare) were germinated on wet filter paper at room temperature. After germination, the seedlings were transferred to a 20-L plastic container containing a nutrient solution with the following composition: 0.7 mM K2SO4, 0.1 mM KCl, 0.1 mM KH2PO4, 2.0 mM Ca(NO3)2, 0.5 mM MgSO4, 10 M H3BO3, 0.5 μM MnSO4, 0.2 μM CuSO4, 0.5 μM ZnSO4, 0.05 μM Na2MoO4, and 0.1 mM Fe-EDTA for rice, while the nutrient solution for the maize, barley and wheat was prepared as described (Kanazawa et al. 1994). The pH of nutrient solution was adjusted daily to 5.5 with 1 M HCl and was renewed weekly. For the Fe-deficiency treatments, 2-week-old plants were transferred to a nutrient solution without Fe and grown for 2 week. Symptoms of Fe-deficiency were apparent after 2 weeks, at which point the roots and leaves were harvested, frozen in liquid nitrogen, and stored at −80°C until use.

Northern blot analysis

Total RNA was extracted from roots and shoots, and 10 μg were electrophoresed in 1.2% (w/v) agarose gels containing 0.66 M formaldehyde and transferred to Hybond-N+ membranes (Amersham, Piscataway, NJ). For GR1, the full length ORF of HvGR1 was labeled with digoxigenin (DIG) and used as a probe for Northern blot analysis. For GR2, Northern blot analysis was performed with full length ORF of HvGR2 labeled with DIG. For OsGR2, a full length ORF was cloned from a rice cDNA library with the forward and reverse primers 5′-ATGGCTAGGAAGATGCTCAAGG-3′ and 5′-CTACAAGTTTGTCTTTGGCTTGG-3′, then sequenced using a Thermo Sequenase Cycle Sequencing Kit (Shimadzu) and a DSQ-2000L DNA sequencer (Shimadzu), and labeled with DIG. Each probe was individually incubated with each membrane at 68°C and processed as described previously (Engler-Blum et al. 1993; Yoshihara et al. 2003).

Determination of GR activity

To subclone HvGR1 into pMAL-c2 (New England Biolabs, Beverly, MA), the ORF was amplified with primers such that it contained an XbaI site at the 5′ end and a HindIII site at the 3′ end. The ORF was amplified, subcloned into pCR-Blunt II-TOPO (Invitrogen, Carlsbad, CA), and sequenced. The resulting plasmid was then digested with XbaI and HindIII and the excised fragment containing HvGR1 was subcloned into pMAL-c2. This plasmid was designated pMAL-c2-HvGR1. HvGR2 and AB063249 were also subcloned into pMAL-c2 in the same manner. These pMAL-c2 plasmids were introduced into E. coli XL1-Blue, which were induced to produce the recombinant fusion proteins. The proteins were purified as described by Higuchi et al. (1999).

The GR activity was determined using 1 μg of recombinant HvGR1, HvGR2, and AB063249 bound to a maltose-binding protein with a GR assay kit (Trevigen, Gaithersburg, MD) according to the manufacturer’s instructions. To determine the effect of Fe-deficiency on total GR activity in rice, wheat, barley, and maize, the crude extract from plants grown under Fe-sufficient and Fe-deficient conditions was extracted using the buffer provided with the kit according to the manufacturer’s instructions (Trevigen). The proteins were quantified using the Bradford assay (Bradford 1976) and diluted to 1 μg μl−1 with dilution buffer. Each assay was performed at least in duplicate with 40 μg of total protein.

Results

Cloning of HvGR1 and HvGR2

The GenBank cDNA clone AB063249 is highly homologous to rice and wheat GR. We screened the full-length cDNA library using AB063249 as a probe and isolated two cDNA clones: one corresponding to AB063249 and one for cytosolic GR (HvGR2). The HvGR2 contained the full length of AB063249 with some variation at the 5′ UTR region. The chloroplast GR (HvGR1) was cloned with primers designed to amplify the HvGR1 ORF. The sequence of GR was highly conserved among graminaceous plants and HvGR2 showed high homology to wheat (Triticum monococum; 96%) and rice (89%) cytosolic GR (Fig. 1). Likewise, HvGR1 showed 86% homology to rice chloroplastic GR. Three genes for GR exist in rice (Fig. 1). The gene for cytosolic GR (OsGR2) is located on rice chromosome 2 and is split into 16 exons (http://www.tigr.org/), whereas rice chloroplastic GR (OsGR1) is located on chromosome 3 and split into 10 exons. Rice was also found to contain a partial sequence for GR1 (LOC_Os10g28000) located on chromosome 10. This sequence lacked a FAD-binding domain and part of the NADPH-binding domain.

Fig. 1
figure 1

Phylogeny and genomic structure of GR. (a) Unrooted phylogenic tree of GR. Hv, Hordeum vulgare (Hv1: AB277096; Hv2: AB277097); Tm, T. monococum (Tm2: AY364467); Os, Oryza sativa (Os1: AK066398; Os2: AK103418; Os3: AK108799); Zm, Zea mays (Zm1: AJ006055); Bo, Brassica oleracea (Bo2: AB125639); Br, Brassica rapa (Br2: AF008441); At, Arabidopsis thaliana (At1: AY054677; At2: AF360228); Ze, Zinnia elegans (Ze2: AB158513); Ps, Pisum sativum (Ps1: X90996; Ps2: Q43621); Mc, Mesembryanthemum crystallinum (Mc2: AJ400816); Nt, Nicotiana tabacum (Nt1: P80461); Vv, Vitis vinifera (Vv1: AF019907); Bj, Brassica juncea (Bj1: AF349449). Vu, Vigna unguiculata (Vu1: DQ267474; Vu2: DQ267475); Gm, Glycine max (Gm1: P48640); GR1 represents chloroplastic while GR2 represents cytosolic GR. (b) Genomic structure of rice cytosolic GR (OsGR2). (c) Genomic structure of rice chloroplastic GR (OsGR1). (d) Genomic structure of rice partial chloroplastic GR (OsGR3). —, Introns; ■, Exons; □, Untranslated regions

Full size image

Based on their nucleotide sequences, HvGR1 and HvGR2 were predicted to encode polypeptides of 550 and 497 amino acids, respectively (Fig. 2). The NADPH and substrate-binding domains were identified by aligning the sequences with those of other GRs that had been characterized previously. Similar to other GRs, these domains were conserved for graminaceous GR1 and GR2.

Fig. 2
figure 2

Sequence homology between rice and barley GR. #, GSSG-binding domain; *, NAD(P)H-binding domain; +, FAD-binding domain

Full size image

Enzyme activity

HvGR1 and HvGR2 were expressed in E. coli as a maltose-binding fusion protein, and purified protein was tested for the ability to reduce GSSG. AB063249 was also expressed as a maltose-binding protein; however, it lacked a complete FAD-binding domain and part of the substrate-binding domain, and was therefore used as a negative control. HvGR1 and HvGR2 were shown to catalyze the conversion of GSSG to GSH (Fig. 3) as measured by the decrease in absorbance of NADPH at 340 nm. The enzyme activity for HvGR1 was three times higher than that of HvGR2.

Fig. 3
figure 3

GR activity of recombinant HvGR1 and HvGR2. The recombinant AB063249 protein lacking a FAD-binding domain and part of the NADPH-binding domain was used as a control

Full size image

The expression of GR1 and GR2 under Fe-deficient conditions

To confirm the Fe-deficient status of the plants the expression of DMAS, a gene essential for Fe-deficiency response in graminaceous plants, was observed through Northern and Western blot analysis. Northern blot analysis confirmed that the expression of DMAS in all species examined is upregulated in response to Fe-deficiency (Bashir et al. 2006). The results of Western blot analysis were also in line to support the Fe-deficient status of plants (data not shown). Northern blot analyses confirmed that AB063249 did not express under Fe-sufficient or deficient conditions (data not shown). Further, Northern blot analyses were performed to determine whether the expression of GR1 and GR2 was upregulated under Fe-deficient conditions in graminaceous crops. The expression of both genes was upregulated in barley in response to Fe-deficiency (Fig. 4). In addition, the expression of TaGR1 was not elevated in response to Fe-deficiency, while that of TaGR2 was (Fig. 4). No band was detected for maize, and faint bands were detected for rice when HvGR2 was used as probe (data not shown). Northern blot analysis was also performed using OsGR2 as a probe, which showed that the expression of OsGR2 decreased slightly under Fe-deficient conditions (Fig. 4b). The expression of HvGR1 was mainly localized to shoot tissues and only slight expression was observed in roots. Conversely, the expression of HvGR2 was mainly observed in roots, where it was significantly upregulated in response to Fe-deficiency. The expression of HvGR2 was also upregulated in response to Zn-deficiency in both root and shoot tissues (data not shown). Although we did not attempt a Northern blot analysis using OsGR1 as a probe, microarray analysis showed that OsGR1 is upregulated under Fe-deficient conditions (data not shown).

Fig. 4
figure 4

Northern blot analysis for GR1 and GR2 in graminaceous plants. (a) Northern blot analysis for chloroplastic GR (GR1). HvGR1 was used as a probe. (b) Northern blot analysis for cytosolic GR (GR2). HvGR2 was used as a probe for barley and wheat. OsGR2 was used as a probe for rice

Full size image

Northern blot analysis using the HvGR1 probe detected two bands (∼2 and ∼1.5 kb; Fig. 4b). The smaller fragment may have been the partial sequence of HvGR1. Only one band representing a ∼2-kb fragment was detected using HvGR2 as a probe.

GR activity increases in response to Fe-deficiency

Total protein was extracted from Fe-sufficient and Fe-deficient barley, wheat, rice, and maize to check the GR activity in response to Fe-deficiency. GR activity increased in barley (×1.85), wheat (×1.24), and in maize (×3.01) root tissues with Fe-deficiency (Table 1). The barley and wheat results are in line with those of the Northern blot analysis, whereas GR activity slightly decreased in rice roots (×0.87). GR activity increased in rice (×1.52), barley (×1.27), and maize (×2.95) shoot tissues. These results correspond with the Northern blot analyses for wheat, barley, and rice. We did not detect a band for maize, likely because we lacked a maize-specific probe. This means that we do not know whether these genes are post-transcriptionally regulated in maize. The enzyme activity was higher in root tissue than shoot tissue for all species examined.

Table 1 Glutathione reductase activity in response to Fe-deficiency stress
Full size table

Discussion

We cloned the genes for chloroplastic (HvGR1) and cytosolic (HvGR2) GR in barley. GR is a ubiquitous enzyme that has been cloned from a number of prokaryotic and eukaryotic species (e.g., Greer and Perham 1986; Edwards et al. 1990; Foyer et al. 1991, Creissen et al. 1992). The sequences of cytosolic and chloroplastic GR are conserved among graminaceous crops (Fig. 1) and it appears that the expression of GR is regulated by similar biotic and abiotic stresses. The HvGR2 and TaGR2 genes were mainly expressed in root tissue and a weak band was observed in shoot tissue. Similar expression patterns were observed for OsGR1 in that expression was mainly observed in root and calli, rather than in leaves (Kaminaka et al. 1998). Conversely, the expression of PsGR2 was very low in root tissue and weak in shoots and flowers (Stevens et al. 1997; Romero-Puertas et al. 2006). These results suggest that the expression of graminaceous and non-graminaceous GR may be regulated differently. It has also been reported that the expression of PsGR1 and PsGR2 are regulated differentially in response to various stresses such as mechanical wounding and increased temperature (Romero-Puertas et al. 2006). In barley, both HvGR1 and HvGR2 were upregulated by Fe-deficiency, whereas wheat exposed to the same conditions only exhibited increased levels of TaGR2 (Fig. 4).

Although we did not examine the expression of GR in response to abiotic stresses other than Fe-deficiency, studies using the differential display protocol showed that the expression of HvGR2 is upregulated under salt stress (Ueda et al. 2002). GR specific activity was also measured to determine whether this enzyme is regulated post-transcriptionally. The ZmGR1 is post-transcriptionally regulated in bundle sheath cells (Pastori et al. 2000), as are PsGR1 and PsGR2 in response to various abiotic stresses (Romero-Puertas et al. 2006). We observed increased GR activity in Fe-deficient barley, wheat, and maize roots and Fe-deficient rice, barley, and maize leaves. The expression of HvGR1 was mainly localized to shoot tissues and that of HvGR2 to root tissues. Therefore, it may be concluded that the increased GR activity in shoot tissue was due to the upregulation of HvGR1, whereas the increase in root tissue may have been a result of the upregulation of HvGR2. Northern blot analyses for barley and wheat confirmed this hypothesis by revealing that the upregulation of gene expression is directly proportional to the increase in GR-specific activity. Moreover, the expression and enzyme activity for OsGR2 decreased in response to Fe-deficiency. These results suggest that GR1 and GR2 genes are not post-transcriptionally regulated in rice, wheat, or barely. We did not detect clear bands for maize using Northern blot analysis, likely because HvGR1 and HvGR2/OsGR2 were used as probes. The GR specific activity, however, clearly showed that the gene is upregulated in Fe-deficient maize root and shoot tissues.

GR may allow plants to cope with Fe-deficiency in a variety of ways. Redox systems using heme and non-heme proteins such as cytochromes, catalase, peroxidases, and ferredoxin require Fe. Thus, under Fe-deficiency stress, the activity of these enzymes declines. GSH scavenges ROS via the Asc–GSH cycle and glutathione peroxidase (GPX). Although the Asc–GSH cycle is downregulated under Fe-deficient conditions (Zaharieva and Abadía 2003), given that the activity of SOD increases (Tewari et al. 2005), GR may play a role in coping with Fe-deficiency-induced oxidative stress through the GPX cycle in combination with SOD (Fig. 5). It should be noted that the oxidative stress induced by Fe-deficiency is fundamentally different from that induced by salt or heavy metal stress, whereby the production of ROS as well as the activities of catalase and peroxidases increase (Patra and Panda 1998; Kim et al. 2005). Conversely, under conditions of Fe-deficiency, the production of ROS does not increase significantly, but rather, oxidative stress occurs as the activities of some enzymes involved in scavenging ROS decrease.

Fig. 5
figure 5

The role of glutathione in coping with Fe-deficiency-induced oxidative stress. CAT, catalase; GPX, glutathione peroxidase; APX, ascorbate peroxidase; MDA, monodehydroascorbate; AsA, ascorbic acid; MDAR, MDA reductase; DHA, dehydroascrobate; DHAR, DHA reductase; N.E., non-enzymatic reaction. The activity/amount of the enzymes/compounds shown in red increases under Fe-deficiency, while that of shown in blue decreases

Full size image

GR may also play a role in internal Fe homeostasis. Mobilization of Fe may help plants cope with Fe-deficiency stress. Graziano and Lamattina (2005) and Graziano et al. (2002) recently discussed the role of NO in Fe mobilization in plants. Interestingly, the ability of NO to induce Fe mobilization is dependent on GSH while that of chelators is independent of GSH (Watts and Richardson 2001). This suggests that NO alone does not have the capacity to remove Fe from intermediates and that it may require the reducing capacity of GSH. Alternatively, or in combination with this latter mechanism, GSH may form a mixed Fe complex with NO in order to acquire the appropriate lipophilicity and charge to diffuse or be transported from the cell (Watts and Richardson 2002). The presence of dithiol dinitrosyl–Fe complexes within cells has been previously demonstrated (Vanin 1991). As part of this complex, GSH may play an important role in signal transduction for the regulation of gene expression in response to different stresses and Fe-homeostasis. However, a S-nitrosylated form of GSH has been suggested to act as a transport molecule for NO, thereby increasing its half-life and allowing for effective biological activity (Lipton et al. 2001; Lipton 2001). GSH is also essential for the activity of glutathione S-transferase, which specifically interacts with the dinitrosyl-diglutathionyl–Fe complex and behaves like a storage protein for this complex in vivo and in vitro (Maria et al. 2003; Turella et al. 2003). These results suggest that GR may play a role in internal Fe homeostasis in graminaceous plants and thereby allow plants to cope with Fe-deficiency.