Modulation of macronutrient metabolism in barley leaves under iron-deficient condition

Abstract

Iron (Fe) deficiency primarily damages photosystems. A large proportion of the reducing equivalents in mesophyll cells are dependent on the photosynthetic apparatus in the chloroplasts; thus, Fe deficiency profoundly affects nitrogen (N) and sulfur (S) assimilation. Microarray data suggested that nutrient metabolisms concerning the assimilation and re-utilization of essential elements were modulated in barley under Fe-deficient condition. We compared the concentrations of the major essential elements between Fe-deficient barley leaves and Fe-deficient rice leaves. The Fe-deficient barley and rice plants had the same Fe and chlorophyll concentrations, but growth was reduced more in Fe-deficient rice than in Fe-deficient barley. The accumulation of ribulose-1,5-bisphosphate carboxylase/oxygenase protein and nitrite reductase, sulfite reductase, and ferredoxin-dependent glutamate synthase mRNAs was reasonably decreased in the chlorotic young leaves of Fe-deficient barley. , , , , and Ca²⁺ concentrations were altered to a larger extent in Fe-deficient rice leaves than in Fe-deficient barley leaves, even in the non-chlorotic old leaves. In this paper, we discuss the adaptation of macronutrient metabolism to Fe deficiency in barley leaves.

Keywords:

• barley
• Fe deficiency
• nitrogen assimilation
• rice
• sulfur assimilation.

Introduction

Deficiency of iron (Fe)—an essential plant nutrient—results in a remarkable decrease in yield. Iron-deficient plants display visual symptoms as well. Thus, a great deal of effort is being directed toward solving the problem of Fe deficiency in plants (Marschner 1995), and investigating ways to increase Fe acquisition has been a central theme. To date, Fe acquisition and transport mechanisms have been explained thoroughly at the molecular biological level (Jeong and Guerinot 2009). Several attempts to reinforce Fe acquisition have successfully produced Fe deficiency-tolerant transgenic plants (Takahashi et al. 2001; Ishimaru et al. 2007; Ogo et al. 2007). However, Fe concentration is not positively correlated with Fe-deficiency tolerance. We have previously reported that Fe deficiency-tolerant barley had low Fe concentration and that Fe deficiency-sensitive rice had high Fe concentration (Maruyama et al. 2005; Hirai et al. 2007). Further, Fe concentrations are less in some Fe deficiency-tolerant transgenic rice lines than in wild-type rice (Ogo et al. 2007).

Chloroplasts are the largest sink of Fe in leaves as well as the specialized organelles for photosynthesis (Terry and Abadia 1986). In the process termed photophosphorylation, electrons activated by light energy in photosystem II (PSII) are transferred to photosystem I (PSI), coupled with adenosine triphosphate (ATP) synthesis. These electrons move through an electron transport chain and are finally received by ferredoxin, which acts as a major reducing agent (Allen 2002) in the processes of carbon (C), nitrogen (N), and sulfur (S) assimilation. Many kinds of Fe-containing proteins are involved in the electron transport or biological redox reactions. PSI is the primary system affected by Fe deficiency, probably because of its high Fe concentration (12 Fe atoms per PSI unit) (Moseley et al. 2002). Consequently, Fe deficiency decreases the production of ATP (Winder and Nishio 1995) and reducing power derived from ferredoxin (Tognetti et al. 2007). Therefore, plant chloroplasts should be equipped with adaptive mechanisms for Fe deficiency. However, such mechanisms have not been studied extensively in higher plants. Recently, we reported an adaptation mechanism by the light-harvesting antenna system in barley, which was related to heat dissipation to compensate for excess light energy under prolonged Fe deficiency (Saito et al. 2010).

Although the mechanism highly contributed to the protection of young leaves of barley from photoinhibition caused by long-term Fe deficiency, the electron transport rate (ETR) from PSII to PSI in Fe-deficient barley leaves was markedly decreased like that in Fe-deficient rice leaves (Saito et al. 2010). The decrease in ETR may immediately limit production of ATP and reductants, negatively affecting subsequent assimilation pathways under Fe deficiency. Despite this, young chlorotic leaves of barley could continue to grow slowly under prolonged Fe deficiency, which was in contrast to the observation in rice in which continuous development of young leaves could not be observed even under short-term Fe-deficient conditions (Saito et al. 2010). This different growth response between the two graminaceous plants implies that barley possesses an additional adaptation mechanism for energy use efficiency downstream of the photosynthetic electron transport chain. However, the mechanism related to the compensation of decrease in energy supplied from the electron transport chain to subsequent photosynthetic reactions, including Calvin cycle and assimilation of N and S, has not been elucidated. Since barley requires assimilated C, N, and S to synthesize and release large amounts of mugineic acids (Mori and Nishizawa 1987), it is important to evaluate the alteration in fundamental metabolic pathways under Fe-deficient conditions.

Most N and S assimilation occurs in the plastids, and the expression of N and S assimilation-related genes is closely regulated to avoid the accumulation of intermediates (Buchanan et al. 2000). In addition, Fe is required by many enzymes involved in N and S assimilation, including nitrate reductase (NR, containing heme Fe), nitrite reductase (NIR, containing a [4Fe-4S] cluster), glutamate synthase (GOGAT, containing a [3Fe-4S] cluster), adenosine 5′-phosphosulfate reductase (APR, containing a [4Fe-4S] cluster), sulfite reductase (SIR, containing a [4Fe-4S] cluster), and the protein ferredoxin (containing a [2Fe-2S] cluster) (Campbell 1996; Balk and Lobréaux 2005). Thus, prolonged Fe deficiency markedly reduces N and S assimilation, leading to cessation of growth even if carbon dioxide (CO₂) assimilation is maintained.

The interaction between plant nutrients and metabolic pathways has been discussed recently (Ohkama-Ohtsu and Wasaki 2010). However, knowledge about the relationships between Fe nutrition and other macronutrients within plant tissues is limited. In this study, we investigated the levels of transcripts involved in C, N, and S photoassimilation pathways in barley. We also compared the concentrations of C, N, and S as well as other major essential elements between barley leaves and rice leaves. The modulation of metabolite distribution in barley leaves to adapt to Fe deficiency is discussed.

Materials and Methods

Plant materials and growth conditions

Plants were hydroponically grown in a growth chamber (14-h light/10-h dark; photosynthetic photon flux density (PPFD), approximately 230 µmol m⁻² s⁻¹; for barley, 24°C light/20°C dark; for rice, 28°C light/25°C dark). The composition of the standard culture solution has been described by Maruyama et al. (2005). Barley (Hordeum vulgare L. cv. Ehimehadaka No. 1) and rice (Oryza sativa L. cv. Nipponbare) seeds were germinated on wet paper for three days and six days, respectively. The seedlings were transferred to culture solutions containing half the concentrations of all the ingredients of the standard culture solution [supplemented with 30 µM Fe-ethylenediaminetetraacetic acid (EDTA)] and grown for four days (barley) or six days (rice). When the first true leaf appeared, the plants were transferred to the standard culture solution supplemented with 30 µM Fe-EDTA and grown for four days. When the second true leaf appeared in the barley plants, or the third true leaf appeared in the rice plants, nine barley plants or 15 rice plants were transplanted into a 5-L container. One group was grown with 30 µM Fe-EDTA and the other group was grown without Fe for eight days. The pH of the culture solution was adjusted to 5.5 with 0.5 M sulfuric acid (H₂SO₄) and 1 M sodium hydroxide (NaOH), and a fresh solution was prepared every 3 d.

At harvest, the leaves were numbered as follows: cotyledon, 1; first true leaf, 2; second true leaf, 3; and so on. Only the leaf blade was harvested. Three barley plants or five rice plants were pooled as one replicate. The chlorotic leaves (or parts of leaves) (leaf 5, leaf 4, and the basal part of leaf 3 of the barley plants; leaf 5 and the basal part of leaf 4 of the rice plants) of Fe-deficient plants were pooled and designated as “younger-Fe leaves.” The green leaves (or parts of leaves) (leaf 1, leaf 2, and the tip of leaf 3 of the barley plants; leaf 2, leaf 3, and the tip of leaf 4 of the rice plants) of Fe-deficient plants were pooled and designated as “older-Fe leaves.” Leaves 4 and 5 of the control barley plants or leaves 5 and 6 of the control rice plants were pooled and designated as “younger control leaves.” Leaves 1, 2, and 3 of the control barley plants or leaves 2, 3, and 4 of the control rice plants were pooled and designated as “older control leaves.” The pooled leaves were cut into about 1-cm segments, and the segments were divided into two groups. One group was frozen in liquid nitrogen and stored in a −80°C freezer until use for the measurement of chlorophyll and ion concentrations, western blot analysis, and reverse transcriptase-polymerase chain reaction (RT-PCR). The other group was dried and stored at room temperature until use for elemental analyses.

Microarray analysis

Total RNA was extracted from leaf 8 (the youngest fully expanded leaf) of Fe-deficient (0.3 µM Fe-EDTA) and Fe-sufficient (10 µM Fe-EDTA) plants from which the leaves were harvested during daylight at 14 days or 25 days after initiating treatment with each Fe concentration (Hirai et al. 2007). The prepared samples were hybridized to the Affymetrix GeneChip® Barley1 Genome Array, as described previously (Saito et al. 2010).

Chlorophyll concentrations

Chlorophylls were extracted with 80% acetone from liquid nitrogen-ground leaves; absorbance of the supernatants was measured at 663.6, 646.6, and 750 nm using a spectrophotometer (UV-1600; Shimadzu Co. Ltd., Kyoto, Japan), according to the method described by Porra et al. (1989).

Elemental analyses

Total N and total C concentrations were measured using an elemental analyzer (CE-440; Exeter Analytical Inc., USA). Dried leaves were digested in a solution of concentrated nitric acid (HNO₃) at 150°C and dissolved in 1% HNO₃. Iron concentration was measured using an atomic absorption spectrophotometer (AA-6300; Shimadzu Co. Ltd); S, phosphorus (P), and calcium (Ca) concentrations were measured using an inductively coupled plasma-atomic emission spectrophotometer (ICPS-8000; Shimadzu Co. Ltd).

Ion concentrations

Frozen leaves were extracted with water, and the extracts were filtered through an ultrafiltration membrane (Microcon Ultracel YM-10; 10,000 molecular weight cut-off (MWCO); MILLIPORE Co. Ltd, USA). The flow-through solution was analyzed using an ion chromatograph (PIA-1000; Shimadzu Co. Ltd).

Reverse transcriptase-polymerase chain reaction (RT-PCR)

Total RNA was extracted using an RNeasy Plant Mini Kit (Qiagen K.K., Tokyo, Japan), and cDNA was synthesized from the total RNA by using the PrimeScript® II 1st strand cDNA Synthesis Kit (TaKaRa Bio Inc., Shiga, Japan). KOD FX DNA polymerase (TOYOBO Co. Ltd, Osaka, Japan) was used for amplification. Primers were designed from highly conserved sequences among graminaceous plant species. Thus, these primers could amplify the majority of the homologous genes in barley. The primers and reaction conditions are shown in Table 1. Signal intensity was determined using ImageJ v. 1.43.

Table 1 Primers and reaction conditions

Gene	Forward primer sequence (5′→3′)	Reverse primer sequence (5′→3′)	Product length (kbp)	Annealing temperature (°C)	Number of cycles
EF-1α	ACCACTGGTGGTTTTGAGGC	CCAGGGTGGTTCATGATGAT	0.70	56	25
NR	GTCGACGCCGAGCTCGCCAA	GCGCACCTCGGACATGGT	0.65	60	28
NIR	TCAAGTGGCTCGGCCTCTT	ACGCACACGTTCCACTTCCT	0.40	58	28
GS2	TGCTCGACATGGACACCA	CGTTTGTTAGTAGGGATGGGT	0.25	56	28
Fd-GOGAT	TGCATGGAGCACCGTGGT	CCATCTAGGGCTTGTATTGGTACT	0.70	58	28
APR	TCGCCGCTGGAGATCATGGAT	CCTGGTGCACGGCTCGCA	0.60	60	28
SIR	TTGGAGGAACACCAAACCAGA	CACTTGTTCACCACCTCTTTCAA	0.20	55	28
GS1	GGCGGATCTGGCATGGAT	AGCAATCGCACATGACAAGG	0.18	57	25
NADH-GOGAT	CGTTTCATTGGTGATGAAAATGGAAA	CAGTGATGGCCCAAACAAC	0.25	56	28
GDH	GAGATCAAGGTTGAATGCAC	CTTCCAGTAGCTGCATCTCT	0.45	55	28
G6P/P T	TCTGTCAGTGGCATGAATTACTA	CTCACTGCTTTGCCTGAGA	0.35	55	28
2-oxo/malate T (mitochondria)	TCATCCAGCCCATCGACATG	GCATCCTAATGAGTGCCAAATC	0.25	58	23
2-oxo/malate T (plastid)	TGCTGATGTACTTCTACTCCCACTA	CTCCAGCACCAAGCCAGATG	0.25	59	25
EF-1α, elongation factor 1α; NR, nitrate reductase; NIR, nitrite reductase; GS, glutamine synthase; Fd-GOGAT, ferredoxin (Fd)-dependent glutamate synthase; APR, adenosine 5′-phosphosulfate reductase; SIR, sulfite reductase; NADH-GOGAT, reduced nicotinamide adenine dinucleotide (NADH)-dependent glutamate synthase; GDH, glutamate dehydrogenase; G6P/P T, glucose-6-phosphate/phosphate translocator; 2-oxo/malate T, 2-oxoglutarate/malate translocator.

Western blot analysis

Proteins were extracted from liquid nitrogen-ground leaves by using an extraction buffer containing 62.5 mM Tris-HCl at pH 6.8, 2.5% sodium dodecyl sulfate (SDS), and 10% glycerol. Homogenates were boiled and centrifuged, and protein concentrations in the supernatants were measured using the bicinchoninic acid (BCA) method (Smith et al. 1985). Sample aliquots were supplemented with 2-mercaptoethanol at a final concentration of 5%. SDS-polyacrylamide gel electrophoresis was performed with 12.5% gel. Proteins were electrotransferred to polyvinylidene difluoride (PVDF) membranes (Hybond™-P; GE Healthcare UK Ltd, Buckinghamshire, UK), and the membranes were probed with 1:1000 diluted anti-ferredoxin antibodies (Agrisera, Vännäs, Sweden). Detection was performed using the enhanced chemiluminescence (ECL) Plus Western Blotting Detection System (GE Healthcare UK Ltd) according to the manufacturer's protocols, with a 1:30,000 diluted peroxidase-labeled anti-rabbit antibody as the secondary antibody. Signal intensity was determined using ImageJ v. 1.43.

Results

Analysis of microarray data

In our previous study, we analyzed the microarray data of barley, focusing on photosynthesis under Fe-deficient condition (Saito et al. 2010). In this study, we re-analyzed those data, searching for genes related to the photoassimilation, transport, utilization, and recycling of C, N, and S among those upregulated or downregulated by Fe deficiency. Further, we focused on glutathione (GSH) metabolism as an intracellular redox buffer (Rouhier et al. 2008). GSH is also a major cysteine reservoir and thus greatly affects S metabolism.

Table 2 (14-days Fe deficiency) and Table 3 (25-days Fe deficiency) show the upregulated and downregulated genes related to the photoassimilation and transport of C, N, and S in Fe-deficient barley. As expected, ferredoxin gene was downregulated by Fe deficiency in the young leaves (Tables 2 and 3). Nitrogen and S assimilation-related genes were downregulated by 14-days Fe deficiency (Table 2) but not by 25-days Fe deficiency (Table 3). Upregulation of thioredoxin/glutaredoxin gene (Tables 2 and 3) was consistent with the previous reports that indicated that thioredoxin/glutaredoxin is important under reducing power-starved conditions (Stenbaek et al. 2008; Forner-Giner et al. 2010). GSH transferases (GSTs) were also upregulated up to three-fold by Fe deficiency in the leaves (Tables 2 and 3). Furthermore, several sugar transporter genes were upregulated by Fe deficiency (Tables 2 and 3).

Table 2 Changed expression of genes related to photoassimilation and transport of carbon (C), nitrogen (N), and sulfur (S)

Probe ID	Annotation	Tissue	Expression fold	Function
For sources of reducing power
HV_CEa0013J19f_at	Photosystem I P700 apoprotein A1 (PsaA)	Young leaf	0.5	Photosystem I
Contig5299_at	Ferredoxin	Young leaf	0.3	Ferredoxin
Contig1307_at	Ferredoxin	Young leaf	0.3	Ferredoxin
HU08P15u_s_at	Ferredoxin-NADP^+ reductase	Young leaf	0.4	Ferredoxin reductase
Glutathione-related redox homeostasis
Contig23773_at	Putative glutaredoxin	Young leaf	5.7	Thioredoxin/Glutaredoxin
Contig5236_at	Putative thioredoxin	Young leaf	2.8	Thioredoxin/Glutaredoxin
Contig9527_at	Glutathione reductase (HvGR1, chloroplastic)	Young leaf	0.8 (NC)	Glutathione reductase
HT05E09u_s_at	Glutathione reductase (HvGR2, cytosolic)	Young leaf	0.6 (NC)	Glutathione reductase
Contig3519_at	Glutathione reductase (HvGR2, cytosolic)	Young leaf	0.7 (NC)	Glutathione reductase
Contig5618_at	Thioredoxin-NADP^+ reductase	Young leaf	0.8 (NC)	Thioredoxin reductase
Contig15951_at	Glutathione S-transferase	Young leaf	5.7	Glutathione conjugation
Contig22280_at	Glutathione S-transferase (maize GST16 like)	Young leaf	4.6	Glutathione conjugation
Contig14387_at	Glutathione S-transferase (OsGSTU9 like)	Young leaf	3.2	Glutathione conjugation
Contig9944_s_at	Glutathione S-transferase	Young leaf	2.5	Glutathione conjugation
Contig2456_at	Glutathione transferase (wheat GSTF5 like)	Young leaf	2.0	Glutathione conjugation
Contig9944_at	Glutathione S-transferase	Young leaf	2.0	Glutathione conjugation
Nitrogen assimilation-related
Contig8716_at	NAD(P)H-nitrate reductase (NADH-NR)	Young leaf	0.4	N assimilation
Contig13776_at	NAD(P)H-nitrate reductase (NADH-NR)	Young leaf	0.3	N assimilation
EBma05_SQ003_C16_s_at	NAD(P)H-bispecific nitrate reductase (barley NAR7)	Young leaf	0.2	N assimilation
Contig8185_at	Ferredoxin-nitrite reductase (ferredoxin-NIR)	Young leaf	0.3	N assimilation
Contig22563_at	Ammonium transporter (OsAMT3;1-like)	Young leaf	3.0	transporter
Sulfur assimilation-related
Contig4024_at	Adenosine 5′-phosphosulfate reductase	Young leaf	0.4.	S assimilation
Contig4025_s_at	Adenosine 5′-phosphosulfate reductase	Young leaf	0.3	S assimilation
Contig4025_x_at	Adenosine 5′-phosphosulfate reductase	Young leaf	0.3	S assimilation
Contig4570_s_at	Ferredoxin-sulfite reductase precursor (ferredoxin-SIR)	Young leaf	0.4	S assimilation
Contig4569_at	Ferredoxin-sulfite reductase precursor (ferredoxin-SIR)	Young leaf	0.4	S assimilation
Contig20387_at	Low-affinity sulfate transporter 3	Young leaf	0.5	transporter
Carbon assimilation-related
Contig346_at	Ribulose-1,5-bisphosphate carboxylase/oxygenase small subunit (rbcS)	Young leaf	2.0	Calvin cycle
EBem05_SQ003_O11_at	Ribulose-1,5-bisphosphate carboxylase/oxygenase large subunit (rbcL)	Young leaf	2.0	Calvin cycle
Contig6706_at	Putative sugar transporter	Young leaf	2.6	Sugar transporter
Contig14224_at	Putative sugar transporter	Young leaf	0.4	Sugar transporter
HVSMEa0008H21r2_at	Similar to hexose carrier protein	Young leaf	55.7	Sugar transporter
Contig16113_at	Dof zinc finger protein	Young leaf	2.1	Transcriptional factor
Note: Upregulated genes (fold increase, boldfaced values) and downregulated genes (fold decrease, italicized values) in young leaves after 14 days of iron (Fe) deficiency. The fold values are in comparison with the value for Fe-sufficient barley leaves. NC means “no change” between the different Fe treatments.

Table 3 Changed expression of genes related to photoassimilation and transport of carbon (C), nitrogen (N), and sulfur (S)

Probe ID	Annotation	Tissue	Expression fold	Function
For sources of reducing power
Contig1317_s_at	Ferredoxin	Young leaf	0.4	Ferredoxin
Glutathione-related redox homeostasis
Contig23442_at	Putative glutaredoxin	Young leaf	2.3	Thioredoxin/Glutaredoxin
rbaal31m19_s_at	Putative thioredoxin	Young leaf	2.5	Thioredoxin/Glutaredoxin
Contig9527_at	Glutathione reductase (HvGR1, chloroplastic)	Young leaf	1.2 (NC)	Glutathione reductase
HT05E09u_s_at	Glutathione reductase (HvGR2, cytosolic)	Young leaf	1.4 (NC)	Glutathione reductase
Contig3519_at	Glutathione reductase (HvGR2, cytosolic)	Young leaf	1.0 (NC)	Glutathione reductase
Contig5618_at	Thioredoxin-NADP^+ reductase	Young leaf	1.0 (NC)	Thioredoxin reductase
Contig21604_at	Glutathione synthetase	Young leaf	3.0	Glutathione synthesis
Contig7448_s_at	Glutathione S-transferase	Young leaf	3.2	Glutathione conjugation
Contig14387_at	Glutathione S-transferase (OsGSTU9 like)	Young leaf	2.1	Glutathione conjugation
HVSMEa0014H14r2_s_at	Glutathione S-transferase (maize GST 37 like)	Young leaf	2.1	Glutathione conjugation
Contig10408_at	Glutathione S-transferase (maize GST 41 like)	Old leaf	4.0	Glutathione conjugation
HW01K14T_at	Glutathione transferase (wheat GSTF3 like)	Old leaf	2.3	Glutathione conjugation
Contig2456_at	Glutathione transferase (wheat GSTF5 like)	Old leaf	2.0	Glutathione conjugation
Nitrogen assimilation-related
EBro08_SQ006_D11_at	High-affinity nitrate transporter	Old leaf	2.3	transporter
Carbon assimilation-related
Contig1017_at	Ribulose-1,5-bisphosphate carboxylase/oxygenase activase A	Young leaf	0.4	Calvin cycle
Contig6707_at	Putative sugar transporter	Young leaf	2.8	Sugar transporter
Contig9662_at	Hexose transporter-like protein	Young leaf	2.0	Sugar transporter
Contig24019_at	Putative hexose carrier protein (putative monosaccharide transporter)	Old leaf	0.1	Sugar transporter
Contig481_at	Sucrose synthase 2 iSucrose-UDP gluccsvltransferase 2)	Young leaf	2.0	Sucrose synthesis
Contig23306_at	Sucrose synthase 2	Old leaf	3.7	Sucrose synthesis
Contig10722_at	Putative starch synthase III	Young leaf	2.5	Starch synthesis
Contig16389_at	Starch-branching enzyme-like protein	Young leaf	2.0	Starch synthesis
Note: Upregulated genes (fold increase, boldfaced values) and downregulated genes (fold decrease, italicized values) in leaves after 25 days of iron (Fe) deficiency. The fold values are in comparison with the value for Fe-sufficient barley leaves. NC means “no change” between the different Fe treatments.

Table 4 (14-days Fe deficiency) and Table 5 (25-days Fe deficiency) show the upregulated and downregulated genes related to the utilization and recycling of C, N, and S in Fe-deficient barley. Cell wall and protein degradation-related genes were remarkably upregulated. The fact that several senescence- and ubiquitination-related genes were upregulated suggests controlled degradation of leaf cells. Upregulation of both protease and protease inhibitor genes also supports this suggestion (Solomon et al. 1999). These microarray data led us to conceive that extensive modulation of metabolism occurred in barley under Fe-deficient condition.

Table 4 Changed expression of genes related to utilization and recycling of carbon (C), nitrogen (N), and sulfur (S)

Probe ID	Annotation	Tissue	Expression fold	Function
Carbohydrate metabolism
HVSMEa0009F12r2_s_at	Pyrophosphate-fructose-6-phosphate 1-phosphotransferase alpha subunit [PFP)	Young leaf	0.5	Glycolysis
Contig9121_at	Glucose-6-phosphate/phosphate translocator precursor homolog	Young leaf	0.4	Pentose phosphate pathway
HT09A17u_at	UDP-glucose 4-epimerase	Young leaf	2.3	Carbohydrate metabolism
Contig21324_at	Putative trehalose-6-phosphate phosphatase	Young leaf	2.0	Carbohydrate metabolism
Sugar starvation & cell wall degradation
Contig2738_s_at	Beta-glucosidase	Young leaf	0.4	Glycosyl hydrolase
Contig7938_at	Alpha-glue osidase	Young leaf	2.0	Glycosyl hydrolase
Contig6481_at	Beta-galactosidase	Young leaf	2.0	Glycosyl hydrolase
Contig11289_at	Endo-1,3-beta-glucanase/glucosidase	Young leaf	36.8	Glycosyl hydrolase
HVSMEm0003C15r2_s_at	Endo-1,3-beta-glucanase/glucosidase	Young leaf	26.0	Glycosyl hydrolase
Contig1637_s_at	Endo-1,3-beta-glucanase/glucosidase	Young leaf	17.1	Glycosyl hydrolase
Contig1637_at	Endo-1,3-beta-glucanase/glucosidase	Young leaf	14.9	Glycosyl hydrolase
Contig2346_at	Endo-1,3-beta-glucanase/glucosidase	Young leaf	2.3	Glycosyl hydrolase
HVSMEb0004L16r2_at	Endo-1,3-beta-glucanase/glucosidase	Young leaf	0.4	Glycosyl hydrolase
HU10K22u_s_at	Glucan 1,3-beta-glucosidase	Young leaf	0.4	Glycosyl hydrolase
Contig1639_at	Beta-glucanase	Young leaf	0.5	Glycosyl hydrolase
Contig18863_at	Cellulase/endo-beta-1,4-glucanase	Young leaf	2.0	Glycosyl hydrolase
Contig12692_s_at	Xyloglucan endo-1,4-beta-D-glucanase/xyloglucan endotransglycosylase	Young leaf	4.3	Glycosyl hydrolase
Contig21684_at	Xyloglucan endo-1,4-beta-D-glucanase/xyloglucan endotransglycosylase	Young leaf	3.5	Glycosyl hydrolase
Contig2671_at	Xyloglucan endo-1,4-beta-D-glucanase/xyloglucan endotransglycosylase	Young leaf	2.0	Glycosyl hydrolase
Contig2672_at	Xyloglucan endo-1,4-beta-D-glucanase/xyloglucan endotransglycosylase	Young leaf	0.2	Glycosyl hydrolase
Contig13674_at	Beta-D-xylosidase	Young leaf	2.0	Glycosyl hydrolase
Contig26387_at	Glycosyl hydrolase family 10	Young leaf	4.9	Glycosyl hydrolase
Contig10084_at	Putative 1,4-beta-mannan endohydrolase	Young leaf	2.8	Glycosyl hydrolase
Proteolysis, senescence acceleration, programmed cell death, & amino acid reutilization
Contig4887_s_at	Vignain/bean endopeptidase/cysteine proteinase/sulfhydryl-endopeptidase	Young leaf	0.4	Peptidase
HV_CEa0009O07r2_s_at	Putative acyltransferase/Putative serine carboxypeptidase	Young leaf	9.8	Peptidase
Contig5275_at	Putative carboxyl-terminal peptidase	Young leaf	2.3	Peptidase
Contig5272_s_at	Putative carboxyl-terminal peptidase	Young leaf	2.8	Peptidase
Contig10893_at	Putative serine carboxypeptidase II	Young leaf	2.0	Peptidase
HV05G16u_at	Putative serine carboxypeptidase II	Young leaf	2.0	Peptidase
Contig23913_at	Putative aspartyl protease	Young leaf	3.0	Protease
EBpi01_SQ004_E18_at	Subtilisin-like serine protease	Young leaf	3.0	Protease
Contig18723_at	Putative subtilisin-like protease	Young leaf	2.5	Protease
Gontig12029_s_at	Putative subtilisin-like serine protease	Young leaf	0.4	Protease
Contig20270_at	Putative cucumisin-like serine protease	Young leaf	2.3	Protease
Contig14748_at	Putative serine protease	Young leaf	2.8	Protease
Contig12417_at	Putative serine protease	Young leaf	0.4	Protease
HVSMEk0006O10r2_at	Thiol protease aleurain	Young leaf	2.0	Protease
Contig9738_at	Putative senescence-associated protein	Young leaf	0.5	Senescence
Contig6229_s_at	Ubiquitin-specific protease 5 (UBP5)	Young leaf	2.5	Ubiquitination
rbaal38e14_x_at	Ubiquitin-protein ligase	Young leaf	0.4	Ubiquitination
HV_CEa0002D05f_s_at	Ubiquitin-conjugating enzyme	Young leaf	0.5	Ubiquitination
Contig3381_s_at	Subtilisin-chymotrypsin inhibitor 2	Young leaf	2.5	Protease inhibitor
HD04G07u_s_at	Putative protease inhibitor [Bsi1, serine-type endopeptidase inhibitor activity)	Young leaf	6.5	Protease inhibitor
Note: Upregulated genes (fold increase, boldfaced values) and downregulated genes (fold decrease, italicized values) in young leaves after 14 days of iron (Fe) deficiency. The fold values are in comparison with the value for Fe-sufficient barley leaves.

Table 5 Changed expression of genes related to utilization and recycling of carbon (C), nitrogen (N), and sulfur (S)

Probe ID	Annotation	Tissue	Expression fold	Function
Carbohydrate metabolism
HVSMEa0009F12r2_s_at	Pyrophosphate-fructose-6-phosphate 1-phosphotransferase alpha subunit (PFP)	Young leaf	0.5	Glycolysis
HVSMEg0005J01r2_x_at	Glucose-6-phosphate/phosphate translocator precursor	Old leaf	0.6	Pentose phosphate pathway
HVSMEg0015N22r2_at	Putative carboxyphosphonoenolpyruvate mutase	Old leaf	3.0	Carbohydrate metabolism
Sugar starvation & cell wall degradation
Contig4954_at	Putative sugar starvation-induced protein	Young leaf	2.6	Sugar starvation response
Contig4954_s_at	Putative sugar-starvation induced protein	Young leaf	2.3	Sugar starvation response
EBro08_SQ005_O13_at	Oligo-1,4→1,4-glucantransferase/amylo-1,6-glucosidase	Old leaf	11.3	Glyco-syl hydrolase
M96941_at	Endo-1,3-beta-D-glucosidase	Young leaf	2.1	Glycosyl hydrolase
Contig1637_s_at	Endo-1,3-beta-glucanase/glucosidase	Young leaf	2.0	Glycosyl hydrolase
Proteolysis, senescence acceleration, programmed cell death, & amino acid reutilization
Contig600_at	Carboxypeptidase C	Young leaf	2.0	Peptidase
Contig18723_at	Putative subtilisin-like protease	Young leaf	2.6	Protease
Contig12029_s_at	Putative subtilisin-like serine protease	Young leaf	2.5	Protease
Contig19015_at	Putative cysteine proteinase	Young leaf	3.0	Protease
Contig6776_s_at	Putative cysteine proteinase	Young leaf	2.1	Protease
Contig6777_at	Putative cysteine proteinase	Young leaf	2.1	Protease
Contig5759_at	Probable cysteine proteinase/cysteine endopeptidase EP-A	Old leaf	2.3	Pro-tease
Contig7199_at	Putative ATP-dependent Clp protease regulatory subunit ClpX	Old leaf	3.3	Protease
Contig8356_s_at	Senescence-associated protein 6	Young leaf	2.0	Senescence
Contig8355_at	Senescence-associated protein 6	Young leaf	4.6	Senescence
HT11C04u_s_at	Putative senescence-associated protein	Young leaf	4.6	Senescence
Contig12125_at	Senescence-associated protein	Young leaf	2.8	Senescence
Contig1010_at	Putative proteinase inhibitor	Old leaf	2.8	Protease inhibitor
Contig3383_at	Subtilisin-chymotrypsin inhibitor 2	Young leaf	0.2	Protease inhibitor
Contig3380_s_at	Subtilisin-chymotrypsin inhibitor 2	Young leaf	2.8	Protease inhibitor
Note: Upregulated genes (fold increase, boldfaced values) and downregulated genes (fold decrease, italicized values) in leaves after 25 days of iron (Fe) deficiency. The fold values are in comparison with the value for Fe-sufficient barley leaves.

Conditions of plant materials used in biochemical analyses

Iron deficiency is more serious in young organs than in those that develop in the early stage of plant development. An effective way to monitor the disturbance in various metabolic pathways due to Fe deficiency at the whole-plant level is to differentiate the organs according to their age, depending on which the organs show different degrees of Fe-deficiency symptoms. Thus, we compared the concentrations of various components between younger and older leaves in this study. The culture solution used in this work contained only Ca(NO₃)₂ as the N source. This enabled us to compare barley and rice, which were grown under the same condition regarding the cost of reducing inorganic N.

First, we confirmed the growth of and chlorophyll and Fe concentrations in the plant materials used in all the experiments. Plant height and shoot fresh weight were more reduced in rice than in barley during the eight day Fe-deficiency treatment (Table 6): barley and rice shoot fresh weight decreased to 70.4% and 44.4%, respectively. However, the reduction in chlorophyll and Fe concentrations were similar between barley and rice (Figure 1). These findings were consistent with our previous observations (Maruyama et al. 2005; Saito et al. 2010). These plants were used in subsequent experiments.

Figure 1 Chlorophyll and iron (Fe) concentrations in leaf blade. (A, B) Chlorophyll a + b concentrations are represented as average ± standard error (SE) (n = 3). (C, D) Iron concentrations are represented as average ± SE (n = 3). (A, C) Barley. (B, D) Rice. See Materials and Methods section for sample names.

Table 6 Growth of barley and rice

Species	Treatment	Plant height (cm)	Shoot fresh weight/plant (g)
Barley	Control	27.8 ± 0.472 (100%)	2.749 ± 0.121 (100%)
	Iron deficiency	25.8 ± 0.383 (92.8%)	1.935 ± 0.069 (70.4%)
Rice	Control	45.3 ± 0.470 (100%)	0.917 ± 0.028 (100%)
	Iron deficiency	32.6 ± 0.648 (72.0%)	0.407 ± 0.019 (44.4%)
Note: Iron-deficient plants were grown without iron for eight days. The values are expressed as average ± standard error (SE). Barley, n = 9; Rice, n = 30. Values in parentheses represent the percentage, which is taken as 100% for the leaves of control plants.

Expression of molecules related to carbon, nitrogen, and sulfur assimilation in leaves

The accumulation of ribulose-1,5-bisphosphate carboxylase/oxygenase (RuBisCO) protein was decreased by Fe deficiency in the younger leaves of barley (Figure 2A). However, the accumulation was higher in the older leaves of Fe-deficient barley than in the older control leaves. RuBisCO accumulation was not altered considerably by Fe deficiency in rice. Thus, no relationship was found between chlorophyll concentration and RuBisCO accumulation. Next, we compared the accumulation of ferredoxin protein between barley and rice (Figure 2A). Predictably, it was dramatically reduced in both Fe-deficient barley and Fe-deficient rice.

Figure 2 Expression of molecules related to carbon (C), nitrogen (N), and sulfur (S) assimilation. (A) Western blot analysis of ferredoxin. Each lane was applied with 10 µg of proteins. Coomassie brilliant blue (CBB)-stained ribulose-1,5-bisphosphate carboxylase/oxygenase (RuBisCO) bands are also shown. (B) Reverse transcriptase-polymerase chain reaction (RT-PCR) of N and S assimilation-related genes from barley. See Table 1 for gene names and reaction conditions. Values indicated in (A) and (B) represent the signal intensity relative to that for younger leaf of control plant. Values on the image of (B) represent the normalized values to intensity of elongation factor 1α (EF1-α). Western blot analysis and RT-PCR were performed at least twice. Typical images are presented. See Materials and Methods section for sample names.

RT-PCR revealed that the accumulation of mRNAs of S assimilation-related APR and SIR was decreased in Fe-deficient barley (Figure 2B). The accumulation of mRNAs of N assimilation-related NR, NIR, glutamine synthase 2 (GS2, chloroplastic GS), and ferredoxin-dependent GOGAT (Fd-GOGAT) was decreased in Fe-deficient barley, especially in the younger leaves (Figure 2B). The accumulation of mRNAs of reduced nicotinamide adenine dinucleotide (NADH)-dependent GOGAT (NADH-GOGAT) and glutamate dehydrogenase (GDH) was decreased in both the younger and older leaves of Fe-deficient barley (Figure 2B). The expression of other genes involved in the metabolism of organic acids that support N assimilation, such as mitochondrial 2-oxoglutarate/malate translocator gene, was decreased in the younger leaves of Fe-deficient barley, while that of glucose-6-phosphate/phosphate (G6P/P) translocator gene was increased in both the younger and older leaves of Fe-deficient barley (Figure 2B). The accumulation of mRNA of N translocation-related GS1 (cytosolic GS) was increased in both the younger and older leaves of Fe-deficient barley (Figure 2B).

Leaf blade components

Molecular and biochemical analyses suggested that Fe deficiency affects the metabolism of major essential elements. Thus, we analyzed the total concentrations and concentrations of the major inorganic forms of several elements in barley and rice leaves.

Total C concentration was decreased in both the younger and older leaves of Fe-deficient rice (Table 7), whereas it was decreased in the younger leaves of Fe-deficient barley but increased in the older leaves (Table 7).

Table 7 Components of leaf blade

		Barley				Rice
		Control		Fe deficiency		Control		Fe deficiency
Element		Younger	Older	Younger	Older	Younger	Older	Younger	Older
Carbon	Total (%, W/DW)	37.69 ± 0.159	35.15 ± 0.147	34.97 ± 0.301	36.82 ± 0.176	43.66 ± 0.040	43.54 ± 0.278	40.36 ± 0.142	41.88 ± 0.147
Nitrogen	Total (%, W/DW)	5.35 ± 0.112	5.19 ± 0.112	5.71 ± 0.217	4.71 ± 0.071	5.49 ± 0.047	5.55 ± 0.012	5.14 ± 0.144	4.92 ± 0.043
	(µmol/g FW)	2.05 ± 0.13	1.98 ± 0.61	1.69 ± 0.04	1.94 ± 0.27	1.52 ± 0.11	1.43 ± 0.04	2.71 ± 0.09	3.95 ± 0.94
	(µmol/g FW)	70.67 ± 2.13	98.30 ± 3.00	102.00 ± 2.89	64.50 ± 7.48	8.87 ± 0.33	26.40 ± 0.72	45.50 ± 1.86	8.35 ± 0.13
Sulfur	Total (µmol/g DW)	78.73 ± 1.09	88.88 ± 2.59	78.92 ± 0.95	102.00 ± 3.44	115.40 ± 4.08	133.90 ± 2.89	145.30 ± 3.23	143.40 ± 1.83
	(µmol/g FW)	3.73 ± 0.16	4.07 ± 0.40	3.59 ± 0.33	6.01 ± 0.43	10.10 ± 0.71	17.50 ± 0.27	19.60 ± 0.17	25.10 ± 1.08
Phosphorus	Total (µmol/g DW)	194.9 ± 2.49	257.9 ± 11.00	290.4 ± 3.74	358.4 ± 14.00	118.4 ± 4.79	101.4 ± 1.81	181.1 ± 1.96	159.3 ± 1.64
	(µmol/g FW)	20.3 ± 1.11	33.4 ± 1.63	25.6 ± 0.56	53.8 ± 4.55	13.8 ± 1.02	17.3 ± 0.40	31.9 ± 0.62	42.5 ± 1.26
Calcium	Total (µmol/g DW)	48.92 ± 2.12	155.10 ± 5.15	59.16 ± 1.82	191.90 ± 7.88	60.49 ± 4.06	216.30 ± 7.31	73.30 ± 4.45	232.70 ± 8.87
	Ca^2+ (µmol/g FW)	4.810 ± 0.11	21.200 ± 0.29	4.470 ± 0.43	26.600 ± 1.39	0.606 ± 0.12	0.968 ± 0.09	2.270 ± 0.10	6.630 ± 2.14
Note: The values are expressed as average ± standard error (SE) (n = 3). See Materials and Methods section for sample names. Statistical significance is shown in Table 8. W, weight; DW, dry weight; FW, fresh weight.

Total N and nitrate ion () concentrations were increased in the younger leaves and decreased in the older leaves of Fe-deficient barley. Ammonium ion () concentration was decreased in the younger leaves of Fe-deficient barley. Total N concentration was decreased in both the younger and older leaves of Fe-deficient rice. Nitrate ion () concentration was dramatically increased in the younger leaves and decreased in the older leaves of Fe-deficient rice. Ammonium ion () concentration was increased in both the younger and older leaves of Fe-deficient rice (Table 7).

Total S and sulfate ion () concentrations were increased by Fe deficiency in the older leaves of barley but in the younger leaves of rice (Table 7). Total P and total Ca concentrations were increased by Fe deficiency in both the younger and older leaves of barley and rice. Dihydrogen phosphate ion () and Ca²⁺ concentrations were also increased by Fe deficiency, but the increases were greater in rice than in barley (Table 7). Concerning other macronutrients, total potassium (K), total magnesium (Mg), K⁺, and Mg²⁺ concentrations were not altered significantly by Fe deficiency in both barley and rice (data not shown).

Thus, the effects of Fe deficiency on the metabolism of major essential elements were different between barley and rice leaves.

Discussion

In this study, a reasonable decrease was observed in ferredoxin protein by Fe deficiency. Total C concentration decreased by Fe deficiency in the younger leaves of barley and rice. However, the effects of Fe deficiency on the growth of and components of leaf blades of barley were opposite to or lesser than those on the growth of and components of leaf blades of rice (Table 8). In particular, growth inhibition in barley was lesser than that in rice (Table 6). Nevertheless, the decrease in NR, NIR, GS2, and Fd-GOGAT mRNAs was remarkable in the younger-Fe leaves of barley. These results suggest that barley could suppress the fluctuation in nutrient concentrations caused by Fe deficiency and regulate the assimilation rate to become rather low under the Fe-deficient condition. Since the energy cost of reducing to corresponds to 12–26% of the photosynthetically generated reductant (Bloom 1997; Noctor and Foyer 1998), downregulation of N assimilation may be one of the important adaptations to Fe deficiency.

Table 8 Summary of the impacts of iron (Fe) deficiency

	Barley		Rice
	Younger	Older	Younger	Older
Total carbon	−**	+**	−**	−**
Total nitrogen	+	−*	−	−**
	−	na	+**	+ +
	+**	−*	+ +**	− −**
Total sulfur	na	+*	+**	na
	na	+*	+**	+**
Total phosphorus	+**	+**	+**	+**
	+*	+*	+ +**	+ +**
Total calcium	+*	+*	+	+
Ca^2+	na	+*	+ +**	+ +
RuBisCO	−	+	na	na
Ferredoxin	− −	−	− −	− −
NR	− −	−
NIR	− −	−
GS2	− −	+
Fd-GOGAT	−	na
APR	− −	−
SIR	− −	− −
GS1	+ +	+ +
NADH-GOGAT	− −	− −
GDH	−	−
G6P/P translocator	+	+
2-Oxoglutarate/malate translocator (mitochondria)	−	na
2-Oxoglutarate/malate translocator (plastid)	na	na
Note: na, not altered by Fe deficiency; +, increased by Fe deficiency; −, decreased by Fe deficiency.
Significant difference between control plants and Fe-deficient plants (Student's t-test, *P < 0.05, **P < 0.01).
RuBisCO, ribulose-1,5-bisphosphate carboxylase/oxygenase; NR, nitrate reductase; NIR, nitrite reductase; GS, glutamine synthase; Fd-GOGAT, ferredoxin (Fd)-dependent glutamate synthase; APR, adenosine 5′-phosphosulfate reductase; SIR, sulfite reductase; NADH-GOGAT, reduced nicotinamide adenine dinucleotide (NADH)-dependent glutamate synthase; GDH, glutamate dehydrogenase; G6P/P, glucose-6-phosphate/phosphate.

We have previously reported that visible senescence of older leaves of barley is accelerated by Fe deficiency (Maruyama et al. 2005). Total N and concentrations were decreased by Fe deficiency in the older leaves of barley (Tables 7 and 8), and proteolysis-, senescence-, and amino acid re-utilization-related genes were upregulated by Fe deficiency in both the younger and older leaves of barley (Tables 4 and 5). Therefore, turnover and redistribution of N could be induced in Fe-deficient barley. Our finding that the expression of GS1, which contributes to N remobilization (Tobin and Yamaya 2001), was induced by Fe deficiency (Figure 2 and Table 8) also supports this hypothesis. In the younger-Fe leaves of barley, the genes related to N assimilation were greatly downregulated as mentioned earlier. Thus, the ability of N assimilation from newly acquired via the roots should be markedly reduced in the younger-Fe leaves of barley, which would result in an increase in concentration in these leaves. However, the rate of increase in concentration in the younger-Fe leaves was lower in barley (144%) than in rice (513%), compared with the younger control leaves (Table 7). These results seem to indicate active remobilization of organic N from the older leaves to the younger leaves of Fe-deficient barley. Ferredoxin protein and mRNAs of N assimilation-related genes could be well detected in the older leaves of Fe-deficient barley (Figure 2 and Table 8), as also the concentrations of chlorophylls and Fe (Figure 1). Thus, the older-Fe leaves of barley may be able to assimilate newly acquired . Interestingly, high-affinity transporter gene was upregulated in the older leaves of barley after 25 days of Fe deficiency by approximately two-fold that in the leaves of control barley (Table 3). This seems to support active N assimilation in the older-Fe leaves. Further, Arabidopsis thaliana nitrate transporter NRT1.7 is induced in senescent leaves and contributes to remobilize (Fan et al. 2009). In Fe-deficient barley, both organic N and could be transported from the older leaves to the younger leaves. Taken together, there is a possibility that the older leaves of Fe-deficient barley simultaneously assimilated and supplied N compounds, supported by enhanced proteolysis, senescence, and amino acid recycling at the transcriptional level (Tables 4 and 5).

In contrast to the increased total N concentration in the younger leaves of Fe-deficient barley, the total N concentration decreased in both the older and younger leaves of Fe-deficient rice. However, concentration increased in both the younger and older leaves of Fe-deficient rice but not in Fe-deficient barley. Because the concentration in the culture solution used in this study was extremely low (4 mM as a calcium salt and 60 nM as a molybdenum salt) and the primary assimilation pathway should be strongly inhibited under Fe-deficient conditions, increased in the leaves of Fe-deficient rice was thought to be derived from the photorespiration pathway. Indeed, handled by the photorespiratory N cycle in a photorespiring leaf can be up to 20 times than that resulting from the reduction of (Guo et al. 2007). However, photorespiration requires an extra amount of NADH and ATP for the reassimilation process of , which is consequently produced in the photorespiration process. As a result, it can be considered that Fe-deficient rice could not appropriately reassimilate from the photorespiration pathway because of energy shortage during Fe deficiency. However, an increase in GS1 in the whole shoots of Fe-deficient barley might prevent an increase in concentration in barley, even though the production of photorespiratory was enhanced during Fe deficiency. It is suggested that there is sufficient GS1 to support primary assimilation in barley leaves (Tobin and Yamaya 2001).

In the younger leaves of Fe-deficient barley, S assimilation-related genes and transporter gene were downregulated (Table 2). It is reasonable because the reducing power in these leaves could be remarkably decreased in Fe deficiency. However, Fe deficiency-induced increase in S assimilation in Strategy II plants may contribute to the biosynthesis of mugineic acid phytosiderophores (Astolfi et al. 2006). Total S and concentrations increased in the older leaves of Fe-deficient barley (Tables 7 and 8). Since ferredoxin protein and mRNAs of APR and SIR could be detected in the older-Fe leaves of barley (Figure 2), it is indicated that barley may assimilate S in the older-Fe leaves. Increased total C concentration in the older leaves of Fe-deficient barley may provide the basis for S assimilation. Unlike the total S and concentrations in the younger leaves of Fe-deficient barley, total S and concentrations were increased in the younger leaves of Fe-deficient rice (Tables 7 and 8). However, ferredoxin protein could not be detected and concentration increased two-fold in the younger-Fe leaves of rice (Table 7); thus, active S assimilation is questionable in Fe-deficient rice.

Our microarray analysis revealed the upregulation of mitochondrial monothiol glutaredoxin (GRX) homologs in Fe-deficient barley leaves (Tables 2 and 3). In yeast, Grx5 knockout leads to Fe accumulation in the cell and inactivation of [Fe-S]-containing enzymes in mitochondria, suggesting that GRXs play a role in [Fe-S] protein synthesis or repair of the scaffoldings (Wingert et al. 2005; Rouhier et al. 2008). Recently, it has been reported that plastidic plant GRXs can act as scaffold proteins in the transfer of [Fe-S] clusters from the synthesis machinery to target apoproteins (Bandyopadhyay et al. 2008). The upregulation of GRXs in Fe-deficient barley leaves (Tables 2 and 3) might be important not only to overcome the limited reducing power from photosynthesis (Forner-Giner et al. 2010) but also to facilitate synthesis of [Fe-S] proteins under Fe-deficient conditions.

GSTs were also upregulated up to three-fold by Fe deficiency (Tables 2 and 3). Glutathionylation of protoporphyrinogen, which is a precursor of chlorophylls and heme, can be catalyzed by maize GSTs in vitro (Dixon et al. 2008), suggesting the possibility of direct involvement of GSTs in Fe homeostasis and photosynthesis. Thus, GSTs in Fe-deficient barley may contribute to better tolerance to Fe deficiency in leaves. Increased level of a GST transcript has been found in Fe-starved alfalfa (Arakawa et al. 2002), indicating that a common scenario of GST function under Fe deficiency may be observed in land plants.

In addition to GRXs and GSTs, 2 glutathione reductase (GR) genes, HvGR1 (chloroplastic) and HvGR2 (cytosolic), have been reported as Fe deficiency-responsive genes (Bashir et al. 2007). The investigators reported that the genes might be involved in internal Fe homeostasis and scavenging reactive oxygen species. However, we could not detect the responses of HvGR1 and HvGR2 to Fe deficiency in leaves in the present study (Tables 2 and 3). This discrepancy was thought to be due to the weak response of HvGR1 to Fe deficiency and the weak expression level of HvGR2 in shoot tissues, as indicated in the report of Bashir et al. (2007) (i.e. GR activity was increased by only 1.27-fold in the shoot tissues of Fe-deficient barley). Nevertheless, the response of the GR genes to Fe deficiency was insignificant even in the barley leaves. The steady levels of the HvGR transcripts under Fe-deficient conditions may indicate the importance of these genes in maintaining the cellular redox status under Fe-deficient conditions.

The expression patterns of the majority of genes shown in Tables 2–5 were similar between 14-day and 25-day Fe-deficiency treatments. However, N and S assimilation-related genes were downregulated by 14-days Fe deficiency (Table 2) and by eight-days Fe deficiency (Figure 2B) but not by 25-days Fe deficiency (Table 3). RuBisCO protein accumulation was decreased by eight-days Fe deficiency (Figure 2A), while both its small and large subunit genes were upregulated by 14-days Fe deficiency (Table 2) and showed no change in expression after 25 days of Fe deficiency (Table 3) in barley leaves. A cycle of leaf-color change, severe chlorosis, and re-greening can be observed during long-term Fe deficiency in barley but not in rice (Maruyama et al. 2005; Hirai et al. 2007). In our growth chamber or greenhouse, severe chlorosis was observed at about two weeks, and new green leaves developed at about three weeks in Fe-deficient barley after transplantation in Fe-deficient nutrient solution. The highest chlorophyll a/b ratio was observed at 12 days, and this ratio decreased to the control level at 20 days (Saito et al. 2010). Primary C, N, and S assimilation-related genes could be transiently downregulated to reduce the consumption of reducing power during the early stage of adaptation to Fe deficiency and restored to the normal level after complete adaptation to Fe deficiency.

Alteration of metabolism as mentioned earlier could contribute to mugineic acid synthesis in Fe-deficient barley. In the younger-Fe leaves of barley, N and S assimilation may be suppressed, and insufficient organic N in these leaves may be supplemented with remobilized organic N from the older-Fe leaves, accompanying senescence enhanced by Fe deficiency. The organic N released from the older-Fe leaves may also be remobilized to the roots. Moreover, the older-Fe leaves may assimilate S, and the assimilated S may be remobilized to the roots and contribute to mugineic acid synthesis with remobilized C and N. Mori et al. (1991) reported that the secretion of mugineic acids by rice ceased within 10 days after Fe-deficiency treatment. Rice could not remobilize and supply enough substrates to synthesize a large amount of mugineic acids. Tracer experiments and metabolomic analyses will clarify the adaptive mechanisms for substance distribution in barley in Fe deficiency.

Disturbance in the metabolism of fundamental organic compounds could influence other minerals. Total P and total Ca concentrations were increased by Fe deficiency in both barley and rice (Tables 7 and 8). Such an increase has also been reported previously (Fleming and Foy 1982). However, water-extractable and Ca²⁺ concentrations were dramatically increased in both the younger and older leaves of Fe-deficient rice, but not in Fe-deficient barley. Thus, the metabolism of major essential elements was significantly disturbed in rice by only the eight day of Fe-deficiency treatment not only in the younger leaves but also in the older leaves. Phosphorus decreases the concentration of water-soluble Fe and promotes Fe-deficiency symptoms in rice (Ladouceur et al. 2008; Zheng et al. 2009). We have previously reported that Fe deficiency dramatically decreases the ratio of water-soluble Fe to total Fe in rice shoots, but not in barley (Maruyama et al. 2005). The increase in concentration in rice may also be a reason for the susceptibility of rice to Fe deficiency.

Our findings suggest that barley possesses a comprehensive mechanism for Fe-deficiency tolerance. Elucidation of such a mechanism will provide a cue for improving low Fe-deficiency tolerance and potentially enhancing Fe utilization efficiency in crops.