Mugineic acid family phytosiderophores (MAs) are metal chelators that are produced in graminaceous plants in response to iron (Fe) deficiency, but current evidence regarding secretion of MAs during zinc (Zn) deficiency is contradictory. Our studies using HPLC analysis showed that Zn deficiency induces the synthesis and secretion of MAs in barley plants. The levels of the HvNAS1, HvNAAT-A, HvNAAT-B, HvIDS2 and HvIDS3 transcripts, which encode the enzymes involved in the synthesis of MAs, were increased in Zn-deficient roots. Studies of the genes involved in the methionine cycle using microarray analysis showed that the transcripts of these genes were increased in both Zn-deficient and Fe-deficient barley roots, probably allowing the plant to meet its demand for methionine, a precursor in the synthesis of MAs. In addition, HvNAAT-B transcripts were detected in Zn-deficient shoots, but not in those that were deficient in Fe. Increased synthesis of MAs in Zn-deficient barley was not due to a deficiency of Fe, because Zn-deficient barley accumulated more Fe than did the control plants, ferritin transcripts were increased in Zn-deficient plants, and Zn deficiency promoted Fe transport from root to shoot. Moreover, analysis using the positron-emitting tracer imaging system (PETIS) confirmed that more 62Zn(II)-MAs than 62Zn2+ were absorbed by the roots of Zn-deficient barley plants. These data suggest that the increased biosynthesis and secretion of MAs arising from a shortage of Zn are not due to an induced Fe deficiency, and that secreted MAs are effective in absorbing Zn from the soil.
Zinc (Zn) is an essential element for living organisms, including higher plants, being a necessary co-factor in many enzymes and proteins involved in cell division, nucleic acid metabolism, protein synthesis and gene expression (Marschner, 1995). In plants, Zn deficiency results in extensive oxidative damage to membranes, lipids, proteins, chlorophyll and nucleic acids (Cakmak, 2000). In many regions of the world, however, particularly those with calcareous soils, little Zn is present in the soil solution, which results in reduced plant growth. The Food and Agriculture Organization has reported that 30% of the world's cultivated soils are deficient in Zn (Hacisalihoglu and Kochian, 2003; Sillanpää, 1990).
Higher plants take up Zn from the rhizosphere via transporters, and the molecular aspects of this phenomenon have begun to be clarified. Several Zn2+ transporter genes have been isolated from Arabidopsis thaliana (Grotz et al., 1998) and Glycine max (Moreau et al., 2002). Halvorson and Lindsay (1977) hypothesized that, in most graminaceous plants, it is free Zn2+ ions that participate in the Zn influx into the root symplast. Recently, three functional Zn2+ transporters were identified from Oryza sativa (Ishimaru et al., 2005; Ramesh et al., 2003). In contrast, Welch (1995) proposed that mugineic acid family phytosiderophores (MAs) play a major role in iron (Fe) acquisition, and may also contribute to the acquisition of Zn and other metal nutrients by graminaceous plants. The YS1 gene, which encodes the Fe(III)-MAs transporter, has been isolated from Zea mays; Fe deficiency increases the expression of the gene in both shoots and roots (Curie et al., 2001). von Wirén et al. (1996) reported that the ys1 mutant absorbs a lower amount of 65Zn(II)-MAs than the wild-type does, and recently Schaaf et al. (2004) showed that ZmYS1 transports Zn(II)-MAs in addition to Fe(III)-MAs in yeast and Xenopus oocytes.
It was reported that Zn deficiency increases the secretion of MAs from wheat and barley roots into the rhizosphere (Cakmak et al., 1994; Walter et al., 1994; Zhang et al., 1989). However, other studies have shown that significant secretion of MAs from barley or wheat only occurs in plants deficient in Fe, and not in those suffering Zn deficiency (Gries et al., 1995; Pedler et al., 2000). Walter et al. (1994) hypothesized that impairment of Fe transport from the roots to the shoot under Zn-deficient conditions results in Fe deficiency, leading to the secretion of MAs. They also suggested that Zn-deficient plants accumulate high levels of phosphorus (P), which precipitates as an iron phosphate, causing Fe deficiency.
MAs are synthesized from methionine (Mori and Nishizawa, 1987). S-adenosyl-methionine (SAM) synthetase converts methionine into SAM. Subsequently, three molecules of SAM are combined to form one molecule of nicotianamine (NA) by nicotianamine synthase (NAS). NA is then converted to 3′′-keto acid by NA aminotransferase (NAAT), and 2′-deoxymugineic acid (DMA) is synthesized by the subsequent action of a reductase. In barley, a further series of hydroxylations of DMA is catalyzed by two dioxygenases, IDS2 and IDS3 (iron-deficiency-specific clones 2 and 3; Kobayashi et al., 2001; Nakanishi et al., 2000). We have isolated the genes that encode most of the enzymes in the biosynthetic pathway of MAs from barley roots: HvSAMS (Takizawa et al., 1996), HvNAS1-7 (Higuchi et al., 1999), HvNAAT-A, HvNAAT-B (Takahashi et al., 1999), HvIDS2 (Okumura et al., 1994) and HvIDS3 (Nakanishi et al., 1993). The expression of each of these genes is dramatically enhanced in roots by Fe deficiency, resulting in increased secretion of MAs. Ma et al. (1995) proposed that the methionine cycle functions to recycle the methionine that is required for the continuous synthesis of MAs. In fact, Fe deficiency upregulates the genes that encode Fe-deficiency-induced protein 1 (IDI1; Yamaguchi et al., 2000a), formate dehydrogenase (FDH; Suzuki et al., 1998) and adenine phosphoribosyltransferase (APRT; Itai et al., 2000), enzymes that are closely related to the methionine cycle in barley roots.
In this study, we show that roots of Zn-deficient barley plants have increased expression of genes involved in the biosynthesis of MAs, and that there is increased secretion of MAs from the roots of these plants. In addition to genes previously reported to be involved in the methionine cycle, other genes putatively involved in the methionine cycle are upregulated by deficiencies of Zn or Fe. To investigate whether Zn-deficient barley is deficient in Fe, as hypothesized by Walter et al. (1994), we measured the concentration of Fe, the pattern of ferritin expression, and the translocation of Fe in Zn-deficient barley. From these results, we conclude that Zn-deficient barley is not deficient in Fe. Furthermore, we show that more Zn(II)-DMA than Zn2+ was absorbed in Zn-deficient barley, suggesting that DMA secreted as a result of Zn deficiency is effective in absorbing Zn from the soil.
Zn-deficient and Fe-deficient barley plants: features of growth and secretion of MAs
Under the conditions used in this study, we observed no marked differences in the appearance or dry weight of the Zn-deficient and Zn-sufficient barley plants after 2 weeks of treatment (Figure 1). However, barley plants grown under low Zn conditions for 2 weeks secreted 4.7 times more MAs than control plants (Table 1). Brown necrotic spots, which are typical of Zn deficiency, were observed in the older leaves of plants after 17 days of growth under low Zn conditions. After 4 weeks, the dry weight of Zn-deficient barley shoots was about half that of control plants, but the dry weight of the roots was similar to that of the controls. The total amount of MAs secreted from Zn-deficient barley was two to three times higher than that from Zn-sufficient barley.
Table 1. The amounts of 2′-deoxymugineic acid (DMA), mugineic acid (MA), epi-hydroxymugineic acid (epi-HMA), and the total amount of mugineic acid family phytosiderophores (MAs) calculated as the sum of DMA, MA and epi-HMA secreted from barley roots of control plants adequately supplied with nutrients and those suffering deficiency of Zn (−Zn) or Fe (−Fe) after 2 and 4 weeks of treatment
Values are nmol plant−1. The data are the mean ± SE of three independent replicates. The values marked with asterisks are significantly different compared to their respective control plants using a one-tailed Student's t-test: *P < 0.05, **P < 0.01.
0.3 ± 0.3
0.0 ± 0.0
3.5 ± 1.7
3.8 ± 2.0
1.6 ± 0.3*
0.2 ± 0.1*
15.9 ± 4.5*
17.7 ± 4.9*
53.5 ± 10.2**
15.5 ± 3.6**
395.8 ± 50.2**
464.8 ± 57.6**
4.8 ± 0.8
0.2 ± 0.1
18.8 ± 4.3
23.7 ± 5.0
7.6 ± 2.6
0.9 ± 0.1**
53.2 ± 13.7*
61.8 ± 16.0*
283.3 ± 65.2
236.7 ± 58.7**
894.8 ± 73.0**
1414.7 ± 171.2**
In contrast to the effects of low Zn supply, barley plants began to show signs of chlorosis, a typical symptom of Fe deficiency, after 1 week of Fe depletion. The dry weight of the Fe-deficient roots was much lower than that of the control roots after 4 weeks, and these roots secreted 60 times more MAs than did the roots of control plants. While Zn deficiency did increase the secretion of MAs, the total amount of MAs secreted by the roots of Fe-deficient plants was >20 times higher than that from roots of Zn-deficient plants.
Expression of genes involved in the synthesis of MAs in barley
MAs are synthesized from SAM through several steps (Figure 2a). Northern blot analysis was used to examine the production of transcripts of five genes involved in the synthesis of MAs in Zn-deficient barley. The levels of HvNAS1, HvNAAT-A, HvNAAT-B, HvIDS2 and HvIDS3 were increased in barley roots that had been grown under low Zn conditions for 2 and 4 weeks, although they were lower than the amounts observed in Fe-deficient plants (Figure 2b). These results are strong indications that MAs are synthesized in barley roots in response to Zn deficiency. The expression in roots of NASHOR2, one of the barley NAS genes (Herbik et al., 1999), differed from that of the other genes, in that it was increased by Zn deficiency but not by Fe deficiency (Figure 2c).
NASHOR2 was expressed in control barley shoots, with expression being increased by Zn deficiency but decreased by Fe deficiency (Figure 2c). The expression of HvNAAT-B was also detected in Zn-deficient shoots, but not in those of the control or Fe-deficient plants (Figure 2b). Slight expression of HvNAS1 and HvNAAT-A was detected in Zn-deficient shoots. These results suggest that DMA is produced in Zn-deficient barley shoots, but not in those suffering Fe deficiency. It should be noted that HvIDS2 and HvIDS3 were not expressed in Zn-deficient shoots, suggesting that only DMA may be produced in Zn-deficient barley shoots despite other MAs, such as mugineic acid (MA) and epi-hydroxy mugineic acid (epi-HMA), being produced in barley roots grown at low Zn and Fe levels.
Microarray analysis of gene expression in barley roots in response to Zn and Fe deficiency
A microarray containing 8987 rice ESTs (9 K microarray) was used to analyze the gene expression profile of barley roots grown under low-Zn or low-Fe stress (Table 2). Genes of cereal crops tend to be highly conserved at the DNA sequence level (Devos and Gale, 2000), and this conservation allows the use of heterologous probes to identify orthologous DNA sequences in DNA hybridization experiments in barley roots (Negishi et al., 2002). The barley plants used in microarray analysis were 20 days old, and had been grown under a Zn- or Fe-deficient treatment for 2 weeks. The induction ratios were calculated for Zn- or Fe-deficient conditions, compared with the control. The profile of the genes with ratios >1.70 or ratios <0.60 is shown in Tables S1 and S2). The results of the microarray analysis showed that the ratios of both NAS and NAAT under Zn deficiency were 1.7. As Northern blot analysis clearly showed upregulation of these genes in Zn-deficient roots (Figure 2b), we considered that the genes for which the ratio was >1.70 were ‘upregulated’. In addition to the increased transcription of genes that participate in MA synthesis, the expression of a gene homologous to YS1, which encodes the transporter of the Fe(III)-MAs and Zn(II)-MAs complexes in maize, was also increased. The IDI1, APT and FDH genes, which have been reported to be involved in the methionine cycle (Itai et al., 2000; Suzuki et al., 1998; Yamaguchi et al., 2000a), were also upregulated in Zn-deficient barley roots. Upregulation was also evident for methylthioribose-1-phosphate isomerase (IDI2) (Ashida et al., 2003; Yamaguchi et al., 2000b) and dehydrase–enolase–phosphatase, which are putatively involved in the methionine cycle. The expression of these genes also increased dramatically in response to Fe deficiency, with induction ratios higher than those in response to Zn deficiency. Putative aspartate/tyrosine/aromatic aminotransferase (IDI4, accession number AB206815), which has been predicted to be the enzyme that catalyze the conversion of 2-keto-4-methylthiobutyric acid (KMTB) to methionine by transamination (Berger et al., 2003; Heilbronn et al., 1999), was also upregulated by Zn deficiency and Fe deficiency. Methylthioribose (MTR) kinase was upregulated by Fe deficiency but not by Zn deficiency.
Table 2. The ratio of Zn or Fe deficiency-inducible genes related to the synthesis of MAs and the putative transporter of Fe(III)-MAs and Zn(II)-MAs in barley roots on microarray analysis
Rice accession number
Putative gene identification
Barley accession number
The induction ratios were calculated as the relative increases in expression under conditions of Zn or Fe deficiency, compared with the expression under control conditions. The ratios are the mean of two independent replicates. Accession numbers were predicted based on the rice sequences used in microarray analysis. The E-value indicates the differences in sequences between rice and barley using blast. Accession numbers marked with an asterisk indicates full-length cDNAs previously isolated from Fe-deficient barley roots.
Barley plants were examined to determine whether Zn or Fe deficiency affects the concentrations of other metals (Figure 3). After 2 weeks of low Zn supply, the Fe concentration in barley shoots was similar to that in control shoots. After 4 weeks, however, the Fe concentration in shoots was about three times higher in Zn-deficient barley than in control plants. This is consistent with previous reports that Fe concentration in shoots is increased in Zn-deficient plants (Pedler et al., 2000; Walter et al., 1994). Low Fe supply increased the Zn concentration in barley shoots, with a 1.4-fold increase after 2 weeks, and a 13-fold increase after 4 weeks. The manganese (Mn) concentration in both Zn- and Fe-deficient barley shoots was higher than in control plants after 4 weeks of treatment. The copper (Cu) concentration was also increased in Fe-deficient shoots, but was lower in Zn-deficient shoots. The Fe concentration in Zn-deficient roots was almost the same as that in control roots (data not shown).
The ferritin gene is expressed in Zn-deficient barley
Microarray analysis showed that a gene encoding a putative ferritin was upregulated twofold in Zn-deficient barley shoots and was reduced by half in Fe-deficient barley shoots (data not shown). Although the barley ferritin gene has not been isolated, the sequence of the barley EST (AF415241) for ferritin has 88% homology to that of the rice ferritin full-length cDNA (AK059354), and the partial sequence of AK059354 is present on this rice 9 K array. Fe regulates ferritin synthesis at the transcriptional level, and an increase in available Fe induces ferritin expression in plants (Lescure et al., 1991; Petit et al., 2001). To confirm the results of the microarray analysis, expression of the ferritin gene was analyzed by Northern blotting, which revealed that transcripts of the ferritin gene increased in Zn-deficient barley shoots (Figure 4a). The transcript level in Zn-deficient barley shoots was higher after 4 weeks than after 2 weeks, consistent with the increase in Fe concentration shown in Figure 3. In addition, although the transcript level of the ferritin gene in roots was much lower than in shoots, the expression was slightly increased in Zn-deficient roots after 4 weeks of treatment (Figure 4b). This increase indicated that Zn-deficient barley was not deficient in Fe, but rather contained an increased amount of Fe in shoots.
Zn deficiency promotes translocation of Fe from roots to shoots
The translocation of 52Fe in Zn-deficient and Zn-sufficient barley was monitored using petis to investigate whether Fe transport from roots to shoots is affected by Zn deficiency. Typical images of the plants used for the petis analysis are shown in Figure 5(a). Zn-deficient barley shoots was found to accumulate much more 52Fe than control barley shoots (Figure 5b). The 52Fe radioactivity in Zn-deficient barley roots was almost the same as in control barley roots. These results indicate that Fe transport is promoted in Zn-deficient barley, resulting in the overall increase in the Fe concentration evident in Figure 3. Real-time imaging of the translocation of 52Fe is shown in Figure 5(c); continuous imaging is shown in Movie S1.
Zn(II)-DMA is the preferable form for Zn absorption in Zn-deficient barley
The roots of Zn-deficient barley plants were supplied with Zn(II)-DMA and Zn2+ ions to examine whether or not Zn(II)-DMA is absorbed. In both cases, most of the Zn was accumulated in the discrimination center, which is the basal part of the shoot, and the sheath (Figure 6a), with about 1.6 times as much 62Zn accumulated in the shoot when 62Zn(II)-DMA was supplied (Figure 6b). The 62Zn radioactivity in roots when 62Zn(II)-DMA was supplied was slightly higher than when Zn2+ was supplied. Real-time imaging of the translocation of 62Zn is shown in Figure 6(c); continuous imaging is shown in Movie S2. Real-time imaging also indicated that more 62Zn was translocated to shoots when 62Zn(II)-DMA was supplied to Zn-deficient barley roots than when 62Zn2+ ions were supplied.
Increased secretion of phytosiderophores has been associated with Fe deficiency in graminaceous plants (Takagi, 1976; Takagi et al., 1984), but there is conflicting data as to whether or not this occurs under conditions of low Zn supply. Our studies have shown that Zn deficiency induces both the secretion of MAs and the expression of genes involved in the synthesis of MAs in barley roots (Table 1, Figure 2b). It is noteworthy that the expression of genes involved in the synthesis of MAs was induced in barley grown for 2 weeks in low Zn conditions (Figure 2b), indicating both increased synthesis and secretion of MAs in barley before the appearance of Zn-deficiency symptoms or growth inhibition.
The amount of secreted MAs in Zn-deficient barley was lower than that in Fe-deficient barley. Detection of low amounts of MAs requires a highly sensitive method for analysis. In the present study, we used HPLC analysis that we have developed to measure the amount of secreted MAs (Mori et al., 1987). This HPLC analysis enabled us to detect a very low amount of MAs and distinguish each kind of MAs secreted from Zn-deficient barley. In contrast, a Cu-mobilization assay or Fe-binding assay was used to detect the secretion of phytosiderophore in the previous reports that did not find an increased secretion of phytosiderophore in Zn-deficient wheat and barley (Gries et al., 1995, 1998; Pedler et al., 2000). These assays are indirect methods to measure the amount of phytosiderophore and cannot distinguish MAs from other metal-chelating substances. This could be one reason that there is conflicting data as to whether or not MA secretion increases under conditions of low Zn supply.
Genes involved in the methionine cycle are upregulated by Zn and Fe deficiencies
Methionine is the precursor of MAs (Mori and Nishizawa, 1987). Ma et al. (1995) proposed that the methionine cycle functions to recycle the methionine that is required for continuous synthesis of MAs. Recently, the all enzymes involved in the methionine salvage pathway have been characterized in Bacillus subtilis (Sekowska et al., 2004). In higher plants, however, the pathway from methylthioribose (MTR)-1-phosphate (MTR-1-P) to KMTB in the methionine cycle (Figure 7) has not been clarified.
In contrast to Bacillus subtilis, higher plants might convert methylthioribulose-1-phosphate to 1,2-dihydroxy-3-keto-5-methylthiopentene by one enzyme. Sekowska et al. (2004) noted that the putative enolase–phosphatase in A. thaliana has a domain with dehydrase activity, while enolase–phosphatase and dehydrase are separated in other organisms such as bacteria, yeasts and mammals. The putative enolase–phosphatase in rice also contains both domains of enolase–phosphatase and dehydrase. Therefore, we presume that these three steps, dehydration, enolization and dephosphorylation, may be performed by a single enzyme in higher plants. In addition, we suppose that IDI4, the putative aspartate/tyrosine/aromatic aminotransferase induced by Fe deficiency, is a candidate enzyme for the final step of the methionine cycle (Berger et al., 2003; Heilbronn et al., 1999). It has been confirmed by Northern blot analysis that transcript levels of all the candidate genes participating in this pathway (Figure 7) were increased in Fe-deficient rice roots (Kobayashi et al., 2005).
Our microarray analysis suggests that all enzymes involved in the proposed methionine cycle are upregulated by Fe deficiency, and while the ratios are lower with Zn deficiency, they are nevertheless upregulated also. The difference in the ratios may reflect the amount of MA synthesis based on the demand for methionine.
DMA is synthesized in Zn-deficient shoots
In barley shoots, the expression of NASHOR2 and HvNAAT-B was increased in response to Zn deficiency, but this did not occur in response to Fe deficiency (Figure 2b,c). Interestingly, no HvIDS2 or HvIDS3 expression was detected in Zn-deficient barley shoots. Many kinds of MAs, including DMA, MA and epi-HMA, were synthesized in barley roots in response to Fe or Zn deficiency. In contrast, our Northern blot analysis suggested that only DMA is synthesized in the shoots of Zn-deficient barley, indicating that only Zn deficiency induced DMA synthesis in barley shoots, although both Zn and Fe deficiency induced the synthesis and secretion of MAs in barley roots. Mizuno et al. (2003) proposed that NAS enzymes in graminaceous plants be classified into two groups based on the phylogenic tree. So far, 15 NAS genes have been reported in barley, rice and maize. NASHOR2, OsNAS3 and ZmNAS3, which belong to the same group, are expressed constitutively in shoots, and their expression is decreased by Fe deficiency (Figure 2c; Inoue et al., 2003; Mizuno et al., 2003). In contrast, Zn deficiency increased the expression of NASHOR2 (Figure 2c) and OsNAS3 (data not shown). Inoue et al. (2003) showed that OsNAS3 was expressed preferentially in pericycle and companion cells, and suggested that the physiological function of OsNAS3 differs from that of OsNAS1 and OsNAS2, which are involved in MA synthesis for secretion into the rhizosphere. It has been hypothesized that the synthesis of MAs for secretion occurs in particular vesicles derived from the endoplasmic reticulum (Negishi et al., 2002). Mizuno et al. (2003) showed that ZmNAS1 and ZmNAS2 were localized at these vesicles, while ZmNAS3 was localized in the cytoplasm. These data suggest that OsNAS3 and ZmNAS3 are not involved in MA secretion from the roots, and, similarly, that NASHOR2 may not be involved in MA secretion. We speculate that NA or DMA synthesized in shoots in response to Zn deficiency is involved either in the long-distance transport of Zn or in increasing the availability of Zn in the cytoplasm; a survival response to Zn-deficiency stress.
Zn-deficient barley is not deficient in Fe
Walter et al. (1994) suggested that Zn deficiency induces Fe deficiency by impairing Fe transport or precipitating Fe as iron phosphate, thus increasing the secretion of MAs. Our data, however, do not support this contention, as Zn-deficient barley plants accumulated more Fe in shoots than did the controls (Figure 3). Furthermore, Northern blot analysis showed that expression of the ferritin gene increases in Zn-deficient barley shoots and roots (Figure 4a,b). Plant ferritins are plastid proteins whose abundance is strictly controlled at the transcriptional level by the Fe status of the cells. These results indicate that the secretion of MAs by Zn-deficient plants was not caused by an induced Fe deficiency. Moreover, our data from the PETIS analysis showed that Fe transport is increased in Zn-deficient barley (Figure 5). These results lead us to propose that Zn-deficient barley is not deficient in Fe, and that the increased secretion of MAs as a result of Zn deficiency is not due to Fe deficiency. On the other hand, recent work in A. thaliana has shown that nitric oxide mediates iron-induced ferritin accumulation (Murgia et al., 2002). Therefore, we cannot exclude the possibility that Zn deficiency directly induces nitric oxide formation and that this results in ferritin accumulation.
Zn uptake and transport in graminaceous plants
More Zn(II)-DMA than Zn2+ ions was absorbed by Zn-deficient barley plants (Figure 6b,c) suggesting that MAs secreted in response to Zn deficiency are effective in absorbing Zn from the soil. In general, the concentration of Zn in the soil solution, particularly at high soil pH, is very low because most of the Zn occurs on the surface of clays, as hydrous oxides, and in organic matter (Marschner, 1993). We have also demonstrated that Zn accumulation is increased in Fe-deficient barley and vice versa (Figure 3). Both Fe and Zn deficiencies cause the secretion of MAs, with greater amounts secreted with Fe deficiency. The ratios of the amounts of MAs secreted from Fe-deficient and Zn-deficient barley plants were similar to the ratio of the Zn and Fe concentrations in barley shoots (Table 1, Figure 3). This suggests that the secretion of MAs from barley roots contributes to both Zn and Fe uptake. In addition, the PETIS experiment showed that Zn-deficient barley promotes the translocation both of Fe(III)-DMA and Zn(II)-DMA (Figures 5b and 6b). Northern blot analysis suggests that Zn-deficient barley plants synthesize DMA in shoots (Figure 2b,c). Therefore, MAs synthesized in Zn-deficient plants might contribute to metal translocation as well as metal uptake.
In maize, Schaaf et al. (2004) showed that ZmYS1 transports Zn(II)-MAs in addition to Fe(III)-MAs, and that ZmYS1 also transports Cu(II)-MAs. Gries et al. (1998) reported that Cu deficiency induced MA secretion in the calcicole grass Hordelymus europaeus.
Our petis analysis showed that Zn2+ ions were also absorbed into roots (Figure 6b). von Wirén et al. (1996) showed that the ys1 mutant absorbs Zn2+ in maize, indicating that Zn is also absorbed as Zn2+ ions via a specific transporter. At least nine genes encoding putative ZIP family transporters of barley are found in the database (http://harvest.ucr.edu/). Five rice ESTs of ZIP family transporter are present on the 9 K array used in our microarray analysis. The ratios of the genes encoding ZIP family transporters were 1.1 (OsZIP1) and 1.5 (OsZIP5 and OsIRT2). The data for two genes (OsZIP3 and OsZIP4) were filtered because of low expression signal or no reproduction on the duplicated spotted array. Therefore, we could not identify the Zn-deficiency-induced ZIP family transporters in barley. Presumably, there are genes encoding the functional Zn2+ transporter in barley that function in the uptake of Zn2+ from the rhizosphere. Zn may be absorbed from soil as both Zn2+ ions and Zn(II)-MAs, with the ratio of uptakes differing among species.
Plant material and growth conditions
Barley seeds (Hordeum vulgare L. cv. Ehimehadaka no. 1) were germinated for 3 days at room temperature on paper soaked with distilled water. After germination, the seedlings were transferred to a Saran net floating on distilled water in a growth chamber (day: 25°C, 14 h of light at 320 μmol photons m−2 sec−1; night: 10 h at 20°C). After 3 days, 14 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 μm H3BO3, 0.5 μm MnSO4, 0.2 μm CuSO4, 0.5 μm ZnSO4, 0.05 μm Na2MoO4, and 0.1 mm Fe-EDTA. The ZnSO4 was omitted from the solution to induce Zn deficiency, and the Fe-EDTA was omitted from the solution to induce Fe deficiency. The pH of the nutrient solution was adjusted daily to 5.5 using 1 m HCl, and the nutrient solution was renewed weekly. Root exudates were collected and the plants were harvested after treatment with the various nutrient solutions for 2 or 4 weeks.
Collection of root exudates and HPLC analysis of MAs
Four plant roots were rinsed with deionized water and then soaked in 800 ml of deionized water 1 h before illumination. Root exudates were collected for 5 h. The deionized water was renewed 2 h after starting the collection. The antimicrobial agent Micropur (Katadyn Products Inc., Wallisellen, Switzerland) was added to the water to prevent microbial degradation of the MAs after collection. After the second sampling, both root exudates were combined and filtered through filter paper (Advantec 5C, Toyo Roshi Kaisha Ltd, Tokyo, Japan). The cationic fraction of the root exudate was prepared as a 2 m NH4OH eluate from Amberlite IR(H+)120 (Rohm and Haas Co., Philadelphia, PA, USA). Condensed and microfiltered samples were subjected to HPLC analysis as described previously (Mori et al., 1987). The total amount of MAs was calculated as the sum of DMA, MA and epi-HMA. DMA, MA and epi-HMA were purified from root exudates and used as standards. The experiments were performed independently three times.
Determination of metal concentrations
After collection of the root exudates, the plants were dried for 1 week at 65°C. The plants (30–50 mg) were then wet-ashed with 2 ml of 11 m HNO3 for 5 h at 150°C. The metal concentrations were measured using inductively coupled plasma atomic emission spectrometry (SPS1200VR; Seiko, Tokyo, Japan) at wavelengths (nm) of 238.204 (Fe), 213.856 (Zn), 293.930 (Mn) and 324.754 (Cu).
Northern blot analysis
Total RNA was extracted from roots and shoots, and 10 or 20 μg per lane were electrophoresed in 1.2% w/v agarose gels containing 0.66 m formaldehyde, transferred to Hybond-N+ membrane (Amersham, Biosciences UK Ltd., Buckinghamshire, UK), and hybridized with probes at 65°C. Northern blots were analyzed using BAS-3000 (FujiFilm, Tokyo, Japan). The following specific primers were designed for each gene and used to prepare specific probes: HvNAAT-A, 5′-CATATTGTAATGGTTCTGTTGTAGCTGTCC-3′ and 5′-CACAACCATTTTTATTGAAATTGATGCAA-3′; HvNAAT-B, 5′-TGGAACTTTTAGTTCTCTATGAATAGA-3′ and 5′-AACATAACATTCACTATGTTTCGTCCAGG-3′; HvIDS2, 5′-GATTTGCATCACTTCCTCGATCGTTCG-3′ and 5′-CAGGTACGTACGATCTCATCACATGTCACG-3′; HvIDS3, 5′-CAAATCTACAAGACTCCTTCAAGATCTGGA-3′ and 5′-TTTATTGCTATAAATCATCAAAG-3′, NASHOR2 5′-GTGATGAATTCACCCACCGATTA-3′ and 5′-AAATAATCCAGCCATATGCAGAC-3′. The accession numbers of the genes are D88273 (HvNAAT-A), AB005788 (HvNAAT-B), D10057 (HvIDS2), AB024058 (HvIDS3) and AF136942 (NASHOR2). The ORF sequence of HvNAS1 (AB010086) was used to design the probe, which may detect the expression of other HvNAS genes. The probe used for HvNAAT-B may hybridize with HvNAAT-A. The EST of the ferritin gene of rice (clone SS6280 from the Rice Genome Resource Center, Tsukuba, Japan) was also used to design probes. The EST of barley ferritin was cloned using ReverTra Ace (Toyobo, Osaka, Japan) and the primers 5′-AGGCTCCAGTCAATTGTCAC-3′ and 5′-CAACCTGCTCCTGAAGGAAT-3′.
Total RNA was extracted from roots of barley grown in nutrient solution for 2 weeks, as described above. Microarray analysis was performed as described on the Rice Microarray Opening Site (http://cdna01.dna.affrc.go.jp/RMOS/pdf/array_protocol_e.pdf). The total RNA was reverse-transcribed with Cy3 or Cy5 dCTP (GE Healthcare Bio-sciences Corp., Piscataway, NJ, USA), and hybridized with microarray glass slides containing 8987 rice ESTs (g_array). Microarray slides were scanned in both Cy5 and Cy3 channels with an Array Scanner Generation III (GE Healthcare Bio-sciences Corp.). The fluorescence intensity for each element on the array was captured using ArrayGauge version 1.21 (FujiFilm). Data processing was performed using the PRIM program (Kadota et al., 2001), and median normalization was performed according to the method of the MGED Data Transformation and Normalization Working Group (http://genome-www5.stanford.edu/mged/normalization.html). The hybridization was performed twice, and the mean induction ratio of genes in Zn- or Fe-deficient roots was calculated.
52Fe translocation using PETIS analysis
Hamamatsu Photonics of Japan and the Takasaki Ion Accelerators for Advanced Radiation Application (tiara) group have developed a dynamic image measurement system termed petis, which uses positron-emitting nuclides and enables the investigator to study biological processes in plants in real time (Kume et al., 1997). The growth conditions of plants used for petis analysis were as described above, except that the temperature was 22°C during the day and 17°C at night. Prior to the analysis, barley plants were grown for 2 weeks in complete nutrient solution or in solution lacking Zn. 52Fe (half life: 8.27 h) was produced using the method described by Watanabe et al. (2001). After the pH of the 52Fe3+ solution had been adjusted to about three using 1 m KOH, the 52Fe3+ was chelated with 197.4 μmol DMA in darkness for 1 h. Zn-deficient and Zn-sufficient barley plants were each supplied with 16 ml of culture solution lacking Fe in a polyethylene bag. The plants and bags were fixed between two acrylic boards and placed between a pair of petis detectors in a chamber, under conditions of 25°C, 65% humidity, and a light density of 320 μmol m−2 sec−1. 52Fe3+-DMA (62.8 fmol) was added to each culture solution. The final concentration of Fe-DMA was 3.9 pm. After PETIS analysis for 8 h, the plants were removed from the polyethylene bags, and the roots were gently washed for 1 min in 50 ml 0.01 mm EDTA. Next, each plant was placed under a bio-imaging plate inside a cassette. The plate was scanned using an image analysis system (BAS-1500, FujiFilm). The plants were then cut into parts, and the absolute amount of radioactivity was analyzed using γ-ray spectrometry with an ORTEC HPGe detector (Seiko EG & G Co. Ltd, Tokyo, Japan).
62Zn translocation using petis analysis
The method for 62Zn was similar to that described for the 52Fe translocation PETIS analysis. Before the analysis, barley plants were grown for 2 weeks in nutrient solution lacking Zn. 62Zn (half life: 9.13 h) was produced according to the method described by Watanabe et al. (2001). After the pH of the 62Zn2+ solution has been adjusted to about 4 using 1 m KOH, the 62Zn2+ was chelated with DMA in darkness for over 3 h. 62Zn (0.8 pmol) with DMA (3.36 μmol) or without DMA was added to each culture solution. The final concentrations of Zn-DMA or Zn2+ were 50 pm. The period of 62Zn absorption was 8 or 10 h. The analysis was conducted as described for the PETIS analysis of 52Fe. All 52Fe and62Zn experiments were repeated at least three times to confirm the reproducibility of the results.
We thank Dr Yoshiaki Nagamura (National Institute of Agrobiological Sciences, Tsukuba, Japan) and the Rice Genome Project (Tsukuba, Japan) for providing cDNA clones, Dr Haruhiko Inoue and Dr Takanori Kobayashi for discussion, and Dr Pax Blamey for assistance with English.