An Al-inducible MATE gene is involved in external detoxification of Al in rice

Authors


(fax +81 86 434 1209; e-mail maj@rib.okayama-u.ac.jp).

Summary

A number of plant species, including rice, secretes citrate from roots in response to Al stress. Here we characterized the functions of a gene, OsFRDL4 (Os01g0919100) that belongs to the multidrug and toxic compound extrusion (MATE) family in rice (Oryza sativa). Heterologous expression in Xenopus oocyte showed that the OsFRDL4 protein was able to transport citrate and was activated by Al. The expression level of the OsFRDL4 gene in roots was very low in the absence of Al, but was greatly enhanced by Al after short exposure. Furthermore, the OsFRDL4 expression was regulated by ART1, a C2H2-type zinc finger transcription factor for Al tolerance. Transient expression of OsFRDL4 in onion epidermal cells showed that it localized to the plasma membrane. Immunostaining showed that OsFRDL4 was localized in all cells in the root tip. These expression patterns and cell specificity of localization of OsFRDL4 are different from other MATE members identified previously. Knockout of OsFRDL4 resulted in decreased Al tolerance and decreased citrate secretion compared with the wild-type rice, but did not affect citrate concentration in the xylem sap. Furthermore, there is a positive correlation between OsFRDL4 expression level and the amount of citrate secretion in rice cultivars that are differing in Al tolerance. Taken together, our results show that OsFRDL4 is an Al-induced citrate transporter localized at the plasma membrane of rice root cells and is one of the components of high Al tolerance in rice.

Introduction

Ionic aluminum (mainly Al3+) is highly toxic to plants; it rapidly inhibits root elongation at low concentrations, and results in reduced uptake of water and nutrients (Kochian et al., 2004; Ma, 2007; Poschenrieder et al., 2008). Some plant species have evolved strategies to detoxify Al both internally and externally (Ma et al., 2001). One of the most documented mechanisms of Al tolerance is the secretion by roots of organic acid anions, including citrate, malate and oxalate. These organic acid anions are able to chelate Al, thereby preventing Al binding to the root cells (Ma, 2000). This mechanism has been found in a number of monocot and dicot plant species, although the kind and the secretion patterns of organic acid anions differ among plant species (Ma, 2005).

Secretion of organic acid anions is mediated by transporters in roots; those for malate and citrate have been identified, but the one for oxalate is still unknown. The malate transporter gene ALMT1 was first isolated in wheat (Sasaki et al., 2004) and its homologous genes have been isolated in Arabidopsis and oilseed rape (Hoekenga et al., 2006; Ligaba et al., 2006). Genes responsible for the Al-induced citrate secretion have also been identified in barley, sorghum, Arabidopsis, maize, wheat and rye (Furukawa et al., 2007; Magalhaes et al., 2007; Liu et al., 2009; Ryan et al., 2009; Maron et al., 2010; Yokosho et al., 2010). Different from ALMT1, the transporters encoded belong to the multidrug and toxic compound extrusion (MATE) protein family. MATE proteins have 400–700 amino acid residues, but without consensus sequences; they are present widely in bacteria, fungi, plants and mammals, and are able to transport numerous substrates (Omote et al., 2006). A group of MATE proteins in plants has been reported to transport citrate (Durrett et al., 2007; Furukawa et al., 2007; Magalhaes et al., 2007; Yokosho et al., 2009; Maron et al., 2010). These proteins form a cluster in the phylogenetic tree of the MATE family (Yokosho et al., 2009, 2010). Interestingly, citrate-transporting MATE proteins play two different physiological roles in plants. Some members, which include FRD3 in Arabidopsis, OsFRDL1 in rice, GmMATE in soybean (Glycine max L.) and ScFRDL1 in rye (Secale cereale L.), are localized in the pericycle cells and are involved in loading of citrate into xylem vessels (Green and Rogers, 2004; Durrett et al., 2007; Rogers et al., 2009; Yokosho et al., 2009, 2010). In xylem sap citrate forms complexes with iron (Fe) and thus facilitates Fe translocation from roots to shoots. By contrast, other MATE members that include HvAACT1, SbMATE, AtMATE, TaMATE, ScFRDL2 and ZmMATE1 are involved in Al-activated citrate secretion in barley (Hordeum vulgare L.), sorghum (Sorghum bicolour L.), Arabidopsis, wheat (Triticum aestivum L.), rye and maize (Zea mays), respectively (Furukawa et al., 2007; Magalhaes et al., 2007; Liu et al., 2009; Ryan et al., 2009; Maron et al., 2010; Yokosho et al., 2010). They are localized in the root epidermis cells (Furukawa et al., 2007) and are required for external Al detoxification.

In the rice genome, there are at least 40 MATE genes (Yokosho et al., 2009). Among them, three members (OsFRDL1, OsFRDL2 and OsFRDL4) belong to the group of citrate transporters (Yokosho et al., 2010). OsFRDL1 has been characterized by function and is involved in the efficient translocation of Fe from the roots to the shoots in rice (Yokosho et al., 2009). In this study we report a non-characterized member, OsFRDL4 (Os01g0919100). The expression of this gene is very low under normal conditions, but is enhanced greatly by Al. A detailed functional study showed that this gene is not involved in Fe translocation, but responsible for external detoxification of Al in rice.

Results

OsFRDL4 encodes a citrate transporter

OsFRDL4 (Os01g0919100) consists of nine exons and eight introns, and encodes a peptide of 599 amino acids (Figure S1). The closest homolog of OsFRDL4 is SbMATE, an Al-induced citrate transporter in sorghum (Magalhaes et al., 2007), that shares 70% identity at the amino acid level. We expressed this gene in Xenopus oocyte to examine whether OsFRDL4 is also able to transport citrate. Compared with the oocytes injected with water (control), the oocytes that expressed OsFRDL4 showed a higher efflux activity for citrate (Figure 1a). This result indicated that OsFRDL4, similar to the other members in this group, is also a citrate efflux transporter.

Figure 1.

 Citrate efflux transport activity in Xenopus oocytes.
(a) Citrate efflux transport activity. OsFRDL4 cRNA, or water was injected into Xenopus oocytes. After 1-day cultivation, the oocytes were injected with 14C-labeled citrate. The release of 14C-labeled citrate from the oocytes was determined 2 h later. Data are given as means ± standard deviation (SD; n = 3). (a,b) Different letters indicate significant difference at P-values < 0.05 (Student’s t-test).
(b) Activation of OsFRDL4 by Al. Citrate was injected into oocytes with or without OsFRDL4 expression, and the inward current was recorded at −120 mV to 0 mV membrane potential in the presence and absence of 100 μm Al. Data are given as means ± SD (n = 3–7).

To examine whether OsFRDL4 is activated by Al, we performed two-electrode voltage clamp analysis. The result showed that the current of oocytes that expressed OsFRDL4 was twofold higher in the presence of Al than that in the absence of Al (Figure 1b). This result indicated that OsFRDL4 is activated by Al.

Expression patterns of OsFRDL4

Previous studies (Durrett et al., 2007; Furukawa et al., 2007; Magalhaes et al., 2007; Rogers et al., 2009; Ryan et al., 2009; Yokosho et al., 2009, 2010; Maron et al., 2010) have shown that homologs of OsFRDL4 are involved in either Fe translocation or Al tolerance depending on individual gene. Here we investigated the expression of OsFRDL4 in response to either Fe deficiency or Al exposure using semi-quantitative reverse transcriptase polymerase chain reaction (RT-PCR). The expression of OsFRDL4 under normal conditions was detected only in the roots at a very low level (Figures 2a and S2b). Fe deficiency did not affect the expression of OsFRDL4 (Figure 2a). However, exposure to Al greatly enhanced the expression of this gene in the roots (Figure 2a). Spatial expression analysis showed that OsFRDL4 was expressed in both the root tip (0–1 cm) and the mature root zones (1–2 cm) (Figure 2b), with a slightly higher expression in the former.

Figure 2.

 Tissue-specific expression pattern of OsFRDL4.
(a) OsFRDL4 expression in the roots and shoots with Fe deficiency or Al treatment. For Fe deficiency treatment, plants were cultured for 8 days in a solution with or without Fe. For Al treatment, roots were exposed to a 0.5 mm CaCl2 solution (pH 4.5) containing 50 μm AlCl3 for 6 h. RNA samples extracted from the shoots and roots were used for semi-quantitative reverse transcription polymerase chain reaction (RT-PCR). Cycle number of PCR is shown on the righthand side.
(b) Expression of OsFRDL4 mRNA at different root regions. The roots were exposed to a 0.5 mm CaCl2 solution (pH 4.5) containing 50 μm AlCl3 for 6 h. RNA was extracted from different root segments; root tip (0–1 cm) or basal root region (1–2 cm). The expression level was determined by quantitative real-time RT-PCR. Expression relative to the root tip is shown. Histone H3 was used as an internal control. Data are given as means ± standard deviation (SD) of three biological replicates.

A time-course experiment showed that the induction of OsFRDL4 occurred at 2 h after the exposure to Al, and at 6 h the expression increased by about 130 times compared with the level before Al exposure (Figure 3a). In the rice mutant of ART1, a C2H2-type Zn finger transcription factor, no Al-induced expression was observed (Figure 3a). A dose–response experiment showed that increasing the external Al concentration did not further increase the OsFRDL4 transcript level (Figure 3b). This result is in agreement with a previous report that Al-induced secretion of citrate is not increased with increasing external Al concentrations in rice (Ma et al., 2002).

Figure 3.

 Expression pattern of OsFRDL4 in response to Al.
(a) Time-dependent expression of OsFRDL4. The roots of wild-type rice and ART1 mutant were exposed to a 0.5 mm CaCl2 solution (pH 4.5) containing 50 μm AlCl3 for 0, 0.5, 1, 2 and 6 h.
(b) Dose–response expression of OsFRDL4. The roots were exposed to a 0.5 mm CaCl2 solution (pH 4.5) containing 0, 10, 30 and 50 μm AlCl3 for 2 h. The expression level was determined by quantitative real-time reverse transcription polymerase chain reaction (RT-PCR). Expression relative to the express at 0 h (a) and 0 μm Al (b) treatment in wild-type is shown. Histone H3 expression was used as an internal control. Data are given as means ± standard deviation (SD) of three biological replicates.

Cellular and subcellular localization of OsFRDL4 protein

The cellular localization of OsFRDL4 was investigated using immunostaining with an anti-OsFRDL4 antibody. In the roots without Al treatment, the signal was very weak in both the root tip and the mature zone (Figure 4a,c). A much stronger signal was observed in the Al-treated roots, which indicated the specificity of this antibody (Figure 4b,d). OsFRDL4 was localized in all cells in the root tips (Figure 4b), and in all root cells except for the epidermal cells in the basal region (Figure 4d).

Figure 4.

 Cell specificity localization of OsFRDL4 protein in roots.
Immunostaining with anti-OsFRDL4 antibody in the root tip at 3 mm (a, b); and basal root at 15 mm (c, d); of wild-type rice in the absence (a, c) or presence of 20 μm Al for 2 h (b, d). Scale bars = 100 μm.

The cellular expression of OsFRDL4 was further confirmed with transgenic plants that carried green fluorescent protein (GFP) under the control of the OsFRDL4 promoter. Immunostaining with anti-GFP antibody showed a similar localization as observed in Figure 4(b) (Figure S2).

To determine the intracellular localization of OsFRDL4, a C-terminal GFP fusion of OsFRDL4 was introduced into onion epidermal cells by particle bombardment. The fluorescence signal from GFP alone was observed both in the cytosol and the nucleus (Figure 5a,g), whereas that from the fused OsFRDL4–GFP was only observed at the outer layer of the cell (Figure 5b,c,h,i). In order to distinguish localization in the plasma membrane from that in the cell wall, plasmolysis was induced by adding 1 m mannitol. In the plasmolysed cells, the fluorescence of OsFRDL4–GFP was observed exclusively in the plasma membrane (Figure 5c,i), further indicating that OsFRDL4 is localized to the plasma membrane.

Figure 5.

 Subcellular localization of OsFRDL4.
Gene fusion between OsFRDL4 and green fluorescent protein (GFP) (b, c, e, f, h, i) or GFP alone (a, d, g) was introduced into onion epidermal cells. (a–c) GFP fluorescence; (d–f) bright field images; and (g–i) merged image. (c, f, i) Plasmolysed cell transformed with OsFRDL4–GFP protein. Arrows indicate nucleus. Scale bars = 100 μm.

OsFRDL4 is not involved in Fe translocation from roots to shoots

To investigate the physiological role of OsFRDL4, we obtained a Tos-17 insertion line (NF8537, osfrdl4) with an insertion of Tos-17 in the second intron in OsFRDL4 (Figure S1). No transcript of OsFRDL4 was detected in the homozygous osfrdl4 line (Figure S1). As there is a possibility that OFRDL4 is also involved in Fe translocation as its homolog, OsFRDL1, we grew both osfrdl4 and wild-type plants under the low Fe conditions. However, no chlorosis in the leaves was observed in either plant. There were also no significant difference between osfrdl4 and its wild-type in Fe and citrate concentrations in the xylem sap (Figure 6a,b). Using Perls’ staining, Fe precipitation was observed in the stele of osfrdl1 only as a positive control, but not in osfrdl4 or wild-type roots (Figure 6c–e). These results showed that OsFRDL4, unlike OsFRDL1, is not involved in Fe translocation.

Figure 6.

 Role of OsFRDL4 in Fe translocation from the roots to the shoots.
(a, b) Concentration of Fe (a) and citrate (b) in the xylem sap of wild-type rice (WT) and a knockout line of OsFRDL4 (osfrld4). Plants were cultured with 0.2 μm FeSO4 for 3 weeks. Xylem sap was collected from cut ends of the shoot for 1 h. Concentration of citrate was determined by HPLC and that of Fe was determined by atomic absorption spectrometry (AAS). Data are given as means ± standard deviation (SD; n = 3). (a,b) Different letters indicate significant difference at P < 0.05 (Student’s t-test). (c–e) Fe precipitation in the root of wild-type (WT) (c), a knockout line of OsFRDL4 (osfrdl4) (d), and a knockout line of OsFRDL1 (osfrld1) (e) as a positive control. Plants were cultured for 10 days in a nutrient solution containing 10 μm FeSO4. Fe precipitation was detected by Perls’ staining.

Knockout of OsFRDL4 resulted in decreased citrate secretion and Al tolerance

We then examined whether OsFRDL4 is involved in the Al-induced citrate secretion to the rhizosphere in rice. A time-course experiment showed that there was no difference in the citrate secretion between the wild-type and osfrdl4 up to 3 h after the exposure to Al with both lines showing small amounts of citrate secretion. However, at 6 h after Al exposure, the Al-induced citrate secretion in osfrdl4 was only 40% of that in the wild-type (Figure 7a). This result suggested that there was a lag between Al exposure and citrate secretion in the wild-type rice, which is consistent with the OsFRDL4 expression pattern (Figure 3a). In contrast, there was no difference in the citrate concentration of the root cell sap extracted from osfrdl4 or the wild-type (Figure 7b). This result indicated that the difference in citrate secretion between osfrdl4 and wild-type was not caused by altered citrate synthesis in the mutant.

Figure 7.

 Phenotypic analysis of OsFRDL4 knockout line.
(a) Al-induced citrate secretion from the roots in wild-type (WT) and knockout line (osfrdl4).
(b) Citrate concentration in root cell sap of the wild-type and osfrdl4. Seedlings of wild-type rice (WT) and knockout line of OsFRDL4 (osfrdl4) were exposed to a 0.5 mm CaCl2 (pH 4.5) solution containing 50 μm AlCl3. Root exudates were collected for 1, 3 and 6 h after Al treatment and then roots were excised and immediately frozen at −80°C. Root cell sap was obtained by centrifugation before completely thawing of the frozen samples. Data are means ± standard deviation (SD; n = 3).
(c) Relative root elongation in root of wild-type and knockout line of OsFRDL4 (osfrdl4). Seedlings of wild-type rice (WT) and a knockout line of OsFRDL4 (osfrdl4) were exposed to a 0.5 mm CaCl2 (pH 4.5) solution containing 0, 5 10, 30 or 50 μm AlCl3 for 24 h. The root length was measured before and after the treatment. Relative root elongation (root elongation with Al/root elongation without Al × 100) is shown. Data are given as means ± standard deviation (SD) of 10 biological replicates.
(d) Al accumulation in the root tip. Root apexes (1 cm) were excised from the roots exposed to 10 and 50 μm Al for 24 h. The concentration was determined by atomic absorption spectrometry (AAS) after digestion in 2 N HCl. Data are given as means ± SD of three biological replicates. (a–c) Different letters indicate significant difference at < 0.05 (Tukey’s test).

We further compared Al tolerance between wild-type and osfrdl4. Both lines grew similarly in the absence of Al or in the presence of a low Al concentration (10 μm) (Figure 7c). However, at higher Al concentrations (30–50 μm), the relative root elongation was significantly smaller in osfrdl4 than in wild-type rice (Figure 7c). In the presence of 50 μm Al, the relative elongation of osfrdl4 was 34% smaller than the wild-type line. The Al concentration in the root apex (0–1 cm) was 1.6 times higher in osfrdl4 than in wild-type rice (Figure 7d); this difference was not observed at 10 μm Al. These results support a role of citrate secretion in Al detoxification in rice, even though the amount of citrate secretion was small.

Complementation test

To determine if OsFRDL4 is responsible for the phenotypes observed above (Figure 7), we introduced OsFRDL4 into osfrdl4 mutant under the control of the OsFRDL4 promoter. Furthermore, we inserted a 35S enhancer prior to the OsFRDL4 promoter to enhance the expression of OsFRDL4. The expression level of OsFRDL4 was only increased slightly in the two independent transgenic lines than that of WT (wild-type) (Figure 8a). The Al-induced citrate secretion in transgenic lines was recovered to the similar level of the WT rice (Figure 8b). These results indicated that OsFRDL4 was responsible for the Al-induced secretion of citrate.

Figure 8.

 Complementation test and phenotypic analysis.
(a) The expression level of OsFRDL4 in the wild-type line (WT), knockout line (osfrdl4) and two independent transgenic lines carrying OsFRDL4 promoter–ORF construct (9 and 19). The seedlings were exposed to 50 μm Al for 24 h. The expression level was determined by quantitative real-time reverse transcription polymerase chain reaction (RT-PCR). Data are given as means ± standard deviation (SD) of three replicates.
(b) Al-induced citrate secretion. Root exudates were collected by exposing different lines to a 0.5 mm CaCl2 (pH 4.5) solution containing 50 μm AlCl3 for 24 h. Citrate was determined by enzymatic method. Data are means ± SD of three replicates. (a, b) Different letters indicate significant difference at P < 0.05 (Tukey's test).

Genotypic difference in OsFRDL4 expression

To investigate whether OsFRDL4 potentially underlies genotypic difference in Al tolerance, the expression levels and citrate secretion were compared in seven rice cultivars that differed in Al tolerance (Figure S3). A good correlation was found between the OsFRDL4 expression level and Al tolerance (relative root elongation) (r = 0.77) (Figure 9a). There was also a good correlation between Al-induced citrate secretion and Al tolerance (r = 0.81) (Figure 9b), between the OsFRDL4 expression level and citrate secretion (r = 0.83) (Figure 9c).

Figure 9.

 Correlation between OsFRDL4 expression, citrate secretion and Al tolerance in seven rice cultivars.
(a) Correlation between Al tolerance and OsFRDL4 expression level.
(b) Correlation between Al tolerance and Al-induced citrate secretion.
(c) Correlation between Al-induced citrate secretion and OsFRDL4 expression level. Al tolerance was evaluated by exposing seedlings to a 0.5 mm CaCl2 (pH 4.5) solution containing 0 or 30 μm AlCl3 for 24 h. The root length was measured before and after the treatment and the relative root elongation (root elongation with Al/root elongation without Al × 100) is shown. Data are given as means ± standard error (SE) of 10 biological replicates. The expression level of OsFRDL4 was determined by quantitative real-time reverse transcription polymerase chain reaction (RT-PCR). Expression relative to Nipponbare root with +Al is shown. Histone H3 was used as an internal control. Data are given as means ± SE of three biological replicates. Al-induced citrate secretion was determined by enzymatic method after the roots were exposed to 30 μm Al for 24 h. Data are given as means ± SE of three biological replicates.

Discussion

OsFRDL4 is an Al-inducible citrate transporter

Our results showed that the rice MATE gene OsFRDL4 encodes a plasma membrane-localized efflux transporter for citrate (Figures 1 and 5). Its subcellular localization and transport substrate are similar to other characterized members in the same group that include HvAACT1, SbMATE, AtMATE1 and ZmMATE1 (Furukawa et al., 2007; Magalhaes et al., 2007; Liu et al., 2009; Maron et al., 2010). Nevertheless, OsFRDL4 shows distinct features in its expression pattern and cellular localization. Its expression is greatly and rapidly induced by Al (Figure 3), which is different from other homolog members. For example, the expression of HvAACT1 in barley was not induced by Al (Furukawa et al., 2007). Although the expression of SbMATE was also induced by Al, the lag time leading to the induction (several days) (Magalhaes et al., 2007) was much longer compared with OsFRDL4 (2 h) (Figure 3). The lag time for OsFRDL4 is similar to those for AtMATE, ZmMATE1 and ScFRDL2 (Liu et al., 2009; Maron et al., 2010; Yokosho et al., 2010), but the induction enhancement appeared to be greater for OsFRDL4 (approximately 80-fold) than for others (only a few folds) (Magalhaes et al., 2007; Liu et al., 2009; Maron et al., 2010; Yokosho et al., 2010). The mechanism that underlies these differences is unknown. One possibility is that OsFRDL4 might be required only for the external Al detoxification and plays no role in the absence of Al. In contrast, other homologous genes are known to be also required for yet uncharacterized functions in the absence of Al. This situation is supported by a much lower expression of OsFRDL4 in the absence of Al (Figure 2). This rapid and great induction of OsFRDL4 in response to Al also provides a good model to study the molecular mechanisms of Al response in the future.

The cell specificity of localization also differs between OsFRDL4 and HvAACT1. In the root tip, OsFRDL4 is localized in all cells (Figure 4), compared with the localization of HvAACT1 to the epidermal cells (Furukawa et al., 2007). As the cell specificity of localization of SbMATE, AtMATE and ZmMATE has not been examined, it is unknown whether these proteins have similar or different cellular localization. The difference in the cell specificity of localization between OsFRDL4 and HvAACT1 suggested that OsFRDL4 may also play a role in the detoxification of Al in the apoplast in addition to the rhizosphere, although the mechanism that underlies different localization remains to be examined in the future.

OsFRDL4 is regulated by ART1

Recently, ART1, a C2H2-type zinc finger transcription factor for Al tolerance, has been identified in rice (Yamaji et al., 2009). ART1 is expressed constitutively in the root and the ART1 protein is localized in the nuclei of all root cells. Microarray analysis has revealed that ART1 regulates 31 genes that are implicated in Al tolerance at different cellular levels (Yamaji et al., 2009). However, a probe for OsFRDL4 was not included in the 44K oligo microarray slides, probably due to its extremely low expression level under the normal growth conditions. Therefore it was unknown whether OsFRDL4 was regulated by ART1. The fact that Al-induced expression of OsFRDL4 did not occur in the art1 mutant (Figure 3) indicated that OsFRDL4 is also regulated by ART1. This conclusion was also supported by similar cell specificity of localization between OsFRDL4 and ART1 (Figure 4; Yamaji et al., 2009). Furthermore, recently, the cis-acting element of ART1 was identified to be GGN(T/g/a/C)V(C/A/g)S(C/G) (Tsutsui et al., 2011), which is present in the OsFRDL4 promoter region (Figure S4).

OsFRDL4 may be partially responsible for genotypic difference in Al tolerance

Rice secretes citrate from the roots in response to Al (Ma et al., 2002). However, despite a high Al tolerance of rice, the amount of citrate secreted is small and there is no difference in the Al-induced citrate secretion between Al-sensitive and -tolerant rice cultivars (Ma et al., 2002). Furthermore, the secretion does not show a dose–response correlation (Ma et al., 2002). Based on these findings, it was concluded that, unlike other gramineous crops such as wheat, barley and rye, secretion of organic acid anions is not an important mechanism of Al tolerance in rice. However, in the present study, we found that knockout of OsFRDL4 resulted in decreased citrate secretion and increased Al sensitivity, as well as increased Al concentration in the root tip (Figure 7). These findings indicate that OsFRDL4 does play a role in the external detoxification of Al through secretion of citrate from the roots. Furthermore, it is likely that OsFRDL4 will be responsible for the genotypic difference in Al tolerance, by further examination of more rice cultivars that differ in Al tolerance. There were good correlations between OsFRDL4 expression level and citrate secretion, and Al tolerance among seven rice cultivars (Figure 9). Previous studies by using different populations detected a quantitative trait locus (QTL) for Al tolerance on chromosome 1 (Wu et al., 2000; Nguyen et al., 2001; Ma et al., 2002). OsFRDL4 (Os01g0919100) is located at 41.8 Mb in chromosome 1, which is flanked by this QTL (Yamaji et al., 2009). Therefore, OsFRDL4 might be the gene responsible for this QTL although further studies are required to confirm this hypothesis. It will be interesting in the future to examine the mechanisms that underlie differential expression of OsFRDL4.

Rice is the most Al tolerant plant species among the small-grain cereal crops (Foy, 1988). This high Al tolerance seems to be achieved by multiple genes that are involved in detoxification of Al, which are regulated by ART1 (Yamaji et al., 2009). Among the downstream genes regulated by ART1, three genes (STAR1, STAR2 and Nrat1) have been characterized by function. STAR1 and STAR2 encode an ATP-binding and a transmembrane domain, respectively, of a bacterial-type ABC transporter complex (Huang et al., 2009). The complex of STAR1–STAR2 transports UDP–glucose and is proposed to be involved in the modification of cell wall (Huang et al., 2009). In contrast, Nrat1 belongs to Nramp family and encodes a transporter for Al (Xia et al., 2010). Knockout of STAR1 and Nrat1 resulted in greater reduction in Al tolerance than that of OsFRDL4 (Huang et al., 2009; Xia et al., 2010; Figure 7), which indicated that the contribution of OsFRDL4 to rice Al tolerance is relatively small.

In conclusion, OsFRDL4 is an Al-inducible citrate transporter that is involved in the Al-induced citrate secretion in rice with distinct expression patterns and cell specificity of localization. OsFRDL4 is one of the components of high Al tolerance in rice.

Experimental Procedures

Citrate transport assay in Xenopus oocytes

The open reading frame (ORF) of the OsFRDL4 cDNA fragment was amplified using the following primers: 5′-ATGGCGAGGAGTTCATCAGCT-3′ (forward) and 5′-TCATTTGCGAAGAAACTTCCACG-3′ (reverse). The fragment that contained the ORF was inserted into a Xenopus laevis oocyte expression vector, pXbG–ev1 (Preston et al., 1992). cRNA preparation, micro-injection into oocytes and citrate transport activity assay were performed as described previously (Yokosho et al., 2009).

For electrophysiological studies, the OsFRDL4 cRNA- or water-injected oocytes were incubated in the Modified Barth’s Saline (MBS) solution at 18°C. After a 2 days incubation, 50 nl of 25 mm sodium citrate was injected into the oocytes and then incubated for 0.5–2 h in ND96 buffer (also referred to as ‘High NaCl’) containing 0 or 100 μm Al at pH 4.5 (Furuichi et al., 2011). The net current across the oocyte membrane was measured using the two-electrode voltage clamp system with the amplifier (MEZ-7200 and CEZ-1200, Nihon Kohden, http://www.nihonkohden.co.jp/) at different membrane voltages. The electrical potential difference across the membrane was clamped from 0 mV to −120 mV.

Plant materials and growth conditions

Rice (Oryza sativa) (cvs. Nipponbare, Fukoku, Nagoyashiro, Mu Bang, Nepal 8, POKKARI, Kasalath and art1 mutant) was used. A Tos-17 insertion line (NF8537) for OsFRDL4 was provided by the Rice Genome Resource Center, which was developed by the Rice Genome Project of the National Institute of Agrobiological Sciences, Japan. The homozygous lines were screened by PCR using OsFRDL4 specific primers (5′-CAATGATGATCCCTGCACTG-3′ and 5′-GGATGCGGTAAGCAGAAAAA-3′). Seeds were soaked in water overnight at 25°C in the dark and then transferred to a net floating on a 0.5 mm CaCl2 solution. On day 7, seedlings were transferred to a 3.5-L plastic pot containing one-half-strength Kimura B solution and grown in a greenhouse at 22–25°C. After 10 days, the seedlings (five plants per pot) were transferred to a 1.2-L pot that contained freshly prepared nutrient solution. The nutrient solution contained the following macronutrients (mm): (NH4)2SO4 (0.18), MgSO4·7H2O (0.27), KNO3 (0.09), Ca(NO3)2·4H2O (0.18) and KH2PO4 (0.09), and the micronutrients (μm): Fe–EDTA (20), MnCl2·4H2O (0.5), H3BO3 (3), (NH4)6Mo7O24·4H2O (1), ZnSO4·7H2O (0.4) and CuSO4·5H2O (0.2). The pH of this solution was adjusted to 5.6 and the nutrient solution was renewed every 2 days. After 10–20 days further growth, the seedlings were used for root exudate collection. For xylem sap analysis, plants were cultured in one-half-strength Kimura B solution with 0.2 μm FeSO4. Each experiment was repeated at least three times with three replicates each.

RNA extraction and quantitative real-time RT-PCR

To examine the effect of Fe deficiency on the expression level, seedlings prepared as described above were cultured for 8 days in a nutrient solution with or without Fe and then the roots were harvested for RNA extraction. Root samples were also taken from the WT and art1 mutant after the seedlings were exposed to a 0.5 mm CaCl2 (pH4.5) solution that contained 50 μm Al for 0, 0.5, 1, 2 and 6 h. For spatial experiment, different root regions (0–1 cm and 1–2 cm) were sampled, respectively. Dose–responsive samples were taken from the roots exposed to 0, 10, 30 and 50 μm Al for 2 h. All samples were frozen immediately in liquid nitrogen and stored −80°C until use.

Total RNA was extracted using the RNeasy mini kit (Qiagen, http://www.qiagen.com/). One microgram of total RNA was used for first-strand cDNA synthesis using a SuperScript 2 kit (Invitrogen, http://www.invitrogen.com) and an oligo(dT)12–16 primer (Invitrogen, following the manufacturers instruction. The transcript levels of OsFRDL4, and Histone H3 (internal control) were determined by using SYBR premix ExTaq (Takara Bio Inc., http://www.takara-bio.com/) with the following primers: OsFRDL4, 5′-CGTCATCAGCACCATCCACAG-3′ (forward) and 5′-TCATTTGCGAAGAAACTTCCACG-3′ (reverse); Histone H3, 5′-AGTTTGGTCGCTCTCGATTTCG-3′ (forward) and 5′-TCAACAAGTTGACCACGTCACG-3′ (reverse). Expression data were normalized with the expression level of HistoneH3 by the ΔΔCt method.

Immunostaining of OsFRDL4 protein

The synthetic peptide C-CENADAAGGGGGDNGDH (positions 60–75 of OsFRDL4) was used to immunize rabbits to obtain antibodies against OsFRDL4. The roots exposed to a 0.5 mm CaCl2 solution with or without 50 μm Al (pH 4.5) for 6 h were used for immunostaining. Immunostaining was performed as described previously (Yamaji and Ma, 2007). Fluorescence from the secondary antibody (Alexa Fluor 555 goat anti-rabbit IgG; Molecular Probes, available from Invitrogen) was observed with a confocal laser scanning microscopy (LSM700; Carl Zeiss, http://www.zeiss.com/).

Immunostaining with an anti-GFP antibody was also performed in transgenic rice carrying OsFRDL4 promoter–GFP as described below.

Subcellular localization of OsFRDL4

The ORF of the OsFRDL4 cDNA fragment was amplified using primers: 5′-AGTCGACATGGCGAGGAGTTCATCAGCT-3′ (forward) and 5′-TGCATGCCTTTGCGAAGAAACTTCCACGGC-3′ (reverse). The OsFRDL4 fragment was ligated to the 3′ end of GFP and placed under the control of the CaMV 35S promoter in pBluescript SK – (Stratagene, http://www.stratagene.com). The resulting plasmid was designated OsFRDL4–GFP. Gold particles with a diameter of 1 μm, and coated with OsFRDL4–GFP or GFP alone, were introduced into onion epidermal cells using particle bombardment (PDS-1000/He particle delivery system, Bio-Rad, http://www.bio-rad.com/) using 1100 psi rupture disks. To induce plasmolysis, cells were incubated with 1 m mannitol for 10 min. GFP fluorescence was observed with a confocal laser scanning microscopy.

Collection and analysis of xylem sap

The nutrient solution was renewed before collection of xylem sap. After 6 h, the seedlings were decapitated at 2 cm above the roots and xylem sap was collected from the cut end for 1 h with a micropipette. The concentration of Fe in the xylem sap was measured using flameless atomic absorption spectrometry (AAS) (Z-2000; Hitachi, http://www.hitachi-hitec.com/). Citrate concentration in xylem sap was measured by high-pressure liquid chromatography (HPLC) using a reverse-phase column (Cosmossil Packed Column 5C18-PAQ 4.6 I. D.x 250 mm, Nakalai tesque, http://www.nacalai.co.jp/). The mobile phase was a dilute perchloric acid solution (pH 2.1) run at 40°C, and peaks were detected by post column method with bromothymol blue at the wavelength of 425 nm (Ma et al., 1997).

Perls’ staining

Perls’ staining was performed with plants grown in a solution containing 10 μm FeSO4 for 10 days according to Yokosho et al. (2009). Briefly, equal amounts of solutions of 4% (v/v) HCl and 4% (w/v) potassium ferrocyanide were mixed immediately prior to use. Rice roots were exposed to the staining solution and vacuum infiltrated for 15 min. The seedlings were then rinsed with water and approximately 200-μm cross-sections were prepared by free hand. The staining was observed under an optical microscope.

Al-induced citrate secretion and citrate concentration in cell sap

Before collection of root exudates, seedlings of both WT and osfrdl4 were exposed to a 0.5 mm CaCl2 (pH 4.5) solution overnight. Root exudates were collected by exposing the roots to a 0.5 mm CaCl2 (pH 4.5) solution containing 50 μm AlCl3. At 1, 3, and 6 h after exposure, the root exudates were collected and passed through a cation-exchange resin column (16 × 14 mm) filled with 5 g of Amberlite IR-120B resin (H+ form; Rohm and Haas, http://www.rohmhaas.com/), followed by an anion-exchange resin column (16 × 14 mm) filled with 2 g of AG 1 × 8 resin (100–200 mesh; formate form; Bio-Rad, http://www.bio-rad.com/). Organic acids retained on an anion-exchange resin were eluted with 2 N HCl and the eluate was concentrated to dryness with a rotary evaporator (40°C). After evaporation, the residue was dissolved in 1 ml of milli-Q water, and the concentration of organic acids was analyzed by enzymatic method according to Delhaize et al. (1993).

For root cell sap collection, the roots harvested after Al exposure were placed into a syringe and frozen at −80°C. After thawing at room temperature for a short time, a syringe with sample was put into a test tube and centrifuged at 5000 g for 20 min to obtain the cell sap.

Measurement of root elongation and Al content in root tip

Seedlings (3-day-old) of both WT and osfrdl4 lines were exposed to 0.5 mm CaCl2 solutions that contained 0, 10, 30 and 50 μm Al for 24 h. Root length was measured at 0 and 24 h with a ruler. After treatment, the roots were placed in a 0.5 mm CaCl2 solution for 15 min, and then roots tips (0–1 cm) were excised with a razor blade. Excised roots were placed in plastic tube with 1 ml 2 N HCl. The tubes were shaken for 24 h. The Al concentration in solution was measured using flameless atomic absorption spectrometry.

Generation of transgenic plants

For tissue localization analysis and complementation of osfrdl4, the promoter region of OsFRDL4 (2 kb) was amplified from rice genomic DNA using primers 5′-GGTACCGCCCGAGTCACTCGCATATGAGTA-3′ and 5′-CTCGAGGCAGGTGTTGCTTAGCTCGTCGAT-3′. The gene of GFP and Nopaline Synthase (NOS) terminator were isolated from the plasmid used for GFP transit assay as described above and fused with the promoter to construct the plasmid of pOsFRDL4–GFP–NOS. For complementation test, the promoter region was fused with the ORF of OsFRDL4 (pOsFRDL4–OsFRDL4–NOS). These plasmids were cloned to a binary vector (pPZPHa–lac). To make an enhanced construct, a 35S enhancer was cloned from pGA2772 (Jeong et al., 2002) using primer 5′-CTCGACTCTAGAGGATCCCCAA-3′ and 5′-AAGCTTGAATTCACTAGTGATTGCTCTAGA-3′. The 35S enhancer fragment was inserted in front of pOsFRDL4–OsFRDL4–NOS region. pOsFRDL4–GFP–NOS was transformed into calluses derived from the Nipponbare and enhancer –pOsFRDL4–OsFRDL4–NOS was transformed into calluses derived from the osfrdl4 by Agrobacterium-mediated method (Hiei et al., 1994).

Root exudates were collected by exposing the transgenic plants to a 0.5 mm CaCl2 (pH 4.5) solution that contained 50 μm AlCl3 for 24 h as described above. Samples were also taken for RNA extraction as described above.

Genotypic variation in OsFRDL4 expression and citrate secretion

Seven rice cultivars differing in Al tolerance as described above were used for OsFRDL4 expression analysis and citrate secretion. For root elongation comparison, the seedlings (5-day-old) were exposed to a 0.5 mm CaCl2 containing 0 and 30 μm Al for 24 h. The root length was measured with a rule before and after the treatment. After the treatment, the roots from each cultivar were sampled for RNA extraction and used for quantitative analysis of the gene expression as described above. To compare the Al-induced citrate secretion, seedlings (14-day-old) were exposed to a 0.5 mm CaCl2 (pH 4.5) solution containing 30 μm AlCl3 for 24 h. Other procedures for root exudation collection and citrate determination were the same as described above.

Accession numbers

Sequence data from this article can be found in the Arabidopsis Genome Initiative or GenBank/EMBL databases under the accession number, AB608020.

Acknowledgements

This research was supported by a Grant-in-Aid for Scientific Research on Innovative Areas from the Ministry of Education, Culture, Sports, Science and Technology of Japan (22119002 to J.F.M.), a grant from the Ministry of Agriculture, Forestry and Fisheries of Japan, Genomics for Agricultural Innovation IPG-0006 (to J.F.M.) and Ohara Foundation for Agricultural Research. We thank Fangjie Zhao for his critical reading of this manuscript.

Ancillary