Isolation and characterization of IRO2, a novel iron ... - Semantic Scholar
Journal of Experimental Botany, Vol. 57, No. 11, pp. 2867–2878, 2006 doi:10.1093/jxb/erl054 Advance Access publication 3 August, 2006 This paper is available online free of all access charges (see http://jxb.oxfordjournals.org/open_access.html for further details)
RESEARCH PAPER
Isolation and characterization of IRO2, a novel iron-regulated bHLH transcription factor in graminaceous plants Yuko Ogo1,*, Reiko Nakanishi Itai1,2,*, Hiromi Nakanishi1, Haruhiko Inoue1,2, Takanori Kobayashi1,2, Motofumi Suzuki1, Michiko Takahashi1, Satoshi Mori1 and Naoko K. Nishizawa1,2,† 1
Received 10 April 2006; Accepted 10 May 2006
Abstract To clarify the molecular mechanism that regulates iron (Fe) acquisition in graminaceous plants, a time-course analysis of gene expression during Fe deficiency stress was conducted using a rice 22K oligo-DNA microarray. Twenty-one genes for proteins that function in gene regulation were induced by Fe deficiency. Of these genes, a putative basic helix–loop–helix (bHLH) transcription factor gene, named OsIRO2, was strongly expressed in both roots and shoots during Fe deficiency stress. The expression of OsIRO2 was induced exclusively by Fe deficiency, and not by deficiencies in other metals. Expression of the barley HvIRO2 gene, which is a homologue of OsIRO2, was also induced by Fe deficiency. An in silico search revealed that IRO2 is highly conserved among graminaceous plants, which include wheat, sorghum, and maize. The cyclic amplification and selection of targets (CASTing) technique revealed that OsIRO2 bound preferentially to the sequence 59-ACCACGTGGTTTT39, and the electrophoretic mobility shift assay revealed 59-CACGTGG-39 as the core sequence for OsIRO2 binding. Sequences similar to the OsIRO2binding sequence were found upstream of several genes that are involved in Fe acquisition, such as OsNAS1, OsNAS3, OsIRT1, OsFDH, OsAPT1, and IDS3. The core sequence of the OsIRO2-binding sequence
occurred more frequently in the upstream regions of Fe deficiency-inducible genes than in the corresponding regions of non-inducible genes. These results suggest that IRO2 is involved in the regulation of gene expression under Fe-deficient conditions. Key words: Basic helix–loop–helix, iron deficiency, microarray, mugineic acids, phytosiderophore, rice, transcription factor.
Introduction Iron (Fe) is an essential element for plant growth. Although abundant in soil, Fe is mainly present as oxidized Fe(III) compounds, which are poorly soluble in neutral-to-alkaline soils. Consequently, plants grown in calcareous soils often exhibit severe chlorosis due to Fe deficiency; this is a major agricultural problem that results in reduced crop yields (Marschner, 1995). Gaining an understanding of the molecular and cellular events related to Fe acquisition and its regulation is a fundamental issue in plant biology, and is a prerequisite for the genetic improvement of plants cultivated in extreme environments. Graminaceous plants have evolved a strategy for Fe acquisition that utilizes Fe(III) chelators, i.e. mugineic acid family phytosiderophores (MAs). Graminaceous plants secrete MAs from their roots to solubilize rhizospheric Fe(III) (Takagi, 1976). Subsequently, Fe(III)–MA complexes are
* These authors contributed equally to this work. y To whom correspondence should be addresssed. E-mail: [email protected] Abbreviations: ABRE, ABA-responsive element; bHLH, basic helix–loop–helix; CASTing, cyclic amplification and selection of targets; EMSA, electrophoretic mobility shift assay; EST, expressed sequence tag; MAs, mugineic acid family phytosiderophores; MBP, maltose-binding protein; ORF, expressed sequence tag. ª 2006 The Author(s). This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/2.0/uk/) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.
Downloaded from http://jxb.oxfordjournals.org/ at Pennsylvania State University on February 21, 2013
Graduate School of Agricultural and Life Sciences, The University of Tokyo, 1-1 Yayoi, Bunkyo-ku, Tokyo 113-8657, Japan 2 Core Research for Evolutional Science and Technology (CREST), Japan Science and Technology Corporation (JST), Tokyo, Japan
2868 Ogo et al.
mobility shift assay (EMSA). Several genes that are involved in Fe acquisition in plants were found to possess sequences homologous to the OsIRO2-binding sequence in their upstream regions. Materials and methods Plant materials Both rice (Oryza sativa L. cv. Nipponbare) and barley (Hordeum vulgare L. cv. Ehimehadaka no. 1) were grown hydroponically, as described previously (Kobayashi et al., 2005). Rice was grown under 14/10 h light/dark cycles and 30/25 8C, and barley was grown under 14/10 h light/dark cycles and 22/17 8C. For the rice microarray analysis, Fe deficiency was initiated 3 weeks after germination by omitting Fe(III)-EDTA from the culture medium. Plants were harvested on days 0, 1, 2, 3, 5, and 7 of Fe deficiency. On day 7 of Fe deficiency, Fe was re-supplied by adding Fe(III)-EDTA at the usual concentration, and the plants were harvested 4 d later (day 7+4). For the northern blot analysis of rice plants, Fe deficiency treatment was initiated 5 weeks after germination, and plants were harvested on days 0, 1, 3, 5, 7, 9, and 11. As a control, plants grown under Fesufficient conditions were harvested on day 7. For the micronutrient deficiency treatments, 2-week-old plants were transferred to nutrient media that lacked zinc, Fe, manganese, or copper, and grown for an additional 2 weeks. For the northern blot analysis of barley, 3week-old plants were transferred to a nutrient medium that lacked Fe. Plants were harvested on days 1, 3, 5, 7, and 9. Plants that were re-supplied with Fe on day 9 were grown for an additional 2 d (day 9+2). As a control, plants grown under Fe-sufficient conditions were harvested on day 1. Oligo-DNA microarray analysis The rice 22K oligo-DNA microarray (Agilent Technology) contained 21 938 unique 60mer oligonucleotides that were synthesized based on sequence data from the Rice Full-length cDNA Project (http:// cdna01.dna.affrc.go.jp/cDNA/). Total RNA samples were prepared from the roots and shoots of three plants at each time point using the guanidium isothiocyanate/CsCl density gradient centrifugation method. The integrity of each RNA sample was checked using the Agilent 2100 bioanalyser (Agilent Technology). Total RNA samples (200 ng) from plants harvested on day 0, and at each subsequent time point, were labelled with Cy3 or Cy5 using the Agilent Low RNA Input Fluorescent Linear Amplification Kit (Agilent Technology). Hybridization was performed according to the manufacturer’s instructions. Images of the hybridized microarrays were obtained with the Agilent microarray scanner (Agilent Technology) and analysed with feature-extraction software (Agilent Technology). The reproducibility of the microarray analysis was assessed by a dye swap in each experiment. The signal intensity was normalized by the LOWESS method. The ratios were calculated as a value (the plants harvested at each time point of Fe deficiency)/(the plants on day 0). The point showing the signal value of Fe deficiency-treated plants >500, the P-value 2 in both Cy3 and Cy5 channels was considered as significantly up-regulated. Northern blot analysis The 1051 bp expressed sequence tag (EST) cDNA clone of OsIRO2 in pBluescript was provided by the National Institute of Agrobiological Sciences (Japan) and sequenced for both DNA strands. For the OsIRO2 probe, the fragment with an XhoI site in the multicloning site at the 59 end of the open reading frame (ORF) and a HindIII site located just before the termination codon was used for northern blot analysis of rice plants. For the HvIRO2 probe, the XhoI fragment that contained the complete insert of the longest clone in pCMVFL3 was
Downloaded from http://jxb.oxfordjournals.org/ at Pennsylvania State University on February 21, 2013
taken up through the Fe(III)–MAs transporter (Curie et al., 2001). MAs are biosynthesized from methionine through nicotianamine (NA) (Shojima et al., 1989). Two dioxygenases convert 29-deoxymugineic acid (DMA), the first member of the MAs, into various kinds of MAs (Nakanishi et al., 2000; Kobayashi et al., 2001). The genes that encode key enzymes for MA synthesis have been isolated from graminaceous plants, such as barley and rice. Several NA synthase (NAS) genes (Higuchi et al., 1999, 2001; Douchkov et al., 2002; Inoue et al., 2003), two NA aminotransferase (NAAT) genes (Takahashi et al., 1999), and two dioxygenase genes, IDS2 (Okumura et al., 1994) and IDS3 (Nakanishi et al., 1993), are all up-regulated by Fe deficiency. Many of the genes involved in supplying methionine for the production of MAs under Fe-deficient conditions have been identified in barley and rice (Negishi et al., 2002; Kobayashi et al., 2005). All of these genes are induced in Fe-deficient plants. As an important clue to the gene regulatory mechanisms that operate under Fe-deficient conditions, the novel cisacting iron-deficiency-responsive elements IDE1 and IDE2 were identified in the barley IDS2 promoter (Kobayashi et al., 2003). IDE1 and IDE2 confer Fe deficiency inducibility and root-specific expression in heterogeneous tobacco plants. The results of a study conducted by Kobayashi et al. (2003) indicate that graminaceous plants and other higher plants share conserved gene regulatory mechanisms that are active under Fe-deficient conditions. Transcription factors that bind to IDE1 and IDE2 have not yet been identified. A basic helix–loop–helix (bHLH) protein, FER, was identified using the fer mutant of tomato, which is unable to induce certain genes that are involved in Fe acquisition under Fe-deficient conditions (Ling et al., 2002). A homologue of FER in Arabidopsis, FIT1/FRU/ AtbHLH29, plays a role similar to that of FER (Colangelo and Guerinot, 2004; Jakoby et al., 2004; Yuan et al., 2005). Although it is assumed that these proteins bind to the E-box motif (CANNTG) (Toledo-Ortiz et al., 2003), their DNAbinding sites have not yet been determined. Many stressresponsive transcription factors that regulate the expression of genes that encode key proteins for stress tolerance have been reported (Gilmour et al., 1998; Liu et al., 1998; Chinnusamy et al., 2003). Since the expression of these key transcription factors can be transient, it is important to detect transient changes in gene expression to identify potential regulators of stress-inducible genes. In this study, a time-course analysis of gene expression during Fedeficiency stress was performed using a rice 22K oligoDNA microarray. Genes, the products of which were suspected of involvement in gene regulation, were searched and analysed. Of theses genes, OsIRO2, which encodes a putative bHLH transcription factor, was strongly induced in shoots and roots by Fe deficiency. The OsIRO2-binding sequence was revealed by the cyclic amplification and selection of targets (CASTing) method and electrophoretic
IRO2, a novel iron-regulated transcription factor 32
used. These fragments were labelled with [a- P]dATP using the random labelling method, and the labelled DNA was purified using a ProbeQuant G-50 micro-column (Pharmacia). Total RNA samples from roots and shoots were extracted by guanidium isothiocyanate extraction for rice and by SDS–phenol extraction for barley. Total RNA (20 lg per lane) was separated on 1.4% (w/v) agarose gels that contained 5% (v/v) formaldehyde and blotted onto nylon membranes (Hybond-N+; Amersham). The blots were hybridized at 65 8C overnight with labelled probes in 0.5 M Church phosphate buffer (Church and Gilbert, 1984), 1 mM EDTA, and 7% (w/v) SDS with 100 lg ml1 salmon sperm DNA. After hybridization, the blots were washed twice with 40 mM Church phosphate buffer and 1% (w/v) SDS at 42 8C for 10 min and then washed at high stringency with 23 SSPE plus 0.1% (w/v) SDS at 45 8C for 10 min. Radioactivity was detected using the FLA-3000 imaging analyser (Fuji Film).
Preparation of thioredoxin–OsIRO2 fusion protein for CASTing The ORF of OsIRO2 was amplified using the primers 59-CACCATGGAGCAGCTGTTCGTCGACGACC-39 and 59-GAAAGTCGTCAGGAACAAAGCAAAGCTT-39. The amplified fragment was cloned into pENTR/D-TOPO (Invitrogen) and sequenced. The pET32a(+) vector (Novagen) was converted into the Gateway destination vector pDEST32a by introducing an RfA cassette at the EcoRV site, according to the Gateway Conversion System (Invitrogen). The pDEST32a-OsIRO2 construct was generated by an LR recombination reaction between the destination vector pDEST32a and the entry vector pENTR-OsIRO2, to produce the thioredoxin– OsIRO2 fusion protein. The expression vector pDEST32a-OsIRO2 was introduced into Escherichia coli strain BL21(DE3), and induction of the recombinant protein was performed according to the pET System Manual (Novagen). After 3 h of induction with isopropyl-b-D-thiogalactopyranoside, a crude cellular extract was prepared by sonication using the binding buffer for the CASTing procedure as the lysis buffer. The thioredoxin–OsIRO2 fusion protein in the crude extract was purified using anti-thioredoxin agarose (Sigma) and immunologically fixed onto agarose resin for use in the binding reaction of CASTing. CASTing Oligonucleotides of 75 nt were chemically synthesized that contained 25 nt of a random sequence and an SalI site flanked by the T3 and T7 primer sequences, i.e. 59-AATTAACCCTCACTAAAGGGACGTCGAC(N25)GCCCTATAGTGAGTCGTATTAC-39, where the underlined sequences represent the T3 and T7 sequences, respectively. Preparation of the double-stranded oligomer for CASTing was performed by annealing a synthesized oligonucleotide with the T7 primer and filling in with the Klenow fragment of DNA polymerase (TOYOBO). The binding buffer consisted of 20 mM TRIS– HCl (pH 8.0), 50 mM KCl, 0.5 mM EDTA, 10% (v/v) glycerol, 20 lg ml1 bovine serum albumin, 1 mM 1,4-dithiothreitol, and 50 lg ml1 salmon sperm DNA. Binding reactions were performed by mixing thioredoxin–OsIRO2 on agarose resin with 10 lg of the
double-stranded oligomer in binding buffer and incubating for 1 h on ice. The DNA–protein complexes were precipitated by centrifugation, and the pellets were washed twice with binding buffer. The pellets were suspended in 50 ll of 13 KOD+ polymerase chain reaction (PCR) buffer (TOYOBO) and boiled for 3 min, to unbind the DNA from the proteins and permit elution. For the PCR, five tubes with the same reaction solution that contained 10 ll of the eluted DNA, T3 and T7 primers, and KOD+ DNA polymerase were prepared in a volume of 50 ll. PCR amplification per cycle was performed for 15 s at 94 8C, 15 s at 55 8C, and 30 s at 68 8C. The PCR was stopped by removing a tube from the machine at the 10th, 15th, 20th, 25th, and 30th cycle and placing on ice. To determine at which cycle the DNA began to be amplified, a 10 ll aliquot of the reaction in each tube was separated on a 2% agarose gel, and the gel was stained with ethidium bromide. A 5 ll aliquot of the reacted solution at the earliest cycle at which DNA was detected was used in the subsequent binding reaction. This series of procedures was repeated seven times. In the last PCR, the DNA was amplified using the B1T3 primer, which contained the aatB1 sequence for the Gateway system (Invitrogen) flanked by the T3 sequence (59-GGGGACAAGTTTGTACAAAAAAGCAGGCTAATTAACCCTCACTAAAGGGAC-39), and the B2T7 primer, which contained the aatB2 sequence flanked by the T7 sequence (59-GGGGACCACTTTGTAGAAAGCTGGGTGTAATACGACTCACTATAGGGC-39). Using a BP recombination reaction, the amplified DNA was introduced into the destination vector pDONR221 (Invitrogen), and the generated clones were sequenced. Preparation of maltose-binding protein–OsIRO2 fusion protein (MBP–OsIRO2) for EMSA The ORF of OsIRO2 was amplified using the primers 59- CTCGAGGAATTCATGGAGCAGCTGTTCGTCGA-39 and 59- TCTAGATTAAAGCTTTGCTTTGTTCCTGACG-39. The amplified fragment was cloned into pCR-BluntII-TOPO (Invitrogen) and sequenced. This ORF was subcloned into the EcoRI and XbaI sites of the pMALc2 vector (New England BioLabs), which was then transformed into E. coli strain XL1-Blue. The induction and purification of the recombinant protein were carried out using amylose resin (New England BioLabs) according to the protocol for the Protein Fusion & Purification System (New England BioLabs). EMSA The 21–27 bp oligonucleotides shown in Fig. 4B with the GG dinucleotide at the 59 end were synthesized and used as the forward oligonucleotides. The reverse oligonucleotides consisted of the sequences complementary to the forward oligonucleotides, with the TT dinucleotide at the 59 end and without CC at the 39 end. The double-stranded DNA probes were produced by boiling and slowly cooling the forward and reverse oligonucleotides in 50 mM NaCl, followed by filling in the gaps with [a-32P]dATP using the Klenow fragment. CAS contained a potential OsIRO2-binding site, as identified by the CASTing method. The mutations were introduced by converting A to C, C to A, T to G, and G to T (m1–m14). The NAS1, FDH, and MTN sequences represent the 1601/1622 region of the OsNAS1 promoter, the 1389/1410 region of the OsFDH promoter, and the 1659/1680 region of the MTN promoter, respectively. IDE2 represents the 27-base Fe-deficiencyresponsive element 2 (Kobayashi et al., 2003). The binding reaction was carried out in 20 ll of binding buffer that contained 15 mM HEPES (pH 7.5), 17.5 mM KCl, 0.4 mM EDTA, 6% glycerol, 0.05% IGEPAL CA-630 (Sigma), 500 ng of poly(dI–dC)2 (Amersham), 0.5–2.0 ng of 32P-end-labelled probe and 225–500 ng of MBP–OsIRO2. The binding reaction mixtures were incubated for 20–30 min at room temperature and then electrophoresed at room temperature at a constant 120 V in a 6% native polyacrylamide gel with 22.5 mM TRIS, 22.5 mM boric acid, and 0.5 mM EDTA.
Downloaded from http://jxb.oxfordjournals.org/ at Pennsylvania State University on February 21, 2013
Isolation of HvIRO2 A full-length cDNA library of barley was constructed using mRNA prepared from Fe-deficient roots and cloned into pCMVFL3 (TOYOBO). Approximately 250 000 colonies were screened with the same probe and under the same hybridization conditions as those used for the northern blot analysis of OsIRO2. Thirty positive clones were isolated and divided into two groups according to insert length. After sequencing using the DSQ-2000L DNA sequencer (Shimadzu), both groups of clones were found to encode the same protein. The longest clone, which contained a 1.4 kb insert, was used for northern blot analysis of HvIRO2.
2869
2870 Ogo et al. Searching for OsIRO2-binding sequences in plant promoters The upstream sequences of various rice genes were derived from the rice genome database based on the positions of the putative initiation codon of the longest ORF in the full-length cDNA clones (KOME at http://cdna01.dna.affrc.go.jp/cDNA/). With respect to Fe-deficiencyinducible genes, 16 genes for which Fe deficiency inducibility had been confirmed by northern analysis, as well as 42 genes that showed up-regulated expression from day 3 to day 7 in the microrarray analysis were chosen for this analysis. One hundred genes on the array were selected randomly from the Fe-deficiency-uninducible genes. The HvNAS1 (AB022688), HvNAAT-A (AB024006), HvNAAT-B (AB024006), IDS2 (D15051), and IDS3 (AB024007) genes from barley in addition to Fe-deficiency-inducible rice genes were used.
Results Downloaded from http://jxb.oxfordjournals.org/ at Pennsylvania State University on February 21, 2013
Microarray analysis of genes related to gene regulation and induced during Fe deficiency
To investigate changes in gene expression related to Fe deficiency in rice, a microarray analysis was conducted. The oligo-DNA microarray used included 21 938 rice genes, which were represented by their unique 60mer oligonucleotides. These genes were characterized by their homologies to other plant proteins and from predictions of their subcellular localizations (KOME at http://cdna01. dna.affrc.go.jp/cDNA/). Approximately 2000 of the genes in this array were predicted to encode DNA-binding proteins, RNA-binding proteins, and/or nuclear-targeted proteins, which may be components of gene regulatory mechanisms. Among them, several genes were found to be up-regulated under Fe deficiency conditions and considered to be involved in gene regulation under Fe-deficient conditions (Fig. 1; Table 1). Induction of some genes by Fe deficiency was confirmed by northern blot (data not shown). Twenty selected genes showed three expression patterns in the shoot. The nine genes in Fig. 1A were maintained at a high level of expression during Fe deficiency. The expression of the six genes in Fig. 1B tended to be increased with progressive Fe deficiency. The expression of the five genes in Fig. 1C was increased at the initiation of Fe deficiency, but decreased subsequently. The increases in gene expression in Fig. 1A and B were dramatically repressed by re-supplying Fe (Fig. 1; on day 7+4). Some of the genes, such as AK101209 and AK102789 (Fig. 1A), AK073385 and AK103636 (Fig. 1B), and AK059581 (Fig. 1C), were up-regulated in both shoots and roots, while other genes were induced exclusively in the shoots. AK065631 was induced only in the roots (Fig. 1D). The expression of AK065631 in the roots was increased on day 1, decreased to the control level on day 2, retained this low level throughout the period of Fe deficiency, and increased once again when Fe was re-supplied. The AK073385 gene (OsIRO2) is specifically up-regulated under Fe-deficient conditions
Among the Fe deficiency-inducible genes that are putatively involved in gene regulation, AK073385, which
Fig. 1. Expression patterns of genes that are putatively involved in gene regulation. The vertical axis shows the means of the ratio of two replicates by microarray analysis. The numbers of days after the initiation of Fe deficiency are shown on the horizontal axis. (A) Genes that maintain induced expression in the shoots during the period of Fe deficiency. (B) Genes that show increased expression levels in the shoots with the progression of Fe deficiency. (C) Genes that show increased expression levels in the shoots only at the beginning of Fe deficiency. (D) Fedeficiency-inducible genes that are expressed exclusively in the roots.
IRO2, a novel iron-regulated transcription factor
2871
Table 1. Genes that are putatively involved in gene regulation under conditions of Fe deficiency The Fe-deficiency-inducible genes, categorized as shown in Fig. 1, are listed. Organs in which the expression was up-regulated are shown as R (root) or S (shoot). Accession no.
Organ
Myb transcription factor bHLH transcription factor DRE-binding protein like Nuclear localizing protein bHLH transcription factor Nuclear localizing protein RING finger protein Nuclear localizing protein Nuclear localizing protein
S, R S S S S S S, R S S
bHLH transcription factor (IRO2) Nuclear localizing protein OsNAC1 Zinc finger protein RING finger protein NAC-domain protein
S, R S, R S S S S
WRKY transcription factor OsNAC5 Myb transcription factor Nuclear localizing protein RNA-binding protein (RGGA)
S S S S S, R
C3HC4-type RING finger protein
R
encodes a bHLH transcription factor, was strongly induced (Fig. 1B), with a peak expression ratio in shoots of 18.4 on day 5, which was the second highest ratio on day 5 of all the genes in the microarray; the highest AK073385 expression ratio in roots was 4.5 on day 3, which was the sixth highest on day 3 of all the genes examined. This bHLH protein was named rice iron-related transcription factor 2 (OsIRO2), and it was analysed in detail. To confirm the induction of OsIRO2 in response to Fe deficiency, northern blot analysis was performed with different samples from those subjected to microarray analysis. OsIRO2 was not expressed in shoots, but was expressed constitutively at low levels in the roots of Fesufficient rice (Fig. 2A). The expression of OsIRO2 was strongly induced by Fe deficiency in both shoots and roots, and this increased expression was maintained throughout the period of Fe deficiency. In contrast, the expression of OsIRO2 was not induced by zinc, manganese, or copper deficiency in either roots or shoots (Fig. 2B). Interestingly, the expression of OsIRO2 in roots was reduced by zinc deficiency. Homologues of OsIRO2 are present exclusively in graminaceous plants
A homologue of OsIRO2, HvIRO2, was isolated from barley (Fig. 3A). The deduced amino acid sequence of HvIRO2 was 64% identical to that of OsIRO2. An in silico
Fig. 2. Northern blot analysis of OsIRO2. Each lane contains 20 lg of total RNA extracted from the roots or shoots of rice plants. A fragment from the 1051 bp region of the EST cDNA clone of OsIRO2 was used as the hybridization probe. (A) Time-course analysis of Fe deficiency in plants. Plants were harvested at 1, 3, 5, 7, 9, and 11 d after the initiation of Fe deficiency (Fe), and Fe-sufficient plants (+Fe) were harvested 0 and 7 d after treatment. (B) Expression patterns under conditions of metal deficiency. Control plants (C) were supplied with all the necessary nutrients.
search revealed that other graminaceous plants also carry bHLH genes that are highly homologous to OsIRO2. ESTs from wheat, sorghum, and maize appear to contain partial sequences of the OsIRO2 ORF (Fig. 3A). The deduced proteins coded by these bHLH genes were found to be highly homologous to each other, even outside the bHLH domain. In contrast, proteins with high levels of homology to OsIRO2 were not found in dicotyledonous plants, such as Arabidopsis, Brassica napus, tobacco, tomato, and soybean (data not shown). The expression of HvIRO2 was strongly induced by Fe deficiency in both shoots and roots and was dramatically reduced by the re-supply of Fe (Fig. 3B). The onset of induction of HvIRO2 in roots by Fe deficiency was earlier than that of OsIRO2 (Figs 2A, 3B). The expression of HvIRO2 increased more rapidly in roots than in shoots. A sequence comparison of the plant bHLH proteins revealed that IRO2 belongs to a different group from tomato FER and the Arabidopsis FIT1/FRU/AtbHLH29 proteins (Fig. 3C). The presence of three amino acid residues, histidine, glutamate, and arginine, in the basic region of the bHLH domain of IRO2 (Fig. 3A) suggests that IRO2 can bind to the G-box (CACGTG), whereas FER and FIT1/FRU/AtbHLH29 can bind to the E-box (CANNTG; non-G-box) (Toledo-Ortiz et al., 2003). Preferred recognition sequence for OsIRO2
The recognition sequence for OsIRO2 was identified by the CASTing method. Potential OsIRO2-binding sites were
Downloaded from http://jxb.oxfordjournals.org/ at Pennsylvania State University on February 21, 2013
Group A AK101209 AK059273 AK106041 AK099921 AK068704 AK101217 AK102789 AK072238 AK106365 Group B AK073385 AK103636 AK108080 AK068028 AK108361 AK063703 Group C AK066255 AK063399 AK112056 AK109364 AK059581 Group D AK065631
Putative identification
2872 Ogo et al.
enriched by seven cycles of DNA binding, isolation of nucleoprotein complexes, and re-amplification of selected oligonucleotides. Of the amplified and cloned sequences, 50 were sequenced and revealed to be independent of each other. Eighty-eight per cent of the sequenced clones contained the G-box (59-CACGTG-39) sequence. The consensus sequence was obtained by arranging the sequences of clones that contained the G-box (Fig. 4A). Interestingly, all the clones containing a G-box possessed the G residue next to the G-box at the 39 end. OsIRO2 was found to bind preferentially to 59-ACCACGTGGTTTT-39, which is an incomplete palindromic sequence that flanks the core G-box. In order to confirm the DNA-binding site of OsIRO2, one of the oligonucleotides that contained the potential OsIRO2-binding site (obtained by the CASTing method), with the sequence 59-GTTAACCACGTGGTTTTCAGC39 (the G-box is underlined), was named the CAS oligonucleotide and used in EMSA to detect the specific DNA-binding activity of OsIRO2. The end-labelled CAS oligonucleotide was used as DNA probe. Two shifted
bands that corresponded to the DNA–protein complex were observed when MBP–OsIRO2 was added, whereas these bands were not observed when MBP alone was added (Fig. 4C). The lower and upper shifted bands may correspond to the DNA–OsIRO2 monomer and DNA–OsIRO2 dimer, respectively. When IDE2, one of the Fe-deficiencyresponsive elements and not similar to CAS (Kobayashi et al., 2003), was used as a probe, no shifted band was observed (Fig. 4C). The oligonucleotides m3, m4, and m5, with mutations in the G-box, did not compete for CAS–OsIRO2 complex formation, whereas m1, m2, m6, m7, and m8 competed (Fig. 4B, D). This result indicates that the G-box is necessary for OsIRO2-binding activity and that the sequences flanking the G-box affect binding to some extent. The oligonucleotides m9, m10, and m11, with single mutations of the G-box, did not compete for the complex formation. Oligonucleotide m12, with a single mutation of the G residue next to the G-box at the 39 end, competed for complex formation to a low degree (Fig. 4D). Interestingly,
Downloaded from http://jxb.oxfordjournals.org/ at Pennsylvania State University on February 21, 2013
Fig. 3. IRO2 from other graminaceous plants. (A) Alignment of IRO2 sequences. The y symbol indicates the position of the putative bHLH domain. The filled inverted triangle indicates the consensus three amino acid residues of bHLH protein that bind to the G-box (Toledo-Qriz et al., 2003). The shaded areas represent either identical residues found in most of the proteins (black) or similar character residues (grey). The putative IRO2 sequences from sorghum, wheat, and maize come from EST clones (sorghum, CN126528; wheat-a, CD872621; wheat-b, CA502657; maize, BM347640). (B) Northern blot analysis of HvIRO2 in Fe-deficient barley. Plants were harvested 1, 3, 5, 7, and 9 d after the initiation of Fe deficiency (Fe), or 2 d after Fe was re-supplied on day 9 of Fe deficiency (9+2 d). The control plants (+Fe) were harvested 1 d after treatment. (C) Phylogenic tree for IRO2 proteins and other bHLHs in plants. The underlined proteins are predicted to recognize the G-box, while the other proteins are thought to bind to the E-box (non-Gbox), based on the predictions of Toledo-Qriz et al. (2003). The proteins studied have the following accession numbers: ORG2, AF488576; ORG3, AF488577; MYC146, AF027732; OSB1, BAB64301; RAP-1, X99548; PIF3, AF251693; FER, AAN39037; FIT1/FRU/AtbHLH29, AT2G28160; ICE1, AY079016.
IRO2, a novel iron-regulated transcription factor
2873
involved in photosynthesis (AK067730) and various kinds of metabolism (AK067591, AK071404, and AK072595), contained OsIRO2-binding core sequences in their promoter regions. Several unknown genes that were induced by Fe deficiency also possessed the core sequences. The three 21 bp native promoters of OsNAS1, OsFDH, and MTN (Fig. 4B, E) were end-labelled and used as DNA probes. All the native promoters bound OsIRO2, although the shifted bands were weaker than that observed with the CAS oligonucleotide (Fig. 4C, D, E). Oligonucleotides m4 and m5 did not compete for the complex formation, whereas the native sequences used as end-labelled probes, as well as m14 and CAS, did compete for the complex formation.
Potential target genes for OsIRO2 In the search for putative target genes regulated by IRO2 proteins, the upstream regions (;2000 bp long) of the 16 known Fe-deficiency-inducible genes and the 42 genes that showed up-regulated expression from day 3 to day 7 in the microarray analysis were used. Initially, these promoter regions were searched for the presence of Gboxes (59-CACGTG-39) (Fig. 5). G-boxes were found in 43% of the Fe-deficiency-inducible genes and in 31% of the randomly selected uninducible genes. Chi-squared analysis showed that G-box frequency in Fe-deficiencyinducible promoters was not significantly higher than in randomly selected uninducible promoters. In contrast, the OsIRO2-binding core sequence (G-box plus G; 59CACGTGG-39) was found in 31% of the Fe-deficiencyinducible genes and in 17% of the randomly selected genes. The frequency of 59-CACGTGG-39 was significantly higher in Fe-deficiency-inducible gene promoters than in randomly selected uninducible gene promoters. Fe-deficiency-inducible genes that carried an OsIRO2binding core sequence are listed in Table 2. Several genes related to MA synthesis and Fe acquisition had the OsIRO2-binding core sequence. The protein products of OsNAS1, OsNAS3 (Higuchi et al., 2001), and barley IDS3 (Kobayashi et al, 2001) participate in synthesis of MAs. Formate dehydrogenase (OsFDH), methylthioadenosine/Sadenosyl-homocysteine nucleosidase (MTN), and adenine phosphoribosyltransferase 1 (OsAPT1) participate in the methionine cycle, through which methionine is efficiently supplied for synthesis of MAs (Suzuki et al., 1998; Itai et al., 2000; Kobayashi et al., 2005). The Fe(II) transporter that is encoded by OsIRT1 functions in the absorbance of Fe(II) ions, which are abundant in the soils of paddy fields (Bughio et al., 2002; Ishimaru et al., 2006). The OsIRO2binding core sequences were also found in AK063399 (OsNAC5), which encodes a transcription factor (Fig. 1). Furthermore, AK103890, the function of which is related to protein degradation through ubiquitination, and genes
Discussion This study shows the gene regulators that would participate in the Fe acquisition mechanism in graminaceous plants (Fig. 1). The genes for three NAC-domain proteins (AK108080 for OsNAC1, AK063703 for NAC domain protein, and AK063399 for OsNAC5) were up-regulated under Fe-deficient conditions in shoots. NAC-domain transcription factors are unique to plants and comprise a large gene family (Ren et al., 2000; Xie et al., 2002; Duval et al., 2002; Hegedus et al., 2003). Several studies have reported that the expression of genes for NAC-domain proteins is induced by wounding, pathogen infection, and/ or drought (Collinge and Boller, 2001; Hegedus et al., 2003; Tran et al., 2004; Ohnishi et al., 2005). Our study is the first to show that genes that encode NAC-domain proteins are up-regulated by metal deficiency. There are some reports of Fe-deficiency-inducible transcription factor genes revealed by microarray analysis. Previously, it had been found that two zinc finger genes and one leucine zipper protein gene were up-regulated by Fe deficiency in barley roots (Negishi et al., 2002). In Arabidopsis, in addition to FIT1, three RING finger protein genes and a putative DNAbinding protein gene are up-regulated by Fe deficiency (Colangelo and Guerinot, 2004). The genes shown in Fig. 1 are reported here for the first time to be Fe deficiency inducible. This wide variety of transcription factors appears to participate in the complex regulatory networks that function under Fe-deficiency stress. Among the Fe-deficiency-inducible genes that are putatively involved in gene regulation, the AK073385 gene, the expression of which was highly Fe deficiency inducible, encodes a novel bHLH protein, which has been named OsIRO2. The induction pattern of OsIRO2 under conditions of Fe deficiency resembled those of other genes induced by Fe deficiency (Inoue et al., 2003; Kobayashi et al., 2005; Ishimaru et al., 2006). The induction of OsIRO2 expression did not occur under conditions of zinc, manganese, or copper deficiency (Fig. 2B). These results suggest
Downloaded from http://jxb.oxfordjournals.org/ at Pennsylvania State University on February 21, 2013
oligonucleotide m13, in which all of the nucleotides outside the G-box were mutated, did not compete for DNA– OsIRO2 complex formation. On the other hand, m14, in which the nucleotides outside the G-box plus the G next to the G-box (59-CACGTGG-39) were mutated, competed for complex formation (Fig. 4D). These results indicate that the G-box is not sufficient for OsIRO2 binding, and the G next to the G-box is also required. Substantial competition by m6 and m12, which are mutated at the G at the seventh position of 59-CACGTGG-39, yet containing 59CCACGTG-39 in their complementary strands, indicates that either ‘G’ at the 39 flank or ‘C’ at the 59 flank of G-box, but not both, is sufficient for OsIRO2 binding. Consequently, the ‘G-box plus G’ sequence (59-CACGTGG-39) was determined to be the OsIRO2-binding core sequence.
2874 Ogo et al.
Downloaded from http://jxb.oxfordjournals.org/ at Pennsylvania State University on February 21, 2013
Fig. 4. OsIRO2-binding sequence. (A) Identification of the consensus nucleotide sequence that acts as the binding site for OsIRO2 using the CASTing method. The binding sequences of clones that contain the G-box are arranged by centring the G-box. The frequencies at which individual nucleotides appear are shown in percentages on the vertical axis. The letter F indicates sequences that could not be confirmed. (B–E) EMSA using MBP–OsIRO2 and the oligonucleotide that contains the OsIRO2-binding site, obtained by CASTing. (B) Nucleotide sequences used for the probes and competitors. One of the oligonucleotides (CAS) that contained a potential OsIRO2-binding site, as identified by the CASTing method, was used as a probe. The putative OsIRO2-binding site determined by CASTing is underlined. The G-box is shaded in grey and the mutated sequences for the competition assays are shaded in black. The mutations were introduced by converting A to C, C to A, T to G, and G to T. The NAS1, FDH, and MTN sequences represent the 1601/1622 region of the OsNAS1 promoter, the 1389/1410 region of the OsFDH promoter, and the 1659/1680 region of the MTN promoter, respectively. IDE2 represents the 27 base Fe-deficiency-responsive element 2 (Kobayashi et al., 2003). (C) The end-labelled CAS or IDE2 was incubated with MBP–OsIRO2 or MBP. Arrowheads indicate two shifted bands and the free probe. (D) The effect of mutations in the binding site on the formation
IRO2, a novel iron-regulated transcription factor
Fig. 5. Frequencies of the OsIRO2-binding core sequence. The frequencies of identified G-box (CACGTG) and OsIRO2-binding core sequence (CACGTGG) in the promoter regions (within the 2 kb upstream regions) of Fe-deficiency-inducible genes (black bars) and randomly selected Fe-deficiency-uninducible genes (white bars).
abolished OsIRO2 binding (Fig. 4D). In addition, the Gbox alone is not sufficient for OsIRO2 binding, as evidenced by a strict requirement for a G next to the Gbox for successful binding (Fig. 4D). The residue next to the G-box at the 39 end must be G, but not T, A, or C, because all the clones containing the G-box obtained by CASTing possessed the G residue next to the G-box at the 39 end (Fig. 4A). Therefore, the ‘G-box plus G’ sequence of 59-CACGTGG-39 is considered to be the OsIRO2-binding core sequence. Furthermore, the sequences flanking the OsIRO2-binding core sequences affect OsIRO2 binding activity to some extent, since probes mutated in nucleotides outside the core sequence competed, to a low degree, for the complex formation (Fig. 4D). Although not many transcription factors recognizing a long DNA sequence such as 8 or 9 bp have been reported, a further analysis of the binding sequence of OsIRO2 might be needed. The determined OsIRO2-binding sequence is not included in the previously identified Fe-deficiency-responsive cis-elements
of the OsIRO2–DNA complexes, as determined by competition assays. EMSA was performed using 0.5 ng of end-labelled CAS oligonucleotide and 225 ng of MBP–OsIRO2 protein. A 50-, 200-, or 500-fold excess of CAS, or of each mutated oligonucleotide, was added to the reaction mixture as unlabelled competitor. The shifted bands are indicated. (E) Binding of OsIRO2 to the native promoters of OsNAS1, OsFDH, and MTN, and the competition assay. The end-labelled native promoters (2 ng) were incubated with 500 ng of MBP–OsIRO2. A 50- or 100-fold excess of each native promoter and mutated oligonucleotide was added to the reaction mixture as unlabelled competitors.
Downloaded from http://jxb.oxfordjournals.org/ at Pennsylvania State University on February 21, 2013
that OsIRO2 plays a specific role in plant responses to Fe deficiency. We isolated an OsIRO2 homologue from barley, encoded by HvIRO2, the expression of which in Fe-deficient barley was almost the same as that in rice and lacked root specificity (Figs 2A, 3B). An in silico search for proteins homologous to IRO2 revealed that IRO2 is widely conserved among graminaceous plants (Fig. 3A). In contrast, it was difficult to find bHLH proteins that were highly similar to IRO2 outside of the bHLH domain of dicot plants (data not shown). IRO2 appears to function in a gene regulatory system that is particular to graminaceous plants that grow under Fe-deficient conditions. However, this may not relate to root-specific gene regulatory systems, such as the regulation of HvNASs and NASHOR1 expression in Fe-deficient barley (Higuchi et al., 1999; Douchkov et al., 2002). In dicotyledons, the bHLH transcription activators FER and FIT1/FRU/bHLH29 are implicated in Fe nutrition (Ling et al., 2002; Colangelo and Guerinot, 2004; Jakoby et al., 2004; Yuan et al., 2005). The expression of FIT1/ FRU/bHLH29 is detectable in Fe-sufficient roots and is increased by Fe deficiency, while the expression of FER is not up-regulated by Fe deficiency. FER and FIT1/FRU/ bHLH29 are predicted to bind to the E-box (59-CANNTG-39; non-G-box; Fig. 3C). A FER-like protein was also found in the rice genome database (XP_472244; ‘OsFER-like’ in Fig. 3C). However, this putative bHLH gene may not be expressed in rice, as the respective EST clone has not been found to date. In contrast, this study shows that OsIRO2 binds preferentially to the G-box (Fig. 4). Furthermore, both OsIRO2 and HvIRO2 were strongly up-regulated by Fe deficiency in both roots and shoots (Figs 2A, 3B), while FER and FIT1/FRU/bHLH29 were expressed only in roots (Ling et al., 2002; Colangelo and Guerinot, 2004; Jakoby et al., 2004). These results suggest that IRO2 in graminaceous plants and FER and FIT1/FRU/bHLH29 in dicotyledonous plants probably have different functions. The bHLH family of proteins is a group of functionally diverse transcription factors that are found in both plants and animals. The bHLH proteins bind as either homo- or heterodimers to specific DNA target sites. Our EMSA results revealed the possibility that OsIRO2 binds to DNA as a homodimer, as two shifted bands were detected (Fig. 4). The preferred DNA-binding sequence of OsIRO2, 59ACCACGTGGTTTT-39, was determined experimentally (Fig. 4A). EMSA revealed that the G-box in the OsIRO2binding site determined by CASTing is necessary for the binding of OsIRO2 to DNA; a mutation in even one base
2875
2876 Ogo et al. Table 2. Promoter analysis of the OsIRO2-binding sequence Listed are Fe- deficiency-inducible genes whose 2 kb upstream regions contain the OsIRO2-binding core sequence. The positions are numbered from the putative translation start site. The –Fe/day 0 ratios reflect the results from day 5 of the microarray analysis. The following four genes are not included in the array, although their expression levels under conditions of Fe deficiency have been reported to be up-regulated ("), down-regulated (#), or unchanged (–): OsNAS3 (Inoue et al., 2003); barley IDS3 (Nakanishi et al., 1993); MTN (Kobayashi et al., 2005); and OsIRT1 (Bughio et al., 2002). Accession no.
Gene name or putative function
Sequence
Position
ACCACGTGGTTTT MAs and Fe acquisition-related AB046401 Nicotianamine synthase 1 (OsNAS1) AB023819 Nicotianamine synthase 3 (OsNAS3)
AK107681
Iron-regulated metal transporter 1 (OsIRT1)
Expressional regulation-related AK063399 OsNAC5 Others AK067730 Chlorophyll b synthase AK067591 Cytochrome P450 AK071404 Lipase AK072595 AK103890 Unknown AK063263 AK068661 AK100816 AK107748 AK108093 AK110584
Tryptophan synthase Ubiquitin extension protein 1 (UBQ1) Unknown Unknown Unknown Unknown Unknown Unknown
IDE1 and IDE2, nor is it similar to these elements. The OsIRO2-binding sequence includes the high affinity sequence (59-ACCACGTGGT-39) of the c-Myc protein, which is implicated in animal cancers (Halazonetis and Kandil, 1991; Blackwell et al., 1993). Among the plant ciselements that have been reported, the sequence of 59GCCACGTGGG-39, which is located in the promoter region of the maize ABA-inducible rab28 gene (Pla et al., 1993), is very similar to the OsIRO2-binding sequence. Many of the cis-elements known as ABA-responsive elements (ABREs) share the 59-(C/T)ACGTGGC-39 motif, and this consensus sequence is sometimes called the G-boxlike ABRE (Choi et al., 2000). However, the G-box-like ABRE is generally bound by basic leucine zipper proteins, and two or more G-box-like ABREs, or the coupling elements, are needed for function (Shen and Ho, 1995; Shen et al., 1996; Hobo et al., 1999). The frequencies of the OsIRO2-binding sequence were estimated using the upstream sequences of 58 Fedeficiency-inducible genes and 100 Fe-deficiencyuninducible genes in rice (Fig. 5). The frequency of the OsIRO2-binding core sequence was significantly higher in Fe-deficient inducible genes, while the fre-
Shoot
Root
tgCACGTGGTggg gaCACGTGGgcca AtCACGTGGgTcT gCCACGTGGCaTg gaCACGTGGTTTa tCCACGTGGTggg
1610 152 365 1097 1395 1668
6.14 #
3.21 "
– 0.91 –
" 2.78 "
gaCACGTGGgccc gaCACGTGGgccc
242 133
2.99
2.40
caCACGTGGaacg
1234
–
"
cCCACGTGGcTga
1317
2.32
1.34
gCCACGTGGTgcg gaCACGTGGaaat gaCACGTGGaaaa tgCACGTGGcgTg gaCACGTGGgaga gaCACGTGGcacg
832 1271 371 309 357 293
2.43 3.15 3.42
5.43 2.75 2.60
4.42 4.58
1.30 3.41
AgCACGTGGgggg tgCACGTGGTact ctCACGTGGccgc gaCACGTGGcggc caCACGTGGTcag cgCACGTGGTTgT cCCACGTGGGagT
850 984 1950 140 641 1040 186
2.33 11.09 4.66 2.41 3.20 10.26
1.25 1.59 0.90 0.91 1.70 1.78
quency of the G-box alone was not significantly different for Fe-deficiency-inducible genes and Fe-deficiencyuninducible genes (Fig. 5). Many transcription factors with the ability to bind to the G-box have been reported (Siberil et al., 2001), and the G-box may occur frequently in the promoter regions of various genes, where it regulates the expression of genes not related to Fe nutrition. The significant difference observed for the frequency of the OsIRO2-binding core sequence suggests that this sequence is an important cis-element in the promoter regions of Fe-deficiency-inducible genes, and that OsIRO2 is one of the key transcription factors under conditions of Fe deficiency. It is noteworthy that many genes related to Fe acquisition have the OsIRO2-binding core sequence in their promoter regions (Table 2), which suggests that OsIRO2 plays a role in the transcriptional regulation of genes that participate in Fe acquisition under Fe-deficient conditions. Our EMSA results revealed that OsIRO2 binds to the native promoters of OsNAS1, OsFDH, and MTN with strict sequence specificity (Fig. 4E). Since oligonucleotides m4 and m5 did not compete for the complex formation, the G-box of the native promoter sequences was bound by OsIRO2. The strong competition shown by
Downloaded from http://jxb.oxfordjournals.org/ at Pennsylvania State University on February 21, 2013
AK073627
Barley mugineic acid synthase IDS3 Formate dehydrogenase (OsFDH) Methylthioadenosine/S-adenosyl homocysteine nucleosidase (MTN) Adenine phosphoribosyltransferase 1 (OsAPT1)
AB024007 AK065872 AK066172
Ratio (–Fe/day 0)
IRO2, a novel iron-regulated transcription factor
Accession numbers The sequence of HvIRO2 has been deposited in DDBJ, EMBL, and GenBank with the accession number AB206536.
Acknowledgements We thank Dr Yoshiaki Nagamura and his colleagues at the National Institute of Agrobiological Sciences (NIAS) for assistance with the microarray experiments.
References Blackwell TK, Huang J, Ma A, Kretzner L, Alt FW, Eisenman RN, Weintraub H. 1993. Binding of myc proteins to canonical and noncanonical DNA sequences. Molecular and Cellular Biology 13, 5216–5224. Bughio N, Yamaguchi H, Nishizawa NK, Nakanishi H, Mori S. 2002. Cloning an iron-regulated metal transporter from rice. Journal of Experimental Botany 53, 1677–1682. Chinnusamy V, Ohata M, Kanrar S, Lee BH, Hong X, Agarwal M, Zhu JK. 2003. The basic helix–loop–helix trans-
cription factor family in plants: a genome-wide study of protein structure and functional diversity. Molecular Biology and Evolution 20, 735–747. Choi H, Hong J, Ha J, Kang J, Kim SY. 2000. ABFs, a family of ABA-responsive element binding factors. Journal of Biological Chemistry 275, 1723–1730. Church GM, Gilbert W. 1984. Genomic sequencing. Proceedings of the National Academy of Sciences, USA 81, 1991–1995. Colangelo EP, Guerinot ML. 2004. The essential basic helix–loop– helix protein FIT1 is required for the iron deficiency response. The Plant Cell 16, 3400–3412. Collinge M, Boller T. 2001. Differential induction of two potato genes, Stprx2 and StNAC, in response to infection by Phytophthora infestans and to wounding. Plant Molecular Biology 46, 521–529. Curie C, Panaviene Z, Loulergue C, Dellaporta SL, Briat JF, Walker EL. 2001. Maize yellow stripe1 encodes a membrane protein directly involved in Fe(III) uptake. Nature 409, 346–349. Douchkov D, Herbik A, Koch G, Mock HP, Melzer M, Stephan UW, Baumlein H. 2002. Nicotianamine synthase: gene isolation, gene transfer and application for the manipulation of plant iron assimilation. Plant and Soil 241, 115–119. Duval M, Hsieh TF, Kim SY, Thomas TL. 2002. Molecular characterization of AtNAM: a member of the Arabidopsis NAC domain superfamily. Plant Molecular Biology 50, 237–248. Gilmour SJ, Zarka DG, Stockinger EJ, Salazar MP, Houghton JM, Thomashow MF. 1998. Low temperature regulation of the Arabidopsis CBF family of AP2 transcriptional activators as an early step in cold-induced COR gene expression. The Plant Journal 16, 433–442. Halazonetis TD, Kandil AN. 1991. Determination of the c-MYC DNA-binding site. Proceedings of the National Academy of Sciences, USA 88, 6162–6166. Hegedus D, Yu M, Baldwin D, Gruber M, Sharpe A, Parkin I, Whitwill S, Lydiate D. 2003. Molecular characterization of Brassica napus NAC domain transcriptional activators induced in response to biotic and abiotic stress. Plant Molecular Biology 53, 383–397. Higuchi K, Suzuki K, Nakanishi H, Yamaguchi H, Nishizawa NK, Mori S. 1999. Cloning of nicotianamine synthase genes, novel genes involved in the biosynthesis of phytosiderophores. Plant Physiology 119, 471–479. Higuchi K, Watanabe S, Takahashi M, Kawasaki S, Nakanishi H, Nishizawa NK, Mori S. 2001. Nicotianamine synthase gene expression differs in barley and rice under Fe-deficient conditions. The Plant Journal 25, 159–167. Hobo T, Asada M, Kowyama Y, Hattori T. 1999. ACGTcontaining abscisic acid response element (ABRE) and coupling element 3 (CE3) are functionally equivalent. The Plant Journal 19, 679–689. Inoue H, Higuchi K, Takahashi M, Nakanishi H, Mori S, Nishizawa NK. 2003. Three rice nicotianamine synthase genes, OsNAS1, OsNAS2, and OsNAS3 are expressed in cells involved in long-distance transport of iron and differentially regulated by iron. The Plant Journal 36, 366–381. Ishimaru Y, Suzuki M, Tsukamoto T, et al. 2006. Rice plants take up iron as an Fe3+ phytosiderophore and as Fe2+. The Plant Journal 45, 335–346. Itai R, Suzuki K, Nakanishi H, Yamaguchi H, Nishizawa NK, Mori S. 2000. Induced activity of adenine phosphoribosyltransferase (APRT) in iron-deficient barley roots: a possible role for phytosiderophore production. Journal of Experimental Botany 51, 1179–1188. Jakoby M, Wang HY, Reidt W, Weisshaar B, Bauer P. 2004. FRU (BHLH029) is required for induction of iron mobilization genes in Arabidopsis thaliana. FEBS Letters 577, 528–534.
Downloaded from http://jxb.oxfordjournals.org/ at Pennsylvania State University on February 21, 2013
m14 indicates that the OsIRO2-binding core sequence is essential for the binding activity in these native promoters. The core sequences were also found in several Fedeficiency-inducible genes that participate in transcription, photosynthesis, and metabolism. The roles of these genes under Fe-deficient conditions are currently unknown. The previously identified IDE1 and IDE2 cis-acting elements mediate gene expression that is both Fe-deficiencyinducible and root-specific (Kobayashi et al., 2003). Recently, many of the Fe-deficiency-inducible genes, the products of which participate in the production of MAs, were found to have sequences homologous to IDE1 and IDE2 in their promoter regions (Kobayashi et al., 2005). Of the genes shown in Table 2, OsNAS1, barley IDS3, OsFDH, MTN, OsAPT1, OsIRT1, and the AK068661 gene of unknown functionality all carry IDEs in their promoter regions (Kobayashi et al., 2003, 2005). The OsIRO2 gene promoter has an IDE2-like sequence at 1148/1174 (Kobayashi et al., 2005), and contains sequences that are moderately homologous to IDE1 (data not shown). Therefore, the following model is proposed for transcription regulation under Fe-deficient conditions in graminaceous plants. Under conditions of Fe deficiency, unknown IDEbinding proteins activate the expression of genes that possess IDEs in their promoter regions, such as IRO2. The induced IRO2 protein not only induces the expression of genes that lack IDEs in their promoter regions, but also further activates the expression of genes that are directly regulated by IDEbinding proteins, such as OsNAS1, barley IDS3, and OsFDH, through binding to the 59-CACGTGG-39 motifcontaining sequences in their upstream regions. IRO2 and IDE-binding proteins may function as part of a coordinated gene network under conditions of Fe deficiency.
2877
2878 Ogo et al. Pla M, Vilardell J, Guiltinan MJ, Marcotte WR, Niogret MF, Quatrano RS, Pages M. 1993. The cis-regulatory element CCACGTGG is involved in ABA and water-stress responses of the maize gene rab28. Plant Molecular Biology 21, 259–266. Ren T, Qu F, Morris TJ. 2000. HRT gene function requires interaction between a NAC protein and viral capsid protein to confer resistance to turnip crinkle virus. The Plant Cell 12, 1917–1926. Shen Q, Ho TH. 1995. Functional dissection of an abscisic acid (ABA)-inducible gene reveals two independent ABA-responsive complexes each containing a G-box and a novel cis-acting element. The Plant Cell 7, 295–307. Shen Q, Zhang P, Ho TH. 1996. Modular nature of abscisic acid (ABA) response complexes: composite promoter units that are necessary and sufficient for ABA induction of gene expression in barley. The Plant Cell 8, 1107–1119. Shojima S, Nishizawa N, Mori S. 1989. Establishment of a cell-free system for the biosynthesis of nicotianamine. Plant and Cell Physiology 30, 673–677. Siberil Y, Doireau P, Gantet P. 2001. Plant bZIP G-box binding factors. Modular structure and activation mechanisms. European Journal of Biochemistry 268, 5655–5666. Suzuki K, Itai R, Suzuki K, Nakanishi H, Nishizawa NK, Yoshimura E, Mori S. 1998. Formate dehydrogenase, an enzyme of anaerobic metabolism, is induced by iron deficiency in barley roots. Plant Physiology 116, 725–732. Takagi S. 1976. Naturally occurring iron-chelating compounds in oat- and rice-root washings. I. Activity measurement and preliminary characterization. Soil Science and Plant Nutrition 22, 423–433. Takahashi M, Yamaguchi H, Nakanishi H, Shioiri T, Nishizawa NK, Mori S. 1999. Cloning two genes for nicotianamine aminotransferase, a critical enzyme in iron acquisition (Strategy II) in graminaceous plants. Plant Physiology 121, 947–956. Toledo-Ortiz G, Huq E, Quail PH. 2003. The Arabidopsis basic/ helix–loop–helix transcription factor family. The Plant Cell 15, 1749–1770. Tran LS, Nakashima K, Sakuma Y, et al. 2004. Isolation and functional analysis of arabidopsis stress-inducible NAC transcription factors that bind to a drought-responsive cis-element in the early responsive to dehydration stress 1 promoter. The Plant Cell 16, 2481–2498. Xie Q, Guo HS, Dallman G, Fang S, Weissman AM, Chua NH. 2002. SINAT5 promotes ubiquitin-related degradation of NAC1 to attenuate auxin signals. Nature 419, 167–170. Yuan YX, Zhang J, Wang DW, Ling HQ. 2005. AtbHLH29 of Arabidopsis thaliana is a functional ortholog of tomato FER involved in controlling iron acquisition in strategy I plants. Cell Research 15, 613–621.
Downloaded from http://jxb.oxfordjournals.org/ at Pennsylvania State University on February 21, 2013
Kobayashi T, Nakayama Y, Itai RN, Nakanishi H, Yoshihara T, Mori S, Nishizawa NK. 2003. Identification of novel cis-acting elements, IDE1 and IDE2, of the barley IDS2 gene promoter conferring iron-deficiency-inducible, root-specific expression in heterogeneous tobacco plants. The Plant Journal 36, 780–793. Kobayashi T, Nakanishi H, Takahashi M, Kawasaki S, Nishizawa NK, Mori S. 2001. In vivo evidence that Ids3 from Hordeum vulgare encodes a dioxygenase that converts 29deoxymugineic acid to mugineic acid in transgenic rice. Planta 212, 864–871. Kobayashi T, Suzuki M, Inoue H, Itai RN, Takahashi M, Nakanishi H, Mori S, Nishizawa NK. 2005. Expression of iron-acquisition-related genes in iron-deficient rice is coordinately induced by partially conserved iron-deficiency-responsive elements. Journal of Experimental Botany 56, 1305–1316. Ling HQ, Bauer P, Bereczky Z, Keller B, Ganal M. 2002. The tomato fer gene encoding a bHLH protein controls iron-uptake responses in roots. Proceedings of the National Academy of Sciences, USA 99, 13938–13944. Liu Q, Kasuga M, Sakuma Y, Abe H, Miura S, YamaguchiShinozaki K, Shinozaki K. 1998. Two transcription factors, DREB1 and DREB2, with an EREBP/AP2 DNA binding domain separate two cellular signal transduction pathways in drought- and low-temperature-responsive gene expression, respectively, in Arabidopsis. The Plant Cell 10, 1391–1406. Marschner H. 1995. Mineral nutrition of higher plants, 2nd edn. London: Academic Press. Nakanishi H, Okukmura N, Umehara Y, Nishizawa NK, Chino M, Mori S. 1993. Expression of a gene specific for iron deficiency (Ids3) in the roots of Hordeum vulgare. Plant and Cell Physiology 34, 401–410. Nakanishi H, Yamaguchi H, Sasakuma T, Nishizawa NK, Mori S. 2000. Two dioxygenase genes; Ids3 and Ids2; from Hordeum vulgare are involved in the biosynthesis of mugineic acid family phytosiderophores. Plant Molecular Biology 44, 199–207. Negishi T, Nakanishi H, Yazaki J, et al. 2002. cDNA microarray analysis of gene expression during Fe-deficiency stress in barley suggests that polar transport of vesicles is implicated in phytosiderophore secretion in Fe-deficient barley roots. The Plant Journal 30, 83–94. Ohnishi T, Sugahara S, Yamada T, Kikuchi K, Yoshiba Y, Hirano HY, Tsutsumi N. 2005. OsNAC6, a member of the NAC gene family, is induced by various stresses in rice. Genes and Genetic Systems 80, 135–139. Okumura N, Nishizawa NK, Umehara Y, Ohta T, Nakanishi H, Yamaguchi H, Chino M, Mori S. 1994. A dioxygenase gene (Ids2) expressed under iron deficiency conditions in the roots of Hordeum vulgare. Plant Molecular Biology 25, 705–719.
RESEARCH PAPER
Isolation and characterization of IRO2, a novel iron-regulated bHLH transcription factor in graminaceous plants Yuko Ogo1,*, Reiko Nakanishi Itai1,2,*, Hiromi Nakanishi1, Haruhiko Inoue1,2, Takanori Kobayashi1,2, Motofumi Suzuki1, Michiko Takahashi1, Satoshi Mori1 and Naoko K. Nishizawa1,2,† 1
Received 10 April 2006; Accepted 10 May 2006
Abstract To clarify the molecular mechanism that regulates iron (Fe) acquisition in graminaceous plants, a time-course analysis of gene expression during Fe deficiency stress was conducted using a rice 22K oligo-DNA microarray. Twenty-one genes for proteins that function in gene regulation were induced by Fe deficiency. Of these genes, a putative basic helix–loop–helix (bHLH) transcription factor gene, named OsIRO2, was strongly expressed in both roots and shoots during Fe deficiency stress. The expression of OsIRO2 was induced exclusively by Fe deficiency, and not by deficiencies in other metals. Expression of the barley HvIRO2 gene, which is a homologue of OsIRO2, was also induced by Fe deficiency. An in silico search revealed that IRO2 is highly conserved among graminaceous plants, which include wheat, sorghum, and maize. The cyclic amplification and selection of targets (CASTing) technique revealed that OsIRO2 bound preferentially to the sequence 59-ACCACGTGGTTTT39, and the electrophoretic mobility shift assay revealed 59-CACGTGG-39 as the core sequence for OsIRO2 binding. Sequences similar to the OsIRO2binding sequence were found upstream of several genes that are involved in Fe acquisition, such as OsNAS1, OsNAS3, OsIRT1, OsFDH, OsAPT1, and IDS3. The core sequence of the OsIRO2-binding sequence
occurred more frequently in the upstream regions of Fe deficiency-inducible genes than in the corresponding regions of non-inducible genes. These results suggest that IRO2 is involved in the regulation of gene expression under Fe-deficient conditions. Key words: Basic helix–loop–helix, iron deficiency, microarray, mugineic acids, phytosiderophore, rice, transcription factor.
Introduction Iron (Fe) is an essential element for plant growth. Although abundant in soil, Fe is mainly present as oxidized Fe(III) compounds, which are poorly soluble in neutral-to-alkaline soils. Consequently, plants grown in calcareous soils often exhibit severe chlorosis due to Fe deficiency; this is a major agricultural problem that results in reduced crop yields (Marschner, 1995). Gaining an understanding of the molecular and cellular events related to Fe acquisition and its regulation is a fundamental issue in plant biology, and is a prerequisite for the genetic improvement of plants cultivated in extreme environments. Graminaceous plants have evolved a strategy for Fe acquisition that utilizes Fe(III) chelators, i.e. mugineic acid family phytosiderophores (MAs). Graminaceous plants secrete MAs from their roots to solubilize rhizospheric Fe(III) (Takagi, 1976). Subsequently, Fe(III)–MA complexes are
* These authors contributed equally to this work. y To whom correspondence should be addresssed. E-mail: [email protected] Abbreviations: ABRE, ABA-responsive element; bHLH, basic helix–loop–helix; CASTing, cyclic amplification and selection of targets; EMSA, electrophoretic mobility shift assay; EST, expressed sequence tag; MAs, mugineic acid family phytosiderophores; MBP, maltose-binding protein; ORF, expressed sequence tag. ª 2006 The Author(s). This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/2.0/uk/) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.
Downloaded from http://jxb.oxfordjournals.org/ at Pennsylvania State University on February 21, 2013
Graduate School of Agricultural and Life Sciences, The University of Tokyo, 1-1 Yayoi, Bunkyo-ku, Tokyo 113-8657, Japan 2 Core Research for Evolutional Science and Technology (CREST), Japan Science and Technology Corporation (JST), Tokyo, Japan
2868 Ogo et al.
mobility shift assay (EMSA). Several genes that are involved in Fe acquisition in plants were found to possess sequences homologous to the OsIRO2-binding sequence in their upstream regions. Materials and methods Plant materials Both rice (Oryza sativa L. cv. Nipponbare) and barley (Hordeum vulgare L. cv. Ehimehadaka no. 1) were grown hydroponically, as described previously (Kobayashi et al., 2005). Rice was grown under 14/10 h light/dark cycles and 30/25 8C, and barley was grown under 14/10 h light/dark cycles and 22/17 8C. For the rice microarray analysis, Fe deficiency was initiated 3 weeks after germination by omitting Fe(III)-EDTA from the culture medium. Plants were harvested on days 0, 1, 2, 3, 5, and 7 of Fe deficiency. On day 7 of Fe deficiency, Fe was re-supplied by adding Fe(III)-EDTA at the usual concentration, and the plants were harvested 4 d later (day 7+4). For the northern blot analysis of rice plants, Fe deficiency treatment was initiated 5 weeks after germination, and plants were harvested on days 0, 1, 3, 5, 7, 9, and 11. As a control, plants grown under Fesufficient conditions were harvested on day 7. For the micronutrient deficiency treatments, 2-week-old plants were transferred to nutrient media that lacked zinc, Fe, manganese, or copper, and grown for an additional 2 weeks. For the northern blot analysis of barley, 3week-old plants were transferred to a nutrient medium that lacked Fe. Plants were harvested on days 1, 3, 5, 7, and 9. Plants that were re-supplied with Fe on day 9 were grown for an additional 2 d (day 9+2). As a control, plants grown under Fe-sufficient conditions were harvested on day 1. Oligo-DNA microarray analysis The rice 22K oligo-DNA microarray (Agilent Technology) contained 21 938 unique 60mer oligonucleotides that were synthesized based on sequence data from the Rice Full-length cDNA Project (http:// cdna01.dna.affrc.go.jp/cDNA/). Total RNA samples were prepared from the roots and shoots of three plants at each time point using the guanidium isothiocyanate/CsCl density gradient centrifugation method. The integrity of each RNA sample was checked using the Agilent 2100 bioanalyser (Agilent Technology). Total RNA samples (200 ng) from plants harvested on day 0, and at each subsequent time point, were labelled with Cy3 or Cy5 using the Agilent Low RNA Input Fluorescent Linear Amplification Kit (Agilent Technology). Hybridization was performed according to the manufacturer’s instructions. Images of the hybridized microarrays were obtained with the Agilent microarray scanner (Agilent Technology) and analysed with feature-extraction software (Agilent Technology). The reproducibility of the microarray analysis was assessed by a dye swap in each experiment. The signal intensity was normalized by the LOWESS method. The ratios were calculated as a value (the plants harvested at each time point of Fe deficiency)/(the plants on day 0). The point showing the signal value of Fe deficiency-treated plants >500, the P-value 2 in both Cy3 and Cy5 channels was considered as significantly up-regulated. Northern blot analysis The 1051 bp expressed sequence tag (EST) cDNA clone of OsIRO2 in pBluescript was provided by the National Institute of Agrobiological Sciences (Japan) and sequenced for both DNA strands. For the OsIRO2 probe, the fragment with an XhoI site in the multicloning site at the 59 end of the open reading frame (ORF) and a HindIII site located just before the termination codon was used for northern blot analysis of rice plants. For the HvIRO2 probe, the XhoI fragment that contained the complete insert of the longest clone in pCMVFL3 was
Downloaded from http://jxb.oxfordjournals.org/ at Pennsylvania State University on February 21, 2013
taken up through the Fe(III)–MAs transporter (Curie et al., 2001). MAs are biosynthesized from methionine through nicotianamine (NA) (Shojima et al., 1989). Two dioxygenases convert 29-deoxymugineic acid (DMA), the first member of the MAs, into various kinds of MAs (Nakanishi et al., 2000; Kobayashi et al., 2001). The genes that encode key enzymes for MA synthesis have been isolated from graminaceous plants, such as barley and rice. Several NA synthase (NAS) genes (Higuchi et al., 1999, 2001; Douchkov et al., 2002; Inoue et al., 2003), two NA aminotransferase (NAAT) genes (Takahashi et al., 1999), and two dioxygenase genes, IDS2 (Okumura et al., 1994) and IDS3 (Nakanishi et al., 1993), are all up-regulated by Fe deficiency. Many of the genes involved in supplying methionine for the production of MAs under Fe-deficient conditions have been identified in barley and rice (Negishi et al., 2002; Kobayashi et al., 2005). All of these genes are induced in Fe-deficient plants. As an important clue to the gene regulatory mechanisms that operate under Fe-deficient conditions, the novel cisacting iron-deficiency-responsive elements IDE1 and IDE2 were identified in the barley IDS2 promoter (Kobayashi et al., 2003). IDE1 and IDE2 confer Fe deficiency inducibility and root-specific expression in heterogeneous tobacco plants. The results of a study conducted by Kobayashi et al. (2003) indicate that graminaceous plants and other higher plants share conserved gene regulatory mechanisms that are active under Fe-deficient conditions. Transcription factors that bind to IDE1 and IDE2 have not yet been identified. A basic helix–loop–helix (bHLH) protein, FER, was identified using the fer mutant of tomato, which is unable to induce certain genes that are involved in Fe acquisition under Fe-deficient conditions (Ling et al., 2002). A homologue of FER in Arabidopsis, FIT1/FRU/ AtbHLH29, plays a role similar to that of FER (Colangelo and Guerinot, 2004; Jakoby et al., 2004; Yuan et al., 2005). Although it is assumed that these proteins bind to the E-box motif (CANNTG) (Toledo-Ortiz et al., 2003), their DNAbinding sites have not yet been determined. Many stressresponsive transcription factors that regulate the expression of genes that encode key proteins for stress tolerance have been reported (Gilmour et al., 1998; Liu et al., 1998; Chinnusamy et al., 2003). Since the expression of these key transcription factors can be transient, it is important to detect transient changes in gene expression to identify potential regulators of stress-inducible genes. In this study, a time-course analysis of gene expression during Fedeficiency stress was performed using a rice 22K oligoDNA microarray. Genes, the products of which were suspected of involvement in gene regulation, were searched and analysed. Of theses genes, OsIRO2, which encodes a putative bHLH transcription factor, was strongly induced in shoots and roots by Fe deficiency. The OsIRO2-binding sequence was revealed by the cyclic amplification and selection of targets (CASTing) method and electrophoretic
IRO2, a novel iron-regulated transcription factor 32
used. These fragments were labelled with [a- P]dATP using the random labelling method, and the labelled DNA was purified using a ProbeQuant G-50 micro-column (Pharmacia). Total RNA samples from roots and shoots were extracted by guanidium isothiocyanate extraction for rice and by SDS–phenol extraction for barley. Total RNA (20 lg per lane) was separated on 1.4% (w/v) agarose gels that contained 5% (v/v) formaldehyde and blotted onto nylon membranes (Hybond-N+; Amersham). The blots were hybridized at 65 8C overnight with labelled probes in 0.5 M Church phosphate buffer (Church and Gilbert, 1984), 1 mM EDTA, and 7% (w/v) SDS with 100 lg ml1 salmon sperm DNA. After hybridization, the blots were washed twice with 40 mM Church phosphate buffer and 1% (w/v) SDS at 42 8C for 10 min and then washed at high stringency with 23 SSPE plus 0.1% (w/v) SDS at 45 8C for 10 min. Radioactivity was detected using the FLA-3000 imaging analyser (Fuji Film).
Preparation of thioredoxin–OsIRO2 fusion protein for CASTing The ORF of OsIRO2 was amplified using the primers 59-CACCATGGAGCAGCTGTTCGTCGACGACC-39 and 59-GAAAGTCGTCAGGAACAAAGCAAAGCTT-39. The amplified fragment was cloned into pENTR/D-TOPO (Invitrogen) and sequenced. The pET32a(+) vector (Novagen) was converted into the Gateway destination vector pDEST32a by introducing an RfA cassette at the EcoRV site, according to the Gateway Conversion System (Invitrogen). The pDEST32a-OsIRO2 construct was generated by an LR recombination reaction between the destination vector pDEST32a and the entry vector pENTR-OsIRO2, to produce the thioredoxin– OsIRO2 fusion protein. The expression vector pDEST32a-OsIRO2 was introduced into Escherichia coli strain BL21(DE3), and induction of the recombinant protein was performed according to the pET System Manual (Novagen). After 3 h of induction with isopropyl-b-D-thiogalactopyranoside, a crude cellular extract was prepared by sonication using the binding buffer for the CASTing procedure as the lysis buffer. The thioredoxin–OsIRO2 fusion protein in the crude extract was purified using anti-thioredoxin agarose (Sigma) and immunologically fixed onto agarose resin for use in the binding reaction of CASTing. CASTing Oligonucleotides of 75 nt were chemically synthesized that contained 25 nt of a random sequence and an SalI site flanked by the T3 and T7 primer sequences, i.e. 59-AATTAACCCTCACTAAAGGGACGTCGAC(N25)GCCCTATAGTGAGTCGTATTAC-39, where the underlined sequences represent the T3 and T7 sequences, respectively. Preparation of the double-stranded oligomer for CASTing was performed by annealing a synthesized oligonucleotide with the T7 primer and filling in with the Klenow fragment of DNA polymerase (TOYOBO). The binding buffer consisted of 20 mM TRIS– HCl (pH 8.0), 50 mM KCl, 0.5 mM EDTA, 10% (v/v) glycerol, 20 lg ml1 bovine serum albumin, 1 mM 1,4-dithiothreitol, and 50 lg ml1 salmon sperm DNA. Binding reactions were performed by mixing thioredoxin–OsIRO2 on agarose resin with 10 lg of the
double-stranded oligomer in binding buffer and incubating for 1 h on ice. The DNA–protein complexes were precipitated by centrifugation, and the pellets were washed twice with binding buffer. The pellets were suspended in 50 ll of 13 KOD+ polymerase chain reaction (PCR) buffer (TOYOBO) and boiled for 3 min, to unbind the DNA from the proteins and permit elution. For the PCR, five tubes with the same reaction solution that contained 10 ll of the eluted DNA, T3 and T7 primers, and KOD+ DNA polymerase were prepared in a volume of 50 ll. PCR amplification per cycle was performed for 15 s at 94 8C, 15 s at 55 8C, and 30 s at 68 8C. The PCR was stopped by removing a tube from the machine at the 10th, 15th, 20th, 25th, and 30th cycle and placing on ice. To determine at which cycle the DNA began to be amplified, a 10 ll aliquot of the reaction in each tube was separated on a 2% agarose gel, and the gel was stained with ethidium bromide. A 5 ll aliquot of the reacted solution at the earliest cycle at which DNA was detected was used in the subsequent binding reaction. This series of procedures was repeated seven times. In the last PCR, the DNA was amplified using the B1T3 primer, which contained the aatB1 sequence for the Gateway system (Invitrogen) flanked by the T3 sequence (59-GGGGACAAGTTTGTACAAAAAAGCAGGCTAATTAACCCTCACTAAAGGGAC-39), and the B2T7 primer, which contained the aatB2 sequence flanked by the T7 sequence (59-GGGGACCACTTTGTAGAAAGCTGGGTGTAATACGACTCACTATAGGGC-39). Using a BP recombination reaction, the amplified DNA was introduced into the destination vector pDONR221 (Invitrogen), and the generated clones were sequenced. Preparation of maltose-binding protein–OsIRO2 fusion protein (MBP–OsIRO2) for EMSA The ORF of OsIRO2 was amplified using the primers 59- CTCGAGGAATTCATGGAGCAGCTGTTCGTCGA-39 and 59- TCTAGATTAAAGCTTTGCTTTGTTCCTGACG-39. The amplified fragment was cloned into pCR-BluntII-TOPO (Invitrogen) and sequenced. This ORF was subcloned into the EcoRI and XbaI sites of the pMALc2 vector (New England BioLabs), which was then transformed into E. coli strain XL1-Blue. The induction and purification of the recombinant protein were carried out using amylose resin (New England BioLabs) according to the protocol for the Protein Fusion & Purification System (New England BioLabs). EMSA The 21–27 bp oligonucleotides shown in Fig. 4B with the GG dinucleotide at the 59 end were synthesized and used as the forward oligonucleotides. The reverse oligonucleotides consisted of the sequences complementary to the forward oligonucleotides, with the TT dinucleotide at the 59 end and without CC at the 39 end. The double-stranded DNA probes were produced by boiling and slowly cooling the forward and reverse oligonucleotides in 50 mM NaCl, followed by filling in the gaps with [a-32P]dATP using the Klenow fragment. CAS contained a potential OsIRO2-binding site, as identified by the CASTing method. The mutations were introduced by converting A to C, C to A, T to G, and G to T (m1–m14). The NAS1, FDH, and MTN sequences represent the 1601/1622 region of the OsNAS1 promoter, the 1389/1410 region of the OsFDH promoter, and the 1659/1680 region of the MTN promoter, respectively. IDE2 represents the 27-base Fe-deficiencyresponsive element 2 (Kobayashi et al., 2003). The binding reaction was carried out in 20 ll of binding buffer that contained 15 mM HEPES (pH 7.5), 17.5 mM KCl, 0.4 mM EDTA, 6% glycerol, 0.05% IGEPAL CA-630 (Sigma), 500 ng of poly(dI–dC)2 (Amersham), 0.5–2.0 ng of 32P-end-labelled probe and 225–500 ng of MBP–OsIRO2. The binding reaction mixtures were incubated for 20–30 min at room temperature and then electrophoresed at room temperature at a constant 120 V in a 6% native polyacrylamide gel with 22.5 mM TRIS, 22.5 mM boric acid, and 0.5 mM EDTA.
Downloaded from http://jxb.oxfordjournals.org/ at Pennsylvania State University on February 21, 2013
Isolation of HvIRO2 A full-length cDNA library of barley was constructed using mRNA prepared from Fe-deficient roots and cloned into pCMVFL3 (TOYOBO). Approximately 250 000 colonies were screened with the same probe and under the same hybridization conditions as those used for the northern blot analysis of OsIRO2. Thirty positive clones were isolated and divided into two groups according to insert length. After sequencing using the DSQ-2000L DNA sequencer (Shimadzu), both groups of clones were found to encode the same protein. The longest clone, which contained a 1.4 kb insert, was used for northern blot analysis of HvIRO2.
2869
2870 Ogo et al. Searching for OsIRO2-binding sequences in plant promoters The upstream sequences of various rice genes were derived from the rice genome database based on the positions of the putative initiation codon of the longest ORF in the full-length cDNA clones (KOME at http://cdna01.dna.affrc.go.jp/cDNA/). With respect to Fe-deficiencyinducible genes, 16 genes for which Fe deficiency inducibility had been confirmed by northern analysis, as well as 42 genes that showed up-regulated expression from day 3 to day 7 in the microrarray analysis were chosen for this analysis. One hundred genes on the array were selected randomly from the Fe-deficiency-uninducible genes. The HvNAS1 (AB022688), HvNAAT-A (AB024006), HvNAAT-B (AB024006), IDS2 (D15051), and IDS3 (AB024007) genes from barley in addition to Fe-deficiency-inducible rice genes were used.
Results Downloaded from http://jxb.oxfordjournals.org/ at Pennsylvania State University on February 21, 2013
Microarray analysis of genes related to gene regulation and induced during Fe deficiency
To investigate changes in gene expression related to Fe deficiency in rice, a microarray analysis was conducted. The oligo-DNA microarray used included 21 938 rice genes, which were represented by their unique 60mer oligonucleotides. These genes were characterized by their homologies to other plant proteins and from predictions of their subcellular localizations (KOME at http://cdna01. dna.affrc.go.jp/cDNA/). Approximately 2000 of the genes in this array were predicted to encode DNA-binding proteins, RNA-binding proteins, and/or nuclear-targeted proteins, which may be components of gene regulatory mechanisms. Among them, several genes were found to be up-regulated under Fe deficiency conditions and considered to be involved in gene regulation under Fe-deficient conditions (Fig. 1; Table 1). Induction of some genes by Fe deficiency was confirmed by northern blot (data not shown). Twenty selected genes showed three expression patterns in the shoot. The nine genes in Fig. 1A were maintained at a high level of expression during Fe deficiency. The expression of the six genes in Fig. 1B tended to be increased with progressive Fe deficiency. The expression of the five genes in Fig. 1C was increased at the initiation of Fe deficiency, but decreased subsequently. The increases in gene expression in Fig. 1A and B were dramatically repressed by re-supplying Fe (Fig. 1; on day 7+4). Some of the genes, such as AK101209 and AK102789 (Fig. 1A), AK073385 and AK103636 (Fig. 1B), and AK059581 (Fig. 1C), were up-regulated in both shoots and roots, while other genes were induced exclusively in the shoots. AK065631 was induced only in the roots (Fig. 1D). The expression of AK065631 in the roots was increased on day 1, decreased to the control level on day 2, retained this low level throughout the period of Fe deficiency, and increased once again when Fe was re-supplied. The AK073385 gene (OsIRO2) is specifically up-regulated under Fe-deficient conditions
Among the Fe deficiency-inducible genes that are putatively involved in gene regulation, AK073385, which
Fig. 1. Expression patterns of genes that are putatively involved in gene regulation. The vertical axis shows the means of the ratio of two replicates by microarray analysis. The numbers of days after the initiation of Fe deficiency are shown on the horizontal axis. (A) Genes that maintain induced expression in the shoots during the period of Fe deficiency. (B) Genes that show increased expression levels in the shoots with the progression of Fe deficiency. (C) Genes that show increased expression levels in the shoots only at the beginning of Fe deficiency. (D) Fedeficiency-inducible genes that are expressed exclusively in the roots.
IRO2, a novel iron-regulated transcription factor
2871
Table 1. Genes that are putatively involved in gene regulation under conditions of Fe deficiency The Fe-deficiency-inducible genes, categorized as shown in Fig. 1, are listed. Organs in which the expression was up-regulated are shown as R (root) or S (shoot). Accession no.
Organ
Myb transcription factor bHLH transcription factor DRE-binding protein like Nuclear localizing protein bHLH transcription factor Nuclear localizing protein RING finger protein Nuclear localizing protein Nuclear localizing protein
S, R S S S S S S, R S S
bHLH transcription factor (IRO2) Nuclear localizing protein OsNAC1 Zinc finger protein RING finger protein NAC-domain protein
S, R S, R S S S S
WRKY transcription factor OsNAC5 Myb transcription factor Nuclear localizing protein RNA-binding protein (RGGA)
S S S S S, R
C3HC4-type RING finger protein
R
encodes a bHLH transcription factor, was strongly induced (Fig. 1B), with a peak expression ratio in shoots of 18.4 on day 5, which was the second highest ratio on day 5 of all the genes in the microarray; the highest AK073385 expression ratio in roots was 4.5 on day 3, which was the sixth highest on day 3 of all the genes examined. This bHLH protein was named rice iron-related transcription factor 2 (OsIRO2), and it was analysed in detail. To confirm the induction of OsIRO2 in response to Fe deficiency, northern blot analysis was performed with different samples from those subjected to microarray analysis. OsIRO2 was not expressed in shoots, but was expressed constitutively at low levels in the roots of Fesufficient rice (Fig. 2A). The expression of OsIRO2 was strongly induced by Fe deficiency in both shoots and roots, and this increased expression was maintained throughout the period of Fe deficiency. In contrast, the expression of OsIRO2 was not induced by zinc, manganese, or copper deficiency in either roots or shoots (Fig. 2B). Interestingly, the expression of OsIRO2 in roots was reduced by zinc deficiency. Homologues of OsIRO2 are present exclusively in graminaceous plants
A homologue of OsIRO2, HvIRO2, was isolated from barley (Fig. 3A). The deduced amino acid sequence of HvIRO2 was 64% identical to that of OsIRO2. An in silico
Fig. 2. Northern blot analysis of OsIRO2. Each lane contains 20 lg of total RNA extracted from the roots or shoots of rice plants. A fragment from the 1051 bp region of the EST cDNA clone of OsIRO2 was used as the hybridization probe. (A) Time-course analysis of Fe deficiency in plants. Plants were harvested at 1, 3, 5, 7, 9, and 11 d after the initiation of Fe deficiency (Fe), and Fe-sufficient plants (+Fe) were harvested 0 and 7 d after treatment. (B) Expression patterns under conditions of metal deficiency. Control plants (C) were supplied with all the necessary nutrients.
search revealed that other graminaceous plants also carry bHLH genes that are highly homologous to OsIRO2. ESTs from wheat, sorghum, and maize appear to contain partial sequences of the OsIRO2 ORF (Fig. 3A). The deduced proteins coded by these bHLH genes were found to be highly homologous to each other, even outside the bHLH domain. In contrast, proteins with high levels of homology to OsIRO2 were not found in dicotyledonous plants, such as Arabidopsis, Brassica napus, tobacco, tomato, and soybean (data not shown). The expression of HvIRO2 was strongly induced by Fe deficiency in both shoots and roots and was dramatically reduced by the re-supply of Fe (Fig. 3B). The onset of induction of HvIRO2 in roots by Fe deficiency was earlier than that of OsIRO2 (Figs 2A, 3B). The expression of HvIRO2 increased more rapidly in roots than in shoots. A sequence comparison of the plant bHLH proteins revealed that IRO2 belongs to a different group from tomato FER and the Arabidopsis FIT1/FRU/AtbHLH29 proteins (Fig. 3C). The presence of three amino acid residues, histidine, glutamate, and arginine, in the basic region of the bHLH domain of IRO2 (Fig. 3A) suggests that IRO2 can bind to the G-box (CACGTG), whereas FER and FIT1/FRU/AtbHLH29 can bind to the E-box (CANNTG; non-G-box) (Toledo-Ortiz et al., 2003). Preferred recognition sequence for OsIRO2
The recognition sequence for OsIRO2 was identified by the CASTing method. Potential OsIRO2-binding sites were
Downloaded from http://jxb.oxfordjournals.org/ at Pennsylvania State University on February 21, 2013
Group A AK101209 AK059273 AK106041 AK099921 AK068704 AK101217 AK102789 AK072238 AK106365 Group B AK073385 AK103636 AK108080 AK068028 AK108361 AK063703 Group C AK066255 AK063399 AK112056 AK109364 AK059581 Group D AK065631
Putative identification
2872 Ogo et al.
enriched by seven cycles of DNA binding, isolation of nucleoprotein complexes, and re-amplification of selected oligonucleotides. Of the amplified and cloned sequences, 50 were sequenced and revealed to be independent of each other. Eighty-eight per cent of the sequenced clones contained the G-box (59-CACGTG-39) sequence. The consensus sequence was obtained by arranging the sequences of clones that contained the G-box (Fig. 4A). Interestingly, all the clones containing a G-box possessed the G residue next to the G-box at the 39 end. OsIRO2 was found to bind preferentially to 59-ACCACGTGGTTTT-39, which is an incomplete palindromic sequence that flanks the core G-box. In order to confirm the DNA-binding site of OsIRO2, one of the oligonucleotides that contained the potential OsIRO2-binding site (obtained by the CASTing method), with the sequence 59-GTTAACCACGTGGTTTTCAGC39 (the G-box is underlined), was named the CAS oligonucleotide and used in EMSA to detect the specific DNA-binding activity of OsIRO2. The end-labelled CAS oligonucleotide was used as DNA probe. Two shifted
bands that corresponded to the DNA–protein complex were observed when MBP–OsIRO2 was added, whereas these bands were not observed when MBP alone was added (Fig. 4C). The lower and upper shifted bands may correspond to the DNA–OsIRO2 monomer and DNA–OsIRO2 dimer, respectively. When IDE2, one of the Fe-deficiencyresponsive elements and not similar to CAS (Kobayashi et al., 2003), was used as a probe, no shifted band was observed (Fig. 4C). The oligonucleotides m3, m4, and m5, with mutations in the G-box, did not compete for CAS–OsIRO2 complex formation, whereas m1, m2, m6, m7, and m8 competed (Fig. 4B, D). This result indicates that the G-box is necessary for OsIRO2-binding activity and that the sequences flanking the G-box affect binding to some extent. The oligonucleotides m9, m10, and m11, with single mutations of the G-box, did not compete for the complex formation. Oligonucleotide m12, with a single mutation of the G residue next to the G-box at the 39 end, competed for complex formation to a low degree (Fig. 4D). Interestingly,
Downloaded from http://jxb.oxfordjournals.org/ at Pennsylvania State University on February 21, 2013
Fig. 3. IRO2 from other graminaceous plants. (A) Alignment of IRO2 sequences. The y symbol indicates the position of the putative bHLH domain. The filled inverted triangle indicates the consensus three amino acid residues of bHLH protein that bind to the G-box (Toledo-Qriz et al., 2003). The shaded areas represent either identical residues found in most of the proteins (black) or similar character residues (grey). The putative IRO2 sequences from sorghum, wheat, and maize come from EST clones (sorghum, CN126528; wheat-a, CD872621; wheat-b, CA502657; maize, BM347640). (B) Northern blot analysis of HvIRO2 in Fe-deficient barley. Plants were harvested 1, 3, 5, 7, and 9 d after the initiation of Fe deficiency (Fe), or 2 d after Fe was re-supplied on day 9 of Fe deficiency (9+2 d). The control plants (+Fe) were harvested 1 d after treatment. (C) Phylogenic tree for IRO2 proteins and other bHLHs in plants. The underlined proteins are predicted to recognize the G-box, while the other proteins are thought to bind to the E-box (non-Gbox), based on the predictions of Toledo-Qriz et al. (2003). The proteins studied have the following accession numbers: ORG2, AF488576; ORG3, AF488577; MYC146, AF027732; OSB1, BAB64301; RAP-1, X99548; PIF3, AF251693; FER, AAN39037; FIT1/FRU/AtbHLH29, AT2G28160; ICE1, AY079016.
IRO2, a novel iron-regulated transcription factor
2873
involved in photosynthesis (AK067730) and various kinds of metabolism (AK067591, AK071404, and AK072595), contained OsIRO2-binding core sequences in their promoter regions. Several unknown genes that were induced by Fe deficiency also possessed the core sequences. The three 21 bp native promoters of OsNAS1, OsFDH, and MTN (Fig. 4B, E) were end-labelled and used as DNA probes. All the native promoters bound OsIRO2, although the shifted bands were weaker than that observed with the CAS oligonucleotide (Fig. 4C, D, E). Oligonucleotides m4 and m5 did not compete for the complex formation, whereas the native sequences used as end-labelled probes, as well as m14 and CAS, did compete for the complex formation.
Potential target genes for OsIRO2 In the search for putative target genes regulated by IRO2 proteins, the upstream regions (;2000 bp long) of the 16 known Fe-deficiency-inducible genes and the 42 genes that showed up-regulated expression from day 3 to day 7 in the microarray analysis were used. Initially, these promoter regions were searched for the presence of Gboxes (59-CACGTG-39) (Fig. 5). G-boxes were found in 43% of the Fe-deficiency-inducible genes and in 31% of the randomly selected uninducible genes. Chi-squared analysis showed that G-box frequency in Fe-deficiencyinducible promoters was not significantly higher than in randomly selected uninducible promoters. In contrast, the OsIRO2-binding core sequence (G-box plus G; 59CACGTGG-39) was found in 31% of the Fe-deficiencyinducible genes and in 17% of the randomly selected genes. The frequency of 59-CACGTGG-39 was significantly higher in Fe-deficiency-inducible gene promoters than in randomly selected uninducible gene promoters. Fe-deficiency-inducible genes that carried an OsIRO2binding core sequence are listed in Table 2. Several genes related to MA synthesis and Fe acquisition had the OsIRO2-binding core sequence. The protein products of OsNAS1, OsNAS3 (Higuchi et al., 2001), and barley IDS3 (Kobayashi et al, 2001) participate in synthesis of MAs. Formate dehydrogenase (OsFDH), methylthioadenosine/Sadenosyl-homocysteine nucleosidase (MTN), and adenine phosphoribosyltransferase 1 (OsAPT1) participate in the methionine cycle, through which methionine is efficiently supplied for synthesis of MAs (Suzuki et al., 1998; Itai et al., 2000; Kobayashi et al., 2005). The Fe(II) transporter that is encoded by OsIRT1 functions in the absorbance of Fe(II) ions, which are abundant in the soils of paddy fields (Bughio et al., 2002; Ishimaru et al., 2006). The OsIRO2binding core sequences were also found in AK063399 (OsNAC5), which encodes a transcription factor (Fig. 1). Furthermore, AK103890, the function of which is related to protein degradation through ubiquitination, and genes
Discussion This study shows the gene regulators that would participate in the Fe acquisition mechanism in graminaceous plants (Fig. 1). The genes for three NAC-domain proteins (AK108080 for OsNAC1, AK063703 for NAC domain protein, and AK063399 for OsNAC5) were up-regulated under Fe-deficient conditions in shoots. NAC-domain transcription factors are unique to plants and comprise a large gene family (Ren et al., 2000; Xie et al., 2002; Duval et al., 2002; Hegedus et al., 2003). Several studies have reported that the expression of genes for NAC-domain proteins is induced by wounding, pathogen infection, and/ or drought (Collinge and Boller, 2001; Hegedus et al., 2003; Tran et al., 2004; Ohnishi et al., 2005). Our study is the first to show that genes that encode NAC-domain proteins are up-regulated by metal deficiency. There are some reports of Fe-deficiency-inducible transcription factor genes revealed by microarray analysis. Previously, it had been found that two zinc finger genes and one leucine zipper protein gene were up-regulated by Fe deficiency in barley roots (Negishi et al., 2002). In Arabidopsis, in addition to FIT1, three RING finger protein genes and a putative DNAbinding protein gene are up-regulated by Fe deficiency (Colangelo and Guerinot, 2004). The genes shown in Fig. 1 are reported here for the first time to be Fe deficiency inducible. This wide variety of transcription factors appears to participate in the complex regulatory networks that function under Fe-deficiency stress. Among the Fe-deficiency-inducible genes that are putatively involved in gene regulation, the AK073385 gene, the expression of which was highly Fe deficiency inducible, encodes a novel bHLH protein, which has been named OsIRO2. The induction pattern of OsIRO2 under conditions of Fe deficiency resembled those of other genes induced by Fe deficiency (Inoue et al., 2003; Kobayashi et al., 2005; Ishimaru et al., 2006). The induction of OsIRO2 expression did not occur under conditions of zinc, manganese, or copper deficiency (Fig. 2B). These results suggest
Downloaded from http://jxb.oxfordjournals.org/ at Pennsylvania State University on February 21, 2013
oligonucleotide m13, in which all of the nucleotides outside the G-box were mutated, did not compete for DNA– OsIRO2 complex formation. On the other hand, m14, in which the nucleotides outside the G-box plus the G next to the G-box (59-CACGTGG-39) were mutated, competed for complex formation (Fig. 4D). These results indicate that the G-box is not sufficient for OsIRO2 binding, and the G next to the G-box is also required. Substantial competition by m6 and m12, which are mutated at the G at the seventh position of 59-CACGTGG-39, yet containing 59CCACGTG-39 in their complementary strands, indicates that either ‘G’ at the 39 flank or ‘C’ at the 59 flank of G-box, but not both, is sufficient for OsIRO2 binding. Consequently, the ‘G-box plus G’ sequence (59-CACGTGG-39) was determined to be the OsIRO2-binding core sequence.
2874 Ogo et al.
Downloaded from http://jxb.oxfordjournals.org/ at Pennsylvania State University on February 21, 2013
Fig. 4. OsIRO2-binding sequence. (A) Identification of the consensus nucleotide sequence that acts as the binding site for OsIRO2 using the CASTing method. The binding sequences of clones that contain the G-box are arranged by centring the G-box. The frequencies at which individual nucleotides appear are shown in percentages on the vertical axis. The letter F indicates sequences that could not be confirmed. (B–E) EMSA using MBP–OsIRO2 and the oligonucleotide that contains the OsIRO2-binding site, obtained by CASTing. (B) Nucleotide sequences used for the probes and competitors. One of the oligonucleotides (CAS) that contained a potential OsIRO2-binding site, as identified by the CASTing method, was used as a probe. The putative OsIRO2-binding site determined by CASTing is underlined. The G-box is shaded in grey and the mutated sequences for the competition assays are shaded in black. The mutations were introduced by converting A to C, C to A, T to G, and G to T. The NAS1, FDH, and MTN sequences represent the 1601/1622 region of the OsNAS1 promoter, the 1389/1410 region of the OsFDH promoter, and the 1659/1680 region of the MTN promoter, respectively. IDE2 represents the 27 base Fe-deficiency-responsive element 2 (Kobayashi et al., 2003). (C) The end-labelled CAS or IDE2 was incubated with MBP–OsIRO2 or MBP. Arrowheads indicate two shifted bands and the free probe. (D) The effect of mutations in the binding site on the formation
IRO2, a novel iron-regulated transcription factor
Fig. 5. Frequencies of the OsIRO2-binding core sequence. The frequencies of identified G-box (CACGTG) and OsIRO2-binding core sequence (CACGTGG) in the promoter regions (within the 2 kb upstream regions) of Fe-deficiency-inducible genes (black bars) and randomly selected Fe-deficiency-uninducible genes (white bars).
abolished OsIRO2 binding (Fig. 4D). In addition, the Gbox alone is not sufficient for OsIRO2 binding, as evidenced by a strict requirement for a G next to the Gbox for successful binding (Fig. 4D). The residue next to the G-box at the 39 end must be G, but not T, A, or C, because all the clones containing the G-box obtained by CASTing possessed the G residue next to the G-box at the 39 end (Fig. 4A). Therefore, the ‘G-box plus G’ sequence of 59-CACGTGG-39 is considered to be the OsIRO2-binding core sequence. Furthermore, the sequences flanking the OsIRO2-binding core sequences affect OsIRO2 binding activity to some extent, since probes mutated in nucleotides outside the core sequence competed, to a low degree, for the complex formation (Fig. 4D). Although not many transcription factors recognizing a long DNA sequence such as 8 or 9 bp have been reported, a further analysis of the binding sequence of OsIRO2 might be needed. The determined OsIRO2-binding sequence is not included in the previously identified Fe-deficiency-responsive cis-elements
of the OsIRO2–DNA complexes, as determined by competition assays. EMSA was performed using 0.5 ng of end-labelled CAS oligonucleotide and 225 ng of MBP–OsIRO2 protein. A 50-, 200-, or 500-fold excess of CAS, or of each mutated oligonucleotide, was added to the reaction mixture as unlabelled competitor. The shifted bands are indicated. (E) Binding of OsIRO2 to the native promoters of OsNAS1, OsFDH, and MTN, and the competition assay. The end-labelled native promoters (2 ng) were incubated with 500 ng of MBP–OsIRO2. A 50- or 100-fold excess of each native promoter and mutated oligonucleotide was added to the reaction mixture as unlabelled competitors.
Downloaded from http://jxb.oxfordjournals.org/ at Pennsylvania State University on February 21, 2013
that OsIRO2 plays a specific role in plant responses to Fe deficiency. We isolated an OsIRO2 homologue from barley, encoded by HvIRO2, the expression of which in Fe-deficient barley was almost the same as that in rice and lacked root specificity (Figs 2A, 3B). An in silico search for proteins homologous to IRO2 revealed that IRO2 is widely conserved among graminaceous plants (Fig. 3A). In contrast, it was difficult to find bHLH proteins that were highly similar to IRO2 outside of the bHLH domain of dicot plants (data not shown). IRO2 appears to function in a gene regulatory system that is particular to graminaceous plants that grow under Fe-deficient conditions. However, this may not relate to root-specific gene regulatory systems, such as the regulation of HvNASs and NASHOR1 expression in Fe-deficient barley (Higuchi et al., 1999; Douchkov et al., 2002). In dicotyledons, the bHLH transcription activators FER and FIT1/FRU/bHLH29 are implicated in Fe nutrition (Ling et al., 2002; Colangelo and Guerinot, 2004; Jakoby et al., 2004; Yuan et al., 2005). The expression of FIT1/ FRU/bHLH29 is detectable in Fe-sufficient roots and is increased by Fe deficiency, while the expression of FER is not up-regulated by Fe deficiency. FER and FIT1/FRU/ bHLH29 are predicted to bind to the E-box (59-CANNTG-39; non-G-box; Fig. 3C). A FER-like protein was also found in the rice genome database (XP_472244; ‘OsFER-like’ in Fig. 3C). However, this putative bHLH gene may not be expressed in rice, as the respective EST clone has not been found to date. In contrast, this study shows that OsIRO2 binds preferentially to the G-box (Fig. 4). Furthermore, both OsIRO2 and HvIRO2 were strongly up-regulated by Fe deficiency in both roots and shoots (Figs 2A, 3B), while FER and FIT1/FRU/bHLH29 were expressed only in roots (Ling et al., 2002; Colangelo and Guerinot, 2004; Jakoby et al., 2004). These results suggest that IRO2 in graminaceous plants and FER and FIT1/FRU/bHLH29 in dicotyledonous plants probably have different functions. The bHLH family of proteins is a group of functionally diverse transcription factors that are found in both plants and animals. The bHLH proteins bind as either homo- or heterodimers to specific DNA target sites. Our EMSA results revealed the possibility that OsIRO2 binds to DNA as a homodimer, as two shifted bands were detected (Fig. 4). The preferred DNA-binding sequence of OsIRO2, 59ACCACGTGGTTTT-39, was determined experimentally (Fig. 4A). EMSA revealed that the G-box in the OsIRO2binding site determined by CASTing is necessary for the binding of OsIRO2 to DNA; a mutation in even one base
2875
2876 Ogo et al. Table 2. Promoter analysis of the OsIRO2-binding sequence Listed are Fe- deficiency-inducible genes whose 2 kb upstream regions contain the OsIRO2-binding core sequence. The positions are numbered from the putative translation start site. The –Fe/day 0 ratios reflect the results from day 5 of the microarray analysis. The following four genes are not included in the array, although their expression levels under conditions of Fe deficiency have been reported to be up-regulated ("), down-regulated (#), or unchanged (–): OsNAS3 (Inoue et al., 2003); barley IDS3 (Nakanishi et al., 1993); MTN (Kobayashi et al., 2005); and OsIRT1 (Bughio et al., 2002). Accession no.
Gene name or putative function
Sequence
Position
ACCACGTGGTTTT MAs and Fe acquisition-related AB046401 Nicotianamine synthase 1 (OsNAS1) AB023819 Nicotianamine synthase 3 (OsNAS3)
AK107681
Iron-regulated metal transporter 1 (OsIRT1)
Expressional regulation-related AK063399 OsNAC5 Others AK067730 Chlorophyll b synthase AK067591 Cytochrome P450 AK071404 Lipase AK072595 AK103890 Unknown AK063263 AK068661 AK100816 AK107748 AK108093 AK110584
Tryptophan synthase Ubiquitin extension protein 1 (UBQ1) Unknown Unknown Unknown Unknown Unknown Unknown
IDE1 and IDE2, nor is it similar to these elements. The OsIRO2-binding sequence includes the high affinity sequence (59-ACCACGTGGT-39) of the c-Myc protein, which is implicated in animal cancers (Halazonetis and Kandil, 1991; Blackwell et al., 1993). Among the plant ciselements that have been reported, the sequence of 59GCCACGTGGG-39, which is located in the promoter region of the maize ABA-inducible rab28 gene (Pla et al., 1993), is very similar to the OsIRO2-binding sequence. Many of the cis-elements known as ABA-responsive elements (ABREs) share the 59-(C/T)ACGTGGC-39 motif, and this consensus sequence is sometimes called the G-boxlike ABRE (Choi et al., 2000). However, the G-box-like ABRE is generally bound by basic leucine zipper proteins, and two or more G-box-like ABREs, or the coupling elements, are needed for function (Shen and Ho, 1995; Shen et al., 1996; Hobo et al., 1999). The frequencies of the OsIRO2-binding sequence were estimated using the upstream sequences of 58 Fedeficiency-inducible genes and 100 Fe-deficiencyuninducible genes in rice (Fig. 5). The frequency of the OsIRO2-binding core sequence was significantly higher in Fe-deficient inducible genes, while the fre-
Shoot
Root
tgCACGTGGTggg gaCACGTGGgcca AtCACGTGGgTcT gCCACGTGGCaTg gaCACGTGGTTTa tCCACGTGGTggg
1610 152 365 1097 1395 1668
6.14 #
3.21 "
– 0.91 –
" 2.78 "
gaCACGTGGgccc gaCACGTGGgccc
242 133
2.99
2.40
caCACGTGGaacg
1234
–
"
cCCACGTGGcTga
1317
2.32
1.34
gCCACGTGGTgcg gaCACGTGGaaat gaCACGTGGaaaa tgCACGTGGcgTg gaCACGTGGgaga gaCACGTGGcacg
832 1271 371 309 357 293
2.43 3.15 3.42
5.43 2.75 2.60
4.42 4.58
1.30 3.41
AgCACGTGGgggg tgCACGTGGTact ctCACGTGGccgc gaCACGTGGcggc caCACGTGGTcag cgCACGTGGTTgT cCCACGTGGGagT
850 984 1950 140 641 1040 186
2.33 11.09 4.66 2.41 3.20 10.26
1.25 1.59 0.90 0.91 1.70 1.78
quency of the G-box alone was not significantly different for Fe-deficiency-inducible genes and Fe-deficiencyuninducible genes (Fig. 5). Many transcription factors with the ability to bind to the G-box have been reported (Siberil et al., 2001), and the G-box may occur frequently in the promoter regions of various genes, where it regulates the expression of genes not related to Fe nutrition. The significant difference observed for the frequency of the OsIRO2-binding core sequence suggests that this sequence is an important cis-element in the promoter regions of Fe-deficiency-inducible genes, and that OsIRO2 is one of the key transcription factors under conditions of Fe deficiency. It is noteworthy that many genes related to Fe acquisition have the OsIRO2-binding core sequence in their promoter regions (Table 2), which suggests that OsIRO2 plays a role in the transcriptional regulation of genes that participate in Fe acquisition under Fe-deficient conditions. Our EMSA results revealed that OsIRO2 binds to the native promoters of OsNAS1, OsFDH, and MTN with strict sequence specificity (Fig. 4E). Since oligonucleotides m4 and m5 did not compete for the complex formation, the G-box of the native promoter sequences was bound by OsIRO2. The strong competition shown by
Downloaded from http://jxb.oxfordjournals.org/ at Pennsylvania State University on February 21, 2013
AK073627
Barley mugineic acid synthase IDS3 Formate dehydrogenase (OsFDH) Methylthioadenosine/S-adenosyl homocysteine nucleosidase (MTN) Adenine phosphoribosyltransferase 1 (OsAPT1)
AB024007 AK065872 AK066172
Ratio (–Fe/day 0)
IRO2, a novel iron-regulated transcription factor
Accession numbers The sequence of HvIRO2 has been deposited in DDBJ, EMBL, and GenBank with the accession number AB206536.
Acknowledgements We thank Dr Yoshiaki Nagamura and his colleagues at the National Institute of Agrobiological Sciences (NIAS) for assistance with the microarray experiments.
References Blackwell TK, Huang J, Ma A, Kretzner L, Alt FW, Eisenman RN, Weintraub H. 1993. Binding of myc proteins to canonical and noncanonical DNA sequences. Molecular and Cellular Biology 13, 5216–5224. Bughio N, Yamaguchi H, Nishizawa NK, Nakanishi H, Mori S. 2002. Cloning an iron-regulated metal transporter from rice. Journal of Experimental Botany 53, 1677–1682. Chinnusamy V, Ohata M, Kanrar S, Lee BH, Hong X, Agarwal M, Zhu JK. 2003. The basic helix–loop–helix trans-
cription factor family in plants: a genome-wide study of protein structure and functional diversity. Molecular Biology and Evolution 20, 735–747. Choi H, Hong J, Ha J, Kang J, Kim SY. 2000. ABFs, a family of ABA-responsive element binding factors. Journal of Biological Chemistry 275, 1723–1730. Church GM, Gilbert W. 1984. Genomic sequencing. Proceedings of the National Academy of Sciences, USA 81, 1991–1995. Colangelo EP, Guerinot ML. 2004. The essential basic helix–loop– helix protein FIT1 is required for the iron deficiency response. The Plant Cell 16, 3400–3412. Collinge M, Boller T. 2001. Differential induction of two potato genes, Stprx2 and StNAC, in response to infection by Phytophthora infestans and to wounding. Plant Molecular Biology 46, 521–529. Curie C, Panaviene Z, Loulergue C, Dellaporta SL, Briat JF, Walker EL. 2001. Maize yellow stripe1 encodes a membrane protein directly involved in Fe(III) uptake. Nature 409, 346–349. Douchkov D, Herbik A, Koch G, Mock HP, Melzer M, Stephan UW, Baumlein H. 2002. Nicotianamine synthase: gene isolation, gene transfer and application for the manipulation of plant iron assimilation. Plant and Soil 241, 115–119. Duval M, Hsieh TF, Kim SY, Thomas TL. 2002. Molecular characterization of AtNAM: a member of the Arabidopsis NAC domain superfamily. Plant Molecular Biology 50, 237–248. Gilmour SJ, Zarka DG, Stockinger EJ, Salazar MP, Houghton JM, Thomashow MF. 1998. Low temperature regulation of the Arabidopsis CBF family of AP2 transcriptional activators as an early step in cold-induced COR gene expression. The Plant Journal 16, 433–442. Halazonetis TD, Kandil AN. 1991. Determination of the c-MYC DNA-binding site. Proceedings of the National Academy of Sciences, USA 88, 6162–6166. Hegedus D, Yu M, Baldwin D, Gruber M, Sharpe A, Parkin I, Whitwill S, Lydiate D. 2003. Molecular characterization of Brassica napus NAC domain transcriptional activators induced in response to biotic and abiotic stress. Plant Molecular Biology 53, 383–397. Higuchi K, Suzuki K, Nakanishi H, Yamaguchi H, Nishizawa NK, Mori S. 1999. Cloning of nicotianamine synthase genes, novel genes involved in the biosynthesis of phytosiderophores. Plant Physiology 119, 471–479. Higuchi K, Watanabe S, Takahashi M, Kawasaki S, Nakanishi H, Nishizawa NK, Mori S. 2001. Nicotianamine synthase gene expression differs in barley and rice under Fe-deficient conditions. The Plant Journal 25, 159–167. Hobo T, Asada M, Kowyama Y, Hattori T. 1999. ACGTcontaining abscisic acid response element (ABRE) and coupling element 3 (CE3) are functionally equivalent. The Plant Journal 19, 679–689. Inoue H, Higuchi K, Takahashi M, Nakanishi H, Mori S, Nishizawa NK. 2003. Three rice nicotianamine synthase genes, OsNAS1, OsNAS2, and OsNAS3 are expressed in cells involved in long-distance transport of iron and differentially regulated by iron. The Plant Journal 36, 366–381. Ishimaru Y, Suzuki M, Tsukamoto T, et al. 2006. Rice plants take up iron as an Fe3+ phytosiderophore and as Fe2+. The Plant Journal 45, 335–346. Itai R, Suzuki K, Nakanishi H, Yamaguchi H, Nishizawa NK, Mori S. 2000. Induced activity of adenine phosphoribosyltransferase (APRT) in iron-deficient barley roots: a possible role for phytosiderophore production. Journal of Experimental Botany 51, 1179–1188. Jakoby M, Wang HY, Reidt W, Weisshaar B, Bauer P. 2004. FRU (BHLH029) is required for induction of iron mobilization genes in Arabidopsis thaliana. FEBS Letters 577, 528–534.
Downloaded from http://jxb.oxfordjournals.org/ at Pennsylvania State University on February 21, 2013
m14 indicates that the OsIRO2-binding core sequence is essential for the binding activity in these native promoters. The core sequences were also found in several Fedeficiency-inducible genes that participate in transcription, photosynthesis, and metabolism. The roles of these genes under Fe-deficient conditions are currently unknown. The previously identified IDE1 and IDE2 cis-acting elements mediate gene expression that is both Fe-deficiencyinducible and root-specific (Kobayashi et al., 2003). Recently, many of the Fe-deficiency-inducible genes, the products of which participate in the production of MAs, were found to have sequences homologous to IDE1 and IDE2 in their promoter regions (Kobayashi et al., 2005). Of the genes shown in Table 2, OsNAS1, barley IDS3, OsFDH, MTN, OsAPT1, OsIRT1, and the AK068661 gene of unknown functionality all carry IDEs in their promoter regions (Kobayashi et al., 2003, 2005). The OsIRO2 gene promoter has an IDE2-like sequence at 1148/1174 (Kobayashi et al., 2005), and contains sequences that are moderately homologous to IDE1 (data not shown). Therefore, the following model is proposed for transcription regulation under Fe-deficient conditions in graminaceous plants. Under conditions of Fe deficiency, unknown IDEbinding proteins activate the expression of genes that possess IDEs in their promoter regions, such as IRO2. The induced IRO2 protein not only induces the expression of genes that lack IDEs in their promoter regions, but also further activates the expression of genes that are directly regulated by IDEbinding proteins, such as OsNAS1, barley IDS3, and OsFDH, through binding to the 59-CACGTGG-39 motifcontaining sequences in their upstream regions. IRO2 and IDE-binding proteins may function as part of a coordinated gene network under conditions of Fe deficiency.
2877
2878 Ogo et al. Pla M, Vilardell J, Guiltinan MJ, Marcotte WR, Niogret MF, Quatrano RS, Pages M. 1993. The cis-regulatory element CCACGTGG is involved in ABA and water-stress responses of the maize gene rab28. Plant Molecular Biology 21, 259–266. Ren T, Qu F, Morris TJ. 2000. HRT gene function requires interaction between a NAC protein and viral capsid protein to confer resistance to turnip crinkle virus. The Plant Cell 12, 1917–1926. Shen Q, Ho TH. 1995. Functional dissection of an abscisic acid (ABA)-inducible gene reveals two independent ABA-responsive complexes each containing a G-box and a novel cis-acting element. The Plant Cell 7, 295–307. Shen Q, Zhang P, Ho TH. 1996. Modular nature of abscisic acid (ABA) response complexes: composite promoter units that are necessary and sufficient for ABA induction of gene expression in barley. The Plant Cell 8, 1107–1119. Shojima S, Nishizawa N, Mori S. 1989. Establishment of a cell-free system for the biosynthesis of nicotianamine. Plant and Cell Physiology 30, 673–677. Siberil Y, Doireau P, Gantet P. 2001. Plant bZIP G-box binding factors. Modular structure and activation mechanisms. European Journal of Biochemistry 268, 5655–5666. Suzuki K, Itai R, Suzuki K, Nakanishi H, Nishizawa NK, Yoshimura E, Mori S. 1998. Formate dehydrogenase, an enzyme of anaerobic metabolism, is induced by iron deficiency in barley roots. Plant Physiology 116, 725–732. Takagi S. 1976. Naturally occurring iron-chelating compounds in oat- and rice-root washings. I. Activity measurement and preliminary characterization. Soil Science and Plant Nutrition 22, 423–433. Takahashi M, Yamaguchi H, Nakanishi H, Shioiri T, Nishizawa NK, Mori S. 1999. Cloning two genes for nicotianamine aminotransferase, a critical enzyme in iron acquisition (Strategy II) in graminaceous plants. Plant Physiology 121, 947–956. Toledo-Ortiz G, Huq E, Quail PH. 2003. The Arabidopsis basic/ helix–loop–helix transcription factor family. The Plant Cell 15, 1749–1770. Tran LS, Nakashima K, Sakuma Y, et al. 2004. Isolation and functional analysis of arabidopsis stress-inducible NAC transcription factors that bind to a drought-responsive cis-element in the early responsive to dehydration stress 1 promoter. The Plant Cell 16, 2481–2498. Xie Q, Guo HS, Dallman G, Fang S, Weissman AM, Chua NH. 2002. SINAT5 promotes ubiquitin-related degradation of NAC1 to attenuate auxin signals. Nature 419, 167–170. Yuan YX, Zhang J, Wang DW, Ling HQ. 2005. AtbHLH29 of Arabidopsis thaliana is a functional ortholog of tomato FER involved in controlling iron acquisition in strategy I plants. Cell Research 15, 613–621.
Downloaded from http://jxb.oxfordjournals.org/ at Pennsylvania State University on February 21, 2013
Kobayashi T, Nakayama Y, Itai RN, Nakanishi H, Yoshihara T, Mori S, Nishizawa NK. 2003. Identification of novel cis-acting elements, IDE1 and IDE2, of the barley IDS2 gene promoter conferring iron-deficiency-inducible, root-specific expression in heterogeneous tobacco plants. The Plant Journal 36, 780–793. Kobayashi T, Nakanishi H, Takahashi M, Kawasaki S, Nishizawa NK, Mori S. 2001. In vivo evidence that Ids3 from Hordeum vulgare encodes a dioxygenase that converts 29deoxymugineic acid to mugineic acid in transgenic rice. Planta 212, 864–871. Kobayashi T, Suzuki M, Inoue H, Itai RN, Takahashi M, Nakanishi H, Mori S, Nishizawa NK. 2005. Expression of iron-acquisition-related genes in iron-deficient rice is coordinately induced by partially conserved iron-deficiency-responsive elements. Journal of Experimental Botany 56, 1305–1316. Ling HQ, Bauer P, Bereczky Z, Keller B, Ganal M. 2002. The tomato fer gene encoding a bHLH protein controls iron-uptake responses in roots. Proceedings of the National Academy of Sciences, USA 99, 13938–13944. Liu Q, Kasuga M, Sakuma Y, Abe H, Miura S, YamaguchiShinozaki K, Shinozaki K. 1998. Two transcription factors, DREB1 and DREB2, with an EREBP/AP2 DNA binding domain separate two cellular signal transduction pathways in drought- and low-temperature-responsive gene expression, respectively, in Arabidopsis. The Plant Cell 10, 1391–1406. Marschner H. 1995. Mineral nutrition of higher plants, 2nd edn. London: Academic Press. Nakanishi H, Okukmura N, Umehara Y, Nishizawa NK, Chino M, Mori S. 1993. Expression of a gene specific for iron deficiency (Ids3) in the roots of Hordeum vulgare. Plant and Cell Physiology 34, 401–410. Nakanishi H, Yamaguchi H, Sasakuma T, Nishizawa NK, Mori S. 2000. Two dioxygenase genes; Ids3 and Ids2; from Hordeum vulgare are involved in the biosynthesis of mugineic acid family phytosiderophores. Plant Molecular Biology 44, 199–207. Negishi T, Nakanishi H, Yazaki J, et al. 2002. cDNA microarray analysis of gene expression during Fe-deficiency stress in barley suggests that polar transport of vesicles is implicated in phytosiderophore secretion in Fe-deficient barley roots. The Plant Journal 30, 83–94. Ohnishi T, Sugahara S, Yamada T, Kikuchi K, Yoshiba Y, Hirano HY, Tsutsumi N. 2005. OsNAC6, a member of the NAC gene family, is induced by various stresses in rice. Genes and Genetic Systems 80, 135–139. Okumura N, Nishizawa NK, Umehara Y, Ohta T, Nakanishi H, Yamaguchi H, Chino M, Mori S. 1994. A dioxygenase gene (Ids2) expressed under iron deficiency conditions in the roots of Hordeum vulgare. Plant Molecular Biology 25, 705–719.
