Here we report the function of a general regulatory factor, GENERAL REGULATORY FACTOR11 (GRF11), in terms of the iron (Fe) deficiency response.
Physiological and molecular responses of the loss-of-function Arabidopsis thaliana grf11 mutant to Fe supply were investigated. Genes involved in posttranscriptional regulation of FER-LIKE FE DEFICIENCY-INDUCED TRANSCRIPTION FACTOR (FIT) were also analyzed. In addition, the molecular link between the signaling molecule nitric oxide (NO) and Fe deficiency responses was further dissected.
Our results suggest that GRF11 is necessary for induction of Fe-deficiency-tolerance mechanisms. The FIT protein can bind to the promoter of GRF11, which contains an E-box motif. GRF11 also positively affects FIT transcription but has no influence on the genes involved in posttranscriptional regulation of FIT. Furthermore, NO positively regulates GRF11 induction upon the onset of Fe deficiency.
We propose that, upon the onset of Fe deficiency, induction of FIT expression is dependent on GRF11, which acts downstream of NO to mediate Fe deficiency responses.
Iron (Fe) is an essential micronutrient for all living organisms. Although the total Fe content in soil generally exceeds the plant's requirement, bioavailable Fe is only presented in limited amounts, especially when the soil is alkaline. Therefore, Fe deficiency is one of the major factors limiting plant growth and development in calcareous soils, which represent 30% of the world's arable land (Guerinot & Yi, 1994). However, plants have evolved sophisticated mechanisms to cope with Fe limitation. These mechanisms are grouped into Strategy I, found in nongraminaceous monocots and dicots, and Strategy II, found in graminaceous monocots (Römheld & Marschner, 1981). In response to Fe deficiency, Strategy I plants can stimulate the activity of ferric chelate reductase (FCR) (Chaney et al., 1972), activate the high-affinity ferrous transporter IRON-REGULATED TRANSPORTER1 (IRT1) (Eide et al., 1996), and excrete protons and reduced organic compounds to the rhizosphere (Römheld & Marschner, 1986; Jin et al., 2006, 2007; Santi & Schmidt, 2009). Furthermore, they can also regulate root development, for example inducing the formation of subapical root hairs (Jin et al., 2009) and increasing root branching (Jin et al., 2008, 2011), to enhance Fe uptake. Among these mechanisms, the FCR-catalyzed reduction of ferric chelates has been demonstrated to be the rate-limiting step for Fe uptake under Fe-limiting conditions in dicots (Connolly et al., 2003).
In Arabidopsis thaliana, reduction of ferric iron by FERRIC REDUCTION OXIDASE2 (FRO2) (Robinson et al., 1999) and uptake of ferrous iron by the metal transporter IRT1 (Eide et al., 1996; Vert et al., 2002) have been documented. The basic helix-loop-helix (bHLH) transcription factor FER-LIKE FE DEFICIENCY-INDUCED TRANSCRIPTION FACTOR (FIT), a functional homolog of FER in tomato (Lycopersicon esculentum; Brown & Embler, 1974) (Ling et al., 2002), is required for the expression of FRO2 and IRT1 (Yuan et al., 2005). Fe deficiency has been found to induce the accumulation of transcripts of FIT (Colangelo & Guerinot, 2004; Jakoby et al., 2004), suggesting that a trans-acting factor is interacting with the promoter of FIT to mediate its transcription. However, little information is available on the sensing of Fe deficiency signal and transcriptional regulation of FIT. In contrast, post-transcriptional regulation of FIT has attracted much attention. For example, heterodimerization of FIT with bHLH038 or bHLH039 can up-regulate Fe acquisition genes even under Fe-sufficient conditions (Yuan et al., 2008). Sivitz et al. (2011)proposed that proteasome-mediated turnover of ‘exhausted’ FIT is necessary for newly synthesized active FIT binding to promoters of its target genes. Meiser et al. (2011) also found that the stability and activity of FIT are regulated by Fe status and the signal molecule nitric oxide (NO) might be involved in this regulation.
Plant hormones are also involved in Fe-deficiency-induced responses (Romera et al., 2011); among these hormones, auxin (Chen et al., 2010), NO (Graziano & Lamattina, 2007), and ethylene (Romera & Alcántara, 2004; García et al., 2010). are accumulated in the roots of various dicot plants upon the onset of Fe deficiency, and their accumulation has been shown to be related to the up-regulation of Fe acquisition genes, with NO and ethylene acting downstream of auxin. Recently, the transcription factors ETHYLENE INSENSITIVE3 (EIN3) and ETHYLENE INSENSITIVE3-LIKE1 (EIL1), two components of the ethylene signaling pathway, have been demonstrated to link ethylene signaling and Fe deficiency responses through interaction with FIT and prevention of FIT degradation by 26S proteasome (Lingam et al., 2011). However, the underlying mechanisms by which other hormones regulate Fe acquisition remain unclear.
The 14-3-3 proteins are a family of phosphoserine-binding proteins that regulate a wide array of target proteins involved in signal transduction, the cell cycle, metabolism, and the stress response (MacKintosh, 2004). The expression of 14-3-3 genes responds to various abiotic stresses, such as low temperature (Jarillo et al., 1994), salt stress (Xu & Shi, 2006), drought (Porcel et al., 2006), phosphorus (P) deficiency (Cao et al., 2007), and Fe deficiency (Wang et al., 2002; Xu & Shi, 2006), indicating that 14-3-3 proteins may be the potential common regulators in plants responding to a wide range of abiotic stresses. In addition, 14-3-3 proteins are also involved in signaling processes triggered by plant hormones such as abscisic acid (ABA), Brassinolide BR and gibberellin GA (Oecking & Jaspert, 2009). As both hormones and 14-3-3 proteins are involved in plant Fe deficiency responses, it is reasonable to hypothesize that there may be interactions or linkages between hormones and 14-3-3 proteins in Fe-deficiency-induced signaling processes.
In A. thaliana, there are 15 members of the 14-3-3 protein family (Rosenquist et al., 2001). Because A. thaliana 14-3-3 proteins regulate a large number of cellular processes, they are also known as general regulatory factors (GRFs) (Rooney & Ferl, 1995). Based on the available microarray data for A. thaliana, a gene encoding GENERAL REGULATORY FACTOR11 (GRF11) is frequently reported to be induced by Fe deficiency (Colangelo & Guerinot, 2004; Dinneny et al., 2008; Buckhout et al., 2009; Long et al., 2010; Yang et al., 2010). In a detailed microarray analysis, Dinneny et al. (2008) found that GRF11 was significantly induced after subjecting plants to Fe deficiency for 24 h and the induction mainly occurred in the elongation and maturation zone. However, the biological functions of this protein in the Fe deficiency response have yet to be identified. In the present study, we investigated the role of GRF11 in the Fe deficiency response in A. thaliana. Our results demonstrate that the expression of core Fe acquisition genes under Fe deficiency is dependent on GRF11. Furthermore, GRF11 acts downstream of NO to mediate FIT expression.
Materials and Methods
For NO-related treatments, the NO donor S-nitrosoglutathione (GSNO) was synthesized according to Stamler & Loscalzo (1992) with glutathione purchased from Sigma, while the NO scavenger 2-(4-carboxyphenyl)-4,4,5,5-tetramethylimidazoline-1-oxyl-3-oxide (cPTIO) was obtained from Dojin Laboratories (Shanghai, China), Nω-nitro-L-arginine methyl ester hydrochloride (L-NAME) from Sigma, and sodium tungstate from Sangon (Shanghai, China). For determination of FCR activity, bathophenanthrolinedisulfonic acid disodium salt hydrate (BPDS) and 4-morpho-lineethanesulfonic acid (MES) were purchased from Sigma. For the yeast two-hybrid assay, YPD medium and synthetic dropout supplements (e.g. –Leu, –Trp, –Leu/–Trp, and –Ade/–His/–Leu/–Trp), and minimal synthetically defined (SD) medium (yeast nitrogen base without amino acids) were all acquired from Clontech (CA, USA), and agar from Sigma.
Plant materials and growth conditions
Arabidopsis thaliana (L.) Heynh. ecotype Columbia-0 (Col-0) was used as the wild type (WT) in the experiments. The bHLH029 T-DNA insertion mutant fit, the NITRIC OXIDE-ASSOCIATED1 (NOA1)-defective mutant noa1, and the nitrate reductase (NR) double mutant nia1 nia2 were the same as in our previous experiments (Chen et al., 2010). The T-DNA insertion line SALK_123168C (grf11) was bought from the Arabidopsis Biological Resource Center (ABRC; Ohio State University, Columbus, OH, USA), and the FIT-Ox line was provided as a gift from H. Q. Ling (Institute of Genetics and Developmental Biology, Beijing, China). To generate transgenic plants over-expressing GRF11, the coding region of GRF11 was amplified using primers 5′- ATTCCATGGATGGAGAACGAGAGAGCGAAG-3′ (forward, with the added NcoI restriction site underlined) and 5′- CGGACTAGTTTAGTTCTCATCTTGTTGATT ATGGCC-3′ (reverse, with the added SpeI restriction site underlined), and cloned into the pMDT19 vector (Takara, Dalian, China) for sequence verification. Then the plasmid was cleaved using NcoI/SpeI and ligated into the binary vector pCAMBIA1304 to generate a pCAMBIA1304-35S::GRF11 construct. The construct was transformed into WT plants (Col-0) via Agrobacterium (EHA 105)-mediated floral dip transformation (Clough & Bent, 1998), and two homozygous lines of transgenic T3 plants with hygromycin resistance, named GRF11-Ox2 and GRF11-Ox5, were used for further analysis.
Seeds were surface-sterilized in 5% (v:v) NaOCl for 5 min, then 95% ethanol for 10 min, and subsequently rinsed thoroughly with sterile water. Seeds were placed onto Petri dishes with half-strength Murashige and Skoog (MS) medium supplemented with 1% sucrose and 0.8% agar (pH 5.5) and kept at 4°C in the dark for 2–4 d, before the plates were transferred to a growth chamber with a temperature of 21°C, relative humidity of 70%, and a 14-h photoperiod (200 μmol m−2 s−1). On day 7, seedlings uniform in size were transferred to vermiculite supplemented with fresh nutrient solutions every other day. After 3 wk, seedlings were transplanted to 1-l pots (nine holes per holder, and one seedling per hole) filled with aerated, complete nutrient solution at pH 6.5. The solution was renewed every 4 d. The nutrient solution had the following composition (in μM): Ca(NO3)2 (300), MgSO4 (50), NaH2PO4 (30), K2SO4 (50), H3BO3 (3), ZnSO4 (0.4), CuSO4 (0.2), MnCl2 (0.5), (NH4)6Mo7O24 (1), and Fe-EDTA (20). After 6 wk of culture, plants were transferred to nutrient solution that contained 20 μM (+Fe) or 0 μM (−Fe) Fe-EDTA for 7 d. For experiments with extrinsic NO treatment, the application of either the NO donor or scavengers has been described previously (Chen et al., 2010).
Acidification capacity was examined according to Yi & Guerinot (1996). Briefly, 12-d-old seedlings grown on agar plates were transferred to Fe-deficient medium for 3 d, and then seedlings were transferred to a 1% agar plate containing 0.006% bromocresol purple and 0.2 mM CaSO4 (pH 6.5) for 24 h.
The WT and the grf11 mutant were grown on half-strength MS, Fe-sufficient or Fe-deficient plates. Root and shoot tissues were harvested and dried in an oven at 70°C for 2 d. Elemental analysis was performed using inductively coupled plasma spectroscopy after the plant samples had been digested with concentrated HNO3 and HClO4 according to a previous report (Jin et al., 2006, 2007).
Perls blue staining
To localize Fe3+, 12-d-old seedlings grown on Fe-sufficient medium were vacuum-infiltrated with Perls stain solution (equal volumes of 4% (v/v) HCl and 4% (w/v) potassium ferrocyanide) for 30 min. Seedlings were then rinsed with water, observed, and photographed with a Nikon AZ100 microscope.
Root FCR activity measurement
FCR activity was determined according to Grusak (1995). Briefly, the whole excised root (c. 0.1 g) was placed in a tube filled with 5-ml of assay solution, which consisted of 0.5 mM CaSO4, 0.1 mM MES, 0.1 mM BPDS and 100 μM Fe-EDTA at pH 5.5 adjusted with 1 M NaOH. The tubes were placed in a dark room at 25°C for 1 h, with periodic hand-swirling at 10-min intervals. The absorbance of the assay solutions was measured at 535 nm, and the concentration of Fe(II)[BPDS]3 was quantified using a standard curve.
Yeast two-hybrid analysis
The yeast two-hybrid analysis was performed according to the manufacturer as described in the Yeast Protocols Handbook (Clontech). The coding sequences of genes FIT and GRF11 were obtained by RT-PCR using primers with introduced restriction sites: 5′-CATATGATGGAAGGAAGAGTCAACGC-3′ (NdeI site underlined) and 5′-GCGGATCCTCAAGTAAATGACTTGATG-3′ (BamHI site underlined) for FIT; 5′-CATATGATGGAGAACGAGAGAGCGAAGC-3′ (NdeI site underlined) and 5′-GTCGACTTAGTTCTCATCTTGTTGATTATGGC-3′ (SalI site underlined) for GRF11. They were then cloned into the pMDT19 vector (Takara). After verification of DNA sequencing, the open reading frame (ORF) of FIT was excised using NdeI/BamHI and inserted into the bait vector pGADT7 to generate the plasmid pAD-FIT, whereas that of GRF11 was cleaved by NdeI/SalI and integrated into the prey vector pGBKT7 to generate the plasmid pBD-GRF11. Then the plasmids were separately transformed into the yeast strains AH109 and Y187, and yeast strain Y187 containing plasmid pBD-GRF11 was mated with AH109 containing the plasmid pAD-FIT. Mated strains were spread on low-stringency SD–Leu/–Trp and high-stringency SD–Ade/–His/–Leu/–Trp to detect protein interactions qualitatively. The interaction between pBD-53 and pAD-T was used as the positive control, and a pair of plasmids with different combinations (Supporting Information Fig. S2) was used as the negative control.
Yeast one-hybrid assay
The yeast one-hybrid assay was performed using the MATCHMAKER Gold Yeast One-Hybrid Library Screening System (Clontech) and YEASTMAKER Yeast Transformation System 2 (Clontech). To investigate the interaction between the FIT protein and GRF11 promoter, we amplified the promoter sequence of GRF11 (Fig. 4a) by PCR from A. thaliana genomic DNA. Primer pairs used for amplification and introduction of restriction sites were 5′-GGGTACCATCTTCTTCGTCGGCCAT-3′ and 5′-GCCTCGAGCAAAATCAAAATTCAACC-3′ (promoter region A), 5′-GGGTACCCCGTATTTCGGTTTAGATT-3′ and 5′-GCCTCGAGAACCATTAAT TGTTGACA-3′ (promoter region B), and 5′-GCGGTACCTTTGGAGATTAA T GACTT-3′ and 5′-GCCTCGAGTTTTCCTTTGTACCTTAT-3′ (promoter region C). The amplified promoter region was cloned upstream of the Aureobasidin A resistance gene (AUR1-C) reporter gene in the pAbAi vector (pAbAi-ProGRF11). The ORF of FIT was cloned in frame after the transcriptional activation domain of the yeast transcription factor GAL4 in pGADT7 (pAD-FIT). A pair of these plasmids, pAbAi-ProGRF11 and pAD-FIT, or the positive control p53-AbAi and pAD-p53 were introduced into yeast strain Y1HGold and cultured on SD medium without Ura containing 0 or 150 ng ml−1 Aureobasidin A (AbA150) at 30°C for 3 d according to the manufacturer's instructions.
Quantitative real-time PCR
Plant material was ground in liquid nitrogen and total RNA was extracted using the RNAplant_plus kit (Tiangen, Shanghai, China). First-strand cDNAs were synthesized from 1 μg of total RNA using the PrimeScript™ RT-PCR Kit (Takara), and diluted to 50–100 ng μl−1. Quantitative real-time PCR was performed on the LightCycler480 machine (Roche) using a SYBR PremixEx Taq kit (Takara). Three biological replicate RNA/cDNA samples were generated, and analysis of each cDNA sample was performed with triplicate technical replicates, from which the relative expression was calculated against that of the internal control gene ubiquitin 10 (UBQ10) using the formula 2−ΔΔCp. Primer pairs used for each gene are listed in Table S1.
Physiological responses of the GRF11 knockout line to Fe deficiency
To examine the biological function of GRF11 in Fe deficiency responses in A. thaliana, we used a T-DNA insertion line (Salk_123168C). The T-DNA was inserted in the promoter region of GRF11 in Salk_123168C (−153 bp from ATG; Fig. S1a), and the full-length transcript of GRF11 was not detected in the homozygous lines (Fig. S1b), indicating that this mutant was a null allele of GRF11.
Because Fe deficiency is generally associated with rhizosphere acidification in Strategy I plants, we first investigated the involvement of GRF11 in acidification capacity by growing both the grf11 mutant and WT plants in Fe-sufficient and Fe-deficient conditions. After being cultured on Fe-sufficient medium for 12 d, seedlings were transferred to Fe-deficient medium for 3 d, and then placed on agar plates containing the pH indicator bromocresol purple for 24 h. A yellow color indicates acidification of the medium. WT plants showed obvious acidification of the rhizosphere, especially in the −Fe treatment, while the grf11 mutant showed no visible acidification (Fig. 1a), indicating that the mutation of GRF11 affects proton release induced by Fe deficiency.
The FCR activity was measured using a BPDS assay. When plants were grown in Fe-sufficient conditions, the root FCR activities of the WT and the grf11 mutant line were similar. However, under Fe-deficient conditions, the root FCR activity was increased by > 4 times in WT plants, but not in the grf11 mutant line (Fig. 1b).
The Fe contents in the seedlings grown under standard half-strength MS, Fe-deficient, and Fe-sufficient conditions were measured. Fe contents in both roots and shoots were always less in the grf11 plants than in the WT plants (Fig. 1c), with the most significant difference found under Fe-deficient conditions. However, there was no significant difference in shoot Fe contents when plants were grown under either MS conditions or Fe-sufficient conditions. We further measured root Fe3+ precipitation with Perls blue staining according to Green & Rogers (2004), where the intensity of blue color indicates the amount of Fe deposited. In the WT plants, heavy staining of cortex cells in the maturation zone was observed, but no staining in the epidermal cells. However, hardly any such staining was observed in grf11 plants (Fig. 1d). All these findings suggested that Fe accumulation in the grf11 mutant was decreased.
To confirm that disruption of GRF11 is responsible for the observed phenotype, a complementation test was performed. A 2.7-kb DNA fragment harboring the 837-bp GRF11 promoter and the candidate gene was introduced into the knockout line by the Agrobacterium (EHA 105)-mediated transformation method. Examination of T2 transgenic lines carrying GRF11 showed that their responses to Fe deficiency, for example, their acidification ability, FCR induction, and Fe uptake (Fig. 1), were all rescued to a similar level to that of the WT plants, indicating that the observed phenotype in the mutant lines was caused by loss of function of the GRF11 gene.
Loss of GRF11 function affects the transcription of Fe acquisition genes
Based on the lower Fe acquisition ability and rhizosphere acidification capacity of the grf11 mutant, we suspected that the expression patterns of genes involved in Fe uptake and proton release in this mutant might be abnormal. Quantitative RT-PCR (qRT-PCR) analysis of FRO2, IRT1 and Arabidopsis H+-ATPase gene AHA2 showed that, in the WT plants, their expression levels were increased by 19-, 30- and 3.6-fold, respectively, by Fe deficiency (Fig. 2), while in the grf11 mutant, they were 35%, 29% and 23% of those of the WT plants under Fe-sufficient conditions; moreover, although the levels of FRO2 and IRT1 were increased by Fe deficiency, they were just 10% and 16% of those of the WT plants under Fe-deficient conditions (Fig. 2). The interesting thing is that the expression of AHA7, another major Fe-deficiency-responsive H+-ATPase isoform that is responsible for the Fe-deficiency-induced differentiation of rhizodermic cells (Santi & Schmidt, 2009), was not changed in the grf11 mutant under Fe-sufficient conditions (Fig. S2). The expression of AHA7 was responsible for Fe deficiency in the mutant, although the expression level was not as high as that of WT plants (Fig. S2). AHA1 expression was found not to be affected in the mutant lines in comparison with WT plants, and not transcriptionally regulated by Fe deficiency (Fig. S2).
35S:GRF11 transgenic plants have no obvious phenotype
Because decreasing the amount of GRF11 mRNA has dramatic effects on Fe acquisition, it was of interest to determine whether increasing the GRF11 copy number had effects on Fe content or the expression of Fe-deficiency-response genes. In two independently generated transgenic lines, GRF11-Ox2 and GRF11-Ox5, the expression of GRF11 was increased by > 270- and 500-fold, respectively, in comparison with the WT (data not shown). In both transgenic lines, the expression of IRT1, FRO2, and AHA2 was higher than that of WT plants under Fe-sufficient conditions and could be further increased by Fe deficiency, except for AHA2, whereas the expression levels of all genes did not exceed that of WT plants under Fe-deficient conditions (Fig. 2). The expression pattern of AHA7 in both transgenic lines was similar to that of IRT1 and FRO2. However, no significant difference in the expression of AHA1 was observed between transgenic lines and WT under Fesufficient conditions, except for the GRF11-OX5 line, which showed a slight increase in AHA1 expression (Fig. S2). Transgenic plants overexpressing either GRF11 or FIT showed similar FCR activity under both Fe-deficient and Fe-sufficient conditions, and similar root and shoot Fe contents to that of WT plants under Fe-sufficient conditions (Fig. S3).
Transcriptional regulation of FIT by GRF11
To elucidate the underlying mechanisms by which GRF11 regulates the expression of Fe acquisition genes, the expression of FIT was monitored. In the fit mutant, the FIT gene could not be detected (Fig. 3a). In the grf11 mutant, the transcript abundance of FIT was significantly reduced compared with that of WT plants under both Fe-sufficient and Fe-deficient conditions, but Fe deficiency could still induce the expression of FIT to a level similar to that of the WT under Fe-sufficient conditions (Fig. 3a), indicating some type of dependence of FIT expression on GRF11. However, in the fit mutant, the transcript abundance of GRF11 was also decreased in comparison with the WT, and the transcriptional induction of GRF11 by Fe deficiency was abolished (Fig. 3b).
Expression levels of FIT in both GRF11 overexpressing lines were higher than that of WT plants under Fe-sufficient conditions, and they could be further increased by Fe deficiency, but did not exceed that of WT plants under Fe-deficient conditions (Fig. 3a). The expression of GRF11 in the FIT-OX line was consistently higher than in WT plants under both Fe-sufficient and Fe-deficient conditions (Fig. 3b).
We examined the direct interaction between the FIT protein and the promoter region of GRF11 using the yeast one-hybrid assay. FIT belongs to the bHLH family and is predicted to recognize a specific sequence, namely E-box motif 5′-CANNTG-3′, in the promoter region of the regulated genes. There are four E-box motifs in the promoter of GRF11, with three motifs in region A (−1038 to −706 from the start codon) and one in region C (−599 to −335) (Fig. 4a). We found that the FIT protein could interact with promoter regions A and C but not with region B (−812 to −559) in the yeast assay (Fig. 4b). These results suggest that FIT could, via a feedback mechanism, regulate the transcription of GRF11 through the E-box motifs of the potential cis-acting elements in the promoter region of GRF11.
GRF11 could not interact with FIT in yeast
Based on the specific expression patterns of GRF11 and FIT in the mutant lines under Fe deficiency described in the previous section, one possibility is that GRF11 could physically interact with FIT to influence FIT transcription. To test this hypothesis, we examined the protein interaction between GRF11 and FIT in a yeast two-hybrid system. The full-length cDNA of FIT was fused to the C-terminus of the activation domain on the pGADT7 vector, while that of GRF11 was fused to the C-terminus of the DNA-binding domain on the pGBKT7 vector; an interaction between these two fusion proteins would allow yeast cells to grow on high-stringency SD–Ade/–His/–Leu/–Trp plates. When yeast strain Y187 containing the plasmid pBD-GRF11 was pairwise mated with AH109 containing the plasmid pAD-FIT, the mated yeast could not survive on the high-stringency plates (Fig. S4a), indicating that GRF11 and FIT could not physically interact with each other. For comparison, yeast cells growing on low-stringency SD–Leu/–Trp plates are also shown (Fig. S4b).
Induction of genes encoding other transcriptional factors involved in Fe acquisition upon the onset of Fe deficiency did not require GRF11
It was reported that heterodimerization of FIT with bHLH038 or bHLH039 can up-regulate the expression of Fe acquisition genes (Yuan et al., 2008), so we further measured the expression levels of all four Fe-deficiency-inducible subgroup Ib bHLH transcription factor genes. As shown in Fig. 5(a), under Fe-sufficient conditions, the expression levels of all four genes in the grf11 mutant did not differ from those in the WT plants, and were all strongly enhanced by Fe deficiency in both the WT and the grf11 mutant. The expression level of bHLH039 was even significantly higher in the grf11 mutant than in the WT. Thus, this rules out the possible involvement of bHLH transcriptional factors in GRF11-dependent regulation of FIT.
Moreover, EIN3 and EIL1 were reported to interact physically with FIT to protect FIT from proteasomal degradation, thereby regulating the expression of Fe acquisition genes (Lingam et al., 2011). However, the expression levels of both genes were very similar between the WT and the grf11 mutant and also not affected by Fe supply (Fig. 5b).
GRF11 acts downstream of NO to participate in Fe deficiency responses
We previously demonstrated that NO was involved in the induction of FCR activity through, at least in part, transcriptional regulation (Chen et al., 2010). In order to examine whether GRF11 acts upstream or downstream of NO in the signaling pathway, we first analyzed the effect of exogenous application of NO on the expression of GRF11 in WT plants. As shown in Fig. 6(a), application of GSNO did not affect the mRNA abundance of GRF11 under Fe-sufficient conditions, whereas it was significantly increased under Fe-deficient conditions. However, the accumulation of the GRF11 transcript induced by Fe deficiency was completely abolished when roots were treated with cPTIO (an NO scavenger), tungstate (a nitrate reductase inhibitor), or L-NAME (an NOA1 inhibitor). The expression of other Fe-deficiency-responsive genes also showed similar responses to the cellular NO status (Fig. 6a). These results indicate that both NOA-dependent and NR-dependent production of NO regulates the expression of GRF11 and Fe acquisition genes under Fe-deficient conditions.
We further analyzed the expression level of GRF11 in the noa1 and nia1 nia2 mutants, both of which have reduced NO synthesis ability (Guo, 2006; Wilson et al., 2008; Chen et al., 2010). Under Fe-sufficient conditions, the expression of GRF11 in the WT and NO synthesis defective mutants was similar, but it was not induced by Fe deficiency in either mutant (Fig. 6b), indicating that GRF11 acts downstream of NO in the signaling pathway.
To further unravel the relationship between NO and GRF11, we compared the effects of exogenous application of NO on FCR activity in the WT and the grf11 mutant. Exogenous application of NO did further enhance the root FCR activity in the WT, but had no effect in the grf11 mutant (Fig. 6c), confirming that GRF11 acts downstream of NO to activate FCR activity under Fe deficiency.
Plants have evolved sophisticated mechanisms to enable them to survive when local Fe bioavailability is limited. The bHLH transcription factor, FIT, plays a central role in mediating the expressions of the major Fe acquisition genes, FRO2, IRT1 and AHA2. However, the signaling pathway involved in transcriptional regulation of FIT expression remains unknown. Here, we demonstrated that a general regulatory factor, GRF11, appears to be involved in this signaling pathway.
First, we demonstrated that GRF11 is required for the proper operation of Fe acquisition mechanisms at physiological and gene expression levels in A. thaliana. This conclusion is based on the following evidence. First, loss of function of GRF11 resulted in the failure of acidification (Fig. 1a) and FCR induction (Fig. 1b), and thus decreased Fe uptake (Fig. 1c,d), while in the complemented line all these features were rescued (Fig. 1). Secondly, expression of genes mediating Fe uptake and acidification, namely IRT1, FRO2, and AHA2, was suppressed in the grf11 mutant both under Fe-deficient and under Fe-sufficient conditions (Fig. 2). This is completely in accordance with the lower Fe content and lack of acidification observed in the mutant. The expression of another Fe-deficiency-responsive gene, AHA7, was not affected in the grf11 mutant (Fig. S2), indicating that GRF11 is somewhat specifically involved in the processes of Fe uptake. Thirdly, the expression of GRF11 in WT plants was induced by Fe deficiency (Figs 3, 6). This is consistent with the results of previous microarray analyses (Colangelo & Guerinot, 2004; Dinneny et al., 2008; Buckhout et al., 2009; Long et al., 2010; Yang et al., 2010). Finally, over-expression of GRF11 can positively affect the expression of FIT under Fe-sufficient conditions (Fig. 3).
FIT has been documented to be a key player in mediating Fe deficiency responses in A. thaliana (Colangelo & Guerinot, 2004; Yuan et al., 2005), and here we found that the expression of GRF11 and that of FIT were interrelated. Therefore, it will be interesting to elucidate their relationships. Knockout of GRF11 resulted in a significant decrease in the expression of FIT and over-expression of GRF11 enhanced the expression of FIT under Fe-sufficient conditions (Fig. 3a), implying a positive regulatory effect of GRF11 on FIT expression. The question then arises: how could a protein that is not a transcription factor modulate the transcription of FIT? A possible answer is that GRF11 is necessary for the action of an as yet unidentified transcription factor or FIT itself to regulate FIT transcription. Indeed, the FIT protein appears to act as a positive regulator of its own gene, as suggested by the analysis of FIT expression in the fit-4 mutant, which cannot produce FIT protein as a result of the presence of an early stop codon (Jakoby et al., 2004). However, knockout of FIT or over-expression of FIT caused a significant decrease or increase of expression of GRF11 under Fe-sufficient conditions (Fig. 3b), supporting the hypothesis that FIT is a transcriptional regulator of GRF11. It has been reported that there is an E-box motif in the promoter region of GRF11 (Colangelo & Guerinot, 2004), which could be a potential FIT binding site. The yeast one-hybrid assay verified the direct interaction between the FIT protein and the promoter region of GRF11 (Fig. 4). Taking these results together, a model can be proposed in which Fe deficiency results in GRF11 accumulation, and GRF11 regulates the expression of FIT, which in turn promotes the expression of GRF11 (Fig. 7). As the expression of either FIT or GRF11 is attenuated but not completely abolished in the grf11 mutant or fit mutant, respectively (Fig. 3), it is likely that other factors act together with either FIT or GRF11 in a combinatorial fashion or that other redundant factors are also involved. Further investigation of this is needed.
Posttranscriptional regulation of FIT is also involved in the Fe deficiency response (Lingam et al., 2011; Sivitz et al., 2011). The finding that over-expression of GRF11 or FIT had no effects on root FCR activity and Fe uptake (Fig. S3) also support the view of the posttranscriptional regulation of the Fe uptake system. Thus, we investigated whether, in addition to transcriptional regulation, GRF11 is also involved in the posttranscriptional regulation of FIT. To address this question, we first investigated whether GRF11 could interact with FIT to affect FIT protein expression levels. However, the yeast two-hybrid analysis ruled out the possibility of a direct interaction (Fig. S4). Secondly, we investigated the expression of four subgroup Ib bHLH genes, namely bHLH038, bHLH039, bHLH100, and bHLH101, and two central transcription factors of the ethylene signaling pathway, namely EIN3 and EIL1, in the grf11 mutant to determine whether the expression of these genes is reduced. Again, expression of genes encoding these transcription factors was not reduced under either Fe-sufficient or Fe-deficient conditions in the mutant (Fig. 5). Thus, it appears that GRF11 is not involved in the posttranscriptional regulation of FIT.
We previously demonstrated that NO is involved in Fe-deficiency-induced FIT expression (Chen et al., 2010). However, the mechanism of this regulation is still not known. Here, we further demonstrated that NO-dependent regulation of FIT transcription under Fe deficiency requires the GRF11 function. In WT plants, Fe-deficiency-induced expression of GRF11 could be further enhanced by exogenous application of NO and eliminated by NO scavengers (Fig. 6a). In both the noa1 and nia1 nia2 mutants, where endogenous NO concentrations are much reduced, the transcriptional response of GRF11 to Fe deficiency was eliminated (Fig. 6b). However, in the grf11 mutant, exogenous application of NO could not restore the Fe-deficiency-induced FCR activity (Fig. 6c). The essence of NOA1 in plants has recently been demonstrated as a functional GTPase (Moreau et al., 2008). However, the finding that plant NO concentrations were significantly reduced when plants were grown in the absence of sucrose (Guo, 2006; Chen et al., 2010; Ree et al., 2011) suggests that this mutant is still of value for investigating the biological function of NO. These results indicated that GRF11 acts downstream of NO to trigger Fe-deficiency-tolerance mechanisms, and both NOA1-dependent and NR-dependent NO production is involved in the regulation of GRF11 expression under Fe-deficient conditions.
In conclusion, we have demonstrated for the first time that a member of the 14-3-3 protein family, GRF11, is involved in the expression of Fe-deficiency-tolerance mechanisms in A. thaliana, and that the activity of FIT is necessary for the transcriptional activation of GRF11 expression. Furthermore, we present a novel signaling pathway in which GRF11 acts downstream of NO to regulate FIT expression (Fig. 7).
This work was supported by a Changjiang Scholarship from the Ministry of Education, and by grants from the Innovative Research Team (IRT1185) and the Fundamental Research Funds for the Central Universities.