Identification of OsbHLH133 as a regulator of iron distribution between roots and shoots in Oryza sativa

Authors

  • LU WANG,

    1. State Key Laboratory of Plant Physiology and Biochemistry
    2. Joint Research Laboratory in Genomics and Nutriomics, College of Life Sciences, Zhejiang University, Hangzhou 310058, China
    Search for more papers by this author
  • YINGHUI YING,

    1. State Key Laboratory of Plant Physiology and Biochemistry
    2. Joint Research Laboratory in Genomics and Nutriomics, College of Life Sciences, Zhejiang University, Hangzhou 310058, China
    Search for more papers by this author
  • REENA NARSAI,

    1. Australian Research Council Centre of Excellence in Plant Energy Biology
    2. Centre for Computational Systems Biology, University of Western Australia, Crawley 6009, Western Australia, Australia
    Search for more papers by this author
  • LINGXIAO YE,

    1. State Key Laboratory of Plant Physiology and Biochemistry
    Search for more papers by this author
  • LUQING ZHENG,

    1. State Key Laboratory of Plant Physiology and Biochemistry
    Search for more papers by this author
  • JINGLUAN TIAN,

    1. State Key Laboratory of Plant Physiology and Biochemistry
    Search for more papers by this author
  • JAMES WHELAN,

    1. Joint Research Laboratory in Genomics and Nutriomics, College of Life Sciences, Zhejiang University, Hangzhou 310058, China
    2. Australian Research Council Centre of Excellence in Plant Energy Biology
    Search for more papers by this author
  • HUIXIA SHOU

    Corresponding author
    1. State Key Laboratory of Plant Physiology and Biochemistry
    2. Joint Research Laboratory in Genomics and Nutriomics, College of Life Sciences, Zhejiang University, Hangzhou 310058, China
      H. Shou. E-mail: huixia@zju.edu.cn
    Search for more papers by this author

H. Shou. E-mail: huixia@zju.edu.cn

ABSTRACT

Iron (Fe) is an essential micronutrient element for plant growth. Regulation of Fe-deficiency signalling networks is one of the many functions reported for basic helix-loop-helix (bHLH) transcription factors in plants. In the present study, OsbHLH133 was found to be induced by Fe-deficiency conditions in Oryza sativa. Insertional inactivation of OsbHLH133 (bhlh133) resulted in growth retardation, with enhanced Fe concentration seen in shoots, and reduced Fe concentration in roots. Overexpression of OsbHLH133 had the opposite effect, that is resulted in an enhanced Fe concentration in roots and reduced Fe concentration in shoots and also in xylem sap. Microarray analysis showed that some of the genes encoding Fe-related functions were up-regulated under Fe-sufficient conditions, in bhlh133 mutant plants compared to wild-type plants. Significant differential expression of a number of signalling pathways, including calcium signalling, was also seen in bhlh133 plants compared to wild-type plants, independent of Fe conditions.

INTRODUCTION

Iron (Fe) is an essential mineral element for both plants and animals. Although it is abundant in soil, the low solubility in neutral to alkaline soil limits its uptake and utilization. Either Fe excess or deficiency is harmful to plants (Halliwell & Gutteridge 1992). Thus, Fe homeostasis must be precisely maintained in plants (Hindt & Guerinot 2012; Kobayashi & Nishizawa 2012). Plants have two distinct strategies for Fe acquisition. Strategy I is used by non-graminaceous plants such as Arabidopsis thaliana (Arabidopsis). In these species, ferric iron (Fe3+) is reduced to the more soluble ferrous iron (Fe2+) by the ferric-chelate reductase FRO2 (Robinson et al. 1999). Subsequently, the major metal transporter, iron-regulated transporter (IRT1), takes up Fe2+ into plant roots (Eide et al. 1996; Vert et al. 2002). Graminaceous plant species use the strategy II mechanism. In these species, mugineic acids (MAs) are synthesized and secreted from plant roots by transporters, that is OsTOM1 in rice (Ishimaru et al. 2006; Nozoye et al. 2011). MAs can chelate Fe3+ to form the Fe3+–MA complexes, which are transported into roots by a specific family of transporters called Yellow Stripe transporters, which were first characterized in Zea mays (maize) as ZmYS1, and later in a variety of other plant species such as barley (HvYS1) and rice (OsYSL15) (Curie et al. 2001; Murata et al. 2006; Inoue et al. 2009). Although rice has the molecular components to take up Fe using both strategies (Ishimaru et al. 2006; Cheng et al. 2007), it primarily utilizes strategy I, because the synthesis and diffusion of MAs is limited in rice (Ishimaru et al. 2006). A Fe2+ transporter, OsIRT1, is found to mediate the transport of Fe2+ from the external solution to the root cells in rice (Bughio et al. 2002; Ishimaru et al. 2006).

A number of genes, including transcription factors, have been identified to be involved in the regulation of Fe acquisition and homeostasis in Arabidopsis and rice (Hindt & Guerinot 2012; Kobayashi & Nishizawa 2012). Basic helix-loop-helix (bHLH) proteins comprise one of the largest transcription factor families in eukaryotic organisms. In total, there are 158 bHLH genes in Arabidopsis and 173 bHLH genes in rice that have been identified; however, only a few of these have been functionally characterized (Li et al. 2006; Pires & Dolan 2010). The regulation of signalling networks during Fe deficiency is one of the many functions reported for bHLH transcription factors in plants. The first bHLH transcription factor related to Fe supply was identified in tomato and named FER (Ling et al. 2002). A fer mutant failed to activate the strategy I pathway of Fe acquisition and displayed severe chlorosis in tomato (Ling et al. 2002; Brumbarova & Bauer 2005). The Arabidopsis protein Fe-deficiency-induced transcription factor (FIT) is the ortholog to FER (Yuan et al. 2005; Bauer, Ling & Guerinot 2007) and forms heterodimers with other bHLH proteins, encoded by AtbHLH38 or AtbHLH39 (Yuan et al. 2008). The complex of FIT/AtbHLH38 or FIT/AtbHLH39 positively regulates the epidermal expression of the Fe reductase gene FRO2, and further induces expression of the Fe2+ transporter IRT1 (Yuan et al. 2008). Another bHLH transcription factor in Arabidopsis, named POPEYE (PYE), interacts with the bHLH transcription factor IAA-Leu Resistant3 (ILR3) to regulate metal ion homeostasis in Arabidopsis (Rampey et al. 2006; Long et al. 2010). PYE positively regulates the growth and development of Arabidopsis under Fe-deficient conditions, under which the mutant pye displayed inhibition of root growth, and severe chlorosis (Long et al. 2010). In addition, the Fe-chelate reductase activity and acidification of rhizospheres in the pye mutant were decreased (Long et al. 2010).

In rice, regulation of the strategy II MA pathway has been shown to be under the control of the bHLH protein OsIRO2 (Ogo et al. 2007). OsIRO2 is induced by Fe deficiency and positively regulates the expression of genes involved in MA biosynthesis, including nicotianamine (NA) synthase genes OsNAS1/2, NA aminotransferase gene OsNAAT1, deoxymugineic acid synthase gene OsDMAS1, metallothionein-like gene OsIDS1 and the Fe3+-DMA transporter OsYSL15 (Inoue et al. 2003; Bashir et al. 2006; Ogo et al. 2006, 2007; Cheng et al. 2007). A previous study in our laboratory identified a negative regulator, involved in Fe homeostasis, named OsIRO3, and overexpression of OsIRO3 was seen to suppress in the expression of genes normally induced under Fe deficiency in rice (Zheng et al. 2010). Overexpression of OsIRO3 also led to hypersensitivity to Fe deficiency, with lower chlorophyll content and Fe concentration observed in the rice shoots (Zheng et al. 2010).

Two other transcription factors, IDEF1 and IDEF2, which specifically bind the iron deficiency-responsive element 1 (IDE1) and 2 (IDE2), respectively, are constitutively expressed in rice roots and shoots (Kobayashi et al. 2003, 2007; Ogo et al. 2008). IDEF1 is essential for the early response to Fe deficiency and positively regulates a number of Fe-responsive genes (Kobayashi et al. 2009). Recently, it was reported that IDEF1 directly binds to divalent metals, enabling the Fe nutritional status to be ‘sensed’ (Kobayashi et al. 2011).

In this study, we characterized a novel bHLH transcription factor, named OsbHLH133 (LOC_Os12g32400.1) in rice (Li et al. 2006). This transcription factor is strongly up-regulated under Fe-deficiency conditions and functional analysis suggests that it acts as an important regulator of Fe distribution in rice plants.

MATERIALS AND METHODS

Plant materials

The rice cultivar Nipponbare (NIP) and Dongjin (DJ) served as the wild-type (WT) controls for the overexpression lines (OE) and bhlh133 mutant, respectively. The seeds of OE lines and the WT counterparts were harvested from the winter nursery, Hainan province, China. Seeds of mutant and DJ were harvested from the Zhejiang province, China.

To construct the overexpression vector of OsbHLH133, the full length of cDNA sequence was amplified using the primer pairs of 5′-ATGGAATGCAGCTCCTTTGAAGCA-3′ and 5′-ACGGATCCCTACTCTTGTGACAGAGAGTTCTG-3′. The amplified fragment was cloned into pTF101-ubi at the BamHI site (Zheng et al. 2010). Then, Agrobacterium-mediated rice transformation was carried out as described previously (Chen et al. 2003). To verify the transgenic plants, quantitative RT-PCR (qRT-PCR) analysis was conducted.

A T-DNA insertion mutant bhlh133 was obtained from the rice T-DNA insertion sequence database (http://signal.salk.edu/cgi-bin/RiceGE) on DJ genotype (Jeon et al. 2000; Jeong et al. 2006). The insertion site of the T-DNA is in the start codon of the gene, just between the A and T base of the start (ATG) codon (Supporting Information Fig. S2a). The insertion was confirmed using PCR with T-DNA left boarder primer (5′-ACGTCCGCAATGTGTTATTAA-3′) and gene-specific primers (5′-TAGTGTCACTCTCCTGTGCT-3′ and 5′-CGAAGCTATCATCTTGGTAA-3′). qRT-PCR analysis was performed to further confirm the effect of T-DNA insertion on the expression of OsbHLH133.

Growth conditions

Rice seeds were germinated in water for 3 d and then transferred to culture solution containing 1.425 mM NH4NO3, 0.323 mM NaH2PO4, 0.513 mM K2SO4, 0.998 mM CaCl2, 1.643 mM MgSO4, 0.009 mM MnCl2, 0.075 µM (NH4)6Mo7O24, 0.019 mM H3BO3, 0.155 µM CuSO4, 0.152 µM ZnSO4, 100 µM or 2 µM ethylenediaminetetraacetic acid (EDTA)-Fe (II) (Yoshida et al. 1976). Plants were grown in a growth chamber with 12 h 30 °C light/12 h 22 °C dark. The pH of the nutrient solution was adjusted to 5.5 with 2 N HCl, and the solution was renewed every two days.

Fe-deficiency treatments were performed as described before (Zheng et al. 2010). For time-course analysis, 2-week old seedlings were transferred to Fe-deficient nutrient solution. Plants were harvested on days 0, 1, 2, 3, 4, 5 and 6 of Fe deficiency. On day 6 of Fe deficiency, Fe was re-supplied by adding 100 µM EDTA-Fe (II), and plants were harvested 3 d later. To investigate the phenotypes of overexpressing and mutant plants, seedlings were germinated and grown on the culture solution supplied with 100 µM or 2 µM EDTA-Fe (II). After 3 weeks of treatment, shoot lengths and SPAD value (total chlorophyll content) were measured. For Fe concentration measurement, seedlings were grown in the normal nutrient solution for 3 weeks, and then were transferred to nutrient solution with or without EDTA-Fe (II) for 10 d.

Measurement of Fe concentration

To determine the concentration of Fe in plants, samples were dried for 3 d at 80 °C before being weighed. Afterward, they were digested in 5 mL of 11 N HNO3 for 5 h at 150 °C.

For xylem sap collection, rice plants were cut at a height of 3 cm above the roots, a column filled with filter paper was placed on the cut end. Xylem sap was collected for 1 h and centrifuged into new tubes. Ten individual plants were pooled into one sample and samples were digested with 0.5 mL of 11 N HNO3 at 80 °C for 60 min. After digestion, samples were diluted to a 3 mL volume.

Fe concentration was measured using inductively coupled plasma mass spectrometry (ICP-MS, Agilent 7500ce, Santa Clara, CA, USA).

Measurement of chlorophyll content

A portable chlorophyll meter (SPAD-502; Konica Minolta Sensing, JP) was used to measure the SPAD value (total chlorophyll content) of the fully expanded youngest leaves.

Histochemical β-glucuronidase (GUS) staining assay

To investigate the expression pattern, 2009 bp upstream of the ATG codon of the OsbHLH133 gene was amplified using the primers pair: 5′-ATCTCCGTTAGGCCATGTCCAAC-3′ and 5′-GCAGATTGCTTCAAAGGAGCTGCA-3′. The promoter was cloned into pBI101.3, connected to a GUS gene (Jefferson, Kavanagh & Bevan 1987) to produce a pOsbHLH133:GUS reporter construct. Transgenic plants were generated via Agrobacterium-mediated transformation. For the histochemical GUS staining assay, T1 transgenic seeds were germinated and grown under Fe-replete and Fe-deprivation solutions for 10 d. Roots and leaves were subjected to GUS-staining buffer containing 100 mM sodium phosphate (pH 7.0), 1 mM X-Gluc (5-bromo-4-chloro-3-indolyl-β-D-glucuronidase), 1 mM K4[Fe(CN)6], 1 mM K3[Fe(CN)6], 0.5% (v/v) TritonX-100 and 20% (v/v) methanol. After staining, roots and leaves were imbedded in 5% (w/v) low-melting-point agarose (Sigma, St Louis, MO, USA). Sections of 50 or 60 µm thickness were cut by vibrating microtome (VT 1000 S, Leica, Bensheim, Germany) and the images were examined under a microscope (Eclipse 90i, Nikon, Tokyo, Japan).

Affymetrix microarray analysis

Seedlings were germinated and grown on the culture solution supplied with 100 µM or 2 µM EDTA-Fe (II) and after 3 weeks of treatment, roots of the seedlings were sampled, snap frozen in liquid nitrogen and stored at −80 °C until RNA isolation. The RNA isolation and microarrays were performed according to manufacturers' instructions, as described in (Zheng et al. 2009). Three biological replications were used for the analysis. All CEL files were first normalized by MAS5 to determine present/absent calls and only those genes present in = >2 replicates were used for further analysis. Data were normalized by GeneChip Robust Multiarray Averaging (GC-RMA) using the Partek Genomics Suite software (version 6.5) and differential expression analysis was carried out using Cyber-T (Baldi & Long 2001; Choe et al. 2005). Genes were defined as significantly differentially expressed if the fold-change had an associated P value of <0.05 and false discovery rate was posterior probability of differential expression (PPDE) >0.96, as carried out in (Narsai, Castleden & Whelan 2010; Narsai et al. 2011). Significant over-representation of functional categories was determined by a z-score analysis as described in (Narsai et al. 2009). Over-representation of functional categories was determined using Fisher's test (ORA Cut-off value = 1) within the Pageman software (Usadel et al. 2006).

qRT-PCR

Total RNA was extracted from root or leaf samples using Trizol reagent according to manufacturer's instruction (Invitrogen, Carlsbad, CA, USA). RNA samples were then treated with RNase-free DNase I (TaKaRa Biomedicals, Tokyo, Japan) to remove the residues of genomic DNA. First-strand cDNA was synthesized using M-MLV reverse transcriptase (Promega, Madison, WI, USA). cDNA was amplified by PCR in LightCycler 480 machine (Roche Diagnostics, Basel, Switzerland) with SYBR Premix Ex Taq (Perfect Real Time) Kit (TaKaRa Biomedicals). Triplicate quantitative assays were performed on each sample and reference gene OsActin was used as an internal control. Transcript levels were calculated relative to OsActin using the formula 2−ΔCt or 2−ΔΔCt. All primer sequences used for the PCR reactions are shown in Supporting Information Table S1.

Statistical analysis of data

For comparisons of treatments in Figs 3–5, significance was determined by a one-way analysis of variance and for differences between groups, with least significant difference (LSD) t-test (P ≤ 0.05).

RESULTS

Fe deficiency specifically induces expression of OsbHLH133

Previous microarray analyses showed that the transcript abundance of OsbHLH133 was significantly induced by Fe deficiency in rice roots (Zheng et al. 2009). To confirm this expression pattern, 3-week-old rice seedlings were transferred to Fe-, Mn-, Cu- or Zn-deprived nutrient solutions for 10 d. qRT-PCR analysis showed that the transcript abundance of OsbHLH133 was specifically induced under Fe and copper (Cu) deficiency conditions in roots, but not by other mineral nutrient deficiencies (Fig. 1a). A 60-fold induction of OsbHLH133 expression was observed in response to Fe deficiency, with a 4.6-fold induction was observed also in response to Cu deficiency in roots (Fig. 1a). The transcript level of OsbHLH133 in roots was higher than that in leaves regardless of Fe supplies (Fig. 1b). In leaves, the basal level of OsbHLH133 expression was very low under normal conditions. However, under Fe-deficient conditions, the expression of OsbHLH133 was seen to be significantly induced in both leaves and roots (Fig. 1b).

Figure 1.

Expression analysis of OsbHLH133 transcript abundance in WT rice plants. (a) Transcript abundance of OsbHLH133 was determined by qRT-PCR under different nutrient deficient conditions. All values are expressed relative to the expression level under nutrient-sufficient conditions (control-set to 1.0) as appropriate. (b) Transcript levels of OsbHLH133 relative to OsActin in root and leaf during Fe-deficiency treatment over 6 d, followed by a return to Fe-sufficient conditions for 3 d. (c) Kinetic analysis showing the induction of OsIRO2, OsIRO3, OsbHLH133 and IDEF1 during Fe-deficiency treatment.

As a transcription factor, OsbHLH133 was targeted to the nucleus (Supporting Information Fig. S1). Kinetic analysis of OsbHLH133 induction and the induction of three other transcription factors involved in Fe-deficiency response in rice, OsIRO2, OsIRO3 and IDEF1, were carried out to gain insight into the role of these during Fe deficiency. In roots, OsbHLH133 was induced approximately 10-fold after just 1 d of Fe-deficiency treatment and reached peak expression at 5 d post-treatment (Fig. 1c). In leaves, the transcript abundance of OsbHLH133 was very low under Fe-sufficient conditions, and was slightly induced under Fe-deficiency treatment (Fig. 1c). Notably, the pattern of OsbHLH133 induction was seen to be comparable to that seen for OsIRO2 (Fig. 1c).

Growth performance of OsbHLH133 overexpression lines and T-DNA insertion mutants

To examine the function of OsbHLH133 in rice plants, a T-DNA insertion mutant bhlh133 was obtained from the rice T-DNA insertion sequence database (Jeon et al. 2000; Jeong et al. 2006). Transgenic lines that constitutively overexpress the OsbHLH133 gene in the Nipponbare genotype (NIP) were also developed to gain further insight into its function. Using qRT-PCR analysis, two independent overexpression lines (OE) of OsbHLH133 were confirmed and named OE-1, OE-2 (Fig. 2). As expected, the transcript levels of OsbHLH133 in the OE lines were significantly higher than the WT counterparts, while no OsbHLH133 transcript was detected in the bhlh133 mutant, regardless of Fe supply (Fig. 2).

Figure 2.

Expression of OsbHLH133 in the root of overexpression (OE) and mutant (bhlh133) plants in Fe-sufficient (+Fe) and -deficient (−Fe) conditions. As expected, no transcripts encoding OsbHLH133 could be detected in the bhlh133 mutant plants and over 100-fold induction is seen in the OE plants.

Under Fe-sufficient (+Fe) conditions, bhlh133 mutant plants displayed a small but significantly (P < 0.05) reduced shoot length (Fig. 3a, Supporting Information Fig. S2b,c). Seed set from bhlh133 plants was reduced to 30% compared to WT under Fe-sufficient conditions (Supporting Information Fig. S2d). In contrast, when seeds were germinated and grown in Fe-deficient conditions for 3 weeks, no significant differences in shoot length were observed between the bhlh133 mutant plants and WT plants (Fig. 3a). Interestingly, under Fe-sufficient conditions, no significant differences in growth performance were observed between the OsbHLH133-OE and the WT seedlings (Fig. 3b). However, under Fe-deficient conditions, the shoot length of the OsbHLH133-OE lines was significantly shorter than the WT (Fig. 3b). To assess the degree of leaf chlorosis caused by Fe deficiency in the OsbHLH133-OE lines, the chlorophyll content (in SPAD units) of the youngest fully expanded leaves of the WT, bhlh133 mutant and OsbHLH133-OE plants was measured. Results showed that the chlorophyll content in the bhlh133 mutant was significantly (P < 0.05) higher than the WT, while the chlorophyll content in the OsbHLH133-OE lines was significantly (P < 0.05) lower than that of the WT plants (Fig. 3c). These results suggest that overexpressing OsbHLH133 leads to hypersensitivity to Fe deficiency, while the bhlh133 knockout plants show improved tolerance to Fe deficiency in rice.

Figure 3.

Growth performance of overexpression (OE) and mutant (bhlh133) plants in Fe-sufficient (+Fe) and -deficient (−Fe) conditions. Seedlings were germinated and grown on the culture solution supplied with 100 µM (+Fe) or 2 µM EDTA-Fe (II) (−Fe) for 3 weeks. (a) Growth performance and shoot length of mutant plants under Fe-sufficient (+Fe) and -deficient (−Fe) conditions. (b) Growth performance and shoot length of OE lines under Fe-sufficient (+Fe) and -deficient (−Fe) conditions. (c) SPAD value of the youngest and fully expanded leaves. Bar = 3 cm. Data shown as the mean ± SD (n = 5). Significance (P < 0.05) of differences in (longer/shorter or higher/lower) compared to WT is indicated by an asterisk.

Fe concentrations of OsbHLH133 overexpression lines and T-DNA insertion mutants

To observe the effect of OsbHLH133 abundance on Fe concentrations in plants, the Fe concentrations in the leaves, stems and roots of the OsbHLH133 overexpression lines, bhlh133 mutant and WT plants were analysed (Fig. 4). Under Fe-sufficient conditions, the Fe concentrations in the leaves and stems of the OsbHLH133-OE plants were significantly lower than WT (Fig. 4a). In contrast, the Fe concentration in the roots of the OE-1 and OE-2 plants was significantly higher than WT, 2.2- and 1.8-fold, respectively (Fig. 4a). Notably, under Fe-deficient conditions, OsbHLH133-OE seedlings showed the same overall pattern of Fe distribution between roots and leaves, as seen under Fe-sufficient conditions (Fig. 4a).

Figure 4.

Analysis of Fe concentration in OsbHLH133-OE and mutant plants. Three weeks old seedlings were transferred to nutrient solutions containing 100 µM (+Fe) or 0 µM (−Fe) EDTA-Fe (II) and after 10 d of treatment, Fe concentrations in leaves, stems and roots were measured by ICP-MS. Data are shown as the mean ± SD (n = 3). Significance (P < 0.05) of differences in Fe concentration (higher/lower) compared to WT is indicated by an asterisk respectively.

Under Fe-sufficient conditions, the Fe concentration in the leaves of the bhlh133 mutant plants was significantly higher than that of the WT counterpart, while the root Fe concentration in the bhlh133 mutant was lower than that found in WT plants (Fig. 4b). These observations in the bhlh133 mutant plants are in agreement with the observations seen in the OsbHLH133-OE lines (Fig. 4).

Fe concentrations in xylem sap were also measured and the results showed that Fe concentration in the xylem sap from OsbHLH133-OE lines was decreased under both Fe-sufficient and deficient conditions (Fig. 5). The Fe concentration in the xylem sap of both OE-1 and 2 lines were either marginally significant (P < 0.06) or significantly (P < 0.05) less than that of the WT, respectively (Fig. 5). In bhlh133 mutant plants, Fe concentration in the xylem sap showed no significant difference from the WT counterpart under Fe-deficient, Fe-sufficient or excess Fe conditions, although a slight increase was seen under the excess Fe conditions (Supporting Information Fig. S3).

Figure 5.

Fe concentrations in xylem sap of OE plants. Three-week-old seedlings were transferred to nutrient solutions containing 100 µM (+Fe) or 2 µM EDTA-Fe (II) (−Fe), after 10 d treatment, xylem sap was collected. Fe concentrations were measured by ICP-MS. Data are shown as the mean ± SD (n = 3). Asterisks indicates significantly lower Fe concentration (P < 0.05), and a hat (^) indicates marginally significantly lower Fe concentration (P < 0.06).

Note that there was no major difference in the zinc (Zn), manganese (Mn) or copper (Cu) concentrations among the leaves, roots or xylem sap of the mutant, OsbHLH133-OE lines and WT (Supporting Information Figs S4 & S5).

Spatial expression pattern of OsbHLH133

To determine the pattern of transcriptional regulation of OsbHLH133 in rice, we assayed the GUS expression of pOsbHLH133:GUS reporter construct in transgenic plants. Four independent transgenic lines were analysed. Under Fe-sufficient conditions, weak GUS expression was mainly observed in the maturation zone of the roots (Fig. 6a), but not at the root tip (Fig. 6c). Cross-sections of the roots exhibited that GUS activity was present in the endodermal cells of the maturation zone, which was 4 cm above the root tip (Fig. 6b), but not in the section of the zone 0.5 cm above the root tip (Fig. 6d). In contrast, under Fe-deficient conditions, GUS expression was observed throughout the roots, including the epidermis, exodermis, cortex, endodermis and the stele (Fig. 6e–h), indicating expression of OsbHLH133 across these cell types. In leaves, GUS activity was barely detectable when plants were grown in the presence of Fe (data not shown). However, under Fe-deficient conditions, GUS expression was observed in the vascular bundles of the leaf blade and leaf sheath (Fig. 6i,j), confirming the induction of OsbHLH133 expression under these conditions.

Figure 6.

Spatial expression patterns of OsbHLH133 in roots and leaves. Histochemical GUS staining was performed using 10-day-old seedlings grown on Fe-sufficient (a, b, c, d) or Fe-deficient (e, f, g, h, i, j) nutrient solutions. (a, b, e, and f), Gus staining on the maturation zones of pbHLH133:GUS transgenic roots; (c, d, g, and h), Gus staining on the zones of pbHLH133:GUS transgenic root tips; (i and j), Gus staining on Fe-deficient leaf blade and leaf sheath, respectively. (b, d, f and h), images of the root sections that were 0.5 cm away from the root tip (d, h) and 4 cm away from the root tip (b, f). Scale bars = 500 µm for (a, c, e, g) and 100 µm for (b, f, d, h, i, j).

Microarray analysis of bhlh133

To examine the gene expression changes seen in the WT and bhlh133 mutant plants, genome-wide expression analysis was carried out on root tissues using Affymetrix rice genome microarrays. As expected, both the qRT-PCR and microarray analysis showed that the expression of OsbHLH133 was significantly (P < 0.05, PPDE >0.96) suppressed in the roots of the bhlh133 mutant plants, and induced under Fe deficiency in the WT plants (Table 1). In the WT and bhlh133 mutant comparisons, only 809 genes were seen to be differentially expressed under Fe-sufficient conditions (Fig. 7a, orange circle; Supporting Information Table S2), compared to nearly double this number of genes (1597 genes) differentially expressed under Fe-deficient conditions (Fig. 7a, dark pink circle, Supporting Information Table S2), thereby supporting a specific role for OsbHLH133 under Fe-deficient conditions in roots.

Table 1.  Differential expression (in roots) of genes encoding Fe-related functions
TIGR identifiersGene nameFold-change
bhlh133 versus WT−Fe versus +Fe
+Fe−FeWT bhlh133
  1. Fold-changes are shown, ‘X’ indicates no significant differential expression (significance was determined by a P-value of <0.05 and false discovery rate was PPDE >0.96).

LOC_Os12g32400.1OsbHLH133X-15.158.97X
Strategy II
DMA biosynthesis     
 LOC_Os03g19427.1OsNAS11.65X3.131.92
 LOC_Os03g19420.1OsNAS21.38X2.361.74
 LOC_Os07g48980.1OsNAS31.151.261.221.33
 LOC_Os02g20360.1OsNAAT11.29X2.602.06
 LOC_Os03g13390.1OsDMAS11.36X3.382.58
 LOC_Os06g02220.1OsMTNXX1.741.76
 LOC_Os04g57400.1OsMTKXX1.571.64
 LOC_Os01g22010.1OsSAM2XX1.231.26
 LOC_Os12g39860.1OsAPT1XX1.251.28
 LOC_Os04g24140.1OsRPI1.33X3.132.56
 LOC_Os11g29370.1OsDEPXX1.831.80
 LOC_Os09g28050.1OsIDI4XX1.741.71
Transporter     
 LOC_Os11g04020.1OsTOM12.29X7.223.25
 LOC_Os02g43410.1OsYSL152.751.147.212.99
Transcription factor     
 LOC_Os01g72370.1OsIRO2XX13.937.78
 LOC_Os03g26210.1OsIRO3XX33.7835.06
Strategy I     
 LOC_Os03g46470.1OsIRT1XX5.504.78
 LOC_Os03g46454.1OsIRT2XX7.598.09
Fe transport and mobilization     
 LOC_Os07g15460.1OsNramp11.77X6.013.48
 LOC_Os04g38940.1OsVIT1;23.14X11.833.97
 LOC_Os02g43370.1OsYSL21.64X4.162.61
 LOC_Os12g18410.1OsMIRXX2.902.36
 LOC_Os11g05390.1OsENA11.37X3.802.78
 LOC_Os06g48060.1OsENA2XX1.301.25
 LOC_Os03g11734.1OsFRDL11.59X4.572.95
Figure 7.

Differential expression analyses in response to +/−Fe conditions in WT and bhlh133 mutant. (a) (i) Venn diagram showing the number of significantly differentially expressed genes (P < 0.05, PPDE >0.96) in the bhlh133 mutant compared to the WT in the presence of Fe (+Fe WT versus bhlh133-orange) and under Fe-deficient conditions (−Fe WT versus bhlh133-dark pink). (ii) Venn diagram showing the number of significantly differentially expressed genes (P < 0.05, PPDE >0.96) in response to Fe-deficient conditions in the bhlh133 mutant (+Fe versus −Fe bhlh133-light pink) and WT (+Fe versus −Fe WT-blue). (b) Heatmap showing the fold-changes in expression, from each of the four comparisons in shown in the Venn diagrams in (a). Genes that were differentially expressed genes in Osbhlh133 mutants, independent of Fe conditions, are indicated in the green boxes. Genes that were differentially expressed only in response to Fe-deficient (−Fe) conditions in the Osbhlh133 mutant plants are indicated in the pink boxes. Significantly over-represented functional categories in these gene-sets [as determined by Pageman statistical analysis, Fisher's test (ORA cut-off value = 1)] are indicated with an asterisk.

In the bhlh133 mutant plants, 377 genes (321 up-regulated and 56 down-regulated genes) were seen to be differentially expressed compared to the WT, independent of Fe conditions (Fig. 7a, orange and dark pink overlap, Supporting Information Table S2). Closer examination of these revealed that the 321 up-regulated genes were over-represented in transcripts encoding proteins across a number of functional categories including major CHO metabolism, lipid metabolism, stress and signalling functions (green boxes; Fig. 7b). Interestingly, these signalling processes largely included transcripts encoding receptor kinases and calcium signalling (green boxes; Fig. 7b), suggesting that the intercellular communication processes are altered in the bhlh133 mutant, independent of Fe conditions. It was also seen that of the 56 down-regulated genes responsive in the mutant, independent of Fe conditions, there was a significant enrichment of transcripts encoding secondary metabolism functions, protein synthesis and elongation functions (green box; Fig. 7b).

Two sets of genes were identified as exclusively responsive to Fe-deficient conditions in the bhlh133 mutant (Fig. 7ai, exclusive genes in dark pink circle; 616 up-regulated genes and 604 down-regulated genes) and light pink circles (Fig. 7aii, 495 up-regulated genes and 706 down-regulated genes, Supporting Information Table S2). The over-represented functional categories are shown for these genes, which were responsive to Fe-deficient conditions only in the bhlh133 mutants (Fig. 7b, pink boxes). This analysis revealed that under Fe-deficient conditions, an up-regulation of genes encoding amino acid metabolism functions, secondary metabolism functions and ammonium transporters is observed only in the bhlh133 mutant (pink box, Fig. 7b), while the down-regulated genes were significantly enriched in photosynthesis-related functions, tetrapyrrole synthesis, ribosomal proteins, as well major CHO synthesis and lipid transfer functions (Fig. 7b, pink boxes, Supporting Information Table S2). In addition, three genes encoding chloroplast-localized ferredoxin and ferredoxin reductases were also seen to be down-regulated exclusively in the bhlh133 mutant in response to Fe-deficient conditions, and this was not seen in the WT (Supporting Information Table S2).

Under Fe-sufficient conditions, some of the genes encoding Fe-related functions were up-regulated in bhlh133 mutant plants (Table 1, Supporting Information Table S2). Namely, Yellow Stripe-Like transporter genes OsYSL2 and OsYSL15; the DMA transporter gene OsTOM1, as well as OsNAAT1, OsNAS1, OsNAS2, OsNAS3 and OsDMAS1, which are essential for DMA biosynthesis. Three metal transporter genes, OsNramp1 (natural resistance-associated macrophage protein) (Takahashi et al. 2011), OsVIT1;2 (ortholog to the Arabidopsis vacuolar iron influxer AtVIT1) (Kim et al. 2006) and the rice citrate transport gene OsFRDL1 (Yokosho et al. 2009), were all also up-regulated (Table 1). Two other transcription factors, known to be associated with transcriptional regulation under Fe-deficient conditions (OsIRO2, OsIRO3, Ogo et al. 2006, 2007; Zheng et al. 2010), were also up-regulated in response to Fe deficiency in both the bhlh133 and WT plants (Fig. 7b; Table 1). Thirteen of the 28 iron-responsive genes (shown in Table 1) that are induced in response to Fe deficiency in the WT and bhlh133 mutant were also induced in the bhlh133 mutant plants under Fe-sufficient conditions. Thus, a Fe-deficiency signal appears to be initiated in the bhlh133 plants (even in the presence of Fe), that is consistent with the lower Fe concentration seen in the bhlh133 roots. Fe deficiency also appeared to result in the significant induction of these listed Fe-responsive genes in both the WT and bhlh133 (Table 1). Notably, the 13 Fe-responsive genes induced in the bhlh133 under Fe-sufficient condition were expressed at a similar level as that present in the WT under Fe-deficient conditions (Table 1).

DISCUSSION

Among all the predicted members of the bHLH-domain containing transcription factors (Li et al. 2006; Pires & Dolan 2010), the bHLH transcription factors of tomato (FER), Arabidopsis (FIT, AtbHLH38, AtbHLH39 and PYE) and rice (OsIRO2 and OsIRO3) have been previously reported to be involved in the Fe-deficiency response and homeostasis (Ling et al. 2002; Colangelo & Guerinot 2004; Ogo et al. 2006; Yuan et al. 2008; Long et al. 2010; Zheng et al. 2010). Similar to these known Fe-responsive bHLH transcription factors, the transcript abundance of OsbHLH133 has also been observed as induced by Fe deficiency (shown in this study and in data from Zheng et al. 2009). Phylogenetic analysis showed that Arabidopsis bHLH transcription factors, AtRHD6 and AtRSL1 were closely related with OsbHLH133 (Supporting Information Fig. S6). In Arabidopsis, AtRHD6 and AtRSL are reportedly known to be involved in root hair formation (Masucci & Schiefelbein 1994; Menand et al. 2007; Yi et al. 2010). However, unlike the root hairless or short root hair phenotype seen in AtRHD6 or AtRSL, rice bhlh133 mutant and the transgenic plants overexpressing OsbHLH133 have normal root hair formation (data not shown). Physiological analysis showed that the alteration of OsbHLH133 expression in the T-DNA insertional mutant plants or overexpressing plants affected the Fe distribution between shoots and roots. As OsbHLH133 is not closely related to other known Fe-responsive transcription factors (Supporting Information Fig. S6), it is likely that bHLH133 is involved in Fe homeostasis by, as yet, an unknown mechanism.

As little as 10 µM EDTA-Fe (II) is sufficient for healthy growth of wild-type rice and when plants were supplied with 10 times more Fe (II) (i.e. 100 µM EDTA-Fe (II)), most of the Fe was retained in the roots, while the shoot Fe concentration was strictly regulated (Zheng et al. 2010). These findings indicate that there is an important role for roots in preventing excess Fe uptake and translocation to shoots. However, the molecular mechanism(s) controlling these processes are largely unknown. In this study, our results from the physiological analysis of OsbHLH133-OE lines and bhlh133 mutant plants indicate the involvement of OsbHLH133 in controlling Fe homeostasis or distribution between roots and shoots. Overexpression of OsbHLH133 resulted in a lower shoot Fe concentration, higher root Fe concentration (Fig. 4a) and hypersensitivity to Fe deficiency (Fig. 3b,c). In contrast, the Fe concentration between leaf and root was altered in the bhlh133 mutant plants, in an opposite manner to that seen in the OsbHLH133-OE lines (Fig. 4b).

The possible role of OsbHLH133 as a negative regulator of Fe translocation from roots to shoots, together with the specific induction of OsbHLH133 seen under Fe deficiency, gives new insights into how plants respond to Fe deficiency at a physiological level. Nutrient deficiency in general, including under Fe deficiency, is known to result in altered root architecture and root hair formation (López-Bucio, Cruz-Ramirez & Herrera-Estrella 2003). These responses occur so that the plant can increase in surface area and explore the soil for additional nutrients. However, root and root hair growth requires Fe (Guerinot & Yi 1994; López-Bucio et al. 2003), and, thus, under limited Fe conditions, the available Fe is retained in the roots to allow plants to respond to Fe deficiency.

Under Fe-sufficient conditions, OsbHLH133 is expressed at a low level in endodermal cells of the maturation zone of roots, but not in root tips and shoots (Fig. 6a–d); supporting the idea that OsbHLH133 might be involved in the regulation of root to shoot long-distance transportation of Fe. Indeed, the Fe concentration in the xylem sap of OsbHLH133-OE lines was also seen to be lower than that in the WT plants (Fig. 5). Under Fe-deficiency conditions, OsbHLH133 was strongly expressed in roots (Fig. 6e–h), not only in endodermal cells, but also in other cells, suggesting that OsbHLH133 may play a role other than in xylem transportation. In addition, in the aerial parts of the plant, OsbHLH133 is only expressed in the vascular tissues under Fe-deficient conditions, suggesting that OsbHLH133 is involved in Fe translocation in Fe-deficient leaves.

To date, a number of transporters have been characterized to be involved in xylem and phloem Fe loading, including the Arabidopsis FRD3, which mediates efflux of citrate into the root vasculature (Durrett, Gassmann & Rogers 2007), the rice FRD3-like protein OsFRDL1 (Yokosho et al. 2009), phenolics efflux zero 1 (PEZ1) (Ishimaru et al. 2011), Arabidopsis ferroportin 1/iron regulated 1 (AtFPN1/AtIREG1) (Morrissey et al. 2009), YSLs, NA, IRT1, NRAMP and a number of others (see review: Guerinot 2010; Kobayashi & Nishizawa 2012). Analysis of global transcript expression changes in rice roots of the bhlh133 mutant plants and the WT transcriptome in the presence and absence of Fe (Fig. 7a) was expected to detect the differential expression in these transporters. However, no significant difference in the expression of the above transporters genes was found between the WT and the mutant. It is possible that there is differential expression of one (or a few) transporter in a specific tissue or cell type; however, the microarray analysis involved the use of full root samples and may have masked the differences in cell type-specific expression. Further analysis using trace Fe analysis could clarify how OsbHLH133 affects Fe distributions.

A number of genes were altered in expression in the rice bhlh133 mutant plants, irrespective of Fe growth conditions, and, among these, a significant enrichment of signalling functions were seen in these up-regulated genes. These up-regulated genes encoding signalling functions, (largely calcium signalling and receptor kinases) and represent a novel list of target genes which might be used to investigate how Fe homeostasis may be controlled in rice (Supporting Information Table S3). Furthermore, the reduction in the expression of genes encoding photosynthesis-related functions in bhlh133 under Fe-deficient growth condition was notable. Plastids and chloroplasts have a high demand for Fe, as co-factors in variety of enzymes; thus, it is envisaged that these functions are limited, and, thus, expression is repressed, in the roots of bhlh133 mutant plants under Fe-deficient conditions. The down-regulation of genes encoding proteins involved in tetrapyrrole metabolism is also consistent with the reduction in transcript abundance for genes encoding photosynthetic genes, and may also play a role in signalling the down-regulation of photosynthetic genes in this system.

In conclusion, this study has identified OsbHLH133 as a regulator of Fe homeostasis in rice, most likely by negatively regulating Fe transportation from root to shoot. It is shown that under Fe deficiency, the roots prioritize the retention of Fe. In addition, alterations in the expression of genes encoding proteins involved in calcium signalling in bhlh133 mutant plants, under both Fe-sufficient and deficient conditions, suggest a role for calcium signalling in root-to-shoot translocation of Fe.

ACKNOWLEDGMENTS

This work was supported by the Sina-Australia Science Cooperation Fund (2010DFA31080), the Key Basic Research Special Foundation of China (2011CB100303), National Natural Science Foundation (31172024), Harvestplus-China (Project 8237), the Ministry of Agriculture (2011ZX08004) and the Government of Zhejiang Province (R3090229). The authors thank Gynheung An who developed the rice T-DNA insertion sequence database from where we obtained the seeds of T-DNA insertion mutant. The authors also thank Dr Jian Feng Ma for the critical reading of the manuscript. Authors have no conflicts of interest to declare.

Ancillary

Advertisement