The dual benefit of a dominant mutation in Arabidopsis IRON DEFICIENCY TOLERANT1 for iron biofortification and heavy metal phytoremediation.

One of the goals of biofortification is to generate iron-enriched crops to combat growth and developmental defects especially iron (Fe) deficiency anaemia. Fe-fortification of food is challenging because soluble Fe is unstable and insoluble Fe is nonbioavailable. Genetic engineering is an alternative approach for Fe-biofortification, but so far strategies to increase Fe content have only encompassed a few genes with limited success. In this study, we demonstrate that the ethyl methanesulfonate (EMS) mutant, iron deficiency tolerant1 (idt1), can accumulate 4-7 times higher amounts of Fe than the wild type in roots, shoots and seeds, and exhibits the metal tolerance and iron accumulation (Metina) phenotype in Arabidopsis. Fe-regulated protein stability and nuclear localisation of the upstream transcriptional regulator bHLH34 were uncovered. The C to T transition mutation resulting in substitution of alanine to valine at amino acid position 320 of bHLH34 (designated as IDT1A320V ) in a conserved motif among mono- and dicots was found to be responsible for a dominant phenotype that possesses constitutive activation of the Fe regulatory pathway. Overexpression of IDT1A320V in Arabidopsis and tobacco led to the Metina phenotype; a phenotype that has escalated specificity towards optimising Fe homeostasis and may be useful in Fe-biofortification. Knowledge of the high tolerance and accumulation of heavy metals of this mutant can aid the development of tools for phytoremediation of contaminants.

Introduction

Micronutrient deficiency causes a significant threat to human health. Fe deficiency is the most prevalent micronutrient deficiency in the world. Fe‐deficiency anaemia affects physical and mental health and the immune system, as well as stunting growth and impairing learning capacity (Black et al., 2008). It has been calculated that over 50% and 80% of pregnant women in developing countries and South Asia, respectively, are at severe risk of anaemia due to low dietary Fe bioavailability, especially where plants are the main source of Fe (WHO/UNICEF, 2017). Fe is one of the most abundant elements in the Earth’s crust but has poor bioavailability for plants in neutral or alkaline soil (Zohlen and Tyler, 2000). Fe food fortification is considered the most practical and sustainable solution to abate Fe deficiency, but it often reacts with food components to cause off‐flavours (Underwood, 2002). Biofortification of plants through genetic engineering to improve Fe concentration is a direct approach that can be used to alleviate Fe deficiency. Genetic engineering has been applied to modify Fe transporters to generate Fe‐enriched crops (Kumar et al., 2019; Masuda et al., 2013; Narayanan et al., 2019).

Due to the diverse nature of soil, concentrations of micronutrients and heavy metals can vary substantially (Tripathi et al., 2015). Under Fe‐limiting conditions, the broad substrate‐specific Fe transporter iron‐regulated transporter1 (IRT1) is responsible for the uptake of mineral pollutants such as Cd, Ni, Cu, Zn as well as Fe (Al Khateeb and Al‐Qwasemeh, 2014; Morrissey and Guerinot, 2009). Some of these heavy metals can inhibit plant growth and affect Fe uptake and homeostasis, and frequently accumulates in crops (Toppi and Gabbrielli, 1999). This limits the Fe‐biofortification efficiency in plants. Hence, it is imperative to obtain insight into Fe uptake, translocation, storage machinery and IRT1 nonspecificity for successful implementation of phytoremediation strategies in unfavourable soil conditions. Plants have acquired several strategies to uptake Fe from soils. Graminaceous species have evolved a chelation‐based strategy where Fe3+‐phytosiderophore complex is transported via an oligopeptide of the YSL family (Curie et al., 2009). Arabidopsis and flowering species follow a reduction‐based Fe uptake mechanism, in which insoluble Fe3+ is reduced by the ferric reduction oxidase 2 (FRO2) to Fe2+ and the reduced Fe2+ is subsequently transported across the root by IRT1 (Robinson et al., 1999; Vert, 2002). Recalcitrant Fe is solubilised by the H+‐ATPase, AHA2 (Santi and Schmidt, 2009). In Arabidopsis, Fer‐like iron deficiency‐induced transcription factor (FIT) and POPEYE (PYE), bHLH‐type transcription factors regulate different subsets of genes involved in Fe acquisition and cellular homeostasis (Colangelo and Guerinot, 2004; Long et al., 2010). FIT regulates downstream genes by forming heterodimers with bHLH38, bHLH39, bHLH100 and bHLH101 (Naranjo‐Arcos et al., 2017). Both FIT and PYE are directly activated by bHLH34, bHLH104, bHLH115 and bHLH105 (ILR3) (Li et al., 2016; Zhang et al., 2015). The abundance of bHLH104 and bHLH105 is regulated by E3‐ligase BTS (BRUTUS) (Selote et al., 2015).

In the current study, to enhance Fe‐biofortification and attain phytoremediation of nonessential metals, we targeted the upstream regulation of IRT1 in Arabidopsis. EMS mutagenesis of the P:Luc reporter line was used to identify novel regulators in the Fe signalling pathway, which can help plants to withstand imbalanced homeostasis, and the idt1 mutant was identified. Overexpression of this mutated allele in Arabidopsis and tobacco showed the metal tolerance and iron accumulation (Metina) phenotype with constitutive activation of Fe‐deficiency response. These results suggest the usefulness of this allele for phytoremediation and Fe‐biofortification.

Results

Screening and identification of Fe‐specific EMS mutant idt1

Forward genetics was adopted to identify novel regulators in the Fe signalling pathway, which may help plants to maintain Fe homeostasis. First, the P:LUC reporter line (WT), which shows luminescence in roots under Fe‐deficient conditions, was constructed. Screening of 30 500 mutagenized seeds of the reporter line under Fe‐sufficient conditions revealed that iron deficiency tolerant1 (idt1) showed constant luminescence (Figure 1a), and there was constitutive IRT1 protein expression in idt1 and the 3×‐backcrossed line idt1‐1 (Figure 1b). Root ferric‐chelate reductase (FCR) activity was observed to be constitutively active in idt1‐1 (Figure 1c). These data suggested that the mutation affected the upstream Fe regulatory system. Constitutive strategy‐I response enhanced Fe uptake and led to substantially higher Fe accumulation in idt1‐1 embryos than in the WT (Figure 1d). Elemental analysis showed 2.31 times higher Fe accumulation in mutant seeds than the WT (Figure 1e). Due to elevated Fe content, the mutant exhibited Fe‐deficiency tolerant phenotype (Figure 1f) with enhanced fresh shoot weight (Figure 1g) and significantly higher chlorophyll content under Fe deficiency (Figure 1h). Physiological and molecular observation of idt1‐1 and idt1 outcross with Landsberg erecta (Ler) revealed the dominant nature of the mutation. Mapping by sequencing identified the recombinant region on AT3G23210 (IDT1, bHLH34) where the amino acid alanine residue was substituted by valine (IDT1A320V). The region of point mutation along with PPVA motif was conserved among agronomic crops, and phylogenetic analysis shows evolutionary relationships of this gene among different species including mono‐ and dicots (Figure S1).

Figure 1

Screening and identification of Fe‐specific EMS mutant idt1. (a) Observation of luminescence using an IVIS Lumina system in 7‐day‐old WT and idt1 grown on Fe10 transferred to Fe10 or –Fe for 2 days. (b) IRT1 protein analysis in (a). (c) Ferric‐chelate reductase (FCR) activity of 5‐day‐old WT and idt1‐1 on Fe10 and transferred to Fe10 or −Fe for 2 days, n = 3 (8 roots/replicate). (d) Perl staining of embryo WT and idt1‐1. (e) Fe concentration analysis of seeds collected from soil‐grown WT and idt1‐1, n = 3 (5 mg seeds/replicate). (f) Phenotypic analysis of 7‐day‐old WT and idt1 seedlings grown on Fe10 transferred to Fe10 or –Fe for 6 days. (g) Fresh shoot weight of (f). (h) Chlorophyll content of (f) n = 3 (eight plants/replicate). The data represent the means ± SD from three independent replicates. Statistical tests were performed with two‐sided student t‐test related to WT. *P < 0.05, **P < 0.005, Scale bar 1 cm in (f). WT (wild‐type reporter line); Fe10 (½ Hoagland with 10 µM Fe); −Fe (Fe0 + 30 µm Ferrozine).

The idt1‐1 exhibits the Metina phenotype

The transgenic lines solely overexpressing the main Fe transporter IRT1 accumulate higher levels of other heavy metals than wild‐type plants with accompanying toxic effects (Connolly, 2002). In this study, the toxicity assessment of excess Fe, Zn, Cd, Ni, Cu, Co, Pb and −Fe showed that idt1‐1 exhibits the Metina phenotype with elevated fresh shoot weight and high chlorophyll content compared with the wild type (Figure 2a–c) despite constitutive IRT1 expression in idt1‐1. Elemental analysis showed only higher Fe concentration with no significant difference in Zn and Mn and 3–6 times higher Fe accumulation in idt1‐1 under different heavy metal treatments than in WT (Figure 2d–f). The mutant could tolerate the accumulation of 1.1–2 times higher Zn, Cd and Ni in shoots than the wild type (Figure 2g). This suggests that the specificity of IDT1A320V for Fe homeostasis in addition to IRT1 up‐regulation can also contribute to heavy metal tolerance.

Figure 2

The idt1 exhibits the Metina phenotype. (a) Seven‐day‐old WT and idt1‐1 seedlings grown on Fe10 and transferred to Fe50, –Fe, Zn100, Cd30, Ni20, Cu20, Co20 and Pb20 for 6 days. (b) Fresh shoot weight. (c) Chlorophyll analysis of shoots (a) n = 3 (six plants/replicate). (d) Metal concentration in 9‐day‐old WT and idt1‐1 seedlings grown on Fe10 and transferred to Fe10 for 7 days. (e, f) Fe concentration analysis in shoots and roots 9‐day‐old WT and idt1‐1 seedlings grown on Fe10 and transferred to Fe10, Fe50, –Fe, Zn100, Cd30 and Ni20 for 7 days. (g) Metal concentration in (e, f). n = 3, 20 individual plants/replicate. The data represents the means ± SD from three independent replicates. Statistical tests were performed with two‐sided student t‐test related to WT. *P < 0.05, **P < 0.005, Scale bar 1 cm in (a). WT (wild‐type reporter line); Fe10 (½ Hoagland 10 µm Fe); Fe50 (50 µm Fe‐EDTA), −Fe (Fe0 + 30 µm Ferrozine); 100Zn (100 µm ZnSO4); Cd30 (30 µm CdSO4) and Ni20 (20 µm NiCl2); Cu20 (½ Hoagland 20 µm CuSO4); Co20 (½ Hoagland 20 µm CoCl2) and Pb20 (½ Hoagland 20 µm PbNO3).

Constitutive Fe‐deficiency response is observed in idt1‐1 and IDT1 overexpression lines

Arabidopsis transgenic lines overexpressing N‐terminus GFP fusion of IDT1A320V and wild‐type IDT1 driven by the ubiquitin‐10 promoter (PUB), P and P, were conducted for further study. By screening T1 (n = 15) to select T3 (n = 5) of both P and wild‐type P lines, we found constitutive IRT1 expression under Fe sufficiency in P lines as in the idt1‐1 mutant, but not in P lines (Figure 3a,b). Transgenic lines of both constructs with similar IDT1 expression were selected for further investigation (Figure 3b). The majority of Fe deficiency‐induced genes involved in acquisition (IRT1, FRO2, AHA2, FIT and FIT partners bHLH‐Ib group), translocation (PYE, BTS, NAS4 and FRD3), storage (FER1) and IMA1 were all found to be expressed at high levels in idt1‐1 and P selected lines under Fe‐sufficient conditions (Figure 3c). These results confirmed that this unique mutation produced a strong allele of IDT1 and activate the overall Fe regulatory machinery.

Figure 3

Constitutive Fe‐deficiency response is observed in idt1‐1 and IDT1 overexpression lines. (a) IRT1 protein expression in roots of WT, idt1‐1 and transgenic OE lines P and Pgrown under Fe10 condition for 9 days. (b) Quantification of IRT1 protein band intensity with ImageJ software and compared similar root IDT1 gene expression of transgenic lines. (c) qRT‐PCR of Fe deficiency‐induced genes found to be up‐regulated in microarray analysis under Fe10 conditions for 9 days, n = 3 biological repeats (eight roots/replicate). The data represents the means ± SD from three independent replicates. Statistical tests were performed with two‐sided student t‐test related to WT. *P < 0.05, **P < 0.005. WT (wild‐type reporter line).

Overexpression of IDT1 results in the Metina phenotype

Analyses under different metal treatments showed that P lines also exhibited the Metina phenotype both in the media and soil like the idt1‐1 mutant (Figure 4a,b) with high fresh weight and chlorophyll content (Figure S2a,b). Similar results were obtained with Arabidopsis transgenic lines overexpressing N‐terminus HA fusion of IDT1A320V and wild‐type IDT1 driven by the 35S promoter, 35S:HA‐IDT1 and 35S:HA‐IDT1/WT, suggesting that the mutation has efficient heavy metal tolerance and Fe accumulation (Figure S3a–c). Elemental analysis revealed 3–6 times higher Fe under normal and alkaline soil conditions (Figure 4c), about five times higher Fe in seeds of P (Figure 4d) and higher Fe concentration in excess Zn and Cd‐treated soil plants than the WT (Figure 4e). In addition, P could also accumulate more Cd (Figure 4f) and Zn than the WT (Figure 4g).

Figure 4

Overexpression of IDT1 in WT results in the Metina phenotype. (a) Phenotype of 7‐day‐old WT and idt1‐1 and GFP‐IDT1 OE lines grown on Fe10 and transferred to –Fe, Fe50, Zn100, Cd10 for 6 days. (b) Phenotype of 9‐day‐old WT and idt1‐1 seedlings grown on Fe10 and transferred to normal soil, alkaline soil and soil watered with 200 µm Fe‐EDDHA (Fe200), 500 µm ZnSO4 (Zn500) and 50 µm CdSO4 (Cd50) for 15 days. (c) Fe content analysis of 9‐day‐old seedlings transferred to normal or alkaline soil for 15 days. (d) Seed Fe concentration in WT, idt1‐1 and mutant OE line #5. (e) Fe (f) Cd (g) Zn, concentration analysis of (b). The data represents the means ± SD from three independent replicates. Statistical tests were performed with two‐sided student t‐test related to WT. *P < 0.05, **P < 0.005.

Overexpression of IDT1 results in the Metina phenotype in tobacco

Overexpression of idt1 dominant mutation in Nicotiana tabacum cultivar W38 resulted in the Metina phenotype with higher biomass and chlorophyll content under Fe deficiency and excess cadmium (Figure 5a,b). More than fifteen T1 lines were screened by antibiotic selection, and 7 T2 lines had high Fe accumulation in shoots and fruits (Figure S4). Soil experiments revealed similar phenotypic results with loss of sensitivity towards alkaline soil and cadmium toxicity in GFP:IDT1 lines (Figure 6a,b). In addition, shoot elemental analysis of three selected transgenic lines showed higher Fe accumulation under normal soil and 3–4 times more cadmium accumulation (Figure 6c,d). This result implies that this allele is functional among higher species and could potentially be useful as a source of phytoremediation and Fe‐biofortification.

Figure 5

Overexpression of IDT1 results in the Metina phenotype in tobacco. (a) Phenotype of 15‐day‐old P:GFP/W38 (GFP) #3, P (GFP‐IDT1) #25 and P/W38 (GFP‐IDT1) #5 transgenic lines grown on Fe10 and transferred to Fe10, –Fe, Fe50, Cd30 for 9 days. (b) Total biomass (whole plant) and chlorophyll content analysis of shoots (a). The data represent the means ± SD from three independent repeats. Statistical tests were performed with two‐sided student t‐test related to P. *P < 0.05, **P < 0.005.

Figure 6

Overexpression of IDT1 led to the Metina phenotype in tobacco in soil. (a) Phenotype of 15‐day‐old (GFP), P (GFP‐IDT1) and P/W38 (GFP‐IDT1) transgenic lines were transferred to soil containing normal soil, alkaline soil and soil watered with 500 µm Fe‐EDTA (Fe500), 100 µm CdSO4 (Cd100) for 15 days. (b) Total biomass and chlorophyll content among different treatments of (a). (c) Fe content analysis of shoots (a) under normal soil conditions. (d) Cd accumulation measurement in the shoot of (a) plants treated with excess cadmium. The data represent the means ± SD from three independent repeats. Statistical test was performed with two‐sided student t‐test related to GFP. *P < 0.05, **P < 0.005.

High IDT1 accumulation and stability is positively correlated with the Metina phenotype

The GFP‐IDT1A320V protein showed constitutive expression and enhanced accumulation in nuclei of transfected protoplast cells (Figure S5a) with significantly higher nucleus/cytoplasmic ratio of GFP‐signal intensity (Figure S5b). The anti‐GFP‐specific immunoblot showed higher accumulation of GFP‐IDT1A320V than wild‐type GFP‐IDT1 (Figure S5c). Similar results were obtained with stable Arabidopsis transgenic lines showing constitutive GFP signal in the nucleus of root tissues of P/WT (Figure S6a,b). Interestingly, the accumulation and nuclear localisation of wild‐type GFP‐IDT1 were highly induced by Fe deficiency only (Figure S6a–c). These data suggested that the nuclear specificity may be due to protein stability or property changes for nuclear localisation. CHX‐protein stability assay showed that the degradation rate of IDT1 protein was lower in idt1‐1 (T 1/2 = 3.3 h) than WT (T 1/2 = 1.2 h) (Figure S6d,e). Taken together, the high stability and nuclear accumulation of IDT1A320V were able to contribute to the constitutive up‐regulation of downstream Fe‐deficiency response, which corresponds to heavy metal tolerance.

Discussion

In this study, we report a dominant mutant idt1 with constitutive expression of Fe‐deficiency pathway genes enabling Fe homeostasis and leading to high accumulation of Fe in roots, shoots and seeds (Figures 1, 2, 3). Its Fe homeostatic specificity resulted in high tolerance under Fe deficiency and excess heavy metal exposure (Figures 1 and 2). Several studies have pointed out that Fe homeostasis can be jeopardised by heavy metal stresses (Leskova et al., 2017; Shanmugam et al., 2013). Intriguingly, the stability of the major Fe transporter in Arabidopsis, IRT1, is regulated by heavy metals (Dubeaux et al., 2018). Here, using a genetic approach, a strong bHLH34 allele IDT1 that turns on downstream Fe homeostatic genes was revealed. The bHLH34 belongs to bHLH‐IVc group upstream of the Fe signalling pathway. Transgenic lines overexpressing IDT1 led to the Metina phenotype in Arabidopsis (Figure 4 and Figure S3) (Shanmugam et al., 2011; Vert, 2002) suggesting the central role of this gene in controlling Fe homeostasis and usage of this allele in Fe‐biofortification and phytoremediation. The mutation site is highly conserved among agronomic crops (Figure S1), hence introducing the idt1 mutation in tobacco resulted in the Metina phenotype (Figure 5 and Figure S4). Transgenic tobacco overexpressing IDT1 accumulated high Fe and tolerated Fe deficiency, and Cd toxicity implying that the regulation of Fe homeostasis is similar among different species. This suggests that this allele might work in other higher edible crops as well and can enhance Fe accumulation. Therefore, the current result may well be valuable for challenging Fe deficiency. However, the regulation of Fe homeostasis may be different in monocots or other plants. It will be worth further investigating the feasibility of the use of this novel allele in other target plant systems.

The production of nutritious, sufficient and safe food is a key objective of sustainable agriculture to feed the undernourished among the world’s population and includes crops such as Golden rice (Tang et al., 2009) to prevent vitamin‐A deficiency and iron‐biofortification of beans, sweet potato, legumes and cassava to combat the hidden hunger in developing countries (Garg et al., 2018). A biofortification study on cassava engineered Fe transporter IRT1 and ferritin to co‐express to enhance Fe levels (Narayanan et al., 2019). Other studies have manipulated components of the Fe homeostasis response, such as bHLH‐IVc (bHLH104/115/ILR3) and the downstream bHLH‐1b (bHLH38/39/100/101), where overexpression of bHLH104 was studied for Cd tolerance by expressing metal detoxification‐associated genes, while high expression of bHLH104 transgene enhanced sensitivity to Fe deficiency (Yao et al., 2018). Co‐expression of FIT/bHLH38 and FIT/bHLH39 showed tolerance to Cd via increasing its sequestration but not in individual bHLH38 and bHLH39 overexpression (Wu et al., 2012). Co‐expression of NICOTIANAMINE SYNTHASE (NAS) and FERRITIN (FER) increased the rice Fe content specifically in endosperms, while solely expressing NAS or ferritin alone did not work (Wirth et al., 2009). Elevated accumulation of nicotianamine enhanced sensitivity to Fe deficiency but could tolerate Ni toxicity (Klatte et al., 2009). Transgenic plants overexpressing IRON MAN 1 (IMA1) showed necrotic spots on the leaves indicating sensitivity to both Fe‐sufficient and Fe‐deficient conditions (Grillet et al., 2018). Hence, targeting single genes for genetic engineering appears to be an unfavourable approach for biofortification. In contrast, IDT1A320V has an upstream and central role in Fe homeostasis and it regulates both FIT‐mediated (IRT1 and FRO2) and PYE‐mediated (NAS4 and FRD3) pathways. In addition, IMA1 and FER1 are highly up‐regulated in idt1‐1. IDT1A320V regulates downstream pathways in a coordinated manner from Fe uptake to storage, which enhances Fe homeostasis to combat Fe deficiency and heavy metal toxicity.

The role of bHLH34 in the Fe signalling mechanism has not been studied comprehensively despite it being a homolog of bHLH104 and forming heterodimers with other members of the bHLH‐IVc group. It also acts independently (Gao et al., 2019). Genetic engineering targeting genes upstream of the Fe signalling pathway could be favoured as they alter the endogenous machinery allowing plants respond to imbalanced homeostasis in a coordinated manner. This unique mutation enhanced IDT1A320V protein accumulation in the protein stability possibly by making a stable interaction with bHLH‐IVc members in the nucleus to activate Fe homeostasis genes, which are directly related to the Metina phenotype (Figure S5 and Figure 6).

Incidental effects on unwanted metals can also be beneficial. For example, transgenic tobacco with enhanced Fe accumulation can show tolerance to high levels of nickel (Douchkov et al., 2005). Similarly, IDT1 overexpressing Arabidopsis and tobacco lines are more tolerant of the toxicity of heavy metals (Figure 4) than WT‐like idt1‐1. They also exhibit prolonged survival as in the case of hyperaccumulator plants which absorb large amounts of metals and can survive under harsh conditions in metalliferous soils. Natural hyperaccumulators have low biomass when they absorb high amounts of heavy metals unlike ‘engineered hyperaccumulators’, which tend to possess higher biomass with enhanced accumulation ability and potential for phytoremediation (Peer et al., 2005). The high specificity of idt1 towards Fe homeostasis limits the competition of other heavy metals and enhances tolerance under Fe‐limiting soil and heavy metal contaminated soil. Scientists are still struggling to integrate biofortification, green manuring and phytoremediation for sustainable agriculture. In this study, we report not only the mechanistic role of IDT1 in Fe homeostasis signalling, but also the 3‐in‐1 action of idt1 which can be utilised to mitigate Fe deficiency by Fe‐fortification, green manuring of Fe‐laden plants and to enhance plants potential to survive under heavy metal toxicity thus affirming it for phytoremediation.

Methods

Plant materials and growth conditions

Seeds were surface sterilized and, after a 3‐day stratification, were sown on ½ Hoagland medium (pH 5.75) with 1% sucrose (J. T. Baker), 0.5 g/L of MES (J.T. Baker) and 0.7% agar (type A, Sigma Aldrich). For Fe10 [10 µm Fe (II) EDTA], −Fe (30 µm Ferrozine to 0 Fe medium), Fe50 [50 µm Fe (II) EDTA], Zn100 (100 µm ZnSO4), Cd10/30 (10/30 µm CdSO4), Cu20 (20 µm CuSO4), Co20 (20 µm CoCl2) and Ni20 (20 µm NiCl2) and Pb10 (10 µm PbNO3). For soil experiments, 1% lime was added to prepare alkaline soil. For heavy metal treatments, soil was watered two times a week with 200 µm Fe‐EDDHA (Fe200), 500 µm ZnSO4 (Zn500) or 50 µm CdSO4 (Cd50). Plants were grown under 70–100 µmol photons/m2/s, 16 h light/8 h dark cycle at 22 °C.

Generation of transgenic constructs and plant transformation

The P:LUC reporter line (Kailasam et al., 2018) homozygous seeds were mutagenized by using 20 mm ethylmethanesulfonate (EMS) as described in Arabidopsis protocols (Salinas and Sanchez‐Serrano, 2006). Approximately 30 500 M2 seeds were screened under ½ Hoagland medium with 10 μm Fe (II) EDTA (Fe10) to observe root luciferase activity, and luminescence imaging was performed by spraying 0.5 mm luciferin along with 0.01% triton X‐100 and incubated for 5 min in the dark. Root luminescence was imaged using an IVIS Lumina imaging system (Xenogen, USA) with exposure time of 1 min. The overexpression lines were generated by amplifying the coding sequence of IDT1 and IDT1 using gene‐specific primers (Table S1) and cloned into entry vector pCR™8/GW/TOPO and expressed into PUB‐GFP and pEG201 by gateway cloning (Invitrogen). Agrobacterium tumefaciens strain C58C1 was used for in planta transformation. P and 35S:HA‐IDT1, 35S:HA‐IDT1 were introduced in WT plants using the floral dip method (Clough and Bent, 1998). P and P were introduced into N. tabacum cultivar W38 explant generated from leaf disc. The resulting T0–T2 transgenic lines were screened on selection media containing 25 mg/L Basta. The P reporter line is referred to as wild type (WT) for simplicity.

Genetic mapping of idt1

A plant mapping population was built by crossing homozygous idt1 and L. erecta, according to the MutMap protocol (Lehrbach et al., 2017). Tolerant phenotypes of the F2 population were observed under –Fe suggesting the trait is dominant. To obtain a homozygous line for gene mapping, the F2 population was re‐screened in F3, and genomic DNA was extracted by DNeasy plant mini kit (Qiagen, Hilden, Germany) from 60 homozygous F2 plants, pooled and sequenced using Illumina HiSeqTM 2500 sequencer. Raw reads were processed by CLC Genomics Workbench 7 (CLC bio). Assemble reads were mapped to the Col‐0 (Tair10) genome, and SNP calling was performed in PuTTY platform via SAMtools (Li et al., 2009). To identify and evaluate the recombinant region linked to the mutant phenotype, a pipeline using VCFtools (Danecek et al., 2011), SNPeff and SNPshif (Cingolani et al., 2012) was used. NGM software (http://bar.utoronto.ca/ngm) was used to visualise recombinant regions.

Determination of biomass and chlorophyll content

Fresh shoot weight of seedlings was weighed immediately after harvest and pooling 8 shoots/technical repeats for Arabidopsis and three seedlings/technical repeats for tobacco. Total chlorophyll content was determined as described previously (Kailasam et al., 2018).

Gene expression and analysis

Total RNA was extracted using GeneDireX RNA isolation kit from 9‐day‐old roots grown under Fe10 conditions. The RT‐qPCR was performed with Applied Biosystems 7500 real‐time PCR system using SYBR Green detection method and gene‐specific primers (Table S1). Transcript level was normalised to the internal control UBC21 using the relative standard‐melt curve method.

Antibody generation and immunoblot analysis

Total protein was extracted, and the anti‐IRT1‐specific antibody was developed in a previous study (Shanmugam et al., 2011). Immunoblots of GFP constructs were performed using commercially raised anti‐GFP antibody (Roche). IDT1‐specific antibody was generated by LTK BioLaboratories according to Arabidopsis thaliana peptide sequence PLPPADRDTSRDLKNLPPC indicated (Figure 1k). Protein (10–20 µg) was separated on 4–12% NuPAGE gel (Invitrogen, Carlsbad, CA) and transferred on polyvinylidene difluoride (PVDF) membrane and blocked by 5% fat‐free milk dissolved in 1% PBST (1× PBS + 2% tween 20) for 45 min, followed by incubation with 1:10 000 αIRT1, 1:1000 αGFP or 1:5000 αIDT1 primary antibody for 1–2 h, washed with 1× PBST, further incubation with 1:20,000‐diluted secondary antibody (HRP‐conjugated anti‐mouse for αIRT1 and αGFP whereas anti‐rabbit for αIDT1) for 1 h. After washing, specific protein bands were visualised by Immobilon Western Chemiluminescence HRP Substrate (Millipore, Burlington, MA) with X‐ray film and later quantified using ImageJ software.

Ferric‐chelate reductase activity

Ferric‐chelate reductase activity was detected as described in a previous study (Grillet et al., 2014) using roots from 5–7 seedlings grown for 5 days under Fe10 and −Fe. Seedlings were incubated in the dark for 1‐hour with mild shaking. Final concentration was measured by reading the absorbance at 535 nm on a Synergy H1 plate reader (BioTek, Winooski, VT).

Measurement of mineral concentration

Elemental analysis was conducted as previously described (Shanmugam et al., 2011). Harvested samples were washed with 10 mm CaCl2, thoroughly washed with water and dried for 72 h at 70 °C predigestion. Microwave‐digested material was analysed by an ICP‐OES (Optima 5300; Perkin‐Elmer, Waltham, MA) and ICP‐MS (Agilent 7800; Agilent, Santa Clara, CA).

Histochemical staining of Fe

Arabidopsis seedlings grown under Fe10 were vacuum infiltrated using Perl’s solution (2% HCl and 2% potassium ferrocyanide) for 15–30 min. Samples were washed 2–3 times with distilled water. For embryo staining, Perl’s staining was intensified with diaminobenzidine DAB as described (Grillet et al., 2018).

Fluorescence imaging in protoplast and stable transgenic lines

Protoplast preparation and transfection were performed as described (Wu et al., 2009). Protoplasts were transfected with GFP tagged constructs of IDT1 and IDT1. For nuclear localisation, nRFP‐VirD2NLS construct was co‐transfected (Wu et al., 2009). Florescence emission of GFP and RFP in protoplasts was observed under a confocal microscope (Zeiss LSM780 META). Transgenic lines with IDT1 and IDT1 overexpression in WT were grown under Fe10 for 7 days and transferred to Fe10 and −Fe for 2 days, and GFP signals were analysed in different root parts. GFP mean intensity was quantified using Zen 2011 software (Carl Zeiss Microscopy GmbH, Jena, Germany).

CHX chemical treatment

Sigma‐Aldrich (St. Louis, MI) CHX stock solution (200 mm) was prepared in DMSO. For control experiments, 0.1% DMSO solution was used. The experiment was conducted as described in a previous study (Liu et al., 2012).

Competing interest

There is no conflict of interest.

Author contribution

KCY conceived the research, interpreted data and was responsible for financial funding. RS performed experiments and data analysis. RS and KCY composed the manuscript.

Figure S1 Multiple sequence alignment and phylogenetic tree for IDT1 and its homologs in agronomic species.

Figure S2 IDT1 overexpression lines possess high fresh shoot weight and chlorophyll content.

Figure S3 35S:HA‐IDT1 overexpression lines show the Metina phenotype.

Figure S4 Overexpression of IDT1 results in high Fe accumulation in tobacco leaf and fruit.

Figure S5 Protoplast expression assay shows more IDT1 accumulation.

Figure S6 IDT1 enhances protein accumulation and localisation in nuclei.

Click here for additional data file.

Table S1 Primer list used in the experiments.

Click here for additional data file.

Click here for additional data file.