Brassica napus cultivar Zhongshuang NO.9, Arabidopsis wild type (Col-0) and meb2 mutant (GK-059D30, a T-DNA insertion in the second exon of MEB2) were used in this study. Seedlings of rapeseed cultivar Zhongshuang NO.9 were grown in 10 l boxes containing Hoagland nutrient solution in a greenhouse with 25°C light/22°C dark cycles under normal light conditions. The Arabidopsis wild type (Col-0), meb2 and transgenic lines were grown on Murashige and Skoog (MS) plates containing 1% sucrose and 0.8% agar. Seedlings were grown under a 16-h-light/8-h-dark condition at 22°C in a growth chamber with light.

The B. napus genome was downloaded from the Genoscope Genome Database¹ (Chalhoub et al., 2014). To identify homologs in B. napus, AtMEB2 (At5g24290) was used as initial protein queries to search against Genoscope Genome Database based on BLASTP with the score value of ≥100 and e-value ≤ e^-10. Conserved VIT domain in the candidate sequences were detected by Pfam database² on line. At last, two candidate genes were identified such as BnaC07g30170D and BnaA06g26800D. According to the sequence information of these two genes, specific primers (Forward 5′-ATGGACCGACCAGCTGACG-3′ and Reverse 5′-TTAAAGGGAAGTAAACCGGTATTC-3′) were designed with the Primer Premier 5.0 (PREMIER Biosoft International, USA). The coding sequence (CDS) was amplified using B. napus cultivar Zhongshuang NO.9 cDNA as the template and sequenced. To illustrate the structure of intron, exon and VIT domain of the candidate genes, fancygene database³ was used to draw the gene structure according to the full-length genome from Genoscope Genome Database and coding sequence from sequencing. The sequences of candidate genes were aligned by using Clustal W (Thompson et al., 1994; Rambaldi and Ciccarelli, 2009). The 3D protein structure model of the homologs regions was predicted by SWISS-MODEL and was analyzed by using SPDBV4.10 (Guex and Peitsch, 1997).

Cellular localization of the BnMEB2 protein was examined by transient expression of a BnMEB2::GFP fusion protein. The open reading frames (ORFs) of BnMEB2 without a stop codon were amplified by using the following primers (Forward 5′-GACTGGTACGAGCTCGGTACCATGGACCGACCAGCTGACG-3′; Reverse 5′-GCCCTTGCTCACCATGGATCCAAGGGAAGTAAACCGGTATTCCTC-3′). The underlined sections in primer sequences indicated KpnI and BamHI restriction sites respectively. PCR products were recombined into pUC19-35Spro::GFP (previously modified and constructed by our lab) to generate the expression vector 35Spro::BnMEB2::GFP by using ClonExpress MultiS One Step Cloning Kit (Vazyme). The detailed information about these vectors was shown in the Supplementary Figure S1. The resulted constructs together with the positive control 35Spro::GFP were transiently expressed in Arabidopsis protoplasts via polyethylene glycol-mediated transformation (Yoo et al., 2007). Furthermore, the plasmid constructs were also delivered into onion epidermal cells for transiently expression by DNA particle bombardment using the PDS-1000/He Biolistic Particle Delivery System (Bio-Rad) as described by Xie et al. (2014). GFP fluorescence was then observed with a confocal microscope (Nikon A1) at 16 h after transfection.

Due to the high similarity between BnaC07g30170D and BnaA06g26800D, only CDS fragment of BnaC07g30170D was constructed into pBWA(V)BS vector⁴ and verified by sequencing to create a 35Spro::BnMEB2 over-expression vector, in which target gene expression was under the control of CaMV 35S promoter. The full-length cDNAs of BnMEB2 were amplified with the following primers (Forward 5′-CGGAATTCATGGACCGACCAGCTGACG-3′ and Reverse 5′-CGGGATCCTTAAAGGGAAGTAAACCGGTATTC-3′). The underlined sections in primer sequences indicated EcoRI and BamHI restriction sites respectively. Plasmid of 35S::BnMEB2 was introduced into Agrobacterium tumefaciens strain GV3101 by electroporation and was transformed into Arabidopsis wild type (col-0) and meb2 mutant by floral dip method (Clough and Bent, 1998). Transgenic plants were selected by herbicide spraying. BAR-resistant and PCR-positive transgenic plants were transferred to a greenhouse and maintained up to T2 generation. Transgenic lines with high expression level of BnMEB2 were selected and used for further analysis.

Total RNA for quantitative RT-PCR was extracted from the rapeseed and Arabidopsis samples using the RNeasy Plant Minikit (Qiagen, USA) and was reverse transcribed with SuperScript^TM III reverse transcriptase (Invitrogen) following the manufacturer’s instructions. qRT-PCR was performed by using SYBR Green Real-time PCR Master Mix (Bio-Rad, USA) in 20 μl reaction mixture and run on CFX96 Real-time PCR system (Bio-Rad). Oilseed rape and Arabidopsis β-actin genes (AF111812, AT1G13180) were used as internal standard. All assays for target genes were conducted with three biological repeats, each with three technical repeats. Primers used in qRT-PCR are listed in Supplementary Table S1. The quantification of threshold cycle (CT) value analysis was achieved using the 2^(-ΔCT) method (Livak and Schmittgen, 2001).

Seeds of B. napus cultivar Zhongshuang NO.9 were germinated and grown in 1/2 Hoagland nutrient solution. At the eighth day, one half of the seedlings were transferred to Hoagland nutrient solution while the other half was transferred to Hoagland nutrient solution with 0.5 mM FeSO₄ for high iron concentration treatments. Roots and leaves were collected at 1, 2, and 3 days after transferred under the normal and high iron concentration conditions and immediately frozen in liquid nitrogen. Total RNA was isolated to determine the mRNA level of BnMEB2. Moreover, at least three replicates of Arabidopsis transgenic lines (35S::BnMEB2/WT and 35S::BnMEB2/meb2), wild type (col-0) and meb2 were grown on MS medium with a series of high iron concentration treatments (0.2-1 mM FeSO₄) for 20 days. These seedlings were collected and their total RNAs were isolated to detect the expression level of other key VIT genes.

Arabidopsis roots were stained with Perls’ solution according to previous study (Green and Rogers, 2004). Equal volume of 4% (v/v) HCl and 4% (w/v) potassium ferrocyanide were mixed immediately prior to use. The mixed solution was vacuum infiltrated into 10-day-old Arabidopsis seedlings for approximately 15 min. Seedlings were rinsed in ultrapure water for three times. The Perls’ staining samples were observed immediately in whole roots using an Olympus IX71 microscope. The method for iron quantification was minor modified from those described by Cassin et al. (2009) and Li et al. (2013). Harvested seedlings were washed for 5 min in a solution containing 5 mM CaSO₄ and 10 mM EDTA, then dried at 70°C for 3 days and weighted. Approximately 300 mg of dry tissue was digested completely in 10 mL 70% HNO₃ at 200°C for 2 h. The iron contents in these samples were measured by inductively coupled plasma spectrometry SPS3000 (Seiko, Tokyo, Japan) and conducted with three biological repeats.

To identify the homologs in B. napus, AtMEB2 was used as protein queries for BLASTP and two copies (BnaC07g30170D and BnaA06g26800D) were identified from B. napus genome. Based on the information in Genoscope Genome Database, full-length cDNA fragments in rapeseed cultivar Zhongshuang NO.9 were amplified by high-fidelity RT-PCR. BnaC07g30170D locates at C07 chromosome, which consists of six exons and five introns. BnaA06g26800D locates at A06 chromosome, which consists of seven exons and six introns. More importantly, same as the well-known VIT family members in Arabidopsis such as AtVIT1 and AtMEB2, BnaC07g30170D and BnaA06g26800D also contained VIT domain. Interestingly, VIT domains of AtVIT1 and AtMEB2 were separated by three or two introns but were fully integrated in the last exon of BnaC07g30170D and BnaA06g26800D (Figure 1A). BnaC07g30170D and BnaA06g26800D encode putative protein with 534 and 537 amino acids respectively, and show a high similarity with AtMEB2. Especially, VIT domain shared over 85% sequence similarity among AtMEB2, BnaC07g30170D and BnaA06g26800D (Figure 1B). Phylogenetic analysis indicated that BnaC07g30170D and BnaA06g26800D had a high level of similarity to these putative plant VIT proteins in Arabidopsis thaliana and Brassiceae family (Figure 1C). BnaC07g30170D, BnaA06g26800D and AtMEB2 were clustered into one group and BnaC07g30170D was more similar to AtMEB2 than that of BnaC07g30170D in the phylogenetic analysis. In addition, BnaC07g30170D has a higher similarity of protein sequence to AtMEB2 than that of BnaA06g26800D in their highly conserved sequence of VIT domains, thus BnaC07g30170D was designated as BnMEB2 for further study. Furthermore, protein 3D structure model showed that AtVIT1, AtMEB2 and BnMEB2 shared similar structures within VIT domain (Supplementary Figure S2). Therefore, these results suggest that BnMEB2 belongs to the VIT family.

The VITs were previously found to locate on vacuolar membrane to mediate vacuolar sequestration of iron storage and acted as metal transporters, such as TgVIT1 (Momonoi et al., 2009), AtVTL1 and AtVTL2 (Gollhofer et al., 2014). The subcellular localization of BnMEB2 was investigated by transiently expression in Arabidopsis protoplasts and onion epidermal cells. BnMEB2 was fused in frame to GFP (green fluorescent protein) in the expression vector pUC19-35Spro::GFP. 35Spro::BnMEB2::GFP fusion construct and the GFP alone control 35Spro::GFP, both driven by 35S promoter were introduced into Arabidopsis protoplasts. BnMEB2::GFP fusion protein was observed exclusively in the vacuolar membrane while GFP alone was found throughout the control cells (Figure 2A). Furthermore, the subcellular localization of BnMEB2 was also investigated by expression of 35Spro::BnMEB2::GFP in onion epidermal cells. BnMEB2::GFP fluorescence was observed in the vacuolar membrane too (Figure 2B). The above results showed that BnMEB2 was localized on the vacuolar membrane and may also be a vacuolar membrane transporter.

To provide deep insight into the function of BnMEB2 during different physiological processes, expression pattern was examined in various tissues. Transcription levels of BnMEB2 in all tissues of rapeseed, including root, young stem (14-day-old stem), mature stem (60-day-old stem), young leaf (14-day-old leaf), mature leaf (60-day-old leaf), flower and silique were detected by qRT-PCR (Figure 3A). BnMEB2 was substantially higher expressed in mature leaves and weakly detected in other tissues including young stems and young leaves. With the storage role of leaves, this result suggested that BnMEB2 may function and coordinate to mediate iron storage in the source tissues for human health. Interestingly, the expression of BnMEB2 in mature stems was higher than that of young stems, and it was the same between mature and young leaves, indicating that BnMEB2 may be highly expressed in mature tissues for advantage of iron storage during plant life cycle.

To further investigate whether BnMEB2 is associated with iron transportation, transcriptional abundance of BnMEB2 in B. napus roots and leaves were analyzed under normal Hoagland nutrient solution and high iron concentration treatment (Hoagland nutrient solution added with 500 μM FeSO₄). The rape seedlings when treated with 500 μM FeSO₄ showed a slow growth, yellowing of leaves and shrinking of roots when compared with the control. Three days after treatment (3 DAT), leaves of rape seedlings were obviously shrinked (Figure 3B). The expression of BnMEB2 was markedly induced by high iron concentration treatment in both roots and leaves. The transcription level of BnMEB2 was up-regulated at least 10-fold after 2 days treatment, and more than 70-fold for 3 days after treatment in seedling roots (Figure 3C). Similar results were obtained in leaves, where significant induction of BnMEB2 was observed after 1, 2, and 3 days of high iron concentration treatment (Figure 3D). Furthermore, in comparison with roots, the induction of BnMEB2 expression was faster in young leaves. In young leaves, at the first day after high iron concentration treatment did BnMEB2 show remarkably higher expression than the second and third days under the same treatment. This phenomenon might be responsible for iron storage in leaves to detoxify the iron stress, while roots are mainly responsible for iron acquisition. Moreover, the expression of BnMEB2 was higher in young roots than in young leaves (Figures 3C,D). Taken together, these data imply that BnMEB2 was highly induced by high iron concentration treatment and was involved in iron transportation.

To better understand the role of BnMEB2 in regulation the tolerance of high iron concentration stress, BnMEB2 over-expression vector driven by 35S promoter was constructed and transformed into Arabidopsis wild type (col-0) and meb2 (defined as 35S::BnMEB2/WT and 35S::BnMEB2/meb2, respectively). A total of three 35S::BnMEB2/WT (line 1#, 12#, and 30#) and three 35S::BnMEB2/meb2 (line 2#, 5#, and 21#) transgenic lines were selected and purified. qRT-PCR analysis showed that the expression levels of BnMEB2 in transgenic plants with either WT or meb2 background were significantly increased as compared with wild type and meb2 mutant (Figures 4A,B). To verify the functions of BnMEB2 in vivo, these transgenic lines were challenged with high iron concentration stress. 35S::BnMEB2/WT transgenic line 1#, 35S::BnMEB2/meb2 transgenic line 21#, wild type and meb2 were sowed on MS medium under a series of high iron concentration conditions (0.2 mM, 0.3 mM, 0.5 mM, 0.8 mM and 1 mM FeSO₄) and cultured for 20 days. As shown in Figure 4E, the high iron concentration significantly impacted on seeds germination at the seedling stage. 35S::BnMEB2/WT and 35S::BnMEB2/meb2 transgenic plants had a higher seeds germination rate and more tolerated to high iron environment than that of meb2, especially when the high iron concentration increasing to more than 0.5 mM FeSO₄. Over-expression of BnMEB2 restored and increased the germination rate of meb2 in the complementary lines 35S::BnMEB2/meb2. These results showed that 35S::BnMEB2/WT and 35S::BnMEB2/meb2 transgenic plants had stronger high iron tolerance than that of meb2, indicating that BnMEB2 was involved in high iron tolerance.

The enhanced tolerance of BnMEB2 over-expression plants to iron stress prompts us to evaluate the expression level of other key VIT genes such as AtVIT1, AtNRAMP3 and AtNRAMP4 in all four above lines. It was previously revealed that AtVIT1 functioned as the main transporter allowing iron uptake into vacuoles when exposed to excess iron (Kim et al., 2006). AtNRAMP3 and AtNRAMP4 genes specifically acted as a retriever to export iron from vacuoles when at the deficient iron conditions (Lanquar et al., 2005). The transcription level of AtVIT1,AtNRAMP3 and AtNRAMP4 showed no significant differences among 35S::BnMEB2 transgenic lines and meb2 when compared with wild type plants (Figures 4C,D), while the expression of MEB2 gene was dramatically increased in the 35S::BnMEB2/WT and 35S::BnMEB2/meb2 transgenic plants and strongly suppressed in meb2 (Figure 4A). These results suggested that the enhanced iron tolerance in 35S::BnMEB2/WT and 35S::BnMEB2/meb2 transgenic plants was mainly caused by the high expression of BnMEB2. Over-expression of BnMEB2 restored the iron tolerance ability of meb2 in the complementary lines, indicating that abundant expression of BnMEB2 facilitated the iron storage in plant vacuoles and thus leading to better tolerate the high iron concentration stress. Hence, BnMEB2 plays important roles in the tolerance of high iron concentration stress.

To deeply characterize the mechanism of enhanced iron tolerance in BnMEB2 over-expressed Arabidopsis transgenic plants, iron storage in roots of WT, meb2, 35S::BnMEB2/WT and 35S::BnMEB2/meb2 plants were determined by representative Perls’ staining. Iron and zinc contents in these seedlings were also measured by using inductively coupled plasma spectrometry. Under high iron concentration condition (0.5 mM FeSO₄), 35S::BnMEB2/WT and 35S::BnMEB2/meb2 transgenic plants showed a better growth status with bigger rosette leaves and developed root system when compared with meb2 (Figure 5A). Iron localization in roots were very similar among WT, meb2 and 35S::BnMEB2/meb2, only a slight higher in 35S::BnMEB2/WT when grown on MS medium (Figures 5B–E). While under the MS medium with 0.5 mM FeSO₄ condition, distinct blue staining was observed and became intensified in the roots of WT, 35S::BnMEB2/meb2 and 35S::BnMEB2/WT transgenic plants (Figures 5F–I). However, iron distribution in roots of meb2 had a little alternation between normal and high iron concentration conditions but root hairs elongation was obviously inhibited under iron stress condition (Figures 5C,G). Furthermore, when challenged by high iron concentration treatment, the iron and zinc contents were significantly increased in all types plants when compared with the WT under normal condition (Figures 5J,K). Interestingly, with high iron concentration treatment, iron contents in the seedlings of 35S::BnMEB2/meb2, 35S::BnMEB2/WT transgenic plants and WT were significantly higher than that of meb2 (Figure 5J), but zinc contents in all four seedlings were similar with each other (Figure 5K). These results indicated that over-expression of BnMEB2 increased iron storage in the vacuole of root cells and iron content in the Arabidopsis seedlings in response to high iron concentration stress, resulting in increased detoxification ability in BnMEB2 transgenic plants which ensured the normal growth and development.

Over the past decade, the important biological functions of several VITs in Arabidopsis, rice and other plant species had been characterized as vacuolar sequestration of iron storage, such as AtVIT1, OsVIT1, OsVIT2 and TgVIT1(Kim et al., 2006; Momonoi et al., 2009; Zhang et al., 2012). In Arabidopsis, AtVTL1, AtVTL2 and AtVTL5 which showed significant homology to AtVIT1, also catalyzed transportation of iron into vacuoles and thus contributed to the regulation of iron homeostasis in the plant (Gollhofer et al., 2014). Another protein AtMEB2 was a vital member of VIT family, and located on endoplasmic reticulum (ER) body membrane involving in iron homeostasis (Yoshida and Negishi, 2013). In general, proteins containing the conserved iron transportation motif of VIT domain are assigned to the VIT family, being responsible for the regulation of cytosolic iron homeostasis (Yoshida and Negishi, 2013). In this study, we identified a homolog of AtMEB2 in B. napus and assigned as BnMEB2. BnMEB2 also contained a VIT domain in the C-terminal region (Figure 1A). By clustal and phylogenetic analysis, BnMEB2 and BnaA06g26800D shared a high similarity with AtMEB2 and also shared similar 3D structures of VIT domain with AtVIT1 and AtMEB2 (Figures 1B,C; Supplementary Figure S2). Notably, it is peculiar that VIT family members were located on the vacuolar membrane. BnMEB2 also was located on vacuolar membrane (Figures 2A,B), indicating a conservative vacuolar metal transport function for BnMEB2. Therefore, it is reasonable to propose that BnMEB2 is a novel VIT member with underlying ability of metal iron transportation.

Incapable for escaping from unwanted changes in the environment, the sessile plants have developed a series of tolerance strategies such as vacuolar sequestration, metal chelation, and metal effluxion to cope with the negative consequences of metal toxicity (Clemens, 2006; Singh et al., 2016). AtMEB2 was involved in the metal ion homeostasis and enhanced iron resistance in yeast ccc1 mutant (Yamada et al., 2013). In B. napus seedlings, expression of BnMEB2 was strongly induced under high iron concentration treatment in both roots and leaves (Figures 3C,D), suggesting that BnMEB2 might play important role in iron homeostasis when plants grew under such conditions. At the stages of germination and growth, over-expression BnMEB2 transgenic seedlings had stronger iron tolerance while the seedlings of meb2 were more sensitive to high iron concentration stress (Figure 4E). The expression of MEB2 gene was dramatically up-regulated while other key VIT genes showed no significant differences among them (Figures 4A–D). These observations lead to the conclusion that the alteration of iron tolerance is closely related to the expression level of BnMEB2.

As far as metal toxicity is concerned, the most important mechanism of metal tolerance in plants is vacuolar detoxification. Although iron can be stored in ferritin protein in chloroplast and mitochondria, vacuolar detoxification serves as a principal and safe strategy in iron storage. In Arabidopsis, only 5% of iron was incorporated in ferritin protein, while vacuoles appear to be the major compartment for iron storage (Ravet et al., 2009). BnMEB2 did not affect the root hair development when grown on MS medium (Figures 5B–E). However, root hair elongation was obviously restrained in meb2 and iron content of meb2 was lower when compared with other seedlings under 0.5 mM FeSO₄ condition (Figures 5G,J), suggesting that loss function of MEB2 resulted in iron toxicity of roots cell and affect roots development which in turn to decrease iron acquisition. On the contrary, over-expression BnMEB2 transgenic seedlings were grown and developed normally even if iron distribution and acquisition significantly increased about five folds (Figures 5F–I). Simultaneously, BnMEB2 located on vacuolar membrane (Figure 2) and the expression level of BnMEB2 was critical for iron tolerance, showing that the high expression level of BnMEB2 acted as a vacuolar sequestration of iron storage to detoxify iron stress. Zinc contents in all types of plants increased from about 30–70 μg/g DW (dry weight), while no significant differences were detected among each other under low zinc concentration condition (Figure 5K), indicating that the increasing of one metal ion might activate roots to acquire other divalent metal too which further harden toxic metal stress, such as over-expression of NAS causing to concomitant increasing in iron and zinc content in rice grains (Johnson et al., 2011). Thus, BnMEB2 enhances iron tolerance which significantly plays a role in plants adapting to adverse environment.

The complex process of iron homeostasis must be highly coordinated by iron uptake, long-distance transport and distribution to different tissues and cell compartments. In the process of iron transport in plant, the VIT members contribute to iron trafficking in vacuoles. Vacuoles are the crucial compartment for storage and detoxification of iron excess and a buffering pool for iron remobilization in periods of iron deficiency. Based on the results above, this study showed that BnMEB2 was involved in high iron tolerance of B. napus. Thus, it is reasonable to propose a model for BnMEB2 regulation in iron homeostasis and enhancing iron tolerance in B. napus, which are partially schematically represented in Figure 6. When challenged by a high level of iron environment, abundant Fe³⁺ is firstly reduced to more soluble Fe²⁺ by FRO2 and is acquired into the root cells with poorly selective by membrane transporter IRT1, causing to iron excess and toxic in cytoplasm of cell. The status of iron excess triggers a significant up-regulated expression of iron transporter genes such as BnMEB2. Then, vacuolar membrane located BnMEB2 traffics of iron into the vacuoles for storage and to avoid iron toxicity. Finally, the iron level in cytoplasm rebalances within an optimal physiological range and the plants result in more tolerant to high iron concentration stress. Briefly, BnMEB2 enhances iron tolerance by the vacuolar sequestration and vacuolar detoxification strategy. It provides a new insight on iron homeostasis in the tolerance of high iron concentration stress.

With the tremendous rise of industrialization emissions of toxic metals, the increasing serious toxic metal pollution strongly interferes with the crops metabolism, development, and thus productivity and yield (Lutts and Lefèvre, 2015). The VIT members play important roles in maintaining the metal ion availability in the cytosol within an optimal range by appropriate sequestration into or retrieval from the vacuoles. For instance, over-expression of AtPCS1 in tobacco activated high levels of phytochelatin synthesis to increase detoxification capacity to toxic metal stress (Zanella et al., 2016). Over-expression of AtVIT1 increased accumulation of iron in the edible part of storage roots and stems of cassava (Narayanan et al., 2015). This study finds BnMEB2 conservatively localizing on the vacuolar membrane and enhancing tolerance to high iron concentration stress. Therefore, over-expression of BnMEB2 can be applied to minimize the serious threat of metal toxic stress and improve the survival and yield of B. napus and other crops. The current study highlights the significance of BnMEB2 potential application.

Iron is a required micro-nutrient for all living organisms. Iron deficiency anemia is one of the most prevalent micro-nutrient deficiencies in the world, affecting almost 25% of the world population (Clemens et al., 2002; Cvitanich et al., 2010). Thus, increasing iron biofortification in vegetable crops of Brassica family could have a positive significance on human health. In higher plants, to maintain the structural and functional integrity of thylakoid membranes in chloroplasts (Kim and Guerinot, 2007), approximately 90% of iron abundance is required to allocate in the leaves (Terry and Abadía, 1986). When B. napus was challenged by high iron environment, considerable iron amounts were accumulated in stomata guard cells of leaves (Tewari et al., 2013). Meanwhile, in oilseed rape mature leaves, BnMEB2 is highly expressed and can be induced by iron stress which revealed a close relationship between BnMEB2 and iron storage (Figure 3A). In our previous study, Bol022443 and Bra009746, significant homologies to AtMEB2 were identified in B. rapa and B. oleracea respectively. They are highly conserved in evolution and all strongly expressed in leaves by previous trancriptome detection (Tong et al., 2013; Liu et al., 2014; Zhou et al., 2014), suggesting that MEB2 was involved in iron distribution and storage in Brassicaceae leaves. For Brassica vegetables, their leaves are the main edible source for supplying vitamins, cellulose, minerals and other kinds of nutrition for human health. Study on MEB2 sheds light on a potential strategy for iron biofortification in Brassica crops.