A node-localized efflux transporter for loading iron to developing tissues in rice

Abstract

Iron (Fe) is an essential micronutrient for plant growth and development. It plays crucial roles in various organs and tissues of plants, but the molecular mechanisms governing its distribution to the above-ground parts after root uptake remain unclear. In this study, we identify OsIET1 (Oryza sativa Iron Efflux Transporter 1), a rice gene highly expressed in the nodes. OsIET1 encodes a plasma membrane-localized protein, which shows efflux transport activity for ferrous iron. It is predominantly expressed in the xylem regions of diffuse vascular bundles, and its expression is upregulated under high Fe conditions. Disruption of OsIET1 impairs Fe allocation, reducing Fe transport to developing tissues (young leaves and grains), while increasing accumulation in nodes and older leaves. This misdistribution causes chlorosis in young leaves and decreases grain yield, especially under Fe-deficient conditions. Furthermore, we detect excessive Fe deposition around the xylem of diffuse vascular bundles in the nodes. Given the pivotal role of nodes in mineral distribution, our results indicate that OsIET1 mediates inter-vascular Fe transfer by facilitating Fe loading into the xylem of diffuse vascular bundles. This process ensures preferential Fe delivery to developing tissues, thereby promoting optimal plant growth and productivity.

Che et al. identify a rice gene OsIET, which encodes a ferrous iron transporter in the node. Its disruption impairs Fe distribution to developing tissues, causing leaf chlorosis and reduced grain yield.

Introduction

Iron (Fe) is an essential micronutrient for plant growth and development, playing critical roles in key physiological processes such as photosynthesis, respiration, and chlorophyll synthesis^1,2. It is also a vital component of numerous enzymes such as catalase, peroxidase^1,2. Although plants require only a small amount of Fe (60 to 200 mg/kg)², Fe deficiency is a common issue in agriculture, particularly in alkaline soils, where its low solubility limits bioavailability despite its abundance. This deficiency adversely affects crop productivity. However, certain plant species have evolved specialized mechanisms to acquire insoluble Fe from the soil^2–5. In rice (Oryza sativa), a staple food for half the population of the world, two different strategies for Fe acquisition have been reported depending on growth conditions.

Under upland conditions, where ferric Fe (Fe(III)) is the prevalent form in soil, rice secretes 2’-deoxymuginec acid (DMA), a kind of phytosiderophore, from the roots to chelate insoluble Fe in soil. The chelated DMA-Fe(III) complex is taken up by the roots without reduction. Genes involved in biosynthesis, secretion of phytosiderophore, and uptake of its Fe complex have been identified^6–8. By contrast, under flooded conditions, where ferrous Fe (Fe(II)) is the major form due to soil reduction, rice roots take up Fe(II) likely through transporters such as IRT1 and OsNramp5^9,10. IRT1 and OsNramp5 transports Fe(II) in yeast, but their role in Fe uptake in plants needs to be further examined. On the other hand, since high Fe shows toxicity to plants, rice has also developed a sophisticated system to maintain internal Fe homeostasis by forming an intricate transcriptional network¹¹. For example, several basic Helix-Loop-Helix (bHLH) transcription factors have been reported to regulate Fe uptake either positively or negatively^12–14.

Following Fe uptake by the roots, it needs to be translocated to different organs, tissues, and cells in the above-ground part to play its multiple physiological functions. Furthermore, due to the limited mobility of Fe within plants, Fe taken up by the roots needs to be preferentially delivered to developing tissues, including young leaves and panicles. However, little is known about how Fe is preferentially delivered to these organs and tissues. Recent studies have established that the node in rice is a hub responsible for the distribution of mineral elements^15,16. Nodes have a complicated but well-developed vascular system, mainly including enlarged vascular bundle (EVB) and diffuse vascular bundle (DVB)^15,16. EVB is connected to the leaf associated with the node and the lower nodes, while DVB is connected to the upper nodes linking to the developing organs, such as new leaves and panicles. The xylem flow in EVB and DVB, which are connected by nodal vascular anastomosis (NVA) at the base of the node, is simply divided following the transpiration rate of each upper organ. Therefore, for preferentially delivering mineral elements, including Fe, an intervascular transfer from EVB to DVB in the node is required¹⁵. In addition, an apoplastic barrier localized at the bundle sheath (BS) cells and the development of the enlarged vascular bundles in the node are also required for efficient intervascular transfer¹⁵.

The inter-vascular transfer includes at least unloading of a mineral nutrient from xylem flow in EVB (xylem unloading, uptake from the xylem apoplast into adjacent cells), transfer through the symplast or apoplast of the parenchyma cell bridge (PCB), characterized by dense plasmodesmata, and reloading to xylem (xylem loading) or phloem of DVB in rice^15,16. Several mineral transporters involved in these processes have been identified. For example, Lsi6, OsNIP3;1, and OsOPT7 localized at the xylem parenchyma cells of EVB, are responsible for xylem unloading of silicon (Si), boron (B), and Fe(II), respectively^17–19. A phosphorus (P) transporter SPDT localized at the xylem transfer cells of EVBs is also responsible for the unloading of P from the xylem of EVBs²⁰, while Lsi3 localized at the parenchyma tissues between EVB and DVB mediates the release of Si to the xylem of DVB²¹. However, the transporter responsible for releasing Fe to the xylem of DVB for preferential distribution is unidentified not only in rice, but also in all other plant species.

Based on our previous gene expression profiles in node I of rice²², we found that a gene OsIET1 (Oryza sativa iron efflux transporter 1) belonging to the Vacuolar Iron Transporter (VIT) family showed higher expression. Detailed functional analysis in this study showed that, unlike other VIT members localized to the tonoplast, OsIET1 is localized to the plasma membrane and shows efflux transport activity for Fe(II). Furthermore, we found that this transporter is required for releasing Fe to the xylem of DVB for preferential distribution to the developing tissues at both vegetative and reproductive stages in rice.

Results

Cloning and phylogenetic analysis of OsIET1

The full-length cDNA of OsIET1 (Os04g0686800) was amplified by PCR using primers designed from the rice genome sequence (http://rapdb.dna.affrc.go.jp/). The open reading frame (ORF) is identical to that in the database, which composed of 627 bp, encoding a protein with 208 amino acids (Supplementary Fig. 1). OsIET1 belongs to Vacuolar Iron Transporter-like (VTL) group and there are five homologs in rice genome, which share 43-46% identity with each other (Supplementary Fig. 1). Among them, OsIET1 shares the highest identity to Os04g0538400, uncharacterized gene, while OsIET1 only shares 14-15 % identity with VIT group (OsVIT1 and OsVIT2), which are involved in vacuolar Fe transport²³.

Expression pattern of OsIET1

We examined the expression pattern of OsIET1 in various rice organs at different growth stages of rice grown in a paddy field. At the vegetative growth stage (tillering), OsIET1 was predominantly expressed in the shoot basal region, including the basal nodes, whereas the expression was very low in the roots and shoots (Fig. 1a). At the flowering stage, high expression of OsIET1 was found in the upper nodes, including node I, II, and III (Fig. 1a). At the grain-filling stage, the highest expression was observed in node I (Fig. 1a).

Fig. 1. Expression pattern of OsIET1.

a Expression pattern of OsIET1 in various organs at different growth stages. Rice was grown in a paddy field, and various organs were sampled at different growth stages. b Response of OsIET1 expression in the shoot basal regions to metal deficiency. Rice was grown in a nutrient solution with or without Zn, Fe, Mn, or Cu for 1 week. c Dose-responsive expression of OsIET1. d Time-dependent expression of OsIET1 in response to Fe supply. Seedlings (17-d-old) precultured without Fe were exposed to different Fe concentrations (0, 0.2, 2, and 20 μM) as FeSO₄ for 24 h (c) or different times (0, 2, 6, and 24 h) at 20 μM Fe (d). The shoot basal region was sampled for RNA extraction. The expression level was determined by quantitative RT-PCR. Expression relative to roots at the tillering stage (a), to the control (b), to -Fe (c, d) is shown. HistoneH3 was used as internal standard. Data are means ± SD of three (a, b) and four (c, d) biological replicates. Asterisks show a significant difference when compared with control in b; to 0 mM Fe (II) in c and to 0 h in d (*, P < 0.05; **, P < 0.01) by two tailed Studentʼs t-test. Source data are provided as a Source Data file.

We also investigated the response of OsIET1 expression to the deficiency of essential metals in the shoot basal region. The OsIET1 expression was significantly down-regulated by Fe-deficiency treatment, but remained unchanged under Mn, Zn, or Cu deficiency conditions (Fig. 1b). A dose-response analysis further confirmed that the OsIET1 expression in the shoot basal regions was upregulated by Fe supply (Fig. 1c), and the expression was similar at different Fe concentrations. A time-course experiment showed that Fe-induced expression of OsIET1 was observed at 2 h upon Fe supply and kept at a similar level until 24 h (Fig. 1d).

Subcellular and cellular localization of OsIET1

To observe the subcellular localization of OsIET1, the ORF of OsIET1 was fused with GFP at the N-terminus and expressed in rice leaf protoplasts under the control of a 35S promoter. GFP-OsIET1 signals were detected at the cell periphery, extending outward from the chloroplast (Fig. 2a-d). In contrast, GFP alone was localized to the cytosol and nucleus (Fig. 2e-h). The subcellular localization was further confirmed by co-expression of GFP-OsIET1 with DsRed, a marker of cytosol and nucleus in onion epidermal cells. The GFP-OsIET1 signal was not overlapped with that from DsRed (Fig. 2i-l), in contrast to GFP alone, with full overlap of the two signals (Fig. 2m-p). Collectively, unlike other homologs of OsIET1, OsIET1 is a plasma membrane-localized protein.

Fig. 2. Subcellular localization of OsIET1.

OsIET1 fused with GFP at the N-terminus (a-d, i–l) or GFP alone (e–h, m–p) was transiently expressed in rice protoplast (a–h) or co-transformed with 35S:DsRed into onion epidermal cells (i–p), respectively. GFP signal (a, e, i, m), chloroplast signal (b, f), DsRed signal (j, n), bright field image (c, g, k, o) and a merged image (d, h, l, p) are shown. Arrows indicate the nucleus. Scale bars = 25 µm.

To examine the tissue- and cell-specific expression of OsIET1, we generated transgenic rice lines expressing GFP under the control of the OsIET1 promoter. At the vegetative stage, immunostaining with GFP antibody revealed that OsIET1 was mainly expressed in the xylem region of DVB/regular VB (RVB), xylem-phloem boundary of EVB, and around the nodal vascular anastomosis (Fig. 3a, d, g). At the reproductive stage, GFP signal was mainly detected in the xylem region of DVB of node I (Fig. 3b, e, h). No GFP signal was detected in the basal node or node I of non-transgenic plants, confirming the specificity of the antibody (Fig. 3c, f). This result is consistent with the expression pattern of OsIET1 investigated by laser microdissection (LMD)-combined RNA-seq. Different from OsOPT7 and OsVIT2, which were highly expressed in the xylem of EVB and PCB, respectively, OsIET1 was highly expressed in the DVB (Supplementary Table 1, Fig. 3h).

Fig. 3. Tissue- and cell-specificity of OsIET1 expression in basal node (a,d) and node I (b,e).

Immunostaining with a GFP antibody was performed in transgenic rice carrying the OsIET1 promoter fused with GFP (a, b, d, e) or wild-type rice (c, f) as negative control. Red color indicates the GFP antibody-specific signal. Blue color indicates cell wall auto-fluorescence. The regular vascular bundle (R), enlarged vascular bundle (E) and nodal vascular anastomosis (NVA), xylem and phloem region of enlarged vascular bundle (X_E, P_E), and phloem area of diffuse vascular bundle (P_D) are shown. Based on the picture in (d, e), different kinds of vascular bundles (VBs) are illustrated in different colors (g, h). Open and solid areas represent xylem and phloem, respectively, and dotted lines surrounding the EVBs indicate the bundle sheath (BS) cell layer of the EVB, which has an apoplastic diffusion barrier. The basal node was sampled from rice cultured with 1/2 Kimura B solution for one month. Node I was sampled from rice grown in soil at the grain filling stage. Scale bars, 200 µm.

Transport activity of OsIET1

To investigate the transport activity of OsIET1, we employed an assay system with proteoliposomes²⁴. The purified protein fraction exhibited a major protein band at the expected molecular weight of OsIET1 (Fig. 4a), which was confirmed by western blot with anti-6 × His antibody (Fig. 4a). Transport assays with radioisotope ⁵⁵Fe showed that proteoliposomes containing recombinant OsIET1 exhibited significantly higher transport activity for Fe(II) at external pH 7.5 compared to control liposomes (without OsIET1 protein) (Fig. 4b). However, no transport activity was detected at external pH 6.0 (without ΔpH) (Fig. 4b). Furthermore, the transport activity was abolished in the presence of 2 μM CCCP, a protonophore. These results indicate that OsIET1 is a proton-coupled efflux transporter for Fe(II) (Fig. 4b). Moreover, Fe(II) transport activity was not inhibited by the presence of Cu(II), Zn(II) or Mn (II) at 0.5 mM, but was significantly decreased by the presence of 0.5 mM cold Fe(II) (Fig. 4c), indicating high specificity of OsIET1 for Fe(II) ion.

Fig. 4. Transport activity of OsIET1 protein.

a Purification of OsIET1 protein. The purified fraction was analyzed by SDS–PAGE and visualized by Coomassie Brilliant Blue staining (left) and by western blot with anti-6 × His antibody (right). b, c OsIET1-mediated Fe transport labeled with ⁵⁵Fe. The proteoliposomes containing recombinant OsIET1, or liposomes (negative control) were incubated in a buffer solution containing 10 µM radioisotope ⁵⁵Fe(II) (b) at either pH 6.0 or pH 7.5, and assayed after 2 min. The proteoliposomes containing OsIET1 were also incubated in a buffer solution containing 10 µM radioisotope ⁵⁵Fe(II) at pH 7.5 in the presence of 2 µM CCCP. c Effect of Cu(II), Zn(II) or Mn (II) on ⁵⁵Fe(II) uptake. The cis-inhibition of Fe(II) transport was assayed at 1 min, and determined in the presence of the 0.5 mM non-labeled Fe(II), Zn(II), Mn(II), or Cu(II). Data are mean ± SE of replicates (n = 6). *, P < 0.05; **, P < 0.01 (two-tailed paired Student’s t-test); ns, not significant. Source data are provided as a Source Data file.

Phenotypic analysis of OsIET1 knockout lines at the vegetative stage

To investigate the physiological role of OsIET1, we generated knockout lines of OsIET1 using CRISPR/Cas9-mediated targeted mutagenesis. We obtained two independent mutant lines, iet1-1 and iet1-2, which has 1 bp insertion (adenine) or 1 bp deletion (cytosine) at the targeted site, respectively (Supplementary Fig. 2). When these lines together with their wild-type line (WT) were exposed to a solution containing different Fe(II) concentrations (from 0 to 20 µM), chlorosis was observed in newly expanded leaf (leaf 7) of the knockout line at 0, 0.2, and 2 μM Fe (Fig. 5a), but not in the WT. At 20 µM Fe, leaf chlorosis was recovered in the knockout lines (Fig. 5a). The SPAD value, which indicates chlorophyll content, was lower in the iet1 mutants in leaf 7 at low Fe concentration (0-2 μM Fe(II)) compared with that of WT, but was similar to WT at 20 μM Fe(II) (Fig. 5b). Analysis of mineral elements showed that there was no difference in root Fe concentration between knockout lines and WT grown at different Fe conditions although most Fe in the roots was thought to be deposited on the root surface (Fig. 5c). The shoot Fe concentration did not differ between WT and iet1 mutants at 0 and 0.2 μM Fe, but was higher in the mutants than in the WT at 2 and 20 μM Fe conditions (Fig. 5d).

Fig. 5. Phenotype of iet1 mutants at vegetative growth stage.

Phenotype (a) and SPAD value (b) of the youngest expanded leaf blade at different Fe supply conditions. Fe concentration in the root (c) and shoot (d). 3-week-old seedlings of iet1 mutants and WT were cultured in a nutrient solution containing 0, 0.2, 2, or 20 μM FeSO₄ for 9 days. Data represent the means ± SD (n = 3). Asterisks indicate a significant difference (*, P < 0.05; **, P < 0.01) between WT and iet1 mutants using two-sided Tukey’s test. Source data are provided as a Source Data file.

Effect of OsIET1 knockout on Fe distribution and deposition at the vegetative stage

The above results showed that although iet1 knockout lines contained similar Fe levels in the whole shoots as the WT, they showed typical Fe-deficiency symptoms at low Fe (Fig. 5a). Given that OsIET1 is highly expressed in the nodes (Fig. 1a), we hypothesized that it may be involved in the distribution of Fe. We therefore compared the Fe concentration in different tissues of the shoots, including the shoot basal region, old leaves (leaf 2-5), middle leaf (leaf 6), and new leaves (leaf 7-8) of WT and iet1 mutants. The results showed the shoot basal region of the mutants contained 2.8-5.8 times more Fe than the WT at all Fe conditions tested (Fig. 6a–d). By contrast, Fe concentration in the new leaves was significantly lower in the mutants than in the WT (Fig. 6a–d). We calculated the distribution ratio of Fe to different tissues. As a result, distribution of Fe to the shoot basal region and old leaves in the mutants was increased, whereas that to the new leaves was decreased at all Fe conditions compared with the WT (Fig. 6e–h). By contrast, the distribution of other elements in different tissues did not differ between WT and mutants (Supplementary Figs. 3, 4). These results indicate that knockout of OsIET1 altered the distribution of Fe to different tissues.

Fig. 6. Concentration and distribution of Fe in different tissues of the shoots.

Concentration (a–d) and distribution (e–h) of Fe in different tissues of the shoots of the wild-type rice (WT) and iet1 mutants. 3-week-old seedlings of iet1 mutants and WT were cultured in a nutrient solution containing 0 (a, b), 0.2 (c, d), 2 (e, f), or 20 (g, h) μM FeSO₄. After 9 days, different tissues were separated and subjected to Fe determination by inductively coupled plasma-mass spectrometry (ICP-MS). Data represent the means ± SD (n = 3). Asterisks indicate a significant difference (*, P < 0.05; **P < 0.01) between WT and iet1 mutants using two-sided Tukey’s test. Source data are provided as a Source Data file.

Since the shoot basal region contains both leaf sheath and basal nodes, we performed the Perls’ blue staining to observe the tissue-specific deposition of Fe. At 0.2 μM, the blue signal was not detected in all lines due to low Fe accumulation (Fig. 7), however, at 2 and 20 μM Fe, a much stronger blue signal was found in PCB in the iet1 mutants than in the WT (Fig. 7). No signal was observed in the leaf sheath of all lines (Fig. 7). These results indicate that high Fe accumulation in the shoot basal region resulted from Fe deposition in PCB of the basal nodes.

Fig. 7. Iron deposition pattern in the basal nodes.

Iron deposition in the basal node of WT and iet1 mutants with Perls’ blue staining. 19-d-old seedlings of iet1 mutants and WT were cultured in a nutrient solution containing 0.2 (a), 2 (b), or 20 (c, d) μM FeSO₄ for 2 weeks. The basal node was sampled for Perls’ blue staining after being sliced. Blue color shows Fe deposition. The regular vascular bundle (RVB), enlarged vascular bundle (EVB), and parenchyma cell bridge (PCB) in the node are shown. Scale bars = 200 μm.

Short-term labeling experiment with ⁵⁷Fe

To confirm the role of OsIET1 in the Fe distribution, we performed a short-term labeling experiment with ⁵⁷Fe stable isotope. After plants precultured with or without Fe were exposed to 2 μM ⁵⁷Fe(II) for 24 h, both the knockout lines and WT accumulated similar ∆⁵⁷Fe (newly acquired Fe) in the roots, although most Fe in the roots was deposited on the root surface (Fig. 8a), irrespective of the pretreatment with or without Fe. In the shoots, the mutants showed similar ∆⁵⁷Fe accumulation as the WT in plants precultured without Fe, however in plants precultured with Fe, the mutants even showed higher level of ∆⁵⁷Fe in the shoots than the WT (Fig. 8b). The total ∆⁵⁷Fe accumulation showed no difference between WT and iet1 mutants (Fig. 8c), but the root-to-shoot translocation rate of ∆⁵⁷Fe was slightly enhanced in iet1 mutant only precultured with 2 μM Fe condition (Fig. 8d). This is probably due to decreased distribution of Fe to the developing tissues of the mutants. Comparison of ∆⁵⁷Fe distribution showed a clear trend in plants both with and without Fe preculture; more ∆⁵⁷Fe was distributed to the shoot basal region, but less to the new leaves (leaf 7-8) of the mutants compared with the WT (Fig. 8e-f). These results confirmed that OsIET1 mediates the preferential distribution of Fe to the developing tissues (new leaves).

Fig. 8. A short-term labeling experiment with the stable isotope of ⁵⁷Fe at the vegetative growth stage.

Net concentration (Δ⁵⁷Fe) in the roots (a) and shoots (b) of plants with and without Fe preculture. Total net ∆⁵⁷Fe uptake (c) and the root to shoot translocation rate of net ∆⁵⁷Fe (d). Distribution of net ∆⁵⁷Fe without (e) and with Fe preculture (f). 3-week-old seedlings of both wild-type rice and iet1 mutant lines were exposed to a nutrient solution containing 2 μM ⁵⁷Fe stable isotope for 24 h. Different tissues were separately harvested and subjected to determination of Fe by inductively coupled plasma-mass spectrometry (ICP-MS). Distribution of each element was calculated based on (net ⁵⁷Fe increase in each organ/whole shoot ⁵⁷Fe increase) x 100. Data represent the means ± SD (n = 3). Asterisks indicate a significant difference (*, P < 0.05; **, P < 0.01) between WT and iet1 mutants using two-sided Tukey’s test. Source data are provided as a Source Data file.

Phenotypic analysis of iet1 mutants at the reproductive stage

To investigate the role of OsIET1 at the reproductive stage, we grew the mutant lines and their WT until maturity in both hydroponic solution and paddy field conditions. In hydroponic conditions containing 1 μM FeSO₄, iet1 mutants exhibited significant growth retardation compared with WT at the grain filling stage (Supplementary Fig. 5). The fresh weight and length of both the root and shoot were markedly lower or shorter in iet1 mutants than in WT (Fig. 9a, b). Chlorosis was observed in the leaf blade of flag leaf and leaf II (second upper leaf) of the mutants, but not in the WT (Fig. 9c). The SPAD value was significantly reduced in these leaves of the mutants compared to WT (Fig. 9d).

Fig. 9. Phenotype of iet1 mutants at grain filling stage.

Fresh weight (a), and length of root and shoot (b) of iet1 mutants and the wild-type rice (WT). Phenotype (c) and SPAD value (d) of flag leaf and leaf II blade of iet1 mutants and WT. Concentration (e) of Fe in different tissues in the WT rice and iet1 mutants. WT rice and iet1 mutants were cultured in a nutrient solution containing 1 μM FeSO₄. At grain filling stage (2 months after transplanting), different tissues were separately harvested and subjected to Fe determination. Data represent the means ± SD (n = 3). Asterisks indicate a significant difference (P < 0.05) between WT and iet1 mutants using two-sided Tukey’s test. Source data are provided as a Source Data file.

Mineral analysis showed that knockout of OsIET1 resulted in decreased Fe accumulation in the leaf II, flag leaf, and panicle, but increased Fe levels in internode II, node I, and node II compared with the WT (Fig. 9e).

When grown in a paddy field, the growth of iet1 mutants was significantly retarded compared to WT at harvest, exhibiting reduced plant height and tiller numbers, and lower 1000-grain weight, leading to a decreased yield by 38–48% (Fig. 10a–f). The mutants accumulated higher Fe levels in node II, leaf II, flag leaf, and other parts, with no significant differences in Fe concentration in the husk and brown rice (Fig. 10g).

Fig. 10. Phenotype of iet1 mutants at mature stage.

Growth (a), plant height (b), tiller number (c), grain number per panicle (d), 1000-grain weight (e), and yield per plant (f) of iet1 mutants and the wild-type rice (WT). Fe concentration (g) in different tissues. The WT rice and iet1 mutants were grown in field until maturity. At harvest, growth parameters were recorded, and different tissues were separately harvested and subjected to Fe determination. Data represent the means ± SD (n = 3). Asterisks indicate a significant difference (P < 0.05) between WT and iet1 mutants using two-sided Tukey’s test. Source data are provided as a Source Data file.

Discussion

To play crucial roles of Fe, it must be allocated to different tissues of the above-ground parts, depending on their requirements, following uptake and xylem loading. Developing tissues, such as young leaves, usually have low transpiration compared with old leaves, but require more Fe for their active growth; therefore, a preferential distribution system for Fe is required. In rice, the preferential distribution of mineral elements is mediated by transporters expressed in the nodes. So far, OsOPT7 has been reported to be involved in the preferential distribution of Fe in rice node¹⁷. OsOPT7 functions as an influx transporter for Fe(II) and is localized in the xylem parenchyma cells of the EVB in the nodes. Knockdown of OsOPT7 altered Fe distribution; less Fe was distributed to the new leaf, upper nodes, and developing panicle, but more Fe was distributed to the old leaves. Therefore, OsOPT7 facilitates xylem unloading of Fe from EVB, the first step of inter-vascular transfer for Fe. After the unloading of Fe by OsOPT7, part of Fe will be sequestered by OsVIT2, a tonoplast-localized transporter for Fe(II). OsVIT2 is mainly expressed in the parenchyma cell bridges of nodes²⁵. Knockout of OsVIT2 resulted in decreased Fe distribution to the nodes, but increased Fe to the leaf blade and grains due to decreased vacuolar sequestration of Fe. In the present study, we found that OsIET1 mainly facilitates Fe²⁺ efflux from xylem parenchyma cells of the diffuse vascular bundles into the xylem vessels, the last step of inter-vascular transfer of Fe in the nodes (Supplementary Fig. 6). This is supported by its expression pattern, tissue localization, and phenotypic characteristics. OsIET1 is highly expressed in the nodes at both vegetative and reproductive growth stages (Fig. 1). Furthermore, it is specially expressed in the xylem region of DVB (Fig. 3). Unlike OsOPT7 and OsVIT2, knockout of OsIET1 resulted in increased Fe deposition in the PCB around the xylem region of DVB in the nodes (Fig. 7). As a result, knockout of OsIET1 decreased growth and seed fertility (Fig. 10). Therefore, OsIET1 plays a vital role in preferential distribution of Fe, which is crucial for healthy growth and grain development.

VITs form a unique transporter family, which is found in plants, fungi, and protists such as the malarial parasite Plasmodium falciparum²⁶. In this family, there are two distinct clades; VIT and VTL (Supplementary Fig. 1). The difference between VIT and VTL is that VTL lacks a cytosolic loop thought to mediate Fe²⁺/H⁺ antiport in VIT. Several members belonging to VTL clade have been functionally characterized in Arabidopsis, Lotus japonicus, soybean (Glycine max), Medicago truncatula^27–32. For example, overexpressing the tonoplast-localized transporter genes AtVTL1, AtVTL2 or AtVTL5 in the nramp3/nramp4 double mutant background restored the root growth and increased seed Fe concentration, indicating their contribution to Fe homeostasis in Arabidopsis³⁰. LjSEN1 and GmVTL1a are essential for nitrogen-fixing symbiosis in Lotus japonicus and Glycine max, respectively^28,29,31,32. MtVTL8, the closest homolog of LjSEN1, mediates the transport of Fe to symbiotic rhizobia²⁷. GmVTL1 mediates the transport of Fe across the symbiosome membrane to bacteroids, although its subcellular localization needs to be further addressed^28,29. However, different from these homologs, we found that OsIET1, belonging to the VTL group, is localized to the plasma membrane, the only member showing different subcellular localization (Fig. 2). This is consistent with its role in (re-)loading Fe(II) to the xylem of DVB. Among the five homologs of OsIET1 in rice, except for OsIET1 characterized in this study, other members remain uncharacterized. Investigation of the subcellular localization of these homologs may provide further insights into the different roles of this clade in Fe transport in the future.

Most Fe-related genes involved in Fe uptake, such as IRT1, OsYSL15, are upregulated by Fe-deficiency in rice^7,8,10,33. Among genes expressed in the rice nodes, the expression of OsOPT7 is up-regulated by Fe-deficiency¹⁷, whereas that of OsVIT2 and OsIET1 is up-regulated by Fe supply²⁵ (Fig. 2). These different responses to Fe-deficiency and -supply at the transcriptional level could be attributed to their different roles in Fe transport (Supplementary Fig. 6). For example, upregulation of OsOPT7 in response to Fe-deficiency could facilitate more Fe unloaded from the xylem of EVB to meet the requirement of Fe in young tissues. By contrast, up-regulation of OsVIT2 by Fe-supply could sequester more Fe into the vacuoles to avoid cytotoxicity of excess Fe. In the case of OsIET1, up-regulation by Fe-supply is probably to reduce excess Fe transfer to the phloem by increasing export to the xylem, which is necessary to maintain Fe homeostasis in rice.

In conclusion, our study underscores the essential role of OsIET1 expressed in the nodes in regulating Fe distribution in rice. OsIET1 is localized to the plasma membrane and functions as an efflux transporter for Fe(II). It is crucial for the efficient allocation of Fe to developing tissues by reloading Fe into the xylem of DVB, which is crucial for the healthy growth and productivity in rice.

Methods

Cloning of the full-length ORF of OsIET1

To clone the full-length ORF of OsIET1, total RNA was extracted from rice node I (cv. Nipponbare) using an RNeasy Plant Mini Kit (Qiagen). cDNA was synthesized using ReverTra Ace qPCR RT Master Mix with gDNA remover (TOYOBO), following the manufacturer’s protocol. The full-length OsIET1 ORF was amplified by PCR using primers (supplementary Table 2) designed based on the putative cDNA sequence (Os04g0686800) from the Rice Annotation Project Database (http://rapdb.dna.affrc.go.jp/) with putative translation start and stop sites. The PCR product was cloned into the pGEM-T vector (Promega) and sequenced using an ABI Prism 3130 sequencer (Applied Biosystems).

Expression pattern analysis with quantitative RT-PCR

To investigate the expression pattern of OsIET1 in different organs at various growth stages, cDNA prepared from a previous study was used³⁴. For mineral deficiency experiments, 4-week-old seedlings were grown in a nutrient solution with or without Zn, Fe, Mn, or Cu for 7 days, and the shoot basal region (0-1 cm from the root-shoot junction) was used for RNA extraction. To compare gene expression under different Fe concentrations, seedlings (2-week-old, cv. Nipponbare) were exposed to a solution free of Fe. After 3 days of exposure, the plants were exposed to a solution containing 0, 0.2, 2, and 20 μM FeSO₄ for 24 h. The shoot basal region was harvested for RNA extraction. For the time-course experiment, similar seedlings as the dose-response experiment were exposed to 20 μM FeSO₄. At 0, 2, 6, and 24 h, the shoot basal region was sampled for RNA extraction. Total RNA extraction and cDNA synthesis were performed as described above. All experiments were performed with 4 biological replicates.

Subcellular localization of OsIET1

To investigate the subcellular localization of OsIET1, we constructed a translational fusion of OsIET1 with green fluorescent protein (GFP). The OsIET1 ORF was amplified by PCR from cDNA derived from rice node tissue. The amplified fragment was then inserted between the cauliflower mosaic virus (CaMV) 35S promoter and GFP-NOS terminator in the pBluescript vector. Rice protoplasts were prepared from leaves of 2-week-old seedlings. GFP-OsIET1 construct or GFP alone were mixed with 100 μl protoplasts, followed by adding 100 μl of 40% polyethylene glycol (PEG). The transfected protoplasts were then incubated with gentle shaking for 16 h at 22 °C. For transient expression in onion epidermal cells, onion epidermal cells were co-transfected with either construct and DsRed, a marker of the cytosol and nuclei. The transfected cells were incubated for 14 h in the dark. The GFP fluorescence was observed by a confocal laser scanning microscope (TCS SP8x; Leica Microsystems). The primer sequences used to amplify the OsIET1 ORF for subcellular localization are provided in Supplementary Table 2.

Tissue- and cell- specificity of OsIET1 expression

To examine the tissue-specific expression of OsIET1, transgenic rice lines carrying OsIET1 promoter fused to GFP were generated. The 2.0 kb OsIET1 promoter region was amplified from rice genomic DNA using primers listed in Supplementary Table 2. The amplified promoter was then fused to the GFP gene isolated from the pBluescript vector. The resulting construct was cloned into the pPZP2H-lac vector³⁵ and transformed into calli of Nipponbare rice via Agrobacterium-mediated transformation. Tissue-specific localization in the basal node and node I was observed by immunostaining with an anti-GFP antibody (A11122, 1:1000; Thermo Fisher Scientific, Waltham, MA, USA) in both transgenic rice (T2 generation) and non-transgenic rice as a negative control³⁶. Alexa Fluor 555 goat anti-rabbit IgG (A21428, 1:2000; Thermo Fisher Scientific, Waltham, MA, USA) was used as a secondary antibody. The basal node was sampled from rice cultured in 1/2 Kimura B solution for one month, while node I was sampled from rice grown in soil at the grain-filling stage.

To further confirm the tissue-specific expression of OsIET1, different tissues of node I were divided by laser microdissection (Arcturus Laser Capture Microdissection System, Life Technologies, Carlsbad, CA, USA). Five different tissues were isolated for RNA-seq analysis, including the xylem region of EVB without bundle sheath, the xylem region of EVB with bundle sheath, the phloem region of both EVB and DVB, DVB, and PCB.

Transport activity assay of OsIET1

To determine transport activity of OsIET1, a proteoliposome assay system was employed²⁴. Briefly, the recombinant baculovirus containing the full length of OsIET1 gene with N-terminal 6×His-tag fusion protein was generated using the Bac-to-Bac baculovirus expression systems (Invitrogen). High Five cells (5 × 10⁶ cells/10-cm dish) were infected with recombinant baculoviruses at a multiplicity of infection of 1 and cultured an additional 48 h. The cells (1.6 × 10⁸ cells) were harvested and suspended in a buffer consisting of 20 mM Tris-HCl (pH 8.0), 0.1 M sodium acetate, 10% glycerol, 0.5 mM dithiothreitol, 1 mM phenylmethanesulfonyl fluoride, 10 μg/ml pepstatin A, and 10 μg ml⁻¹ leupeptin and disrupted by sonication with a TOMY UD-211 tip sonifier (OUTPUT4). The cell lysates were centrifuged at 800 × g for 10 min to remove debris, and the resultant supernatant was centrifuged at 160,000 × g for 1 h. The pellet (membrane fraction) was suspended in a buffer consisting of 20 mM MOPS-Tris (pH 8.0), 10% glycerol, 10 μg/ml pepstatin A, and 10 μg ml⁻¹ leupeptin at 10 mg protein ml⁻¹ and solubilized with 1.6% Fos-choline-14. After centrifugation at 260,000 × g for 30 min, the supernatant was applied to 1.25 ml of Ni-NTA Superflow resin (Qiagen), and the mixture was incubated for 4 h at 4 °C. The resin was washed with 10 ml of washing buffer consisting of 20 mM MOPS-Tris (pH 8.0), 20 mM imidazole, 20% glycerol, and 0.1% n-decyl-β-D-thiomaltopyranoside. The OsIET1 protein was eluted from the resin with 3 ml of the same buffer containing 250 mM imidazole.

For the transport assay, purified OsIET1 (30 μg) was incubated with liposomes (500 μg) and frozen at -80 °C for at least 10 min. The mixture was diluted with reconstitution buffer containing 20 mM MES-KOH (pH 6.0), 0.1 M potassium acetate, and 5 mM magnesium acetate. The reconstituted proteoliposomes were pelleted by centrifugation at 200,000 g for 1 h at 4 °C and resuspended in reconstitution buffer. Liposomes were prepared at a concentration of 10 mg ml⁻¹ with a lipid composition of 40% phosphatidylcholine, 30% phosphatidylethanolamine, 10% phosphatidylserine, and 20% cholesterol by weight. For Fe transport assays, FeCl₃ was labeled with ⁵⁵Fe (PerkinElmer) and reduced to FeCl₂ by the addition of ascorbate (200x concentration of FeCl₃). Reaction mixtures (130 μl) containing 0.6 μg of protein incorporated into proteoliposomes, 20 mM MES-KOH (pH 6.0) or 20 mM MOPS-KOH (pH 7.5), 0.1 M potassium acetate, 5 mM magnesium acetate, 10 mM KCl, and the indicated concentrations of reduced ⁵⁵FeCl₂ (30 MBq μmol⁻¹) were incubated at 27 °C. For competition assays, 0.5 mM non-labeled Fe(II), Mn (II), Zn(II), or Cu(II) was added to the reaction mixture. After 2 min incubation (or 1 min for competition assays), the transport assay was terminated by separating the proteoliposomes from the external mixture using centrifuge columns packed with Sephadex G-50 (fine). The Fe incorporated into the liposomes was quantified by liquid scintillation counting (Aloka, Tokyo, Japan). All experiments were performed with six biological replicates.

Generation of knockout lines of OsIET1

OsIET1 knockout lines were generated using the CRISPR/Cas9 method as follows: the Cas9 plant expression vector (pU6gRNA) and sgRNA expression vector (pZDgRNA_Cas9ver.2_HPT) were provided by Dr. Endo (National Institute of Agrobiological Sciences, Japan)³⁷. According to the design principles of the target sequences in the CRISPR/Cas9 system, a target sequence was prepared for OsIET1 (Supplementary Table. 2). Double-stranded target sequences (made by annealing the paired single oligonucleotides after 95 °C for 5 minutes denature) were cloned into the pU6gRNA vector (linearized by BbsI). OsU3-gYSA in pZDgRNA-Cas9ver.2_HPT was replaced by synthetic gRNA expression constructs using AscI and PacI sites. The derived construct was transformed into calluses (cv. Nipponbare) by Agrobacterium-mediated transformation³⁸. Target and genotyping sequences for the iet1 mutants are listed in Supplementary Table 2.

Phenotypic analysis of OsIET1 knockout lines

Seeds of wild-type rice (cv. Nipponbare) and two CRISPR/Cas9 mutant lines (iet1-1 and iet1-2) were soaked overnight at 30 °C in the dark for 2 days, followed by transferring onto a net floating in a 0.5 mM CaCl₂ solution. On day 7, the seedlings were transferred to 3.5-L plastic pots containing half-strength Kimura B solution³⁹. The solution pH was adjusted to 5.6, and the nutrient solution was renewed every 2 days. Plants were grown in a closed greenhouse at 25-30 °C under natural light. All experiments were performed with three biological replicates.

For comparison at the vegetative growth stage, 3-week-old seedlings of both iet1 mutants and WT were cultured in a nutrient solution containing 0, 0.2, 2, or 20 μM FeSO₄. Solution of FeSO₄ was freshly prepared before use. After 9 days, the roots, shoot basal region (0-5 mm from the root-to-shoot junction), old leaves (leaf 2–5), leaf 6, and young leaves (leaf 7, 8) were sampled for growth parameter analysis and mineral element determination as described below. The distribution ratio of Fe and other mineral elements in different organs was calculated based on the biomass and Fe concentration. At the harvest, the SPAD values of leaf 7 (the newest fully expanded leaf) were measured with a SPAD-5 chlorophyll meter (Konika Minolta, Tokyo, Japan). The leaves were also photographed.

For comparison at the reproductive growth stage, both the WT rice and iet1 mutants were cultured in a nutrient solution containing 1 μM FeSO₄. At the grain-filling stage, plants were photographed. The fresh weight, length of roots and shoots, and SPAD value of the flag leaf blade and leaf II blade were recorded. The above-ground parts were separated into node II, leaf II, node I, flag leaf, and panicle, and subjected to the determination of mineral elements as described below.

We also grew WT and iet1 mutants in a field in Zhenjiang, Jiangsu, where the soil type is meadow soil. At the ripening stage, plants were harvested, and the yield was recorded. The plants were separated into brown rice, husk, rachis, peduncle, node I and II, flag leaf, and straw (containing the internodes and leaves II-VI), and subjected to mineral element determination as described below.

Perls’ blue staining

Perls’ blue staining was performed on basal nodes. The 19-day-old seedlings of iet1 mutants and WT were cultured in a nutrient solution containing 0.2, 2, or 20 μM FeSO₄ for 2 weeks. The shoot basal region was sliced with LinearSlicer PRO10 (Dosaka EM). Cross-sections (100 μm) were placed on microscope slides and incubated with equal volumes of 4% (v/v) HCl and 4% (w/v) potassium ferrocyanide solution at room temperature for 1 h. The signal was observed under an optical microscope (VHX-S90BE; KEYENCE).

Short-term labeling experiment with ⁵⁷Fe

A short-term labeling experiment with stable isotope ⁵⁷Fe was performed. Three-week-old seedlings of WT and iet1 mutant lines were precultured in a nutrient solution with or without 2 μM Fe for 8 days, followed by exposure to 2 μM ⁵⁷Fe (96.1% ⁵⁷Fe, TRACE Sciences International), prepared by dissolving ⁵⁷Fe in 2 M HCl. After 24 h-exposure, the above-ground parts of each plant were separated into the shoot basal region, old leaves (leaf 2–5), leaf 6, and young leaves (leaf 7-8). The concentration of ⁵⁷Fe isotope in each organ was determined by ICP-MS as described below. The distribution ratio of ∆⁵⁷Fe within the above-ground part was calculated based on the accumulation in different organs.

Mineral nutrient elements concentration determination

All samples were dried at 70 °C for at least 3 days. Dried samples were digested with concentrated HNO₃ (60%) at a temperature up to 135 °C. The concentration of mineral nutrient elements in the digested solution was determined by inductively coupled plasma-mass spectrometry (ICP-MS, 7700X; Agilent Technologies). The concentrations of ⁵⁷Fe were determined with isotope mode. ΔFe (net Fe increase) was calculated according to Yamaji et al. (2013).

Statistical analysis

Data are presented as mean ± standard deviation (SD) (n = 3–6). Differences among treatments were tested using Tukey’s test and two-tailed paired Student’s t-test.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

Author contributions

J.C. and J.F.M. conceived and designed the experiments. J.C., S.H., Y.T.Q., Y.Y., C.T., T.M., Z.Y.L., N.Y. and J.F.M. performed the experiments. J.C., S. H., Y.T.Q., T.M., Z.Y.L., N.Y., R.F.S. and J.F.M. analyzed the data. J.C., S. H. and J.F.M. wrote the manuscript. All authors discussed the results and commented on the manuscript.

Peer review

Peer review information

Nature Communications thanks the anonymous reviewer(s) for their contribution to the peer review of this work. A peer review file is available.

Data availability

The RNA-seq data generated in this study have been deposited in the DDBJ BioProject under accession number PRJDB37649. The experimental data that support the findings of this study are available from the corresponding author upon request. The source data for Figs. 1, 4, 5, 6, 8, 9, 10 and Supplementary Figs. 3, 4 are provided as a Source Data file. Source data are provided with this paper.

Competing interests

The authors declare no competing interests.

Supplementary information

The online version contains supplementary material available at 10.1038/s41467-025-64863-4.