Selective induction of putative iron transporters, OPT8a and OPT8b, in maize by mycorrhizal colonization

Preparation of two-compartment system for Iron-59 (⁵⁹Fe)  uptake assay

To examine the Fe uptake via mycorrhizal pathway, we developed a two-compartment culture system which spatially separates the root compartment (RC) and hyphal compartment (HC) (Fig. S1). HC is a cylindrical column (90 mm in height; 35 mm in diameter) containing the same soil as RC, and separated from RC by a three-layered barrier comprising an outer polyethylene filter (122 µm opening, PE-180, Sansho Co. ltd, Tokyo, Japan), a medial polyethylene mesh (1.5 mm in thickness; 2-mm opening to make an air gap to prevent capillarity action, Netlon protector C-16, Lite corporation, Tokyo, Japan), and an inner nylon filter (42 µm opening, N-No.330T, Tokyo Screen Co. ltd. Tokyo, Japan) (Fig. S1a). The inner filter prevents roots from penetrating into HC and allows mycorrhizal hyphae penetration. The bottom of HC is covered with a polystyrene dish (10 mm in height; 37 mm in diameter) to prevent the water in HC from flowing out (Fig. S1a). The gap between the outer polyethylene filter and the bottom dishes was sealed with a commercially available hot melt adhesive. The HC was introduced into a 500-mL polyethylene pot (120 mm in height; 100 mm max. diameter; two 6-mm holes in diameter at the bottom) from the initiation of cultivation. The bottom of HC was placed about 30 mm above the bottom of the pot. Maize plants were grown in RC, and 10 mL aqueous solution containing Iron-59 (III) chloride (⁵⁹FeCl₃) (925 kBq) was gently applied to each HC at 14 days post planting (dpp) using a pipette.

Plant cultivation

In ⁵⁹Fe uptake analysis, soils and containers were prepared as follows. Akadama soil (tuff loam) (100 g, Setogahara Kaen, Gunma, Japan) was placed on the bottom of RC. About 1000 spores of Rhizophagus irregularis (formerly called Glomus intraradices DAOM 197198, Premier Tech, Riviere-du-Loup, Canada) was mixed thoroughly with upper-layer soil [Kanuma soil (weathered volcanic lapillus)/Ezo sand (small pumice)/Nippi soil (granular potting soil) (Nihon Hiryo, Tokyo, Japan) mixture (15:7:4, by weight)]. The total of 260 g upper-layer soil was placed on top of the Akadama soil. HC contained 52 g of the soil without fungal inoculant. Surface-sterilized seeds of maize (sweet corn, Kaneko Seeds, Co. ltd., Gunma, Japan) were sown in RC, and were watered every other day from the bottom by maintaining a water level less than 5 mm in depth. A single plant was left in each container and received 50 mL nutrient solution (1/2 strength Hoagland without Fe, pH 5.7) once a week from the bottom. HC were also watered (5–10 mL) every day from the top using a pipette when the soil surface was dried (Kanuma soil changes in color from dark yellow ocher to light yellow ocher when it is dried). Plants were cultivated in a growth chamber at 26°C with 16 h day/8 h night cycle.

Plants for gene expression analysis were cultivated using the same soil and the same pots as for ⁵⁹Fe uptake analysis. Each pot contained 250 g of soil mix as an upper layer and about 1000 spores of R. irregularis. The plants were watered every other day from the bottom by maintaining a water level less than 5 mm in depth. No nutrient solution was added. The cultivation was conducted in a greenhouse with a temperature of 16 h day/8 h night cycle at 26°C day/23°C night.

Measurement of ⁵⁹Fe radioactivity

Leaves (blades) of each position harvested at 44 dpp or soils from RC and HC were placed in test tubes. Soils of each compartment were homogenized by stirring and a portion (10–12 g) was subjected to measurement. ⁵⁹Fe activities in each sample were quantified using a well-type NaI(Tl) scintillation counter (2480 WIZARD², PerkinElmer, Massachusetts, USA). To detect the radioactivity of leaves, high-purity germanium detectors were used. Each leaf placed in test tubes was recovered, combined, dried for at least 3 d at 60°C, and subjected to gamma-ray spectrometry using germanium detectors (GAMMA-X(GMX) N-type coaxial HPGe, SEIKO EG&G, Tokyo, Japan).

Total RNA extraction and quantitative Reverse Transcription Polymerase Chain Reaction (RT-PCR)

Total RNA was extracted from mycorrhizal or non-mycorrhizal roots using NucleoSpin Plant RNA extraction kit according to the manufacturer’s instructions (Macherey-Nagel, Düren, Germany). The yield and purity of the RNA were determined spectrophotometrically (NanoDrop, Thermo Scientific, Wilmington, DE, USA). First-strand cDNA was synthesized from 100–290 ng of total RNA using PrimeScript^TM RT Master Mix (Perfect Real Time, Takara, Tokyo, Japan) according to the manufacturer’s instructions. Specific cDNA were amplified by SYBR Premix Ex Taq II (Takara) and gene-specific primers (Table S1) using real-time PCR (Thermal Cycler Dice, Takara). PCR reactions were one at 95°C for 30 s heat and 40 cycles of 95°C for 5 s and 60°C for 30 s. Specificity of each PCR product was assessed with the dissociation curve (95°C for 15 s, 60°C for 30 s, and 95°C for 15°C). The transcript levels of the target gene were calculated and expressed as fold expression in comparison to the constitutively expressed ubiquitin gene of maize (ZmUBQ, GRMZM2G118637) using the standard curve method.

AM fungi delivered ⁵⁹Fe to maize

We examined the Fe transport by AM fungi to maize using a two-compartment culture system (Fig. S1a and b). All leaves (blades and sheaths) of each plant were combined, dried and subjected to ⁵⁹Fe determination. ⁵⁹Fe was detected in the shoots of mycorrhizal plants, but not in non-mycorrhizal plants (Fig. 1). ⁵⁹Fe levels detected in the soils of RC were much lower than those of HC, and were similar between mycorrhizal and non-mycorrhizal treatments (Fig. S2), indicating that ⁵⁹Fe did not move from HC to RC. These results suggested that AM fungi took up Fe from the soils and delivered it to the host, similar to the observation by Caris et al. (Citation1998), who also demonstrated that mycorrhizal peanut (Arachis hypogea L.) and sorghum (Sorghum bicolor L.) could take up⁵⁹Fe via AM fungi.

Expression analysis of several genes involved in Fe homeostasis

To obtain insights into the possible mechanism of Fe uptake in mycorrhizal roots of maize, we examined the expression levels of 10 genes that are possibly involved in Fe uptake/responses. The selection criteria of 10 genes were as follows. (i) The transcript was observed in the genome-wide atlas of the transcription of Maize (eFP Browser, Winter et al. Citation2007; Sekhon et al. Citation2011). (ii) The gene was the ortholog or the close homolog of a functionally characterized gene of the other species. (iii) The gene was mainly expressed in roots. The details of the selection process are described in the Supporting Information (Gene information). The accession numbers of the genes, the expression levels in eFP Browser and the phylogenetic trees are provided in Table S2, and Fig. S3 and S4. Total RNA was isolated from roots at 20 dpp. Roots were also stained with wheat germ agglutinin-fluorescein isothiocyanate (WGA-FITC), confirming the successful colonization (Fig. S5). Quantitative RT-PCR (qRT-PCR) analyses showed that the symbiotic phosphate transporter ZmPT6, which expresses in mycorrhizal roots and is required for phosphate uptake via AM fungi (Willmann et al. Citation2013), was specifically expressed in the inoculated roots (Fig. 2). The transcript levels of ZmNASa, ZmNASf and ZmNAATa were significantly elevated in mycorrhizal roots compared to non-mycorrhizal roots (Fig. 2). However, the expression levels of ZmDMASa, ZmYS1, ZmIRTa, ZmIRTb and ZmIRO2a were similar between the two treatments. ZmIRO2b was not detected in our qRT-PCR (data not shown). Rice (Oryza sativa L.) IRO2 encodes a basic helix-loop-helix (bHLH) transcription factor which positively regulates various genes related to Strategy II (Ogo et al. Citation2007), and the expression of maize ortholog ZmIRO2 (Nozoye et al. Citation2013) is also strongly induced under Fe-deficient conditions. The soils used in this study (Kanuma soil, Akadama soil and Nippi soil) are rich in Fe; the leaves of neither non-mycorrhizal (Myc-) plants nor mycorrhizal (Myc+) plants showed any Fe-deficiency symptoms (Fig. S5a). Therefore, it is unlikely that ZmNASa, ZmNASf and ZmNAATa were induced by Fe deficiency. These results suggested that nicotianamine synthesis was transcriptionally up-regulated in mycorrhizal roots of maize, whereas DMA synthesis and its uptake in the form of Fe(III)-DMA might not be important for Fe uptake in mycorrhizal roots. Supporting this, a ys1 mutant inoculated with AM fungi showed an enhanced growth compared with non-inoculated ys1 (see supporting information for the materials and method, Fig. S6a), increasing the dry weight of shoots and roots 3.2 times and 2.6 times, respectively (Fig. S6b and c). Chlorophyll levels were increased in mycorrhizal plants (Fig. S7) and the total Fe contents in shoots and roots were also increased to 3.6 times and 1.6 times, respectively, as well as the increased accumulation of phosphorus (Fig. S8). Clark and Zeto (Citation1996a, Citation1996b) also reported similar results using acid and alkaline soils. We found that nicotianamine synthesis was likely to be up-regulated in maize mycorrhizal roots. Because it is thought that nicotianamine is not secreted but plays a role in the internal transport of Fe (von Wiren et al. Citation1999; Takahashi et al. Citation2003), nicotianamine synthesized in mycorrhizal roots may chelate Fe delivered by AM fungi within the roots or shoots.

Intriguingly, the transcript levels of oligopeptide transporter genes ZmOPT8a and ZmOPT8b in mycorrhizal roots were induced 194- and 62-fold, respectively (Fig. 2). The OPT family is thought to have important roles in metal uptake and translocation (Bogs et al. Citation2003; Wintz et al. Citation2003; Cagnac et al. Citation2004; Stacey et al. Citation2007; Pike et al. Citation2009; Hu et al. Citation2012). In rice, OsOPT1, OsOPT3 and OsOPT4 transport Fe-nicotianamine conjugates when heterologously expressed in yeast fet3fet4 defective in Fe uptake (Vasconcelos et al. Citation2008). It is thus possible that ZmOPT8s are involved in Fe delivery in mycorrhizal roots. In mycorrhizal roots, highly branched fungal structures, arbuscules, in which exchange of resources occurs between the fungus and the host plant, are formed in the cortical cells. Arbuscules are surrounded by plants’ periarbuscular membranes, on which symbiotic phosphate and ammonium transporters are localized to take up the nutrients (Harrison et al. Citation2002; Kobae et al. Citation2010). If Fe uptake also occurs at arbuscules, it is possible that ZmOPT8s are localized on periarbuscular membrane and implicated in Fe uptake. Alternatively, ZmOPT8s may have a role in long-distance transport of Fe in mycorrhizal plants. Considering the broad substrate specificity of the OPT family (glutathione, peptides of 4–13 amino acids, phytochelatin, heavy metals; Lubkowitz Citation2011), ZmOPT8s may also have other roles in addition to Fe delivery in mycorrhizal roots.