Mutually exclusive alterations in secondary metabolism are critical for the uptake of insoluble iron compounds by Arabidopsis and Medicago truncatula

Abstract

The generally low bioavailability of iron in aerobic soil systems forced plants to evolve sophisticated genetic strategies to improve the acquisition of iron from sparingly soluble and immobile iron pools. To distinguish between conserved and species-dependent components of such strategies, we analyzed iron deficiency-induced changes in the transcriptome of two model species, Arabidopsis (Arabidopsis thaliana) and Medicago truncatula. Transcriptional profiling by RNA sequencing revealed a massive up-regulation of genes coding for enzymes involved in riboflavin biosynthesis in M. truncatula and phenylpropanoid synthesis in Arabidopsis upon iron deficiency. Coexpression and promoter analysis indicated that the synthesis of flavins and phenylpropanoids is tightly linked to and putatively coregulated with other genes encoding proteins involved in iron uptake. We further provide evidence that the production and secretion of phenolic compounds is critical for the uptake of iron from sources with low bioavailability but dispensable under conditions where iron is readily available. In Arabidopsis, homozygous mutations in the Fe(II)- and 2-oxoglutarate-dependent dioxygenase family gene F6'H1 and defects in the expression of PLEIOTROPIC DRUG RESISTANCE9, encoding a putative efflux transporter for products from the phenylpropanoid pathway, compromised iron uptake from an iron source of low bioavailability. Both mutants were partially rescued when grown alongside wild-type Arabidopsis or M. truncatula seedlings, presumably by secreted phenolics and flavins. We concluded that production and secretion of compounds that facilitate the uptake of iron is an essential but poorly understood aspect of the reduction-based iron acquisition strategy, which is likely to contribute substantially to the efficiency of iron uptake in natural conditions.

Figure 1.

The transcriptome of Arabidopsis and M. truncatula roots as affected by iron deficiency. Numbers in white circles denote differentially expressed genes and their orthologs in both species.

Figure 2.

Flow scheme for the construction of an ortholog coexpression network comprising differentially expressed genes in roots from M. truncatula and Arabidopsis. See text for a detailed description. [See online article for color version of this figure.]

Figure 3.

Coexpression network of differentially expressed genes in Arabidopsis and M. truncatula as affected by iron deficiency. The network was constructed with the MACCU toolbox following the procedure shown in Figure 2. Square boxes and red edges indicate genes and coexpression relationships from the M. truncatula data set, and diamond-shaped boxes and blue edges denote genes and coexpression relationships derived from the Arabidopsis data set. Oval boxes refer to genes that were common in both networks. Yellow and blue boxes indicate up- and down-regulated genes, respectively. Gray boxes denote genes that were common in both data sets but oppositely regulated in M. truncatula and Arabidopsis. Green circles indicate the presence of the hhbhAAACCAAv motif in the promoter, and yellow circles denote regulation by FIT.

Figure 4.

Effects of iron availability on the growth of the wild-type, f6′h1-1, and pdr9-2 mutant plants. A to F, Wild-type (A and D), f6′h1-1 (B and E), and pdr9-2 (C and F) seedlings were grown on avFe media [10 µm Fe(III)-EDTA, pH 5.5; A–C] and navFe (10 µm FeCl₃, pH 7.0; D–F). G, Seedling fresh weight (FW) under the different iron regimes. H, Chlorophyll content. I, Iron content. Error bars indicate se from five replicates. J to O, Quantitative reverse transcription-PCR analysis of the expression of iron-responsive genes in leaves (J) and roots (K–O). Three independent replicates were performed for each sample. The ΔΔC_T (cycle threshold) method was used to determine the relative amount of gene expression, with the expression of elongation factor 1 used as an internal control. Error bars represent se. Statistical differences were assessed by Duncan test (P < 0.05). Col-0, Columbia-0.

Figure 5.

Visualization and relative quantification of fluorescent phenolic compounds produced and excreted by Arabidopsis roots. Fluorescence pictures were taken before (p) and after (np) removing the plants from the agar using 365 nm as excitation wavelength and 9 s as exposure time (A). Plants were grown for 6 d on avFe media or on media with navFe. Relative quantification of phenolic compounds was based on pixel average intensity (p.a.i.) in equal areas around plants (B) and used for calculating the distribution of phenolics (C). Statistical differences were assessed by Duncan test (P < 0.05). col-0, Columbia-0.

Figure 6.

Effects of pH-regulated Rbfl secretion on iron concentrations and growth in M. truncatula. A, Fresh weight of leaves, stems, and roots. B, Leaf iron concentration. C to E, Plants grown on +Fe [45 µm Fe(III)-EDTA, pH 5.5; C], iron-deplete media (0 µm Fe, pH 5.5; E), or iron-deplete media adjusted to pH 7.7 (0 µm Fe, pH 7.7; D). F, Growth media of iron-deficient plants after 6 d of growth, showing yellowing of the nutrient solution caused by Rbfl and Rbfl derivatives secretion. Measurements were performed after 6 d of treatment. Data represent means of five replicates, and error bars denote se. Statistical differences were assessed by Duncan test (P < 0.05).

Figure 7.

Partial rescue of the f6′h1-1 (A) and pdr9-2 (B) phenotype by growth alongside Arabidopsis or M. truncatula plants. Percentages of total chlorophyll (Chl), iron content, and leaf fresh weight (FW) of mutant plants grown alone or in mixed population with wild-type (col-0) or M. truncatula (Mt) plants when compared with col-0 under navFe conditions. Significant differences (asterisks) were assessed by t test (P < 0.05).

Figure 8.

Mechanisms and regulation of the reduction-based iron acquisition mechanism. The transcription factors FIT and bHLH38/39 regulate proton extrusion, Fe(III) reduction, and uptake of Fe(II). In addition, FIT and bHLH38/39 control the mutually exclusive induction of the phenylpropanoid pathway in Arabidopsis and flavin synthesis in M. truncatula, leading to the production of putatively iron-binding compounds (IBCs) that are excreted via specific transporters in the plasma membrane. The consensus sequence hhbhAAACCAAv (indicated by green circles) is likely to be involved in the regulation of the pathways. Iron-binding compounds chelate Fe(III) from nonsoluble ferrihydroxides. The resulting chelates are split by reduction via FRO2, and the released ferrous iron is taken up via IRT1. Yellow circles indicate genes that are regulated by AtFIT.