The forms of iron available to plants are the ferric iron cation (Fe³⁺), or ferric-ion chelates (Fe³⁺-chelates), and the ferrous iron cation (Fe²⁺) (Figure 1). Two main strategies of root iron acquisition are described: strategy I (the reduction-based strategy), and strategy II (the chelation strategy) (Curie and Briat, 2003; Curie et al., 2009; Kobayashi and Nishizawa, 2012; Connorton et al., 2017). Strategy I is found in non-graminaceous monocots and dicots. It relies on the reduction of Fe³⁺ by a ferric reduction oxidase (encoded by a FRO gene), and the incorporation of the resulting Fe²⁺ into the root by an iron-regulated transporter (encoded by an IRT or a RIT gene). The pH is decreased in the rhizosphere (Hinsinger et al., 2003), as a result of proton extrusion by plasma membrane proton pumps (encoded by an AHA gene); this acidification increases Fe³⁺ solubility.

Strategy II is found in grasses. It relies on the excretion of phytosiderophores (PSs, structural derivatives of mugineic acid) by a transporter of mucigenic acid (encoded by a TOM gene) that chelates Fe³⁺ (Fe³⁺-PS) before incorporation into the root by an oligopeptide transporter belonging to the Yellow Stripe/Yellow Stripe Like family (YS/YSL) (Curie et al., 2009).

Differentiating plant species according to their iron uptake strategy has its own limitations, as both strategies are found in rice, barley and peanut (Ishimaru et al., 2006; Pedas et al., 2008; Xiong et al., 2013). In Arabidopsis (a strategy I plant), chelating agents (e.g., coumarins) may contribute to iron nutrition in addition to the reduction strategy, especially in alkaline environments (Fourcroy et al., 2014; Schmid et al., 2014; Schmidt et al., 2014). These fluorescent phenolic compounds (FPCs) are synthesized via the phenylpropanoid pathway and secreted via an ABC transporter (e.g., PDR9 in Arabidopsis). Chlorotic phenotypes of IRT1 and FRO2 mutants were not restored by FPCs from plant exudates, suggesting that strategy I is implied in the uptake of Fe³⁺chelated to PC (Fourcroy et al., 2016). However, based on a review of results obtained under phosphate- (Pi-) deficient conditions, Tsai and Schmidt (2017) hypothesized the existence of an auxiliary IRT1-independent iron chelate transporter (ICT) that would bypass the IRT1 uptake system and internalize the Fe-FPC complex under Pi-deficient conditions. Flavins, another family of phenolic compounds, also increased iron solubilization in the rhizosphere of other plant species (e.g., barrel medic, sugar beet) than Arabidopsis (Rodríguez-Celma et al., 2013; Sisó-Terraza et al., 2016). Iron solubilization by phenolic compounds has also been reported in rice, a strategy II plant (Bashir et al., 2010; Ishimaru et al., 2011). Besides PSs and FPCs, organic acids (e.g., citrate or succinate) chelate Fe³⁺ (Fe³⁺-OC) and contribute to plant iron nutrition (Abadía et al., 2011; Adeleke et al., 2017). Synthetic ferric chelates (Fe³⁺-SC)—e.g., Fe-EDDHA or Fe-EDTA –, may provide iron to strategy I plants; entire chelates have been found in roots and shoots (Orera et al., 2009, 2010). These small hydrophilic molecules are suspected to use the transpiration stream as the driving force of entry (Abadía et al., 2011). Iron uptake by the leaves follows similar mechanisms as those described in the roots (Malhotra et al., 2019). Thus, even if strategies I and II remain essential pathways for iron acquisition by plants, they are not necessarily exclusive and may be complemented by additional mechanisms.

Iron is a central cofactor of enzymatic reactions involving electron transfer in essential metabolic pathways such as respiration or photosynthesis (Balk and Schaedler, 2014). Yet, its redox properties may also induce toxic effects. Free Fe²⁺ catalyzes the Fenton reaction in which reactive oxygen species (ROS) characteristic of oxidative stress are generated, and ROS may cause irreparable damage to cellular components when they are present in excessive concentrations (Winterbourn, 1995). The intracellular concentrations and forms of iron are therefore tightly regulated at the cellular level. The mechanisms involved in iron homeostasis have many common features in strategy I and strategy II plants (Connorton et al., 2017; Grillet and Schmidt, 2019; Kobayashi et al., 2019). In planta, chelation of Fe²⁺ to Nicotianamine (NA), and Fe³⁺ to citrate, allows iron mobilization and the control of its high reactivity (Curie et al., 2009; Connorton et al., 2017). In strategy II plants, ferric iron is additionally present in the form of Fe³⁺-PS (Zhang et al., 2019). Iron is also stored in unreactive forms such as ferritins that represent a major iron pool in plants mostly found in chloroplasts and mitochondria. They behave like a buffer that stores iron to avoid overload and the resulting ROS formation, and releases it when needed. To avoid oxidative stress, iron storage in vacuoles also contributes to iron homeostasis. Candidate transporters for moving cytosolic iron into vacuoles are members of the IRT, FPN/IREG (ferroportin/iron regulated) and VIT (vacuolar iron transporter) families (Morrissey et al., 2009; Roschzttardtz et al., 2009; Vert et al., 2009). VIT transporters are particularly important for vacuolar iron storage in seeds (Roschzttardtz et al., 2009; Zhang et al., 2012). In seeds, iron is mainly found under an insoluble form poorly available for nutrition because it is complexed with phosphate bound to inositol in phytates (Mary et al., 2015). Iron remobilization from the vacuoles is mediated by members of the natural resistance-associated macro-phage protein family (NRAMP) (Curie et al., 2000; Nevo and Nelson, 2006). In addition to ferritins and vacuoles, the root apoplast appears as a third level of iron storage by plants (Curie and Mari, 2017). Bienfait et al. (1985) demonstrated that a pool of 500–1,000 nanomoles of Fe per gram fresh weight could be formed in the root apoplast where it is adsorbed on the cell wall whose net charge is negative (Shomer et al., 2003). The cell wall composition, more particularly the respective proportions of celluloses, hemicelluloses, pectins, and lignins, varies depending upon plant genotypes, and influences the amount of adsorbed iron (Chen, 2014; Shi et al., 2018). The plant capacity to store iron in its root apoplast is positively correlated to the amount of hemicelluloses in the cell wall and would be a key determinant in the IDC tolerance of graminaceous plant species such as maize (Shi et al., 2018). IDC tolerance in soybean is also associated with an increased amount of root apoplastic iron (Longnecker and Welch, 1990). Remobilization of precipitated apoplastic iron relies on phenolic compounds (Jin et al., 2007; Bashir et al., 2011; Ishimaru et al., 2011; Lei et al., 2014), putrescine synthesis, and decreased cell wall suberization (Zhu et al., 2016; Curie and Mari, 2017).

Proteins involved in iron acquisition—F6′H1, PDR9, and members of the HA, FRO, and IRT family—also mediate cellular iron trafficking (Connorton et al., 2017). Members of the oligopeptide transporter (OPT) protein family (e.g., YS/YSL transporters) are also key determinants of iron transport in planta (Su et al., 2018; Grillet and Schmidt, 2019; Kumar et al., 2019). This transporter family is particularly important for the transport of Fe to the seeds (Grillet et al., 2014; Curie and Mari, 2017; Su et al., 2018; Kumar et al., 2019).

The identification of plant traits mediating plant iron uptake and iron homeostasis in planta offers opportunities for plant breeders to promote iron nutrition and content in agricultural products (Waters and Sankaran, 2011).

Conventional plant selection and new breeding strategies are both applied to enhance iron acquisition, storage and nutritional availability in edible parts of crops. There exists a natural genetic variation in the level of expression of the mechanisms regulating iron uptake and homeostasis among plant genotypes. The efficiency of plant iron nutrition is highly variable across plant species (Hansen et al., 2006). The level of tolerance or the susceptibility to IDC highly differs according to plant species and even to cultivars (e.g., Gildersleeve and Ocumpaugh, 1989; Zribi and Gharsalli, 2002; Mahmoudi et al., 2009; Helms et al., 2010). The Fe levels in grains can vary significantly depending on cultivars (e.g., from 10 to 160 mg/kg in maize, 15–360 mg/kg in wheat, 23–105 mg/kg in pea and 34–157 mg/kg in bean) (White and Broadley, 2005, 2009). Assimilability of Fe for human beings correlates positively with the iron content in edible parts of crops (Welch et al., 2000) and varies with the forms of iron. The most assimilable forms of iron in plants are Fe²⁺-nicotianamine (Fe²⁺-NA) and Fe³⁺-ferritin (Zielińska-Dawidziak, 2015; Beasley et al., 2019). In addition, iron nutritional availability is decreased by antinutrient molecules (e.g., phytates and tannins) that hamper its assimilation. Sufficient iron availability in food products is only possible when the concentration of these molecules is low (Sandberg, 2002; Delimont et al., 2017). Conventional breeding has led to the selection of cultivars showing better resistance to iron stress (i.e., IDC tolerant) and a higher iron content in edible parts, but also a decreased content of antinutrients (reviewed in Garcia-Oliveira et al., 2018). This was the case in species displaying high natural variability in their iron content (e.g., bean and pearl millet) (Manwaring et al., 2016; Lockyer et al., 2018). The selection of IDC-tolerant cultivars has further improved yields under iron stress conditions. However, the corresponding selection process relies on a long and costly screening of inbred lines.

Taking that limitation into account, transgenesis has been proposed as an option to promote plant iron nutrition and content by overexpressing or silencing genes mediating plant iron acquisition, transport and/or storage. The corresponding strategy has been followed through the targeting of one gene or several ones in combination and has led to genotypes with an increased iron content (i.e., from <2 to 6-fold) (Kawakami and Bhullar, 2018; Connorton and Balk, 2019). However, transgenesis raises public concerns (Lassoued et al., 2018). Furthermore, positive effects recorded in controlled conditions may be lost in field conditions because iron bioavailability varies among soils (Gregory et al., 2017). For example, iron uptake by soybean was increased by overexpressing FRO in controlled iron stress conditions (Vasconcelos et al., 2006) but not in high-calcareous soil environments (Kocak, 2014); even more, this genetic transformation appeared to be deleterious under non-iron stressed conditions due to toxic effects of the iron overload (Vasconcelos et al., 2006).

More generally, results from cultivars obtained from conventional and new breeding strategies vary depending upon soil iron bioavailability (Gregory et al., 2017; Garcia-Oliveira et al., 2018; Lockyer et al., 2018; Connorton and Balk, 2019). Alternative strategies based on QTL (quantitative trait loci) identification and on genome-wide association (GWAS) have been proposed to identify putative traits and genes mediating plant iron nutrition, and include them in plant breeding programs. The first step of this strategy confirmed the importance of genes implied in (i) iron uptake strategies I and II, (ii) the synthesis of phenolic compounds, and (iii) iron homeostasis. They further underlined the multigenic character of traits related to the plant iron status and the crucial importance of environmental conditions (Garcia-Oliveira et al., 2018; Connorton and Balk, 2019). Gene expression profiling of soybean plants sensitive or tolerant to IDC pinpointed key roles for phenylpropanoids (Waters et al., 2018). The major contribution of iron storage in the root apoplast and of fluorescent phenolics to remobilize this extracytoplasmic iron was confirmed and represents potential breeding targets (Curie and Mari, 2017; Waters et al., 2018). The complex and interregulated mechanisms of plant iron uptake and homeostasis has also been emphasized. A key role has been given to (i) phosphorus known to be in close relation with the iron status (e.g., Vansuyt et al., 2003; Tsai and Schmidt, 2017; Shi et al., 2018; Filiz and Kurt, 2019), (ii) ISR through the root-specific transcription factor MYB72 and beta-glucosidase BGLU42 (Zamioudis et al., 2014), and (iii) hormone signaling, especially IAA, Et, NO and ABA signaling (Lei et al., 2014; Li et al., 2015; Curie and Mari, 2017; Filiz and Kurt, 2019). The complex interrelations between the plant iron status, the P status, defense reactions and hormone signaling make the promotion of plant nutrition via plant breeding a difficult task.

The rhizosphere microbiota impacts the physico-chemical properties of the root environment by acidifying the soil through the release of organic acids and protons, and chelating iron with organic acids and siderophores (Figure 1B➊, ➋). These modifications prompt iron extraction from the soil matrix and thus modify its solubility and availability for the host plant.

Solubilization of iron in the rhizosphere is promoted by acidification. Protons are released during microbial activities such as nitrification (Kuypers et al., 2018). Protons may also be released from carboxylic groups when the pH of the soil solution is higher than the pKa of organic acids exuded by microorganisms (Glasauer et al., 2003). The protons released by microbial and plant activities acidify the rhizosphere (Hinsinger et al., 2003; Norton and Ouyang, 2019). Iron initially bound in scarcely soluble minerals (e.g., hematites, goethites) and amorphous solids [e.g., Fe(OH)₃] is replaced by protons at the sorption sites and released in the soil solution (Figure 1B➊; Albrecht-Gary and Crumbliss, 1998; Glasauer et al., 2003). Acidification also results from phosphate solubilization (Sharma A. et al., 2013). Organic acids are produced by rhizosphere bacteria that solubilize phosphates (e.g., Pseudomonas, Bacillus, Rhizobium, and Enterobacter) (Werra et al., 2009; Adeleke et al., 2017). Fe and P are often sequestered in soils together in low-solubility minerals like strengite or phosphosiderite. Thus, increased solubility of iron is associated with increased solubility of P (Marschner et al., 2011; Rijavec and Lapanje, 2016).

Solubilization of iron in the rhizosphere is also promoted by its chelation with organic acids, and by siderophores that scavenge ferric iron immobilized in scarcely soluble or insoluble forms and make it available to plants (Figure 1B➋; Kraemer, 2004; Jin et al., 2010; Ferret et al., 2014). The great majority of aerobic microorganisms synthesize small molecules with a high affinity for ferric iron—called siderophores—for their nutrition in iron stress conditions. Microbial siderophores (MSs) present high but variable affinity for Fe³⁺, and are also diverse in size and chemical composition (Budzikiewicz, 2004; Hider and Kong, 2010; Saha et al., 2016; Khan et al., 2019). It has long been known that plants use iron chelated to MSs for their nutrition in Fe-limiting conditions; in particular, pyoverdines (pvds), a major class of siderophores produced by fluorescent pseudomonads, show a high affinity for ferric iron (reviewed by Crowley, 2006; Vansuyt et al., 2007; Jin et al., 2010; Shirley et al., 2011; Nagata et al., 2013; Radzki et al., 2013; Trapet et al., 2016). Rhizoferrin, ferrocrocin, fusigen, and coprogen, all produced by fungal root symbionts, also display high affinity for ferric iron (Winkelmann, 2017; Haselwandter et al., 2020). Microorganisms are expected to be highly competitive for Fe compared to plant roots because they can (i) use Fe bound to phytosiderophores (PSs) (microbial siderophores like pyoverdine have much higher affinity for Fe than PSs do), (ii) decompose PSs, and (iii) acquire iron more efficiently (Figure 1B⓫; Marschner et al., 2011; Sadrarhami et al., 2021). However, while Fe-pvds are more stable than Fe-PSs, they do not depress plant iron nutrition but, even more, promote it (Vansuyt et al., 2007; Jin et al., 2010; Shirley et al., 2011) in contrast with the early report of Becker et al. (1985b). The mechanisms underlying the beneficial effect of microbial siderophores on plant nutrition remain to be elucidated, even if some insights have been given (Vansuyt et al., 2007; Gonzàlez-Guerrero et al., 2016). Organic acids present much lower affinity for iron than siderophores do, but in circumneutral and alkaline environments such as calcareous soils, organic acids may be deprotonated and thus act as metal-complexing agents (Dehner et al., 2010). Hydrogen cyanide (HCN) produced by microorganisms may also contribute to iron mobilization by chelation (Frey et al., 2010; Rijavec and Lapanje, 2016).

Availability of soil nutrients, including Fe, can also be enhanced by increasing the volume of soil explored (Figure 1B➓). This is achieved by root fungal symbionts which greatly extend the scope of the roots through their fine hyphae. Increases in plant iron content have been ascribed to a better access to soil nutrients via fungal networks (Caris et al., 1998; Liu et al., 2000; Farzaneh et al., 2011). In addition, iron is transported into root cells by endosymbionts (Gonzàlez-Guerrero et al., 2016).

Rhizosphere microorganisms modulate plant iron uptake mechanisms. The expression of genes involved in strategies I and II is modulated in the presence of microorganisms (Figure 1B➏; Zhang et al., 2009; Nagata et al., 2013; Nagata, 2017; Kobae et al., 2014; Zhou et al., 2016, 2018; Wang et al., 2017; Kabir et al., 2020). Rhizosphere acidification and ferric reductase activity (implied in strategy I) and plant PS synthesis (implied in strategy II) are enhanced (Reza, 2017; Wang et al., 2017; Zhou et al., 2018; Prity et al., 2020). The production of plant phenolics known to impact plant iron uptake and remobilization (Figure 1B➎; Fourcroy et al., 2016; Curie and Mari, 2017; Waters et al., 2018) and iron assimilability in food due to their antinutrient properties (Delimont et al., 2017) is also modified. Fluorescent pseudomonads induce the ISR (Van Loon et al., 2008; Berendsen et al., 2015) which regulates the expression of the root-specific transcription factor MYB72 and the MYB72-controlled beta-glucosidase BGLU42 (Zamioudis et al., 2014). These factors control the synthesis and excretion of iron-mobilizing FPCs in Arabidopsis (Palmer et al., 2013). Coumarins improve plant performance by eliciting microbe-assisted iron nutrition (Harbort et al., 2020). The concentration and composition of phenolic compounds in edible parts of plants is regulated by associated microorganisms (e.g., Basha et al., 2006; Lavania et al., 2006; Baslam et al., 2011; Jain et al., 2014; Singh et al., 2014; Baker et al., 2015).

Microorganisms modulate plant hormone signaling, which in turn impacts the plant iron physiology by modulating iron acquisition and homeostasis (Figure 1B➑; Lei et al., 2014; Li et al., 2015; Filiz and Kurt, 2019). Plant hormone signaling is under the control of microorganisms through the elicitation of the induced systemic response (ISR), a plant response interrelated with the plant iron deficiency response (Zamioudis et al., 2014; Romera et al., 2019). Changes in the plant indole acetic acid (IAA) and iron contents are observed concomitantly after bacterial inoculation, suggesting that microorganisms impact together hormone signaling and iron nutrition (Zhou et al., 2016, 2018; Housh et al., 2021). Emission of volatile organic compounds (VOCs) by Bacillus amyloliquefaciens promotes plant iron nutrition in Arabidopsis, and this promotion requires nitric oxide (NO) regulation (Wang et al., 2017), suggesting that the beneficial effect of the rhizosphere bacterial strain is regulated by plant hormone signaling. Rhizosphere microorganisms may either synthesize or degrade phytohormones [i.e., abscisic acid (ABA); IAA; gibberellic acid (GA); cytokinins (CKs); salicylic acid (SA); ethylene (Et); NO] and therefore modulate phytohormone concentrations (Horchani et al., 2011; Bakker et al., 2014; Egamberdieva et al., 2017; Ravanbakhsh et al., 2018). For example, the concentration of ethylene, a key regulator of root apoplastic iron remobilization under Fe shortage (Curie and Mari, 2017), is regulated by microbial 1-aminocyclopropane-1 carboxylic acid (ACC) deaminase that degrades the Et precursor (Ravanbakhsh et al., 2018).

The influence of microorganisms on the plant antioxidant defense has been associated to an increase of the plant iron content in a series of studies performed on sorghum, sunflower and alfalfa (Figure 1B➐; Kabir et al., 2020; Prity et al., 2020; Rahman et al., 2020). According to these authors, the promotion of ROS-scavenging activities by arbuscular mycorrhizal fungi (AMF) is part of the mechanisms involved in alleviation of Fe-deficiency symptoms.

Rhizosphere microorganisms can modify the plant iron status via their influence on the plant P status (Figure 1B➒; Tsai and Schmidt, 2017; Shi et al., 2018; Filiz and Kurt, 2019). Phosphate solubilizers and AMF have long been described to promote plant P nutrition and growth (Brown, 1974; Smith et al., 2011). More recently, microbial promotion of P nutrition was showed to impact iron partitioning in the roots and shoots of Thypha angustifolia (Ghosh et al., 2014), and to enhance iron nutrition in chicory through an AMF (Glomus irregulare syn. Rhizophagus irregularis; Labidi et al., 2012).

Other activities of rhizosphere microorganisms may also influence iron storage in the root apoplast and its remobilization. The hemicellulose composition of the root cell wall influences the amount of stored Fe, while putrescine, a diamine excreted by the roots, is involved in the iron remobilization process (Figure 1; Zhu et al., 2016; Shi et al., 2018). Since cellulose and hemicellulose are degraded by microbial activities, iron storage in the root apoplast is likely to be impacted by the corresponding microorganisms (Figure 1B➍; Lasa et al., 2019). Metatranscriptomic data indicate that the proportion of cellulose degraders is increased in the rhizosphere of cereals (Turner et al., 2013). Also, putrescine is one of the most commonly used substrate by wheat rhizosphere microorganisms (Gała̧zka et al., 2019): microbial degradation of this diamine (e.g., by pseudomonads) is thus likely to regulate plant remobilization of apoplastic iron (Kuiper et al., 2001; Song et al., 2015; Liu et al., 2018).

In sum, MSs play a key role in plant physiology related to iron uptake and homeostasis (Table 1 and Figures 1, 2) through (i) phosphorus solubilization and thus the plant P status (Sharma A. et al., 2013), (ii) elicitation of plant defense reactions through Microbial Associated Molecular Patterns (MAMPs) inducing ISR (De Vleesschauwer et al., 2006; Höfte and Bakker, 2007; Van Loon et al., 2008), (iii) plant hormone signaling and the synthesis of fluorescent root phenolics via ISR (Pieterse et al., 2014; Zamioudis et al., 2014), and (iv) the expression of genes mediating iron uptake and homeostasis (Table 1). The importance of microbial siderophores in the rhizosphere is also evidenced by results showing that their synthesis and activities are enhanced in the rhizosphere. Protein families related to siderophore production increased in barley root- and rhizosphere-associated bacterial taxa (Bulgarelli et al., 2015), and sequences encoding bacterial siderophore synthesis were highly enriched within bacterial endophytes in rice roots (Sessitsch et al., 2012).

In addition, discussion is running on the possible contribution of microbial siderophores to the remobilization of root apoplastic iron and in a non-reductive process of iron uptake by plants (Figure 1B➌). A large-scale transcriptomic study in Arabidopsis suggests that remobilization of root apoplastic iron is promoted by the pyoverdine synthesized by P. fluorescens strain C7R12 (Trapet et al., 2016). Plants grown in iron-deficient conditions in the presence of apo-pyoverdine (a siderophore uncomplexed with iron) exhibited a phenotype similar to that of plants grown in iron-containing medium and incorporated more iron than the untreated plants did. In these conditions, pyoverdine repressed the expression of root genes related to ABA signaling (Trapet et al., 2016), suggesting that the MS may promote the remobilization of root apoplastic iron (which indeed implies ABA regulation) (Lei et al., 2014; Curie and Mari, 2017). Possible non-reductive uptake of bacterial ferrisiderophores would represent an additional and major influence of microorganisms on the plant iron status. Various findings support the existence of a transport system whereby the plant would internalize bacterial ferrisiderophores in the form of integral ferric chelates. Pyoverdines of fluorescent pseudomonads chelated to iron (Fe³⁺-pvd) contribute to the plant iron nutrition of both dicots (strategy I) and graminaceous monocots (strategy II) more efficiently than the synthetic ferric chelate Fe³⁺-EDTA does (Vansuyt et al., 2007; Jin et al., 2010; Shirley et al., 2011). The stability constant of the Fe³⁺-pvd complex is significantly higher (10³²) than that of Fe³⁺-EDTA (10²⁵) or Fe-PS (10¹⁸) (Vansuyt et al., 2007; Shirley et al., 2011). This suggests that dissociation and ligand exchange between Fe³⁺-pvd and Fe³⁺-PS might not be the sole process accounting for the enhanced iron nutrition by Fe-MS and suggests incorporation of Fe³⁺-pvd by the roots. The presence of pvd in planta was confirmed by measurements of ¹⁵N-labeled Fe³⁺-pvd and by immunodetection using anti-pyoverdine antibodies (Vansuyt et al., 2007; Trapet et al., 2016). The use of IRT1 knock-out mutants indicated that this membrane transporter of Fe²⁺ (IRT1) is not involved in iron uptake from Fe³⁺-pvd in strategy I plants (Vansuyt et al., 2007). Thus, Fe-pvd clearly contributes to plant iron nutrition, with evidence of the presence of pvd in planta but no proof of the direct uptake of the entire Fe³⁺-pvd complex. The description of a non-reductive iron uptake system in a phytoplankton organism supports a possible incorporation of bacterial ferrisiderophores kept throughout evolution (Kazamia et al., 2018). This hypothesis is also supported by the presence of vesicles in the roots of Arabidopsis supplemented with Fe-pvd (Lemanceau et al., 2009); these vesicles mediate the internalization of ferrisiderophores by endocytosis in diatoms (Kazamia et al., 2018).