Many trace metals are essential micronutrients, but also potent toxins. Due to natural and anthropogenic causes, vastly different trace metal concentrations occur in various habitats, ranging from deficient to toxic levels. Therefore, one focus of plant research is on the response to trace metals in terms of uptake, transport, sequestration, speciation, physiological use, deficiency, toxicity, and detoxification
Essential trace elements have to be transported across membranes for uptake, long-distance transport and intracellular sequestration, which is often accomplished by highly specific transport proteins (see Figs 1–۳). For this reason, the current review is grouped by metals. Nevertheless, not all transporters are completely specific, but ions with similar chemical properties often share some of the transport routes.
Fe acquistion and transport have recently been covered in a number of excellent reviews, including those of Kobayashi and Nishizawa (2012), Thomine and Vert (2013), and Curie and Mari (2017). Despite being the most abundant transition metal in the earth’s crust, in aerated soils Fe is poorly bioavailable because it is always present in the trivalent form, which precipitates and forms poorly soluble hydroxides and oxides that are not readily available for plant uptake (Lindsay and Schwab, 1982). To ensure Fe aquisition, two distinct physiological strategies have evolved in plants, one based on Fe(III) reduction (Strategy I) and the other on the chelation of Fe(III) (Strategy II) (Marschner and Römheld, 1994). All plants, with the exception of grasses, rely on the reduction-based Fe acquisition Strategy I, which involves an acidification of root apoplast and rhizosphere, mediated by P-type ATPases (Santi and Schmidt, 2009). An acidification by one pH unit increases Fe(III) solubility by a factor of 1000. Fe3+ in the soil solution is then chelated by phenolic compounds, particularly catechol-type coumarins, that are released by the roots (Schmid et al., 2014). Release of chelators may also be uncoupled from Fe limitation and under control of Pi starvation (Tsai and Schmidt, 2017). Subsequently, chelated Fe(III) gets reduced to Fe(II) by a Fe(III) chelate reductase (Yi and Guerinot, 1996), which is encoded by FRO2 in A. thaliana (Robinson et al., 1999). This enzyme reduces Fe(III) with electrons derived from cytosolic NADH via a haem cofactor. The reductase has a low pH optimum and thus depends on the increased H+-ATPase activity (Susin et al., 1996). Exudated coumarins are also likely to contribute to Fe(III) reduction in the apoplast (Schmid et al., 2014). After reduction, Fe2+ is translocated across the plasma membrane of rhizodermal and cortical cells by proteins of the ZIP family, primarily IRT1, a rather unspecific divalent ion transporter (Eide et al., 1996; Vert et al., 2002). Under Fe-limiting conditions, the Fe acquisition machinery, including IRT1 and FRO2, as well as coumarin synthesis and release, is transcriptionally up-regulated by FER/FIT and other basic helix–loop–helix (bHLH)-type transcription factors (see ‘Effects of and strategies against non-optimal metal nutrition: limitation and toxicity’). In addition, IRT1 is under tight post-transcriptional control and undergoes constant turnover (Barberon et al., 2011). The transporter is not constitutively localized in the plasma membrane, but is mostly present in the trans-Golgi network/early endosomes, from where it moves to the distal side of the plasma membrane for Fe uptake or to the lytic vacuole for degradation (Barberon et al., 2011). Trafficking to the vacuole requires mono-ubiquitination of the transporter and is induced by the availabiliy of non-Fe substrates of IRT1, such as Zn and Mn (Barberon et al., 2014). This mechanism probably serves to balance Fe acquisition and trace metal stress under Fe deficiency (see ‘Effects of and strategies against non-optimal metal nutrition: limitation and toxicity’).
In graminaceous monocots (grasses), another mechanism, called Strategy II, has evolved. This includes the synthesis and release of phytosiderophores, which are metal ligands belonging to the mugineic acid family. They are synthesized from methionine via NA (Römheld, 1991; Negishi et al., 2002). Although both monocots and dicots produce NA, only grasses produce phytosiderophores. The critical step of their biosynthesis is accomplished by the nicotianamine aminotransferase (NAAT). Accordingly, expression of a barley NAAT in rice increased tolerance to Fe-deficient conditions (Takahashi et al., 2001). Membrane transporters for both uptake and release of phytosiderophores have been identified. Mugineic acid is released by the TOM1 transporter (Nozoye et al., 2011), which is related to the NA-transporting ZIF1 protein (see ‘Effects of and strategies against non-optimal metal nutrition: limitation and toxicity’, Zinc). Both transporters belong to the Major Facilitator Superfamily. Hexadentate chelates of the phytosiderophores with Fe3+ are taken up by the root cells via Yellow Stripe1 (YS1 in maize) and YSL (e.g. in rice) proteins (Curie et al., 2001), named after the chlorotic phenotype of the ys1 mutant that cannot take up Fe–phytosiderophore complexes (Basso et al., 1994; Von Wirén et al., 1994). YS1 mediates H+-coupled Fe3+–phytosiderophore symport, and its expression, like that of TOM1, is up-regulated upon iron stress.
As a grass, rice belongs to the Strategy II plants, but also employs the Strategy I-related IRT1 transporter, which is transcriptionally up-regulated under Fe limitation (Bughio et al., 2002; Ishimaru et al., 2006). Therefore, rice is able to use a combination of both strategies, albeit it has a relatively poor phytosiderophore release. Rice mutants with a defect in the NAMT are unable to synthesize phytosiderophores, and become Fe limited when only Fe(III) is available (Cheng et al., 2007). Upon exposure to Fe(II), however, they do not suffer from Fe limitation; the Fe deficiency response is not induced, but Fe(II) is taken up. Unlike Strategy I plants, rice apparently does not perform Fe(III) reduction because Fe2+ is highly abundant in hypoxic paddy soils, which have a low redox potential (see ‘Effects of and strategies against non-optimal metal nutrition: limitation and toxicity’, Iron toxicity in rice).
Another Fe entry route apart from the ‘default transporters’ might operate during phosphate starvation. It is pH-independent and may have driven the speciation of plants inhabiting calcareous soils (see Tsai and Schmidt, 2017). Though the transporters involved are still unidentified, the strong down-regulation of IRT1 in Pi-deficient plants indicates a different Fe entry route. Further, the root architecture/morphology of Fe-limited and Pi-limited plants is very similar, with root growth correlated to internal Fe, but not Pi concentrations (Reymond et al., 2006).
Radial transport of Fe in the root is regulated by the endodermis, which initially forms an apoplastic discontinuity due to the lignin-impregnated radial walls of the Casparian strip. At later developmental stages, endodermal cells are completely enveloped by a suberin layer, blocking access of the endodermal cell’s plasma membrane to the external root apoplast. Interestingly, this suberization is delayed and more discontinuous under Fe deficiency and in irt1 or nramp1 mutants affected in Fe uptake (Barberon et al., 2016). The decreased extent of this blocking, or at least decelerating, layer under nutrient deficiency may improve nutrient uptake.
For transport into the shoot and distribution to various tissues, Fe is bound to a range of chelators, notably citrate and NA (Schuler et al., 2012). Fe(III)–citrate has been localized in the xylem, and the pH in the xylem would favour citrate over NA, which is the main chelator of Fe in the phloem (Stephen and Scholz, 1993; Durrett et al., 2007; Curie et al., 2009). The citrate transporter FRD3 (Ferric Reductase Defective 3) is localized in the plasma membrane of the pericycle and of cells surrounding the vascular tissue (Green and Rogers, 2004). Mutants with defects in this transporter have decreased citrate concentrations in xylem and shoot, and accumulate Fe in the root, indicating the requirement for FRD3 (in rice: OsFRDL1) for Fe translocation (Durrett et al., 2007; Yokosho et al., 2009). Involvement of the plant homologues of ferroportin (FPN1) expressed in the vascular tissue was discussed (Morrissey et al., 2009).
Expression of the Fe(III) chelate reductase gene FRO1 in leaves of pea suggests Fe(III) reduction also at the sink location (Waters et al., 2002). Phloem transport of Fe is required for seed loading, and, although not all transporters are fully discovered, members of the YSL family of Fe–NA transporters may be involved (DiDonato et al., 2004; Schaaf et al., 2005; Schuler et al., 2012). Plants that are mutated or silenced in the respective genes not only accumulate less Fe in the seeds, but the seeds also have reduced germination rates on soils with limiting bioavailable Fe (Le Jean et al., 2005). Studies on plants that overexpress the vacuolar ZIF1 transporter, which removes NA from the cytosol, indicate that this chelator also plays a crucial role in Fe distribution within the leaf (Haydon et al., 2012).
Although Fe is part of electron transport chains in chloroplasts and mitochondria, ~80–۹۰% of the total Fe is accumulated in the plastids (see ‘Functions of the metals in the plant’). Yeast complementation assays suggest involvement of the permease PIC1, which is localized in the inner chloroplast envelope, in Fe uptake by this organelle (Duy et al., 2007). Further, Fe(III) reduction by FRO7 seems to be required for Fe transport into the plastids (Jeong et al., 2008). Not only do plants with defective FRO7 accumulate less Fe in their plastids, leading to decreases in photosynthetic performance due to lack of Fe in the plastid, but fro7 mutants also have severe growth defects under Fe-deficient conditions and recover only when supplied with additional Fe.
Ferritin, the universal Fe storage protein, has different locations and different roles in animals compared with plants (Briat et al., 2010). In plants, it is mostly found in non-green plastids and developing tissues (Theil and Hase, 1993). However, localization in green plastids after stress and in mitochondria has also been shown, indicating a defence mechanism against Fe overload and resultant oxidative stress (Briat et al., 2010). Arabidopsis thaliana knock-out mutants devoid of ferritin in seeds, vegetative tissues, and reproductive organs (fer1-3-4) indicate that ferritins are not a Fe source, but rather involved in Fe detoxification (Ravet et al., 2009).
The electron transport chain in mitochondria also requires high amounts of Fe. For A. thaliana mitochondria, a molar ratio of 26:8:6:1 for Fe:Zn:Cu:Mn has been detected (Tan et al., 2010). How Fe enters the organelle is not yet fully elucidated. Similar to chloroplasts, reduction of ferric iron to ferrous iron may take place at the outer membrane. Proteins encoded by the FRO genes FRO3 and FRO8 have a predicted mitochondrial localization in A. thaliana (Mukherjee et al., 2006; Jeong and Connolly, 2009). However, the subcellular localization of the FROs in rice is not yet known (Vigani, 2012). Similarly, how iron crosses the mitochondrial outer membrane is not known; no transporters have been identified yet. In the inner membrane, the Fe-specific transporter MIT1 (Mitochondrial Iron Transporter 1) has been identified in rice (Bashir et al., 2011). MIT1 belongs to the conserved mitochondrial carrier family (MCF). These are small proteins of ~30 kDa, localized in the inner membrane of mitochondria, which were also identified in yeast, zebrafish, humans, and Drosophila (summarized in Jain and Connolly, 2013). ABC transporters may play a role in the export of Fe–S clusters (Kushnir et al., 2001). The mitochondrial protein frataxin seems to be involved in Fe homeostasis, acting as a chaperone and delivering Fe for synthesis of haem and FeS clusters (Ramírez et al., 2010; Maliandi et al., 2011). Mutants with disfunctional frataxin suffer from Fe toxicity, oxidative stress, and Fe accumulation in the mitochondria, while overexpression enhances resistance to oxidative stress (Møller et al., 2011).
In developing seeds, Fe needs to be unloaded from the symplast of the maternal tissue and reabsorbed by the embryo. In the apoplastic space between both tissues, Fe(III) is complexed by citrate and malate (Grillet et al., 2014). To allow uptake by the embryo as Fe2+, the chelated Fe(III) is reduced by ascorbate released by the embryo (Grillet et al., 2014), not by a membrane-bound Fe(III) chelate reductase, as in roots. After uptake by the embryo, Fe is stored in vacuoles of endodermal cells surrounding the vascular tissues (Lanquar et al., 2005). Loading of those vacuoles is mediated by the tonoplast-localized VIT1 protein (Kim et al., 2006). During germination, this Fe store is unloaded by transporters of the NRAMP family, NRAMP3 and NRAMP4 (Lanquar et al., 2005). Those transporters are essential for seedling establishment on calcareous soils with low Fe availability. VIT1 and NRAMP3/4 thus function as an import–export module for Fe (Mary et al., 2015). During imbibition prior to germination, the remobilized Fe is transiently sequestrated by the MTP8 transporter in vacuoles of subepidermal cells (Eroglu et al., 2017). This may serve as a buffer to prevent Fe overload.