ZmYS1 Functions as a Proton-coupled Symporter for Phytosiderophore- and Nicotianamine-chelated Metals*

Among higher plants graminaceous species have the unique ability to efficiently acquire iron from alkaline soils with low iron solubility by secreting phytosiderophores, which are hexadentate metal chelators with high affinity for Fe(III). Iron(III)-phytosiderophores are subsequently taken up by roots via YS1 transporters, that belong to the OPT oligopeptide transporter family. Despite its physiological importance at alkaline pH, uptake of Fe-phytosiderophores into roots of wild-type maize plants was greater at acidic pH and sensitive to the proton uncoupler CCCP. To access the mechanism of Fe-phytosiderophore acquisition, ZmYS1 was expressed in an iron uptake-defective yeast mutant and in Xenopus oocytes, where ZmYS1-dependent Fe-phytosiderophore transport was stimulated at acidic pH and sensitive to CCCP. Electrophysiological analysis in oocytes demonstrated that Fephytosiderophore transport depends on proton cotransport and on the membrane potential, which allows ZmYS1-mediated transport even at alkaline pH. We further investigated substrate specificity and observed that ZmYS1 complemented the growth defect of the zinc uptake-defective yeast mutant zap1 and transported various phytosiderophore-bound metals into oocytes, including zinc, copper, nickel, and, at a lower rate, also manganese and cadmium. Unexpectedly, ZmYS1 also transported Ni(II), Fe(II), and Fe(III) complexes with nicotianamine, a structural analog of phytosiderophores, which has been shown to act as an intracellular metal chelator in all higher plants. Our results show that ZmYS1 encodes a proton-coupled broad-range metal-phytosiderophore transporter that additionally transports Fe- and Ni-nicotianamine. These biochemical properties indicate a novel role of YS1 transporters for heavy metal homeostasis in plants.

which are hydroxamate-, hydroxypyridonate-, catechol-, or aminocarboxylate-type chelators with a particular high affinity for Fe(III) (1,2). After mobilization of Fe(III), e.g. from sparingly soluble precipitates, siderophore-bound iron is subsequently taken up by specific transport systems, which have been well characterized in prokaryotic organisms (3). Yeast or dicotyledonous and non-graminaceous monocotyledonous plants increase proton secretion by H ϩ -ATPases as well as Fe(III) reduction capacity by ferric chelate reductases at the root plasma membrane and induce the Fe(II) transporter IRT1 (4). This reduction-based strategy, however, is prone to inhibition by high pH and elevated bicarbonate levels in calcareous soils, due to the pH dependence of chemical Fe(III) solubilization and enzymatic Fe(III) reduction (5). By contrast, graminaceous plant species constitutively express the reduction-dependent pathway for iron uptake at a comparatively low level, but enhance capacities for phytosiderophore release and Fe(III)-phytosiderophore uptake under iron deficiency (6). Physiological studies showed that Fe-phytosiderophore transport in graminaceous plants is unaffected by the presence of strong Fe(II) chelators, such as BPDS (6), but recognizes the backbone structure of mugineic acid-derived phytosiderophores, which are aminocarboxylate-type, hexadentate Fe(III) chelators (Refs. 1 and 7 and Fig. 1). In roots of maize highaffinity Fe(III)-phytosiderophore uptake is required to produce non-chlorotic, healthy plants, and is strongly dependent on the YS1 gene product (8). Transposon-tagging of the YS1 gene in maize allowed to identify ZmYS1 as a highly hydrophobic protein with 14 putative transmembrane-spanning domains that confers growth of the iron uptake-defective yeast mutant fet3 fet4 on phytosiderophore-bound iron, even in the presence of the Fe(II) chelator BPDS 1 (9). This study confirmed earlier uptake experiments with double-labeled Fe-phytosiderophores in the ys1 mutant indicating a stochiometric uptake of metal and ligand (10). The non-reductive transport pathway of Fephytosiderophores seems especially effective at alkaline pH and very low concentrations of Fe(III)-phytosiderophore complexes (11,12).
Although showing high affinity for ferric, phytosiderophores of the mugineic acid family do also chelate other heavy metals. Phytosiderophore-enriched root exudates efficiently mobilized copper, manganese, and zinc as well as other non-essential heavy metals (13,14). Moreover, uptake studies with labeled metal-phytosiderophores indicated that barley roots possess uptake activities for phytosiderophorecomplexed zinc and copper (7,15) and that ZmYS1 is responsible for the import of Zn(II)-phytosiderophores in maize roots (16). Thus, the plant phytosiderophore system could potentially also serve for the acquisition of metals other than iron, which is supported by the observation that phytosiderophore release is not only triggered under iron-deficient growth conditions, but also induced under zinc deficiency (15,17). In plants, metal-phytosiderophore uptake is not only important for covering the micronutrient demand of the plant, but might also be important for employing plants in future for phytoremediation purposes.
In the dicotyledonous plant Arabidopsis thaliana eight ZmYS1 homologs (AtYSL1-8) have been identified (9). Since the release of phytosiderophores has so far not yet been observed in dicotyledonous plants, it is possible that YSL proteins function to transport metal-nicotianamine (NA) chelates. Nicotianamine, which occurs in all higher plants, is structurally similar to phytosiderophores and serves as a precursor in their biosynthetic pathway (Ref. 18 and Fig. 1). A strong link between NA and iron metabolism is reflected in the NA-less tomato mutant chloronerva that shows severe symptoms of iron deficiency due to a deletion in a gene encoding nicotianamine synthase (19). Furthermore, NA has been shown to chelate not only Fe(III) but also Fe(II), in particular at higher pH, which may allow it to serve as an intracellular Fe(II) scavenger thereby protecting the cell from Fe(II)-mediated oxidative damage (20). The recent success of chemical synthesis of NA (21) allowed us also to investigate metal-NA chelate transport.
In order to understand iron-phytosiderophore transport under alkaline conditions and to characterize metal and ligand specificity by the ZmYS1 transporter in maize, we expressed ZmYS1 in two heterologous systems and verified its transport properties in the ys1 maize mutant. We found that ZmYS1 mediates transport of phytosiderophore-bound Fe(III) and other metals by H ϩ -coupled cotransport and permeates also NA-complexed Fe(II), Fe(III), and Ni(II).
To support growth of the fet3 fet4 mutant, solid YNB medium contained additionally 30 M FeCl 3 , while the liquid YPD medium was acidified to pH 5.0 with HCl. Yeast was transformed with either pDR196 or pDR196-ZmYS1 using the LiAc method (27), and transfor-mants were selected on uracil-deficient YNB medium containing 0.1% arginine as nitrogen source and the appropriate supplements.
For growth tests a single colony from BY4741 and fet3 fet4 transformants was washed in 1 ml of 10ϫ TE (pH 7.5) and resuspended in sterile MQ water to obtain an OD of 0.09. Cells were streaked on uracil-deficient YNB medium, containing 0.1% arginine, 3% glucose, 0.01% tryptophane, histidine, methionine, leucine, and the respective iron source. For complementation of the zap1 mutant, cells were streaked on a modified LIM medium (28) without sodium citrate, zinc sulfate, and ammonium, but supplemented with 0.1% arginine, 3% glucose, 0.01% histidine, methionine, and leucine, 50 mM MES/Tris (pH 7.0) and the respective zinc source.
For uptake experiments single colonies of transformed BY4741 cells were cultured in liquid YNB medium containing 3% glucose, the appropriate supplements and 30 M NiSO 4 , harvested at mid-log phase, washed at 4°C in ice cold uptake buffer consisting of 3% glucose, 40 mM NaH 2 PO 3 (pH 6.5) and resuspended on ice in uptake buffer to obtain OD 12. Before the uptake experiment, cells were incubated for 15 min at 30°C (500 rpm). Uptake was started by addition of 1 volume of solution containing 3% glucose, 100 mM MES/Tris (pH 6.5) and the respective 63 Ni(II)-chelate (specific activity 143 Ci mol Ϫ1 ). 100 l of uptake reaction was diluted in 10 ml of 10 mM EDTA (pH 6.5) and filtered (Whatman, 1822 025) before analyzing filter membranes by scintillation counting.
Electrophysiological Studies in Xenopus laevis Oocytes-Capped cRNA was transcribed from pOO2-ZmYS1 in vitro using the mMessage mMachine kit (Ambion, Austin, TX), after linearization of the plasmid with Eco72I. Oocyte experiments were performed as described in (29). Metal-chelate-induced currents in oocytes were detected by two electrode voltage clamp 3 days after cRNA injection.
Preparation of Metal Chelates-Deoxymugineic acid (DMA) was produced, and NA was chemically synthesized as described previously (16,21). Fe-DMA was prepared by mixing appropriate amounts of a 10 mM FeCl 3 solution, pH Ͻ2, and MES/Tris buffer and, if not indicated otherwise, with a 10-fold excess of DMA for 2-3 h at RT. Fe(II)-NA was obtained by mixing a freshly prepared Fe(II) stock solution consisting of 10 mM FeCl 2 and 20 mM ascorbate in 200 mM MES/Tris, pH 7.0 or 7.5, with the appropriate amount of a 10 mM NA solution and MES/Tris buffer, pH 7.0 or 7.5, followed by 10 min of incubation at 65°C. According to this procedure successful complex formation had been verified (20,35). Formation of the Fe(II)-NA complex was monitored by disappearance of the purple Fe-ascorbate complex. The Fe(III)-and Ni(II)-NA complexes were prepared similarly by dissolving FeCl 3 or NiSO 4 in 10 mM HCl to obtain a 10 mM metal solution, and mixing with the appropriate amount of MES/Tris buffer and NA, followed by incubation at 65°C for 10 min. In all solutions NA was in 20% excess if not indicated otherwise. All chelate solutions were centrifuged or filtered (0.2 m, Schleicher & Schuell) to remove precipitated iron hydroxides. Iron concentrations of supernatants or filtrates were verified by AAS or scintillation counting.

59
Fe-labeled Fe-phytosiderophore Uptake Studies in Maize-Maize plants were precultured and subjected to uptake studies using 59 Felabeled Fe-DMA (8,16). CCCP was added from 1 mM or 10 mM stock solutions in methanol to uptake solutions to final concentrations of 10 or 100 M (solutions w/o CCCP contained methanol alone). Uptake solutions were buffered with MES/Tris or HEPES.

Fe-Phytosiderophore Transport in Maize Roots
Is pH-dependent-Since graminaceous plant species are dependent on Fe(III)-phytosiderophore uptake in particular at alkaline soil pH, uptake of 59 Fe-labeled Fe(III)-DMA was compared in wild type (Alice) and ys1 mutant plants at different pH values of the uptake solution. At alkaline pH (pH 7.5) uptake rates of Fe(III)-DMA were 5-fold higher in wild type maize compared with ys1 mutant plants. However in wild type maize decreasing pH caused an increase in uptake which was almost 2-fold higher at pH 4.5 compared with pH 7.5 ( Fig. 2A). The majority of this Fe(III)-DMA uptake could be ascribed to ZmYS1, since uptake rates in the ys1 mutant were five to ten times lower and did not change with external pH. To confirm a pH dependence of Fephytosiderophore uptake further, the effect of the proton uncoupler CCCP on root uptake of labeled Fe(III)-DMA was tested in Fe-deficient wild-type plants. Addition of CCCP at concentrations of 10 and 100 M led to a 10 -20-fold decrease in Fe(III)-DMA uptake (Fig. 2B), suggesting a dependence of Fe(III)-DMA transport on the proton gradient across the root plasma membrane.
Transport Properties of ZmYS1 in Yeast-ZmYS1 has been shown to complement growth of the fet3 fet4 mutant (9), which is defective in high and low affinity iron transport and cannot grow on Fe-limited medium (30). As expected, fet3 fet4 cells transformed with pDR196-ZmYS1 grew well on medium containing 7.5 M Fe(III)-2Ј-DMA, whereas transformants with the empty vector did not (Fig. 3A). Growth complementation was pH-dependent. ZmYS1 transformants appeared earlier (data not shown), and colony number increased with decreasing pH (Fig. 3A).
To test ligand specificity of ZmYS1-mediated transport, ZmYS1-transformed fet3 fet4 cells were grown on NA-chelated iron. After 7-days growth complementation was achieved by 40 M of either Fe(II)-NA or Fe(III)-NA (Fig. 3B), while growth was not complemented using 7.5 M Fe-NA (data not shown). With supply of 30 M FeCl 3 differences between yeast transformants disappeared (Fig. 3C).
Zn(II)-phytosiderophore transport was investigated employing the zinc uptake-defective yeast strain zap1, which carries a deletion in the zinc-responsive transcriptional activator Zap1 (31) and does not grow on modified LIM medium containing less than 400 M zinc when transformed with pDR196 alone (Fig. 3D). Transformation of zap1 with pDR196-ZmYS1, however, permitted growth on 0.7 mM Zn(II)-DMA as zinc source, showing that ZmYS1 also transports phytosiderophore-chelated zinc (Fig. 3D).
To investigate ZmYS1-mediated transport of nickel complexes, we first examined time-dependent uptake of 63 Ni-DMA in the wt yeast strain BY4741. Little uptake was detected at 4°C for either ZmYS1-expressing or control cells (Fig. 4A). At 30°C, ZmYS1 expression led to an almost linear increase in 63 Ni accumulation over 20 min, while control cells hardly accumulated any 63 Ni. Nickel(II)-DMA uptake activity stopped immediately after addition of CCCP, while cells receiving the solvent alone continued to accumulate 63 Ni(II)-DMA (Fig. 4B). Uptake activity of 63 Ni(II)-NA was also CCCP-sensitive and increased almost linearly over time, though at a 3-fold lower rate (Fig. 4C). Concentration-dependent uptake of both nickel complexes in a range between 1 and 300 M yielded saturable kinetics, and Lineweaver-Burk transformation allowed calculatation of a K m of ZmYS1 for Ni(II)-NA of 165 M compared with a K m of 19 M for Ni(II)-DMA (Fig. 4D). Hence nickel was transported by ZmYS1 in a similar manner to iron, i.e. as a DMA-or NA-chelate and in a CCCP-sensitive manner.
Electrophysiological Characterization of ZmYS1-mediated Transport in Oocytes-To study the transport mechanism employed by ZmYS1 in more detail, the protein was expressed heterologously in Xenopus laevis oocytes. Oocytes were voltageclamped to Ϫ70 mV in choline-based buffer solution at pH 6.0 and then superfused with buffer containing different concentrations of Fe(III)-DMA. In oocytes injected with cRNA encoding ZmYS1, currents induced by Fe(III)-DMA corresponded to the net influx of positive charge, reversed after withdrawal of substrate and were voltage and concentration-dependent (Fig.  5A). Currents were absent in water-or non-injected control oocytes (data not shown). Fe(III)-DMA-dependent currents increased with more negative voltage and increasing substrate concentrations up to ϳ25 M Fe(III)-DMA (Fig. 5B). Thus, Fe(III)-DMA transport was saturable, and the Fe(III)-DMA concentration that permitted half-maximal uptake was between 5 and 10 M with the tendency to decrease at more positive voltage (Fig. 5C). A decreasing K m with less negative voltage indicates binding of a negatively charged substrate to the transport protein.
Fe(III)-DMA-induced currents strongly depended on external pH. Highest currents were obtained at pH 5.0 and gradually decreased by increasing the pH to 9.5 (Fig. 6A). Plotting substrate-induced currents against H ϩ concentrations showed that Fe(III)-DMA-induced currents increased up to proton concentrations of 100 nM, corresponding to a pH of 7, but saturated at higher proton concentrations (Fig. 6B). Replacing choline by sodium (100 mM, pH 9.5) did not stimulate transport, indicating that sodium could not substitute protons to drive metalchelate uptake (Fig. 6A).
Since yeast uptake experiments provided evidence that ZmYS1 might also transport other heavy metals chelated with phytosiderophores ( Fig. 4), different metal-chelates were tested for the induction of currents in ZmYS1-injected oocytes. When supplied at 100 M external concentration, Ni(II)-, Zn(II)-, and Cu(II)-DMA provoked currents of similar size to those achieved with Fe(III)-DMA (Fig. 7A). By contrast, Mn(II)-and Cd(II)bound DMA induced significantly smaller currents and negligible currents were induced by the unchelated ligand. Ferrous and ferric complexes with the synthetic chelator EDTA or unchelated metals (data not shown) induced little if any current. Since ZmYS1 might be saturated at Ͻ100 M, currents were also assayed at 5 M corresponding to the K m for Fe(III)-DMA (Fig. 5C). Currents induced by Ni(II)-, Zn(II)-, and Cu(II)-DMA were smaller relative to those induced by Fe(III)-DMA, while currents induced by Mn(II)-and Cd(II)-DMA approached zero (Fig. 7B), indicating that Cd-and Mn-phytosiderophore transport by ZmYS1 will be relatively insignificant in the presence of more competitive metals like iron, nickel, zinc, and copper. Smaller currents induced by 5 M Ni(II)-DMA compared with 5 M Fe(III)-DMA were consistent with the higher K m value of ZmYS1 for Ni(II)-DMA (Fig. 4D) compared with Fe(III)-DMA (Fig. 7A). Currents induced by DMA-chelated Fe(III) were then compared with NA-chelated iron at a concentration of 5 M, clearly showing that ZmYS1 transports Fe(III)-DMA more efficiently than Fe(II)-or Fe(III)-NA (Fig. 7B). Together with the observation in yeast, where ZmYS1-mediated uptake of Ni(II)-DMA was higher than uptake of Ni(II)-NA, we concluded that metal-DMA complexes are more favored substrates for ZmYS1 compared with metal-NA complexes. At higher concentration and more negative membrane potential a discrimination between Fe(II)-and Fe(III)-NA transport could be achieved. Currents induced by 100 M Fe(II)-NA increased steeply with voltage and were 3-fold larger than those for Fe(III)-NA (Fig.   FIG. 2. Fe(III) 7C). The induction of currents by NA complexes with both iron forms clearly agreed with the growth complementation of ZmYS1-transformed yeast cells on Fe(II)-and Fe(III)-NA (Fig. 3B).

ZmYS1
Mediates H ϩ -coupled High Affinity Fe-Phytosiderophore Transport-In three different systems we observed that ZmYS1-mediated Fe(III)-DMA transport depends on the availability of protons: (i) in maize roots Fe(III)-DMA uptake was severely inhibited by CCCP, and the contribution of ZmYS1 to transport of the iron chelate increased with the concentration of external protons (Fig. 2); (ii) in yeast, low pH values improved growth complementation by ZmYS1 of the iron uptake-defective yeast mutant fet3 fet4 (Fig. 3A) and uptake of 63 Ni(II)-DMA was strongly inhibited by CCCP (Fig. 4B); and (iii) oocytes expressing ZmYS1 exhibited a positive inwardly rectifying and voltage-dependent current in response to supply with negatively charged Fe(III)-DMA and this current was proton dependent (Figs. 5 and 6). Considering the almost stochiometric uptake of double-labeled 59 Fe-and 65 Zn-14 C-DMA by maize roots (8,16), we conclude that ZmYS1 represents a H ϩ -Fe(III)-phytosiderophore cotransporter. These data provide the first description of the transport mechanism employed by a member of the OPT family of oligopeptide transporters. In view of the pH and CCCP sensitivity of oligopeptide transport in yeast by ScOPT1 (32,33), which belongs to a different cluster within the OPT family (34), it is likely that energization by H ϩ -cotransport also holds true for other OPT members.
Proton cotransport is surprising in view of the environmental conditions, under which the uptake of Fe(III)-phytosiderophores has to take place. The release of phytosiderophores by graminaceous species is enhanced as a response to low availability of soluble iron, which naturally occurs at high pH, such as is found in calcareous soils. In contrast to plants employing reduction-dependent iron acquisition, iron deficiency in grasses shows little if any correlation with elevated bicarbonate concentrations (12), since iron acquisition does not rely on an Fe(III) reduction step. Our investigations uncovered another factor that contributes to more efficient iron acquisition in grasses: coupling the transport of Fe-phytosiderophores, which are negatively charged at neutral pH (35), to protons allows efficient root uptake even at high soil pH, because transport of the proton-coupled substrate can be driven as long as the overall difference of the electrochemical potential remains negative. This is emphasized by the Fe-DMA-induced currents in ZmYS1-expressing oocytes at pH 9.5, where currents appeared already at membrane potentials below Ϫ120 mV (Fig. 6A).
Since the membrane potentials of root cells in grasses are usually between Ϫ120 and Ϫ180 mV, Fe-phytosiderophore uptake can still be driven until the availability of protons as a co-substrate becomes limiting. In oocytes proton availability limited Fe-DMA transport between pH 7 and 9.5 (Fig. 6B); but in soil-grown plants proton limitation probably occurs even at higher soil pH, because P-type ATPase-mediated proton extrusion may generate a local pH gradient favoring substrateproton cotransport (36). Thus, even at high proton buffering capacities in bicarbonate-rich soils, the electrochemical potential drives accumulation mainly by generating a strongly negative membrane potential. Under these conditions ZmYS1 can exhibit a high substrate affinity (K m of 5-10 M, Fig. 5C) that would address the micromolar phytosiderophore concentrations found in the rhizosphere of grasses (11).
Several plant transporters employ proton cotransport for negatively charged substrates, such as sulfate, nitrate or amino acids (37,38). Fe(III)-DMA complexes are negatively charged at neutral pH, while at pH 5 the majority of complexes is uncharged (35). The slight tendency of ZmYS1 to increase the affinity for DMA-complexed Fe(III) at more positive voltage (Fig. 5C) tempts us to speculate that the negatively charged substrate and protons bind to different sites of ZmYS1 prior to transport. Since Fe(III)-DMA transport generated a positive inward current, at least two protons are cotransported with each complex. Interestingly, the net current with different saturating metal-complex concentrations was similar irrespectively of the valency of substrate charge (Fig. 7A). As phytosiderophore complexes with divalent cations such as copper, zinc, or nickel, are predicted to carry two negative charges at neutral pH (35), the positive inward current demands that at least three protons are cotransported with these substrates.
ZmYS1 Shows a Broad Metal Specificity-Growth complementation and oocyte transport studies clearly showed that ZmYS1 also transports phytosiderophore-chelated Zn(II). This has been substantiated in planta by lower uptake rates of 65 Zn-phytosiderophores in the ys1 mutant compared with wild type maize (16). Since phytosiderophore release is also enhanced under zinc deficiency in several graminaceous species, zinc-phytosiderophore transport by ZmYS1 most likely represents one part of a concerted adaptive response to low zinc availability. Thus, phytosiderophore release, metal chelation and chelate uptake apply for iron and zinc, which is ecologically reasonable, since poor availability of both metal micronutrients can limit plant growth on calcareous soils (12). However, ZmYS1 also transported phytosiderophore-chelated nickel and copper at a similar rate to iron and zinc (Fig. 7A), although low availabilities of nickel and copper are apparently not limiting growth factors in calcareous soils. The specificity of phytosiderophore-mediated metal chelation and transport might thus be regarded as low, suggesting that metal-chelate transport by ZmYS1 primarily depends on the probability of complex formation and on regulation of ZmYS1. Since metal-phytosiderophore transport by ZmYS1 shows a similar poor selectivity for iron, as does metal transport by the major iron transporter IRT1 (30,39), for which a complex regulation at the transcript level by iron and cadmium and at the protein level by iron and zinc was found (40), it might be expected that ZmYS1 also responds to the availability of multiple metal elements. The broad range of metal transport therefore suggests that regulation of ZmYS1 by other metals is required as a prerequisite for plants to maintain metal ion homeostasis. On the other hand, the broad metal specificity of ZmYS1 opens up the possibility that the demand of other micronutrients, such as nickel, might also be satisfied by the phytosiderophore system and paves the way to employ phytosiderophore-releasing species for phytore- mediation. Since phytosiderophore-overproducing rice plants have already been generated and shown to increase iron acquisition on calcareous soils (41), we suggest to test such engineered plant lines also for the hyperaccumulation of other heavy metals.
ZmYS1 Also Transports Fe-and Ni-nicotianamine-In contrast to phytosiderophores NA is not only synthesized by graminaceous plant species but by all higher plants and appears essential for chlorosis-free plant growth (19,42). Due to the structural similarity of NA with phytosiderophores, we reasoned that ZmYS1 or ZmYS1-related proteins might also transport Fe-NA. Indeed, in both heterologous systems ZmYS1 transported also NA-chelated iron and Ni(II) (Figs. 3B, 4, and  7). Fe-and Ni-NA transport by ZmYS1 is of particular physiological importance, since NA is believed to fulfill a physiological role, which is different from phytosiderophores. Based on the formation of stable complexes between NA and Fe(II) or Fe(III) (20), the occurrence of NA in aqueous extracts of leaves and roots (43) and the likely accumulation of Fe-NA complexes in the cytoplasm and the vacuole (44), NA may find its physiological role in iron chelation for the intracellular delivery of iron and long-distance trafficking of iron, zinc, and copper (44). NA-chelated Fe(II), which exhibits a relatively high kinetic stability (20), could even provide the opportunity for membrane transport of a ferrous iron chelate by maintaining a low risk for the generation of free superoxide radicals compared with the transport of free ferrous iron as mediated by IRT-or NRAMPfamily transporters (39,45). Since Arabidopsis possesses eight ZmYS1-like genes (AtYSL1-8) (9) but does not synthesize phytosiderophores, transport of Fe-NA might represent a major function for this transporter family in dicotyledonous plant species. Thus, YS1 and YSL proteins might act in a complementary manner to IRT and NRAMP transporters, since membrane transport could be mediated without involving a reduction step.
An alternative physiological role of YS1 and YSL proteins is indicated by our finding that ZmYS1 mediates transport of Ni(II)-NA (Fig. 4). Supported by the observation that one NA synthase gene from maize, ZmNAS3, is up-regulated by iron sufficiency, which provides evidence for NA synthesis independent of phytosiderophore biosynthesis (46), a novel role for NA in heavy metal transport and detoxification is indicated. Indeed, recent investigations on the physiological basis of nickel tolerance in the hyperaccumulator species Thlaspi caerulescens identified enhanced chelation of Ni(II) by NA in the xylem as a response to toxic levels of external nickel (47). This response implies the existence of membrane transporters for NA-chelated Ni(II), that are most likely represented by YS1 and YSL homologs. Future experiments should therefore focus on the membrane localization of YS1 and YSL proteins in graminaceous and dicotyledonous plants as well as on overexpression of YS1 and YSL genes, which might become an important foundation for future phytoremediation strategies.