Iron delivery to the growing leaves associated with leaf chlorosis in mugineic acid family phytosiderophores-generating graminaceous crops

ABSTRACT

Graminaceous cereal crops as well as all other green plants require iron (Fe), as Fe is a component of heme and [Fe-S] clusters in the photosystems (PSI, PSII) containing green pigments (chlorophylls). The interveinal yellowing (Fe chlorosis) of growing leaves is caused by insufficient Fe uptake and internal delivery. In the 1970s, Sei-ichi Takagi discovered Fe³⁺-chelating mugineic acid family phytosiderophores (MAs), which are released from graminaceous roots to directly absorb Fe-MAs complexes. However, the mechanisms underlying the Fe delivery from root cells to terminal interveinal mesophyll cells need intensive investigation. This review first overviews the roles of metal-chelating compounds, i.e., MAs and nicotianamine (NA), and the Fe-chelate transporters involved in primary Fe partitioning; then, the delivery of Fe to the growing leaves via phloem and into the developing chloroplasts/thylakoids via symplastic diffusion and membrane transport is discussed. Fe-MAs are absorbed into the root epidermis cells by YSL transporters and transformed to Fe-NA for symplastic radial movement between root parenchyma cells and into the xylem. If heavy metal ions such as Co²⁺ and Cu²⁺ are simultaneously present, competition may occur at their complex formation with NA prior to the radial movement that causes leaf chlorosis because of heavy metal-induced restriction of Fe availability. In xylem saps, large fractions of Fe form Fe-MAs, as in barley plants (Hordeum vulgaris), instead of Fe-citrate, which is predominant in rice plants (Oryza sativa). The former complexes are transferred to the phloem at the stem nodes for the growing leaves, while Fe-citrate is partitioned to the mature leaves by transpiration. Although the major routes of Fe supply to the growing leaves by phloem transport have been confirmed, the mechanisms underlying the synthesis of phloem Fe-compounds, delivery through the phloem system, unloading at the growing sink leaves and Fe trafficking and utilization in the chloroplasts/thylakoids still require intensive investigation. Thus, insufficient root uptake of Fe-MAs and insufficient phloem delivery of Fe to the growing leaves and finally to the chloroplasts may cause the reduced chlorophyll‒Fe-protein assembly (chlorosis). However, the schemes of iron phloem delivery presented in this review must be confirmed in future studies.

KEY WORDS:

• Chlorosis
• graminaceous crops
• iron-deficiency
• mugineic acids
• phloem delivery

1. Introduction: historical review of iron chlorosis in graminaceous plants

Iron (Fe) is an essential nutrient for all plants, including graminaceous cereal crops, such as rice (Oryza sativa L.) and barley (Hordeum vulgare L.). Fe is a cofactor of the respiratory cytochromes and aconitase in mitochondria and a main metal component of photosystems I and II (PSI, PSII) and cytochrome-b₆f (Cyt-b₆f), which collect solar energy to convert atmospheric CO₂ to carbohydrates in the green chloroplasts. A deficiency of Fe causes chlorosis in the growing parts of graminaceous plants (Figure 1).

Figure 1. Photos of Fe-deficient barley and rice plants grown hydroponically without Fe nutrition after cultivation with a complete set of nutrients to establish early plant development. The Fe concentrations in the shoots of barley and rice plants were 53 and 81 mg kg⁻¹ DW, respectively (photos courtesy of Dr. Takeshi Shimizu)

Although large quantities of Fe are present in soil as ferric (Fe³⁺) oxides and hydroxides, these forms are insoluble and scarcely available to crops, particularly under H⁺-deficient neutral and alkaline conditions. Fe in submerged anoxic soil is soluble in ferrous form (Fe²⁺) and available to plants such as paddy-grown rice.

About 100 years ago, leaf chlorosis, a condition characterized by yellowing of leaves due to a deficiency of chlorophyll, was observed in the young leaves of upland rice cultivated in the calcareous soil in what is now Puerto Rico (Gile and Carrero 1920). The chlorosis was induced in calcareous soils containing carbonates and in lime-amended soils. The leaves were cured of leaf chlorosis by spraying with a ferrous sulfate solution, but not by applying the same solution to soil, even in considerable quantities, since the ferrous form was immediately converted to the insoluble ferric form in calcareous soils.

Gile and Carrero (1920) also reported that when rice plants with slight chlorosis were submerged, they showed severe chlorosis several days later but they became a perfectly normal green color after 10 days of continuous flooding. Takagi (1966) later clarified the mechanism underlying such leaf-color change from the yellow of severe chlorosis to the normal green during flooding. Namely, he considered that the initial flooding dilutes the root-excreted Fe-solubilizing substances, while the continuous flooding increases the amounts and concentrations of such substances.

Willis and Carrero (1923) demonstrated that the liming of rice soils makes them alkaline and results in the conversion of the soil water-soluble ferrous form to an insoluble ferric form. Alkalization of the culture-solution pH is also induced in the calcium-nitrate nutrition but not in ammonium-sulfate nutrition. This is because the preferential uptake of nitrate by plants results in an excess of alkaline calcium ions, causing a high pH, whereas the preferential uptake of the ammonium from ammonium-sulfate results in an excess of acid, causing a low pH.

Leaf-interveinal chlorosis of upland rice plants was frequently observed in the farm areas (soil pH 6.5–7.3) near the Ashio Copper Mining site in Japan’s Gunma Prefecture, as well as in the Watarase Riverside fields, which were contaminated by copper drains from the Ashio Copper Mining site (Mitsui et al. 1958; Shionoya et al. 1959). The copper concentrations in the yellow leaves were as high as 30 mg g⁻¹ (DW), while their Fe concentrations fluctuated widely. The rice plants with chlorosis were occasionally re-greened by foliar sprays of a diluted ferrous sulfate (Shionoya et al. 1959).

Interveinal chlorosis developed quickly in the young leaves of hydroponically cultured rice plants when the plants were treated simultaneously with heavy metals and Fe-citrate. The chlorosis was most severe in the plants treated with excess cobalt (Co²⁺). Plants treated with high doses of Ni²⁺ or Zn²⁺ showed less severe chlorosis, whereas those treated with a high dose of toxic Cu, which is known to cause severe water stress (Chino 1967), did not exhibit any chlorosis at all. This type of chlorosis is known as ‘heavy metal-induced chlorosis,’ and the plants suffering from this condition are restored to a healthy green color by a foliar spray of diluted ferrous sulfate solution. Such heavy metal-induced chlorosis has also been observed in non-graminaceous plants (Hewitt 1948; Dekock 1956).

In the 1950s, Sei-ichi Takagi began investigations on the acute chlorosis induced in rice seedlings by water flooding (Takagi 1966). His pioneering studies in graminaceous plants led to the finding that under Fe starvation, these plants develop a unique physiological strategy: synthesis and excretion of the phytosiderophores known as mugineic acid family phytosiderophores (MAs) (Takagi, Kamei, and Yu 1988; Kawai, Takagi, and Sato 1988) from the root surface to the rhizosphere to solubilize insoluble Fe in soils. MAs also solubilize the root-apoplast-localized ferric oxides. Thus, the graminaceous plants developed a specific Fe-uptake system exclusively by forming Fe-MAs complexes (Takagi 1976; Takagi, Nomoto, and Takemoto 1984). However, due to the limited secretion of MAs, certain graminaceous crops such as rice and corn are susceptible to Fe shortage and subsequent severe chlorosis, whereas Fe-deficiency-tolerant species such as barley excrete large quantities of MAs.

The MAs released were 2ʹ-deoximugineic acid (DMA) from rice and corn plants and mugineic acid (MA) and its family phytosiderophores from barley and other grasses (Takemoto et al. 1978; Sugiura et al. 1981; Mino et al. 1983; Kawai, Takagi, and Sato 1988; Murakami et al. 1989; Ma and Nomoto 1996). MAs have the ability to specifically solubilize Fe³⁺ in soils at pH 4.0‒9.0, and rice and barley plants take up Fe-DMA and Fe-MAs, respectively, into their roots (Takagi, Nomoto, and Takemoto 1984). MAs are also synthesized and accumulated in the leaves for inter-organ Fe mobilization (Higuchi et al. 2001).

2. Uptake and transport of iron in the roots

As depicted in Figure 2, the MAs produced from nicotianamine (NA) in the root cells of graminaceous plants are secreted from the apical zone (Marschner, Römheld, and Kissel 1987; Yoshida, Kawai, and Takagi 2004) by rice transporter of MAs (TOM1) and HvTOM1 Fe-chelator effluxers (Nozoye et al. 2011). The Fe³⁺-MAs formed via equilibrium complexation with high stability constants (Guelke and von Blanckenburg 2007) or via prior exchanges with soil Fe(III)-MSs (microbial siderophores) (Yehuda et al. 1996) and Fe³⁺-HSs (humic substances) (Cesco et al. 2002) near the root surfaces are taken up by YS/YSL (yellow stripe/yellow stripe-like) Fe-uptake transporters such as ZmYS1 (Curie et al. 2001), HvYS1 (Murata et al. 2006), and OsYSL15 (Inoue et al. 2009; Lee et al. 2009) located in the plasma membrane of epidermal root cells. Leaf interveinal chlorosis appeared in the ZmYS1-deficient mutant (ys1), which releases DMA but cannot absorb Fe-DMA (von Wirén et al. 1994), and OsYSL15-less mutants of rice showed chlorosis and reduced Fe concentrations in their shoots (Lee et al. 2009). The Fe³⁺-MAs taken up are reduced to Fe²⁺-NA by cytosolic ascorbate (Weber, von Wirén, and Hayen 2008). The Fe-NA may be transferred via a symplastic radial movement through the plasmodesmata among parenchyma cells (Barberon and Geldner 2014) and finally unloaded into the xylem. The main carriers of Fe in the xylem fluids are citrate and MAs (Ariga et al. 2014): The citrate and MAs are released to the apoplastic xylem vessels by rice ferric reductase defective-like (OsFRDL1) citrate effluxer (Inoue et al. 2004; Yokosho et al. 2009) and barley Al-activated citrate efflux transporter (HvAACT1; Furukawa et al. 2007), and TOM2 (Nozoye et al. 2015), respectively, at the pericycle. The Fe-NA released to the acidic-apoplast xylem (pH 5.5‒6.5) may be ligand-exchanged to Fe-citrate and Fe-MAs as observed in xylem saps (Ariga et al. 2014). The NA present in xylem saps (Ariga et al. 2014) is released from root parenchyma cells by the NA exporter ENA1 (Nozoye et al. 2019).

Figure 2. Fe uptake at the root-cell plasma membranes and transport to xylem vessels in the roots (lower picture); upward transport through the stem xylem to mature leaves and xylem-to-phloem transfer to growing leaves (top picture) in graminaceous plants. AACT, Al-activated citrate efflux transporter; Cit, citrate; ENA, efflux transporter of NA; FRDL, ferric reductase defective-like; HS, humic substances; MAs, mugineic acid family phytosiderophores; MS, microbial siderophores; NA, nicotianamine; PD, plasmodesmata; TOM, transporter of mugineic acid family phytosiderophores; YSL, yellow stripe-like

3. Heavy metal-induced restriction of Fe transport in the root cells, resulting in leaf chlorosis

When rice seedlings grown hydroponically (pH 5.1) were treated with high doses of Co²⁺, Cu²⁺, and Mn²⁺ for 3 days, the treated plants exhibited a decreased transport of taken-up ⁵⁹Fe [supplied as ⁵⁹Fe-citrate simultaneously for 3 days] to the plant tops, and the levels of specific activity of ⁵⁹Fe accumulation in the tops and specifically in the growing leaves became smaller than those in the non-treated control plants (Table 1; Chino 1967). The appearance of chlorosis in the growing leaves was severe in the Co²⁺-treated plants and negligible in the Mn²⁺-treated plants, and was related to the smaller distribution of ⁵⁹Fe into the growing leaves in the former plants, although the ⁵⁹Fe accumulation in the roots was little affected by the treatment with high doses of heavy metals (Chino 1967).

Table 1. Effect of simultaneous treatment with Co, Cu, and Mn on the partitioning of ⁵⁹Fe to the roots, tops, and growing leaves

Treatment	^59Fe uptake		^59Fe-specific activity
	(cpm)		(cpm mg^−1 FW)
	Roots	Tops	Roots	Tops	Growing leaves
Control	216,053	22,709 (9.5) ^†	227	7.8	11.2
Co	295,302	2,710 (0.9)	298	1.0	1.2
Cu	148,115	3,368 (2.2)	209	2.0	3.3
Mn	309,540	9,310 (2.9)	336	3.4	4.0

The plants treated with 3.0 mg L⁻¹ Co as CoSO₄, 0.6 mg L⁻¹ Cu as CuSO₄, 60 mg L⁻¹ Mn as MnCl₂ and the controls (without Co, Cu, and Mn) were fed 3 mg Fe L⁻¹ as ⁵⁹Fe-labeled Fe-citrate for 3 days. Growth depression was observed in the Cu-treated plants, and chlorosis appeared in the growing leaves of the Co-treated plants.

^†The % distribution of whole ⁵⁹Fe uptake.

Adapted from Chino (1967).

In barley plants grown hydroponically with Fe supplied as Fe-K₂-EDTA, treatments with certain heavy metals caused interveinal chlorosis in growing leaves (Agarwala, Bisht, and Sharma 1977). The chlorosis was most severe in the plants treated with high doses of Co²⁺ or Cu²⁺, and was slight in those treated with high doses of Zn²⁺ or Mn²⁺. The extent of chlorosis was inversely related to the quantity of ⁵⁹Fe [supplied as ⁵⁹Fe-EDTA], which was partitioned to the shoots, with low levels of Fe concentrations also observed in growing leaves.

Heavy metal excess-induced chlorosis ranging from pale green to yellow/white in growing leaves was previously reported for both non-MAs-synthesizing plants such as sugar beet (Hewitt 1948), mustard (Sinapis alba; Dekock 1956) and soybean (Brown and Tiffin 1960) and MAs-synthesizing plants such as oat (Hunter and Vergnano 1953) and rice (Brown et al. 1953) by a similar order of heavy metals, i.e., Co > Cu > Zn > Mn, although this order was somewhat dependent on the plant species and heavy-metal physiology. The appearance of the chlorosis was well ordered, with stability constants of the metal complexes declining in the order Cu²⁺ > Co²⁺ > Fe²⁺ > Zn²⁺ > Mn²⁺, as given by Mellor and Maley (1948), Hewitt (1951) and Irving and Williams (1953).

A working model proposed for the heavy metal-induced Fe deficiency in rice seedlings, which was reported in Table 1, is depicted in Figure 3. In the rice plants, the ligand of Fe-citrate supplied to the culture solution is exchanged by the DMA secreted from the roots and the Fe-DMA complex formed is taken up by the OsYSL15 transporter (Murata et al. 2006; Inoue et al. 2009). Bivalent cations of Co, Cu, and Mn supplied simultaneously with ⁵⁹Fe may be taken up by IRT1 (Bughio et al. 2002; Ishimaru et al. 2006; Barberon and Geldner 2014) and Mn may be taken up by NRAMP5 (Ishimaru et al. 2012; Sasaki et al. 2012). No competition would occur between Fe-DMA and either Co²⁺, Cu²⁺, or Mn²⁺ during their uptake processes, as previously inferred by Chino (1967). The Fe³⁺-DMA taken up into the epidermal cells may be reduced to Fe²⁺ by ascorbate (Weber, von Wirén, and Hayen 2008), and the Fe²⁺ produced competes against Co²⁺, Cu²⁺, and Mn²⁺ during metal-NA complex formation, probably following the stability constants of the metal complexes stated above under a limited quantity of NA in the root-cell cytosols (Higuchi et al. 1999). The metal-NA complexes formed are transported via symplastic diffusion among the parenchyma cells (Barberon and Geldner 2014). Thus, the reduced excretion of Fe to the xylem may result in the reduced availability of Fe to the plant tops and finally to the growing leaves, where chlorosis due to the Fe-deficiency may occur (see Table 1).

Figure 3. Working model of heavy metal-induced restriction of iron transport in the radial symplastic diffusion through parenchyma cells in rice plants under treatment with excess doses of Co²⁺ (A), Cu²⁺ (B), and Mn²⁺ (C), which may compete with Fe²⁺ in the formation of complexes with NA. A decrease in the Fe quantity is marked by Δ. DMA, 2ʹ-deoxymugineic acid; HMA, heavy metal P-type ATPase; IRT, Fe-regulated transporter; NRAMP, natural resistance-associated macrophage protein. For Cit and YSL, see the legend of Figure 2

4. Iron delivery to the leaves via xylem

Both barley and rice plants exhibit interveinal chlorosis under Fe deficiency, but rice shoots have a higher Fe content than barley shoots (Figure 1). In a culture experiment by Maruyama et al. (2005), barley plants showed a higher chlorophyll index (SPAD value), i.e., a higher chlorophyll:Fe ratio in the leaves and a more efficient distribution of Fe to the growing leaves than rice plants, indicating that chloroplast Fe in barley leaves functions more efficiently than that in rice leaves (Hirai et al. 2007).

Fe-citrate and Fe-MAs in the xylem fluids ascend through the stems. Fe-citrate is largely delivered via transpiration to the mature leaves as in Arabidopsis (Schuler et al. 2012) and castor bean (Ricinus communis) (Hazama et al. 2015), whereas Fe-MAs may be efficiently transferred via xylem transfer cells to phloem in the stem nodes by YSL transporters such as OsYSL18 (Aoyama et al. 2009) and then by plasmodesmata, which are located between adjacent parenchyma and phloem companion cells, to the phloem sieve elements in a manner similar to amino-acid xylem-to-phloem transport by amino acid transporters (Tegeder 2014; Yoneyama et al. 2016) (Figure 2). The xylem-to-phloem transfer of Fe at the nodes is more active in barley plants (Tsukamoto et al. 2009) than in rice plants, since the quantity of Fe-MAs in barley xylem sap is large (Table 2). The preferential transport of Fe-MAs, regulated at nodal transfer cells and destined to the growing leaves in barley may be responsible for the more efficient use of Fe in the growing leaves than in rice. In rice, Fe is predicted to be re-translocated to the growing leaves via phloem, after its delivery to the mature leaves via xylem as citrate-bound Fe.

Table 2. The concentrations of iron (Fe)-containing complexes in xylem saps from young rice, barley, and castor bean plants grown on water-submerged (rice) or upland (barley, castor bean) soils

Plant species	Fe concentration
	(μM)
	Fe3-Cit3^†	Fe2-Cit2^‡	Fe-MAs^§	Fe ions	Total
Rice (cv. Nipponbare)	4.2	1.6	0.4	2.8	9.1 ± 1.0
Barley (cv. Amagi-Nijo)	8.4	6.2	8.7	1.5	24.9 ± 4.9
Castor bean (cv. Mizuma)	11.8	6.5	0	19.8	38.1 ± 5.1

†, Tri-Fe³⁺, tri-citrate complex. ‡, Binuclear Fe³⁺-citrate species.

§, The MAs for rice and barley were 4.5 μM DMA and 0.4 μM DMA plus 17.2 μM MA, respectively.

Adapted from Ariga et al. (2014).

5. Synthesis and delivery of phloem Fe-compounds

Figure 4 shows functional schemes of Fe nutrition in the mature and growing leaves in graminaceous plants. The site with the greatest demand for Fe nutrition is the growing leaves, where new organelles such as mitochondria and chloroplasts that contain prosthetic groups of heme and [Fe-S] cluster proteins are constructed. Sixty to ninety percent of the leaf iron is allocated to the chloroplasts (Terry and Abadía 1986; Wada et al. 2015; Kroh and Pilon 2020), and most of the Fe required in the growing leaves during their early developing period is actively supplied by phloem transport.

Figure 4. Schemes of the re-mobilization of Fe in mature leaves (lower picture) and utilization of Fe in growing leaves (top picture) in graminaceous plants. Cyt, cytochrome; FRO, ferric-chelate oxidoreductase; Fds, ferredoxins; LHC, light-harvesting chlorophyll (a/b) protein; MIT, mitochondrial Fe transporter; NiR, nitrite reductase; PD, plasmodesmata; VIT, vacuolar iron transporter. For Cit, PD and YSL, see the legend of Figure 2

5.1. Chemical forms of phloem Fe and other metals

Although the identification of metal compounds in the phloem sap is essential, analysis of the chemical forms (speciation) of the metals is still difficult because the phloem system, a coupling of enucleated sieve elements and companion cells, is usually located at the central sites of stems and leaf veins. We collected 0.5‒2 μL phloem sap from the cut stylets of plant hoppers from young rice plants according to the method of Kawabe, Fukumorita, and Chino (1980) and 5‒20 μL phloem sap of castor bean plants from the surface incision of the stems by the method of Hall and Baker (1972).

Our precise chemical analysis of rice and castor bean phloem saps (Table 3) showed nearly identical chemical forms (speciation) for Fe as well as Cd, Zn, and Cu. Fe was largely bound to DMA as Fe³⁺-DMA in rice and to NA as Fe^2+/3+-NA in castor bean, and the remaining Fe was bound to 10‒20 kDa proteins and citrate. Large fractions of phloem sap Cd in rice were bound to 13 kDa proteins, and a smaller fraction was bound to thio-compounds such as phytochelatins (PCs). Zn was predominantly bound to NA both in rice and castor bean plants. Cu was dominantly bound to NA with the remainder bound to histidine and proteins.

Table 3. Chemical forms of Cd, Zn, Fe, and Cu in phloem saps from rice and castor bean plants

Metal	Phloem sap from rice plants	Phloem sap from castor bean plants
Fe	Largely bound to DMA as Fe^3+-DMA, some bound to proteins and citrate.	Largely bound to NA as Fe^2+/3+-NA, some bound to proteins and citrate.
Cd	Largely bound to 13 kDa proteins, some bound to thio-compounds.	Half bound to 13 kDa and >66 kDa proteins, half to thio-compounds.
Zn	Predominantly bound to NA.	Predominantly bound to NA.
Cu	Dominantly bound to NA, some bound to histidine and proteins.	Dominantly bound to NA, some bound to histidine and proteins.

Informations are from Kato et al. (2010), Nishiyama et al. (2012), Ando et al. (2013), Ariga et al. (2014), Hazama et al. (2015), Yoneyama, Ishikawa, and Fujimaki (2015) and Yoneyama and Ariga (2016).

5.2. Metabolism of Fe in the mature source leaves

In rice, Fe is predicted to be the first allocated to the mature leaves via xylem as Fe-citrate and then re-allocated to the growing leaves through the phloem, as previously discussed in Section 4 and depicted in Figure 4. The Fe-citrate, which arrives at the xylem apoplasts of mature leaf veins of graminaceous plants, may be reduced to Fe²⁺ via a light-dependent Fe-chelate oxidoreductase (FRO) system in the mesophyll-cell membranes using an NADH-dependent FRO such as OsFRO2 (Gross et al. 2003; Ishimaru et al. 2006; Mikami et al. 2011), and the Fe²⁺ produced and present in the xylem apoplast may be absorbed by the iron-regulated transporter (IRT1) into the mesophyll cells (Mikami et al. 2011). Such Fe-citrate reduction and Fe²⁺ transport systems in the leaf mesophyll membranes (Brüggemann, Maas-Kantel, and Moog 1993; Mukherjee et al. 2006; Feng et al. 2006) are operated in non-graminaceous plants, which transport predominantly Fe-citrate in the xylem to the leaves (castor bean in Table 2; Schuler et al. 2012). A hypothesis proposed by Grillet et al. (2014) and Grillet and Schmidt (2017) for dicotyledonous plants suggests that apoplastic Fe-citrate may also be reduced to Fe²⁺ by ascorbate excreted from the mesophyll cells, and the Fe²⁺ produced is also taken up by membrane-localized IRT1 into the mesophyll cells. In the future, it would be worthwhile to examine the operation of such xylem Fe-citrate reduction with excreted ascorbate in graminaceous plants. The intracellular Fe²⁺ may bind to NA as Fe-NA. The Fe-NA could then move through the plasmodesmata of mesophyll cells by symplastic diffusion (Madore, Oross, and Lucas 1986; Stephan et al. 1996; Schuler et al. 2012) and, finally, would be loaded into phloem by OsYSL2 (Koike et al. 2004) for re-translocation to the growing leaves after ligand exchange to MAs in the phloem. Some Fe-NA in mature leaf cells are stored as ferritins (high-Fe proteins) in the chloroplast stroma (Briat et al. 2010) and as Fe-NA in the vacuoles through rice vacuolar iron transporters (OsVIT1/2; Zhang et al. 2012). Vacuolar Fe is taken up to the cytoplasm by NRAMPs (natural resistance-associated macrophage proteins) under low-Fe supply (in Arabidopsis; Lanquar et al. 2005).

Fe-containing proteins, such as the PSI and PSII proteins assembled in the thylakoid membranes, are the sources of Fe for re-mobilization from the mature leaves. The investigation of protein turnover in barley leaves by a proteome analysis combined with in planta isotope labeling indicated Kd (d⁻¹) values of 0.04 for PSI and 0.10 for PSII (Nelson et al. 2014). The ferritin proteins accumulated as soluble storage forms in the chloroplasts are ready for re-mobilization under Fe deficiency (Briat 2008; Briat et al. 2010). The turnover of these Fe-containing compounds, including autophagy at the stage of leaf senescence (Pottier et al. 2014), may also be one of the mechanisms which sustain the re-mobilization and re-cycling of Fe as phloem-delivered Fe-MAs and Fe-proteins. Along with the decreases in leaf Fe and N contents, which are accelerated by senescence in barley plants, the expressions of NAS2 and DMAS1 genes and the synthesis of Fe-chelators DMA and citrate rather than NA are enhanced (Shi et al. 2012).

5.3. Synthesis of phloem Fe-compounds

At present, the mechanisms by which Fe-compounds (Fe-MAs and Fe-proteins) in phloem sap are formed in the source leaves and delivered to the sink leaves are largely unknown, although a possible mechanism can be proposed based on the literature of phloem transport of photosynthates, amino acids, and proteins (Oparka and Santa Cruz 2000; Van Bel 2003; Lee J-Y 2018). Namely, Fe-NA in the mesophyll cells may be released to the apoplastic space and taken up by OsYSL2, which is located in the plasma membranes of companion cells. The Fe-NA would then be transformed to Fe-MAs and transferred to the sieve elements. The Fe-MAs present and Fe-proteins re-mobilized in the mature leaves could then be released to sieve elements through the plasmodesmata (Figure 4). The phloem companion cells are also known as the site of protein synthesis (Van Bel 2003). It is possible that certain specific proteins are synthesized in the companion cells by incorporating phloem-present Fe ions, and are then released to the sieve elements. Such metal-binding activity has been observed for Cd in rice phloem sap by 13 kDa proteins (Kato et al. 2010) and for Fe by the castor bean iron transport protein (Krüger et al. 2002).

5.4. Unloading of phloem Fe-compounds in sink leaves and roots

The phloem Fe-compounds (Fe-MAs and Fe-proteins) delivered to the young-leaf veins may be released to the mesophyll cells through plasmodesmata (Hell and Stephan 2003), which are located between phloem companion cells and mesophyll cells (Figure 4). Upon delivery to the mesophyll cells, such Fe-MAs and Fe-proteins may be utilized as Fe nutrients by converting them to Fe-NA for the synthesis of mitochondria and for the light-involved synthesis of green chloroplasts. Fe import to mitochondria in rice is mediated by the rice mitochondrial Fe transporter (OsMIT), which is localized on the inner mitochondrial membrane (Bashir et al. 2011), and Fe import to chloroplasts and its allocation in the chloroplasts may be regulated by PIC1 (permease in chloroplasts) (in Arabidopsis; Duy et al. 2011), although the chemical forms of their substrates in vivo are not known.

It is important to note that the phloem transport of Fe-compounds to the root apex has the role of long-distance signaling of the Fe demand of shoots, thereby regulating the Fe-acquisition activity in the roots under the conditions of Fe-deficiency or Fe-excess (Kobayashi et al. 2010). In such shoot-to-root signaling, phloem-specific iron transporters such as AtOPT3 (Arabidopsis oligopeptide transporter; Stacey et al. 2008; Zhai et al. 2014) and potentially OsYSL2 (Ishimaru et al. 2010) may play important roles, although the sensing mechanism of the signals, which are unloaded at the root apex (Lee J-Y 2018), has not been elucidated and is under active investigation. As a shoot-to-root signal in rice, certain peptides, called IRON MAN (IMA)/Fe uptake-inducing peptides, were recently reported (Kobayashi, Nagano, and Nishizawa 2021).

6. Fe utilization in chloroplasts and occurrence of chlorosis

During vegetative growth, Fe is mainly used for the synthesis of chloroplasts and mitochondria in the growing leaves (Álvarez-Fernández et al. 2014). As depicted in Figure 4, solar energy is captured by chlorophylls through the process of excitation and is soon transferred to the complexes of PSII (containing 2‒3 Fe), Cyt-b₆f (5 Fe) and PSI (12 Fe) arranged on the chloroplast thylakoids (Raven, Evans, and Korb 1999). The chlorophyll-protein assemblies (CPs) in thylakoids contain heme and [Fe-S] cluster proteins, which function as the prosthetic groups. The [Fe-S] moiety is synthesized in the chloroplasts using imported Fe²⁺ on the SUF (sulfur utilization factor) machinery, which includes scaffold SUFB, and shortage of imported Fe causes downregulation of SUFB expression, resulting in retardation of [Fe-S] synthesis (Liang et al. 2014; Hantzis et al. 2018; Kroh and Pilon 2020). Heme and chlorophyll moieties are formed from a common precursor, protoporphyrin IX (Kroh and Pilon 2020). The heme structure is formed when Fe²⁺, probably supplied as Fe-NA, is inserted into protoporphyrin IX by ferrochelatase. Chlorophylls (a/b) are produced via two major steps: (1) Mg²⁺ insertion into protoporphyrin IX by magnechelatase and (2) the light-dependent reduction of dark-accumulated protochlorophyllide to chlorophyllide a. The synthesis and accumulation of green pigments in leaf cells may induce the transformation of the etioplasts to chloroplasts (Rhodes and Yemm 1966; Bollivar 2006).

The chlorophylls are inserted into specific proteins to form light-harvesting chlorophyll (LHC) proteins by binding with galactolipids on the thylakoids (in sugar beet; Nishio, Taylor, and Terry 1985); these LHC proteins function as antennae, capturing solar energy and transferring it to the photo-systems where solar energy is converted to chemical energy. Thus, chlorophylls are bound noncovalently to LHC proteins and [Fe-S]-containing PSI and PSII (Markwell, Thornber, and Boggs 1979; von Wettstein, Gough, and Kannangara 1995); the uncombined free chlorophylls under illumination may be chemically destroyed due to an excess of energy, which may produce toxic radicals (Rodríguez-Celma et al. 2013). First PSI and then PSII and Cyt-b₆f are assembled on the thylakoid membranes in balanced ratios (Nevo et al. 2012). The ratio of the Fe-containing heme moiety to the chlorophyll moiety in leaves is always in balance (Marsh, Evans, and Matrone 1963), and the ratio of iron and chlorophylls also appears to be balanced (Oertli and Jacobson 1960). The re-arrangement of leaf PSII proteins in barley Fe-deficient young leaves may underlie the reorganization/remodeling of the LHCII proteins, which is unique in Fe-deficient barley leaves and functions to photo-protect the plants by a partial re-allocation of green chlorophylls (Saito et al. 2010, 2014). This re-modeling does not occur in rice leaves, and thus rice leaves undergo severe yellowing under prolonged iron deficiency. It is noteworthy that the contents of ferredoxin (Fd), a chloroplast-soluble [Fe-S]-containing protein, are correlated with the signs of iron chlorosis (Alcaraz et al. 1985). If the amount of Fe²⁺-NA delivered to the developing mesophyll cells is small, it may cause a decrease in the ratio of the green chlorophyll and Fe-protein (C-P) complexes relative to the structural growth (Machold 1971; Spiller and Terry 1980), resulting in interveinal chlorosis.

7. Conclusion and future research

Iron chlorosis occurs in growing leaves due to a local shortage of Fe at chlorophyll-PS assembly. To deliver the root-absorbed Fe to the growing leaves, the substrate Fe is chemically transformed so that it can pass the plasmodesmata by diffusion in slightly alkaline cytosols in the root parenchyma and leaf mesophyll cells, and so that it can move around subcellular organelles such as plastids and mitochondria. To draw a clear picture of iron availability to the sink leaves, further studies will be needed to clarify the speciation of intracellularly delivered Fe and to characterize the phloem delivery via loading and unloading, as described in the sections below.

7.1. Mechanisms of heavy metal-induced restriction of iron transport

During the formation of metal-MAs complexes in the rhizosphere, there is a possibility of competition between Fe-MAs and Cu-MAs or between Fe-MAs and Co-MAs (Sugiura et al. 1981), and between inhibition of the uptake of Fe-MAs by Cu-MAs or Co-MAs. However, the findings of the study by Chino (Table 1) showed that ⁵⁹Fe uptake into the plants was not significantly inhibited by Co²⁺, Cu²⁺, or Mn²⁺, although ⁵⁹Fe transport from the roots to the plant tops was greatly inhibited. It is also possible that Fe-MAs in the xylem fluid compete against heavy metals, since Cu²⁺ and Co²⁺ can form Cu-MAs and Co-MAs (Sugiura et al. 1981; Ando et al. 2013) (Figure 3). However, Fe in the xylem can form abundant levels of Fe-citrate complexes (Ariga et al. 2014). Therefore, these are not the cause of heavy metal-induced restriction of Fe transport in the xylem.

7.2. Mechanisms of loading and unloading of Fe into xylem and phloem

Understanding the speciation of metal ions in specific intracellular and intravascular compartments is crucial to understanding how the metals are kept soluble, possibly for their membrane transport and for delivery to their final target (Grillet et al. 2014).

Regarding the radial transport of Fe to the xylem via parenchyma cells in roots (Figure 2), the mechanism of Fe-NA loading into xylem after the symplastic diffusion via plasmodesmata is not clear, including in regard to whether YSL transporters are involved. Although the mechanisms are not known, the symplastic connection between phloem and mesophyll cells in the growing leaves via the plasmodesmata might function as a series of molecular gates, instead of metal transporters, to take up the metal-containing compounds including Fe-containing peptides (Nishiyama et al. 2012) in a manner similar to the loading and unloading of photosynthates (Madore, Oross, and Lucas 1986).

Previously, light-dependent translocation of ⁵⁹Fe-epihydroxymugineic acid from leaf veins to mesophyll cells in intact barley plants and also its influx into the chloroplasts isolated from barley leaves was demonstrated (Bughio et al. 1997). However, information on the in vivo mechanisms of Fe transport in the chloroplasts is still scarce (Balk and Schaedler 2014), as is information on the mechanisms of the trafficking of Fe to the synthesis site of [Fe-S] clusters and heme cofactors of photosystems and redox-involved proteins such as ferredoxins and nitrite reductase in developing chloroplasts and thylakoids.

7.3. Shoot-to-root signaling of iron-deficiency

Sensing of low-Fe-availability signals produced locally and conveyed by long-distance phloem transport induces the expression of transcription factors (IDEF1/IRO2, IDEF2) specifically at lateral-root parenchyma cells and apical root vascular cells in roots, and at mesophyll cells and vascular cells in the mature leaves (Kobayashi et al. 2010). The speciation of such signaling substances functioning in the root parenchyma and leaf mesophyll cells and in the phloem is not definitely determined, although Fe-NA in the cell cytosols and Fe-MAs in the phloem solute might be such candidates in graminaceous plants (Yoneyama and Ariga 2016). The genes encoding IRON MAN (IMA) peptides, which are able to bind Fe²⁺ and other divalent metals, were recently suggested to be important players in shoot-derived, phloem mobile signals in Arabidopsis (Grillet et al. 2018). IMA homologues were recently confirmed to play such a role in rice (Kobayashi, Nagano, and Nishizawa 2021), although the involved peptides in plant tissues and specifically in phloem saps have not been chemically identified.

Acknowledgments

The author thanks Professors Sei-ichi Takagi and Shigenao Kawai for their input regarding the interaction between iron availability by mugineic acids and chlorosis in graminaceous plants and his colleagues for their contribution to the speciation of metals in phloem saps of rice and castor bean, as summarized in Table 3. Finally, the author thanks Dr. Takeshi Shimizu for his courteous provision of the photos in Figure 1.

Disclosure statement

No potential conflict of interest was reported by the author(s).