P

Mt

Ga

b

a

ARR1AA

KMEEFF

I

oelathwfawq

ag

CC

0h

Journal of Plant Physiology 171 (2014) 639–647

Contents lists available at ScienceDirect

Journal of Plant Physiology

journa l h om epage: www.elsev ier .com/ locate / jp lph

hysiology

edicago truncatula ecotypes A17 and R108 differed in their responseo iron deficiency

en Lia,1, Baolan Wanga,1, Qiuying Tiana, Tianzuo Wanga, Wen-Hao Zhanga,b,∗

State Key Laboratory of Vegetation and Environmental Change, Institute of Botany, the Chinese Academy of Sciences, Beijing 100093, PR ChinaResearch Network of Global Change Biology, Beijing Institutes of Life Sciences, Chinese Academy of Sciences, Beijing, China

r t i c l e i n f o

rticle history:eceived 17 August 2013eceived in revised form9 December 2013ccepted 20 December 2013vailable online 21 March 2014

eywords:edicago truncatula

cotype A17cotype R108

s u m m a r y

Medicago truncatula Gaertn is a model legume species with a wide genetic diversity. To evaluate theresponses of the two M. truncatula ecotypes, the effect of Fe deficiency on ecotype A17 and ecotypeR108, which have been widely used in physiological and molecular studies, was investigated. A greaterreduction in shoot Fe concentration of R108 plants than that of A17 plants was observed under Fe-deficient conditions. Exposure to Fe-deficient medium led to a greater increase in ferric chelate reductase(FCR) activity in roots of A17 than those of R108 plants, while expression of genes encoding FCR inroots of A17 and R108 plants was similarly up-regulated by Fe deficiency. Exposure of A17 plants toFe-deficient medium evoked an ethylene evolution from roots, while the same treatment had no effecton ethylene evolution from R108 roots. There was a significant increase in expression of MtIRT encodinga Fe transporter in A17, but not in R108 plants, upon exposure to Fe-deficient medium. Transcripts

e deficiencyerric chelate reductase activity

of MtFRD3 that is responsible for loading of iron chelator citrate into xylem were up-regulated by Fedeficiency in A17, but not in R108 plants. These results suggest that M. truncatula ecotypes A17 and R108differed in their response and adaptation to Fe deficiency, and that ethylene may play an important rolein regulation of greater tolerance of A17 plant to Fe deficiency. These findings provide important cluesfor further elucidation of molecular mechanism by which legume plants respond and adapt to low soil

Fe availability.

ntroduction

Iron (Fe) is an essential nutrient for plant growth and devel-pment. Although Fe is the fourth most abundant element in thearth’s crust, it usually exists in the form of Fe (III)-oxides withow availability for plants, especially in calcareous soils (Guerinotnd Yi, 1994). Therefore, Fe deficiency often limits crop produc-ion and nutritional quality. In addition, more than 2 billion peopleave been reported to be at risk of Fe-deficiency-induced anemiaorldwide (WHO, 2007). Given that plants are a primary Fe source

or humans, understanding mechanisms by which plants efficientlycquire Fe from soil, particularly under Fe-deficient conditions,ould allow for breeding crop cultivars with high Fe nutritional

uality, thus alleviating Fe malnutrition of humans.

Two strategies have been developed by plants to mobilize and

cquire Fe from soil (Römheld and Marschner, 1981). These strate-ies are defined as Strategy I in non-graminaceous monocots and

∗ Corresponding author at: State Key Laboratory of Vegetation and Environmentalhange, Institute of Botany, the Chinese Academy of Sciences, Beijing 100093, PRhina. Tel.: +86 10 6283 6697; fax: +86 10 6259 2430.

E-mail address: whzhang@ibcas.ac.cn (W.-H. Zhang).1 These authors contributed equally to this work.

176-1617/$ – see front matter © 2014 Elsevier GmbH. All rights reserved.ttp://dx.doi.org/10.1016/j.jplph.2013.12.018

© 2014 Elsevier GmbH. All rights reserved.

dicots, and Strategy II in graminaceous monocots (Römheld andMarschner, 1981; Kobayashi and Nishizawa, 2012). In Strategy Iplants, numerous changes in morphological and physiological pro-cesses are initiated under Fe-deficient conditions. These includesub-apical swelling with proliferation of root hairs, developmentof transfer cells, increases in ferric chelate reductase (FCR) activity,acidification of the rhizosphere, and up-regulation of Fe (II) trans-porters (Curie and Briat, 2003; Hell and Stephan, 2003; Romeraand Alcántara, 2004; Morrissey and Guerinot, 2009; Conte andWalker, 2011). In contrast, the Strategy II plants are character-ized by enhanced release of phytosiderophores and highly specificuptake system for Fe (III)-phytosiderophores in response to Fe defi-ciency (Römheld and Marschner, 1994; Inoue et al., 2003; Nozoyeet al., 2011).

Fe (III) has to be reduced to Fe (II) by a plasma membrane-bound FCR prior to uptake by roots in Strategy I plants (Romeraand Alcántara, 2004; Kobayashi and Nishizawa, 2012). FRO2, whichwas first isolated from Arabidopsis, encodes a plasma membraneFe (III) chelate reductase and is involved in mediation of Fe (II)acquisition by Strategy I plants (Robinson et al., 1999a,b). Once

reduced, uptake of Fe (II) into roots is mediated by a divalent trans-porter of Iron-Regulated Transport 1 (IRT1) (Vert et al., 2002). TheArabidopsis knockout mutant of IRT1 showed chlorosis and wasunable to take up Fe and other divalent cations in low Fe medium,

6 Physi

ia2aite

atifaRoaab2

r2ebrsCt2aI

oaagew2tTttsbtktocmm

M

P

Rit0Tbt

40 G. Li et al. / Journal of Plant

ndicating that the IRT1 is a major transporter responsible for high-ffinity metal uptake under Fe-deficient conditions (Vert et al.,002). AtIRT1 belongs to ZIP (ZRT/IRT-related protein) family, andn orthologue of ZIP has been also identified in other plant species,ncluding Lycopersicum esculentum (Eckhardt et al., 2000), Medicagoruncatula (López-Millán et al., 2004) and Arachis hypogaea (Dingt al., 2010).

After taken up by root cells, Fe has to be loaded into the xylemnd transported to the shoot via transpiration stream. FRD3, a pro-ein belonging to the multidrug and toxin efflux (MATE) family, isnvolved in loading of citrate, which is an iron chelator necessaryor distribution of Fe throughout the plant, into the xylem (Greennd Rogers, 2004; Durrett et al., 2007; Rellan-Alvarez et al., 2009;oschzttardtz et al., 2011). FRD3 plays a key role in the regulationf Fe translocation from roots to shoots (Durrett et al., 2007). Inddition, FRD3 is expressed at detectable levels in roots, as wells in seeds and flowers in Arabidopsis, but its function appears toe root-specific (Rogers and Guerinot, 2002; Roschzttardtz et al.,011).

Several hormones are involved in the regulation of Fe-deficiencyesponses, including ethylene (Lucena et al., 2006; Waters et al.,007), cytokinins (Séguéla et al., 2008), and brassinosteroids (Wangt al., 2012). Among the phytohormones, ethylene is one of theest characterized phytohormones associated with Fe-deficiencyesponse in strategy I plants. For instance, exposure of cucumbereedlings to Fe-deficient medium up-regulates the expression ofsACS2 and CsACO2 that encode ACC synthase and ACC oxidase,wo enzymes responsible for ethylene biosynthesis (Garcia et al.,011). The Fe deficiency-induced ethylene has been shown to act asn enhancer to regulate FRC activity and IRT in response to Strategy

plants to Fe deficiency (Romera and Alcántara, 1994, 2004).M. truncatula has been used as a model plant to study genomics

f legume plants because of its small diploid genome size and rel-tively easy transformation (Cook, 1999; Wang et al., 2011). Therere different ecotypes (accession lines) of M. truncatula with wideenetic variations (Ellwood et al., 2006). Among the M. truncatulacotypes, M. truncatula Jemalong A17 (A17) has been used for thehole-genome sequencing project (Choi et al., 2004; Young et al.,

005), while the ecotype R108 is often used for gene transforma-ion, due to its superior in vitro regeneration (Hoffmann et al., 1997).here are reports demonstrating that the A17 differs from R108 inheir phenotypes (Schnurr et al., 2007; Bolingue et al., 2010) and inheir responses to abiotic stress (de Lorenzo et al., 2007) and biotictress (Salzer et al., 2004; Gaige et al., 2012). However, there haveeen few studies to investigate whether the two ecotypes differ inheir response to deficiency in mineral nutrients. Identification ofey physiological processes underlying tolerance of legume plantso nutrient deficiency is of importance for molecular improvementf plants grown in nutrient-poor soils. In the present study, weompared the effect of Fe deficiency on the two ecotypes of legumeodel species of M. truncatula at physiological, morphological andolecular levels.

aterials and methods

lant growth

Two Medicago truncatula Gaertn ecotypes (Jemalong) A17 and108 were used in this study. Seeds of both ecotypes were dipped

n sulfuric acid for 10 min to degrade the seed coat, and rinsedhoroughly with sterilized water. The seeds were then sown on

.8% agar to germinate at 25 ◦C until the radicals were about 2 cm.he seedlings were planted in plastic buckets (12 seedlings perucket) filled with 2.5 L aerated nutrient solution. The composi-ion of full-strength nutrient solution is: 0.25 mM KH2PO4, 0.5 mM

ology 171 (2014) 639–647

MgSO4·7H2O, 0.125 mM CaCl2, 1.25 mM KNO3, 0.5 mM NH4NO3,15 �M H3BO3, 2.5 �M MnSO4·H2O, 0.5 �M ZnSO4·7H2O, 0.5 �MCuSO4·5H2O, 0.35 �M NaMoO4·2H2O, and 50 �M Fe (III) EDTA witha pH of 6.0. Seedlings were firstly cultured in half-strength nutri-ent solution for 5 days and then transferred to full strength nutrientsolution for another 10 days. Thereafter, the 15-day-old seedlingswere exposed to the Fe-deficient nutrient solution (−Fe) in whichFe (III) EDTA was removed, while the nutrient solution containing50 �M Fe (III) EDTA is referred to as Fe-sufficient medium (+Fe).Nutrient solution was replaced every 5 days. Plants were grown inan environment-controlled chamber with a photosynthetic photonflux density of 350 �mol m−2 s−2 photo-synthetically active radia-tion, 80% relative humidity, and at a 16 h 26 ◦C/8 h 19 ◦C day/nightregime.

Determination of chlorophyll and root morphology

Plants exposed to Fe-deficient medium for 10 days were usedto measure chlorophyll (Chl) concentration and root morphol-ogy. Secondary root density is defined as the ratio of secondaryroot number to primary root length. Newly formed leaves wereexcised and weighed, then extracted in aqueous acetone (80%, v/v)as described previously (Wang et al., 2012). The extract solutionswere measured in 663 nm and 645 nm with a spectrophotometer(Rio-Rad, USA). Total Chl concentration (mg L−1) was calculated as8.02A663 + 20.21A645, and Chl concentration was expressed as mgChl g−1 fresh weight. Excised roots were washed with deionizedwater and the length of primary root and numbers of secondaryroots were determined.

Determination of Fe, Mn and Zn concentrations

The concentrations of Fe, Zn and Mn in roots and shoots of plantsgrown in Fe-sufficient and Fe-deficient medium were measuredusing an inductively couple plasma-atomic emission spectrometry(ICP-OES) with a detection limit of 0.01 ppm for Fe as describedpreviously (Wang et al., 2012). Samples of shoots and roots ofboth ecotypes cultured in Fe-sufficient (+Fe) and Fe-deficient (−Fe)medium for 10 days were harvested and dried at 75 ◦C and digestedwith the mixture of nitric acid and hydrogen peroxide using amicrowave system (CEM, USA). The digested samples were ana-lyzed by iCAP6300 (Thermo Electron, USA).

Measurements of Fe (III) chelate reductase activity

Activities of FCR in roots grown in Fe-sufficient and Fe-deficientmedium were determined spectrophotometrically following theprotocols described by Yi and Guerinot (1996). Briefly, intactroot systems were submerged in the reductase assay solutioncontaining 0.1 mM Fe (III) EDTA and 0.3 mM ferrozine (pH 5.0).Samples were incubated in a culture chamber for 1 h, and thenthe absorbance of the assay solution was read with a spec-trophotometer (Rio-Rad, USA) at 562 nm. The concentration of Fe(III)-ferrozine was calculated with the molar extinction coefficientof 36.485 mM−1 cm−1 and the results were expressed in nmol Fe(II) per gram fresh weight per hour.

Detection of ethylene production

Ethylene production from roots was detected with a gas chro-matograph as described previously (Sun et al., 2007; Li et al., 2009).Briefly, excised roots of the two ecotypes grown in Fe-sufficient

and Fe-deficient medium for 10 days were rinsed with distilledwater and placed into a sealed glass jar (10 cm3) for 1 h at roomtemperature. To minimize a wounding effect, the excised rootswere incubated in aerated distilled water for 1 h in the glass jar

Physi

ptdfiu0w

Xc

mtFms1RwTurmd2A2d

R

edapl3(GT3GC35GTastswPsat5twa

G. Li et al. / Journal of Plant

rior to sealing up. Ethylene was sampled with 1 cm3 syringes fromhe headspace of the sealed container and the concentration wasetermined with a gas chromatograph (GC-7A, Shimadzu, Japan)tted with a flame ionization detector and an activated glass col-mn filled with GDX502 (Shimadzu, Japan). The detection limit was.013 cm3 m−3. Ethylene production was calculated based on fresheight (FW) of root samples.

ylem sap collection and determination of Fe and citrateoncentration

Xylem sap was collected with a pressure chamber (PMS, Instru-ents, Corvallis, OR) as described previously (Li et al., 2009). Briefly,

he excised roots from the two ecotypes grown in Fe-sufficient ande-deficient medium for 10 days were put into the chamber. Toinimize the diurnal fluctuations in the concentration of xylem

ap contents, xylem sap was collected in the morning between0:00 and 12:00 am, and the collection was made from A17 and108 plants alternatively. A pressure of approximately 1.1 MPaas applied for 15 min to allow for collection of the xylem sap.

he collected xylem sap was used for measuring Fe concentrationssing an ICP after the first few drops of xylem sap from excisedoots were discarded. Citrate concentration in the xylem sap waseasured by a reverse-phase HPLC system following the protocols

escribed by Wang et al. (2006). Separation was conducted on a50 mm × 4.6 mm reverse-phase column (Alltima C18, 5 Micron;lltech Associates, Inc., Deerfield, IL, USA). The mobile phase was5 mM KH2PO4 (pH 2.5) with a flow rate of 1 mL min−1 at 28 ◦C, andetection of organic anions was carried out at 214 nm.

NA isolation and quantitative real-time PCR

RNA isolation was carried out as described preciously by Sunt al. (2010). Briefly, roots and young leaves of plants exposed to Fe-eficient medium for 10 days were sampled. MtActin gene was useds a reference to quantify the relative transcript level. The RT-qPCRrimers designed using Primer Premier5.0 software were as fol-

ows: MtFRO2 (MTR 7g038180) (5′-ATGGAACCGTCCAATGCTTGT-′ and 5′-TTGCCGTACTCTGCCGCTGT-3′), MtFRD3 (MTR 3g029510)5′-AGTGACATTCTGCGTGACCTT-3′ and 5′-ACCATCAGCGAGAAGGA-3′), MtIRT (MTR 4g083570) (5′-AGTGCTCGTCCAAATATGAAGGG-3′ and 5′-TGCTGGGATCGAAGTTGTGAAA-3′), MtACO1 (MTRg083370) (5′-CCAAAGGGCTAGAGGCTGTTC-3′ and 5′-GGTAGGTACGCAAATGGA AA-3′), MtACO2 (MTR 2g025120) (5′-GTTAGTAATACCCTCCTTGTCCT AAGC-3′ and 5′-AAGGATGATGCCACCAGCAT-′), MtACS2 (AY062022) (5′-TGCCTACACCTTACTATCCAG-3′ and′-TCTGTCCATAACTGCCTAA-3′), MtACS3 (MTR 8g101820) (5′-TCTACCAGGTTTCAGAGTTG-3′ and 5′-CTCTTCTTCAATCTTTCCCAT-3′), MtActin (MTR 7g026230) (5′-ACGAGCGTTTCA GATG-3′

nd 5′-ACCTCCGATCCAGACA-3′). All primers for the genes wereelected on the basis of available A17 sequence data. To verify thathe primers are identical to the genes in R108 plants, we cloned andequenced these genes in R108 plants. The sequences in R108 plantsere perfectly matched for A17 plants (Data not shown). Real-time

CR was performed using SYBR GreenERTM qPCR SuperMix Univer-al (Invitrogen) in 96-well reaction plates (Applied Biosystems) on

Step One Plus Real-Time PCR System (Applied Biosystems). Thehermal cycle used was 95 ◦C for 2 min, 40 cycles of 95 ◦C for 30 s,

5 ◦C for 30 s, and 72 ◦C for 30 s. Each sample was carried out byhree independent experiments and the relative expression levelas analyzed by the comparative CT method as described by Livak

nd Schmittgen (2001).

ology 171 (2014) 639–647 641

Statistical analysis

Analysis of variance was conducted between the different treat-ments. Significant differences between the two ecotypes underFe-sufficient and Fe-deficient conditions were evaluated by LSDmultiple range tests (P ≤ 0.05) using the SAS statistical software.

Results

General effect of Fe deficiency on A17 and R108

Distinct symptoms associated with Fe deficiency such as chloro-sis and suppressed shoot growth were observed in the R108seedlings after growth in Fe-deficient medium (0 �M Fe) for 10days (Fig. 1A). By contrast, no such symptoms appeared in A17seedlings when they were exposed to the identical Fe-deficientmedium (Fig. 1A). A reduction in chlorophyll content is one of theresponses of plants to Fe deficiency (Graziano et al., 2002). No sig-nificant difference in chlorophyll concentration was observed inA17 plants after grown in Fe-deficient medium for 10 days (Fig. 1B).In contrast, there was a marked reduction in chlorophyll concentra-tion in R108 plants when exposed to the same Fe-deficient mediumfor 10 days (Fig. 1B). Chlorophyll concentration in A17 and R108plants was comparable under Fe-sufficient conditions (P = 0.512),but it became lower in R108 than that in A17 plants under Fe-deficient conditions (P = 0.002). These phenotypes suggest that A17plants are more tolerant to low Fe availability in growth mediumthan R108 plants.

Exposure of the two ecotypes to Fe deficiency led to significantdecreases in shoot biomass of both ecotypes (Table 1). There wasa significant decrease in root biomass of R108 plants, but not ofA17 plants when transferring them to Fe-deficient medium for 10days (Table 1). In addition to root biomass, we also investigatedthe effect of Fe deficiency on root morphology of the two eco-types. Exposure of R108 seedlings to Fe-deficient medium led toa significant decrease in their primary root length and number ofsecondary root (Table 1). However, the secondary root density ofR108 plants remained unchanged due to a reduction in primaryroot length when challenged by Fe deficiency (Table 1). In con-trast to R108 plants, no changes in primary root length, secondaryroot number, and secondary root density were found in A17 plantswhen challenged by the identical Fe-deficient regimes. Note thatthe secondary root number of A17 plants was significantly higher(P = 0.001) than that of R108 plants under Fe-sufficient conditions(Table 1), and that exposure to Fe-deficient medium led to the sec-ondary root number in R108 plants being much less than that inA17 plants (Table 1).

Effect of Fe deficiency on Fe concentration

Fe concentration in shoots of A17 was significantly higher(P = 0.014) than that in R108 plans under Fe sufficient conditions(Fig. 2A). There was a marked decrease in Fe concentrations ofA17 and R108 plants upon exposure of these plants to Fe-deficientmedium, and the decrease in shoot Fe concentration was greaterin R108 than in A17 plants (Fig. 2A). For example, Fe concentra-tion in A17 shoots was reduced by 43% in response to treatmentwith Fe deficiency, while shoot Fe concentration in R108 plants wasreduced by 68% by the same Fe-deficient treatment, leading to theshoot Fe content in A17 plants being 2.1-fold higher than in R108plants under Fe-deficient conditions. A similar decease in Fe con-

centrations in roots of both A17 and R108 plants was also observedwhen these plants were grown in Fe-deficient medium (Fig. 2B).However, unlike Fe concentration in shoots, a decrease in rootFe concentration was comparable between A17 and R108 plants,

642 G. Li et al. / Journal of Plant Physi

Fig. 1. Chlorophyll concentrations in newly formed leaves of M. truncatula A17 andR108 seedlings grown in Fe-sufficient (+Fe) and Fe-deficient medium (−Fe). Fifteen-day-old seedlings of A17 and R108 plants pre-cultured in Fe-sufficient mediumwere transferred to either Fe-deficient or Fe-sufficient medium and young leaveswere harvested for determination of chlorophyll concentrations after growth ineither Fe-sufficient and Fe-deficient medium for 10 days. Data are means ± SE withfour replicates. *Significant differences between Fe-sufficient (+Fe) and Fe-deficient(−Fe) for each ecotype at the level of P ≤ 0.05.

Table 1Growth and root morphological characteristics of M. truncatula A17 and R108 plants growthat were exposed to Fe-deficient medium (−Fe) for 10 days were used to determine biofour replicates, and significant differences between +Fe and −Fe treatments within each e

Ecotypes A17

+Fe −Fe

Shoot DW (mg) 21.58 ± 0.97 14.94Root DW (mg) 5.41 ± 0.33 4.95Primary root length (cm) 34.63 ± 0.43 35.98Secondary root number 67.67 ± 1.91 69.50Secondary root density 1.9 6 ± 0.06 1.93

ology 171 (2014) 639–647

leading to a similar Fe concentration in roots of A17 and R108plants (Fig. 2B). In Fe-sufficient seedlings, Fe concentration in thexylem of A17 plants was not significantly different (P = 0.142)from that of R108 plants (Fig. 2C). Fe concentration in the xylemof both A17 and R108 plants was reduced by exposure to Fe-deficient medium, and Fe concentration in the xylem of R108 plantswas not significantly different (P = 0.071) from that of A17 plants(Fig. 2D). Similarly, citrate concentration in the xylem was signif-icantly higher (P = 0.045) in A17 plants than in R108 plants whengrown in Fe-deficient medium, while citrate concentration in thexylem of A17 and R108 plants was comparable when grown underFe-sufficient conditions (Fig. 2E).

Effect of Fe deficiency on FCR activity and expression of MtFRO2

Ferric chelate reductase (FCR) activity in roots of A17 and R108plants was not significantly different (P = 0.172) when grown inFe-sufficient medium (Fig. 3A). A marked increase in FCR activ-ity was observed in roots of A17 and R108 plants upon exposureof A17 and R108 to Fe-deficient medium for 10 days, leadingto no significant difference in FCR activity between A17 andR108 plants under Fe-deficient conditions (P = 0.663) (Fig. 3A).For example, FCR activity in A17 plants was increased from47.82 ± 10.85 to 140.50 ± 28.71 nmol Fe(II) g−1 WT−1 h−1 (n = 4)after 10 days of transferring from Fe-sufficient medium to Fe-deficient medium The same treatment led to an increase FCRactivity of R108 plants from 78.96 ± 16.93 to 128.40 ± 8.76 nmolFe(II) g−1 WT−1 h−1 (n = 4). Similar to FCR activity, exposure of A17and R108 plants led to significant increases in expression level ofMtFRO2 that encodes ferric chelate reductase (Fig. 3B).

Effect of Fe deficiency on concentration of Zn and Mn

In addition to Fe, Fe transporters can also mediate transportof other divalent metals such as Zn and Mn in plants (Vert et al.,2002). We thus investigated the effect of Fe deficiency on Zn andMn concentrations in shoots and roots of the two ecotypes. Therewere increases in Zn concentration in shoots and roots of both A17and R108 plants when grown in Fe-deficient medium compared tothose grown in Fe-sufficient medium (Table 2). The increases in Znconcentration in both roots and shoots of R108 plants were greaterthan those of A17 plants (Table 2). Under Fe-sufficient conditions,R108 plants had higher (P = 0.0001) Zn concentration in their rootsthan A17 plants (Table 2). Similar increases in Mn concentrations inroots and shoots of A17 and R108 plants were also observed whenthese plants were grown in Fe-deficient medium (Table 2). UnlikeZn concentrations, A17 and R108 plants had similar (P = 0.432)

Mn concentrations in their roots under Fe-sufficient conditions,while Mn concentrations in R108 shoots were significantly lower(P = 0.006; P = 0.012) than in A17 leaves under Fe-sufficient and Fe-deficient conditions (Table 2).

n in Fe-sufficient (+Fe) and Fe-deficiency (−Fe) medium. Fifteen-day-old seedlingsmass (dry weight, DW) and morphological parameters. Data are means ± SE withcotype were shown as * (P ≤ 0.05), respectively.

R108

+Fe −Fe

± 0.96* 20.27 ± 0.59 15.18 ± 0.48* ± 0.26 6.1 8 ± 0.48 4.40 ± 0.24*

± 0.68 39.48 ± 0.53 31.98 ± 1.30* ± 1.73 45.20 ± 5.00 37.83 ± 2.52*

± 0.04 1.14 ± 0.12 1.18 ± 0.07

G. Li et al. / Journal of Plant Physiology 171 (2014) 639–647 643

Fig. 2. Effect of Fe deficiency on Fe concentration in shoot (A), root (B and C), Fe concentration in the xylem (C) and citrate concentration in the xylem (D) of M. truncatulaecotypes A17 and R108. Fifteen-day-old seedlings pre-cultured in Fe-sufficient medium were transferred to Fe-deficient medium for 10 days, and concentrations of Fe andcitrate in shoots, roots and xylem sap were determined. Data are means ± SE with four replicates. *Significant differences between Fe-sufficient (+Fe) and Fe-deficient (−Fe)f

TCrb

or each ecotype at the level of P ≤ 0.05.

able 2oncentrations of Zn and Mn in leaves and roots of A17 and R108 plants grown in Fe-suffioots of A17 and R108 plants were measured after they were grown in Fe-deficient mediuetween +Fe and −Fe treatments within each ecotype were shown as * (P ≤ 0.05), respect

Concentration (mg g−1 DW) A17

+Fe −Fe

Leaf Zn 0.02 ± 0.00 0.1Root Zn 0.05 ± 0.01 0.6Leaf Mn 0.22 ± 0.01 0.4Root Mn 0.52 ± 0.09 2.6

cient (+Fe) and Fe-deficient (−Fe) medium. Zn and Mn concentrations in leaves andm for 10 days. Data are means ± SE with four replicates, and significant differencesively.

R108

+Fe −Fe

0 ± 0.00* 0.03 ± 0.01 0.19 ± 0.02*4 ± 0.02* 0.14 ± 0.01 0.76 ± 0.01*2 ± 0.02* 0.16 ± 0.00 0.37 ± 0.05*2 ± 0.07* 0.61 ± 0.07 4.26 ± 0.24*

644 G. Li et al. / Journal of Plant Physiology 171 (2014) 639–647

Fig. 3. Ferric chelate reductase (FCR) activity for root of A17 and R108 plants grownin Fe-sufficient and Fe-deficient medium. Fifteen-day-old seedlings of A17 and R108plants pre-cultured in Fe-sufficient medium were exposed to Fe-deficient mediumff(

E

tusFiswicF

EA

rwtpoRFtm

Fig. 4. Effect of Fe deficiency on expression patterns of genes encoding IRT (MtIRT)(A) and FRD3 (MtFRD3) (B). Roots of A17 and R108 seedlings pre-culture in Fe-sufficient medium for 15 days were sampled at 10 days after exposure of A17 andR108 seedlings to Fe-deficient medium were used to extract total RNA for detectionof relative gene expression by real-time PCR. Data are means ± SE with three biolog-ical replicates. *Significant differences between Fe-sufficient (+Fe) and Fe-deficient(−Fe) for each ecotype at the level of P ≤ 0.05.

Fig. 5. Effect of Fe deficiency on ethylene production from roots of A17 and R108plants. Fifteen-day-old seedlings of A17 and R108 pre-cultured in Fe-sufficient

or 10 days, and were used to determine FCR activity. Data are means ± SE withour replicates. *Significant differences between Fe-sufficient (+Fe) and Fe-deficient−Fe) for each ecotype at the level of P ≤ 0.05.

ffects of Fe deficiency on expression of MtIRT and MtFRD3

To further verify the different responses of A17 and R108 plantso Fe deficiency, two genes (MtIRT and MtFRD3) involved in Feptake and transport were studied at the transcript level. Expres-ion of MtIRT that encodes IRT was increased in A17 plants 2-fold bye deficiency (Fig. 4A). In contrast to A17 plants, expression of MtIRTn R108 plants was insensitive to Fe deficiency (Fig. 4A). Under Fe-ufficient conditions, abundance of MtFRD3 transcript in A17 plantsas comparable with that in R108 plants, and expression of MtFRD3

n A17 plants was enhanced by Fe deficiency, while MtFRD3 trans-ripts in R108 plants remained relatively constant in response toe deficiency (Fig. 4B).

ffect of Fe deficiency on ethylene production and expression ofCS and ACO

Ethylene has been shown to be a key regulator to mediateesponse of plants to Fe deficiency (Lucena et al., 2006). To evaluatehether Fe deficiency-induced ethylene production is involved in

he difference in Fe mobilization and transport in A17 and R108lants, the effect of Fe deficiency on ethylene production in rootsf the two plants was studied. Ethylene level in roots of A17 and

108 roots was comparable (P = 0.091) when they were grown ine-sufficient medium, and a marked increase in ethylene produc-ion from A17 roots was observed when exposed to Fe-deficient

edium (Fig. 5). In contrast, exposure of R018 plants to the same

medium were exposed to Fe-deficient medium for 10 days, and ethylene evolu-tion rates were determined. Data are mean ± SE with three replicates. *Significantdifferences between Fe-sufficient (+Fe) and Fe-deficient (−Fe) for each ecotype atthe level of P ≤ 0.05.

G. Li et al. / Journal of Plant Physiology 171 (2014) 639–647 645

Fig. 6. Effect of Fe deficiency on expression of MtACS2 (A), MtACS3 (B), MtACO1 (C) and MtACO2 (D) of A17 and R108 plants. Fifteen-day-old A17 and R108 seedlings weree oring

m

FR

111doteiliutaetibsp

D

R2

xposed to Fe-deficient medium for 10 days, and total RNA was extracted for monitean ± SE with three biological replicates.

e-deficient medium did not evoke ethylene production in roots of108 plants (Fig. 5).

Ethylene production in plants is catalyzed by two enzymes,-aminocyclopropane-1-carboxylic acid synthase (ACS) and-aminocyclopropane-1-carboxylicacid oxidase (ACO) (Kende,993). To further study the differential response of ethylene pro-uction in the two ecotypes to Fe deficiency, effect of Fe deficiencyn ethylene biosynthesis in the two ecotypes was studied at theranscriptional level. The expression levels of MtACS2, MtACS3 thatncode ACS in A17 and R108 plants were comparable when grownn Fe-sufficient medium (Fig. 6A). Exposure to Fe-deficient mediumed to a significant up-regulation of MtACS2 in A17 plants, but notn R108 plants (Fig. 6A). Expression of MtACS3 was significantlyp-regulated by Fe deficiency in both A17 and R108 plants withhe increase being greater in A17 than in R108 plants (Fig. 6B). Inddition to ACS, expression patterns of MtACO1 and MtACO2 thatncode ACO in A17 and R108 plants also differed in their responseo Fe deficiency. For example, up-regulation of MtACO1 expressionn R108, but not in A17 plants, was observed when challengedy Fe deficiency (Fig. 6C), while exposure to Fe-deficient mediumignificantly enhanced expression of MtACO2 in both A17 and R108lants (Fig. 6D).

iscussion

The two widely used ecotypes of legume model plant A17 and108 differ in their tolerance to salt stress (de Lorenzo et al.,007), morphology (Schnurr et al., 2007), dormancy behavior of

transcript levels by real-time PCR. MtActin was used as an internal control. Data are

seeds (Bolingue et al., 2010) and susceptibility to fungal pathogenMacrophomina phaseolina (Gaige et al., 2012). However, there hasbeen no study to compare the responses of the two ecotypes todeficiencies in mineral nutrients. In the present study, we demon-strate that the M. truncatula ecotype A17 and ecotype R108 differedin their response to Fe deficiency, such that there were less changein leaf chlorophyll concentration, root growth and morphology, andshoot Fe concentrations in A17 than in R108 plants in response toexposure to Fe-deficient medium. The physiological mechanismsunderlying the differential responses of the two ecotypes to Fedeficiency were explored.

An increase in activity of root FCR has been reported in StrategyI plants when exposed to Fe-deficient medium (Schmidt, 2003).The enhanced activity of FCR is an adaptive response to Fe defi-ciency by facilitating reduction of rhizosphere Fe (III), thus allowingplants for more efficient acquisition of Fe. We found an increase inroot FCR activity in A17 and R108 roots, but the magnitude of Fedeficiency-induced increase in FCR activity was higher in A17 thanin R108 plants (Fig. 3). A similar Fe deficiency-induced increasein FCR activity of M. truncatula A17 plants has been reported(Andaluz et al., 2009). At the transcriptional level, we observed agreater up-regulation of MtFRO2 in A108 than in A17 plants underFe-deficient conditions (Fig. 3), suggesting that the difference inFCR activity between the two ecotypes is likely to be accounted

for post-transcriptionally and/or other unknown mechanisms. Thegreater increase in FCR activity of A17 roots than R108 plantsunder Fe-deficient conditions may allow A17 plants to be moreefficient reduction of Fe (III) to Fe (II), thus leading to higher Fe

6 Physi

cmgicfc(

aFgts(FtcRAamtaMRfZitmimg

RicctteRoamelAFuctAmaapeKpilp

r

46 G. Li et al. / Journal of Plant

oncentrations in leaves of A17 than R108 plants. A similar argu-ent has been used to explain the differential tolerance of pea

enotypes to Fe deficiency (Kabir et al., 2012). The Fe concentrationsn roots of both ecotypes (approx. 4500 ppm) under Fe-sufficientonditions appeared higher than those reported in the literatureor other dicot plants, but these values were comparable to Fe con-entration in roots of Arabidopsis grown in Fe-sufficient conditionsBarberon et al., 2011).

Fe is taken up by roots after reduction from Fe (III) to Fe (II) by membrane-bound FCR in the rhizosphere. In Strategy I plants,e transport from the rhizosphere to root cells has been sug-ested to be mediated by an iron-regulated transporter belongingo ZIP family, referred to as IRT1 (Vert et al., 2002). A recent studyhowed that MTR 4g083570 is the most likely ortholog of AtIRT1Rodriguez-Celma et al., 2013). In the present study, we found thate deficiency up-regulated expression of MtIRT in A17 plants, whilehe expression of MtIRT in R108 plants was insensitive to Fe defi-iency (Fig. 4A). The difference in MtIRT activity between A17 and108 plants may explain the higher Fe concentrations in shoot of17 than R018 under Fe-deficient conditions (Fig. 2B). This resultlso suggests that the IRT-mediated transport may be a major deter-inant for the observed difference in tolerance of the two ecotypes

o Fe deficiency. In addition to mediation of Fe transport, IRT canlso mediate transport of other divalent cations such as Zn andn into roots (Vert et al., 2002). Less sensitivity of FCR activity in

108 than in A17 plants may result in the lower amount of Fe (II)or acquisition by plants, thus leading to greater accumulation ofn and Mn by R108 plants under Fe-deficient conditions. Anothernteresting observation is that there was a higher Mn concentra-ion in R108 roots than in A17 plants when grown in Fe-deficient

edium (Table 2). Expression of MtIRT1 in R108 was lower thann A17 plants in Fe-deficient medium (Fig. 4A), implying that other

echanisms may exist in R108 plants which underlie the observedreater accumulation of Mn in R108 plants.

Fe concentration in leaves of A17 plants was comparable with108 plants under Fe-sufficient conditions, while Fe concentration

n A17 leaves was higher than in R108 leaves under Fe-deficientonditions (Fig. 2). Fe concentration in roots of the two plants wasomparable regardless of Fe supply (Fig. 2). These results implyhat A17 plant is more efficient than R108 in terms of Fe acquisi-ion under Fe-deficient conditions. In addition, the less inhibitoryffect of Fe deficiency on Fe concentrations in shoots of A17 than108 plants may also be explained by more efficient translocationf Fe from roots to shoots in A17 plants than R108 plants. FRD3 is

key player in controlling Fe translocation from roots to shoots byediating loading of Fe (III) chelator citrate into the xylem (Durrett

t al., 2007). In the present study, we found that the expressionevel of MtFRD3 was significantly up-regulated by Fe-deficiency in17, while expression of MtFRD3 in R108 plants was insensitive toe deficiency (Fig. 4B). These results suggest that A17 plants canp-regulate loading of citrate into the xylem in response to Fe defi-iency, thus conferring efficient Fe translocation of Fe from rootso shoots. Moreover, the more abundance of FRD3 transcripts in17 than in R108 plants under Fe-deficient conditions (Fig. 4B)ay also be responsible for the higher concentrations of citrate

nd Fe in the xylem of A17 plants than that of R108 plants (Fig. 2Cnd D). Therefore, our results highlight an important role of FRD3layed in regulation of Fe translocation, thus determining the tol-rance capacity of M. truncatula to Fe deficiency. In a recent study,abir et al. (2012) reported that citrate concentrations in roots ofea genotype tolerant to Fe deficiency are substantially increased

n response to Fe-deficient treatment, while the same treatment

eads to a marginal decrease in citrate concentrations in roots ofea genotype less tolerant to Fe deficiency.

Ethylene has been shown to be a key regulator to mediateesponse of plants to Fe deficiency. For instance, it has been

ology 171 (2014) 639–647

suggested that Fe deficiency-induced ethylene production acts asa signal to regulate expression of transcription factor FER and FCRactivity (Romera and Alcántara, 2004; Lucena et al., 2006). A markedincrease in ethylene evolution from A17 roots, but not R108 roots,was found upon exposure to Fe-deficient medium (Fig. 5). A similardifference in Fe deficiency-induced in ethylene evolution betweenpea genotypes differing in tolerance to Fe deficiency has beenobserved (Kabir et al., 2012). We further demonstrated that theexpression of MtACS2, MtACS3, and MtACO2, which encode ACS andACO, in A17 differed from R108 plants in response to Fe deficiency(Fig. 6). ACS activity is a limiting enzyme in ethylene biosynthesis(Kende, 1993). Therefore, the greater up-regulation of MtACS2 andMtACS3 induced by Fe deficiency in A17 plants than in R108 plantsmay account for the Fe deficiency-induce ethylene production inroots of A17 plants (Fig. 5). Given that A17 plants had higher Fe con-centration in leaves and the xylem and ethylene production thanR108 plants under Fe-deficient conditions, it is conceivable thatethylene may be an important regulator to modulate Fe translo-cation by targeting FRD-dependent physiological processes, thusconferring A17 plants more efficient utilization of Fe under Fe-deficient conditions. Further studies by screening large numbers ofM. truncatula ecotypes can test the relationship between ethyleneproduction and Fe utilization efficiency.

In summary, we demonstrate that M. truncatula ecotype A17was more tolerant to Fe deficiency than ecotype R108. The highuse efficiency of Fe in M. truncatula A17 plants than in R108 plantsmay result from the more effective acquisition of Fe due to higherFCR activity and greater up-regulation of FRD3. Our findings alsodemonstrate that A17 plants can evolve ethylene in response toFe deficiency, and that the evolved ethylene may act as a signalto enhance mobilization and translocation of Fe under Fe-deficientconditions, thus conferring M. truncatula ecotype A17 more tolerantto Fe deficiency than ecotype R108.

Acknowledgements

This study was supported by the National Natural Science Foun-dation of China (31101594 and 31272234) and State Key Laboratoryof Vegetation and Environmental Change. We thank the two anony-mous reviewers for their constructive suggestions on the previousversion of the manuscript.

References

Andaluz S, Rodríguez-Celma J, Abadía A, Abadía J, López-Millán AF. Time courseinduction of several key enzymes in Medicago truncatula roots in response to Fedeficiency. Plant Physiol Biochem 2009;47:1082–8.

Barberon M, Zelazny E, Robert S, Conejero G, Curie C, Friml J, et al. Monoubiquitin-dependent endocytosis of the Iron-Regulated Transporter 1 (IRT1) transportercontrols iron uptake in plants. Proc Natl Acad Sci USA 2011;108:E450–8.

Bolingue W, Ly Vu B, Leprince O, Buitink J. Characterization of dormancy behaviourin seeds of the model legume Medicago truncatula. Seed Sci Res 2010;20:97–107.

Choi HK, Mun JH, Kim DJ, Zhu HY, Baek JM, Mudge J, et al. Estimating genomeconservation between crop and model legume species. Proc Natl Acad Sci USA2004;101:15289–94.

Cook DR. Medicago truncatula – a model in the making! Commentary. Curr OpinPlant Biol 1999;2:301–4.

Conte SS, Walker EL. Transporters contribute to iron trafficking in plants. Mol Plant2011;4:464–76.

Curie C, Briat JF. Iron transport and signalling in plants. Annu Rev Plant Biol2003;54:183–216.

de Lorenzo L, Merchan F, Blanchet S, Megias M, Frugier F, Crespi M, et al. Differentialexpression of the TFIIIA regulatory pathway in response to salt stress betweenMedicago truncatula genotypes. Plant Physiol 2007;145:1521–32.

Ding H, Duan L, Li J, Yan H, Zhao M, Zhang F, et al. Cloning and functional anal-

ysis of the peanut iron transporter AhIRT1 during iron deficiency stress andintercropping with maize. J Plant Physiol 2010;167:996–1002.

Durrett TP, Gassmann W, Rogers EE. The FRD3-mediated efflux of citrate intothe root vasculature is necessary for efficient iron translocation. Plant Physiol2007;144:197–205.

Physi

E

E

G

G

G

G

G

H

H

I

K

K

K

L

L

L

L

M

N

R

R

R

G. Li et al. / Journal of Plant

ckhardt U, Marques AM, Buckhout TJ. Two iron-regulated cation transportersfrom tomato complement metal uptake-deficient yeast mutants. Plant Mol Biol2000;45:437–48.

llwood SR, D’Souza NK, Kamphuis LG, Burgess TI, Nair RM, Oliver RP. SSR analysis ofthe Medicago truncatula SARDI core collection reveals substantial diversity andunusual genotype dispersal throughout the Mediterranean basin. Theor ApplGenet 2006;112:977–83.

aige AR, Doerksen T, Shuai B. Medicago truncatula ecotypes A17 and R108 showvariations in jasmonic acid/ethylene induced resistance to Macrophomina phase-olina. Can J Plant Pathol 2012;34:98–103.

arcia MJ, Suarez V, Romera FJ, Alcantara E, Perez-Vicente R. A new model involvingethylene, nitric oxide and Fe to explain the regulation of Fe-acquisition genes inStrategy I plants. Plant Physiol Biochem 2011;49:537–44.

raziano M, Beligni MV, Lamattina L. Nitric oxide improves internal iron availabilityin plants. Plant Physiol 2002;130:1852–9.

reen LS, Rogers EE. FRD3 controls iron localization in Arabidopsis. Plant Physiol2004;136:2523–31.

uerinot M, Yi Y. Iron: nutritious, noxious, and not readily available. Plant Physiol1994;104:815–20.

ell R, Stephan UW. Iron uptake, trafficking and homeostasis in plants. Planta2003;216:541–51.

offmann B, Trinh TH, Leung J, Kondorosi A, Kondorosi E. A new Medicago trun-catula line with superior in vitro regeneration, transformation, and symbioticproperties isolated through cell culture selection. Mol Plant-Microbe Interact1997;10:307–15.

noue H, Higuchi K, Takahashi M, Nakanishi H, Mori S, Nishizawa NK. Three ricenicotianamine synthase genes, OsNAS1, OsNAS2, and OsNAS3 are expressed incells involved in long-distance transport of iron and differentially regulated byiron. Plant J 2003;36:366–81.

abir AH, Paltridge NG, Able AJ, Paull JG, Stangoulis JCR. Natural variationfor Fe-efficiency is associated with upregulation of Strategy I mechanismsand enhanced citrate and ethylene synthesis in Pisum sativum L. Planta2012;235:1409–19.

ende H. Ethylene biosynthesis. Annu Rev Plant Physiol Plant Mol Biol1993;44:283–307.

obayashi T, Nishizawa NK. Iron uptake, translocation and regulation in higherplants. Annu Rev Plant Biol 2012;63:131–52.

i YS, Mao XT, Tian QY, Li LH, Zhang WH. Phosphorus deficiency-induced reductionin root hydraulic conductivity in Medicago falcata is associated with ethyleneproduction. Environ Exp Bot 2009;67:172–7.

ivak KJ, Schmittgen TD. Analysis of relative gene expression data using real-timequantitative PCR and the 2−��CT method. Methods 2001;25:402–8.

ópez-Millán A-F, Ellis DR, Grusak MA. Identification and characterization of severalnew members of the ZIP family of metal Ion transporters in Medicago truncatula.Plant Mol Biol 2004;54:583–96.

ucena C, Waters BM, Romera FJ, Garcia MJ, Morales M, Alcantara E, et al. Eth-ylene could influence ferric reductase, iron transporter, and H+-ATPase geneexpression by affecting FER (or FER-like) gene activity. J Exp Bot 2006;57:4145–54.

orrissey J, Guerinot ML. Iron uptake and transport in plants: the good, the bad, andthe ionome. Chem Rev 2009;109:4553–67.

ozoye T, Nagasaka S, Kobayashi T, Takahashi M, Sato Y, Sato Y, et al.Phytosiderophore efflux transporters are crucial for iron acquisition in gram-inaceous plants. J Biol Chem 2011;286:5446–54.

obinson NJ, Procter CM, Connolly EL, Guerinot ML. A ferric-chelate reductase for

iron uptake from soils. Nature 1999a;6721:694–7.

omera FJ, Alcántara E. Ethylene involvement in the regulation of Fe-deficiencystress responses by Strategy I plants. Funct Plant Biol 2004;31:315–28.

ömheld V, Marschner H. Iron deficiency stress induced morphological and physi-ological changes in root tips of sunflower. Plant Physiol 1981;53:354–60.

ology 171 (2014) 639–647 647

Römheld V, Marschner H. Strategies of plants for acquisition of iron. Plant Soil1994;165:261–74.

Rellan-Alvarez R, Giner-Martinez-Sierra J, Orduna J, Orera I, Rodriguez-Castrillon JA,Garcia-Alonso JI, et al. Identification of a tri-Iron(III), tri-citrate complex in thexylem sap of iron-deficient tomato resupplied with iron: new insights into plantIron long-distance transport. Plant Cell Physiol 2009;51:91–102.

Robinson NJ, Procter CM, Connolly EL, Guerinot ML. A ferric-chelate reductase foriron uptake from soils. Nature 1999b;397:694–7.

Rodriguez-Celma J, Lin WD, Fu GM, Abadia J, Lopez-Millan AF, Schmidt W. Mutu-ally exclusive alterations in secondary metabolism are critical for the uptake ofinsoluble iron compounds by Arabidopsis and Medicago truncatula. Plant Physiol2013;162:1473–85.

Rogers EE, Guerinot ML. FRD3, a member of the mutidrug and toxin effluxfamily controls iron deficiency responses in arabidopsis. Plant Cell 2002;14:1787–99.

Romera FJ, Alcántara E. Iron-deficiency stress responses in cucumber (Cucumis-sativus L.) roots – a possible role for ethylene. Plant Physiol 1994;105:1133–8.

Roschzttardtz H, Séguéla-Arnaud M, Briat J-F, Vert G, Curie C. The FRD3 citrateeffluxer promotes iron nutrition between symplastically disconnected tissuesthroughout Arabidopsis development. Plant Cell 2011;23:2725–37.

Salzer P, Feddermann N, Wiemken A, Boller T, Staehelin C. Sinorhizobium meliloti-induced chitinase gene expression in Medicago truncatula ecotype R108-1: acomparison between symbiosis-specific class V and defence-related class IVchitinases. Planta 2004;219:626–38.

Schmidt W. Iron solutions: acquisition strategies and signaling pathways in plants.Trends Plant Sci 2003;8:188–93.

Schnurr JA, Jung H-JG, Samac DA. A comparative study of alfalfa and stemtraits: morphology, chemical composition, and ruminal digestibility. Crop Sci2007;47:1672–80.

Séguéla M, Briat JF, Vert G, Curie C. Cytokinins negatively regulate the root ironuptake machinery in Arabidopsis through a growth-dependent pathway. PlantJ 2008;55:289–300.

Sun P, Tian QY, Zhao MG, Dai XY, Huang JH, Li LH, et al. Aluminum-induced ethyleneproduction is associated with inhibition of root elongation in Lotus japonicus L.Plant Cell Physiol 2007;22:276–88.

Sun P, Tian QY, Chen J, Zhang WH. Aluminum-induced inhibition of root elongationin Arabidopsis is mediated by ethylene and auxin. J Exp Bot 2010;61:347–56.

Vert GA, Grotz N, Dédaldéchamp F, Gaymard F, Guerinot ML, Briat J-F, et al. IRT1,an Arabidopsis transporter essential for Iron uptake from the soil and for plantgrowth. Plant Cell 2002;14:1223–33.

Wang BL, Li YS, Zhang WH. Brassinosteroids are involved in response of cucumber(Cucumis sativus) to iron deficiency. Ann Bot 2012;110:681–8.

Wang Z, Shen JB, Zhang FS. Cluster-root formation, carboxylate exudation and protonrelease of Lupinus pilosus as affected by medium pH and P deficiency. Plant Soil2006;287:247–356.

Wang TZ, Chen L, Zhao MG, Tian QY, Zhang WH. Identification of drought-responsivemicroRNAs and their targets in Medicago truncatula by genome-wide high-throughput sequencing. BMC Genomics 2011;12:367.

Waters BM, Lucena C, Romera FJ, Jester GG, Wynn AN, Rojas CL, et al. Ethyleneinvolvement in the regulation of the H+-ATPase CsHA1 gene and of the newisolated ferric reductase CsFRO1 and iron transporter CsIRT1 genes in cucumberplants. Plant Physiol Biochem 2007;45:293–301.

WHO. Micronutrient deficiency: iron deficiency anaemia; 2007, http://www.who.int/nutrition/topics/ida/en/index.html

Yi Y, Guerinot MG. Genetic evidence that induction of root Fe (III) chelate

reductase activity is necessary for ion uptake under iron deficiency. Plant J1996;10:835–44.

Young ND, Cannon SB, Sato S, Kim D, Cook DR, Town CD, et al. Sequenc-ing the genespaces of Medicago truncatula and Lotus japonicus. Plant Physiol2005;37:1174–81.