at SciVerse ScienceDirect

Plant Physiology and Biochemistry 53 (2012) 46e53

Contents lists available

Plant Physiology and Biochemistry

journal homepage: www.elsevier .com/locate/plaphy

Research article

Constitutive expression of a barley Fe phytosiderophore transporter increasesalkaline soil tolerance and results in iron partitioning between vegetative andstorage tissues under stress

Sonia Gómez-Galera a, Duraialagaraja Sudhakar b, Ana M. Pelacho c, Teresa Capell a, Paul Christou a,d,*

aDepartment of Plant Production and Forestry Science, ETSEA, University of Lleida-CRA, Av. Alcalde Rovira Roure 191, E-25198 Lleida, SpainbDepartment of Plant Molecular Biology and Biotechnology, Centre for Plant Molecular Biology, Tamil Nadu Agricultural University, Coimbatore 641 003, IndiacDepartment of Hortofruticulture, Botany and Gardening, ETSEA, University of Lleida-CRA, Av. Alcalde Rovira Roure 191, E-25198 Lleida, Spaind ICREA, Passeig Lluís Companys 23, 08010 Barcelona, Spain

a r t i c l e i n f o

Article history:Received 13 October 2011Accepted 13 January 2012Available online 21 January 2012

Keywords:Transgenic riceIron uptakeAlkalinityToleranceIron-phytosiderophore transporter

Abbreviations: HvYS1, barley iron-phytosideroinductively coupled plasma mass spectrometry; Mpanesulfonic acid; NA, nicotianamine; PS, phytoubiquitin-1 promoter.* Corresponding author. Department of Plant Prod

ETSEA, University of Lleida-CRA, Av. Alcalde Rovira RouTel.: þ34 973702831; fax: þ34 973238264.

E-mail address: christou@pvcf.udl.es (P. Christou).

0981-9428/$ e see front matter � 2012 Elsevier Masdoi:10.1016/j.plaphy.2012.01.009

a b s t r a c t

Cereals have evolved chelation systems to mobilize insoluble iron in the soil, but in rice this process israther inefficient, making the crop highly susceptible to alkaline soils. We therefore engineered rice toexpress the barley iron-phytosiderophore transporter (HvYS1), which enables barley plants to take upiron from alkaline soils. A representative transgenic rice line was grown in standard (pH 5.5) or alkalinesoil (pH 8.5) to evaluate alkaline tolerance and iron mobilization. Transgenic plants developed secondarytillers and set seeds when grown in standard soil although iron concentration remained similar in leavesand seeds compared to wild type. However, when grown in alkaline soil transgenic plants exhibitedenhanced growth, yield and iron concentration in leaves compared to the wild type plants which wereseverely stunted. Transgenic plants took up iron more efficiently from alkaline soil compared to wildtype, indicating an enhanced capacity to increase iron mobility ex situ. Interestingly, all the additionaliron accumulated in vegetative tissues, i.e. there was no difference in iron concentration in the seeds ofwild type and transgenic plants. Our data suggest that iron uptake from the rhizosphere can be enhancedthrough expression of HvYS1 and confirm the operation of a partitioning mechanism that diverts iron toleaves rather than seeds, under stress.

� 2012 Elsevier Masson SAS. All rights reserved.

1. Introduction

Many plants grow poorly in alkaline soils, which account forw30% of the world’s arable land [21]. One of the problems ofalkaline soils is that they limit iron uptake. Plants need iron tosynthesize chlorophyll and to carry out photosynthesis, so irondeficiency results in chlorosis, poor growth and reduced yields. Ironin alkaline soils is present mostly as Fe3þ, which is insoluble andtherefore inaccessible to the plants. This limitation cannot easily beovercome using Fe2þ fertilizers because the soluble iron is rapidlyconverted into Fe3þ in situ [10]. Crops growing in alkaline soils also

phore transporter; ICP-MS,OPS, 3-(N-morpholino)pro-siderophores; Ubi-1, maize

uction and Forestry Science,re 191, E-25198 Lleida, Spain.

son SAS. All rights reserved.

fail to accumulate iron in edible organs causing nutrient deficiencyin humans [10].

Plants have evolved different mechanisms to overcome ironlimitation in alkaline soils [2,9]. Non-graminaceous species releaseprotons into the rhizosphere to increase the solubility of Fe3þ byacidification, and the Fe3þ is then reduced to Fe2þ by a membrane-bound ferric reductase oxidase [25]. This allows the uptake of ironinto the root cells through iron-regulated transporter 1 (IRT1) [9].Iron absorption in this manner is known as strategy I. Gramina-ceous plants use a mechanism based on iron chelation, whichinvolves the secretion of molecules known as phytosiderophores(PS) and the subsequent absorption of PSeFe3þ complexes [14,21].This is known as strategy II, and it is considered more efficient thanstrategy I.

In graminaceous crops, alkaline tolerance correlates with theamount of PS secreted into the soil, and cereals can be ranked in thefollowing order starting with the most tolerant: barley/wheat > oat/rye > maize/sorghum > rice [27]. Among the staplecereal crops, rice is therefore themost susceptible to iron deficiencyin alkaline soils. Susceptibility of rice to such alkaline conditions is

S. Gómez-Galera et al. / Plant Physiology and Biochemistry 53 (2012) 46e53 47

thought to occur due to the low amounts of PS that are secreted bythe roots. Rice is the main source of calories for nearly half theworld’s population, with 40% and 70% of individuals relying solelyon rice in Asia and Africa, respectively [44]. Therefore rice isarguably the most important target for genetic engineering strat-egies to address alkaline tolerance.

Genetic engineering has been used to introduce genes encodingiron transporters, iron reductases and enzymes involved in PSbiosynthesis into plants, and this has enhanced the uptake andaccumulation of iron in several crops, especially under iron-limiting conditions [2,9,10,45]. ‘Strategy I’ transgenic rice plantshave been engineered to express yeast iron reductases [13], as wellas transcription factors controlling the expression of genes inducedby iron deficiency, such as IDEF1 and OsIRO2 [20,28,29]. ‘Strategy II’transgenic rice plants have also been reported, expressing barleygenes encoding enzymes involved in PS biosynthesis such as IDS3(2-oxoglutarate-dependent dioxygenase; IDS ¼ iron deficiencyspecific), nicotianamine synthase (NAS) and nicotianamineaminotransferase (NAAT). Rice plants expressing these genesproduced and secreted more PS than wild type plants and weremore tolerant to alkaline soils both in the laboratory [12,17,18,22]and under iron-limiting conditions in the field [36]. In addition,seeds from transgenic rice plants expressing IDS3, NAS2 andOsIRO2 contained more iron than their wild type counterparts[15,17,18,28,36].

In strategy II plants, iron-phytosiderophore transporters arerequired to import PSeFe3þ complexes into root cells, and these areencoded by YS1 genes, named after the yellow stripe phenotypeobserved when the first such gene was discovered in maize [7].ZmYS1 transports several ions as PS complexes, including Fe3þ,Fe2þ, Ni2þ, Zn2þ, Cu2þ, Mn2þ and Cd2þ. In addition, ZmYS1 alsotransports the nicotianamine complexes Fe2þeNA and Ni2þeNA[27,34]. Murata et al. [27] cloned and characterized the corre-sponding barley gene (HvYS1) whose expression was shown to beroot-specific and inducible under iron-limiting conditions. Inter-estingly, although HvYS1 is closely related to ZmYS1, the barleyprotein is specific for PSeFe3þ complexes [27].

We reasoned that a heterologous PSeFe3þ transporterexpressed in rice would increase the accumulation of iron underthe limiting conditions imposed by alkaline soils, and that HvYS1

Fig. 1. Northern blot analysis of HvYS1 expression in the leaves and roots of 16 independentthese three lines. Eight of the 16 rice transformants showed HvYS1 expression. WT, wild ty

would be an ideal candidate because its specificity would preventthe displacement of iron by less desirable minerals. We thereforeproduced transgenic rice plants expressing the transgene stronglyand consistently, and expanded one line to test T1 plants for alka-line tolerance and iron accumulation. Although the transgenicplants were more tolerant to alkaline soils than controls and tookup more iron under limiting conditions, additional iron accumu-lated in vegetative tissues rather than seeds, suggesting the oper-ation of a partitioning mechanism for iron under stress.

2. Results

2.1. Identification of transgenic rice lines expressing HvYS1

Bombarded rice embryos were regenerated under hygromycinselection leading to the recovery of 16 independent transgeniclines, each of which was evaluated by northern blot analysis todetect HvYS1 steady state mRNA in the roots and leaves. Eight linesshowed strong and consistent HvYS1 expression in both tissues(Fig. 1). Although all the lines produced seeds, line L16 was selectedfor analysis because it produced enough seeds for the statistically-relevant testing of next generation plants under standard andalkaline soil conditions. The seeds were germinated on mediumcontaining hygromycin to screen out negative segregants, and tenplantlets each were transferred to pots containing standard soilfacilitating iron uptake (pH 5.5) and alkaline soil with low ironmobility (pH 8.5).

2.2. Transgene expression in L16 T1 plants

Transgene expression in T1 plants was determined by northernblot using mRNA extracted from the leaves of two plants from eachgroup at the beginning and end of the 8 week treatment period. Asshown in Fig. 2, high-level transgene expression was observedunder both treatment regimens at both sampling time points,whereas no expression was observed in wild type controls as ex-pected. These results confirmed that the transgene is constitutivelyexpressed in T1 plants and is not influenced by the different soilconditions.

rice transformants. (*) There was insufficient material for mRNA analysis in the roots ofpe control. Numbers indicate independent transgenic lines.

Fig. 2. Northern blot analysis of HvYS1 expression in L16 T1 plants growing in standard (Std) or alkaline (Alk) soil at different time points (0 W, start of treatment; 8 W, after 8 weeksof treatment). M, marker; WT, wild type control. Each lane number corresponds to one independent plant under the treatment regime.

S. Gómez-Galera et al. / Plant Physiology and Biochemistry 53 (2012) 46e5348

2.3. Growth and development parameters

There was no significant difference in the average height of wildtype and transgenic plants grown in standard soil after 8 weeks(wild type ¼ 39.60 cm � 1.92, L16 ¼ 39.53 � 3.25 cm) as shown inFig. 3a. However, there was a significant difference (p< 0.05) underalkaline conditions, where the transgenic plants reached anaverage height of 40.67 � 3.73 cm and the wild type plants werestunted at 26.38 � 2.55 cm (Figs. 3b and 4). The transgenic plantsgained a small height advantage during early growth (weeks 1e4)but the differences became clear from 5 to 8 weeks (Figs. 3b and 4).

Fig. 3. Height of L16 T1 rice plants and wild type (WT) growing in standard (a) oralkaline (b) soils. Data represent mean � standard error. Asterisks indicate a statisti-cally significant difference between L16 and WT (p < 0.05); n (in legend) indicates thenumber of plants assessed weekly for L16 and WT.

L16 plants produced secondary tillers under both standard andiron-limiting conditions, whereas wild type plants failed toproduce secondary tillers even when the plants were growing instandard soil (Fig. 4), probably reflecting a limitation of rice fortaking up part of the iron present in the rhizosphere (Table 1). Thetotal number of seeds and the mean of seeds per plant produced byL16 plants was also higher than that of wild type plants under bothconditions, but this difference was exacerbated and statisticallysignificant (p < 0.05) under iron-limiting conditions (Table 1).These data suggest that L16 plants generally experience a more

Fig. 4. Phenotypes of wild type (WT) and L16 T1 rice plants after 0, 4 and 8 weeks (W)growing in standard or alkaline soils. Red arrows indicate presence of flowering spikes.(For interpretation of the references to colour in this figure legend, the reader isreferred to the web version of this article.)

Table 1Number of tillers and seeds produced by wild type (WT) and L16 T1 rice plants growing in standard and alkaline soils. Numbers in a column followed by different letter aresignificantly different (p > 0.05).

Number oftreated plants

Number ofsecondary tillers

Total numberof seeds

Number of treatedplants setting seeds

Number of seeds/plant(Mean � standard error)

Standard soil WT 7 0 400 a 6 66.67 � 12.75 aL16 8 6 559 a 7 79.86 � 8.86 a

Alkaline soil WT 9 0 74 b 2 37.00 � 5.00 bL16 7 4 340 a 5 68.00 � 13.52 a

S. Gómez-Galera et al. / Plant Physiology and Biochemistry 53 (2012) 46e53 49

permissive environment for enhanced growth and seed set whichis maintained (albeit attenuated) under iron-limiting conditions.

2.4. Chlorophyll content

There were no significant differences in chlorophyll contentbetween L16 and wild type plants after 8 weeks growing in stan-dard soil. The SPAD value for wild type plants was 27.43 � 1.70 andthat for L16 plants was 29.83 � 1.44 (Fig. 5a). In contrast, there wasa significant difference (p< 0.05) between L16 and wild type plantsgrowing in the alkaline soil, which began to emerge during thethird week of treatment (Fig. 5b). The SPAD value of L16 plants fellonly slightly from 40.70 � 1.94 at the beginning of treatment to37.03 � 2.70 after 8 weeks, with no overt changes in phenotype. Incontrast, the SPAD value of wild type plants fell from 36.93 � 2.32to 20.48 � 1.25 over the same period, and the leaves showed clearevidence of chlorosis from week 3 (Figs. 4 and 5b)

2.5. Iron concentration

There was no significant difference in iron concentrationbetween the leaves of L16 and wild type plants that had been

Fig. 5. Chlorophyll content of L16 T1 rice plants and wild type (WT) growing instandard (a) or alkaline (b) soils. Data represent means � standard error. Asterisksindicate a statistically significant difference between L16 and WT (p < 0.05); n (inlegend) indicates the number of plants assessed weekly for L16 and WT.

growing in standard soil for 8 weeks at p� 0.05 (Table 2). However,iron concentration in L16 leaves was significantly higher than inwild type leaves (p < 0.05) after 8 weeks in alkaline soil, with thewild type leaves and L16 leaves containing 36.77 � 0.24 and54.60� 0.61 mg of iron per gram dry leaf weight, respectively. Thesedata indicate that the transgenic plants not only mobilize ironmoreeffectively in alkaline soil but also accumulate more iron than theirwild type counterparts.

Interestingly, there was no significant difference in ironconcentration when comparing the seeds of L16 and wild typeplants regardless of the soil type at p � 0.05 (Table 2). The trans-genic seeds had a higher iron concentration thanwild type seeds instandard soil, although the difference was not significant atp � 0.05, whereas transgenic seeds had lower iron concentrationthan wild type seeds in alkaline soil, although again the differencewas not significant at p � 0.05. These data indicate that althoughtransgenic plants canmobilize and absorb more iron thanwild typeplants in alkaline soil, the iron is diverted to vegetative rather thanstorage tissues.

3. Discussion

The ability of cereals to withstand alkaline soils depends ona mechanism involving the secretion and reabsorbtion of small,high-affinity iron-chelating compounds known as phytosider-ophores [12,21,22]. These are among the strongest soluble Fe3þ-binding agents known in nature and their efficiency stems fromtheir ability to overcome the attraction between Fe3þ and anionicmineral components in soils [41]. Barley is one of the hardiestcereal species in an alkaline environment and this is thought tooccur due to the synthesis and secretion of abundant PS into therhizosphere followed by rapid and efficient reabsorbtion [27,37]. Incontrast, rice produces and absorbs only small amounts of PS andis therefore susceptible to iron deficiency when growing in alka-line soils [23]. Rice is well adapted to submerged conditions(where Fe2þ is more abundant than Fe3þ) and, in addition to thestrategy II [14], can deploy a strategy I mechanism for Fe2þ

acquisition based on the reduction of ferric iron followed byuptake through the OsIRT1 transporter [3]. However, rice plantsgrown in alkaline soils show deficiency symptoms even underwaterlogged conditions because they are unable to induce ferricchelate reductase [13].

Table 2Iron concentration (mg iron per gram dryweight) in the leaves and seeds of wild type(WT) and L16 T1 rice plants growing in standard and alkaline soils. Numbers ina column followed by different letter are significantly different (p > 0.05).

Numberof treatedplants

Leaves(Mean � standarderror)

Seeds(Mean � standarderror)

Standardsoil

WT 7 48.60 � 1.32 b 25.52 � 2.48 aL16 8 50.33 � 0.47 b 30.59 � 3.41 a

Alkalinesoil

WT 9 36.77 � 0.24 c 24.18 � 5.58 abL16 7 54.60 � 0.61 a 19.40 � 1.66 b

S. Gómez-Galera et al. / Plant Physiology and Biochemistry 53 (2012) 46e5350

Several previous reports describe transgenic rice plants engi-neered to produce greater amounts of PS through the expression ofbarley enzymes involved in PS synthesis such as IDS3, HvNAS1 andHvNAAT. The resulting rice plants grew more robustly in alkalineand iron-limiting environments and were able to sequester largeramounts of iron under these conditions [12,17,18,22,36]. It isunlikely that the minimal PS synthesis observed in wild type rice ispaired with a hyper-effective reabsorbtion capability because itwould not be an efficient use of resources to produce more PStransporters than necessary.

The archetypal PS transporter is the maize YS1 protein [7] butthis is remarkably promiscuous and can mobilize a wide range ofdivalent cations in addition to Fe3þ, including Fe2þ, Ni2þ, Zn2þ,Cu2þ, Mn2þ and Cd2þ. The corresponding barley protein (HvYS1) ismuch more specific, and can only mobilize PS complexes contain-ing Fe3þ [27]. Because rice is an important food crop, it is morebeneficial to select for the accumulation of iron, an essentialnutrient which is low in cereal-based diets, than the diverse metalions listed above, especially those that are toxic (i.e. Cd2þ).

We therefore engineered rice with the barley HvYS1 gene andrecovered 16 putative transgenic lines under selection. Eight of thelines expressed HvYS1 mRNA in the leaves and roots at high levels.All lines produced T1 seeds, but L16 produced enough seeds forimmediate testing of the T1 plants for alkaline tolerance and wastherefore chosen for further analysis. We randomly selected 20 L16T1 plants and randomly assigned them to two groups, experimentaland control. The control group was grown in standard soil, with 10wild type plants grown under the same conditions. The experi-mental group was grown in locally-sourced alkaline soil containingimmobilized Fe3þ, again with 10 wild type plants for comparison.

Under standard conditions, the performance of the wild typeand transgenic plants was very similar (Fig. 4) and there were nosignificant differences in height (Fig. 3a), chlorophyll content(Fig. 5a) or iron concentration in the leaves or seeds at p � 0.05(Table 2). Under alkaline conditions, differences between the wildtype and transgenic plants became apparent within 3e4 weeks(Fig. 4), with the transgenic plants maintaining strong growthparameters but wild type plants becoming stunted (Fig. 3b), con-taining less chlorophyll (Fig. 5b) and losing up to 30% of the ironconcentration in the leaves (Table 2). These three factors areprobably linked. The reduction in leaf iron concentration is likely tobe the direct consequence of less efficient iron absorption by theroots, which in turn starves the plant of iron required for chloro-phyll synthesis [24]. Iron is required in the active site of severalenzymes involved in the biosynthesis of chlorophyll, such asglutamyl-tRNA reductase, which synthesizes the precursor 5-aminolevulinic acid [40]. Iron is also part of the ironesulfurcomplex in photosystem I and acts as an electron donor andacceptor in photosystem II [11]. Iron limitation reduces thephotochemical conversion efficiency of photosystem II and there-fore reduces the accumulation of biomass, explaining the stuntedgrowth of wild type rice plants in alkaline soil.

Whereas HvYS1 expression had a negligible effect on height andiron concentration of transgenic plants under standard conditions,there were differences in the number of secondary tillers andnumber of seeds. These differences were exacerbated in the alka-line soil, showing that the transporter had an impact on the growthand development of vegetative and storage tissues even underconditions favorable for iron uptake. The transgenic plantsproduced secondary tillers under both standard and alkalineconditions whereas no secondary tillers were produced by wildtype plants in either case. The absence of secondary tillers in wildtype plants growing in standard soil suggests that rice growth islimited by iron availability, even when plants grow under idealconditions, and increasing iron availability through the expression

of the HvYS1 overcomes this limitation. This suggests that ironmight be critical for tiller formation, and limitation in iron uptakeappears to be a rather under-rated constraint for growth, a hypoth-esis supported by our observations (Table 1). Similarly, the totalnumber of seeds produced in standard soil wasw40% higher for thetransgenic plants (559) compared to wild type plants (400), eventhough they were no statistically different. Differences becamewider and significant in alkaline soil, where the number of seedsproduced by transgenic plants fell to 340 but the number of seedsproduced by wild type plants fell to 74 (Table 1). Reduction in thetotal number of seeds in plants growing in alkaline soil wasa consequence of the limited number of plants setting seeds, whichwas especially critical in wild type plants (only 2 out of 9 plants inthe treatment set seeds; Table 1). Number of seeds per plant wasalso reduced from transgenic (79.86 � 8.86) and wild type(66.67 � 12.75) plants growing in standard soil, compared totransgenic (68.00 � 13.52) and wild type (37.00 � 5.00) plantsgrowing in alkaline soil. Statistically significant differences were notobservedwhen comparing transgenic andwild type plants grown instandard conditions (p � 0.05), whereas significant differences(p < 0.05) were observed between plants grown under alkalineconditions (Table 1). In addition, reproduction (flowering and seeddevelopment) was initiated earlier in transgenic plants than inwildtype plants growing in standard or alkaline soil (Fig. 4). These datasuggest that iron limitation is a key developmental tipping point,and that even under favorable conditions the provision of additionaliron can be beneficial for reproductive development.

We also observed a significant difference in the distribution ofiron between vegetative and storage tissues under iron-limitingconditions. As shown in Table 2, wild type seeds had the sameiron concentration in both soil types, whereas the transgenic seedsshowed a marginal increase in iron concentration in standard soiland a small drop in alkaline soil, in sharp contrast to the increase iniron concentration observed in the leaves of the same plants. Thesedata strongly suggest the existence of an iron partitioning mecha-nism that is induced by the increased uptake of PSeFe3þ

complexes. Separate iron transport mechanisms for vegetativeand storage tissues have been proposed before [8,31,43]. It ispossible that the transgene expression profile may influence thisprocess because the HvYS1 transgene was expressed constitutivelyin ricewhereas the native protein is restricted to barley roots, and istherefore involved in primary iron transport from the rhizosphereinto the root cells but not with long-distance transport or seedloading [6,27,38].

Other studies have shown that increasing iron absorption leadsto the accumulation of iron in both vegetative and storage tissues,although these studies have involved regulatory proteins thatrespond to either iron deficiency or sufficiency [19], particularlydeficiency-specific transcription factors with multiple downstreamtargets such as OsIDEF1, OsIDEF2 [20] and OsIRO2 [28,29]. Forexample, Kobayashi et al. [16,19,20] found that OsIDEF1 is consti-tutively expressed in vegetative and reproductive tissues andactivates multiple iron homeostasis and utilization genes understandard or iron-limiting conditions. These authors suggested thatOsIDEF1 overexpression in rice confers alkaline tolerance byenhancing iron utilization rather than uptake [20] and increasesthe efficiency of CO2 assimilation and biomass production [32].This is consistent with the formation of additional secondary tillersin the transgenic plants in standard and alkaline soils, since in bothcases the HvYS1 transporter would increase the amount ofbioavailable iron within the plant. Similarly, the overexpression ofOsIRO2 also increased the tolerance of rice plants to alkalinesoils, and promoted the accumulation of iron in both vegetativetissues and seeds [37]. These transcription factors are thereforelikely to influence downstream genes that control iron transport

S. Gómez-Galera et al. / Plant Physiology and Biochemistry 53 (2012) 46e53 51

mechanisms for vegetative and storage tissues, whereas the over-expression of HvYS1 appears to affect predominantly the vegetativepathway.

Our results provide evidence that rice has a limited capacity fortaking up iron when grown in standard or alkaline soil, and thatincreasing absorption of iron from the rhizosphere promotes plantgrowth and development in both soils under waterlogged condi-tions. We conclude that the rice strategy II mechanism for ironabsorption is inefficient and unable to confer tolerance to alkalinesoils, but can be improved by the heterologous expression of thebarley HvYS1 transporter which selectively takes up PSeFe3þ

complexes from the rhizosphere. This strategy could be used todevelop new varieties of upland rice adapted to aerobic condi-tions, where iron limitation is a particular problem becausesoluble Fe2þ fertilizers are rapidly oxidized and immobilized[10,30]. Other alkaline-sensitive cereals (e.g. maize and sorghum)could be also engineered with HvYS1, and the more efficientPSeFe3þ uptake would help to counteract iron deficiency causedby microbes in the rhizosphere that obtain iron by degrading PSsecreted by plants [42]. Our data also confirms previous resultswhich concluded that separate and specific transport and sinkmechanisms for accumulating iron in seeds are required in addi-tion to the iron absorption mechanism from the rhizosphere;therefore, an appropriate strategy would be to combine enhancedPSeFe3þ uptake with the regulation of internal pathways toensure the accumulation of iron in edible organs in order toprovide nutritional benefits especially when plants are grown inalkaline soil [1,10,15,43,45].

4. Materials and methods

4.1. Gene cloning and vector construction

The HvYS1 cDNA (GenBank ID: AB214183.1) was cloned from theroots of 2-week-old barley plants (Hordeum vulgare L. cv. Ordalie)growing in vitro on MS medium without iron [26]. Total RNA wasextracted using the RNeasy� Plant Mini Kit (Qiagen, Hilden, Ger-many) and 1 mgwas reverse transcribed using the Omniscript RT Kit(Qiagen, Hilden, Germany). The full-size cDNA (2037 bp) wasamplified by PCR using forward primer HvYS1-BamHI-FOR (50-AGGATC CAT GGA CAT CGT CGC CCC GGA CCG CA-30) and reverse primerHvYS1-HindIII-REV (50-AAA GCT TTT AGG CAG CAG GTA GAA ACTTCA TG-30). The product was transferred to the pGEM�-T Easyvector (Promega, Madison, WI, USA) for sequencing and verifica-tion. The HvYS1 cDNA was then subcloned using the BamHI andHindIII sites and inserted into the expression vector pAL76 [4]which contains the maize Ubi-1 promoter and first intron, and anAgrobacterium tumefaciens nos transcriptional terminator.

4.2. Plant transformation and regeneration

Mature rice embryos (Oryza sativa L. cv EYI 105) were excisedand cultured as previously described [35,39]. After 7 days, theembryos were bombarded with gold particles carrying the trans-gene and selectable marker [5]. The rice embryos were incubatedon high-osmoticum medium (0.2 M mannitol, 0.2 M sorbitol) for4 h prior to bombardment. For transformation, the gold particleswere coated with the pAL76-HvYS1 plasmid and a second vectorcontaining the hygromycin phosphotransferase (hpt) selectablemarker [34,39]. Bombarded embryos were selected on mediumsupplemented with 30 mg L�1 hygromycin and callus pieces weretransferred sequentially to shooting and rooting medium contain-ing hygromycin as above.

Regenerated plantlets were transferred to pots containing soil(Traysubstract; Klasmann-Deilmann GmbH, Geeste, Germany) and

were cultivated under flooded conditions in a growth chamber at26 � 2 �C, with a 12-h photoperiod (900 mmohn m�2 s�1

photosynthetically-active radiation) and 80% relative humidity.Plants were irrigated with a soluble iron solution (Sequestrene 138Fe G-100; Syngenta Agro SA, Madrid, Spain) until they matured.Seeds were collected from line 16 (L16), which produced the largestnumber of seeds.

4.3. Germination and performance under iron-limiting conditions

L16 T1 seeds were sterilized in 70% ethanol for 30 min withagitation and then in commercial bleach (Lejía Conejo, HenkelIberica, Barcelona, Spain) for 3 min, and were germinated on MSmedium containing 30 mgL�1 hygromycin (26 � 2 �C, 16-hphotoperiod, 80% relative humidity). Wild type seeds weregerminated using the same procedure, but in medium lackinghygromycin. After 1 week, L16 T1 and wild type plantlets weretransferred to pots filled with standard cultivation soil (pH 5.6e6.5;Traysubstract, Klasmann-Deilmann GmbH, Geeste, Germany) andwere grown under flooded conditions for 5 weeks in a growthchamber (26 � 2 �C, 12-h photoperiod, 80% relative humidity,irrigation with iron solution as above).

To test alkaline tolerance, 20 six-week-old L16 T1 and wild typeplants were selected and divided into two groups of 10. One(control) group was maintained in standard soil whereas the other(experimental) group was transferred to pots filled with locallyobtained alkaline soil (pH 8.5). The potted plants were placed in1.5-L plastic containers and watered with iron-free nutrient solu-tion (0.7 mM K2SO4, 0.1 mM KCl, 0.1 mM KH2PO4, 0.5 mM MgSO4,2 mM Ca(NO3)2, 10 mMH3BO3, 0.5 mMMnSO4, 0.5 mMZnSO4, 0.2 mMCuSO4, 0.01 mM (NH4)6Mo7O24). In the control group, the pH of thenutrient solution was adjusted daily to 5.5 with 1 M HCl whereasthere was no adjustment in the experimental group. The nutrientsolution was completely renewed twice weekly until seedcollection.

4.4. Monitoring plant growth and development

The performance of wild type and transgenic plants undercontrol and experimental conditions was monitored by scoring forgrowth and development parameters (height, number of tillers,and number of seeds) and chlorophyll content. Plant height wasscored weekly for 8 weeks on the basis of the distance between thecrown and the node of the last fully-expanded leaf, and an averageheight was calculated for the control and experimental groups. Thenumber of tillers and seeds was determined at the experimentalend point. The chlorophyll content was scored weekly for 8 weeksand was calculated as the average of three independent measure-ments performed on the youngest fully-expanded leaf of each plantusing a portable SPAD-502 chlorophyll meter (Minolta Co., Tokyo,Japan). Average values for measurements are presented asmeans � standard error of the mean.

4.5. Expression analysis

Transgene expressionwas confirmed by northern blots of mRNAobtained from leaves (0.15 g) and roots (0.2 g). Plants wererandomly selected for sampling, with wild type and transgenicsamples taken from the same time points. Denatured RNA (15 mg)was fractionated by 1.2% w/v agarose-formaldehyde gel electro-phoresis in MOPS buffer [33] and transferred to a positively-charged nylon membrane (Roche Diagnostics GmbH, Mannheim,Germany) prior to UV cross-linking.

HvYS1 mRNA was detected using a probe generated withprimers HvYS1-BamHI-FOR and HvYS1-HindIII-REV (see above)

S. Gómez-Galera et al. / Plant Physiology and Biochemistry 53 (2012) 46e5352

labeled with digoxigenin using the PCR DIG Probe Synthesis Kit(Roche Diagnostics GmbH, Mannheim, Germany). Labeled probeswere purified using the QIAquick Gel Extraction Kit (Qiagen, Hilden,Germany) and denatured at 68 �C for 10 min prior to hybridizationovernight at 50 �C.

After hybridization, the membranes were washed twice for15 min in 2 � SSC, 0.1% SDS at room temperature, then twice for30 min in 0.5 � SSC, 0.1% SDS at 68 �C, once for 30 min in 0.2� SSC,0.1% SDS at 68 �C, and finally once for 10 min in 0.1 � SSC, 0.1% SDSat 68 �C. The bound probe was detected with an alkalinephosphatase-labeled anti-DIG Fab-fragment and the CSPD�

chemiluminescent substrate (Roche Diagnostics GmbH,Mannheim,Germany). The signal was revealed by exposure to X-ray film(Kodak Biomax light film, SigmaeAldrich, USA) at 37 �C for upto 3 h.

4.6. Determination of mineral levels

The youngest fully-expanded leaves were collected from wildtype and L16 T1 plants from the experimental and control groups 8weeks after treatment. We also selected 15e21 seeds from each ofthe plants. Collected samples (leaves and dehusked seeds) werewashed in milliQ water to remove any external mineral contami-nation and dried at 70 �C for 4 d in an oven. Dried tissues wereground to a fine powder and sent to Servicios Científico-Técnicos atOviedo University (Oviedo, Spain) for mineral analysis. Powdered0.1-g samples were suspended in HNO3 and H2O2 and digested ina microwave, then supplemented with ultra-pure water to a finalvolume of 20 mL. Aliquots were diluted 1:8, spiking with 57Fe, andthe mineral concentration was determined by isotopic dilution andinductively coupled plasma mass spectrometry (ICP-MS). Averagevalues for measurements are presented as means � standard errorof the mean.

4.7. Statistical analysis

Differences between transgenic and wild type plants weretested by ANOVA and subsequent comparison of means using the ttest with the residual mean square in the ANOVA as the estimate ofvariability.

Acknowledgements

We thank Dr. Changfu Zhu for his help and advice on genecloning. S.G.-G. is the recipient of a fellowship from the CatalanRegional Government and European Social Fund (number 2009FIC00136). D. S. is supported by the Biotechnology Overseas Asso-ciateship program of the DBT, Government of India. P.C. acknowl-edges the financial support of the Spanish Ministry of Science andInnovation (grant number BFU2007-61413) and the EuropeanResearch Council for the BIOFORCE advanced grant.

References

[1] K. Bashir, Y. Ishimaru, N.K. Nishizawa, Iron uptake and loading into rice grains,Rice 3 (2010) 122e130.

[2] P. Bauer, Z. Bereczky, Gene networks involved in iron acquisition strategies inplants, Agronomie 23 (2003) 447e454.

[3] N. Bughio, H. Yamaguchi, N.K. Nishizawa, H. Nakanishi, S. Mori, Cloning aniron-regulated metal transporter from rice, J. Exp. Bot. 53 (2002) 1677e1682.

[4] A.H. Christensen, P.H. Quail, Ubiquitin promoter based vectors for high-levelexpression of selectable and/or screenable marker genes in mono-cotyledonous plants, Transgenic Res. 5 (1996) 213e218.

[5] P. Christou, T.L. Ford, M. Kofron, Production of transgenic rice (Oryza sativa L.)plants from agronomically important indica and japonica varieties via electricdischarge particle acceleration of exogenous DNA into immature zygoticembryos, Biotechnology 9 (1991) 957e962.

[6] C. Curie, G. Cassin, D. Couch, F. Divol, K. Higuchi, M.L. Jean, J. Misson,A. Schikora, P. Czernic, S. Mari, Metal movement within the plant: contribu-tion of nicotianamine and yellow stripe 1-like transporters, Ann. Bot. 103(2009) 1e11.

[7] C. Curie, Z. Panavience, C. Loulegue, S.L. Dellaporta, J.F. Briat, E.L. Walker,Maize yellow stripe1 encodes a membrane protein directly involved in Fe(III)uptake, Nature 409 (2001) 346e349.

[8] G. Drakakaki, P. Christou, E. Stöger, Constitutive expression of soybeanferritin cDNA in transgenic wheat and rice results in increased iron levelsin vegetative tissues but not in seeds, Transgenic Res. 9 (2000) 445e452.

[9] A. Ghandilyan, D. Vreugdenhil, M.G.M. Aarts, Progress in the geneticunderstanding of plant iron and zinc nutrition, Physiol. Plant 126 (2006)407e417.

[10] S. Gómez-Galera, E. Rojas, D. Sudhakar, C. Zhu, A.M. Pelacho, T. Capell,P. Christou, Critical evaluation of strategies for mineral fortification of staplefood crops, Transgenic Res. 19 (2010) 165e180.

[11] R. Hänsch, R.R. Mendel, Physiological functions of mineral micronutrients (Cu,Zn, Mn, Fe, Ni, Mo, B, Cl), Curr. Opin. Plant Biol. 12 (2009) 259e266.

[12] K. Higuchi, S. Watanabe, M. Takahashi, S. Kawasaki, H. Nakanishi,N.K. Nishizawa, S. Mori, Nicotianamine synthase gene expression differs inbarley and rice under Fe-deficient conditions, Plant J. 25 (2001) 159e167.

[13] Y. Ishimaru, S. Kim, T. Tsukamoto, H. Oki, T. Kobayashi, S. Watanabe,S. Matsuhashi, M. Takahashi, H. Nakanishi, S. Mori, N.K. Nishizawa, Mutationalreconstructed ferric chelate reductase confers enhanced tolerance in rice toiron deficiency in calcareous soil, Proc. Natl. Acad. Sci. USA 104 (2007)7373e7378.

[14] Y. Ishimaru, M. Suzuki, T. Tsukamoto, K. Suzuki, M. Nakazono, T. Kobayashi,Y. Wada, S. Watanabe, S. Matsuhashi, M. Takahashi, H. Nakanishi, S. Mori,N.K. Nishizawa, Rice plants take up iron as an Fe3þ-phytosiderophore and asFe2þ, Plant J. 45 (2006) 335e346.

[15] A.A.T. Johnson, B. Kyriacou, D.L. Callahan, L. Carruthers, J. Stangoulis, E. Lombi,M. Tester, Constitutive overexpression of the OsNAS gene family revealssingle-gene strategies for effective iron- and zinc-biofortification of riceendosperm, Plos ONE 6 (2011) 1e11.

[16] T. Kobayashi, R.N. Itai, Y. Ogo, Y. Kakei, H. Nakanishi, M. Takahashi,N.K. Nishizawa, The rice transcription factor IDEF1 is essential for the earlyresponse to iron deficiency, and induces vegetative expression of lateembryogenesis abundant genes, Plant J. 60 (2009) 948e961.

[17] T. Kobayashi, H. Nakanishi, M. Takahashi, S. Kawasaki, N.K. Nishizawa, S. Mori,In vivo evidence that Ids3 from Hordeum vulgare encodes a dioxygenase thatconverts 20-deoxymugineic acid to mugineic acid in transgenic rice, Planta212 (2001) 864e871.

[18] T. Kobayashi, H. Nakanishi, M. Takahashi, S. Mori, N.K. Nishizawa, Generationand field trials of transgenic rice tolerant to iron deficiency, Rice 1 (2008)144e153.

[19] T. Kobayashi, Y. Ogo, M.S. Aung, T. Nozoye, R.N. Itai, H. Nakanishi,T. Yamakawa, N.K. Nishizawa, The spatial expression and regulation of tran-scription factors IDEF1 and IDEF2, Ann. Bot. 105 (2010) 1109e1117.

[20] T. Kobayashi, Y. Ogo, R.N. Itai, H. Nakanishi, M. Takahashi, S. Mori,N.K. Nishizawa, The transcription factor IDEF1 regulates the response to andtolerance of iron deficiency in plants, Proc. Natl. Acad. Sci. USA 104 (2007)19150e19155.

[21] S. Mori, Iron acquisition by plants, Curr. Opin. Plant Biol. 2 (1999)250e253.

[22] S. Mori, H. Nakanishi, M. Takahashi, K. Higuchi, N.K. Nishizawa, Geneticengineering of transgenic rice with barley strategy-II genes, in: W. Horst,M.K. Schenk, A. Bürkert, N. Claassen, H. Flessa, W.B. Frommer,H.E. Goldbach, H.W. Olfs, V. Römheld, B. Sattelmacher, U. Schmidhalter,S. Schubert, N. von Wirén, L. Wittenmayer (Eds.), Plant Nutrition: FoodSecurity and Sustainability of Agro-Ecosystems Through Basic and AppliedResearch (Developments in Plant and Soil Sciences), Springer, New York,2001, pp. 14e15.

[23] S. Mori, N. Nishizawa, H. Hayashi, M. Chino, E. Yoshimura, J. Ishihara, Why areyoung rice plants highly susceptible to iron deficiency? Plant Soil 130 (1991)143e156.

[24] J.L. Moseley, T. Allinger, S. Herzog, P. Hoerth, W. Wehinger, S. Merchant,M. Hippler, Adaptation to Fe-deficiency requires remodeling of the photo-synthetic apparatus, EMBO J. 21 (2002) 6709e6720.

[25] I. Mukherjee, N.H. Campbel, J.S. Ash, E.L. Connolly, Expression profiling of theArabidopsis ferric chelate reductase (FRO) gene family reveals differentialregulation by iron and copper, Planta 223 (2006) 1178e1190.

[26] T. Murashige, F. Skoog, A revised medium for rapid growth and bioassays withtobacco tissue cultures, Physiol. Plant 15 (1962) 473e497.

[27] Y. Murata, J.F. Ma, N. Yamaji, D. Ueno, K. Nomoto, T. Iwashita, A specifictransporter for iron(III)-phytosiderophore in barley roots, Plant J. 46 (2006)563e572.

[28] Y. Ogo, R.N. Itai, T. Kobayashi, M.S. Aung, H. Nakanishi, N.K. Nishizawa, OsIRO2is responsible for iron utilization in rice and improves growth and yield incalcareous soil, Plant Mol. Biol. 75 (2011) 593e605.

[29] Y. Ogo, R.N. Itai, H. Nakanishi, T. Kobayashi, M. Takahashi, S. Mori,N.K. Nishizawa, The rice bHLH protein OsIRO2 is an essential regulator of thegenes involved in Fe uptake under Fe-deficient conditions, Plant J. 51 (2007)366e377.

[30] S. Pal, S.P. Datta, R.K. Rattan, A.K. Singh, Diagnosis and amelioration of irondeficiency under aerobic rice, J. Plant Nutr. 31 (2009) 919e940.

S. Gómez-Galera et al. / Plant Physiology and Biochemistry 53 (2012) 46e53 53

[31] L.Q. Qu, T. Yoshihara, A. Ooyama, F. Goto, F. Takaiwa, Iron accumulation doesnot parallel the high expression level of ferritin in transgenic rice seeds, Planta222 (2005) 225e233.

[32] K. Ravet, B. Touraine, J. Boucherez, J.F. Briat, F. Gaymard, F. Cellier, Ferritinscontrol interaction between iron homeostasis and oxidative stress in Arabi-dopsis, Plant J. 57 (2009) 400e412.

[33] J. Sambrook, E.F. Fritsch, T. Maniatis, Molecular Cloning: a LaboratoryManual, second ed. Cold Spring Harbor Laboratory Press, Cold SpringHarbor, 1989.

[34] G. Schaaf, U. Ludewig, B.E. Erenoglu, S. Mori, T. Kitahara, N. von Wirén, ZmYS1functions as a proton-coupled symporter for phytosiderophore- andnicotianamine-chelated metals, J. Biol. Chem. 279 (2004) 9091e9096.

[35] D. Sudhakar, L.T. Duc, B.B. Bong, P. Tinjuangjun, S.B. Maqbool, M. .Valdez,R. Jefferson, P. Christou, An efficient rice transformation system utilizingmature seed-derived explants and a portable, inexpensive particlebombardment device, Transgenic Res. 7 (1998) 289e294.

[36] M. Suzuki, K.C. Morikawa, H. Nakanishi, M. Takahashi, M. Saigusa, S. Mori,N.K. Nishizawa, Transgenic rice lines that include barley genes have increasedtolerance to low iron availability in a calcareous paddy soil, Soil Sci. PlantNutr. 54 (2008) 77e85.

[37] M. Suzuki, M. Takahashi, T. Tsukamoto, S. Watanabe, S. Matsuhashi, J. Yazaki,N. Kishimoto, S. Kikuchi, H. Nakanishi, S. Mori, N.K. Nishizawa, Biosynthesisand secretion of mugineic acid family phytosiderophores in zinc-deficientbarley, Plant J. 48 (2006) 85e97.

[38] D. Ueno, N. Yamaji, J.F. Ma, Further characterization of ferric-phytosiderophore transporters ZmYS1 and HvYS1 in maize and barley,J. Exp. Bot. 60 (2009) 3513e3520.

[39] M. Valdez, J.L. Cabrera-Ponce, D. Sudhakhar, L. Herrera-Estrella, P. Christou,Transgenic Central American, West African and Asian elite rice varietiesresulting from particle bombardment of foreign DNA into mature seed-derived explants utilizing three different bombardment devices, Ann. Bot.82 (1998) 795e801.

[40] D. von Wettstein, S. Gough, C.G. Kannangara, Chlorophyll biosynthesis, PlantCell 7 (1995) 1039e1057.

[41] N. Von Wirén, H. Khodr, R.C. Hider, Hydroxylated phytosiderophores speciespossess an enhanced chelate stability and affinity for iron (III), Plant Physiol.124 (2000) 1149e1157.

[42] N. Von Wirén, V. Römheld, J.L. Morel, A. Guckert, H. Marschner, Influence ofmicroorganisms on iron acquisition in maize, Soil Biol. Biochem. 25 (1993)311e316.

[43] J. Wirth, S. Poletti, B. Aeschlimann, N. Yakandawal, B. Drosse, S. Osorio,T. Tohge, A.R. Fernie, D. Günther, W. Gruissem, C. Sautter, Rice endosperm ironbiofortification by targeted and synergistic action of nicotianamine synthaseand ferritin, Plant Biotechnol. J. 7 (2009) 631e644.

[44] R.S. Zeigler, A. Barclay, The relevance of rice,, Rice 1 (2008) 3e10.[45] C. Zhu, S. Naqvi, S. Gomez-Galera, A.M. Pelacho, T. Capell, P. Christou, Trans-

genic strategies for the nutritional enhancement of plants, Trends Plant Sci. 12(2007) 548e555.