Hun Liu，Aijz Soomro，Yjun Zhu，Xinjin Qiu，Ki Chen，Tinqing Zheng，Longwei Yng，Dnying Xing，*，Jinlong Xu，，d，**

aHubei Collaborative Innovation Center for Grain Industry，Yangtze University，Jingzhou 434025，China

bInstitute of Crop Science/National Key Facility for Crop Gene Resources and Genetic Improvement，Chinese Academy of Agricultural Sciences，Beijing 100081，China

cAgricultural Genomics Institute at Shenzhen，Chinese Academy of Agricultural Sciences，Shenzhen 518120，China

dShenzhen Institute of Breeding for Innovation，Chinese Academy of Agricultural Sciences，Shenzhen 518120，China

?

QTL underlying iron and zinc toxicity tolerances at seedling stage revealed by two sets of reciprocal introgression populations of rice（Oryza sativa L.）

Huan Liua，1，Aijaz Soomrob，1，Yajun Zhuc，Xianjin Qiua，Kai Chenc，Tianqing Zhengb，Longwei Yanga，Danying Xinga，*，Jianlong Xub，c，d，**

aHubei Collaborative Innovation Center for Grain Industry，Yangtze University，Jingzhou 434025，China

bInstitute of Crop Science/National Key Facility for Crop Gene Resources and Genetic Improvement，Chinese Academy of Agricultural Sciences，Beijing 100081，China

cAgricultural Genomics Institute at Shenzhen，Chinese Academy of Agricultural Sciences，Shenzhen 518120，China

dShenzhen Institute of Breeding for Innovation，Chinese Academy of Agricultural Sciences，Shenzhen 518120，China

A R T I C L E I N F O

Article history:

in revised form

14 May 2016

Accepted 6 June 2016

Available online 14 June 2016

Rice

Reciprocal introgression lines

Iron tolerance

Zinc tolerance

Quantitative trait locus/loci（QTL）

A B S T R A C T

Iron and zinc are two trace elements that are essential for rice.But they are toxic at higher concentrations，leadingtoseverericeyieldlossesespeciallyinacidsoilsandinlandvalleys.Inthis study，two reciprocal introgression line（IL）populationssharing the same parentswereused with high-density SNP bin markers to identify QTL tolerant to iron and zinc toxicities.The results indicatedthatthejaponicavariety02，428hadstrongertolerancetoironandzinctoxicitiesthanthe indica variety Minghui 63.Nine and ten QTL contributing to iron and zinc toxicity tolerances，respectively，were identified in the two IL populations.The favorable alleles of most QTL came from 02，428.Among them，qFRRDW2，qZRRDW3，and qFRSDW11 appeared to be independent of genetic background.The region C11S49-C11S60 on chromosome 11 harbored QTL affecting multiple iron and zinc toxicity tolerance-related traits，indicating partial genetic overlap between thetwotoxicitytolerances.Ourresultsprovideessentialinformationandmaterialsfordeveloping excellent rice cultivars with iron and/or zinc tolerance by marker-assisted selection（MAS）.

?2016 Crop Science Society of China and Institute of Crop Science，CAAS.Production and hosting by Elsevier B.V.This is an open access article under the CC BY-NC-ND license（http://creativecommons.org/licenses/by-nc-nd/4.0/）.

1.Introduction

Rice is one of the main crops providing nutrition and trace elements to humans.In the face of the rapidly increasing world population，improving production efficiency is an important measure for increasing rice yield.However，many abiotic and biotic stresses limit crop yield capacity. For example，iron and zinc，acting as cofactors for manyenzymes［1］，are two trace elements essential for humans，plants and animals，but are toxic to most plants and animals at higher concentrations［2-3］.Ferrous iron（Fe2+）and zinc（Zn2+）toxicity harm rice production，especially in acidic soil in southeast Asia，West Africa，and Brazil［4-6］. Ingeneral，yieldlossesassociatedwithirontoxicity commonly range from 15%to 30%，but complete crop failure can occur in response to severe toxicity at early growth stages［7］.Besides iron toxicity，acidic soil is often associated with zinc toxicity［8］.This problem further damages the health of rice plants，increasing rice yield loss［9］.In fact，iron and zinc toxicities may occur in normal fields at low soil pH when harmful organic acids and hydrogen sulfide accumulate，a problem frequently found in the rice sowing and transplanting periods in China.Breeding rice varieties with iron and zinc toxicity tolerance is the most effective and economic means of minimizing yield loss resulting from iron and zinc toxicity stresses.

Tolerances to iron and zinc toxicities in rice are genetically complex traits，and there is large genotypic variation in the primary rice gene pool.Previous studies screening many irontolerantgenotypes［4，10-16］havesuggestedthatenvironmental conditions，timing and level of iron stress，the screening system，and other factors play crucial roles in determining genotype responses to iron toxicity［16］.For this reason，specialized varieties matching individual environments and varieties adapted to a wide range of iron toxicity environments should be bred.For rice zinc tolerance，few screening experiments have been performed.Asominori and Lemont are two relatively tolerant cultivars［9，15］，and TY-167 is a zinc-tolerant genotype［17］.

Genetic of tolerances to iron and zinc toxicities in rice appears to be quantitatively inherited.Many QTL for iron toxicity tolerance［7，10-11，12，15，16，18，19］and several QTL for zinc toxicity tolerance［9，15］have been identified using biparental populations or germplasm resources.Among them，no QTL has been fine-mapped except for tLBS5 for iron tolerance on chromosome 1［20］.Many iron and zinc transporter proteins in rice mediating metal ion uptake，distribution and homeostasis have been cloned，and some play important roles in tolerance to iron or zinc toxicity.For instance，OZT1 confers rice tolerance to Zn and Cd ion stress［21］，and OsFRO1 enhances tolerance to Fe toxicity of rice in the vegetative stage［22］.However，their value for breeding is uncertain.

Advanced-backcross QTL analysis is a method combining QTL analysis with cultivar development［23］.Following this strategy，our research groups have reported several successful applications of a large-scale backcross（BC）breeding strategy to improve abiotic stress tolerance in rice［24-27］，and to identify QTL and mine favorable alleles for complex traits［28-30］.In the present study，QTL tolerances to iron and zinc toxicities were identified using two sets of reciprocal introgression lines（ILs）derived from Minghui 63×02，428 and high-density SNP bin markers.The QTL and the elite lines with iron tolerance and/or zinc tolerance will provide essential information and materials for developing new rice cultivars with iron and/or zinc tolerance by marker-assisted selection（MAS）.

2.Materials and methods

2.1.Development of two sets of ILs population

Minghui 63（MH63），the male parent of the widely adapted hybrid indica variety Shanyou 63，whose distribution covers more than 21°of longitude and 20°of latitude in China［31］，was crossed with 02，428，a typical japonica with tolerance to low CO2concentration stress selected from mutant progenies derived from a cross between two landraces，Pang-Xie-Gu and Ji-Bang-Dao［32］.The F1plants were simultaneously backcrossed to the two parents to develop two BC1F1populations，each of around 80 plants.The BC1F1plants were then used as male parents in backcrosses with the corresponding parents to produce two BC2F1populations.The BC progenies were selfed successively for seven generations with no selection from BC2F1to BC2F8. Ultimately，two sets of reciprocal ILs，consisting of 198 BC2F8introgression lines（ILs）in the02，428 background（designated as 02，428-ILs）and226BC2F8ILsintheMH63background（MH63-ILs），were developed after removal of lines with heading dates too late for QTL detection.In addition，262F8recombinant inbred lines（RILs），derived from the same parents，were developed by single-seed descent from the F2population and used to construct a genetic map.

2.2.Phenotyping iron and zinc toxicity tolerances

Evaluations of iron and zinc toxicity tolerances at the seedling stage were performed in turn in the greenhouse in Agricultural College of Yangtze University in year 2014.The method and workflow described by Zhang et al.［15］were applied with slight changes.Plump seeds of each IL and the parents were selected and placed in an oven at 50°C for 3 days to break dormancy.Seeds were sterilized with 1%sodium hypochlorite solution.After washing twice with distilled water，all seeds were placed into an incubator at 37°C for 24 h.Germinating seeds were selected and placed in holes in a thin Styrofoam board（10 holes in one row were taken as a replication for one line）with a nylon net bottom，which floated on water in a plastic box.Each line had six replications and each replication had 10 holes with two germinating seeds per hole.After 7 days，one seedling per hole was kept so that all seedlings of a line would experience similar growth conditions.Two replicates selected randomly were then defined as controls and transferredtostandardYoshida'sculturesolution［33］，tworeplicates were transferred to the same solution supplemented with 300 mg L-1Fe2+（added as FeSO4·7H2O），and the remaining two replicates were transferred to the same solution supplemented with 200 mg L-1Zn2+（added as ZnSO4）.The temperature in the greenhouse was set to 30°C/25°C（day/night）and the relative humidity was maintained at 50-70%.The solution in plastic boxes was changed every 5 days and the pH was maintained at 4.5.When differences between controls and treatments were clearly observed（about 15 days after treatment），therootdryweight（RDW）andshootdryweight（SDW）of treatment and control plants were measured.The derived trait，total dry weight（TDW），was calculated as the sum of RDW and SDW.The indexes of toxicity tolerance，relative root dry weight（RRDW），relative shoot dry weight（RSDW）and relative total dryweight（RTDW）were calculated according to the following formula:

Relative value（%）

=（trait value in treatment）/（trait value in control）×100:

2.3.Genotyping and map construction

IncomparisonwithreciprocalILs，whichareskewedtowardsone parent or the other in genome due to successive backcrossing with the recurrent parent，RILs，which are a random mixture of MH63 and 02，428 backgrounds，are more suitable for map construction.Genomic DNA of MH63，02，428，the two sets of ILs，and the RILs was isolated using a DNeasy mini Kit（Qiagen）and the genotypes of the RILs were determined based on SNPs generated by whole-genome sequencing with the Illumina Genome Analyzer IIx as described previously［34］.

MH63and02，428weresubjectedtowhole-genome resequencing and a total of 5，336，108，154 and 5，562，905，674 nucleotides of data were obtained.Alignment was performed against the MSU6.1 assembly of the Nipponbare sequence as the reference genome.In total，5，062，106，567 and 5，278，080，725 nucleotides of consistent sequence were obtained for MH63 and 02，428，covering respectively 96.57%and 94.03%of the whole genome.Single-nucleotide polymorphisms（SNPs）between these two sequences were identified.Using evidence from more than 3，4，or 5 reads，respectively 48，498，42，124，and 36，410 SNPs were found between MH63 and 02，428.

A total of 384 SNPs evenlydistributed along the genome were chosen forthedesign ofanIllumina SNPchip［35］forgenotyping the two sets of ILs and RILs using their parents and F1as control. Theframeworkmapwasconstructedbasedongenotypicdataof theRILpopulationusingQTLIciMapping［36］.Furthermapfilling was performed by restriction association site DNA （RAD）sequencing［37］for the two sets of ILs as well as the two parents. Finally，58，936 qualified SNPs were identified and integrated into the framework map，with an average distance of 77 kb between adjacent SNPs.A bin was defined as coverage of series of SNPs with same genotype along a chromosome，resulting in a total of 4568 chromosome bins for 12 chromosomes.

2.4.Data analysis

The phenotypic value of each line in a test environment was taken as the average of two replicates.One-way ANOVA in SAS 9.3（SAS Institute Inc.Cary，NC）was applied to analyze the variances in target traits in the two populations.Before phenotypic data analysis and QTL mapping，the extreme values for each trait were removed from the dataset.

QTL mapping was conducted using the inclusive interval mapping（ICIM）function in the IciMapping 4.0 software［38］. In this function，marker selection is first conducted via stepwise regression considering all SNP marker information simultaneously.Then the phenotypic valuesare adjusted by all markers retained in the regression equation，omitting the two markers flanking the current mapping interval.The adjusted phenotypic values are then used for one-dimensional scanning.Based on experience，the LOD threshold was set at 2.5 for claiming a putative QTL.

3.Results

3.1.Linkage map and characteristics of introgression in the reciprocal ILs

The linkage map was constructed using the RIL population derived from MH63×02，428.The map spanned 1496.3 cM with a mean distance of 0.33 cM between adjacent markers.The inheritance ofparentalsegmentsacrossthegenomesofILswas characterized using high-density informative bin markers.The ILs showed marked variations in introgressed segments from donor parents，and almost all ILs contained more recurrent parent genome than donor parent genome.On average，the introgresseddonor-genomeproportion in02，428-ILs was21.9%，with a range from 1.64%to 88.07%，whereas the introgressed donor-genomeproportioninMH63-ILswas7.98%，rangingfrom 0.02%to 87.06%.The reciprocal sets of ILs were well separated，without overlap，with respect to the frequency distribution of the MH63 genome（Fig.1）.

Fig.1-The frequency distribution of MH63 genome in two sets of ILs derived from a cross between MH63 and 02，428.

3.2.Performance of toxicity tolerance of the parents and their ILs under iron and zinc stress conditions

Iron and zinc stress tests were administered to same populations and parents in the greenhouse.Under normal conditions，the RDW，SDW，and TDW of MH63 were all higher than those of 02，428 in both tests，and differences in RDW and TDW between parentswerehighlysignificant（P≤0.01）andsignificant（P≤0.05），respectively（Table 1）.These results implied that MH63 showed greater growth than 02，428 at seedling stage under normal cultivation conditions.Except that MH63 showed markedly higher RDW under iron stress and SDW under zinc stress，the two parents showed no significant differences in TDW，RDW，and SDW under the two stress conditions. However，the RRDW and RTDW of 02，428 were significantly higherthanthoseofMH63underbothconditionsandtheRSDW of 02，428 was significantly higher than that of MH63 under the iron stress condition（Table 1，F(xiàn)ig.2），suggesting that 02，428 had significantly stronger tolerance to iron and zinc toxicities than MH63.

TransgressivesegregationswereobservedforRRDW，RSDW，and RTDW of iron and zinc stresses to normalconditions in the two IL populations（Fig.3）.MH63-ILs had average RRDW，RSDW，and RTDW of 71.67%，67.21%，and 70.24%withrangesof29.35-94.97%，15.66-92.16%，and 18.10-91.58%，respectively，under iron stress，and average RRDW，RSDW and RTDW of 54.83%，59.66%，and 58.88%with ranges of 19.32-81.40%，24.07-86.19%，and 23.49-85.86%，respectively，under zinc stress（Table 2）.Similarly，02，428-ILs had average RRDW，RSDW，and RTDW of 75.12%，72.86%，and 74.07%with ranges of 36.05-95.00%，23.09-94.16%，and 25.56-94.81%，respectively，under iron stress，and average RRDW，RSDW and RTDW of 74.10%，68.66%，and 69.65%with ranges of 24.74-94.62%，34.82-94.63%，and 38.24-93.76%，respectively，under zinc stress（Table 2）.As shown in Fig.2，variations of relative RDW，SDW，and TDW in the two sets of ILs all showed normal distribution or skewed but continuous distribution under the two stress conditions，indicating that these two sets of IL populations had abundant diversity in iron and zinc toxicity tolerances and suggesting that some lines carried QTL for iron and/or zinc toxicity tolerances，in view of their tolerant phenotypes.

Table 1-Performance of iron and zinc toxicity tolerance-related traits between two parents under iron and zinc stress conditions.

Fig.2-Performance of the two parents at 15 days after treatment with iron and zinc stresses.

Fig.3-Frequency distributions of relative RDW，SDW，and TDW of iron and zinc stresses to normal conditions in two sets of ILs derived from MH63×02，428.Arrows point to the averages of the recurrent parents.

3.3.Detection of QTL for iron toxicity tolerance

FourandfiveQTLunderlying RRDWandRSDWwere identified under iron stress in the MH63-IL and 02，428-IL populations，respectively（Table 3，F(xiàn)ig.4）.Specifically，three QTL（qFRRDW1-1，qFRRDW1-2，and qFRRDW2）for RRDW and one QTL（qFRSDW11）for RSDW were identified on chromosomes 1，2，and 11 in MH63-ILs with favorable alleles all from the donor parent，02，428.Four QTL（qFRRDW2，qFRRDW3，qFRRDW9-1，andqFRRDW9-2）forRRDWandoneQTL（qFRSDW11）for RSDW were identified on chromosomes 2，3， 9，and 11 in 02，428-ILs.The favorable alleles at qFRRDW2 and qFRSDW11 were from 02，428，whereas those at the other three QTL were from MH63.

Among the above QTL，two QTL，qFRRDW2 on chromosome 2 and qFRSDW11 on chromosome 11 were simultaneously detected in both genetic backgrounds，suggesting that expressions of both QTL are independent of genetic background.The average phenotypic variances explained（PVE）of qFRRDW-2 and qFRSDW-11 were 12.20%and 11.66%，respectively，indicating that they were major QTL for tolerance to iron toxicity.

Table　2-Performance　for　iron　and　zinc　toxicity tolerance-related traits in two sets of ILs under iron and zinc stress conditions.

3.4.Detection of QTL for zinc toxicity tolerance

Four and six QTL for the three traits，RRDW，RSDW，and RTDW，were identified under zinc stress in MH63-and 02，428-IL populations，respectively（Table 4，F(xiàn)ig.4）.Two QTL（qZRRDW3 and qZRRDW11）for RRDW，one（qZRSDW2）for RSDW and one（qZRTDW11）for RTDW were mapped on chromosomes 2，3，and 11 in MH63-ILs.The 02，428 alleles at all QTL increased trait values.The qZRRDW11 for RRDW waslocatedintheregionoverlappingwiththatof qZRTDW11 for RTDW，suggesting the two QTL may be allelic.Two QTL（qZRRDW3 and qZRRDW7）for RRDW，two（qZRSDW11-1andqZRSDW11-2）forRSDWandtwo（qZRTDW3 and qZRTDW11）for RTDW were identified on chromosomes 3，7，and 11 in 02，428-ILs.The 02，428 alleles at all QTL except qZRTDW3 increased trait values.The qZRSDW11-2 for RSDW may be allelic to the qZRTDW11 forRTDW，in view of their location in the adjacent regions sharing the same marker C11S115.

Among the above QTL，qZRRDW3 for RRDW was simultaneously detected on chromosome 3 in both backgrounds，suggesting that expression of the QTL is independent of genetic background.

Table 3-QTL for iron toxicity tolerance identified in MH63-IL and 02，428-IL populations.

3.5.Genetic relationship between iron and zinc toxicity tolerances

The QTL mapping results（Tables 3 and 4；Fig.4）indicated that qFRSDW11 for RSDW was simultaneously detected under both iron and zinc stress conditions in 02，428-ILs，suggesting a genetic overlap between the two stress tolerances.Additionally，the region of C11S49-C11S60 on chromosome 11 harbored qFRSDW11，qZRSDW11，qZRRDW11，and qZRTDW11 affecting the iron and zinc toxicity tolerance-related traits RSDW，RRDW，and RTDW and detected in both backgrounds，hinting that the region contains genetically overlapping loci for iron and zinc toxicity tolerances.

4.Discussion

4.1.Detection of QTL for iron and zinc toxicity tolerances

Using RRDW，RSDW，and their derived trait RTDWas indexes of iron and zinc toxicity tolerance as recommended by Wu et al.［16］，nine and ten QTL contributing to iron and zinc toxicity tolerances，respectively，were identified in the two sets of reciprocal IL populations in this study（Tables 2 and 3）. Comparison of QTL identified in this study with previously reported iron toxicity tolerance QTL on the japonica Kato GRAMENE annotation sequence map 2009［39］，revealed that some were located in the same regions as QTL previously reported，or in adjacent regions.For instance，qFRRDW1-1 with flanking markers C1S110 and C1S124 on chromosome 1，which affected RRDW in MH63-ILs，was mapped in the same region as a QTL for RRDW under iron stress［11］，and partially overlapped with QRdw1 affecting root dry weight under iron and zinc stresses［15］.qFRRDW2，located in the region C2S139-C2S143 on chromosome 2，which affected RRDW under iron stress in both MH63-and 02，428-ILs，was mapped in a region adjacent to qRRL2-2 for relative root length under iron stress［40］.QTL regions for the iron and zinc toxicity tolerances mentioned above that were identified in different mapping populations and diverse environments could be beneficial for MAS breeding of iron and zinc toxicity-tolerant cultivars.It is noteworthy that QTL in the region C11S55-C11S59 on chromosome 11 that affected multiple iron and zinc toxicity tolerance-related traits in both MH63-and 02，428-ILs，was an important QTL with a large additive effect and genetic background independence.It merits confirmation in other populations and fine-mapping for map-based cloning.

4.2.Effect of genetic background on detection of QTL for iron and zinc toxicity tolerances

In this study，the reciprocal ILs were skewed towards one parent or the other in genome due to successive backcrossing with the recurrent parent and showed relatively uniform genetic background，thus ensuring that QTL mapping of stress tolerance was not strongly affected by genetic“noise”from cosegregating，non-target traits such as heading date and plant size.Accordingly，background effect on QTL detection can be revealed by comparison of mapping results from the two reciprocal IL populations.Of the 19 QTL affecting iron or zinc toxicity tolerance identified in the reciprocal backgrounds，only two（10.5%）were simultaneously identified in both backgrounds，clearly suggesting that most QTL detected in one background were not identified in another，so that there were genetic background effects on QTL detection for iron and zinc toxicity tolerance.This finding suggests that special care should be taken when QTL mapping informationis applied to breeding for iron and zinc toxicity tolerance using MAS，as genetic backgrounds may differ greatly between mapping and breeding populations.It is essential that QTL mapping be combined with MAS-based breeding in the same population，a practice that has been strongly recommended for complex quantitative traits［23，26］.

Fig.4-Distribution of QTL affecting iron and zinc toxicity tolerances identified in MH63-and 02，428-IL populations on a linkage

Table 4-QTL for zinc toxicity tolerance identified in MH63-IL and 02，428-IL populations.

4.3.Genetic overlap between iron and zinc toxicity

Genetic overlap has been previously found in various fields，such as human disease［41］and plant stress tolerance［29，42-43］.Some chromosome regions harbor QTL for tolerance to more than one metal ion stress［9，18，44-45］.In a previous study in our laboratory，Zhang et al.［15］detected two zinc toxicity tolerance-related QTL，QSdw2a and QSdw5，which were mapped together with a iron toxicity tolerance QTL.In the present study，a 777 kb region flanked by C11S49-C11S60 on chromosome 11 contained one iron toxicity tolerance QTL（qFRSDW11）and three zinc toxicity tolerance QTL（qZRRDW11，qZRTDW11，and qZRSDW11-1.qZRRDW3）associated with RRDW was mapped together with two QTL for tolerance to iron toxicity in a previous study［10，46］.These results show that there is at least partial genetic overlap between iron and zinc stress tolerances.Recently，functional and comparative genomic studies have revealed many multifunctional metal transporters.For instance，OZT1 confers plant tolerance to Zn and Cd ions［21］，and OsHMA3 not only reduces the toxicity of Ca2+to rice seedling but also maintains Zn2+balance in the rice stem［47-48］.These reports provide additional evidence of genetic overlap among different metal toxicity stresses.The molecular mechanisms underlying different stress tolerances await deeper investigation.

4.4.Potential application in breeding of iron and zinc toxicity tolerances

During long domestication and artificial selection，some favorable alleles have been intentionally or inadvertently introgressed into modern varieties from wild rice or landrace［49］.02，428，which was selected from a cross between two landraces，showed tolerance not only to low CO2concentration stress［32］but also to iron and zinc toxicities，as demonstratedinthisstudy.TwoQTL（qFRRDW2and qFRSDW11）fortolerance toirontoxicityandtwo QTL（qZRRDW3 and qZRSDW11）contributing to zinc toxicity tolerance were simultaneously identified in the two genetic backgrounds，and their favorable alleles all came from 02，428.This finding shows that favorable genes for iron and zinc toxicity tolerance“hidden”in 02，428 could be used by introgression and further pyramiding in elite modern cultivar backgrounds using molecular marker technologies.

Acidic soils with iron toxicity are always associated with zinc toxicity［8］.Thus，genetically overlapping loci provide a strategy for development of a cultivar tolerant to both stress toxicities that can not only simplify the process of MAS and reduce its cost but also improve the efficiency of development of a stress-tolerant variety.In this respect，the unique QTL，qFRSDW11，which affected iron and zinc toxicity tolerances in MH63-and 02，428-ILs，could be applied in breeding excellent rice varieties with tolerances to iron and zinc toxicities.

Acknowledgments

This research was funded by the National High Technology ResearchandDevelopmentProgramofChina（No. 2014AA10A601），Hubei Collaborative Innovation Center for Grain Industry（No.LXT-16-01）to JLX，the Shenzhen Peacock Plan（No.20130415095710361）to ZKL，the CAAS Innovative Team Award to JLX's team，and Key Discipline of Crop Science of Yangtze University to HL.The authors thank the students of Professor Danying Xing's group for their kind assistance in experiments for measuring tolerance to iron and zinc stresses.

R E F E R E N C E S

［1］L.E.Williams，R.F.Mills，P1B-ATPases-an ancient family of transition metal pumps with diverse functions in plants，Trends Plant Sci.10（2005）491-502.

［2］Z.L.He，X.E.Yang，P.J.Stoffella，Trace elements in agroecosystems and impacts on the environment，J.Trace Elem.Med.Biol.19（2005）125-140.

［3］P.J.White，P.H.Brown，Plant nutrition for sustainable development and global health，Ann.Bot.105（2010）1073-1080.

［4］A.Audebert，M.Fofana，Rice yield gap due to iron toxicity in West Africa，J.Agron.Crop Sci.195（2009）66-76.

［5］M.Becker，F(xiàn).Asch，Iron toxicity in rice-conditions and management concepts，Plant Nutr.Soil Sci.168（2005）558-573.

［6］N.K.Fageria，N.A.Rabelo，Tolerance of rice cultivars to iron toxicity，J.Plant Nutr.10（2008）653-661.

［7］A.Audebert，K.L.Sahrawat，Mechanisms for iron toxicity tolerance in lowland rice，J.Plant Nutr.23（2000）1877-1885.

［8］H.Obata，Micro essential elements，in:T.Matsuo，K. Kumazawa，R.Ishii，K.Ishihara，H.Hirata（Eds.），Science of Rice Plant，Vol.2:Physiology，F(xiàn)ood and Agriculture Policy Research Center，Tokyo，Japan 1995，pp.402-419.

［9］Y.J.Dong，T.F.Ogawa，D.Z.Lin，H.J.Koh，H.Kamiunten，M. Matsuo，S.H.Cheng，Molecular mapping of quantitative trait loci for zinc toxicity tolerance in rice seedling（Oryza sativa L.），F(xiàn)ield Crops Res.95（2006）420-425.

［10］J.L.Wan，H.Q.Zhai，J.M.Wan，H.Yasui，A.Yoshimura，Mapping QTL for traits associated with resistance to ferrous iron toxicity in rice（Oryza sativa L.），using japonica chromosome segment substitution lines，Acta Genet.Sin.30（2003）893-898.

［11］J.L.Wan，H.Q.Zhai，J.M.Wan，H.Ikehashi，Detection and analysis of QTLs for ferrous iron toxicity tolerance in rice，Oryza sativa L.Euphytica 131（2003）201-206.

［12］A.Shimizu，C.Q.Guerta，G.B.Gregorio，S.Kawasaki，H. Ikehashi，QTLs for nutritional contents of rice seedlings（Oryza sativa L.）in solution cultures and its implication to tolerance to iron-toxicity，Plant Soil 275（2005）57-66.

［13］I.Dufey，P.Hakizimana，X.Draye，S.Lutts，P.Bertin，QTL mapping for biomass and physiological parameters linked to resistance mechanisms to ferrous iron toxicity in rice，Euphytica 167（2009）143-160.

［14］G.Onaga，J.Egdane，R.Edema，I.Abdelbagi，Morphological and genetic diversity analysis of rice accessions（Oryza sativa L.）differing in iron toxicity tolerance，J.Crop Sci.Biotechnol. 16（2013）53-62.

［15］J.Zhang，A.S.Aijaz，L.Chai，Y.R.Cui，X.Q.Wang，T.Q.Zheng，J.L.Xu，Z.K.Li，Mapping of QTL for iron and zinc toxicity tolerance at seedling stage using a set of reciprocal introgression lines of rice，Acta Agron.Sin.39（2013）1754-1765（in Chinese with English abstract）.

［16］L.B.Wu，M.Y.Shhadi，G.Gregorio，E.Matthus，M.Becker，M. Frei，Genetic and physiological analysis of tolerance to acute iron toxicity in rice，Rice 7（2014）8.

［17］A.Song，P.Li，Z.J.Li，F(xiàn).L.Fan，M.Nikolic，Y.C.Liang，The alleviation of zinc toxicity by silicon is related to zinc transport and antioxidative reactions in rice，Plant Soil 344（2011）319-333.

［18］P.Wu，A.Luo，J.Zhu，J.Yang，N.Huang，D.Senadhira，Molecular markers linked to genes underlying seedling tolerance for ferrous iron toxicity，Plant Soil 196（1997）317-320.

［19］I.Dufey，X.Draye，S.Lutts，M.Lorieux，C.Martinez，P.Bertin，Novel QTLs in an interspecific backcross Oryza sativa×Oryza glaberrima for resistance to iron toxicity in rice，Euphytica 204（2015）609-625.

［20］E.Matthus，L.B.Wu，Y.Ueda，S.H?ller，M.Becker，M.Frei，Loci，genes，and mechanisms associated with tolerance to ferrous iron toxicity in rice（Oryza sativa L.），Theor.Appl.Genet.128（2015）2085-2098.

［21］H.X.Lan，Z.F.Wang，Q.H.Wang，M.M.Wang，Y.M.Bao，J. Huang，H.S.Zhang，Characterization of a vacuolar zinc transporter OZT1 in rice（Oryza sativa L.），Mol.Biol.Rep.40（2013）1201-1210.

［22］S.Ruengphayak，V.Ruanjaichon，C.Saensuk，S.Phromphan，S.Tragoonrung，R.Kongkachuichai，A.Vanavichit，F(xiàn)orward screening for seedling tolerance to Fe toxicity reveals a polymorphic mutation in ferric chelate reductase in rice，Rice 8（2015）3.

［23］S.D.Tanksley，J.C.Nelson，Advanced backcross QTL analysis: a method for the simultaneous discovery and transfer of valuable QTLs from unadapted germplasm into elite breeding lines，Theor.Appl.Genet.92（1996）191-203.

［24］J.Ali，J.L.Xu，A.M.Ismail，B.Y.Fu，C.H.M.Vijayakumar，Y.M. Gao，J.Domingo，R.Maghirang，S.B.Yu，G.Gregorio，S. Yanaghihara，M.Cohen，B.Carmen，D.Mackill，Z.K.Li，Hidden diversity for abiotic and biotic stress tolerances in the primary gene pool of rice revealed by a large backcross breeding program，F(xiàn)ield Crops Res.97（2006）66-76.

［25］Y.X.He，T.Q.Zheng，X.B.Hao，L.F.Wang，Y.M.Gao，Z.T.Hua，H.Q.Zhai，J.L.Xu，Z.J.Xu，L.H.Zhu，Z.K.Li，Yield performances of japonica introgression lines selected for drought tolerance in a BC breeding programme，Plant Breed. 129（2010）167-175.

［26］Z.K.Li，B.Y.Fu，Y.M.Gao，J.L.Xu，J.Ali，H.R.Lafitte，Y.Z.Jiang，J.D.Rey，C.H.M.Vijayakumar，R.Maghirang，T.Q.Zheng，L.H. Zhu，Genome-wide introgression lines and their use in genetic and molecular dissection of complex phenotypes in rice（Oryza sativa L.），Plant Mol.Biol.59（2005）33-52.

［27］F.Zhang，H.Q.Zhai，A.H.Paterson，J.L.Xu，Y.M.Gao，T.Q. Zheng，R.L.Wu，B.Y.Fu，J.Ali，Z.K.Li，Dissecting genetic networks underlying complex phenotypes:the theoretical framework，PLoS One 6（2011），e14541.

［28］J.L.Xu，H.R.Lafitte，Y.M.Gao，B.Y.Fu，R.Torres，Z.K.Li，QTLs for drought escape and tolerance identified in a set of random introgression lines of rice，Theor.Appl.Genet.111（2005）1642-1650.

［29］Y.Wang，J.P.Zang，Y.Sun，J.Ali，J.L.Xu，Z.K.Li，Identification of genetic overlaps for salt and drought tolerance using simple sequence repeat markers on an advanced backcross population in rice，Crop Sci.52（2012）1767-1775.

［30］Y.J.Zhu，K.Chen，X.F.Mi，T.X.Chen，J.Ali，G.Y.Ye，J.L.Xu，Z.K. Li，Identification and fine mapping of a stably expressed QTL for cold tolerance at the booting stage using an interconnected breeding population in rice，PLoS One 10（2015），e0145704.

［31］F.X.Wu，Q.H.Cai，Y.S.Zhu，J.F.Zhang，H.A.Xie，Application of indica restorer line，Minghui 63，for rice hybridization，F(xiàn)ujianJ.Agric.Sci.26（2011）1101-1112（in Chinese with English abstract）.

［32］H.B.Li，A wide-compatibility rice germplasm-02428，Crop Variety Resour.（4）（1989）42（in Chinese）.

［33］S.Yoshida，D.A.Forno，J.H.Cock，K.A.Gomez，Laboratory Manual for Physiological Studies of Rice，third ed. International Rice Research Institute，Manila，Philippines，1976.

［34］X.H.Huang，Q.Feng，Q.Qian，Q.Zhao，L.Wang，A.H.Wang，J.P. Guan，D.L.Fan，Q.J.Weng，T.Huang，G.J.Dong，T.Sang，B. Han，High-throughput genotyping by whole-genome resequencing，Genome Res.19（2009）1068-1076.

［35］W.Chen，H.D.Chen，T.Q.Zheng，R.B.Yu，W.B.Terzaghi，Z.K. Li，X.W.Deng，J.L.Xu，H.He，Highly efficient genotyping of rice biparental populations by GoldenGate assays based on parental resequencing，Theor.Appl.Genet.127（2014）297-307.

［36］L.Meng，H.H.Li，L.Y.Zhang，J.K.Wang，QTL IciMapping: integrated software for genetic linkage map construction and quantitative trait locus mapping in bi-parental populations，Crop J.3（2015）269-283.

［37］M.R.Miller，J.P.Dunham，A.Amores，W.A.Cresko，E.A. Johnson，Rapid and cost-effective polymorphism identification and genotyping using restriction site associated DNA（RAD）markers，Genome Res.17（2007）240-248.

［38］H.H.Li，G.Y.Ye，J.K.Wang，Modified algorithm for the improvement of composite interval mapping，Genetics 175（2007）361-374.

［39］Gramene，Gramene Database（Version No.33），Cold Spring Harbor Laboratory，and Cornell University，Ithaca，New York，2011.

［40］C.F.Zhao，L.H.Zhou，D.Ding，Y.D.Zhang，Q.Y.Zhao，X.Yu，Z. Zhu，T.Chen，S.Yao，C.L.Wang，Mapping of quantitative trait loci associated with ferrous iron toxicity tolerance at seedling stage based on chromosome segment substitution line in rice，Jiangsu J.Agric.Sci.29（2013）461-467（in Chinese with English abstract）.

［41］R.S.Desikan，A.J.Schork，Y.Wang，A.Witoelar，M.Sharma，L.K.McEvoy，D.Holland，J.B.Brewer，C.H.Chen，W.K. Thompson，D.Harold，J.Williams，M.J.Owen，M.C.O. Donovan，M.A.Pericak-Vance，R.Mayeux，J.L.Haines，L.A. Farrer，G.D.Schellenberg，P.Heutink，A.B.Singleton，A.Brice，N.W.Wood，J.Hardy，M.Martinez，S.H.Choi，A.DeStefano，M.A.Ilram，J.C.Bis，A.Smith，A.L.Fitzpatrick，L.launer，C.V. Duijn，S.Seshadri，I.D.Ulstein，D.Aarsland，T.Fladby，S. Djurovic，B.T.Hyman，J.Snaedal，H.Stefansson，K. Stefansson，T.Gasser，Q.A.Andreassen，A.M.Dale，IPDGC Investigators，Genetic overlap between Alzheimer's disease and Parkinson's disease at the MAPT locus，Mol.Psychiatr.20（2015）1588-1595.

［42］M.Trujillo，M.Troeger，R.E.Niks，K.H.Kogel，R.Huckelhoven，Mechanistic and genetic overlap of barley host and non-host resistance to Blumeria graminis，Mol.Plant Pathol.5（2004）389-396.

［43］L.Z.Xiong，Y.N.Yang，Disease resistance and abiotic stress tolerance in rice are inversely modulated by an abscisic acid-inducible mitogen-activated protein kinase，Plant Cell 15（2003）745-759.

［44］J.F.Ma，R.F.Shen，Z.Q.Zhao，M.Wissuwa，Y.Takeuchi，T. Ebitani，M.Yano，Response of rice to Al stress and identification of quantitative trait loci for Al tolerance，Plant Cell Physiol.43（2002）652-659.

［45］X.Y.Wang，P.Wu，Y.R.Wu，X.L.Yan，Molecular marker analysis of manganese toxicity tolerance in rice under greenhouse conditions，Plant Soil 238（2002）227-233.

［46］J.L.Wan，H.Q.Zhai，J.M.Wan，H.Yasui，A.Yoshimura，Detection and analysis of QTLs associated with resistance to ferrous iron toxicity tolerance in rice（Oryza sativa L.），using recombinant inbred lines，Acta Agron.Sin.30（2004）329-333（in Chinese with English abstract）.

［47］H.Miyadate，S.Adachi，A.Hiraizumi，K.Tezuka，N.Nakazawa，T.Kawamoto，K.Katou，I.Kodama，K.Sakurai，H.Takahashi，N.Satoh-Nagasawa，A.Watanabe，T.Fujimura，H.Akagi，OsHMA3，a P1B-type of ATPase affects root-to-shoot cadmium translocation in rice by mediating efflux into vacuoles，New Phytol.189（2011）190-199.

［48］A.Sasaki，N.Yamaji，J.F.Ma，Overexpression of OsHMA3 enhances Cd tolerance and expression of Zn transporter genes in rice，J.Exp.Bot.65（2014）6013-6021.

［49］M.J.Kovach，M.T.Sweeney，S.R.McCouch，New insights into the history of rice domestication，Trends Genet.23（2007）578-587.

22 February 2016

*Corresponding author.

**Corresponding author at:Institute of Crop Science/National Key Facility for Crop Gene Resources and Genetic Improvement，Chinese Academy of Agricultural Sciences，Beijing 100081，China.

E-mail addresses:xujlcaas@126.com（J.Xu）xingdy_2006@126.com（D.Xing）

Peer review under responsibility of Crop Science Society of China and Institute of Crop Science，CAAS.1These authors contributed equally to this work.

http://dx.doi.org/10.1016/j.cj.2016.05.007

2214-5141/?2016 Crop Science Society of China and Institute of Crop Science，CAAS.Production and hosting by Elsevier B.V.This is an open access article under the CC BY-NC-ND license（http://creativecommons.org/licenses/by-nc-nd/4.0/）.