Ruili Pei,Jingying Zhang,Ling Tian,Shengrui Zhang,Fenxia Han,Shurong Yan,Lianzheng Wang,Bin Li*,Junming Sun*

The National Engineering Laboratory for Crop Molecular Breeding,MOA Key Laboratory of Soybean Biology(Beijing),Institute of Crop Sciences,Chinese Academy of Agricultural Sciences,Beijing 100081,China

1.Introduction

Soybean(Glycine max L.Merrill)is one of the most important oilseed crops in the world.It provides the world's supply of vegetable protein and oil.Soybean also produces biologically active substances with potential benefit for human health,including isoflavones,soyasaponin,and lunasin[1–3].

Isoflavones belong to a group of secondary metabolites derived from the phenylpropanoid pathway,and are mainly produced in legumes.As precursors of major phytoalexin glyceollins,isoflavones play important roles in plant–microbe interaction[4,5].Isoflavones also function as signal molecules in soybean nodulation[6,7].Isoflavones have attracted increasing attention in recent years owing to their potential benefits for human health.As biologically active substances,isoflavones reduce the risk of menopausal symptoms,breast cancer,osteoporosis,dementia,and cardiovascular diseases[8–12].In view of their important roles,studies of the biosynthesis and accumulation of isoflavones in soybean seeds have been performed[13–16].The ultimate goal of these studies is to clarify the genetic basis of isoflavone accumulation and to develop soybean cultivars with desired isoflavone contents.Given that soybean isoflavone contents are typical quantitative traits influenced by both genetic and environmental factors,identification of stable QTL for isoflavone components across environments will facilitate understanding the genetic basis of isoflavone accumulation in soybean seeds.To date,273 QTL for isoflavones have been detected in soybean,including 61 for daidzein,68 for genistin,71 for glycitein,and 73 for total isoflavones,according to the Soybase database(https://www.soybase.org/)[17–29].However,the genetic network regulating isoflavone accumulation in soybean seed is still unclear.

We previously developed a recombinant inbred line(RIL)population including 200 lines from a cross between the soybean cultivars Luheidou 2(LHD2)and Nanhuizao(NHZ),and performed QTL mapping for isoflavones using 110 ofthese 200 lines [20].Recently,we genotyped the remaining 90 RILs using specific-locus amplified fragment sequencing(SLAF-seq),and combined the genotyping data with those of the original 110 lines to generate an integrated high-density genetic map comprising 3541 SLAF markers[30].In present study,this genetic map was used to identify novel,stable QTL for isoflavone components in the increased population,and the QTL were compared with those previously identified.

2.Materials and methods

2.1.Plant materials and field trials

A total of 200 lines of a RIL population(F5:7–8)derived from a cross between the cultivars LHD2 and NHZ were planted at Changping Experimental Station in 2009,and Shunyi Experimental Stations from 2009 to 2011.Field trials were performed using a randomized complete block design with three replicates.The rows of each plot were 2.0 m in length,with 0.5 m between adjacent rows and 0.1 m between adjacent plants[20].

2.2.Isoflavone extraction and determination in soybean seeds

Extraction and determination of isoflavones were performed following Li et al.[20].Twelve isoflavone standards were provided by Akio Kikuchi(National Agricultural Research Center for Tohoku Region,Japan):daidzin,glycitin,genistin,malonyldaidzin, ma1ony1g1ycitin, ma1ony1genistin,acetyldaidzin,acetylglycitin,acetylgenistin, daidzein,glycitein,and genistein.Identification and quantification of isoflavone components in soybean seeds were based on the retention times and peak areas of 12 standard isoflavone solutions using high-performance liquid chromatography(HPLC),and the precise isoflavone component contents in soybean seeds were calculated following Sun et al.[31].In soybean seeds,isoflavones consist of six major components:daidzin,glycitin,genistin,malonyldaidzin,malonylglycitin,and malonylgenistin[20].Accordingly,the total isoflavone contents were calculated as the sum the contents of these six major components.

2.3.SLAF-seq data genotyping and soybean genetic map construction

SLAF library construction and sequencing,SLAF-seq data grouping and genotyping,and the construction of a soybean genetic map using the 200 lines are described in detail in our previous report[30].The soybean high-density genetic map,comprising 3541 SLAF markers,was used to identify additive and epistatic QTL for isoflavones in the present study.

2.4.QTL mapping for isoflavone contents in soybean seeds

Additive QTL for six isoflavone components were detected using the inclusive composite interval mapping(ICIM)method in QTL IciMapping 4.0 software[32],and the P-value for entering variables(PIN)was set to 0.01.The threshold of the logarithm of odds(LOD)scores was determined using 1000 permutations at the significance level of 0.05.Since QTL for isoflavone contents were affected by environments,only QTL identified in multiple environments were designated as stable QTL and analyzed in the present study.Epistatic QTL were detected using the ICIM-EPI method with LOD threshold=5.0,PIN=0.05,and Step=5 cM.

2.5.Gene annotation in QTL intervals

The genomic sequences corresponding to additive QTL intervals were analyzed based on the genome sequences of cultivar Williams82(Wm82.a2.v1,https://phytozome.jgi.doe.gov/),andthesesequenceswerefurtherannotated against the NR(NCBI non-redundant protein sequences)(https://blast.ncbi.nlm.nih.gov/), KOG/COG (Clusters of Orthologous Groups of proteins)(http://www.ncbi.nlm.nih.gov/COG/),and Swiss-Prot(Manually annotated and reviewed protein sequences)(http://www.ebi.ac.uk/uniprot/)databases using BlastX program in NCBI(https://blast.ncbi.nlm.nih.gov/).

3.Results

3.1.Phenotypic analysis of isoflavone contents in soybean seeds

The six predominant isoflavone components(daidzin,genistin,glycitin,malonyldaidzin,malonylgenistin,and malonylglycitin)of soybean seeds in 200 lines of the RIL population were determined.As shown in Table 1,the contentsofthesix isoflavone componentsand total isoflavones exhibited broad ranges among the 200 lines across four environments,with coefficients of variation(CVs)varying from 0.16 to 0.50.Analysis of variance suggested that isoflavones are affected by both genetic and environmental factors. However, broad-sense heritability across allenvironmentsranged from 0.56 to 0.87 for the six isoflavone components and total isoflavones,suggesting that isoflavones are under mainly genetic control(Table 2).

Table 1–Characteristics of predominant isoflavones in the 200 soybean recombinant inbred lines.

As shown in Fig.1,the distributions of total isoflavone content were continuous and quantitative.Most isoflavone components exhibited normal distributions across the four environments,though those of some components were not normal in specific environments according to the Kolmogorov–Smirnov test(Table 1).Transgressive segregation was observed among the 200 lines(Fig.1),suggesting that both parents contribute to isoflavone content in soybean seeds.

Table 2–Analysis of variance of predominant isoflavones in 200 soybean recombinant inbred lines.

Fig.1–Frequency distributions of total seed isoflavone content in 200 lines in four environments.TIF_2009CP represents total seed isoflavone content in 200 lines at Changping Experimental Station in 2009.TIF_2009SY,TIF_2010SY,and TIF_2011SY represent total isoflavone contents of soybean seeds in 200 lines at Shunyi Experimental Station from 2009 to 2011.Arrows indicate total seed isoflavone contents in the two parent lines(cultivars LHD2 and NHZ).

3.2.QTL mapping for isoflavones in soybean seeds using high-density genetic map

A high-density genetic map comprising 3541 SLAF markers was used for QTL mapping.Based on 1000 permutations for six isoflavones and total isoflavone contents,a LOD score of 3.3 was selected as the threshold for declaring the presence of an additive QTL.Using this genetic map and the seed isoflavone contents in the 200 lines,24 stable QTL across environments were identified for isoflavone components.These QTL were mapped to 13 linkage groups(LG)(Fig.2).The phenotypic variation explained by individual QTL varied from 4.2%to 21.2%,with LOD scores ranging from 4.9 to 17.9(Table 3).The favorable alleles of 16 QTL were derived from LHD2,the parent with higher isoflavone content,whereas the favorable alleles of the remaining eight QTL were derived from NHZ(Table 3).Three genes involved in isoflavone biosynthesis(Table 3),and 13 genes encoding MYB-like transcription factors within the genomic regions corresponding to the 24 stable QTL were identified(data not shown).

Specifically,for daidzin,two stable QTL,qD16 and qD20 explained 7.4%and 8.0%of mean phenotypic variation across environments.The favorable allele of qD20 was derived from cultivar LHD2,while the favorable allele of qD16 was derived from cultivar NHZ.For genistin,three stable QTL(qG8,qG9,and qG20)explained 4.2%–9.4%of mean phenotypic variation.The favorable alleles of all three QTL were derived from LHD2.A 4-coumarate:CoA ligase gene Gm4CL(Glyma.09G211100.1),and an isoflavone reductase gene GmIFR(Glyma.09G211500.1)were found within the genomic region corresponding to qG9.For glycitin,the favorable allele of qGL5 was derived from LHD2,and it explained 9.6%of mean phenotypic variation.For malonyldaidzin,eight stable QTL(qMD2,qMD3,qMD7,qMD13,qMD15–1,qMD15–2,qMD19,andqMD20)were identified,explaining 6.3%–11.8%of mean phenotypic variation.The favorable alleles of qMD3,qMD7,qMD15–1,and qMD20 were derived from LHD2,and the favorable alleles of qMD2,qMD13,qMD15–2,qMD19 were derived from NHZ.For malonylgenistin,four QTL(qMG14,qMG16,qMG18,and qMG20)were detected,explaining 4.7%–21.2%of mean phenotypic variation.The favorable alleles of qMG14,qMG18,and qMG20 were derived from LHD2,and the favorable allele of qMG16 was derived from NHZ.A chalcone reductase gene GmCHR(Glyma.14G005700)was identified within the genomic region corresponding to qMG14.For total isoflavone content,six stable QTL(qTIF2,qTIF7,qTIF16,qTIF18,qTIF19,and qTIF20)were detected.The mean phenotypic variation explained by individual QTL varied from 6.0%to 14.3%.The favorable alleles of qTIF2,qTIF7,qTIF18,and qTIF20 were derived from LHD2,while the favorable alleles of qTIF16 and qTIF19 were derived from NHZ.Additionally,the loci on Gm16,Gm19,and Gm20 contributed to multiple isoflavone components in soybean seeds(Fig.2).

3.3.Epistatic effect on isoflavones in soybean seeds

Epistatic effects on isoflavone content were also analyzed.Nine epistatic QTL were identified for isoflavones in soybean seeds,explaining 4.7%–15.6%of phenotypic variation with LOD scores ranging from 5.2 to 7.3(Table 4).Notably,the epistatic effect between the 10 cM and 85 cM on Gm09 explained phenotypic variation for both malonyldaidzin and malonylgenistin in 2011SY(Table 4),suggesting a pleiotropic effect of the epistatic QTL.

Fig.2–Twenty-four stable QTL for isoflavone content in soybean seeds on 13 linkage groups.SLAF marker distributions are depicted on the groups based on their genetic positions in centiMorgans(cM).The 24 stable QTL for isoflavone content are shown between tightly linked SLAF markers on the right side of each linkage group.Three isoflavone biosynthesis genes are indicated within QTL intervals in red text.

4.Discussion

Previous studies suggested that both population size and marker density affect the accuracy and efficiency of QTL mapping[33,34].In our previous study,we suggested that increasing marker density could increase the efficiency and accuracy of QTL mapping for isoflavone content[20].In the present study,we found that more stable QTL(24 in contrast to 11)were identified for isoflavone components in soybean seeds with an increase of population size from 110 to 200,suggesting that increasing population size could improve detection efficiency for QTL mapping.Moreover,most of these(20 of the 24)corresponded to QTL found in previous studies(Table 3),suggesting the reliability of the QTL mapping in the present study.Most of favorable alleles(for 16 of the 24 QTL)were derived from cultivar LHD2,which contains a higher isoflavone content(3697 μg g?1)than cultivar NHZ(1816 μg g?1)[20].However,the eight favorable alleles derived from the parent with lower isoflavone content suggest that cultivar NHZ also harbors favorable alleles for isoflavones.

Table 3–Descriptions of 24 stable QTL for the predominant isoflavones in soybean seeds.

Table 4–Epistatic QTL for predominant isoflavones insoybean seeds.

Of the 24 stable QTL,four major stable QTL(qMD20,qMG14,qTIF2,and qTIF20)explained much phenotypic variation(＞10%)for isoflavones.Specifically,qMD20 and qTIF20 were mapped to the same locus on Gm20 across four environments,and these two loci explained 11.8%and 14.9%of phenotypic variation for malonyldaidzin and total isoflavone contents,respectively.The high stability and phenotypic variation explained by this locus suggest the presence of a major gene controlling isoflavone accumulation in soybean seeds.

Four novel stable QTL were identified by comparison of stable QTL with previous QTL for isoflavones.qG8 explained 7.69%of mean phenotypic variation for genistin in 2009SY and 2010SY.qMD19 explained 7.29%of mean phenotypic variation for malonyldaidzin in 2009SY,2010SY,and 2011SY.qMG18 explained 4.68% of mean phenotypic variation for malonylgenistin in 2009SY and 2010SY.qTIF19 explained 7.21%of phenotypic variation for total isoflavones in 2010SY and 2011SY.The favorable alleles of the first two loci were derived from cultivar LHD2 and those of the second two from cultivar NHZ.The identification of novel QTL will contribute to the understanding of the genetic basis of isoflavone accumulation and regulation in soybean seeds.

Gene annotation revealed three genes(Gm4CL,GmCHR,and GmIFR)encoding key enzymes involved in isoflavone biosynthesis.Gm4CL encodes a 4-coumarate:CoA ligase,which catalyzes the reaction of 4-coumarate and CoA to form 4-coumaroyl-CoA[35].GmCHR encodes a chalcone reductase.This enzyme catalyzes the transformation from 4-coumaroyl-CoA to isoliquiritigenin,which is the chalcone precursor of daidzin[36].In the present study,however,GmCHR was found within the genomic region corresponding to qMG14,which contributed mainly to malonylgenistin content.This finding may be explained by the close correlations between different isoflavones in soybean seeds.GmCHR might affect malonylgenistin content by regulating daidzin accumulation in soybean seeds.We also cannot exclude the possibility that gene or genes other than GmCHR within the genomic region corresponding to qMG14 conferred the major additive effect of this locus.IFR encodes an isoflavone reductase,which is a key enzyme involved in the synthesis of the phytoalexin glyceollin from daidzein [37].Itcatalyzes a NADPH-dependent reduction of 2′-hydroxyisoflavones to form 2′-hydroxyisoflavanones[38].Owing to their essential roles in isoflavone biosynthesis,these structural genes may contribute the major effects of the corresponding QTL for isoflavone content.

Thirteen MYB-like genes were found within the genomic regions corresponding to the 24 stable QTL.Some MYB transcription factors may affect isoflavone content by regulating the expression level of structural genes involved in isoflavone biosynthesis[15,16].Therefore,these MYB-like genes may suggest candidate genes for isoflavone accumulation in soybean seeds.

In summary,24 stable QTL were identified for isoflavone content using a high-density genetic map.Of these 24 QTL,20 have been reported previously,whereas four(qG8,qMD19,qMG18,and qTIF19)represent novel QTL for isoflavone components in soybean seeds.Three structural genes involved in isoflavone biosynthesis(Gm4CL,GmCHR,and GmIFR)and 13 MYB-like transcription factor genes were found associated with the 24 stable QTL and represent candidate genes regulating isoflavone contents in soybean seeds.The stable and novel QTL will facilitate understanding the genetic bases of isoflavone accumulation and regulation in soybean seeds,and the SLAF markers tightly linked to major QTL will be useful in marker-assisted selection for the improvement of soybean quality.

Acknowledgments

This work was supported by the National Key Technology R&D Program of China during the Twelfth Five-Year Plan Period of China(2014BAD11B01-x02),Beijing Science and Technology Project(Z16110000916005),National Science and Technology Major Project(2016ZX08004-003),National Key R&D Program of China (2016YFD0100504 and 2016YFD0100201),National Natural Science Foundation of China(31671716,31171576),and Agricultural Science and Technology Innovation Project of CAAS.

[1]X.Wan,H.Liu,Y.Sun,J.L.Zhang,X.B.Chen,N.Chen,Lunasin:a promising polypeptide for the prevention and treatment of cancer,Oncol.Lett.13(2017)3997–4001.

[2]Z.Tavassoli,M.Taghdir,B.Ranjbar,Renin inhibition by soyasaponin I:a potent native anti-hypertensive compound,J.Biomol.Struct.Dyn.36(2018)166–176.

[3]I.Mateos-Aparicio,A.Redondo Cuenca,M.J.Villanueva-Suarez,M.A.Zapata-Revilla,Soybean,a promising health source,Nutr.Hosp.23(2008)305–312.

[4]N.Shimada,S.Sato,T.Akashi,Y.Nakamura,S.Tabata,S.Ayabe,T.Aoki,Genome-wide analyses of the structural gene families involved in the legume-specific 5-deoxyisoflavonoid biosynthesis of Lotus japonicus,DNA Res.14(2007)25–36.

[5]V.V.Lozovaya,A.V.Lygin,O.V.Zernova,A.V.Ulanov,S.Li,G.L.Hartman,J.M.Widholm,Modification of phenolic metabolism in soybean hairy roots through down regulation of chalcone synthase or isoflavone synthase,Planta 225(2007)665–679.

[6]J.L.Ferrer,M.B.Austin,C.Stewart Jr.,J.P.Noel,Structure and function of enzymes involved in the biosynthesis of phenylpropanoids,Plant Physiol.Biochem.46(2008)356–370.

[7]E.De Rijke,L.Aardenburg,J.Van Dijk,F.Ariese,W.H.Ernst,C.Gooijer,U.A.Brinkman,Changed isoflavone levels in red clover(Trifolium pratense L.)leaves with disturbed root nodulation in response to waterlogging,J.Chem.Ecol.31(2005)1285–1298.

[8]F.H.Sarkar,Y.Li,Soy isoflavones and cancer prevention,Cancer Investig.21(2003)744–757.

[9]T.Cornwell,W.Cohick,I.Raskin,Dietary phytoestrogens and health,Phytochemistry 65(2004)995–1016.

[10]R.A.Dixon,Phytoestrogens,Annu.Rev.Plant Biol.55(2004)225–261.

[11]A.B.Moore,L.Castro,L.Yu,X.Zheng,X.Di,M.I.Sifre,G.E.Kissling,R.R.Newbold,C.D.Bortner,D.Dixon,Stimulatory and inhibitory effects of genistein on human uterine leiomyoma cell proliferation are influenced by the concentration,Hum.Reprod.22(2007)2623–2631.

[12]X.Di,L.Yu,A.B.Moore,L.Castro,X.Zheng,T.Hermon,D.Dixon,A low concentration of genistein induces estrogen receptor-alpha and insulin-like growth factor-I receptor interactions and proliferation in uterine leiomyoma cells,Hum.Reprod.23(2008)1873–1883.

[13]W.Jung,O.Yu,S.M.Lau,D.P.O'Keefe,J.Odell,G.Fader,B.McGonigle,Identification and expression of isoflavone synthase,the key enzyme for biosynthesis of isoflavones in legumes,Nat.Biotechnol.18(2000)208–212.

[14]S.Dhaubhadel,B.D.McGarvey,R.Williams,M.Gijzen,Isoflavonoid biosynthesis and accumulation in developing soybean seeds,Plant Mol.Biol.53(2003)733–743.

[15]S.S.Chu,J.Wang,Y.Zhu,S.L.Liu,X.Q.Zhou,H.R.Zhang,C.E.Wang,W.M.Yang,Z.X.Tian,H.Chen,D.Y.Yu,An R2R3-type MYB transcription factor,GmMYB29,regulates isoflavone biosynthesis in soybean,PLoS Genet.13(2017)e1006770.

[16]J.Yi,M.R.Derynck,X.Li,P.Telmer,F.Marsolais,S.Dhaubhadel,A single-repeat MYB transcription factor,GmMYB176,regulates CHS8 gene expression and affects isoflavonoid biosynthesis in soybean,Plant J.62(2010)1019–1034.

[17]J.J.Gutierrez-Gonzalez,T.D.Vuong,R.Zhong,O.Yu,J.D.Lee,G.Shannon,M.Ellersieck,H.T.Nguyen,D.A.Sleper,Major locus and other novel additive and epistatic loci involved in modulation of isoflavone concentration in soybean seeds,Theor.Appl.Genet.123(2011)1375–1385.

[18]J.J.Gutierrez-Gonzalez,X.Wu,J.Zhang,J.D.Lee,M.Ellersieck,J.G.Shannon,O.Yu,H.T.Nguyen,D.A.Sleper,Genetic control of soybean seed isoflavone content:importance of statistical model and epistasis in complex traits,Theor.Appl.Genet.119(2009)1069–1083.

[19]Y.Wang,Y.P.Han,X.Zhao,Y.G.Li,W.L.Teng,D.M.Li,Y.Zhan,W.B.Li,Mapping isoflavone QTL with main,epistatic and QTL×environment effects in recombinant inbred lines of soybean,PLoS One 10(2015),e0118447.

[20]B.Li,L.Tian,J.Y.Zhang,L.Huang,F.X.Han,S.R.Yan,L.Z.Wang,H.K.Zheng,J.M.Sun,Construction of a high-density genetic map based on large-scale markers developed by specific length amplified fragment sequencing(SLAF-seq)and its application to QTL analysis for isoflavone content in Glycine max,BMC Genomics 15(2014)1086.

[21]H.J.Zhang,J.W.Li,Y.J.Liu,W.Z.Jiang,X.L.Du,L.Li,X.W.Li,L.T.Su,Q.Y.Wang,Y.Wang,Quantitative trait loci analysis of individual and total isoflavone contents in soybean seeds,J.Genet.93(2014)331–338.

[22]G.L.Zeng,D.M.Li,Y.P.Han,W.L.Teng,J.Wang,L.J.Qiu,W.B.Li,Identification of QTL underlying isoflavone contents in soybean seeds among multiple environments,Theor.Appl.Genet.118(2009)1455–1463.

[23]J.J.Gutierrez-Gonzalez,X.Wu,J.D.Gillman,J.D.Lee,R.Zhong,O.Yu,G.Shannon,M.Ellersieck,H.T.Nguyen,D.A.Sleper,Intricate environment-modulated genetic networks control isoflavone accumulation in soybean seeds,BMC Plant Biol.10(2010)105.

[24]M.A.Kassem,K.Meksem,M.J.Iqbal,V.N.Njiti,W.J.Banz,T.A.Winters,A.Wood,D.A.Lightfoot,Definition of soybean genomic regions that control seed phytoestrogen amounts,J.Biomed.Biotechnol.1(2004)52–60.

[25]V.S.Primomo,V.Poysa,G.R.Ablett,C.J.Jackson,M.Gijzen,I.Rajcan,Mapping QTL for individual and total isoflavone content in soybean seeds,Crop Sci.45(2005)2454–2464.

[26]T.Yoshikawa,Y.Okumoto,D.Ogata,T.Sayama,M.Teraishi,T.M.T.Toda,K.Yamada,K.Yagasaki,N.Yamada,K.Yagasaki,N.Yamada,T.Tsukiyama,T.Yamadand,T.Tanisaka,Transgressive segregation of isoflavone contents under the control of four QTLs in a cross between distantly related soybean varieties,Breed.Sci.60(2010)243–254.

[27]C.J.Smallwood,C.N.Nyinyi,D.A.Kopsell,C.E.Sams,D.R.West,P.Chen,S.K.Kantartzi,P.B.Cregan,D.L.Hyten,V.R.Pantalone,Detection and confirmation of quantitative trait loci for soybean seed isoflavones,Crop Sci.54(2014)595–606.

[28]H.Z.Liang,Y.L.Yu,S.F.Wang,Y.Lian,T.F.Wang,Y.L.Wei,P.T.Gong,X.Y.Liu,F.X.J.,M.C.Zhang,QTL mapping of isoflavone,oil and protein contents in soybean(Glycine max L.Merr.),Agric.Sci.China 9(2010)1108–1116.

[29]Y.P.Han,W.L.Teng,Y.Wang,X.Zhao,L.Wu,D.M.Li,W.B.Li,Unconditional and conditional QTL underlying the genetic interrelationships between soybean seed isoflavone,and protein or oil contents,Plant Breed.134(2015)300–309.

[30]B.Li,S.X.Fan,F.K.Yu,Y.Chen,S.R.Zhang,F.X.Han,S.R.Yan,L.Z.Wang,J.M.Sun,High-resolution mapping of QTL for fatty acid composition in soybean using specific-locus amplified fragment sequencing,Theor.Appl.Genet.130(2017)1467–1479.

[31]J.M.Sun,B.L.Sun,F.X.Han,S.R.Yan,H.Yang,K.Akio,Rapid HPLC method for determination of 12 isoflavone components in soybean seeds,Agric.Sci.China 10(2011)70–77.

[32]H.H.Li,J.M.Ribaut,Z.L.Li,J.K.Wang,Inclusive composite interval mapping(ICIM)for digenic epistasis of quantitative traits in biparental populations,Theor.Appl.Genet.116(2008)243–260.

[33]H.H.Li,S.Hearne,M.Banziger,Z.L.Li,J.K.Wang,Statistical properties of QTL linkage mapping in biparental genetic populations,Heredity 105(2010)257–267.

[34]M.Stange,H.F.Utz,T.A.Schrag,A.E.Melchinger,T.Wurschum,High-density genotyping:an overkill for QTL mapping?Lessons learned from a case study in maize and simulations,Theor.Appl.Genet.126(2013)2563–2574.

[35]A.Uhlmann,J.Ebel,Molecular cloning and expression of 4-coumarate:coenzyme A ligase,an enzyme involved in the resistance response of soybean(Glycine max L.)against pathogen attack,Plant Physiol.102(1993)1147–1156.

[36]H.Du,Y.B.Huang,Y.X.Tang,Genetic and metabolic engineering of isoflavonoid biosynthesis,Appl.Microbiol.Biotechnol.86(2010)1293–1312.

[37]T.L.Graham,J.E.Kim,M.Y.Graham,Role of constitutive isoflavone conjugates in the accumulation of glyceollin in soybean infected with Phytophthora megasperma,Mol.Plant Microbe 3(1990)157–166.

[38]J.D.Cooper,F.Qiu,N.L.Paiva,Biotransformation of an exogenously supplied isoflavonoid by transgenic tobacco cells expressing alfalfa isoflavone reductase,Plant Cell Rep.20(2002)876–884.