Kenji Fukung*,Noko IzukTkehiro HchikenStoshi Mizuguchi Hidemi ItoKtsuyuki Ichitni

aFaculty of Life and Environmental Sciences,Prefectural University of Hiroshima,562 Nanatsuka-cho,Shobara,Hiroshima,Japan,727-0023

bFaculty of Agriculture,Kagoshima University,1-21-24 Korimoto,Kagoshima,Japan,890-8580

A nucleotide substitution at the 5′splice site of intron 1 of rice HEADING DATE 1(HD1)gene homolog in foxtail millet,broadly found in landraces from Europe and Asia

Kenji Fukunagaa,*,Naoko Izukaa,Takehiro Hachikena,Satoshi Mizuguchia, Hidemi Itoa,Katsuyuki Ichitanib

aFaculty of Life and Environmental Sciences,Prefectural University of Hiroshima,562 Nanatsuka-cho,Shobara,Hiroshima,Japan,727-0023

bFaculty of Agriculture,Kagoshima University,1-21-24 Korimoto,Kagoshima,Japan,890-8580

A R T I C L E I N F O

Article history:

Accepted 6 August 2015

Available online 15 August 2015

Foxtail millet Geographicaldistribution HD1(HEADING DATE 1)homolog Setaria italica Splice site

We investigated genetic variation of a rice HEADING DATE 1(HD1)homolog in foxtail millet. First,we searched for a rice HD1 homolog in a foxtail millet genome sequence and designed primers to amplify the entire coding sequence of the gene.We compared full HD1 gene sequences of 11 accessions(including Yugu 1,a Chinese cultivar used for genome sequencing)from various regions in Europe and Asia,found a nucleotide substitution at a putative splice site of intron 1,and designated the accessions with the nucleotide substitution as carrying a splicing variant.We verified by RT-PCR that this single nucleotide substitution causes aberrant splicing of intron 1.We investigated the geographical distribution of the splicing variant in 480 accessions of foxtail millet from various regions of Europe and Asia and part of Africa by dCAPS and found that the splicing variant is broadly distributed in Europe and Asia.Differences of heading times between accessions with wild type allele of the HD1 gene and those with the splicing variant allele were unclear.We also investigated variation in 13 accessions of ssp.viridis,the wild ancestor,and the results suggested that the wild type is predominant in the wild ancestor.

©2015 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

Foxtail millet[Setaria italica(L.)P.Beauv.]is one of the oldest cereals in the Old World.This millet adapts to various environmental conditions such as temperate and tropical climate and high-and low-altitude conditions,and its agronomic traits show large variation as a result of adaptation to local environments and cultivation under various cultural conditions.Given that foxtail millet has some advantages for genetic studies,such as diploidy with small chromosome numbers(2n=2x=18),small genome size(ca.500 Mb),inbreeding habit,and a relatively short growth period,it hasbecome a modelplantfor panicoid grass species such as biofuel grasses(switchgrass and Napier grass)and other millet species such as pearlmillet[1].The foxtailmillet genome sequence has recently been determined[2,3].Owing to its high variation in several agronomic traits as a result of adaptation to variable environmental condition and human selection,this millet will be also a good material for studying crop evolution.

Heading time is one ofthe mostimportanttraits in adaptation to localenvironments.This trait has already been investigated in foxtail millet;landraces show high variability in heading time, and this trait is determined by a combination of length of the basic vegetative growth period and sensitivity to short-day conditions[4,5].A recent phylogenetic analysis has shown that heading time is associated with phylogenetic differentiation of foxtailmillet landraces[6].This trait is known to be variable also in other plant species and has been investigated in detail[7-9]. Recently,molecular mechanisms of this trait have been studied in severalplant species,particularly in model plants such as rice and Arabidopsis[10].Among the most important genes for flowering/heading are CONSTANS(CO)in Arabidopsis[11]and its ortholog HEADINGDATE 1(HD1)in rice[12],which are associated with sensitivity to photoperiod and regulate expression of FT in Arabidopsis and Hd3a in rice.In severalplant species,homologs of this gene have also been isolated and analyzed[13-17],revealing that the gene is important in flowering of all plant species. Although this gene(andalso other genes involvedin heading)has been investigated in detail in the context of domestication and adaptation to local conditions in rice[18,19],only a few genetic studies[20,21]of heading time of foxtail millet,which has been cultivated more broadly than rice since ancient times,have been performed.Ichitani et al.performed genetic analyses of heading time using progenies derived from a single cross and suggested that heading characteristics are controlled by polygenes[20]. Mauro-Herrera etal.performed QTL analyses for flowering time and genetic analyses of candidate genes and reported that HD1/ CO colocalized with one of the QTLs on chromosome 4[21]. More recently,Jia et al.performed genome-wide association mapping and suggested some candidate genes for this trait[6]. However,no detailed work on specific genes involved in heading time has yet been performed.

In the present study,we performed sequencing analyses for the foxtail millet HD1(CO)gene homolog,identified a splicing variant of the gene,and investigated the geographical distribution of the variant.We also investigated the relationships between HD1 genotype and heading time.

2.Materials and methods

2.1.Plant materials and DNA extraction

We used a total of 480 foxtail millet accessions and 13 accessions of S.italica ssp.viridis,the wild ancestor of foxtail millet,as shown in Table S1.Most of the foxtail millet accessions were landraces directly collected in fields,and some were obtained frominstitutes in other countries.These samples cover the entire traditional cultivation area of this millet fromeast Asia to western Europe and part of Africa.All of the foxtail millet samples are maintained at the National Institute of Agrobiological Sciences(NIAS)Genebank,Tsukuba, Japan.All of the ssp.viridis accessions were obtained from the United States Department of Agriculture(USDA)Genebank.The 13 accessions included three from Turkey,two each from Russia,Afghanistan,Iran and mainland China,and one each from Chile and Mongolia.As foxtail millet and ssp.viridis are predominantly self-pollinating species,a single plant for each accession was chosen for the analyses.DNAwas extracted from seedling leaves according to Murray and Thompson[22]with some modifications or using a Qiagen DNaeasy Plant Mini Kit according to the manufacturer's instructions.

2.2.Search for an HD1 gene homolog in foxtail millet,PCR amplification,and comparison of the sequences

A search was performed for a rice HEADING DATE 1(HD1) homolog in the foxtail millet genome database(http://www. plantdg.org/SiGDB/)by BLAST(http://www.plantgdb.org/SiGDB/ cgi-bin/blastGDB.pl,blastn with E-value=1e-20)using a rice HD1 gene sequence,AB041837,as a query.We designed a primer pair to amplify the entire coding sequence of the gene and primers for sequencing(Figs.1,and S1,and Table S2). Primers were designed with Primer3(http://primer3.wi.mit. edu/).We amplified 10 accessions of foxtail millet from various geographical areas of Europe and Asia(two from Japan(NIAS JP71626 and 71640),two from Taiwan of China(JP 73913,222588),one from the Philippines(JP 222569),one from Myanmar(JP 222570),two from India(JP 222981 and 222982), one from ex-Czechoslovakia(JP 222971),and one from France (JP 223004))and determined the sequences by direct sequencing using the primers listed in Table S2.Sequences of 10 landraces and one Chinese accession,Yugu 1,used for genome sequencing by Bennetzen et al.[2]were aligned with CLUSTALW (http://www.genome.jp/tools/clustalw/).Structures of the gene (positions ofexons and intron)were deduced by comparison with those in other plant species including rice,maize(EU302135), wheat(AB094487),and barley(AF490468).

2.3.mRNA extraction,RT-PCR,and comparison between wild type and variant

Sequence analysis indicated that some landraces carry a point mutation on the putative 5′splice site of intron 1resulting in loss of function of this gene(see results).To verify that this point mutation causes aberrant splicing,we performed RT-PCR of the gene including intron 1 in a variant type(AT type)of a landrace from Taiwan of China (JP 73913,see Table S1)and also a wild type(GTtype)from Japan (JP 71640).RNA was extracted from young leaves grown in naturalday conditions according to Wang et al.[23],and cDNA synthesis and RT-PCR were performed with a Takara Prime Script RT-PCR Kit according to manufacturer's instructions.The primer pair ConsNewF1(5′-CAGCAAGGATCCTGACAACA-3′) and ConsNewR2(5′-CTTGATCCT TGGTC GTGCTT-3′)was used for amplification(Figs.1,and S1,Table S2).PCRconditions were 5 min at 94°C;35 cycles of 1 min at 94°C,1 min at 56°C,and 2 min at 72°C;and 2 min at 72°C.After purification with a Wizard SV Gel and PCR Clean-Up System(Promega Co.Ltd.), PCR products were ligated into the pGEM vector(Promega Co. Ltd.)and transformed into JM108 Escherichia coli cells.Sequencing was performed with a BigDye terminator kit v3.1 accordingto the instructions of the manufacturer(Applied Biosystems) and M13 primers.Two clones for the wild type and seven clones for the splicing variant were sequenced.Samples were sequenced on an ABIPRISM 3130xl Genetic Analyzer/ABIPRISM 3500xl Genetic Analyzer at the Institute of Gene Research, Kagoshima University.

Fig.1–Structure of the HD1 gene of foxtail millet,primers used for amplification/sequencing,and indels and SNPs found between accessions.Black boxes indicate exons.Horizontal arrows indicate primers and directions(Table S2).Verticalarrows indicate indels and SNPs.Accessions(Yugu 1,mainland China;JP 71626&71640,Japan;JP 222588&73913,Taiwan of China;JP 222569,the Philippines;JP 222981&222982,India;JP 222570,Myanmar;JP 222971,ex-Czechoslovakia;JP 223004,France)with indels and SNPs are indicated in parentheses(See details in Fig.S1).

2.4.dCAPS analysis of landraces and assessment of geographicaldistribution

After verifying that one nucleotide sequence substitution that occurs at the 5′splice site of intron 1 leads to loss of function of the gene by comparison with the sequence of the ortholog of rice and other grass species and sequencing of RT-PCR products(see results),we analyzed this point mutation in 480 accessions of foxtail millet from various cultivation areas.To detect this single-nucleotide polymorphism(SNP),we developed a dCAPS marker[24]for the SNP.dCAPS primers were designed with dCAPS Finder version 2.0(http://helix.wustl. edu/dcaps/dcaps.html).Primer sequences were HD1dCAPSF (5′-AGCAGTACCGAAGGCA AGAA-3′)and HD1dCAPSR(5′-AGCA TAGTAATAGATGAACGCA-3′)(Figs.1,S1,and S2 and Table S2). PCR was performed with TOYOBO 2×Quick Taq and PCR conditions were 5 min at 94°C;35 cycles of 1 min at 94°C, 1 min at 50°C,and 2 min at 72°C;and 5 min at 72°C.After digestion with Sph I,electrophoresis was performed on 2.0% agarose gel.Approximately 220-bp bands were scored as the wild-type allele and 200-bp bands as the splicing variant allele.

2.5.Cultivation experiments

Six individuals of each of97 accessions were cultivated in pots in a greenhouse in the Faculty of Life and Environmental Sciences, Prefectural University of Hiroshima,Shobara,Hiroshima(34°50′N,132°59′E).These 97 accessions cover a broad cultivation area from East Asia to Europe and fromhigh to low latitude(Table S3). Of the 97,16 carried the wild-type allele and 81 the splicing variant allele.Seeds were sown in pots on May 22,2011 under natural day-length conditions.Days to heading was defined as the number of days until the panicle on the main culm was first visible in the sheath of the flag leaf.

3.Results

3.1.Sequence of HD1 homolog in foxtail millet and sequence variation between landraces

A homolog of rice HEADING DATE 1(HD1)was found on chromosome 4 of the foxtail millet genome sequence[21]. The collinearity of chromosome 4 of foxtail millet with rice chromosome 6[25],on which the rice HD1 gene is located[12], suggested that the gene on foxtail millet chromosome 4 is an ortholog of the rice HD1 gene.We also aligned the HD1 gene in the foxtailmillet genome with predicted Setaria italica zinc finger protein HD1-like(LOC101768789)(XM_004965163)as a query and confirmed that the gene on chromosome 4 is a foxtail millet HD1 homolog.We also found a gene showing much lower identity with the rice HD1 gene on foxtailmillet chromosome 1, but it may be a paralog of the gene.

We determined 2835-2837 bp of the gene of 10 landraces from various regions of Eurasia(DDBJ:AB807720-AB807729) and compared the full gene sequences from Yugu 1 and the 10 landraces to identify polymorphisms in the gene.The most important polymorphism between accessions was a single-nucleotide substitution of GT by AT at the 5′end of intron 1(splice site)found in 7 of the 11 accessions,which may cause loss of function of the gene as shown in Fig.1. Other polymorphisms were also found.Two specific SNPs were found in the 5′region of the gene and in intron 1 in Yugu 1 and a Japanese landrace(JP 71640)(Fig.1).A landrace from Myanmar(JP 222570)had several nucleotide substitutions innon-coding sequences and also a substitution in the coding region leading to an amino acid substitution(Fig.1).This substitution may change the function of the product,but in this study,we focused on the SNP(single nucleotide polymorphism)in intron 1,which apparently causes loss of function of the gene.We designated the accessions with AT at the 5′end of intron 1 as“splicing variant”.

3.2.Verification of splicing variation by RT-PCR

We determined the sequences of two clones ofa wild type and seven clones of a splicing variant,and the results of sequencing are summarized in Fig.2.Expected splicing of intron 1 was observed in both of the two clones of the wild type(JP 71640),whereas three different aberrant splicing products were found in the variant(a Taiwanese accession,JP 73913),designated as types 1-3 as shown in Fig.2.Of the seven clones,four showed a type 1 splicing pattern,two showed a type 2 pattern,and one showed a type 3 pattern(Fig.2).In all three types,intron 1 was spliced at GT in exon 1(33 bp upstream of the 5′splice site of the wild type)instead of the end of intron 1 and also at a different AG instead of the 3′end of intron 1 in types 2-3(68 and 25 bp,respectively,upstream of the 3′splice site of the wild type)(Figs.2 and S1).

3.3.Geographicaldistribution of the splicing variant in foxtail millet and variation in the wild ancestor revealed by dCAPS analysis

Genotyping of wild type and splicing variant was successfully performed by dCAPS analysis as shown in Fig.3.We investigated the geographical distribution of the splicing variant allele in 480 accessions offoxtailmillet fromvarious areas of Europe and Asia and a part of Africa(Fig.4).We also confirmed that results of dCAPS are consistent with G-A substitution by sequencing the region including the 5′end of intron 1 of 8 randomly chosen accessions(data not shown).Surprisingly,of the 480 accessions, only 90 showed the wild-type allele(18.8%)and 390 showed the splicing variant allele(81.2%).The splicing variant allele is not evenly distributedin Europe and Asia:31(27.6%)ofthe accessions from Japan,11(26.8%)from Korea Peninsula,12(14.6%)from mainland China,one(11.1%)from the Philippines,2(28.6%)from Indonesia,1(50%)from Thailand,3(10%)from Myanmar,20 (45.5%)from India,5(83.3%)from Sri Lanka,2(40%)from Turkey, and one each from Primorskaya province of the ex-USSR and Poland.The wild-type allele was observed most frequently in accessions from India and Sri Lanka of south Asia and at lower frequencies in accessions from Japan,Korea Peninsula,and mainland China of east Asia but rarely in accessions from the Nansei Islands of Japan,Taiwan of China,the Philippines,Nepal, Central Asia,West Asia,Europe and Africa(Fig.3).

As for ssp.viridis,2 accessions from Afghanistan had a variant allele,whereas the 11 other accessions from mainland China,Mongolia,Russia,Iran,Turkey and Chile,had a wild-type allele,suggesting that the wild type is predominant in ssp.viridis.

3.4.Relationships between HD1 genotype and heading time

A histogram of the distribution of heading time of accessions with the wild-type allele and those with the splicing variant allele is shown in Fig.5.One very late-heading accession from Halmahera Island,Indonesia(169 days),had the wild-type allele,and a few late heading accessions from Thailand, Myanmar,and Luzon island in the Philippines also had the wild-type allele.Most of the accessions with the wild-type and variant alleles showed early to late heading(34-119 days)but the average heading times of accessions with the wild-type allele and those with the splicing variant allele were 85.8 and 65.7 days,respectively.Comparison(by t-test)between accessions with the wild-type allele and those with the splicing variant allele with all 97 accessions showed a significant difference between the two alleles(P＜0.01),but when the very late-heading accession from Halmahera Island wasexcluded,there was no significant difference between the two types.

Fig.2–Results of sequencing of RT-PCR products of the wild-type and variant alleles.Sequences in boxes indicate cDNA sequences amplified by RT-PCR and sequences between boxes indicate spliced sequences.Stars indicate a nucleotide substitution at the 5′splice site of the intron 1.The wild-type allele was spliced at expected GT–AG sequences,whereas the variant allele was spliced at different GT and AGs.Three different RT-PCR products were obtained and designated as types 1–3. All three types were spliced at the same GT(33 bp upstream of the 5′splice site of the wild type)in exon 1.Type 1 was spliced at the same AG as the wild type,and types 2 and 3 were spliced at different AGs(68 and 25 bp upstream of the 3′splice site of the wild type)in intron 1.

Fig.3–A gel image of the results of genotyping of wild-type and splicing variant type of HD1 gene in foxtail millet.M indicates a 100-bp ladder size marker.Number above photo indicates cultivation number of the samples(see Table S1) and“w”denotes wild type.The 220-and 200-bp bands correspond to wild type and splicing variant,respectively.

4.Discussion

4.1.Aberrant splicing at the 5′end of intron 1 of HD1 gene

We identified a nucleotide substitution at the 5′end of the intron 1 of the HD1 gene by sequencing analysis(Fig.1)and confirmed the occurrence of aberrant splicing by sequencing of the RT-PCR products with the splicing variant(Fig.2).At least three types of splicing(types 1-3)occurred in the splicing variant of the HD1 gene(Fig.2).A nucleotide substitution at the 5′end of intron 1 leads to multiple different aberrant splicing products in the waxy gene of rice[26].Judging from the putative amino acid sequences deduced from these nucleotide sequences,a protein of type 1 lacks 11 amino acids from the HD1 protein compared with the wild type,and a frame shift and a premature stop codon occur in types 2 and 3(Fig.S3).Given that a zinc finger motif is encoded in exon 1 of the HD1 gene,whereas a nuclear localization signal is encoded in exon 2 ofthe gene[12],this nucleotide substitution likely leads to loss of function of the gene in types 2 and 3. However,it is possible that type 1 is functional despite the deletion of 11 amino acids.Given that a simple G-to-A transition in a splicing donor site leads to mis-spliced mRNA with a premature stop codon in the rice waxy gene[27,28]and the pea flower color(bHLH)gene[29],the results of RT-PCR in the present study strongly suggest that mis-splicing occurs in intron 1 because of a single nucleotide substitution at the 5′end of intron 1.

4.2.Geographicaldistribution of the splicing variant in foxtail millet and variation in the wild ancestor

As shown in Fig.4,the splicing variant is predominant in Europe and Asia and the wild type is frequently found in limited regions:east Asia including Japan,Korea Peninsula, and mainland China,south Asia including India and Sri Lanka,and Southeast Asia,and also sporadically in the ex-USSR and Europe.Takei and Sakamoto[5]reported that foxtail millet landraces from intermediate latitudes(between 27°N and 34°N)are sensitive to change in day length,and landraces from an intermediate latitude would be expected to carry the wild-type allele.However,the wild-type allele of the HD1 gene is distributed broadly from high latitudes in Primorskaya province,ex-USSR,and northeast China to low latitudes in Luzon island,Philippines and Halmahera Island, Indonesia.Mauro-Herrera et al.[21]reported that one of three maize candidate genes in the Setaria QTL intervals is an HD1 (CO)homolog,suggesting that the HD1 gene influences foxtail millet heading time.

The results for the geographicaldistribution ofthe wild type of the HD1 gene in Asia,particularly in south and east Asia,are congruent with those of rDNA-RFLP analysis[30],suggesting that some genetic exchanges occurred between countries in east Asia such as Japan,Korea Peninsula,and mainland China and countries in south Asia such as India and Sri Lanka. However,another interpretation is also possible:that this point mutation arose and was selected at multiple times during the spread of this millet in Europe and Asia because of some advantage of the splicing variant in adaptation to local environments.Detailed phylogenetic studies of HD1 gene in foxtail millet will be helpful for determining how the mutant type arose and spread in Europe and Asia.

As for ssp.viridis,two accessions from Afghanistan had the variant allele,whereas the other 11 accessions had the wild-type allele,suggesting that the wild type is predominantin ssp.viridis.Interestingly,our observations in the cultivation experiment showed that these two accessions from Afghanistan had morphological characteristics similar to those of Afghan landraces of foxtail millet(data not shown), which have morphological characteristics similar to ssp.viridis except for large grains and a non-shattering habit[31].These accessions may have been misclassified or may have arisen through introgression between foxtail millet and its wild ancestor.In these cases,the variant allele may have originated after domestication.However,further study of ssp.viridis accessions from a broader geographicalarea is needed.

4.3.Relationships between HD1 genotype and heading time

As shown in Fig.5,differences in heading between the wild type and variant are unclear.Regulation of heading time may be so complex that the effect of a single gene on heading time is masked by the effects of other genes influencing the trait, although the HD1 gene is colocalized with a QTL for heading time[21]and seems a likely candidate for control of heading time variation in foxtail millet.Ehd1 confers short-day promotion offlowering in the absence of a functionalallele of Hd1[32] and several other genes are also involved in heading time in rice[9,33].Recently,in sorghum,another gene,pseudoresponse regulator protein 37(PRR37),involved in regulation of heading time has also been identified[15].This gene is also importantin adaptation to different latitudes.In maize,the HD1 gene has been mapped and QTL analysis for flowering time has been performed[34],but the results do not provide evidence of theeffect ofthe gene on flowering time photoperiod responses.It appears that HD1(CONSTANS)does not always play an important role in latitudinal adaptation in sorghum and maize.Jia et al.[6]identified several genes including NF-YC9 (also known as HAP5C)and FIE1 that influence heading time in foxtail millet by genome-wide association studies.Comparative genetic studies of control of flowering time among multiple grass species willhelp in an understanding of grass evolution.

Fig. 4 – Geographical distribution of the wild-type allele (indicated in black) and variant allele (indicated in white). “n” indicatesnumber of accessions investigated.

Fig.5–Distribution of number of days to heading of 97 accessions.Accessions with the wild-type allele are indicated in gray and those with the variant allele are indicated in black.One very late-heading accession from Halmahera Island,Indonesia is indicated by an arrow.

5.Conclusions

Foxtail millet is cultivated over a broad area in Europe and Asia and is highly variable in heading time.Although heading/flowering is genetically controlled in a complex manner,an HD1(CO)gene homolog was analyzed in the present study as the first step for elucidating the genetic basis of this trait.We found that a splicing variant(putative non-functional)allele of the gene is broadly distributed in Europe and Asia but found no clear association between genotype and heading time,suggesting that several genes influence the trait,as reported in rice,maize,and sorghum. Given that not only conserved but also lineage-specific genes influence latitudinal adaptation in cereal species,it is desirable to analyze variation in other candidate genes influencing this trait in foxtail millet and also to evaluate the effects of genes in segregating populations under different photoperiods.We are also now developing recombinant inbred lines (RILs)from a cross between a landrace from a temperate area and one from a subtropical area[35]for this purpose.

Acknowledgments

We thank J.Bennetzen and R.Percifield for providing information on foxtailmillet genome sequence before publication.This work was partially supported by the NIAS Genebank Project, NIAS,Japan.

Supplementary material

Supplementary material to this article can be found online at http://dx.doi.org/10.1016/j.cj.2015.07.003.

R E F E R E N C E S

[1]A.N.Doust,E.A.Kellogg,K.M.Devos,J.L.Bennetzen,Foxtail millet:a sequence-driven grass model system,Plant Physiol. 149(2009)137-141.

[2]J.L.Bennetzen,J.Schmutz,H.Wang,R.Percifield,J.Hawkins, A.C.Pontaroli,M.Estep,L.Feng,J.N.Vaughn,J.Grimwood,J. Jenkins,K.Barry,E.Lindquist,U.Hellsten,S.Deshpande,X. Wang,X.Wu,T.Mitros,J.Triplett,X.Yang,C.Y.Ye,M. Mauro-Herrera,L.Wang,P.Li,M.Sharma,R.Sharma,P.C. Ronald,O.Panaud,E.A.Kellogg,T.P.Brutnell,A.N.Doust,G.A. Tuskan,D.Rokhsar,K.M.Devos,Reference genome sequence of the model plant Setaria,Nat.Biotechnol.30(2012)555-561.

[3]G.Zhang,X.Liu,Z.Quan,S.Cheng,X.Xu,S.Pan,M.Xie,P. Zeng,Z.Yue,W.Wang,Y.Tao,C.Bian,C.Han,Q.Xia,X.Peng, R.Cao,X.Yang,D.Zhan,J.Hu,Y.Zhang,H.Li,H.Li,N.Li,J. Wang,C.Wang,R.Wang,T.Guo,Y.Cai,C.Liu,H.Xiang,Q. Shi,P.Huang,Q.Chen,Y.Li,J.Wang,Z.Zhao,J.Wang, Genome sequence of foxtailmillet(Setaria italica)provides insights into grass evolution and biofuel potential,Nat. Biotechnol.30(2012)549-554.

[4]E.Takei,S.Sakamoto,Geographical variation of heading response to daylength in foxtailmillet(Setaria italica P. Beauv.),Jpn.J.Breed.37(1987)150-158.

[5]E.Takei,S.Sakamoto,Further analysis of geographical variation of heading response to daylength in foxtail millet (Setaria italica P.Beauv.),Jpn.J.Breed.39(1989)285-298.

[6]G.Jia,X.Huang,H.Zhi,Y.Zhao,Q.Zhao,W.Li,Y.Chai,L. Yang,K.Liu,H.Lu,C.Zhu,Y.Lu,C.Zhou,D.Fan,Q.Weng,Y. Guo,T.Huang,L.Zhang,T.Lu,Q.Feng,H.Hao,H.Liu,P.Lu,N.Zhang,Y.Li,E.Guo,B.Zhang,W.Li,Y.Wang,H.Li,B.Zhao,J. Li,X.Diao,B.Han,A haplotype map of genomic variations and genome-wide association studies of agronomic traits in foxtail millet(Setaria italica),Nat.Genet.45(2013)957-961.

[7]K.Ichitani,Y.Okumoto,T.Tanisaka,Photoperiod sensitivity of Se-1 locus found in photoperiod insensitivity rice cultivars of the northern limit region of rice cultivation,Breed.Sci.47 (1997)145-152.

[8]M.Koornneef,C.Alonso-Blanco,D.Vreugdenhil,Naturally occurring genetic variation in Arabidopsis thaliana,Annu.Rev. Plant Biol.55(2004)141-172.

[9]M.Yano,Genetic controlof flowering time in rice,a short-day plant,Plant Physiol.127(2000)1425-1429.

[10]T.Izawa,Y.Takahashi,M.Yano,Comparative biology comes into bloom:genomic and genetic comparison of flowering pathways in rice and Arabidopsis,Curr.Opin.Plant Biol.6 (2003)113-120.

[11]J.Putterill,F.Robson,K.Lee,R.Simon,G.Coupland,The CONSTANS gene of Arabidopsis promotes flowering and encodes a protein showing similarities to zinc finger transcription factors,Cell 80(1995)847-857.

[12]M.Yano,Y.Katayose,M.Ashikari,U.Yamanouchi,L.Monna, T.Fuse,T.Baba,K.Yamamoto,Y.Umehara,Y.Nagamura,T. Sasaki,Hd1,a major photoperiod sensitivity quantitative trait locus in rice,is closely related to the Arabidopsis flowering time gene CONSTANS,Plant Cell12(2000)2473-2484.

[13]J.Clotault,A.C.Thuillet,M.Buiron,S.De Mita,M.Couderc, B.I.G.Haussmann,C.Mariac,Y.Vigouroux,Evolutionary history of pearl millet(Pennisetum glaucum[L.]R.Br.)and selection on flowering genes since its domestication,Mol. Biol.Evol.29(2012)1199-1212.

[14]T.A.Miller,E.H.Muslin,J.E.Dorweiler,Amaize CONSTANS-like gene,conz1,exhibits distinct diurnalexpression patterns in varied photoperiods,Planta 227(2008)1377-1388.

[15]R.L.Murphy,R.R.Klein,D.T.Morishige,J.A.Brad,W.L.Rooney, F.R.Miller,D.V.Dugas,P.E.Klein,J.E.Mullet,Coincident light and clock regulation of pseudoresponse regulator protein 37 (PRR37)controls photoperiodic flowering in sorghum,Proc. Natl.Acad.Sci.U.S.A.108(2011)16469-16474.

[16]C.Navarro,J.A.Abelenda,E.Cruz-Oró,C.A.Cuéllar,S. Tamaki,J.Silva,K.Shimamoto,S.Prat,Control of flowering and storage organ formation in potato by FLOWERING LOCUS T,Nature 478(2011)119-122.

[17]F.Valverde,CONSTANS and the evolutionary origin of photoperiodic timing of flowering,J.Exp.Bot.62(2011) 2453-2463.

[18]Y.Takahashi,K.Shimamoto,Heading date 1(Hd1),an ortholog of Arabidopsis CONSTANS,is a possible target of human selection during domestication to diversify flowering times of cultivated rice,Genes Genet.Syst.86(2011)175-182.

[19]Y.Takahashi,K.M.Teshima,S.Yokoi,H.Innan,K. Shimamoto,Variations in Hd1 proteins,Hd3a promoters,and Ehd1 expression levels contribute to diversity of flowering time in cultivated rice,Proc.Natl.Acad.Sci.U.S.A.106(2009) 4555-4560.

[20]K.Ichitani,K.Nagao,Y.Narita,K.Fujikawa,M.Samejima,S. Taura,M.Sato,Genetic analysis of heading characters in foxtail millet(Setaria italica(L.)P.Beauv.)using the progeny from the cross between the two diverse strains,Gai 53 and Kuromochi,Mem.Fac.Agr.Kagoshima Univ.38(2003)17-25.

[21]M.Mauro-Herrera,X.W.Wang,H.Barbier,T.P.Brutnell,K.M. Devos,A.N.Doust,Genetic controland comparative genomic analysis of flowering time in Setaria(Poaceae),G3-Genes Genom,Genet.3(2013)283-295.

[22]M.G.Murray,W.F.Thompson,Rapid isolation of high-molecular-weight plant DNA,Nucleic Acids Res.8(1980) 4321-4325.

[23]Z.Y.Wang,F.Q.Zheng,G.Z.Shen,J.P.Gao,D.P.Snustad,M.G. Li,J.L.Zhang,M.M.Hong,The amylose content in rice endosperm is related to the post-transcriptional regulation of the waxy gene,Plant J.7(1995)613-622.

[24]M.M.Neff,E.Turk,M.Kalishman,Web-based primer design for single nucleotide polymorphism analysis,Trends Genet. 18(2002)613-615.

[25]K.M.Devos,Z.M.Wang,J.Beales,T.Sasaki,M.D.Gale, Comparative genetic maps of foxtailmillet(Setaria italica)and rice(Oryza sativa),Theor.Appl.Genet.96(1998)63-68.

[26]X.L.Cai,Z.Y.Wang,Y.Y.Xing,J.L.Zhang,M.M.Hong,Aberrant splicing of intron 1 leads to the heterogeneous 5′UTR and decreased expression of waxy gene in rice cultivars of intermediate amylose content,Plant J.14(1998)459-465.

[27]H.Y.Hirano,M.Eiguchi,Y.Sano,A single base change altered the regulation of the Waxy gene at the posttranscriptional levelduring the domestication of rice,Mol.Biol.Evol.15 (1998)978-987.

[28]M.Isshiki,K.Morino,M.Nakajima,R.J.Okagaki,S.R.Wessler, T.Izawa,K.Shimamoto,A naturally occurring functional allele of the rice waxy locus has a GT to TT mutation at the 5′splice site of the first intron,Plant J.15(1998)133-138.

[29]R.P.Hellens,C.Moreau,K.Lin-Wang,K.E.Schwinn,S.J. Thomson,M.W.E.J.Fiers,T.J.Frew,S.R.Murray,J.M.I.Hofer, J.M.E.Jacobs,K.M.Davies,A.C.Allan,A.Bendahmane,C.J. Coyne,G.M.Timmerman-Vaughan,T.H.N.Ellis, Identification of Mendel's white flower character,PLoS ONE 5 (2010),e13230http://dx.doi.org/10.1371/journal.pone.0013230.

[30]M.Eda,A.Izumitani,K.Ichitani,M.Kawase,K.Fukunaga, Geographicalvariation of foxtail millet,Setaria italica(L.)P. Beauv.based on rDNAPCR-RFLP,Genet.Resour.Crop.Evol.60 (2013)265-274.

[31]S.Sakamoto,Origin and phylogenetic differentiation of cereals in the Southwest Eurasia,in:S.Sakamoto(Ed.), Domesticated Plants and Animals of the Southwest Eurasian Agro-pastroral Culture Complex,Kyoto University,Kyoto 1987,pp.1-45.

[32]K.Doi,T.Izawa,T.Fuse,U.Yamanouchi,T.Kubo,Z. Shimatani,M.Yano,A.Yoshimura,Ehd1,a B-type response regulator in rice,confers short-day promotion of flowering and controls FT-like gene expression independently of Hd1, Gene Dev.18(2004)926-936.

[33]K.Ebana,T.Shibaya,J.Wu,K.Matsubara,H.Kanamori,H. Yamane,U.Yamanouchi,T.Mizubayashi,I.Kono,A. Shomura,S.Ito,T.Ando,K.Hori,T.Matsumoto,M.Yano, Uncovering of major genetic factors generating naturally occurring variation in heading date among Asian rice cultivars,Theor.Appl.Genet.122(2011)1199-1210.

[34]N.D.Coles,M.D.McMullen,P.J.Balint-Kurti,R.C.Pratt,J.B. Holland,Genetic control of photoperiod sensitivity in maize revealed by joint multiple population analysis,Genetics 184 (2010)799-812.

[35]K.Sato,Y.Mukainari,K.Naito,K.Fukunaga,Construction ofa foxtailmillet linkage map and mapping spikelet-tipped bristles 1(stb1)with transposon display markers and simple sequence repeat markers with genome sequence information, Mol.Breed.31(2013)675-684.

3 April 2015

in revised form 4 July 2015

.Tel./fax:+824 74 1714,+824 74 0191.

E-mail address:fukunaga@pu-hiroshima.ac.jp(K.Fukunaga).

Peer review under responsibility of Crop Science Society of China and Institute of Crop Science,CAAS.