Lei Zhng,Ho Zhng,Linyi Qio,Lingfeng Mio,Dong Yn,Pn Liu,Gungyo Zho,Jizeng Ji,*,Lifeng Go,*

a Institute of Crop Sciences,Chinese Academy of Agricultural Sciences,Beijing 100081,China

b Department of Biology,Taiyuan Normal University,Taiyuan 030031,Shanxi,China

c College of Agronomy,Shanxi Agricultural University,Taiyuan 030031,Shanxi,China

Keywords:Agronomic traits Haplotypes Triticum aestivum

ABSTRACT The MADS-box gene plays an important role in regulating plant growth and development.In this study,a SEP3-like MADS-box gene TaSEP3-1 was isolated from bread wheat.The expression patterns of the three homoeologs TaSEP3-A1, TaSEP3-B1,and TaSEP3-D1 were similar,and higher expression levels were detected in floral organs and developing kernels.TaSEP3-D1 was located in the nucleus and cytoplasm and possessed transactivation activity in yeast.Homoeolog sequence polymorphism analysis identified four,three,and four haplotypes of TaSEP3-A1,TaSEP3-B1,and TaSEP3-D1,respectively,and the haplotypes of TaSEP3-D1 had larger effects on agronomic traits than those of TaSEP3-A1 and TaSEP3-B1. D1_h4,significantly associated with heading date,plant height,and other yield-related traits,was the favored haplotype of TaSEP3-D1.Transgenic wheat genotypes overexpressing TaSEP3-D1 exhibited delayed heading and reduced plant height,indicating a role in regulating heading date and plant development.These results shed light on the role of TaSEP3-D1 in wheat plant development.The favored haplotype of TaSEP3-D1 can be applied in breeding to improve plant architecture and yield in wheat.

1.Introduction

MADS-box transcription factors constitute one of the largest transcription factor families in eukaryotes [1].All MADS-box proteins encode a highly conserved DNA-binding MADS domain consisting of approximately 60 amino acids [2].MADS-domain proteins can be divided into two types.Type I proteins are largely uncharacterized in function in plants [3].Type II proteins,also referred as MIKC-type proteins,are characterized by the presence of four characteristic domain structures,namely the MADS (M),intervening (I),keratin-like (K) and C-terminal (C) domains [4].MIKC-type MADS-box genes in plants are involved in diverse aspects of vegetative and reproductive growth and development,as well as different stress responses [5].

So far,MADS-box genes involved in floral and plant development have been identified and characterized in various species,includingArabidopsis[6],soybean[7],maize[8],barley[9],and rice[10].Constitutive expression of theArabidopsis APETALA1(AP1)gene results in early flowering and conversion of inflorescence to the terminal flower[11].In contrast,mutation inAP1causes partial flower transformation into inflorescence shoots with alterations in sepal and petal identities[12].GmAP1regulates flowering time and alters plant height in soybean [13].MIKC-type MADS-box genes are also well characterized in monocots.Maize geneZAG3is expressed in the inflorescence and affects meristem growth [14].Down-regulation ofZmMADS1by RNAi caused delayed flowering,whereas overexpression led to early flowering [15].Barley MADS-box genesHvBM1andHvBM10inhibit floral development and cause floral reversion when ectopically expressed inArabidopsis,andHvBM1overexpression lines also show delayed head emergence [16].Rice genesOsMADS6andOsMADS14are exclusively expressed in inflorescence and developing kernels.Overexpression of these genes caused early flowering and dwarfism,suggesting roles in the early stages of flower development [17].In rice,OsMADS45is not only involved in floral organ identity,but also in regulation of heading date[10,18].Transgenic rice overexpressingOsMADS45displayed earlier heading date than the wild type.On the contrary,heading ofOsMADS45RNAi plants was later.OsMADS45also plays a role in seed development and its inhibition in rice endosperm stabilized amylose content under high temperature stress [19,20].

As an important cereal crop,wheat is grown worldwide.A genome-wide study of the wheat MADS-box gene family showed there are 15 subfamilies of wheat MIKC-type MADS-box genes[21].Previous studies showed that wheat MADS-box genes also play a very important role in flower development and plant architecture.The wheat MADS-box genesFUL1/VRN1,FUL2,andFUL3are crucial for the development of spikelets and spikes,as well as affecting plant height and flowering time.Overexpression ofFUL1andFUL2resulted in early flowering phenotypes [22,23].TaAGL6affects stamen development through transcriptional regulation ofTaAP3[24].WSEPis expressed at the floral organ differentiation stage and subsequent developmental stages in wheat [25].TransgenicArabidopsisplants overexpressingTaMADS1displayed early flowering and terminal flower formation,which is also observed in plants that ectopically expressWSEP,suggesting that these two genes might possess similar functions [26].

In this study,TaSEP3-1,a member of theSEP3subfamily of the MIKC-type MADS-box gene,was isolated and analyzed in bread wheat.Expression analysis showed that the expression patterns of the threeTaSEP3-1homoeologs were similar,and higher expression levels were detected in floral organs and developing kernels.Association analysis showed thatTaSEP3-D1was significantly associated with heading date,plant height and other yield-related traits,andD1_h4was the favored haplotype ofTaSEP3-D1.Transgenic wheat plants overexpressingTaSEP3-D1exhibited delayed heading and reduced plant height.Our overall results indicated thatTaSEP3-D1has pleiotropic functions in regulating heading date and plant architecture,providing a genetic resource for wheat breeding.

2.Materials and methods

2.1.Plant materials and growth conditions

Wheat cultivar Chinese Spring (CS) was used forTaSEP3-1isolation.Kenong 199 (KN199) was used as the recipient genotype for wheat transformation.TheTaSEP3-D1gene sequence was ligated to thepUbi:casvector,and introduced into KN199 using the particle bombardment system of Beijing Genovo Biotechnology Co.,Ltd.Transgenic wheat lines and wild type (WT) KN199 were grown in growth chambers under long day (LD,16 h light/8 h darkness) and short day (SD,8 h light/16 h darkness) photoperiod at (20 ± 2) °C.Transgenic wheat lines and WT were also grown in the experimental field of the Institute of Crop Science in Beijing (39°N,116°E).Wheat materials,including wild materials composed of tetraploid ancestors and synthetic hexaploid wheat lines (SHW),landraces and modern varieties were planted in 2011,2012,2013 and 2014 at three locations,viz.,Beijing (BJ)(40°N,116°E),Jiaozuo (JZ) (35°N,113°E) and Xinxiang (XX) (35°N,113°E) in 4-row,2 m plots with 25 cm between rows.Field management,including irrigation and fertilization was performed according to local practices.

2.2.Isolation and analysis of protein sequences

According to the accession number,the protein sequences of cloned MADS-box genes in wheat and other species includingArabidopsis,rice,barley,and maize were downloaded from the National Center for Biotechnology Information database.Multiple sequence alignments were performed by the Clustal X software[27].The phylogenetic analysis for multiple MADS-box proteins was conducted using the MEGA 6.0 software[28],and the phylogenetic tree was constructed by a Neighbor-Joining method after 1000 bootstrap replications.

2.3.Gene expression analysis of TaSEP3-1

Expression analysis ofTaSEP3-1was performed using various tissues of CS collected at different developmental stages and stored at -80 °C.Total RNA extraction was performed using an RNAprep Pure Plant kit (TIANGEN,China) following the manufacturer’s instructions.One μg of total RNA of each sample was used for first-strand cDNA synthesis using the FastKing RT Kit (TIANGEN,China).qRT-PCR experiments were conducted using the following program:denaturation at 95 °C for 3 min,followed by 55 cycles at 95°C for 10 s and 60°C for 30 s.All qRT-PCR experiments were performed three times and relative transcript levels ofTaSEP3-1were determined using the 2-ΔΔCTmethod [29].Wheat geneTaGAPDHwas used as the internal control.The primers used in this study were summarized in Table S1.

2.4.Subcellular location of TaSEP3-D1

Wheat protoplasts were used to determine the subcellular location of TaSEP3-D1.The coding sequence ofTaSEP3-D1without stop codon was amplified and cloned into theHindIII andBamHI restriction sites of the pJIT163-GFP vector.The TaSEP3-D1-GFP construct and GFP as control were transformed into the wheat protoplasts using a method previously described [30].Fluorescence was observed after incubation for 18 h at 20 °C using the confocal microscope (Leica TCS-NT,Germany).

2.5.Transactivation assay of TaSEP3-D1

To study the transactivation of TaSEP3-D1 in yeast,the fulllength ORF of TaSEP3-D1 was fused in frame to the GAL4-binding domain of the pGBKT7 vector.The constructs,including the empty vector (BD) as negative control,were transformed separately into the yeast strain AH109 and cultured at 30 °C on SD/-Trp medium.After three days’ cultivation,yeast colonies were transferred to fresh SD/-Trp/-His and SD/-Trp/-His/-Ade medium for further cultivation.

2.6.Data analysis

Based on the wheat 660K SNP array,genome-wide haplotype blocks were calculated by using Plink software with default parameters but modification inblocks-max-kbparameter for different chromosomes [31].SNPs surroundingTaSEP3-1genes within the same haplotype block were selected for haplotype analysis.Haplotype frequency and diversity in different material groups were assessed by PowerMarker software [32].Pearson’s correlation coefficients were calculated to analyze the correlation between haplotypes of the threeTaSEP3-1homoeologs and agronomic traits,including days to heading (Hd) and maturity (Md),plant height(PH),spike number per plant (SN),spike length (SL),number of spikelets per spike (SLN),number of grains per spike (GN),grain length(GL),thousand kernel weight (TKW),yield per (4 × 2 × 0.25) m plot,and cold tolerance.The statistically significant differences in phenotype betweenTaSEP3-D1haplotypes and betweenTaSEP3-D1overexpression transgenic lines and WT were evaluated by twotailed independent samplet-tests in SPSS 19.0 software.

3.Results

3.1.Isolation and sequence analysis of TaSEP3-1 homoeologs

Fig.1.Wheat TaSEP3-1 belongs to the MIKC-type MADS-box family.(A)Alignment of MIKC-type MADS-box proteins from different plant species.The characteristic MADS and K domains are marked in the red rectangle and red line,respectively.(B) Phylogenetic tree of MIKC-type MADS-box proteins.

The cDNA ofTaSEP3-1was screened from the sequence library built by the introduction of large-scale wheat transcription factor genes into rice [33].The isolatedTaSEP3-1was designated asTaSEP3-D1(TraesCS7D02G261600)according to its genomic origin after blasting the database of the hexaploid wheat (http://plants.ensembl.org).As bread wheat consists of three genomes,homoeologsTaSEP3-A1located on 7A (TraesCS7A02G260600) andTaSEP3-B1on 7B(TraesCS7B02G158600)were expected.ALL threeTaSEP3-1genes had eight exons and encoded highly similar proteins(>98%identity)with 246 amino acids(Fig.1A).The main differences in gene structure ofTaSEP3-1were caused by transposable elements(TE)in introns 1 and 2(Fig.2A),withTaSEP3-D1(8.5 kb)being larger thanTaSEP3-A1(4.9 kb) andTaSEP3-B1(5.1 kb).As shown in Fig.1,TaSEP3-1 contains typical MADS_MEF2_Like and K-box domains,belonging to the MIKC-type MADS-box protein family.A neighbor-joining phylogenetic tree constructed to determine the relationship between TaSEP3-1 homoeologs and counterparts in other plant species (Fig.1B) showed that TaSEP3-A1,TaSEP3-B1,and TaSEP3-D1 clustered with BM9,ZMM6,OsMADS45,and OsMADS24.Among them,OsMADS45andOsMADS24exhibit similar sequences and expression patterns,and both participate in floral development.Knockdown of both genes caused morphological alterations of floral organs and significantly delayed the heading date[18].ZMM6is strongly expressed during kernel development in maize [34].BM9is also involved in the regulation of flower development [35].The phylogenetic analysis suggested that TaSEP3-1 might possess functions similar to these MADS-box proteins.

3.2.Subcellular localization and transactivation activity of TaSEP3-D1

To determine the subcellular localization of TaSEP3-D1,the coding sequence ofTaSEP3-D1without stop codon was amplified and fused with the GFP gene sequence driven by the 35S promoter.Expression of the (35S::GFP) control and TaSEP3-D1-GFP(35S::TaSEP3-D1::GFP) fusion protein in wheat protoplasts(Fig.3A) showed that that TaSEP3-D1 was located in the nucleus and cytoplasm.MADS-box protein SEPALLATA3 (SEP3) [36],the ortholog of TaSEP3-1 inArabidopsis,was likewise shown to be located in both the cytoplasm and nucleus.It was been reported that MADS domain protein MPF2 inPhysalis floridanawas located in the cytoplasm and imported into the nucleus upon hormone treatment with cytokinin [37].The nucleuscytoplasmic distribution characteristic of these MADS-box proteins is possibly associated with roles other than transcriptional control [36].

To study the transactivation activity of TaSEP3-D1 in yeast,the full-length ORF ofTaSEP3-D1was fused in frame to the GAL4-binding domain of the pGBKT7 vector(Fig.3B).The results showed that the yeast harboring the full length ofTaSEP3-D1was able to grow on SD-Trp-His and SD-Trp-His-Ade medium,indicating that TaSEP3-D1 possessed transactivation activity in yeast.

Fig.2.Gene structure,polymorphisms,and haplotype frequency of TaSEP3-1.(A) Gene structure and sequence polymorphisms of TaSEP3-1.(B) Frequency of TaSEP3-1 haplotypes in wild accessions (W),landraces (L),and modern varieties (M).

Fig.4.Expression analysis of TaSEP3-D1 by qRT-PCR.SR,roots at the seedling stage;SL,leaves at the seedling stage;developing spikes 5 mm to 60 mm in length;5 DPA,developing kernels at 5 days post anthesis.Error bars represent standard deviation from three replicates. TaGAPDH was used as the internal control.The experiments were performed three times with similar results.

Fig.5.Phenotypic analysis of transgenic wheat lines(OE)overexpressing TaSEP3-D1.(A,B)Comparison of heading dates of WT KN199 and OE lines grown in LD conditions.DAG,days after germination.(C,D)Spike lengths(C)and plant heights(D)of the same WT and OE lines.Scale bar,2 cm.Data are means±standard deviation of 20 plants;**,P <0.01 (t-test).

3.3.Expression patterns of TaSEP3-1 in wheat

qRT-PCR was performed to analyze the expression patterns ofTaSEP3-D1at different developmental stages in various tissues and organs.Transcripts ofTaSEP3-D1were detected in young spikes,glumes,lemmas and paleas,stamens,pistils and lodicules as well as developing kernels at 5 days post anthesis (DPA)(Fig.4).The expression levels in tissues such as root and leaf were extremely low.Similar expression patterns were also observed forTaSEP3-A1andTaSEP3-B1(Fig.S1).The results were consistent with a previous study showing thatTaSEP3-1(WSEP) was expressed in young spikes,and predominantly expressed in paleas and pistils with attached lodicules [25].Transcripts ofTaSEP3-1were also detected in developing kernels in this study,indicating that theTaSEP3-1is not only involved in the inflorescence development but also in seed development.

3.4.Overexpression of TaSEP3-D1 influences growth

Fig.6.Phenotypic comparisons of three TaSEP3-D1 haplotypes in seven environments.Traits included plant height,heading date,number of kernels per spike,number of spikelets per spike,and TKW.E1 to E7 indicate environments 2014XX,2014BJ,2014JZ,2013XX,2013BJ,2012XX,and 2011XX,respectively.Heading date,days from April 1.The error bars represent standard error.*, P ≤0.01.

To characterize the functions ofTaSEP3-D1,the coding sequence ofTaSEP3-D1was transformed into the bread wheat cultivar KN199.Two independentTaSEP3-D1-overexpression transgenic lines(line 8 and line 9)with different expression levels were identified and used for phenotypic evaluation (Fig.S2).Agronomic traits,including heading date,plant height,and spike length,were investigated in lines overexpressingTaSEP3-D1and WT KN199(Fig.5).Compared with the WT,plants overexpressingTaSEP3-D1exhibited delayed heading under both LD and SD conditions in the controlled growth chamber.Furthermore,under natural vernalization conditions in the field,theTaSEP3-D1overexpressing lines were also 8–9 days later in heading date than the WT(Fig.S3).In addition to delayed heading,plant height of the overexpression lines was also reduced and spike length was increased.In summary,TaSEP3-D1negatively regulated heading time,and the transgenic lines exhibited significant changes in plant height and spike length,indicating thatTaSEP3-D1had multiple effects in modulating wheat growth and development.

3.5.Frequency and diversity of TaSEP3-1 haplotypes

The Wheat 660K SNP Array was employed to detect sequence polymorphisms ofTaSEP3-A1,TaSEP3-B1,andTaSEP3-D1.Compared with the CS reference genome,TaSEP3-A1,TaSEP3-B1,andTaSEP3-D1were positioned at 253,427,929–253,432,840 bp on chromosome 7A (Chr.7A),215,001,224–215,006,291 bp on Chr.7B and 237,610,173–237,618,708 bp on Chr.7D,respectively.SNPs positioned around theTaSEP3-1genes were used to define haplotype blocks.Haplotype blocks calculated using Plink ranged from 251,523,569 bp to 254,042,554 bp on Chr.7A,213,982,297 bp to 215,361,179 bp on Chr.7B and 237,028,604 bp to 239,363,204 bp on Chr.7D.SNPs surrounding theTaSEP3-1gene within the same haplotype block were then used for haplotype analysis.Detailed genotypes of all the wheat accessions assessed are listed in Table S2.Four,seven,and three SNPs sites were selected,respectively,to identify the haplotypes ofTaSEP3-A1,TaSEP3-B1,andTaSEP3-D1in tested wheat materials.There were four,three,and four haplotypes ofTaSEP3-A1,TaSEP3-B1,andTaSEP3-D1,respectively (Fig.2A).

Previous studies had shown that favorable allelic variants continue to accumulate during artificial selection [38].In order to determine whether particular haplotypes ofTaSEP3-A1,TaSEP3-B1,andTaSEP3-D1have been selected during wheat breeding,we investigated the frequency and diversity of these haplotypes.As shown in Fig.2B,onlyA1_h1andA1_h2ofTaSEP3-A1were present in wild accessions,and the former accounted for 76%of accessions.Landraces contained all four haplotypes whereas only three(A1_h1,A1_h2,andA1_h3) were present in modern varieties.The frequency ofA1_h1decreased from 24%in landraces to 8%in modern varieties.On the contrary,compared with landraces,the proportion ofA1-h3in modern varieties increased about sevenfold,indicating thatA1_h3had undergone positive selection during the breeding.

All three haplotypes ofTaSEP3-B1were found in wild accessions,landraces,and modern varieties.In the wild accessions,the frequencies of the three haplotypes were 41% (B1_h1),31%(B1_h2),and 27% (B1_h3),respectively.The frequency of the first two haplotypes decreased from landraces to modern varieties,whereasB1_h3exhibited the opposite trend,increasing from 64%among landraces to 94% in modern varieties,suggesting thatB1_h3had undergone positive selection under breeding.

ForTaSEP3-D1,D1_h1was predominated in wild accessions.The proportions ofD1_h2andD1_h3decreased from landraces to modern varieties,whereasD1_h4increased from 5% to 45% in modern varieties,respectively,indicating that it had undergone positive selection in breeding,and was the favored haplotype ofTaSEP3-D1.Moreover,D1_h4was found only in landraces and modern cultivars,suggesting that it might play an important role in bread wheat.The geographic distribution ofTaSEP3-D1haplotypes in five major Chinese wheat production zones(Table S3)revealed that the proportion ofD1_h4genotypes among modern cultivars was higher than that in landraces in all five zones,and the proportions ofD1_h4in modern cultivars was consistently higher than those ofD1_h2andD1_h3.

Haplotype diversity (PIC value) ofTaSEP3-1was calculated using PowerMarker software[32].On average,haplotype diversity ofTaSEP3-A1was lower in wild accessions(0.30)than in landraces(0.44)and modern cultivars(0.46).The diversity ofTaSEP3-B1haplotypes was greatly decreased from wild(0.58)and landrace(0.41)accessions to modern cultivars(0.11),implying thatTaSEP3-B1had been subject to positive selection in wheat breeding.The diversity ofTaSEP3-D1was a little higher in modern cultivars(0.50) than in wild (0.41) and landrace (0.41) accessions,indicating diversifying selection onTaSEP3-D1during breeding (Fig.S4).

3.6.Association of TaSEP3-D1 haplotypes with agronomic traits

Association analyses were made betweenTaSEP3-1haplotypes and agronomic traits collected in seven environments.The number of traits significantly correlated with haplotypes ofTaSEP3-A1,TaSEP3-B1,andTaSEP3-D1in at least two environments (P<0.01)was one,three,and nine,respectively (Table S4),indicating the importance ofTaSEP3-D1in regulating agronomic traits.There were significant differences in traits between the threeTaSEP3-D1haplotypes (Fig.6).Compared with the other twoTaSEP3-D1haplotypes,accessions withD1_h4exhibited significantly reduced plant height(about 12.4 to 24.2 cm)and heading date(about 2.2 to 2.6 days)in all seven environments.Significant differences in other traits were also detected between differentTaSEP3-D1haplotypes.As shown in Fig.6,the number of spikelets per spike was significantly higher in accessions withD1_h4than those withD1_h2andD1_h3in environments 2014XX,2014BJ,and 2011XX.The same was true for number of kernels per spike and TKW in six environments.These results implied thatTaSEP3-D1could be involved in multiple aspects of wheat growth and development,and thatD1_h4was a favored haplotype selected in breeding.

4.Discussion

MADS-box genes encode a large transcription factor family and play important roles in flower development and growth regulation in plants.Flowering involves a phase transition from vegetative to reproductive growth in higher plants,and is mainly controlled by vernalization,photoperiod,autonomous,and some other pathways[5].MADS-box genes involved in flowering regulation have been well-studied inArabidopsis,rice,and maize[5].In temperate cereal crops such as wheat and barley,heading time associated with the timing of flowering is an important character because of its influence on adaptation to diverse environments.Several MADS-box related to flowering time in wheat have been cloned and characterized.VRN1is an important regulator of flowering time [39],and different allelic combinations of the homoeologous genesVRN-A1,VRN-B1,andVRN-D1lead to different times of heading [40].In addition,FUL2andFUL3,paralogs ofVRN1,also affect flowering time [22].Ectopic expression ofTaMADSinArabidopsisresults in early flowering [5].TaMADS1,ortholog of the rice OsMADS24 in wheat,is involved in regulation of flower development.Overexpression ofTaMADS1inArabidopsisleads to early flowering and abnormal development of floral organs [41].Overexpression ofTaAGL6inArabidopsispromotes early flowering [24].Although some studies have identified the flowering regulation of some MADS-box genes in wheat,there is still much to be studied compared with other species such asArabidopsisand rice.

In this study,theSEP3-like MADS-box geneTaSEP3-1was analyzed.Transcripts ofTaSEP3-D1were detected in young spikes,glumes,lemmas and paleas,stamens,pistils and lodicules consistent with a previous study [25].We also identified transcripts ofTaSEP3-1in developing kernels.It was reported thatOsMADS45,the ortholog ofTaSEP3-1,was not only involved in the regulation of heading date,but also involved in seed development,and inhibition ofOsMADS45in rice endosperm stabilized amylose content under conditions of high temperature stress [18–20].Therefore,TaSEP3-1might also be involved in seed development.

TaSEP3-1regulated heading time in wheat.Delayed heading of wheat plants overexpressingTaSEP3-D1was not consistent with the function ofOsMADS45in rice andWSEPinArabidopsis[20,25].The likely reason for this discrepancy was attributed to differences in genetic background.A similar phenomenon was observed for theGW2genes in rice and wheat.TaGW2in wheat andGW2in rice have divergent functions in grain development.It was reported thatGW2negatively regulates grain size in rice[42],but bothTaGW2RNAi andTaGW2knockdown plants showed significantly decreased grain size,indicating thatTaGW2positively regulates the grain size [43].Therefore,it would be interesting to examine whetherTaSEP3-1knockout plants show earlier heading.We are currently using CRISPR/Cas9 genome editing to constructTaSEP3-1knockout plants,and the T0plants should be available within a few months.The identifiedTasep3-1genotypes will be a valuable resource for future experiments aimed at further clarifying the function of this gene.Moreover,delayed heading was observed under both controlled LD and SD conditions(10–12 days),as well as field environments(8–9 days).This inconsistent effect ofTaSEP3-D1on flowering time might be due to involvement of theSEP3-like gene in different flowering regulation pathways in dicots and monocots.Moreover,it seemed that the delayed heading was likely related to the duration of low temperature (30 days in growth chambers verses 2.5 months in the field) than the day length (LD and SD).We propose thatTaSEP3-D1might interact with vernalization genes.Further research will be necessary to extend our understanding ofTaSEP3-1gene in regulation of flowering time.

Plant height is a decisive factor of plant architecture[44]and is also an important breeding target as it affects grain yield [45].Recent studies have shown that some genes,involved in regulation of the flowering pathway,may also affect other yield-related traits[46].For example,Ppd-D1affects heading date and has pleiotropic effects on a variety of agronomic traits,such as tiller number,number of spikelets per spike,and plant height [47].The three MADSbox genesVRN1,FUL2,andFUL3,in addition to regulating flowering time,play vital roles in development of spikelets and spikes,and also affect plant height [22].In this study,wheat plants overexpressingTaSEP3-D1exhibited not only delayed heading compared with WT plants,but also had reduced plant height and increased spike length.The observed phenotypic changes indicated that in addition to flowering regulation,TaSEP3-D1might have pleiotropic roles in modulating wheat growth and development,or might regulate genes involved in determination of plant height and spike length.Previous studies showed that TaSEP3-1 proteins interact with VRN1 [25].SinceVRN1can affect plant height,whether this interaction or changes in downstream genes regulated by this interaction leads to changes in plant height needs further investigation.Therefore,further research will help to explore the underlying mechanism ofTaSEP3-1regulating growth and development of wheat.

Wheat is one of the earliest domesticated crops and is widely cultivated.Genes conferring to plant architecture and flowering time are crucial for wheat yield and adaptation[46–49].Discovery of favored allelic variants has significant potential in breeding application.Our findings suggested thatTaSEP3-D1was associated with plant height,heading date,and many other yield-related traits.D1_h4was significantly associated with shorter plant height,reduced heading date,higher spikelet number per spike,number of kernels per spike,and TKW,all favorable traits in wheat breeding.Frequency analysis ofTaSEP3-D1haplotypes in wild accessions,landraces,and modern varieties indicated thatD1_h4had undergone selection and was the favored haplotype ofTaSEP3-D1.In this study,we further investigated the geographic distribution ofTaSEP3-D1haplotypes in five major wheat production zones in China (Table S3).The proportion ofD1_h4in modern varieties was higher than that in landraces in all five zones,and the proportions in modern varieties were higher thanD1_h2andD_h3.These results suggested thatD1_h4had undergone selection and was positively selected in Chinese breeding programs.Therefore,the favored haplotypeD1_h4could be selected for improvement of plant growth and architecture.

Declaration of competing interest

The authors declare that they have no conflict of interests.

CRediT authorship contribution statement

Lifeng Gao,Lei Zhang,and Hao Zhangdesigned the experiments;Lei Zhang and Hao Zhangperformed the experiments;Hao Zhang,Lingfeng Miao,Dong Yan,and Pan Liucollected traits data;Linyi Qiao and Guangyao Zhaohelped for data analysis;Lei Zhang and Hao Zhangwrote the manuscript;Jizeng Jia and Lifeng Gaorevised the manuscript.

Acknowledgments

This work was funded by the Basic Scientific Research Project of Chinese Academy of Agricultural Sciences (Y2019XK10) and the National Key Research and Development Program of China(2016YFD0100102,2016YFD0100302).

Appendix A.Supplementary data

Supplementary data for this article can be found online at https://doi.org/10.1016/j.cj.2020.12.007.