Xioling M,Xing Ouyng,Dongheng Liu,Aimin Zhng

a Key Laboratory of Cultivation and Protection for Non-Wood Forest Trees of Ministry of Education and the Key Laboratory of Non-Wood Forest Products of Forestry Ministry,Central South University of Forestry and Technology,Changsha 410004,Hunan,China

b State Key Laboratory of North China Crop Improvement and Regulation,College of Agronomy,Hebei Agricultural University,Baoding 071000,Hebei,China

c State Key Laboratory of Hybrid Rice,Hunan Hybrid Rice Research Center,Hunan Academy of Agricultural Sciences,Changsha 410125,Hunan,China

Keywords:Triticum aestivum Grain weight TaAGPS Haplotype Starch synthesis

ABSTRACT ADP-glucose pyrophosphorylase(AGPase)influences cereal productivity.There are few reports on the function of cytosolic AGPase small subunit in bread wheat(TaAGPS).In the present study,TaAGPS was preferentially expressed in developing endosperm during grain-filling stages in bread wheat.TaAGPS allelic variations were characterized in 143 wheat accessions by PacBio RS II sequencing.Two haplotypes(TaAGPS-7A-TG and TaAGPS-7A-CT)of TaAGPS-7A were identified and corresponding functional markers were developed,whereas no variants of TaAGPS-7B and TaAGPS-7D were detected.TaAGPS-7A was associated with thousand-kernel weight(TKW)by haplotype–trait association analysis in two populations.Near-isogenic lines(NILs)with TaAGPS-7A-TG showed higher TKW and total kernel starch content than those with TaAGPS-7A-CT,owing to the higher AGPase activity of TaAGPS-7A-TG than TaAGPS-7A-CT both in vitro and in vivo.Overexpression of TaAGPS-7A-TG in bread wheat doubled the transcription levels of TaAGPS and increased AGPase activity by 55.7%,resulting in a 3.0-g higher TKW than in the wild type(WT).Knockdown of TaAGPS led to reduced expression of TaAGPS,AGPase activity,and TKW than in the WT.Thus,owing to the 218th amino acid change of Ser to Ala in TaAGPS-7A,the favorable haplotype TaAGPS-7A-TG showed higher AGPase activity,resulting in higher kernel starch content and grain weight.This finding could be applied to increasing starch content and grain weight in bread wheat.

1.Introduction

Bread wheat(Triticum aestivum L.)is a major cereal crop[1].The demand for wheat is expected to be 900 million metric tons of grain by 2050,owing to human population growth(https://www.fao.org/fileadmin/templates/esa/Global_persepctives/Presentations/Fischer_pres.pdf).The best strategy for meeting this demand is to increase yield potential by exploiting the major yield genes[2].Although the complex and large allohexaploid genome of wheat(17 Gb,genome formula AABBDD)brings huge challenges to molecular genetics studies,the complex genome confers wide adaptability and flexibility[3].

The key yield traits are thousand-kernel weight(TKW),grain number per spike,and spike number per unit area[4].In modern breeding,TKW has been the main target trait for improving grain yield[4,5].Its genetic improvement is expected to increase wheat yield potential[2,6,7].In wheat kernels,the amount and proportion of starch,accounting for 65%–75% of dry weight,strongly affect TKW,thereby affecting yield potential[8].

Starch synthesis is a complex process.At least six enzymes are involved in the synthesis of endosperm starch:ADP-glucose pyrophosphorylase(AGPase),granule-bound starch synthase(GBSS),soluble starch synthesis(SS),branching enzyme(BE),debranching enzyme(DBE),and glucose-6 phosphotransferase(GPT)[9].AGPase catalyzes the conversion of glucose-1-phosphate(G-1-P)to ADP-glucose(ADPG),which is the main substrate for amylose and amylopectin synthesis[10].AGPase is considered[11]the key enzyme regulating the rate and efficiency of starch synthesis in higher plants such as wheat,rice,and maize,and its activity is positively correlated with grain filling rate and starch content accumulation in cereals[12–14].A maize AGPase gene with specific site mutations increased grain weight and starch content in bread wheat[15].Mutation or knockdown of AGPase in cereals also reduced seed starch synthesis and grain weight[10].

AGPase is a heterotetramer that comprises two large and two small subunits[16].Both large and small subunits have two types,cytosolic and plastidial types,based on location in endosperm cells[17].The small subunits of AGPase are relatively conserved and have both catalytic and regulatory activities[1,18].Mutations in rice AGPS2 led to reduction of AGPase enzyme activity and endosperm starch accumulation[19].The T-DNA insertion mutant of AGPase small subunit(APS1)in Arabidopsis thaliana lost most AGPase activity and starch synthesis ability[20].

Although TaAGPS(GenBank:EF405961)was shown to be associated with TKW[21,22],characterization of novel positive alleles of TaAGPS and identifying their precise role in grain development is desirable.The purpose of the present study was to identify haplotypes of TaAGPS and characterize their effects on starch content and grain weight.

2.Materials and methods

2.1.Plant materials,growth conditions,and evaluation of phenotypic traits

A panel of 143 wheat accessions composed of Chinese cultivars,landraces and foreign accessions(hereafter called the germplasm panel)were used for identification of TaAGPS gene allelic variations(Table S1).These accessions were obtained from the Institute of Genetics and Developmental Biology,Chinese Academy of Sciences(IGDB,CAS)and grown at the Institute’s experimental station(40°11′N,116°42′E)in Beijing in the 2015–2016 cropping season.A population of 246 F8recombinant inbred lines(RILs)derived from the cross Zhou 8425B/Chinese Spring[23]was kindly provided by the Institute of Crop Science,Chinese Academy of Agricultural Sciences.The RILs were planted at Zhengzhou(34°78′N,113°66′E)and Zhoukou(33°62′N,114°65′E)of Henan province,during the 2012–2013 and 2013–2014 growing seasons[23].

Luyuan 212 and Jingdong 8 are two high-TKW bread wheat varieties in China and were used as female parents.Yemaizi,a low-TKW landrace,was used as the donor to cross with Luyuan 212 and Jingdong 8.The F1plants were backcrossed once with Luyuan 212 or Jingdong 8,separately.The resulting BC1F1plants were sequentially self-pollinated to produce two BC1F5populations,LY(Luyuan 212/Yemaizi)and JY(Jingdong 8/Yemaizi).Both populations consisted of 300 lines.In each population,three TaAGPS-7A heterozygous lines were identified and further advanced.Six pairs of BC1F8near-isogenic lines(NILs)derived from LY(LY1,LY2 and LY3)and JY(JY1,JY2 and JY3)were used for estimating TKW of TaAGPS-7A-CT and TaAGPS-7A-TG.To investigate the regulation of TKW by the two haplotypes of TaAGPS-7A,the expression levels of TaAGPS genes,AGPase activities,and total starch contents of mature kernels of NILs(LY1,LY2,JY1 and JY2)with different haplotypes were measured.These NILs were generated in IGDB,CAS and grown at two experimental stations,in Beijing(40°11′N,116°42′E)and in Zhaoxian(37°45′N,114°46′E,Hebei,China)during the 2016–2017 growing season.

To further confirm the effect of TaAGPS gene on wheat TKW,overexpression(UBIpro:TaAGPS-7A-TG)and knockdown(TaAGPSRNAi)constructs of TaAGPS were transformed separately into the bread wheat cultivar Kenong 199(haplotype:TaAGPS-7A-TG).The T2generation of the overexpression lines(OE 1 and OE 2),the RNA interference lines(RNAi 1 and RNAi 2),and the wild type(WT)Kenong 199 were planted at Zhaoxian(37°45′N,114°46′E)and Dishang(37°95′N,114°73′E)experimental stations in Hebei,China during the 2017–2018 growing season.At each experimental station,all the accessions,NILs,transgenic lines and WT were planted with three biological replicates.For each replicate,each line was planted in three 1.5-m rows with 20 cm between rows and 5 cm between plants.Field management followed local practices.

Six pairs of BC1F8NILs derived from LY(LY1,LY2 and LY3)and JY(JY1,JY2 and JY3)with different TaAGPS haplotypes were planted in Beijing and Zhaoxian,and 15 plants of each line were randomly selected for the measurement of TKW.The TKWs of transgenics and NILs were measured using SC-G software(Wanshen Detection Technology Co., Ltd.,Zhejiang,China).Three independent samples of 300 kernels were weighed and the means were converted to TKW.Simultaneously,eight plants of individual overexpression and knockdown lines planted in Beijing and Zhaoxian were randomly selected for TKW measurement.Other agronomic traits of transgenic lines were also recorded.

The seeds of 10 randomly selected plants of four NILs(LY1,LY2,JY1,and JY2)with different TaAGPS-7A haplotypes were harvested and cleaned,and 100-g samples were taken for milling.The grain samples were moistened for 24–36 h to adjust moisture content to 17.0%±0.5%and milled with a Brabender Junior mill(Brabender,Duisburg,Germany).The flour was sieved through a 70 GG screen.Total starch content was measured with a Megazyme Total Starch Assay Kit(catalog K-TSTA,Irishtown,Ireland)[1].

2.2.RNA extraction and quantitative reverse transcription PCR(qRTPCR)analysis

Total RNA was extracted from developing endosperms,roots,stems,and leaves with TRNzol Universal Reagent(Tiangen,Beijing,China)according to the manufacturer’s instructions[24].Firststrand cDNA was synthesized from 2.0 μg of total RNA using the Fast Quant RT Kit(Tiangen).qRT-PCR was conducted with a Lightcycler 480(Roche,Basel,Switzerland)using the SYBR Green I Master Kit(Roche).The relative expressions of three homoeologous TaAGPS genes and total TaAGPS in the tissues were calculated from three biological and three technical replicates by the 2-ΔΔCT method[25,26].Ta4045(ubiquinol-cytochrome C reductase iron–sulfur subunit)was adopted as an internal control[27].The primers for qRT-PCR are listed in Table S2.

2.3.Haplotype analysis of TaAGPS gene and development of cleaved amplified polymorphic sequence(CAPS)markers

Pacific Biosciences RS II(PacBio RS II)DNA sequencing system(Pacific Biosciences,Menlo Park,CA,USA)was used to identify the allelic variations of the TaAGPS gene in the 143 wheat accessions.In each library for PacBio RS II sequencing,PCR products of up to 36 accessions(Table S1)and 29 genes involved in starch synthesis(including TaAGPS)were mixed in equal amounts.Two primer pairs conserved in A,B and D subgenomes were applied to amplify the promoter and coding sequences of TaAGPS homeologs(Table S2).Eight-bp barcode sequences were ligated to the 5′ends of the forward and reverse primers to distinguish the sequence information of the same gene from the different accessions in the same library.Six forward and six reverse primers with different barcodes,resulting in 36 combinations,were designed to amplify TaAGPS gene from the 36 accessions(Tables S3,S4).

Genomic DNA was extracted from young seedlings of each accession using the CTAB method[28].PCR amplification with LA-Taq DNA polymerase(Takara Bio,Otsu,Japan)was performed with the following program:95 °C for 5 min,35 cycles of 95 °C for 30 s,56 °C for 30 s and 72 °C for 4 min 30 s,and finally 72 °C for 10 min.PCR products were purified with a TIANgel Midi Purification Kit(Tiangen)and then cloned into the pGEM-T cloning vector(Tiangen),and transformed into E.coli cells by the heat shock method[6].Positive clones were selected based on PCR verification and were sequenced(Beijing Genomics Institute,Beijing,China).Sequence alignments and SNP identification of TaAGPS genes were performed with DNASTAR(https://www.dnastar.com/).

A CAPS marker was developed according to the single nucleotide polymorphism(SNP)at the 5090 bp position for detecting the haplotype of TaAGPS-7A in 143 wheat cultivars.The germplasm panel,RILs,and NILs were amplified with the primer pair ScytoABDF3/Scyto3R(Table S2),and the PCR products were digested with Dde I restriction endonuclease(New England Biolabs(NEB),Ipswich,MA,USA).The digestion products were separated by electrophoresis in 2%agarose gels.The haplotypes of TaAGPS-7A classified by the CAPS marker were validated by Sanger sequencing in 30 randomly chose wheat accessions(Table S1).

2.4.Vector construction and plant transformatio n

In the overexpression construct(UBIpro:TaAGPS-7A-TG),the ubiquitin(Ubi)promoter was applied to drive TaAGPS-7A-TG overexpression in wheat.In the RNAi construct,the fragment(273–644 bp)distributed in the second and third exons of TaAGPS was obtained as the target sequence in the RNAi hairpin and driven by the maize Ubi promoter.This fragment was conserved among the three TaAGPS genes in Kenong 199(haplotype TaAGPS-7A-TG).The fourth intron of bread wheat Wx gene(GenBank:LC373577)was used as a spacer in the RNAi hairpin.Both the overexpression and RNAi plasmids were transformed into immature embryos of Kenong 199 by biolistic particle delivery[29].In these transgenic experiments,the herbicide bialaphos(bar)gene was employed to select positive transformants,and all transgenic lines were obtained from independent transformation events.All primers used for vector construction and transformant screening are listed in Table S2.

2.5.The in vitro and in vivo enzyme activity assay of TaAGPS

For the in vitro enzyme activity assay of TaAGPS,a prokaryotic expression system was used.Firstly,the two types of cDNA fragments of TaAGPS-7A were amplified from the bread wheat cultivars Chinese Spring(TaAGPS-7A-CT)and Xiaoyan 81(TaAGPS-7A-TG),respectively,by using the ScytocdF2 and ScytocdR1 primer pair(Table S2).The PCR product(1444 bp)was inserted in the pGEMT Easy vector(TransGen,Beijing,China),and then the plasmids containing the precisely cloned fragments of respectively TaAGPS-7A-CT and TaAGPS-7A-TG with 1422 bp were used as templates to amplify the two types of TaAGPS full-length cDNA sequences with the primer pair AGPS-F3/AGPS-R3(Table S2).The full-length cDNA fragments were inserted in pGEM-T Easy vectors,and then the plasmids containing the precisely cloned fragments of

TaAGPS-7A-CT and TaAGPS-7A-TG were digested with Sal I and Hind III enzymes.The expected fragments were inserted into the transcription activation expression vectors pMAL-2(NEB)cleaved with the same restriction enzymes(Sal I and Hind III)and the resulting plasmids were named pMAL-2-TaAGPS-7A-CT and pMAL-2-TaAGPS-7A-TG,respectively.These plasmids were transformed into Escherichia coli strain BL21-glgC to obtain the pMAL-2-TaAGPS-7A fusion protein.After the positive strain was obtained,0.3 mmol L-1IPTG was added to induce the expression of the protein at 37 °C for 3 h,and the protein was checked by SDS-polyacrylamide gel electrophoresis(SDS-PAGE)(with 12%acrylamide).The induced fusion proteins,designated as pMAL-2-TaAGPS-7A-CT,pMAL-2-TaAGPS-7A-TG and pMAL-2-MBP(empty vector),were purified and quantified as previously described[30].

For in vivo assay of the total AGPase activity,the endosperms at 10 DPA of transgenic lines(OE 1,OE 2,RNAi 1,RNAi 2)and six pairs of NILs(LY1,LY2,LY3,JY1,JY2,and JY3)were collected,with the endosperms at 10 DPA of non-transgenic Kenong 199 as the control.Each sample had three biological replicates.

The AGPase activity assay was performed using the AGPase Activity Kit(Comin Biotechnology,Jiangsu,China)following the available protocol[31,32].Because AGPase catalyzes the conversion of ADP-glucose into NADPH by interacting with glucose-6-phosphate dehydrogenase and phosphoglucomutase,AGPase activity can be estimated from the amount of degraded ADP-glucose by measuring NADPH concentration with a spectrophotometer at 340 nm.AGPase activity was characterized in micromoles of NADPH per minute(U)per gram of fresh weight(FW).

2.6.Statistical analysis

Statistical analysis was performed by one-way ANOVA in the IBM SPSS 20 package for Windows(IBM,Armonk,NY,USA).

3.Results

3.1.TaAGPS is preferentially expressed in developing endosperms of bread wheat

Expression of TaAGPS was not detected in roots,stems or leaves of Chinese Spring,but was pronounced in developing endosperm(Fig.S1A).TaAGPS showed a peak expression level at 10 DPA and then gradually decreased at 20–25 DPA,suggesting that the TaAGPS was expressed specifically in the developing endosperms and was most highly expressed at the earlier middle stages of grain development.Accordingly,the developing endosperms at 10 DPA of wheat were selected for further qRT-PCR and AGPase activity assay.

To further investigate the expression of the three TaAGPS homeologs(TaAGPS-7A,TaAGPS-7B and TaAGPS-7D),RNA from developing endosperms of Chinese Spring from 5 to 25 DPA were isolated.Each TaAGPS allele showed a similar expression pattern to the total TaAGPS at all stages of endosperm development.However,the relative expression levels of the three TaAGPS genes showed large differences.The expression of TaAGPS-7D was highest,TaAGPS-7A was medium,and TaAGPS-7B was lowest(Fig.S1B).

3.2.Identification of TaAGPS allelic variants and development of a DNA marker

Based on the result of PacBio RS II sequencing of TaAGPS in the 143 wheat accessions,no variant in the promoter and genomic regions of TaAGPS-7B and TaAGPS-7D was detected among the 143 bread wheat accessions examined(data not shown).However,only two SNP variants were detected in the TaAGPS-7A.One variant was located in the promoter region 476 bp 5′to the start codon of ATG(C-476 T),and the other was located in the exon III region,5090 bp 3′to the ATG(T5090G)and led to a predicted change of the 218th amino acid from Ser to Ala.To confirm these two SNPs,30 randomly selected wheat accessions(Table S1)were subjected to the identification of variations in three TaAGPS genes(TaAGPSA/B/D)by cloning and Sanger sequencing(Seven wheat accessions were listed in Table 1).These two variants formed two TaAGPS-7Ahaplotypes(TaAGPS-7A-TG and TaAGPS-7A-CT)in the germplasm panel(Table 1).

Table 1 Haplotypes of three TaAGPS genes in seven wheat accessions.

Sequence analysis showed the haplotype of TaAGPS-7A-CT contained a Dde I endonuclease site(CTCAG)at the second SNP locus(the 5090th bp position),whereas the haplotype of TaAGPS-7A-TG(CGCAG)did not.To provide a fast and effective tool for detecting the favorable haplotype of TaAGPS-7A in bread wheat,a CAPS marker was developed based on the T-G nucleotide polymorphism(the 5090th bp position).By using the primer pair of ScytoABDF3 and Scyto3R(Table S2),a 636 bp amplicon was obtained,which could be digested with Dde I.After digestions,the TaAGPS-7A-CT type produced two short fragments of 234 bp and 402 bp,and there were no short fragments of the TaAGPS-7A-TG type,as shown by agarose electrophoresis(Fig.S2).According to the established DNA marker,all the haplotype of TaAGPS-7A in 143 wheat accessions was classified into two groups(Table S1).

3.3.TaAGPS-7A is associated with TKW and TaAGPS-7A-TG is the superior haplotype

By using the developed CAPS marker,the germplasm panel was separated into two groups based on the TaAGPS-7A loci,confirming the reliability of the initial PacBio RS II data.The TaAGPS-7A-TG had a significantly higher effect on TKW(P＜0.01)than TaAGPS-7A-CT(Fig.1A).Such as the Zhou 8425B with TaAGPS-7A-TG had higher TKW,while Chinese Spring with the TaAGPS-7A-CT haplotype had lower(P＜0.01)TKW.The effects of TaAGPS-7A haplotypes on TKW were further confirmed in the RIL population which derived from the cross between Zhou 8425B and Chinese Spring.Our results showed that the TaAGPS-7A-TG haplotype-containing lines had higher(P＜0.05)mean TKW than TaAGPS-7A-CT in Zhengzhou and Zhoukou during the 2012–2013 and 2013–2014 seasons(Figs.1B,S3).These findings suggested that the most likely candidate gene influencing TKW in the population developed from Zhou 8425B/Chinese Spring was TaAGPS-7A.

Six pairs of BC1F8NILs representing two haplotypes of TaAGPS-7A,which were obtained from the two different cross combinations,Luyuan 212/Yemaizi(LY)and Jingdong 8/Yemaizi(JY)respectively were used for further analysis.The TKW of three NILs of LY with TaAGPS-7A-TG(named LY1-TG,LY2-TG and LY3-TG)were 5.88,4.63,and 2.08 g higher than corresponding NILs of TaAGPS-7A-CT(named LY1-CT,LY2-CT and LY3-CT)in Beijing(Fig.1C).The TKW of TaAGPS-7A-TG NILs were also higher(P＜0.05)than that of TaAGPS-7A-CT by 1.49–5.88 g at Zhaoxian.Similar data were also obtained from the NIL pairs of JY,in which the JY1-TG,JY2-TG and JY3-TG had higher(P＜0.05)TKW than JY1-CT,JY2-CT and JY3-CT,respectively,in both Beijing and Zhaoxian experimental stations(Fig.1C).Thus,the haplotype TaAGPS-7ATG confers higher TKW than TaAGPS-7A-CT.

3.4.TaAGPS-7A haplotype affects total starch content

The different effects on total kernel starch contents between the TaAGPS-7A-TG and TaAGPS-7A-CT were also identified by using the LY and JY high generation NILs.The results showed that the total starch contents of LY1-TG(76.63%),LY2-TG(78.09%),JY1-TG(78.67%)and JY2-TG(81.21%)planted at Beijing were significantly higher than those of LY1-CT(66.15%),LY2-CT(73.02%),JY1-CT(73.80%),and JY2-CT(78.71%),respectively(Fig.1D).The consistent results were also received at Zhaoxian field experiment(Fig.1D).These results suggest that the TaAGPS-7A-TG haplotype showed a stronger effect on total kernel starch content than TaAGPS-7A-CT.

3.5.TaAGPS-7A-TG has higher enzyme activity

To investigate the reason for the TKW difference between the two haplotypes of TaAGPS-7A,the expression levels of TaAGPS-7A in the developing endosperm of NILs were measured by qRT-PCR.Because the highest expression of TaAGPS-7A in developing endosperm was observed at 10 DPA,the endosperms at 10 DPA of four pairs of NILs(LY1,LY2,JY1,and JY2)planted in Zhaoxian were assayed.There were no significant differences in the expression of TaAGPS-7A between the same pairs of NILs(Fig.S4A).Nor did the total expression levels of TaAGPS differ between the same pair of NILs(Fig.S4B).

To determine whether the 218th amino acid change of Ser to Ala between TaAGPS-7A-CT and TaAGPS-7A-TG affected enzyme activity and TKW,both in vitro and in vivo assays were performed.In the in vitro experiment,the purified TaAGPS-7A-TG and TaAGPS-7A-CT enzymes were separately obtained from the prokaryotic expression system(Fig.2A,B)and were further confirmed by mass spectrometry(Table S5).The result of enzyme activity assay showed that the activity of AGPase produced by E.coli with TaAGPS-7A-TG was higher(P＜0.01)than that of the protein produced by E.coli with TaAGPS-7A-CT(Fig.2C).To further verify the effects of the two haplotypes on the protein levels of AGPase in vivo,the activities of AGPase in endosperms at 10 DPA of four pairs of NILs(LY1,LY2,JY1,and JY2)were measured.Our results showed that the AGPase activities of LY1,LY2,JY1 and JY2 with the haplotype TaAGPS-7A-CT were 1.55,1.41,1.90,and 3.27 mU g-1fresh weight,respectively.However,the enzyme activities of the four NILs with TaAGPS-7A-TG were respectively 1.98,3.08,2.28 and 3.69 mU g-1fresh weight,which was significantly higher than those of the corresponding TaAGPS-7A-CT NIL(Fig.2D).Taken together,these results suggest that the variation of TaAGPS-7A can influence the total AGPase activity,and the TaAGPS-7A-TG haplotype has significantly higher AGPase activities than TaAGPS-7ACT.

3.6.TaAGPS-7A-TG positively regulates TKW by increasing TaAGPS activity

To gain further insight into the role of TaAGPS-7A-TG in regulating TKW,the overexpression and RNAi transgenic lines of TaAGPS-7A-TG had been evaluated.The TKW of OE 1 and OE 2 were significantly higher than the WT,but showed significant decrease in RNAi 1 and RNAi 2 both under the Zhaoxian and Dishang field experiment(Fig.3A).Negligible changes were detected for plant height,spike number per plant,and grain number per spike(data not shown).

The expression levels of TaAGPS and the activities of AGPase in the endosperms at 10 DPA of overexpression lines(OE 1 and OE 2),RNAi lines(RNAi 1 and RNAi 2)and WT were also measured in the Zhaoxian field experiment.Relative to the WT,the expression levels of TaAGPS were increased by 2.0 and 3.5 times in OE 1 and OE 2,respectively,by qRT-PCR(Fig.3B).But the TaAGPS expression levels in the RNAi 1 and RNAi 2 declined by 95.2% and 83.6%,respectively(Fig.3B).The activities of AGPase in endosperms were increased by 34.78% in OE 1 and 76.71% in OE 2,and reduced by 79.81% in RNAi 1,60.87% in RNAi 2,compared with the WT(1.07 mU g-1fresh weight)(Fig.3C).Taken together,these findings suggest that the TaAGPS-7A-TG positively regulates TKW bread wheat by increasing the activity of AGPase.

4.Discussion

The cytosolic AGPS plays important role in starch synthesis[1,13,18,22,33–36].In bread wheat,the differences among the expression levels of the three homoeologous TaAGPS genes(Fig.S1B)suggest that the TaAGPS evolved differently during wheat polyploidization,and that TaAGPS-7D and TaAGPS-7A play greater roles in regulating grain starch synthesis and grain weight development.In this study,two new haplotypes(TaAGPS-7A-CT and TaAGPS-7A-TG)of TaAGPS-7A were identified.These two haplotypes were identified associated with TKW.The function of the TaAGPS gene was investigated via gene overexpression and silencing and the mechanism by which the two haplotype influenced TKW was verified.TaAGPS increased the yield of wheat.This yield could be increased by strengthening the expression of TaAGPS using genetic engineering.

Fig.1.TKW and total starch content analysis of TaAGPS-7A haplotypes in multiple populations.(A)TKW of TaAGPS-7A haplotypes in the germplasm panel.(B)TKW of TaAGPS-7A haplotypes in F8 RILs.(C)TKW of TaAGPS-7A haplotypes in BC1F8 NILs.Lanes 1,2,3,4,5 and 6 represent LY1,LY2,LY3,JY1,JY2 and JY3.(D)Comparison of total starch content among TaAGPS haplotypes in BC1F8 NILs.Six pairs of BC1F8 NILs from two populations,Luyuan 212/Yemaizi(LY1,LY2 and LY3)and Jingdong 8/Yemaizi(JY1,JY2 and JY3).*and**indicate differences at P＜0.05 and P＜0.01,respectively.

Fig.2.Induction and purification of TaAGPS::MBP fusion protein and comparison of AGPase activity in vitro and in vivo.(A)Induction of TaAGPS::MBP fusion protein and MBP.(B)Purification of AGPase protein.(C)Comparison of AGPase activity in vitro.(D)The activity(mU g-1 fresh weight)of AGPase in developing endosperms at 10 DPA of NILs with two haplotypes planted in Zhaoxian.The numbers to left of arrows indicate the sizes of the proteins.Lines M,1,2,3,4,5 represent respectively protein marker,the induced protein of MBP,the uninduced protein of TaAGPS::MBP with haplotype TaAGPS-7A-TG,the induced protein of TaAGPS::MBP with haplotype TaAGPS-7A-TG,the uninduced protein of TaAGPS::MBP with haplotype TaAGPS-7A-CT,and the induced protein of TaAGPS::MBP with haplotype TaAGPS-7A-CT.Values are means±SD(n=3).Single and double asterisks indicate differences at P＜0.05 and P＜0.01,respectively.

Previous studies[21,22,37,38]have shown that different positions of SNPs have different effects on phenotypes.Hou et al.[22]and our data both confirmed that the sequence variations of TaAGPS-7A might affect the wheat TKW.But what are the mechanisms of TaAGPS-7A affecting the TKW is still unclear.We found two SNPs in the promoter region and the third exon of TaAGPS-7A,incorporated into the TaAGPS-7A-TG or TaAGPS-7A-CT allele,led to differences in TKW and total starch content(Fig.1C,D).As the expression levels of TaAGPS-7A gene in the NILs with TaAGPS-7A-TG or TaAGPS-7A-CT haplotype showed no significant difference(Fig.S4A),the main reason for the differences of TKW and total starch content may be the difference in enzyme activity between the two haplotypes.It is reported that the nucleotide substitution of T5090G leading to the 218th amino acid change of Ser to Ala might affect AGPase activity[22,39,40].Our in vitro prokaryotic expression enzyme activity test also revealed that TaAGPS-7ATG showed higher enzyme activity than TaAGPS-7A-CT(Fig.2C).There were no significant differences in the expression of TaAGPS-7A between the same pair of NILs(Fig.S4A).These findings suggest that the SNP changes in the promoter region did not affect the expression of TaAGPS-7A.We further confirmed that TaAGPS-7A-TG had higher(P＜0.05)enzyme activity than TaAGPS-7A-CT using several pairs of TaAGPS-7A NILs(Fig.2D).The 218th amino acid change of TaAGPS-7A from Ser to Ala led to higher enzyme activity,which might contribute to higher TKW and total starch content.

Starch,consisting of amylose and amylopectin,is the major storage compound of mature kernel[9].Functional changes in genes acting in the starch synthesis pathway influence starch content,amylose content,and grain weight.Overexpression of TaAGPS-7A-TG produced higher AGPase activity and TKW than in the WT,and knockdown of TaAGPS reduced AGPase activity and TKW to lower levels than observed in the WT(Fig.3A,C).Mutations of GBSSI reduced amylose contents in wheat[38].TaSus1 and TaSus2 with different haplotypes were associated with TKW[41].TaBT1,responsible for unidirectional transmembrane transport of ADP-glucose,showed different effects on starch synthesis and TKW with different haplotypes[42].Polymorphisms in six starch synthesis genes were associated with TKW[21].These examples support the influence of starch synthesis-pathway genes on TKW.Identification of alleles of genes acting in this pathway and of their influence on starch quality and yield traits will be useful in future breeding.

CRediT authorship contribution statement

Xiaoling Ma:Investigation,Data curation,Funding acquisition,Validation,Writing-original draft.Xiang Ouyang:Methodology,Data curation,Writing-review&editing.Dongcheng Liu:Conceptualization,Funding acquisition,Project administration,Writingreview & editing.Aimin Zhang:Conceptualization,Funding acquisition,Project administration,Writing-review & editing.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgments

Fig.3.TKW,expression of TaAGPS,and AGPase activity changed in transgenic wheat.(A)TKW of overexpression and RNAi lines planted in Zhaoxian and Dishang changed compared with WT.(B)Expression of TaAGPS in developing endosperms at 10 days post-anthesis(DPA)of overexpression and RNAi lines planted in Zhaoxian changed with respect to WT.(C)The AGPase activity in developing endosperms at 10 days post-anthesis(DPA)of overexpression and RNAi lines planted in Zhaoxian changed in comparison with WT.Wild type(WT)was used as the negative control.Single and double asterisks indicate differences at P＜0.05 and P＜0.01,respectively.

This work was financially supported by the National Natural Science Foundation of China(31871617,32172066)and the Education Department of Hunan Province(20B615).We thank Dr.Zhonghu He,the Chinese Academy of Agricultural Sciences for providing experimental materials,and Prof.James C.Nelson,Kansas State University for language editing and proofreading of this manuscript.We are grateful to the anonymous reviewers for valuable suggestions.

Appendix A.Supplementary data

Supplementary data for this article can be found online at https://doi.org/10.1016/j.cj.2022.07.006.