A major and stable QTL for wheat spikelet number per spike validated in different genetic backgrounds

DlNG Pu-yang ,MO Zi-qiang ,TANG Hua-ping ,MU Yang ,DENG Mei ,JlANG Qian-tao ,LlU Ya-xi,CHEN Guang-deng,CHEN Guo-yue,WANG Ji-rui,Ll Wei,Ql Peng-fei,JlANG Yun-feng,KANG Hou-yang,YAN Gui-jun,WEl Yu-ming,ZHENG You-liang,LAN Xiu-jin,MA Jian

1 State Key Laboratory of Crop Gene Exploration and Utilization in Southwest China,Ministry of Science and Technology/Triticeae Research Institute,Sichuan Agricultural University,Chengdu 611130,P.R.China

2 College of Resources,Sichuan Agricultural University,Chengdu 611130,P.R.China

3 College of Agronomy,Sichuan Agricultural University,Chengdu 611130,P.R.China

4 UWA School of Agriculture and Environment and The UWA Institute of Agriculture,Faculty of Science,University of Western Australia,Crawley 6009,Australia

Abstract The spikelet number per spike (SNS) contributes greatly to grain yield in wheat. Identifying various genes that control wheat SNS is vital for yield improvement. This study used a recombinant inbred line population genotyped by the Wheat55K single-nucleotide polymorphism array to identify two major and stably expressed quantitative trait loci (QTLs) for SNS. One of them (QSns.sau-2SY-2D.1) was reported previously,while the other (QSns.sau-2SY-7A) was newly detected and further analyzed in this study. QSns.sau-2SY-7A had a high LOD value ranging from 4.46 to 16.00 and explained 10.21-40.78%of the phenotypic variances. QSns.sau-2SY-7A was flanked by the markers AX-110518554 and AX-110094527 in a 4.75-cM interval on chromosome arm 7AL. The contributions and interactions of both major QTLs were further analyzed and discussed. The effect of QSns.sau-2SY-7A was successfully validated by developing a tightly linked kompetitive allele specific PCR marker in an F2:3 population and a panel of 101 high-generation breeding wheat lines. Furthermore,several genes including the previously reported WHEAT ORTHOLOG OF APO1 (WAPO1),an ortholog of the rice gene ABERRANT PANICLE ORGANIZATION 1 (APO1) related to SNS,were predicted in the interval of QSns.sau-2SY-7A. In summary,these results revealed the genetic basis of the multi-spikelet genotype of wheat line 20828 and will facilitate subsequent fine mapping and breeding utilization of the major QTLs.

Keywords:yield potential,QTL detection,QTL validation,predicted genes,tightly linked KASP marker

1.lntroduction

Common wheat (Triticum aestivumL.) is a major crop that is widely cultivated all over the world. More than 20% of the human-consumed calories are provided by wheat (FAO 2015). To meet the increasing demands of a growing human population,it is essential to substantially increase the wheat yield. Therefore,identifying ways to increase the grain yield is a critical and urgent requirement for wheat breeders. Grain yield is a multifactorial trait that is determined by grain number per spike (GNS) and grain weight,the two primary yield components that are typically negatively correlated (Slafer and Miralles 1993;Sadras 2007;Bustoset al.2013;Garcíaet al.2013). GNS can be further assessedviathe grain number per spikelet and the spikelet number per spike (SNS).SNS is reportedly affected by various factors,including plant spacing,the nitrogen nutrition level,day length,temperature,light intensity,and the duration of spike development (Maet al.2019a). However,grain weight is a part of the domestication syndrome in cereal crops and is more stably inherited than final yield (Guanet al.2018). Thus,this different heritabilities of SNS and grain weight may be the reason that fewer stable loci related to SNS have been identified compared with those related to the thousand-grain weight (TGW).

The identification,utilization and pyramiding of various genes for SNS from diverse wheat germplasm resources are essential because of the potential for SNS to increase wheat yield. So far,a number of genes related to wheat SNS have been identified. Prominent examples areTEOSINTE BRANCHED1(TB1) (Dixonet al.2018),FRIZZY PANICLE(WFZP) (Dobrovolskayaet al.2015),PHOTOPERIOD RESPONSElocus (Ppd-1) (Bodenet al.2015),FLOWERING LOCUST2(FT2) (Shawet al.2018),WAPO1(TaAPO-A1) (Kuzayet al.2019;Muqaddasiet al.2019),Q(Greenwoodet al.2017),andGrain Number Increase 1(GNI1) (Sakumaet al.2019).

Because of the complexity of the wheat genome,studies on SNS in wheat have primarily focused on genetic mapping and marker development. To date,quantitative trait loci (QTLs) of SNS have been identified on nearly 21 chromosomes in wheat (Liet al.2007;Echeverry-Solarteet al.2015;Liet al.2015;Luoet al.2016;Kuzayet al.2019;Woldeet al.2019;Yaoet al.2019). However,major QTLs for SNS that have been stably detected and validated across multiple environments,have been reported in a limited number of studies. Therefore,the identification,utilization and pyramiding of various genes for SNS from diverse wheat germplasm resources are essential.

With the development of molecular markers,increasing QTL analysis in wheat have been published. To date,multiple methods have been developed for QTL analysis(Yanget al.2021;Griffithset al.2022),however,these methods all have certain errors to varying degrees (Verbylaet al.2012). In addition to the widely reported false positives of QTLs (Bernardo 2004),there are also numerous false negatives in QTL mapping (Belknapet al.1996;Ramasamyet al.2013). In this study,a false negative QTL (QSns.sau-2SY-7A) was detected and further analyzed.

The winter wheat line 20828 has particularly desirable agronomic traits,and it has been widely used for breeding programs over the last decade. Using a recombinant inbred line (RIL) population,a major and stable QTL (QSns.sau-2D)for SNS was previously mapped on chromosome 2DS,with a positive allele from the line 20828 (Maet al.2019a). Here,another major and stable QTL for SNS was identified using a different RIL population,and the positive allele at this QTL was also from 20828 (Appendix A). This QTL was further validated in different genetic backgrounds. The contributions and interactions of both major QTLs identified here and in the above-mentioned study (Maet al.2019a)(i.e.,QSns.sau-2D) were analyzed and discussed.

2.Materias and methods

2.1.Plant materials

The winter wheat line 20828 (G214-5/3/Chuanyu 19//Lang 9247/50788) was used as the common parent of two populations. One is a RIL population derived from the cross between 20828 and SY95-71 (2SY,194 lines) (Maet al.2019a;Liuet al.2020),and the other one is an F2:3population from 20828 and Chuanmai 60 (CM60,2CM,205 F2:3lines) (Maet al.2019b,2020) (Appendix A). The 2SY population was used for QTL mapping,while the 2CM population was used for QTL validation. A panel of 101 high-generation breeding lines with a background of 20828 was also used for QTL validation in this study (Appendix B).Notably,the 2SY population originally consisted of 194 lines(Maet al.2019a),while 126 of them with two parents used in the present study were randomly selected for the Wheat55K SNP array genotyping as described previously (Liuet al.2020). The SNS of 2SY RILs population was measured in seven environments:Chongzhou (103°38′E,30°32′N),Wenjiang (103°51′E,30°43′N) and Ya’an (103°0′E,29°58′N)in 2017 and 2018 (17CZ,17WJ,17YA,18CZ,18WJ,and 18YA) and Khulna,in Bangladesh (89°33′E,22°49′N) in 2018 (18KB). All trials used a random block design in each field. Each line was planted in a single 1.5 m row with 30 cm between rows;15 seeds were planted in each row.The data of anthesis date (AD),productive tiller number(PTN),spike length (SL) and TGW were measured in our previous studies (Liet al.2020;Liuet al.2020) and further analyzed here. For plant height (PH) and SNS,we added the data from the environment of 2018KB for subsequent analysis based on Liuet al.(2020),where major QTLs for SNS were roughly identified only for detecting correlations between SNS and PTN. For all of the traits analyzed in this study in a given environment,the mean value of at least three plants(i.e.,three biological repeats) in each line was calculated (Liuet al.2021). The mean values for all of the environments were used to estimate the best linear unbiased prediction(BLUP) value for further analysis. The detailed environments in which these traits were measured are listed in Appendix C. All field experiments were well irrigated and managed in accordance with local standard practices.

2.2.Phenotypic data analysis

Phenotypic data analyses,including BLUP for target traits from different environments,frequency distribution,phenotypic variation,broad-sense heritability (H2) across environments,Pearson’s correlations and significant differences within a given trait of parents and between different RIL groups,were conducted according to a previous report (Maet al.2019a). The BLUP dataset was obtained to calculate correlations between SNS with SL,TGW,AD,PH,and PTN in the 2SY population. Lines that carry different alleles at the major QTLs from the 2SY population were screened based on the genotypes of their flanking markers. Thus,the identified lines were further compared for the above seven traits using Student’st-test (with aP＜0.05 cut off).

2.3.QTL mapping

A linkage map for the 2SY RIL population was constructed using the Wheat55K SNP array according to a previously reported study (Liuet al.2020). Specifically,in the entire Wheat55K SNP array,only poly high resolution single nucleotide polymorphism (SNP) with minor allele frequency(＜0.3) among the 2SY RILs population was retained.Redundant marker identification was performed by the BIN function of IciMapping 4.1. JoinMap 4.0 was used to construct linkage maps by these markers using the Kosambi mapping function with LOD≥3. MapChart 2.2 was used to draw genetic maps. According to the results of Liuet al.(2020),the genetic map of the 2SY population contained 38 linkage groups with a total genetic distance of 4 273.03 cM and the average interval between adjacent markers was 1.69 cM. QTL detection (including QTL×environment (QE) analysis) was conducted according to a previous study (Maet al.2019a).Specifically,QTL mapping and epistasis interaction analysis were performed by IciMapping 4.1 based on the biparental population (BIP) module of inclusive composite interval mapping (ICIM) and LOD values ≥2.5 and 5,respectively.The QE analysis was based on the multi-environment trials(MET) with pre-adjusted parameters:Step=1 cM,PIN=0.001 and LOD=4.QTLs detected in at least four environments(including the BLUP dataset) were treated as stable. The QTLs explained more than 10% of phenotypic variation were treated as major QTL,and those with common flanking markers or less than 1 cM apart were treated as a single QTL.

The physical positions and predicted genes within the interval between the flanking markers (referring to‘Chinese Spring’ or ‘CS’ (IWGSC 2018) and the wild emmer(T.turgidumssp.dicoccoides) genome) (Avniet al.2017)were obtained following previous studies (Liuet al.2018;Maet al.2019a). Specifically,the flanking markers of the QTLs were blasted against the ‘CS’ and the wild emmer genome to identify their physical positions. Genes between the flanking markers were retrieved from CDS sequences.The annotation and function of a given gene were analyzedviaUniProt (http://www.unipr ot.org/).

2.4.Validation of major QTLs and comparison of SNS genes on 7A

According to the preliminary QTL mapping results,the flanking markers were converted to flexible Kompetitive Allele Specific PCR (KASP) markers (Appendix D) to validate the genetic effects of the major QTLs with different genetic backgrounds following a previously described method (Maet al.2019a). The PCR reaction mixtures(10 μL) contained 50 ng of genomic DNA,0.2 μmol L-1of each primer and 5 μL of 2×Master Mix (LGC Science Ltd.,Shanghai,China). The cycling parameters were 94°C for 15 min,10 cycles of 94°C for 20 s,61°C for 1 min followed by 26 cycles of 94°C for 20 s,55°C for 1 min,and a final storage at 37°C for 1 min.

For validation,80 lines were randomly selected from the validation population (2CM). According to marker profiles,the 80 lines from the 2CM population and the 101 breeding lines were divided into two groups:one with homozygous alleles from 20828 (designated as A) and the other with homozygous alleles from alternative parents (designated as B). The differences in SNS between the two groups in each population were evaluated by Student’st-test (P＜0.05).

2.5.Sequence isolation of WHEAT ORTHOLOG OF APO1 (WAPO1)

WAPO1(TaAPO-A1),an ortholog of the rice geneABERRANT PANICLE ORGANIZATION 1(APO1),was previously reported to be involved in the regulation of SNS(Kuzayet al.2019;Muqaddasiet al.2019). Here,we isolated sequences ofWAPO1from wheat lines 20828 and SY95-71 for analysis. Genomic DNA was extracted from leaves at tillering stage by the cetyl trimethyl ammonium bromide (CTAB) method (Murray and Thompson 1980).A set of primers was designed to amplify the genomic sequences ofWAPO1based on the reference sequences of ‘CS’ (Appendix D). PCR amplification was performed with the Applied Biosystems Veriti (ThermoFisher Scientific Co.,Ltd.). PCR reaction mixtures (50 μL) contained 100 ng of genomic DNA,10 mmol L-1of each dNTP,20 μmol L-1of each primer,and 1 μL of Phanta?Max Super-Fidelity DNA Polymerase (Vazyme Biotech Co.,Ltd.) with 25 μL of 2×Phanta Max Buffer. The cycling parameters were 95°C for 3 min,35 cycles of 95°C for 15 s,60°C for 15 s,72°C for 2 min,and a final storage at 72°C for 5 min.Amplification products were separated on 1.2% agarose gels,and a FastPure?Gel DNA Extraction Mini Kit (Vazyme Biotech Co.,Ltd.) was used to purify the fragments. The recovered PCR products were used for gene cloning and then sent to Tsingke Biological Technology Co.,Ltd.,China for sequencing. At least nine independent clones of each amplicon were sequenced in both directions.

3.Results

3.1.Phenotypic analyses

Among the seven environments,as well as in the combined analysis by the BLUP dataset,significant differences of SNS between 20828 and CN16 were found at the 0.01 level. As shown in Appendices E and F,the SNS of 20828 ranged from 25.67 to 30.50,that of SY95-71 ranged from 19.00 to 22.00 and the 2SY RIL population ranged from 15.00 to 33.75. The estimatedH2value of SNS was 0.75 (Appendix E),so it was assumed to be affected by genetic factors.In the seven environments,Pearson’s correlation for SNS ranged from 0.17 to 0.77 (Table 1).

3.2.QTL for SNS

QTL analysis of the 128 RILs identified 10 significant QTL on six chromosomes. They were distributed on chromosomes 1D (2 QTLs),2A (1),2B (2),2D (2),5A (2),and 7A (Table 2).QSns.sau-2SY-7A,a major QTL on 7A,was verified in six environments and BLUP (except 2018KB) and explained 10.21-40.78% of the phenotypic variance with LOD values of 4.46-16.00. The positive allele atQSns.sau-2SY-7Awas inherited from the 20828 genotype (Table 2).Three flanking markers (AX-110094527,AX-110518554andAX-110442528) were converted to KASP markers(KASP-AX-110094527,KASP-AX-110518554andKASPAX-110442528) by the initial QTL mapping results. These results indicated thatQSns.sau-2SY-7Awas located within a 4.75-cM interval on chromosome 7AL between markersAX-110518554andAX-110094527(Figs.1 and 2). In addition,QSns.sau-2SY-2D.1was detected in four environments,explaining 11.05-15.30% of the phenotypic variance with LOD values of 2.90-6.01. The remaining QTLs were only detected in a single environment (Table 2).Thirty QTLs were detected in the QE interaction analysis.With the exceptions ofQSns.sau-2SY-7AandQSns.sau-2SY-2D.1which were further demonstrated to be major and stable loci,most of these QTLs explained low phenotypic variance and had low LOD scores (Appendix G). In addition,epistasis effects betweenQSns.sau-2SY-7AandQSns.sau-2SY-2D.1were also analyzed,and the results showed that there was no epistasis effect between them (Appendix H).

Table 1 Correlation coefficients for the spikelet number per spike (SNS) in the 2SY population evaluated in different environments

Table 2 Quantitative trait loci for spikelet number per spike (SNS) identified from different environments1)

A previous study (Maet al.2019a) identified a QTL(QSns.sau-2D) that controlled SNS in the same interval as theQSns.sau-2SY-2D.1detected in this study. Since the positive alleles of both were from 20828,this study assumed that they represent the same locus. In a previous study(Maet al.2019a),we employed thePpd-D1-derived InDel marker (Bealeset al.2007) to re-mapQSns.sau-2D. The results showed thatQSns.sau-2Dwas not linked toPpd-D1with a genetic distance of 5.4-cM (Maet al.2019a).

3.3.Correlation analysis

Phenotypic correlations between SNS and other yieldrelated traits are listed in Appendix H. Positive correlation(P＜0.01) was only observed between SNS and SL. No significant correlation was detected between SNS and either AD,PH,TGW,or PTN (Appendix I). According to the marker profiles,the 2SY RILs population was divided into two groups,one containing homozygous alleles from 20828 and the other containing homozygous alleles from SY95-71 at the locus ofQSns.sau-2SY-7A. Comparison analysis showed that lines with homozygous alleles from 20828 had significantly longer AD,lower TGW and longer SL than those from SY95-71 (P＜0.01). This suggests thatQSns.sau-2SY-7Asignificantly (P＜0.01) affected AD,TGW and SL (Fig.3),while no significant differences were detected for PH and PTN.

Fig.3 Student’s t-test for the two groups of lines carrying the allele from either 20828 or SY95-71 at QSns.sau-2SY-7A from the 2SY recombinant inbred line (RIL) population for anthesis date (AD,A),plant height (PH,B),thousand-grain weight (TGW,C),productive tiller number (PTN,D),and spike length (SL,E).+and -represent lines with and without the positive alleles of QSns.sau-2SY-7A based on the flanking markers KASP-AX-110094527. The boxplots consist of the maximum,minimum,median and two quartiles. The two quartiles are connected to draw the box;then the maximum and minimum are connected with the box,while the median is in the middle of the box. ＊＊,significance at P＜0.01,above the significance levels is the percentage difference between the two groups of lines.

To further assess possible genetic correlations between SNS and other traits including AD,TGW and SL,QTLs for them were detected. The results showed that no QTL clusters for SNS,TGW,SL,and AD were detected on chromosome 7A,indicating thatQSns.sau-2SY-7Ais mostlikely independent of AD and will have great potential in wheat breeding. However,QTL clusters of SNS,AD and SL were detected on chromosome 2D,except for TGW(Appendix J).

Fig.1 Linkage map of QSns.sau-2SY-7A and its effect. A,linkage map of QSns.sau-2SY-7A for SNS. B,effect of QSns.sau-2SY-7A shown as box plots calculated after dividing the 2SY population (a recombinant inbred line (RIL) population derived from the cross between 20828 and SY95-71) into two groups according to the flanking markers.+and -represent lines with and without the positive alleles of QSns.sau-2SY-7A based on the flanking markers KASP-AX-110094527. Phenotypes of lines carrying different alleles at QSns.sau-2SY-7A based on 20828 and SY95-71. 17 and 18,2017 and 2018,respectively. CZ,Chongzhou;WJ,Wenjiang;YA,Ya’an,Sichuan Province,China;KB,Khulna,Bangladesh;BLUP,best linear unbiased prediction. SNS,spikelet number per spike. ＊＊,significant at P＜0.01,the differences between the two groups of lines are shown below the environment names.

3.4.Validation of QSns.sau-2SY-7A in different genetic backgrounds

QSns.sau-2SY-2D.1(i.e.,QSns.sau-2D) was previously validated(Maet al.2019a) and thus,the present study only validatedQSns.sau-2SY-7A. The flanking SNP marker,AX-110094527was successfully converted to the KASP markerKASPAX-110094527to evaluate the effects ofQSns.sau-2SY-7Ain the 2CM population. According to the marker profiles,lines with homozygous alleles from 20828 had significantly higher (P＜0.01) SNS than those from CM60,and the difference was 5.50% in the 2CM population (Fig.4).

Fig.4 Validation of QSns.sau-2SY-7A in an F2:3 population derived from 20828 and Chuanmai 60 population (2CM). A,fluorescence PCR genotyping results in the 2CM population by the KASP marker KASP-AX-110094527. Blue boxes represent lines carrying the alternative alleles from CM60 (HEX fluorescence);orange frames represent lines carrying the allele of 20828 (FAM fluorescence);black spots represent blank controls. B,effects of QSns.sau-2SY-7A in the population of 2CM. 18CZ,2018 Chongzhou,China.The boxplots consist of the maximum,minimum,median and two quartiles. The two quartiles are connected to draw the box;then the maximum and minimum are connected with the box,while the median is in the middle of the box. ＊＊,significance at P＜0.01,the difference between the two groups is given below the environment names.

3.5.Effects of QSns.sau-2SY-7A and QSns.sau-2SY-2D.1 on SNS

The positive alleles of bothQSns.sau-2SY-7AandQSns.sau-2SY-2D.1were derived from 20828. Their effects on SNS were further evaluated by Student’st-test in 2SY populations and the 101 high-generation breeding lines which have the background of 20828 (Fig.5). Compared with lines without any of the positive alleles that increase SNS,lines with the increased allele fromQSns.sau-2SY-2D.1significantly increased SNS by 4.98 and 9.12% in the 2SY population and breeding lines,respectively. Lines with the increased allele fromQSns.sau-2SY-7Asignificantly increased SNS by 7.25 and 11.71% in the 2SY population and breeding lines,respectively. Lines with a combination ofQSns.sau-2SY-7AandQSns.sau-2SY-2D.1significantly increased SNS by up to 15.23 and 18.99% in the 2SY and breeding lines,respectively. In addition,lines with a combination of increased alleles ofQSns.sau-2SY-7AandQSns.sau-2SY-2D.1significantly increased SNS in comparison with those with only the increased allele of eitherQSns.sau-2SY-7A(8.47%) orQSns.sau-2SY-2D.1(10.84%) in the 2SY population,andQSns.sau-2SY-7A(9.05%) orQSns.sau-2SY-2D.1(6.52%) in breeding lines.

Fig.5 The effects of QSns.sau-2SY-7A and QSns.sau-2SY-2D.1 on spikelet number per spike (SNS) in the 2SY population (A) and breeding lines (B).+and -,lines with and without the positive alleles of the target quantitative trait loci (QTL) based on the flanking markers (QSns.sau-2SY-7A and QSns.sau-2SY-2D.1 in 2SY),respectively,or the developed KASP markers KASP-AX-110094527 for QSns.sau-2SY-7A and KASP-AX-94721936 for QSns.sau-2SY-2D.1 in breeding lines of the corresponding QTL,respectively.The boxplots consist of the maximum,minimum,median and two quartiles. The two quartiles are connected to draw the box;then the maximum and minimum are connected with the box,while the median is in the middle of the box. ＊,significance at P＜0.05;＊＊,significance at P＜0.01,differences between the two groups are labelled above the significance levels.

3.6.Sequence analysis of WAPO1

AsWAPO1was reported to be a candidate gene for SNS on chromosome 7AL (Kuzayet al.2019;Muqaddasiet al.2019),its orthologous sequences were isolated from parental lines 20828 and SY95-71 for analysis. Two SNPs were identified in 140 bp (G-T) and 1 284 bp (G-A) in the coding region ofWAPO1,respectively. In addition,these two SNPs caused amino acid changes in 47aa (C-F) and 384aa (D-F).

4.Discussion

4.1.QSns.sau-2SY-7A is a major and stable QTL

This study identified a major QTL for SNS in wheat. This QTL was located on a 4.75-cM interval and physically mapped between 673.87 and 677.70 Mb on chromosome 7AL (Fig.2). Numerous studies have reported QTLs for SNS on chromosome 7A. To identify whether the QTL reported in this study overlaps with previously identified QTLs,these QTLs were physically positioned on chromosome 7A in CS(Appendix K). A comparison of physical maps for these QTLs showed thatQSns.sau-2SY-7A,detected in this study,overlapped withQTsn.cau-7A.3(Zhaiet al.2016),QFsn.cau-7A.1(Zhaiet al.2016),QFss-7A(Xuet al.2015),QTss-7A(Xuet al.2015),andQSns.sau-QZ-7A(Luoet al.2016)(Appendix K). In addition,the geneWAPO1(TaAPO-A1),an ortholog of the rice geneAPO1,is located in the interval ofQSns.sau-2SY-7A. This gene was reported (Muqaddasiet al.2019) to be related to the spikelet number and is the most likely candidate gene for the 7AL SNS QTL reported by Kuzayet al.(2019). The overlapping ofQSns.sau-2SY-7Adetected by the present study,with previously identified QTLs,suggests thatQSns.sau-2SY-7Amay be allelic toQTsn.cau-7A.3(Zhaiet al.2016),QFsn.cau-7A.1(Zhaiet al.2016),QFss-7A(Xuet al.2015),QTss-7A(Xuet al.2015),QSns.sau-QZ-7A(Luoet al.2016),andWAPO1(Kuzayet al.2019;Muqaddasiet al.2019). Given thatWAPO1has been reported (Kuzayet al.2019;Muqaddasiet al.2019),we also isolated sequences ofWAPO1from wheat lines 20828 and SY95-71. Genomic sequence alignment showed that there were two SNPs(G140T and G1284A) in the coding region (Appendix L) with amino acid changes (C47F and D384N),which was consistent with the results from Muqaddasiet al.(2019). According to the previous study (Muqaddasiet al.2019),wheat genotypes with a T base at 140 bp and a G base in 1 284 bp had significantly higher numbers of SNS. Therefore,we deduce thatWAOP1is also likely the candidate gene forQSns.sau-2SY-7A. These results also suggested that chromosome 7A in wheat plays an important role in SNS development.

Fig.2 The predicted genes from the physical interval of QSns.sau-2SY-7A. The physical map of 7A of ‘Chinese Spring’ (CS,A)and wild emmer (B). Corresponding orthologs are indicated by dotted lines.

4.2.Contributions of both QSns.sau-2SY-7A and QSns.sau-2SY-2D.1 to SNS of 20828

By employing the 2SY (20828/SY95-71) population,the present study detected two major and stably expressed QTLs:QSns.sau-2SY-7AandQSns.sau-2SY-2D.1.Their positive alleles were contributed by 20828,while onlyQSns.sau-2Dwas detected at multiple environments in a previous study that used the 2CN (20828/CN16) population(Maet al.2019a). Since different male parents were used in both RIL populations,the expression ofQSns.sau-2SY-7Ais likely inhibited in specific genetic backgrounds or masked by the effect fromQSns.sau-2SY-2D.1as reported in previous studies (Jianget al.2007;Lillemoet al.2013;Ollieret al.2020;Ruanet al.2020). According to previous studies (Belknapet al.1996;Ramasamyet al.2013),QSns.sau-2SY-7Ashould be defined as a false negative QTL.This suggests that several different populations should be developed when focusing on a given wheat genotype.

Previous studies identified pyramiding of elite genes as an effective method to improve yield (Fanet al.2015). The results of the present study showed that the combination ofQSns.sau-2SY-7AandQSns.sau-2SY-2D.1had the strongest effect on SNS,followed byQSns.sau-2SY-2D.1andQSns.sau-2SY-7A(Fig.5). This indicates the potential of pyramiding both QTLs for regulating SNS and thus possibly improving wheat yield.

4.3.Predicted genes in the interval of the major QTL QSns.sau-2SY-7A

Genes in the interval ofQSns.sau-2SY-2D.1(i.e.,QSns.sau-2D) have been previously reported (Maet al.2019a).Thus,the present study focused onQSns.sau-2SY-7Aonly.QSns.sau-2SY-7Awas mapped between 669.18 and 673.06 Mb on 7AL of wild emmer (Fig.2). There were 55 and 75 predicted genes in the intervals of CS and wild emmer,respectively,26 of which were shared (Appendix M). Based on functional annotations,this study identified several genes that are associated with plant growth and development. For instance,TraesCS7A01G484300.1encodes a receptor-like kinase and was reported to control organ growth and flower development by promoting cell proliferation inArabidopsis(Shpaket al.2004). Thus,it may affect the formation of SNS in wheat.TraesCS7A01G483100.1encodes the MYB family transcription factorALTERED PHLOEM DEVELOPMENT(APL).APLis involved in the promotion of phloem development and the repression of xylem development,and thus affects growth and development of plants (Bonkeet al.2003). Interestingly,this study identified three common genes:TraesCS7A01G482400.1,TraesCS7A01G482800.1andTraesCS7A01G482900.1(Appendix M). All three genes encode a cortical cell-delineating protein in this interval,which has been reported to be involved in the spatial control of cell expansion and cell division and is required for plant growth and morphogenesis (Fuet al.2009). In summary,genes involved in plant growth and development may provide further clues for the fine mapping and identification of candidate genes underlying this major QTL in future work.

4.4.Relationships among SNS and yield-related traits

Phenotypic correlations and difference analyses among these spike-related traits identified strong genetic associations of SNS and SL (Appendix I;Fig.3). According to previous studies (Cuiet al.2012;Kuzayet al.2019;Woldeet al.2019),it is likely that a single allele may simultaneously increase both SL and SNS. Significant differences were detected in TGW and AD between lines with different alleles atQSns.sau-2SY-7A(Fig.3). For TGW,this result was in accordance with the finding that an increase in SNS may decrease grain weight,given that a single grain will assimilate fewer nutrients with increased SNS (physiological reasoning) (Pinthus and Millet 1978). Significant differences in AD (Fig.3) validated the findings of previous studies,which reported that late-flowering genotypes may have relatively longer SL with spikelets (Shawet al.2013;Dixonet al.2018;Kuzayet al.2019). These results indicate that the duration from transitions of the spike meristem to the terminal spikelet plays a pivotal role in the increase of SNS (Rahman and Wilson 1977;Guoet al.2018;Woldeet al.2019).

5.Conclusion

In this study,the major and stably expressed QTL for SNSQSns.sau-2SY-7Awas identified and further confirmed in different genetic populations by a developed molecular marker. The other major QTLQSns.sau-2D,which was previously detected,was also identified in this study. The contributions and interactions of these two identified major QTLs were further analyzed and discussed. Predicted genes in the interval ofQSns.sau-2SY-7Awill be valuable for the fine mapping and cloning of candidate genes.The results will also help us to understand the genetic relationships between SNS and the yield potential of crops.The major QTL for SNS identified herein and the developed KASP marker may facilitate the utilization of the wheat line 20828 for breeding.

Acknowledgements

This work was supported by the projects from the Applied Basic Research Programs of Science and Technology Department of Sichuan Province,China (2020YJ0140 and 2021YJ0503),the International Science and Technology Cooperation and Exchanges Program of Science and Technology Department of Sichuan Province,China(2021YFH0083 and 2022YFH0053),the National Natural Science Foundation of China (31971937 and 31970243),and the Key Projects of Scientific and Technological Activities for Overseas Students of Sichuan Province,China. The funding bodies did not play a role in the design of the study,in the collection,analysis,and interpretation of data or in writing the manuscript. We thank MogoEdit Bianji Company (http://www.mogoedit.com/) for editing the English text of this manuscript. We thank the anonymous referees for critically reading and revising this manuscript.

Declaration of competing interest

The authors declare that they have no conflict of interest.

Appendicesassociated with this paper are available on http://www.ChinaAgriSci.com/V2/En/appendix.htm