Co-location of QTL for Sclerotinia stem rot resistance and flowering time in Brassica napus

Jin Wu,Peipei Chen,Qing Zho,Gungqin Ci,Yue Hu,Yng Xing,Qingyong Yng,Youping Wng,Yongming Zhou,*

a Jiangsu Provincial Key Laboratory of Crop Genetics and Physiology,Yangzhou University,Yangzhou 225009,Jiangsu,China

b National Key Laboratory of Crop Genetic Improvement,Huazhong Agricultural University,Wuhan 430070,Hubei,China

c Agricultural Bioinformatics Key Laboratory of Hubei Province,College of Informatics,Huazhong Agricultural University,Wuhan 430070,Hubei,China

d Guizhou Rapeseed Institute,Guizhou Academy of Agricultural Sciences,Guiyang 550008,Guizhou,China

Keywords:Sclerotinia stem rot Brassica napus QTL mapping Flowering time SNP array

A B S T R A C T

1.Introduction

Sclerotinia sclerotiorum is a cosmopolitan fungal pathogen with a broad host range(＞400 plant species),including virtually all dicotyledonous and some monocotyledonous plants[1].This pathogen causes Sclerotinia stem rot(SSR)in rapeseed(Brassica napus)in all major growing regions,including China,Canada,India,Germany,France and Australia.In China,Sclerotinia causes annual yield losses of 10%-20%and as high as 80%in severely infected fields.Rapeseed plants infected by S.sclerotiorum during flowering and seed filling suffer not only a significant loss of seed yield but also a reduction in oil content and changes in fatty acid profile,thus resulting in inferior rapeseed quality[2,3].

Cultivation of resistant cultivars is a preferred option for controlling SSR in B.napus.However,no highly resistant or immune germplasm in B.napus and its closely related species has been identified thus far.In B.napus,SSR resistance is a trait with very complex genetic basis determined by multiple minor quantitative trait loci(QTL)[4].A considerable number of SSR resistance QTL have been identified through QTL mapping and genome-wide association studies(GWAS)in B.napus[4-9].Unfortunately,no fine mapping or map-based cloning of SSR resistance QTL has been reported,which limits the utilization of QTL in breeding and further understanding of the molecular mechanism of SSR resistance in rapeseed.This dilemma may be attributable to the difficulty of identifying the resistance phenotype in QTL fine-mapping since the complexity of plant-microbe-environment interactions and a large segregating population must be considered.Hence,exploring the influencing factors of SSR resistance in B.napus,such as environmental factors,plant growth and development,will help us accurately identify resistance based on plant immune systems.

In the lifecycle of flowering plants,flowering is a major developmental event in which plants switch from the vegetative growth phase to reproductive growth[10].In B.napus,flowering time(FT)is a complex quantitative trait controlled by multiple genes[11,12].QTL mapping and GWAS of FT have greatly enhanced our understanding of the genetic architecture of this complex trait in B.napus[8,10-15].Dozens of FT QTL have been detected,and these QTL are mainly distributed on chromosomes A2,A3,A10,and C2.Two genes,BnFLC.A2 and BnFLC.C2,for QTL mapped on A2 and C2 have been cloned and shown to have redundant functions in controlling rapeseed flowering by mediating the vernalization response[16].

Flowering is also a critical developmental stage in the lifecycle of flowering plants that can be vulnerable to environmental stresses,such as pathogen infection,drought,heat,and salinity[17].In seed-producing plants,exposure to stress during flowering can cause severe yield losses[17].For example,few or no seeds are produced if S.sclerotiorum infection occurs during the early flowering stage of rapeseed plants,while reductions in yield are smaller if infection occurs during the late flowering stage or mature stage.Furthermore,previous QTL analyses of some crop plants(such as rice and maize)have found significant associations between disease resistance and FT[18-21].Significant negative corrections between SSR resistance and FT were also found in B.oleracea[22]and B.napus[6,8].Therefore,there are conflicts between early maturation and SSR resistance in rapeseed breeding.The genetic basis and molecular mechanisms underlying these associations between disease resistance and FT remain elusive.

The current study aimed to understand the link between FT and SSR resistance in B.napus,which is crucial for breeding cultivars with early maturation and SSR resistance and will also benefit map-based cloning of SSR resistance QTL.First,we explored a natural population containing 521 rapeseed inbred lines to investigate the correlation between FT and SSR resistance.Then,the association was further confirmed using a double haploid(DH)population with high FT variation.Moreover,we identified the chromosomal regions affecting both FT and SSR resistance after constructing a high-density single nucleotide polymorphism(SNP)-based genetic map in the DH population.Finally,we evaluated the extent of SSR resistance QTL colocalized with FT QTL by comparing QTL of the two traits identified in the present and previous studies via arrangement onto the physical map of B.napus.

2.Materials and methods

2.1.Plant materials and phenotypic analysis

A total of 521 rapeseed inbred lines were employed to analyze the correlation between FT and SSR resistance.The accessions were collected from the major breeding institutes across China,and detailed information of the accessions was described previously[23].The 521 rapeseed inbred lines were grown in the 2012-2013 season.The experiment was carried out in a randomized block design with two replications.Each line was grown in two rows,and each row harbored 10-12 plants.

A biparental segregation population consisting of 150 individual DH lines named the ZD-DH population was used in this study.The population was developed from microspore culture of F1buds of a cross between two B.napus inbred lines(ZP1×D126).The DH population,along with their parental lines,was grown in two consecutive growing seasons in 2013-2014 and 2014-2015.The field experiments for the DH population were randomized complete block designs with three replications.Each DH line was planted in one row harboring 10-12 plants.

All materials were grown on the experimental farm of Huazhong Agriculture University,Wuhan,China,in the winter-type oilseed rape growing season,with 21 cm between plants within each row and 30 cm between rows.Field management followed normal breeding practices.

2.2.Phenotypic measurements and statistical analysis

The S.sclerotiorum isolate SS-1 was cultured on potato dextrose agar(PDA)containing 25.0%potato,2.5%dextrose and 1.5%agar(pH 5.8)[7].The isolate was maintained at 4°C in the dark and cultured twice at 23°C in the dark prior to inoculation.Mycelial agar plugs(7 mm in diameter)were used as the inoculum,and were punched from the margin of a 2-day-old culture of S.sclerotiorum.

The SSR resistance of the 521 rapeseed inbred lines was assessed using the detached stem inoculation assay as described in our previous research[4].For each accession in each replicate,the stems of six individuals with a length of approximately 30 cm were cut for inoculation.The lesion length was measured at five days post inoculation(dpi).This assessment was conducted when most inbred lines were at the stage of flowering termination.

SSR resistance of the ZD-DH population was assessed using a stem inoculation assay as described by Wu et al.[7].Plants were inoculated in disease nursery plots.For each DH line in each replication,8-10 stems were inoculated with mycelial agar plugs at a height of 50 cm above the ground.The lesion length was measured at 7 dpi.This assessment was conducted with three replications when most DH lines were at the stage of flowering termination.

FT was recorded as the days from sowing to flowering(when the first flower had opened on 25%of the plants in the plot).Microsoft Excel software was used to perform the regression analysis(simple linear regression model)of SSR resistance(as dependent variable)to FT(as independent variable).

2.3.SNP array genotyping

Genomic DNA was extracted from young leaves of the ZD-DH lines,parental lines and an association panel using a modified CTAB method.Genotyping was performed using Brassica 60 K Illumina Infinium HD Assay SNP arrays(Illumina Inc.,San Diego,CA)according to the manufacturer's protocol(Infinium HD Assay Ultra Protocol Guide,http://www.illumina.com/)[24].Illumina GenomeStudio software(Illumina Inc.,San Diego,CA)was used to cluster and visualize all SNP array data for further analysis.Each SNP was reassessed manually to determine if any error was observed during the clustering analysis.

2.4.Construction of the linkage map and QTL mapping

The SNP array data processing for the DH population followed a bifiltering method described by Cai et al.[25].We firstly calculates the percentage of non-parental genotypes(PNPG)based on monomorphic loci in the DH population.Subsequently,the difference in PNPG of single-locus SNPs(Simple SNP and sHemi-SNP)and multi-loci SNPs(mHemi-SNP and Pseudo-simple SNP)was compared to filter multi-loci SNPs among the SNP loci.This procedure utilized both monomorphic and polymorphic SNPs in the DH population and could effectively distinguish the mHemi-SNPs and Pseudo-simple SNPs that resulted from superposition of the signals from multiple SNPs[25].Markers with the same genotypes in the DH population were considered redundant SNPs.To simplify the subsequent analysis,we used only nonredundant SNPs(nrSNPs)for linkage map construction with MAPMAKER/EXP 3.0[26]and MSTmap[27]software as described by Cai et al.[28].The MAPMAKER/EXP 3.0 software can process no more than 101 markers for one group,while the MSTmap software can process＞10,000 markers at one time.So,we first employed MSTmap to process all the loci and grouped the loci at a LOD(likelihood of odds)score of 5.0.Marker orders of each group were then calculated by finding the minimum spanning tree of a graph based on pairwise recombination frequencies[27].At the same time,each group assessed by MSTmap was analyzed again through MAPMAKER/EXP 3.0 with a minimum LOD score of 11.0 and a maximum distance of 25 cM.Then,we compared the marker orders of each group obtained by MAPMAKER/EXP 3.0 and MSTmap.The consistent regions of marker orders were retained,while for inconsistent regions of marker orders,adjustments were made by recalculation with more rigorous parameters(a minimum LOD score＞15 and a maximum distance ＜20 cM)by MAPMAKER/EXP 3.0.The Kosambi mapping function was chosen to calculate the genetic distances between markers[29].

QTL was detected using the composite interval mapping(CIM)procedure with the Windows QTL Cartographer(V2.5)software(http://statgen.ncsu.edu/qtlcart/WQTLCart.htm)as described by Wu et al.[7].A significance threshold(P=0.05)for QTL was determined through permutation analysis using 1000 repetitions.The confidence interval of QTL was determined by 1-LOD intervals surrounding the QTL peak.If QTL detected in different seasons had overlapped confidence intervals,they were considered to be the same QTL.

In order to identify environment-stable QTL,interactions between QTL and environments were performed by the Multi-Environment Trials(MET)functionality in QTL IciMapping software(V4.1)[30,31],at a walking step of 1 cM and with a LOD threshold of 2.5 as manual input.

2.5.Alignment of SSR resistance and FT QTL to the B.napus genome

QTL were detected from previous biparental QTL mapping studies for SSR resistance[6-8]and FT[8,11,15]as well as from GWAS for SSR resistance[4,9]and FT[12].Based on the physical positions of markers in confidence intervals or linkage disequilibrium blocks,all QTL were aligned to the‘Darmor-bzh'reference genome of B.napus[32].The physical locations of the SNPs,SSRs(single sequence repeats)and RFLPs(restricted fragment length polymorphisms)were determined using electronic PCR(e-PCR)or BlastN.For SSRs and SNPs(prefixed with IGF)markers,e-PCR was performed with the primer sequences using the genome sequences as templates[33].The e-PCR parameters for a given primer pair were set to allow three mismatches and one gap.We selected the amplicon according to the QTL linkage group if a pair of primers with multiple amplicons[33].The probe sequences of the RFLP markers were blasted against the reference genome sequences with an E-value of 1×10-20,and the hits were selected according to the linkage group of the QTL from the top several Blast hits.The physical localization of the SNP markers used in the GWAS was determined by Xu et al.[12],Liu et al.[23],and Wei et al.[9].A QTL alignment map was built using Circos software[34].

3.Results

3.1.Correlation between SSR resistance and FT in a natural population and a DH population

To screen resistance against S.sclerotiorum in B.napus,we assayed the resistance performance of a set of 521 rapeseed inbred lines when most lines were at the stage of flowering termination using a detached stem inoculation assay.Continuous and extensive phenotypic variations in SSR resistance were observed among the natural population.The lesion length on the stems at 5 dpi ranged from 1.25 to 13.52 cm(coefficient of variation is 31.2%)(Fig.1).We also recorded the FTs of the inbred lines,which had a wide variation,from 147 to 176 days (coefficient of variation is 2.5%)(Fig.1).Interestingly,we found that SSR resistance was negatively correlated with FT[Pearson's correlation coefficient(r)=-0.29,P=1.01×10-9](Fig.1).

Fig.1-Regression analysis of Sclerotinia stem rot resistance to flowering time in a set of 521 rapeseed inbred lines.r is Pearson’s correlation coefficient between the two traits.

A DH population with high FT variation was also chosen to further confirm the correlation between SSR resistance and FT.The female parent ZP1 showed flowering six days earlier than the male parent D126 in both 2014 and 2015(Fig.2-A).The FT of the DH population showed continuous segregation,ranging from 120 to 167.5 days in 2014 and 112.7 to 170.7 days in 2015(Fig.2-A).The resistance performance of the parental lines and the ZD-DH population was assayed using a stem inoculation assay in the field at the stage of flowering termination.A significant difference in the lesion length on the stem at 7 dpi between the two parents was observed,and ZP1 exhibited higher resistance than D126(Fig.2-B,C).In the ZD-DH population,transgressive segregation and continuous distribution were observed for SSR resistance in both years;the lesion length ranged from 0.6 cm to 13.1 cm in 2014 and 2.3 cm to 9.2 cm in 2015(Fig.2-D).The regression analyses showed that SSR resistance was also significantly negatively correlated with FT in the ZD-DH population in both years,and the correlation coefficients were high[-0.63 in 2014(P=3.95×10-17)and-0.66 in 2015(P=4.11×10-20)].These results further confirmed the link between SSR resistance and FT in B.napus,suggesting that early FT may be positively associated with increased susceptibility to S.sclerotiorum.

3.2.Construction of a high-density SNP-based genetic map

To understand the genetic basis of SSR resistance and FT,the ZD-DH population and its parents were genotyped by the Brassica 60 K SNP Array,and a total of 16,803 SNP markers(32.2%)from the array showed polymorphisms between the parental lines.To simplify the subsequent analysis,we used only 3060 nrSNPs for linkage map construction.A highdensity B.napus SNP bin map comprising 2904 nrSNPs(15,622 SNPs)was constructed.The total length of the genetic map was 2879.7 cM,with an average distance of 1.09 cM between adjacent nrSNPs(bins)(Fig.3,Table S1).The bin number,density and the biggest gap varied considerably across the different chromosomes.The highest bin density was found on A10,with a density of 0.61 cM bin-1.The greatest number of bins and the longest genetic map distance were found on A3,with 377 bins distributed over a genetic map distance of 288.5 cM.The bin density and number on the A genome(0.88 cM bin-1,1987 bins)were higher/larger than those on the C genome(1.33 cM bin-1,917 bins)(Fig.3,Table S1).

3.3.Detection of QTL for SSR resistance and FT in the ZD-DH population

A total of 16 nonredundant QTL for SSR resistance and FT were identified in two years(Table 1,Fig.4).Half of the QTL could be detected in both years(Table 1).Among the QTL,8 for SSR resistance were mapped on chromosomes A2,A3,A6,A10,C2,C3,and C8,which explained 4.73%-17.15%of the phenotypic variations(Table 1,Fig.4).The QTL of qSRA2,qSRA6,qSRA10,and qSRC8 were detected in both years(Table 1,Fig.4).The alleles from the susceptible parent D126 at these four loci increased the SSR resistance,while the resistant alleles of qSRA3a,qSRA3b,and qSRC3 were from the tolerant parent ZP1.

In addition,eight QTL for FT were located on chromosomes A2,A3,A6,A7,C2,C3,and C8,and these QTL accounted for 3.47%-22.19%of the phenotypic variations(Table 1).Four of them(qFTA2,qFTA3,qFTA6,and qFTC8)were detected repeatedly in both years.The alleles from the earlier flowing parent ZP1 at loci qFTA2,qFTA6,and qFTC8 accelerated flowering,while those at qFTA3 postponed flowering(Table 1).

In order to detect the stability of the identified QTL across environments,further analysis was performed with the MET functionality in QTL IciMapping(Table 2).Most QTL(except for qSRC3)identified by CIM method(Windows QTL Cartographer)were confirmed by inclusive composite interval mapping(ICIM)method in QTL IciMapping software(Table 2).No epistatic effect was found in this study.Twelve(six for SSR resistance,six for FT)out of the 15 QTL had a large LOD(A)value(＞2.5)and a small LOD(A by E)value(＜2.5)(Table 2),therefore these QTL showed non-significant additive×environment interaction effects.To these QTL,the additive effect contributed much more phenotypic variation than environment factor.The other three QTL(qSRA3b,qFTA6,and qFTA7a)showed additive×environment interactions[LOD(A by E)＞2.5],accounting for 0.77%,1.08%,and 1.34%of the phenotypic variance(Table 2).

Interestingly,the SSR resistance and FT QTL on A2,A6,C2,and C8 had overlapping confidence intervals(Table 1,Fig.4).In addition,although the confidence intervals of SSR resistance and FT QTL located in A3 and C3 were not overlapped,they were closely linked(Table 1,Fig.4).We found that the overlapped or closely linked QTL for SSR resistance and FT had the opposite of positive and negative additive effect values(Table 1).In other words,if the FT QTL alleles accelerate flowering,the closely linked SSR resistance QTL alleles will increase susceptibility.The QTL mapping results were consistent with the observation that there was a significant negative correlation between SSR resistance and FT(Fig.2-E,F).

Fig.2-Phenotypic characterization of Sclerotinia stem rot(SSR)resistance and flowering time(FT)of the ZD-DH population in 2014 and 2015.(A)Frequency distribution of the FT in the ZD-DH population.(B)Disease lesion on the stem of the two parental lines ZP1 and D126 at seven days post inoculation.Arrows indicate the lesion boundaries on the stem.(C)Lesion lengths of ZP1 and D126 at seven days post inoculation.★ ★indicates significant difference at P＜0.01.(D)Frequency distributions of the lesion length in the ZD-DH population.(E,F)Regression analysis of SSR resistance and FT in 2014(E)and 2015(F).

3.4.Integration analysis of the QTL for SSR resistance and FT

To determine if overlapping or closely linked SSR resistance and FT QTL is a universal phenomenon,we compared the distributions of QTL for SSR resistance and FT identified in the present and previous studies[4,6-9,11,12,15].A total of 54 SSR resistance QTL and 117 FT QTL were collected(Table S2).After aligning all QTL to the B.napus genome based on the physical positions of markers in the confidence intervals or LD blocks,we successfully located 49(90.7%)SSR resistance QTL and 102(87.2%)FT QTL on the physical map(Fig.5,Table S2).Among these located QTL,slightly more SSR resistance QTL were mapped on the C subgenome(28 QTL)than on the A subgenome(21 QTL),while many more FT QTL were mapped on the A subgenome(64 QTL)than on the C subgenome(38 QTL)(Fig.5,Table S2).Notably,four chromosomal regions with multiple QTL for both SSR resistance and FT were identified:A2[0-7.7 Mb,3 SSR resistance and 9 FT QTL(3&9)],A3(0.8-7.5 Mb,3&11),C2(0-15.2 Mb,8&7),and C6(20.2-36.6 Mb,7&5)(Fig.5,Table S2).Hence,SSR resistance and FT QTL are often detected in the same chromosomal regions,especially the above four chromosomal regions.

Fig.3-Overview of the high-density SNP bin map of B.napus.Each horizontal line represents a bin.The genetic distance(cM),bin density(bin cM-1),bin number and SNP number of each linkage group(chromosome)are shown on the right of each chromosome with different colored text.Detailed information on this map is shown in Table S1.

4.Discussion

Although early maturing varieties are often more susceptible to SSR than later maturing varieties in rapeseed breeding and production,whether there is a genetic link underlying this phenomenon is unclear.The prevailing explanation for this phenomenon is that the FT of earlier maturing varieties are consistent with the time when large quantities of S.sclerotiorum ascospores are forcibly discharged into the air under suitable environmental conditions.Late maturing varieties may escape from this stage,known as disease escaping.In this study,we found that SSR resistance QTL and FT QTL were colocalized on some chromosomal regions,thus affecting SSR resistance and FT simultaneously.

Natural variability within a species for FT and stress responses can be utilized to dissect the association between these two plant processes[17].In the present study,we investigated the correlation between FT and SSR resistance using a natural population containing 521 rapeseed inbred lines.Though the correlation of SSR resistance with FT in the natural population was low(r=-0.29,Fig.1),the correlation was highly significant(P=1.01×10-9)when the number of rapeseed inbred lines was 521.The low correlation coefficient also indicated that SSR resistance and FT were not linked in some accessions of the natural population.The association was further confirmed in the ZD-DH population(r=-0.63 to-0.66,Fig.2).These results were consistent with those from previous studies of B.oleracea and B.napus.In B.oleracea,Mei et al.[22]reported significant negative correlations(r=-0.26 to-0.39)between SSR resistance and FT in an F2population of B.oleracea.Similarly,negative correlations of SSR resistance and FT were found by Zhao et al.[6](r=-0.52)and by Wei et al.[8](r=-0.21 to-0.35)in DH populations of B.napus.Significantassociations between disease resistance and FT were also found in other plants.In Arabidopsis,early FT was positively associated with increased susceptibility to Verticillium dahlia[35]and Fusarium oxysporum[36].In rice,bacterial grain rot resistance was negatively correlated with heading date[19],and blast resistance was negatively correlated with FT[37].In chickpea,ascochyta blight resistance was negatively correlated with FT[38].However,the genetic bases and molecular mechanisms underlying the associations between disease resistance and FT are often not clear.

Previous research has demonstrated that the SSR resistance in B.napus showed moderate heritability(57.2%-84.0%),as determined by quantitative resistance[4].In the present study,we dissected its genetic underpinnings by QTL mapping in the ZD-DH population,and eight SSR resistance QTL that explained 4.73%-17.15%of the phenotypic variations were identified(Table 1).The major QTL qSRA2 and qSRC8 were detected in both years,only explaining 7.40%-17.15%and 10.17%-12.78%of the phenotypic variations,respectively(Table 1).This finding is consistent with previous studies showing that SSR resistance in B.napus was determined by multiple minor QTL[4].

After identifying the FT QTL using the ZD-DH population,we found that the SSR resistance and FT QTL on A2,A6,C2,and C8 had overlapping confidence intervals and those located on A3 and C3 were closely linked(Table 1,Fig.4).The colocalized QTL for SSR resistance and FT on A2 and C2 are in accordance with the results of studies by Zhao et al.[6]and Wei et al.[8].The QTL analysis results were consistent with the observation that there were significant negative correlations between SSR resistance and FT.

It is challenging to compare the QTL results of previous studies,largely because the studies were performed with different sets of molecular markers and many different segregating populations.With recent advances in Brassica genome sequencing,it is now possible to map the QTL from different studies onto the physical map of B.napus.Therefore,to place the genetic architecture of SSR resistance and FT in B.napus in the larger context of genetic backgrounds,we integrated QTL mapping and GWAS data from multiple studies by aligning the QTL onto the physical map of B.napus(Fig.5).Four colocalized QTL hotspots on A2(0-7.7 Mb),A3(0.8-7.5 Mb),C2(0-15.2 Mb),and C6(20.2-36.6 Mb)for both SSR resistance and FT were identified(Fig.5).

There are several possible reasons for the associations between disease resistance and FT.Firstly,resistance QTL/genes identified by QTL mapping or GWAS might affect disease either directly or indirectly[20].The QTL/genes affecting FT may fall in the latter class.For example,senescence that coincides with flower development may promote disease development.In particular,necrotrophic disease tends to progress more rapidly in senescing tissues than in younger tissues[20,39].

Fig.4-QTL screening results of the QTL for Sclerotinia stem rot(SSR)resistance and flowering time(FT).Curves of different colors represent QTL screened from different traits and years.QTL nomenclature uses the initials of‘q’and the abbreviation of the trait followed by the chromosome number;an alphabetical letter a or b or c is added if more than one QTL is identified in one chromosome.The horizontal dotted line represents the LOD(likelihood of odds)values of 2.5.

Table 2-Evaluation of additive QTL×environments interactions by the Multi-Environment Trials(MET)functionality in QTL IciMapping software.

A second possibility is that the resistance genes and FT genes were closely linked.BnFLC.A2 and BnFLC.C2 are the causal genes of the FT QTL on A2 and C2 in B.napus[16].In our previous research,we performed dynamic transcriptomic analyses to understand the differential defense response to S.sclerotiorum in a resistant line and a susceptible line of B.napus at 24,48,and 96 h post inoculation[40].Within 5 Mb near BnFLC.A2(134,361-138,212 bp of A2)and BnFLC.C2(208,561-212,139 bp of C2),118 and 42 relative differentially expressed genes in the resistant line compared with the susceptible line were identified respectively[40].Therefore,in these two chromosomal regions with colocalized QTL of SSR resistance and FT,the causative genes for the two traits may be closely linked.Thus,further analyses of QTL for SSR resistance affected by FT using a disease evaluation method and/or genetic population not confounded by FT variation should be performed.

A third possibility is that several genes that affect both disease resistance and flowering have been identified in Arabidopsis.For example,Kidd et al.[41]identified an additional mediator subunit,MED8,that is a regulator of both FT and disease resistance in Arabidopsis.Lai et al.[42]found that Arabidopsis MED18 is a multifunctional protein regulating plant immunity,FT and responses to hormones through interactions with distinct transcription factors.Lyons et al.[43]found that in addition to its role in defense regulation,the Arabidopsis RNA-binding protein FPA also promotes the transition to flowering.The FLOWERING LOCUS D(FLD)functions to repress expression of the flowering repressor FLOWERING LOCUS C(FLC)and is required for the response to systemic acquired resistance signals,leading to systemic accumulation of SA and enhancement of disease resistance[44].Plant U-Box protein 13(PUB13)[45],HOPW1-1-INTERACTING 3(WIN3)[46],SUMO E3 ligase SIZ1[47]and EDM2[48]have also been demonstrated to regulate both flowering and disease defense in Arabidopsis.These findings suggest that some proteins have evolved to play dual roles in defense and flowering.Hence,the possibility of pleiotropy of FT genes to affect SSR resistance in rapeseed cannot be ruled out.

Fig.5-Integration analysis of the QTL for Sclerotinia stem rot(SSR)resistance and flowering time(FT)based on the reference genome sequences of B.napus.From inside to outside,the five inner circles(1-5)with a yellow background color represent FT QTL identified by Xu et al.[12],Wei et al.[8],Long et al.[11]and Quijada et al.[15],and the six outer circles(6-11)with a blue background color represent SSR resistance QTL identified by Wu et al.[4],Wei et al.[9],Wei et al.[8],Wu et al.[7],and Zhao et al.[6].Short bars with color within the 11 circles represent the physical localization of QTL.The outermost circle represents 19 chromosomes of B.napus scaled based on their assembled lengths(Mb).

In conclusion,the present study reveals co-localized QTL for SSR resistance and FT in B.napus,which provide clues to dissect the genetic link between the two important traits.Further studies are required to understand the molecular mechanisms underlying the relationship,which is essential for genetic improvement of SSR resistance and early maturity in rapeseed breeding.

Supplementary data for this article can be found online at https://doi.org/10.1016/j.cj.2018.12.007.

Competing interests

The authors declare no conflict of interest.

Acknowledgments

This work was supported by the National Natural Science Foundation of China(31671725,31601330,31330057),the National Key Basic Research Program of China(2015CB150201),Science&Technology Special Project of Guizhou Academy of Agricultural Sciences([2014]014,[2017]08),and the China Postdoctoral Science Foundation(2015M581867,2016T90514).