Pingping Zhng, Chenjin Guo, Zho Liu, Amy Bernro,Hongxing M,Peng Jing, Guicheng Song, Guihu Bi,,*

aJiangsu Academy of Agricultural Sciences, Jiangsu Collaborative Innovation Center for Modern Crop Production, Nanjing 210014, Jiangsu,China

bDepartment of Agronomy,Kansas State University,Manhattan, KS 66506,USA

cHebei Agricultural University,Baoding 071001,Hebei,China

dUSDA-ARS,Hard Winter Wheat Genetics Research Unit,Manhattan, KS 66506,USA

Keywords:QTL mapping FHB Yangmai 158 Zhengmai 9023 KASP

ABSTRACT Fusarium head blight (FHB) is one of the prevalent fungal diseases of wheat worldwide. Exploring new FHB resistance quantitative trait loci (QTL) in adapted wheat cultivars is a critical step for breeding new FHB-resistant cultivars. In this study, we developed a population of 236 F5:7 recombinant inbred lines (RILs) using two popular Chinese wheat cultivars, Yangmai 158 and Zhengmai 9023, with moderate FHB resistance to identify the QTL for FHB type II resistance. This population was evaluated for percentage of symptomatic spikelets per spike (PSS) using single floret injection in repeated greenhouse experiments. Mean PSSs were 33.2% for Yangmai 158 and 30.3% for Zhengmai 9023. A genetic linkage map of 1002 single nucleotide polymorphisms (SNPs) generated by genotyping-by-sequencing (GBS) was constructed for the RIL population. Six QTL were identified for FHB resistance, and three of them were repeatable in the both experiments. Zhengmai 9023 contributed the resistance allele at one repeatable QTL, designated as Qfhb.7D, whereas Yangmai 158 contributed the resistance alleles at the other two repeatable QTL, Qfhb.3AL and Qfhb.2DS. The additional QTL, Qfhb.4AS was significant in the mean PSS, and Qfhb.2DL and Qfhb.7AS were significant in only one experiment. Replacement of each allele individually at the three repeatable QTL significantly changed PSSs. Qfhb.3AL, Qfhb.2DS, and Qfhb.7D explained 8.35% to 9.89%, 5.13% to 7.43%, and 6.15% to 9.32% of the phenotypic variations, respectively. The three repeatable QTL contributed by the two parents were additive and stacking the resistance alleles from all the three repeatable QTL showed the highest level of resistance in the current RIL population. Ten SNPs in the QTL regions of Qfhb.3AL, Qfhb.2DS, and Qfhb.7D were converted into KBioscience competitive allele-specific PCR (KASP) assays. One KASP marker for Qfhb.3AL was validated in a panel of wheat cultivars from China. Some of these KASP markers could be useful for marker-assisted selection to stack these QTL.

1.Introduction

Fusarium head blight (FHB), mainly caused by the fungus Fusarium graminearum, is a serious wheat disease that causes billion dollars of losses each year worldwide,and is becoming a threat to global wheat production[1-3].In China,FHB is one of the most important diseases in the winter wheat regions of Yangtze-Huai River Valleys and southwestern and northeastern spring wheat regions. The recent changes in climate and cropping systems such as straw retention and minimum or no tillage cultivation make FHB epidemics more frequent,severe,and widely spread.It even has become one of the problematic diseases in the arid and semi-arid wheat growing regions where FHB has not been reported previously [2,4-6]. From 2010 to 2018,more than 3.3 Mha of wheat in average suffered from FHB annually, with the most severe epidemic in 2012 in which more than 10.0 Mha of wheat was infected [7]. More importantly, the FHB pathogen can produce mycotoxin such as deoxynivalenol (DON) in infected kernels, which is a serious safety concern to human and animal health [8].Although FHB epidemics could be reduced by traditional disease management practices such as deep ploughing,improving field drainage, and using fungicides, host resistance has proven to be the most effective and environmentally friendly strategy to minimize FHB damage [9].

Resistance to FHB is a quantitative trait that is governed by multiple QTL and vulnerable to changing environments [2].Wheat FHB resistance has been classified into five types:resistance to initial fungal infection (Type I), resistance to fungal spread in the spike (Type II), low toxin accumulation(Type III), low Fusarium damaged kernel (FDK) (Type IV), and tolerance to yield loss(Type V)[10];however,only type II is the stable type of resistance and the easiest to assess and select in wheat breeding programs [1]. To date, more than 200 QTL have been reported for different types of resistance on all the wheat chromosomes, and meta-analysis clustered them into about 50 QTL with unique chromosome locations for each QTL[11]. Some of these QTL have been associated with plant defense responses [12,13]or plant morphological traits such as plant height,anther extrusion and flower time[14].Among these QTL, only seven are formally named, with Fhb1, Fhb2,Fhb4, and Fhb5 from common wheat on chromosomes 3BS,6BL,4BL,and 5AS,respectively,and Fhb3,Fhb6,and Fhb7 from alien species that were transferred to chromosomes 7AS,1AS,and 7DL of wheat,respectively [2].Among all these QTL,only Fhb1 consistently showed the highest level of resistance,explaining up to 50% of the phenotypic variation for type II resistance in different genetic backgrounds and testing environments.Fhb1 has been successfully applied in breeding programs through marker-assisted selection (MAS) worldwide, especially in the USA, Canada, Europe, China, and the International Maize and Wheat Improvement Center(CIMMYT)in Mexico[6,7,15].To date,many cultivars carrying Fhb1 have been released in the winter wheat regions of the Middle and Lower Valleys of the Yangtze River in China,North Dakota and Minnesota in the USA, and Manitoba in Canada[15]. More recently, Fhb1 has been cloned as a histidine-rich calcium binding protein (TaHRC) and the diagnostic markers have been developed for the gene [9,16,17]. The cloned gene and diagnostic markers for Fhb1 will greatly facilitate the deployment of this gene in new wheat cultivars. However,Fhb1 alone is not sufficient for preventing FHB damage in severe FHB epidemics due to its partial resistance [2,10,16].Thus, pyramiding Fhb1 with other FHB resistance QTL from adapted genetic backgrounds is necessary to quickly achieve acceptable levels of both FHB resistance and other agronomic traits [18,19]. Some Chinese landraces with high or moderate FHB resistance have been identified and used in breeding in the first half of the last century, but they could not be used directly as the breeding parents in modern breeding programs due to their inferior agronomic performances [3,20-23].Exploring new FHB resistance QTL directly from adapted wheat cultivars may speed up the gene pyramiding process.

Yangmai 158 and Zhengmai 9023 have been the most popular Chinese commercial wheat cultivars in the Middle and Lower Valleys of the Yangtze River and Huang-Huai Valleys with the total planting areas of about 8 and 17 Mha,respectively (China Seed Association, http://www.seedchina.com.cn). Both cultivars have desirable agronomic and grain quality traits with moderate FHB resistance [6,24-26]. They have been used extensively as parents in many breeding programs, but QTL for FHB resistance in the two cultivars remain unknown. The objectives of this study were to (1)identify QTL for type II resistance in both Yangmai 158 and Zhengmai 9023 using a population of recombinant inbred lines (RILs); (2) quantify the effects of resistance QTL in the two cultivars;and(3)develop breeder-friendly markers tightly linked to the QTL for MAS.

2. Materials and methods

2.1. Plant materials and FHB evaluation

The RIL population of 236 F5:7lines was derived from a cross between two Chinese wheat cultivars, Yangmai 158 and Zhengmai 9023, by single-seed descent. Yangmai 158 is a hard-red facultative wheat cultivar, and Zhengmai 9023 is a hard-white winter wheat cultivar. A diversity panel of 94 wheat cultivars from the Middle and Lower Valleys of the Yangtze River in China (Table S1) were used to validate the KBioscience Allele-Specific PCR (KASP, LGC Genomics, Beverly, MA, USA) markers linked to an FHB resistance QTL. The panel consisted of newly released cultivars from Jiangsu (67),Anhui (3), and Hubei provinces (24) in the past several decades.

Wheat plants from the RIL population were evaluated for Type II resistance in the greenhouse experiments at Kansas State University, Manhattan, Kansas in spring 2017 and 2018.Eight seeds per RIL and the parents were planted in 96-cell plastic trays filled with a soil mix (Hummert International,Topeka, KS, USA). After 28 days of vernalization at 6 °C in a cold room,the seedlings in each cell were transplanted into a 14×14 cm plastic pot(Hummert International)filled with the soil mix. The pots were placed on greenhouse benches in a randomized complete block design with two replications(pots) per line.The greenhouse temperature was firstly set at(13 ± 2) °C at the night and (17 ± 2) °C in the day with 12 h supplementary daylight for 30 days,then was set at(19 ± 2)°C at night and (22 ± 5) °C in the day under the same daylight setting until the maturity. Conidial spore suspensions of F.graminearum strain GZ 3639 isolated from a Kansas field were prepared following Bai et al. [27]. A 10 μL spore suspension(～1000 spores)was injected into a floret in a central spikelet of a spike at early anthesis. About five spikes per pot were inoculated and enclosed in a moist chamber at 100% relative humidity for 48 h to initiate infection. Then, the plants were moved back to the greenhouse benches for disease development. The total number of spikelets and the number of symptomatic spikelets in each inoculated spike were counted at 15 days post-inoculation (dpi). Percentage of symptomatic spikelets (PSS) was calculated as FHB severity for each plant and the PSS of each RIL in each experiment and mean PSS per line over two experiments were calculated for QTL analysis.

Type II FHB resistance was also evaluated for the diversity panel in three field experiments in Nanjing, China, with two experiments conducted at Xuanwu and Luhe Stations, Nanjing in 2018, and one experiment at Luhe Station, Nanjing in 2019. In the field experiments, 10 spikes per line were inoculated using the highly virulent Fusarium strain F0609 by needle injection of a 10 μL spore suspension (～1000 spores)into a middle spikelet of each spike. After inoculation, the inoculated spikes were covered with a plastic bag for 72 h to maintain moisture. Then, the plastic bag was removed, and the spikes were misted with water to maintain moisture.The total and symptomatic spikelets in each inoculated spike were counted at 21 dpi to calculate PSS.

2.2. DNA extraction and marker analysis

Leaf tissues were collected into 96-deepwell plates at threeleaf stage, and dried in a freeze dryer (ThermoSavant,Holbrook, NY, USA) for 48 h, and ground into a fine powder using a Mixer Mill (MM 400, Retsch, Germany). Genomic DNA was isolated using a modified cetyltrimethyl ammonium bromide protocol [28]. A genotyping-by-sequencing (GBS)library was generated for the RILs and parents according to Mascher et al. [29]and sequenced using an Ion Proton sequencer (Thermo Fisher Scientific, Waltham, MA, USA). In brief, each DNA sample was digested with HF-PstI and MspI and ligated with adaptors using the T4 ligase (New England BioLabs Inc., Ipswich, MA, USA). Ligated samples with different barcodes were pooled, cleaned up using a PCR Purification Kit (Qiagen Inc., Valencia, CA, USA), and then amplified by PCR using 10 μmol L?1Ion primers and 5 μL Taq 5×Master Mix(New England Bio Labs Inc.).The PCR mixture was incubated at 95 °C for 30 s initially, followed by 16 cycles of 95 °C for 30 s,62 °C for 20 s,and 68 °C for 1 min,then at 72 °C for 5 min for a final extension.The PCR products were cleaned up again using the QIA PCR Purification Kit (Qiagen Inc.,Valencia, CA, USA). The fragments of 250-300 bp were size selected in an E-gel system and sequenced in an Ion Proton sequencer (Life Technologies Inc.). The sequence reads generated from Ion Proton were analyzed for SNPs using Universal Network Enabled Analysis Kit (UNEAK) in TASSEL pipeline [30,31]. Sequence reads with less than 64 bp were added with a poly-A tail in the 3′-end to ensure that all reads have at least 64 bp.

2.3. Linkage map construction and QTL analysis

A linkage map of GBS-SNP markers was constructed using the Kosambi mapping function [32]and ‘Regression' mapping algorithm in JoinMap v4.0[33].QTL were detected for PSS from each experiment and mean PSS over both experiments using the ‘Composite Interval Mapping' (CIM) in WinQTL Cartographer v2.5[34]with the threshold of 2.5 that was generated by 1000 permutations. Analysis of variance (ANOVA) and least significant difference (LSD) at P = 0.05, 0.01, and 0.001 were performed using the PROC GLM procedure in SAS 9.0(SAS Inc.,Cary, NC, USA) to compare the effects among genotypic groups that carried different numbers of QTL.

2.4. KASP marker development and validation

Based on the GBS reads harboring the SNPs in the repeatable QTL regions, KASP primers were designed, and used to validate GBS-SNP polymorphisms in the RIL population. The KASP master mix for each reaction comprised of 3 μL of 2×KASP reaction mix, 0.0825 μL of KASP primer mix (including 12 μmol L?1of the two forward primers each,and 30 μmol L?1of the common reverse primer)and 3 μL of DNA(～60 ng).The PCR mix was incubated at 94 °C for 15 min, followed by 11 cycles of 94 °C for 20 s and annealing at 65 °C for 1 min with a decrease of 0.8 °C in each subsequent cycle. Then, the PCR proceeded with additional 39 cycles of 94 °C for 20 s and 57 °C for 1 min, followed by 16 °C for 10 min. PCR plates were then read in a BMG FLUOstar Omega reader (BMG LabTech Inc.,Gary, NC, USA). The mismatches between KASP-SNPs and GBS-SNPs were counted. The GBS-SNPs in the QTL region were replaced by corresponding KASP markers and remapped in the RIL population for QTL analysis. The KASP markers were also validated in the panel of wheat cultivars from the Middle and Lower Valleys of the Yangtze River in China.Genomic DNA was isolated from each line by bulking leaf samples from four individual seedlings.

3. Results

3.1. FHB resistance of the RILs and parents

Fig.1- Frequency distribution of percentage of symptomatic spikelets in a spike (PSS)for the population of 236 Yangmai 158× Zhengmai 9023 recombinant inbred lines(RILs)evaluated in spring 2017 and 2018,and the mean PSS over the two greenhouse experiments.The average PSS for Yangmai 158 and Zhengmai 9023 were 33.2%and 30.3%,respectively,over the two greenhouse experiments.

Both parents showed moderate FHB resistance in spring 2017 and 2018 greenhouse experiments.The mean PSSs were 33.2%for Yangmai 158 and 30.3% for Zhengmai 9023 over the two experiments (Fig. 1, Table 1). The mean PSSs for the RIL population were 32.8% and 38.6%at the ranges of 5.4%-99.0%and 6.1%-100.0% in the two experiments, respectively. There were 97 and 74 moderately resistant RILs (PSS ≤25.0%) in the two experiments,respectively.Frequency distribution of PSSs in the RIL population was continuous with obvious transgressive segregation(Fig.1),suggesting that both parents contributed resistance alleles. The correlation in PSS between the two experiments was highly significant, with a correlation coefficient of 0.52 (P < 0.001). Variance analysis (ANOVA)showed significant variations in genotypes, environments and genotypes by environments(P <0.001).

3.2. Construction of a linkage map

Sequencing the GBS libraries identified 15,409 SNPs with≤80%missing data.Among them,1993 SNPs had ≤20%missing data and were used to construct a linkage map. The initial map was constructed using 1002 SNPs in 231 RILs after removal of five other RILs that had excessive missing data.The map was 2651.0 cM in genetic distance with an average marker density of 2.65 cM per marker.It contained 42 linkage groups representing 21 chromosomes except for chromosome 4D(Table S2)with at least three markers per group.The wheat A, B and D genomes had 42.8%, 47.0%, and 10.2% of total markers, respectively. The chromosomes with the highest and lowest marker density were 2D (0.87 cM per marker) and 3D(6.41 cM per marker),respectively (Table S3).

3.3. QTL for FHB resistance

Six FHB resistance QTL were detected on chromosomes 2D(Qfhb.2DS and Qfhb.2DL), 3A (Qfhb.3AL), 4A (Qfhb.4AS), 7A(Qfhb.7AS), and 7D (Qfhb.7D) (Fig. 2). Zhengmai 9023 contributed the resistance alleles at Qfhb.4AS and Qfhb.7D, and Yangmai 158 provided the resistance alleles at the other four QTL. Qfhb.3AL, Qfhb.2DS, and Qfhb.7D were significant in both experiments and the mean PSS, therefore they were considered repeatable QTL; whereas Qfhb.2DL and Qfhb.7AS were significant only in one experiment, Qfhb.4AS was significant only in the mean PSS. Among the three repeatable QTL,Qfhb.3AL showed the largest effect, and explained 8.35%-9.89% of the phenotypic variation (Table 2), and was mapped near SNP GBS17928 in a 16.97 cM interval flanked by SNPs GBS20758 and GBS781. Qfhb.7D was mapped in a 27.53 cM interval between SNPs GBS979 and GBS20328 and explained 6.15%-9.32%of the phenotypic variation.Qfhb.2DS was located near SNP GBS2375 in a small linkage group (18.35 cM), and it explained 5.13%-7.43% of the phenotypic variation, and Qfhb.2DL explained 10.25% and 6.56% of the phenotypic variation in spring 2017 experiment and for the mean PSS,respectively.Qfhb.7AS was only significant in spring 2018 and explained 5.47% of the phenotypic variation. Qfhb.4AS was significant in the mean PSS and explained 5.08% of the phenotypic variation.

Fig.2-Maps of quantitative trait loci(QTL)for FHB resistance in the population of Yangmai 158×Zhengmai 9023 recombinant inbred lines(RILs) evaluated in spring 2017 and 2018 greenhouse experiments.

?

SNPs GBS17928, GBS2375, and GBS20328 are the closest markers to Qfhb.3AL, Qfhb.2DS, and Qfhb.7D, and their contrasting alleles were designated ‘AA/aa', ‘BB/bb', and ‘CC/cc', respectively, where uppercase letters represented the marker alleles that were associated with FHB resistance, and lowercase letters denoted the marker alleles that were associated with FHB susceptibility. To compare the allelic effect of each individual QTL on type II FHB resistance,all RILs in the population were grouped based on their alleles.Replacement of a susceptibility allele by a corresponding resistance allele at each locus significantly decreased PSS(Fig.3a). However, the levels of significance between two alleles varied with QTL, P < 0.001, P < 0.05, and P < 0.01 for Qfhb.3AL,Qfhb.2DS, and Qfhb.7D, respectively. The combined effects of the three repeatable QTL on FHB resistance were also investigated (Fig. 3b). The average PSSs of two experiments for the eight possible RILs groups were 46.28%(aabbcc),37.44%(AAbbcc), 41.92% (aaBBcc), 36.10% (aabbCC), 29.66% (AABBcc),31.19% (AAbbCC), 38.62% (aaBBCC), 22.89% (AABBCC), respectively. In general, the RIL groups carrying more resistance alleles showed higher FHB resistance. The RIL group carrying resistance alleles at all QTL showed significantly lower PSS than all the other RIL groups (P < 0.001), thus pyramiding the minor FHB resistance QTL from the two cultivars could significantly reduce the FHB severity.

3.4. Conversion of GBS-SNPs to KASP markers

To develop breeder-friendly markers for breeding,14 KASP assays were designed based on the GBS tag sequences in the three repeatable QTL intervals, and 12 of them segregated in the current RIL population(Fig.4a?c).Among them,ten KASPSNPs had SNP calls identical to their corresponding GBS-SNPs,and two SNPs in Qfhb.3AL (GBS516, 99.52%; GBS7621, 97.84%matching) showed a few mismatches. The three KASP SNPs that were mapped to the peaks of the three QTL regions were selected(Tables S1 and S4).All alleles at the three KASP SNPs were unequally distributed (Table S1). Zhengmai 9023 allele was predominant at Qfhb.2DS and Yangmai 158 allele was predominant at Qfhb.7D in the panel, indicating that markers for Qfhb.2DS and Qfhb.7D are likely polymorphic when their donors are crossed to most parents from the Middle and Lower Valleys of the Yangtze River in China. The mean PSS value for the group of the accessions with the resistance marker allele at Qfhb.2DS was lower than these with the susceptibility allele although the difference was not significant.For the GBS17928 at Qfhb.3AL,the frequency of Yangmai 158 allele was higher than Zhengmai 9023, in particular in these accessions developed from Jiangsu province,suggesting this QTL has been deployed in many cultivars from Jiangsu province.The PSS values for the group of the accessions with the resistance marker allele were significantly lower than these with the susceptibility allele in all three experiments(Table S1, Fig. 4d, Table 3). Thus, these KASP markers can be utilized in MAS for stacking these QTL.

4. Discussion

Fusarium head blight is a quantitative trait that is conditioned by multiple QTL. Due to natural and human selection, some Chinese landraces such as Wangshuibai and Haiyanzhong etc. in the Middle and Lower Valleys of Yangtze River where severe FHB epidemics frequently occurred in the history accumulated multiple QTL, thus gained high levels of FHB resistance [22,23,35,36]. Unfortunately, the resistance QTL in these landraces have been difficult to be incorporated into modern cultivars due to many undesirable agronomic traits and poor adaptation of these landraces.Fhb1,a major QTL for type II resistance identified from Chinese landraces,has been successfully moved to some adapted backgrounds and widely used in breeding programs worldwide [6,7,15]; however, Fhb1 only confers partial resistance and it alone is not sufficient to protect wheat from losses under severe epidemics.Therefore,more genes in adapted wheat backgrounds are urgently needed to improve the levels of resistance in modern cultivars.

Fig.3- Boxplots to show the effects of individual quantitative trait locus(QTL)(a)and allelic combinations of the three repeatable QTL(b)measured by the percentage of symptomatic spikelets(PSS,%)in the recombinant inbred population.AA,BB,and CC represent peak marker alleles associated with FHB resistance on Qfhb.3AL(GBS17928), Qfhb.2DS(GBS2375), and Qfhb.7D(GBS20328),respectively;aa,bb,and cc correspond marker alleles associated with FHB susceptibility,respectively.*,**,and***indicate significant at the 0.05,0.01,and 0.001 probability levels,respectively.The horizontal line in the box indicates the median;“x” indicates the mean value.A dot above the boxplot is a outlier.

Fig.4- KASP assay validation profiles of GBS17928(Qfhb.3AL)(a), GBS2375(Qfhb.2DS)(b) and GBS20328(Qfhb.7D)(c)in the recombinant inbred population,and KASP assay validation profile of GBS17928(Qfhb.3AL)(d)in a panel of wheat cultivars from the Middle and Lower Valleys of the Yangtze River in China.The blue dots represent SNP allele in Yangmai 158,the green dots represent SNP allele in Zhengmai 9023,the red dots represent heterozygotes,the yellow dots represent artificial heterozygotes,and the black dots are negative controls.

Fortunately,many breeding lines with moderate resistance or moderate susceptibility have also been identified,and most of them have improved agronomic traits and desired adaptation to local environments. Although the levels of FHB resistance in these lines are not sufficient for full protection under severe FHB epidemics, they carry various numbers of resistance genes that can be more easily integrated into new modern cultivars unlike those from landraces. Therefore,stacking these minor genes from different cultivars with Fhb1 can create new highly resistant cultivars. Yangmai 158 and Zhengmai 9023 have excellent agronomic traits and processing quality and were widely grown in the Middle and Lower Valleys of the Yangtze River and Yellow River Valley,respectively, in the past two decades (http://www.seedchina.com.cn). They also demonstrated stable moderate FHB resistance in multiple field locations [6,24]and can be excellent genetic sources for improvement of FHB resistance in wheat.Marker data showed that neither of them carry Fhb1[6], therefore, identifying the QTL underlining FHB resistance in these sources and developing breeder-friendly markers are critical for breeders to quickly deploy these QTL in breeding.

In this study, we identified two QTL for FHB resistance(Qfhb.4AS and Qfhb.7D) from Zhengmai 9023 and other four from Yangmai 158. One QTL region from Zhengmai 9023 covered most of chromosome arm 7DS and a small portion of 7DL.Li et al.[37]identified a QTL for type IV resistance flanked by Xwmc405 and Xcfd14(91.4-265.0 Mb,IWGSC RefSeq v2.0)in Wangshuibai.Later this QTL was narrowed down to a smaller genomic region between markers Xwmc702 and Xcfd46(135.2-140.0 Mb) in another Chinese landrace Haiyanzhong[22,36].Cativelli et al.[38]identified a QTL for type II resistance near Xwmc702 in Catbird and McCartney et al. [39]also reported a major QTL near Xgwm44 (98.2 Mb) from ‘Kenyon'.All these QTL-linked markers are between Xwmc405 and Xdfd14 on 7D [40], therefore, they are likely the same QTL.Another QTL in Arina was identified near Xwmc488(527.2 Mb)on 7D [41]. In addition, Wu et al. [24]using single marker analysis identified eight FHB resistance associated SSR markers that span a long genomic region from 92.3 to 632.0 Mb between Xbarc126 and Xbarc76. Recently, a QTL with a major effect was reported to be closely linked to Xgwm428 (616.9 Mb) [42], and another QTL for type III and IV resistance was reported near the marker wPt-743601(2.52 Mb)on 7DS[43].In the current study,the Qfhb.7D located between flanking marker GBS979 and GBS20328 at the interval between 107.1 and 521.7 Mb, which appear to overlap with several previously reported QTL [24,41]. However, the Qfhb.7D region covers almost half of chromosome 7D,so its interval needs to be further narrowed down to accurately determine the relationship to previously reported QTL on this chromosome.

?

Another QTL,Qfhb.4AS,from Zhengmai 9023 with a minor effect was significant only in the mean PSS over the two experiments. Previously several QTL for FHB resistance were reported on 4AL [11,39,44], but they are likely different from the 4AS QTL reported in this study based on the closely linked markers in the consensus map [40]. Three QTL were previously mapped on chromosome arm 4AS with the first QTL near wPt-0804 around 18.0 Mb in Frontana[43],the second QTL near Xgwm165 at 51.12 Mb [45], and the third QTL between cfa2256 at 81.7 Mb and Xcfd71 at 146.7 Mb [46]. Qfhb.4AS from the current study was mapped between 132.9 and 310.3 Mb,thus, it partially overlaps with the third QTL reported on the 4AS [46]; however, their allelic relationship remains to be determined.

The current study mapped four QTL from Yangmai158.Qfhb.3AL on chromosome arm 3AL showed the largest effect.Previous studies mapped several QTL on chromosome 3A,but most of them were mapped on the short arm [20,21,46-51].Only two QTL were reported on 3AL. Koller et al. [46]found a QTL near Xcfa2193 between 690.8 and 747.0 Mb, and Li et al.[23]found a QTL near Xwmc428 (597.9 Mb) using association analysis method. In the current study, Qfhb.3AL was mapped between GBS20758 and GBS781(544.5-599.3 Mb)that are close to Xwmc428, therefore, it partially overlaps with the QTL reported by Li et al. [23], but is likely different from the one reported by Koller et al.[46].

Two FHB resistance QTL were identified on 2D of Yangmai 158. Several QTL have been previously reported on 2DL.Wuhan-1 from China was reported to carry a resistance QTL near Xgwm539(515.2 Mb)on 2DL[52],a similar location where QFhb.crc-2D.4 was mapped in the population of Kenyon/86ISMN [39]. In addition, three other QTL were mapped on the same chromosome 2DS in the Kenyon/86ISMN population[39]. Agnes et al. [43]identified a QTL near wPt-732603(599.8 Mb) in the population of GK Mini Manó/Frontana. In the current study, Qfhb.2DL was located near GBS16142(587.6 Mb), which is less than 30 Mb from wPt-732603 and Xcfd233 (560.4 Mb), thus, is most likely the same QTL previously reported[11,43,46,52].

Qfhb.2DS in the current study is a repeatable QTL that mapped close to GBS2375 (75.2 Mb) in a small linkage group and showed a significant effect in the both greenhouse experiments. Yangmai 158 carries the dwarf gene Rht8,which is known to be tightly linked to Xgwm261 [53]. Thus,Qfhb.2DS might be the same QTL that mapped near Xgwm261(20.4 Mb) [21,23,35,39,46,54]. Wu et al. [24]identified a minor QTL near Xgwm296 that is 1 cM from Xgwm261 in ARz/Yangmai 158 population.However,the resistance allele of the QTL was from ARz, suggesting multiple alleles at the QTL.Based on the physical positions of the tightly linked markers,Qfhb.2DS is likely the same QTL as previously reported QTL[11,52]. However, the average distance per marker on the 2D chromosome is 9.09 cM, which suggested the linked markers we identified may be still too far from the QTL. Further fine mapping of the QTL is essential to determine the relationship with other known QTL.

Qfhb.7AS with a minor effect was significant only in spring 2018. Previously, several QTL have been reported on chromosome 7A, but most of them were located on the 7AL. For example, Jia et al. [35]identified a QTL flanked by Xgwm276 and Xgwm282(647.0-686.0 Mb)on 7AL;Wu et al.[55]identified QFHB-7A (638.7-671.0 Mb) in the same region from a Chinese source, which has high frequency in wheat cultivars used in the Middle and Lower Valleys of the Yangtze River in China.This QTL was also reported in other studies [56]. Only one study reported a QTL on 7AS between Xwmc479-Xcfa2049(22.7-49.8 Mb) [46]. The Qfhb.7AS (42.6-61.0 Mb) identified in this study may be the same as the one reported by Koller et al.[46],but different from the QTL reported by Miedaner et al.[57]that was near Xwmc596 at 481.2 Mb.

In the current study,both Yangmai 158 and Zhengmai 9023 showed moderate resistance that was conditioned by at least two QTL in each parent. As indicated by transgressive segregation in the population, these QTL were contributed by both parents. Although each individual QTL showed a minor effect, the three repeatable QTL together provided a high level of resistance (PSS = 22.89%), showing additive effects among QTL from the two parents (Fig. 3b). In the mapping population, some RILs had even lower PSS (<10%,Fig. 1) than the lines with all three repeatable resistance alleles (Fig. 3b), which might be due to the three additional unstable QTL from Yangmai 158 and Zhengmai 9023, or to common QTL between the two parents or some other undetected minor QTL [58]. However, lines with a combination of the three repeatable QTL from the two parents showing high resistance indicates that pyramiding these minor resistance QTL is an effective approach to improve FHB resistance. Many cultivars and improved lines from different regions may have moderate resistance or moderate susceptibility to FHB [38,39,41,46,57]because they may carry various numbers of minor resistance QTL. We demonstrated that one or two minor resistance QTL in a line may not show a visually distinguishable effect on FHB resistance and may have similar or only slightly lower PSS than lines with no resistance QTL (Fig. 3b). Some moderately susceptible lines may also carry useful QTL for FHB resistance that can be used in gene pyramiding. The advantage of using these sources of resistance is that they are improved lines with desirable agronomic and adaptation traits, and the newly developed lines that have FHB resistance genes from gene pyramiding can be quickly selected for variety release. However, these minor genes individually may not be distinguishable by the conventional phenotyping. Therefore, developing breederfriendly markers tightly linked to these QTL is the key step to deploy these genes in breeding programs [2,16,59].

In this study, we developed 10 KASP assays for the three repeatable QTL, and they all showed expected segregation ratios in the RIL population. Three of the KASP markers that were closest to the three QTL were selected to regroup the mapping population, and the resulting groups showed clear phenotypic differences between contrasting allelic groups and among groups with different allele combinations at the three loci. These KASP markers were also evaluated in a panel of wheat cultivars from the Middle and Lower Valleys of the Yangtze River in China. We found that KASP markers are more likely polymorphic when the donor parents were crossed to most of the cultivars in the panel, suggesting that the KASP markers linked to these QTL should be useful for marker-assisted selection. The KASP marker, GBS17928, at Qfhb.3AL showed a significant correlation with FHB resistance in three independent field experiments, indicating effectiveness of this marker in MAS to improve FHB resistance. In addition,a high frequency of the resistance allele detected in the accessions from Jiangsu province suggests that Qfhb.3AL has been transferred to most of the cultivars in this region.Since Fhb1 have been transferred into many adapted backgrounds [3,60,61], stacking Fhb1 with these minor QTL may quickly improve the FHB resistance levels in new wheat cultivars.

5.Conclusions

Yangmai 158 and Zhengmai 9023 have been two most popular commercial wheat cultivars in China due to their superior agronomic and quality traits as well as stable moderate FHB resistance expressed in large wheat growing areas.They have also been extensively used as parents in many breeding programs in China.Six QTL were identified for FHB resistance on chromosomes 3A,4A,2D(2),7D,and 7A,with three of them repeatable in the two experiments.Additive effects of the QTL from different parents suggest that accumulation of enough minor QTL from modern cultivars can also achieve high levels of resistance. Most of the QTL identified in this study have been reported previously in different sources of germplasm from different countries,which indicate that these minor QTL are likely repeatable QTL and should be very useful for improving FHB resistance in modern cultivars since they are already in adapted cultivars. More importantly, A KASP marker at Qfhb.3AL was significantly correlated with FHB resistance in the diversity panel, which should be useful for improve FHB resistance in breeding programs.

Supplementary data for this article can be found online at https://doi.org/10.1016/j.cj.2020.05.007.

Declaration of competing interest

Authors declare that there are no conflicts of interest.

Acknowledgments

This is contribution number 20-241-J from the Kansas Agricultural Experiment Station. F. graminearum strain F0609 conidial spore suspensions were provided by the Institute of Plant Protection, Jiangsu Academy of Agricultural Sciences, China.This material is based upon work supported partially by the US Wheat and Barley Scab Initiative, the National Research Initiative Competitive Grants (2017-67007-25939) from the National Institute of Food and Agriculture, U.S. Department of Agriculture,the National Natural Science Foundation of China(31671690),the Natural Science Foundation of Jiangsu Province(BK20161375),and the National Key Research and Development Program of China (2016YFD0100502). Any opinions, findings,conclusions,or recommendations expressed in this publication are those of the authors and do not necessarily reflect the view of the U.S.Department of Agriculture.

Author contributions

Pingping Zhang and Guihua Bai designed experiments,interpreted results, and wrote the manuscript. Pingping Zhang, Chenjin Guo, Zhao Liu, Amy Bernardo, Hongxiang Ma, Peng Jiang, and Guicheng Song performed experiments.All authors approved the final version of the manuscript.