Genetic architecture of quantitative trait loci (QTL) for FHB resistance and agronomic traits in a hard winter wheat population

Yuzhou Xu, Yogung Li, Ruolin Bin, Guorong Zhng, Alln K.Fritz, Ynhong Dong, Lnfi Zho,Yunfng Xu, Ni Ghori, Amy Brnro, Pul St.Amn, Jssi L.Shoup Rupp, Myron Bru,Wi Wng, Eur Akhunov, Brtt Crvr, Guihu Bi,,*

a Department of Agronomy, Kansas State University, Manhattan, KS 66506, USA

b Department of Plant Pathology, University of Minnesota, St.Paul, MN 55108, USA

c Hard Winter Wheat Genetics Research Unit, USDA, Manhattan, KS 66506, USA

d Department of Plant Pathology, Kansas State University, Manhattan, KS 66506, USA

e Department of Plant and Soil Sciences, Oklahoma State University, Stillwater, OK 74078, USA

Keywords:Wheat Fusarium head blight FHB resistance Developmental and yield traits Pleiotropic QTL

ABSTRACT Wheat resistance to Fusarium head blight (FHB) has often been associated with some undesirable agronomic traits.To study the relationship between wheat FHB resistance and agronomic traits, we constructed a linkage map of single nucleotide polymorphisms (SNPs) using an F6:8 population from G97252W × G97380A.The two hard winter wheat parents showed contrasts in FHB resistance, plant height (HT), heading date (HD), spike length (SL), spike compactness (SC), kernel number per spike(KNS),spikelet number per spike(SNS),thousand-grain weight(TGW)and grain size(length and width).Quantitative trait locus (QTL) mapping identified one major QTL (QFhb.hwwg-2DS) on chromosome arm 2DS for the percentage of symptomatic spikelets (PSS) in the spike, deoxynivalenol (DON) content and Fusarium damaged kernel(FDK).This QTL explained up to 71.8%of the phenotypic variation for the three FHB-related traits and overlapped with the major QTL for HT,HD,SL,KNS,SNS,TGW,and grain size.QTL on chromosome arms 2AL, 2DS, 3AL and 4BS were significant for the spike and grain traits measured.G97252W contributed FHB resistance and high SNS alleles at QFhb.hwwg-2DS, high KNS alleles at the QTL on 2AL and 2DS, and high TGW and grain size alleles at QTL on 3AL; whereas G97380A contributed high TGW and grain size alleles at the QTL on 2AL and 2DS, respectively, and the high KNS allele at the 4BS QTL.Combining QFhb.hwwg-2DS with positive alleles for spike and grain traits from other chromosomes may simultaneously improve FHB resistance and grain yield in new cultivars.

1.Introduction

Wheat (Triticum aestivum L.) is an important cereal crop for human nutrition supply in the world.Continuous increase in wheat productivity is critical to meet the growing demand from a rapidly rising world population.Wheat Fusarium head blight(FHB), mainly caused by Fusarium graminearum, is a devastating disease that reduces not only grain yield but also grain quality,and therefore threatens global wheat production [1].Mycotoxins such as deoxynivalenol (DON) produced by the fungus during infection are detrimental to humans and livestock when the contaminated grain is used as food and feed [2].

Growing resistant cultivars is one of the most effective strategies to reduce FHB damage.Wheat FHB resistance can be active,passive, or both [3].Active resistance is usually expressed physiologically or biochemically by activating internal host plant defense mechanisms to suppress pathogen growth and limit the spread of FHB symptoms within wheat spike tissues after initial infection[4]; however, passive resistance is mainly expressed as disease avoidance due to certain morphological traits that create favorable micro-environments to avoid or reduce fungal initial infection,resulting in low FHB infection in host plants [5].Several morphological and developmental traits including plant height(HT),heading date (HD), anther extrusion, and spike compactness (SC) have been discovered associated with plant reactions to FHB [3,6–8].In general, passive FHB avoidance due to morphological features is usually more vulnerable to changes in testing environments than active resistance.

Based on FHB infection,DON content,disease progress in wheat spikes and kernels,and grain yield losses,wheat FHB resistance has also been described as five types [4].Type I is resistance to fungal initial infection.Type II is the resistance to spread of FHB symptoms within an infected spike[9].Miller et al.[10] described Type III resistance as resistance to deoxynivalenol (DON) accumulation in infected kernels.Later, Mesterhazy [5] proposed Type IV resistance as low Fusarium damaged kernel(FDK)and Type V resistance as low yield loss or FHB tolerance.To date, more than 500 quantitative trait loci (QTL) for Types I, II and III resistance have been reported on all 21 wheat chromosomes from various resistant sources [11], and some of them have been frequently associated with undesired developmental and yield traits [3].However, the genetic relationships between these traits and FHB resistance have not been well characterized.Unveiling the genetic relationships among these traits will provide useful guidelines for selecting wheat cultivars with not only a high level of FHB resistance but also desirable agronomic traits for high yield potential in wheat breeding programs.

Wheat kernel number per spike (KNS), spikelet number per spike (SNS), and thousand-grain weight (TGW) are major grain yield components and have higher heritability than grain yield per se; thus, it is more effective to assess grain yield components,which will increase statistical power for detecting QTL for grain yield [12].The objectives of this study are to identify QTL for FHB resistance and related yield traits using a recombinant inbred line (RIL) population and to characterize the relationships among the QTL for those traits.

2.Materials and methods

2.1.Plant materials

A population of 132 F6:8RILs was developed by single seed descent from a cross between two winter wheat lines G97252W and G97380A from Goertzen Seed Research, Inc in KS.The cross was initially made at Oklahoma State University in the mid-2000s to map Rht8.G97252W was found to be moderately FHB resistant,whereas G97380A was highly FHB susceptible (Table S1).The two parents also showed significant differences in plant HT, HD,SC, KNS, SNS, TGW, spike length (SL), grain width (GW), grain length (GL), and grain area (GA).

2.2.Evaluation of FHB and agronomic traits in greenhouses

Two parents and all the RILs were evaluated for type II FHB resistance in four greenhouse experiments in 2019 spring(FHB_GH2019S),fall(FHB_GH2019F)and winter(FHB_GH2019W),and 2020 spring(FHB_GH2020S)at Kansas State University using a randomized complete block design with two replications.Wheat seedlings were vernalized at 6 °C for 50 d and then were transplanted into 14 × 14 cm Dura pots containing Metro-Mix 360 soil mix(Hummert International,Earth City,MO,USA).The greenhouse temperatures were set at 12 ± 2 °C for daytime and 15 ± 3 °C for night during the seedling stage, and changed to 25 ± 3 °C (at day)and 20±3°C(night)three weeks after transplanting.The daylength was set for 12 h with supplemental light.Five plants per line were transplanted into each pot (replication) and fertilized with Miracle-Gro (The Scotts Miracle-Gro Company, Marysville, OH)weekly for four weeks.A conidial spore suspension of F.graminearum was prepared by culturing the F.graminearum strain GZ3639 from Kansas in mungbean broth [13].The final inoculum concentration was adjusted to about 100,000 conidiospores mL-1by counting them in a microscope.At the flowering stage, a 10-μL conidial suspension (1000 conidia/spike) was injected into a central spikelet of a spike using a syringe (Hamilton, Reno, NV).Five spikes were inoculated in each pot and moved into a moist chamber at 100%relative humidity and 20–25°C to initiate fungal infection.After 48 h of incubation, the plants were moved back to the greenhouse benches for disease development.The number of infected spikelets and total number of spikelets per inoculated spike were determined for each plant at 16 d after inoculation.FHB severity was estimated using the percentage of symptomatic spikelets (PSS) in a spike for QTL analysis.In all trials, plant HT was measured from the ground to the top of the spike of the main stem excluding awns before harvesting.HD was recorded when 50% of plants had 50% spikes emerging from the flag leaf sheaths(Feekes 10.1).SL was measured from the base to the top of a spike excluding awns, and SNS was counted before harvest.SC was calculated by dividing the SNS by SL.

The RIL population was evaluated for KNS, TGW, GW, GL, and GA in two greenhouse experiments in spring (Yld_GH2015S) and fall (Yld_GH2015F) 2015 at Kansas State University.The greenhouse yield experiments were conducted using the same design as used for the greenhouse FHB experiments.Spikes from five primary tillers in each pot were collected after maturity and handthreshed to estimate KNS, GW, GL, GA, and TGW using a Marvin seed analyzer (GTA Sensorik GmbH, Germany).A twodimensional image of a seed sample was extracted and the outline of the shadow area was determined.Then, GW and GL were measured along the cross and vertical sections of each seed, respectively.GA was measured by calculating the pixels inside the shadow area.Mean values from two replications were used for QTL mapping.

2.3.Evaluation of FHB and agronomic traits in field trials

The FHB field trial was conducted in the Rocky Ford FHB nursery, Manhattan, KS in the 2019–2020 wheat growing season(FHB_RF2020S).About 30 seeds per line were sown in a 1-m long single-row plot using a randomized complete block design with two replications.The nursery was inoculated by scattering 4 g of F.graminearum-infested corn kernels per plot on the soil surface twice with the first application before the boot stage (Feekes 8)and the second application two weeks later (Feekes 10.1).The nursery was misted using an overhead impact sprinkler system for 3 min hourly from 7 PM to 6 AM daily between flowering(Feekes 10.5.1) and milky ripe (Feekes 11.1) stages to facilitate FHB infection.We rated FHB disease symptoms by visually estimating PSS in the field,and the FHB rating might include both type I and type II resistance because the two types of resistance are usually mixed in field.When highly susceptible lines showed over 90%PSS on more than 50% spikes, PSS was estimated visually for all RILs and parents heading at the same time window within 2 days along with these susceptible lines.HT, HD, SL, SNS, and SC were measured in the field experiment before harvest using the same method as described for the greenhouse experiments.All wheat plots were hand-harvested after maturity and threshed using an Almaco thresher (Nevada, IA, USA) with the air blower open slightly.The collected seeds were manually cleaned to keep as many infected kernels as possible.Samples from the field trial were visually estimated for FDK by comparing the grain samples with a set of controls at 5%, 10%, 20%, 50%, 80%, and 100% FDK.FDK value was determined by two skilled evaluators and averaged for QTL analysis.Ten grams of grain from each line were randomly sampled and ground to a fine powder for DON assay using a gas chromatography-mass spectrometry (GC–MS) at the University of Minnesota.

The RIL population was evaluated for KNS,TGW,GW,GL and GA in two additional field trials in spring 2020 (Yld_AB2020S) and 2021 (Yld_AB2021S) at the Kansas State University Agronomy Farm in Ashland Bottoms, Manhattan, KS.In these field experiments, the RILs were arranged in a randomized complete block design with two replications.For each RIL, 50 seeds were sown as a single row plot of 1.22-m long.Field management followed local practices without irrigation.KNS, GW, GL, GA, TGW, SL, SNS,SC, HT, and HD traits were estimated using the same methods described for the greenhouse experiments.

2.4.DNA extraction and SNP genotyping

Three pieces of 2.0 cm-long wheat leaf tissues were collected at the two-leaf stage from each RIL and parent into 1.3 mL 96-deepwell plates with a 3.2-mm stainless steel bead in each well.The tissues were dried in a freeze dryer (ThermoSavant, Holbrook, NY,USA)for 48 h and ground into a fine powder by shaking the plates at 30 cycles per sec for 3 min in a Mixer Mill(MM300,Retsch,Germany).Genomic DNA was isolated using a modified cetyltrimethyl ammonium bromide (CTAB) protocol [14].Genomic DNA quality was checked by electrophoresis using a 1%agarose gel and quantified using a Quant-iT PicoGreen dsDNA Assay Kit (Thermo Fisher,Waltham, MA) and a FLUOstar Omega microplate reader (BMG Labtech, Germany).The genotyping-by-sequencing (GBS) libraries were constructed using MspI and PstI restriction enzymes following the protocol from Poland et al.[15] and sequenced for three runs in an Ion Proton sequencer (Thermo Fisher, Waltham, MA,USA).The SNPs were called using the reference-based GBSv2 pipeline implemented in the ‘Trait analysis by association evolution and linkage’ (TASSEL) package [16].Only SNPs called from more than 70%of the RILs with heterozygotes<10%and minor allele frequency (MAF) > 20% were used for linkage map construction.

2.5.Linkage map construction and QTL analysis

Initially,1600 GBS-SNPs were used for the first round of linkage map construction and QTL analysis.After the QTL on chromosome arm 2DS was identified,eight additional kompetitive allele specific polymorphic chain reaction(KASP)markers and 12 SSR markers in the 2DS QTL interval were added to the linkage map(Table S2).The KASP primers were designed based on the flanking sequences for four SNPs identified from exome capture (https://wheat.triticeaetoolbox.org/;) [17], three SNPs from the wheat 55K SNP array[18], and one SNP (AX-111561744) from Xu et al.[19].In addition,12 SSR markers that were mapped in the 2DS QTL region in previous studies [20–23] and one KASP marker for a photoperiod gene Ppd-D1 [24] were added to the QTL region.Redundant GBS-SNP markers were removed using the bin function in QTL IciMapping v4.1[25]by keeping the markers with the least missing datapoints in each bin.The genetic linkage map was constructed with a minimum logarithm of odds value(LOD)value of 3.0 using the IciMapping v4.1.Recombination rates were converted to genetic distances in centiMorgans (cM) using the Kosambi function [26].Linkage groups were assigned to corresponding chromosomes based on the physical positions of these markers in the International Wheat Genome Sequencing Consortium (IWGSC) RefSeq v2.1 [27].

QTL mapping was conducted using the inclusive composite interval mapping of additive function (ICIM-ADD) in IciMapping v4.1 [25].The LOD threshold for each trait was estimated by 1000-time permutations using the best linear unbiased prediction(BLUP)values to claim a significant QTL.The QTL effects were estimated as the phenotypic variation explained(PVE)by the QTL calculated by ICIM in IciMapping v4.1.Peak LOD values in the QTL regions were used to estimate the QTL positions.QTL for different traits that were located to the same region or overlapped within the confidence interval were considered the same QTL.QTL significant in at least two experiments were considered relatively stable QTL.

QTL were named following international nomenclature.All QTL names started with‘Q’,followed by a trait designator,a dot,a laboratory designator (HWWG to represent USDA, Hard Winter Wheat Genetics Research Unit), a hyphen (-) and the symbol for the chromosome or chromosome arm on which the QTL resided.If more than one QTL for a certain trait were identified in the same chromosome, a serial number (1, 2, 3, etc.) was added after the chromosome name to show their order in the chromosome from the short arm to the long arm.

2.6.Conversion of GBS-SNPs to KASP markers

The GBS-SNPs within the major QTL interval for FHB resistance were converted to KASP assays (https://biosearch-cdn.azureedge.net/assetsv6/kasp-explanation-fact-sheet.pdf).The PolyMarker software (https://www.polymarker.info/) was used to design genome-specific primers for the KASP markers.Two tail sequences(GAAGGTGACCAAGTTCATGCT and GAAGGTCGGAGTCAACGGATT)were added to the 5′-end of the two allele-specific-forward primers to match with the FAM- and HEX-fluorescence-dye-labeled sequences in the KASP reaction mix.KASP assays were performed in a ProFlex Dual 384-well PCR system(Applied Biosystems,Foster City, CA, USA) using a 4-μL reaction volume including 1.94 μL 2× PACE Genotyping Master Mix (3CR Bioscience, Harlow, Essex,UK),0.06 μL KASP primer mix and 2 μL genomic DNA at 25 ng μL-1.

The PCR started with an initial denaturation step of 94 °C for 15 min, followed by 10 touch-down PCR cycles at 94 °C for 20 s,and 60 °C for 1 min with -0.5 °C/cycle, and then went through 35 cycles of 94 °C for 20 s and 57 °C for 1 min.The PCR products were scanned in a FLUOstar Omega microplate reader (BMG Labtech Inc., Cary, NC, USA) and the signal data were analyzed using the KlusterCaller software(LGC group,Teddington,UK).The newly designed KASP markers were evaluated for polymorphisms between the two parents,and the polymorphic markers were then used to genotype the mapping population to update the linkage map.The new map was used to re-map the QTL and the QTL map was drawn using MapChart v 2.32 [28].

2.7.Statistical analysis

Experimental locations and years were combined as an environmental variable.Histogram and Pearson’s correlation for each trait were calculated using ggplot2 and corrplot package in R, respectively [29,30].BLUP values for agronomic traits across greenhouse and field experiments were calculated separately using a mixed linear model,Y=R+E+G+G×E,implemented in the R packages lme4 [31].In the model, R was replication, E was environment, G was genotype, and G × E was interaction between genotypes and environments.All the variables were considered as random effects.Mean values of two replications in each experiment and the BLUP values were used for QTL mapping.Analysis of variance (ANOVA)was conducted separately for greenhouse and field experiments in all replicated trials using aov function (R Core Team, 2021).The broad-sense heritability was calculated using the formula H2= VG/[VG+ VG×E/E + Ve/(R × E)], where VGwas genotypic variance, VG×Ewas variance of G × E, Veis the residual variance, E is the number of environments, R is the number of replications.For FHB traits in greenhouse, the same model and analysis were conducted as above.For FHB traits in field, only one environment(FHB_RF2020S)was used to calculate the BLUP value for QTL mapping using the following model, Y = R + G.ANOVA was conducted using the same model.The broad-sense heritability was calculated using the formula H2= VG/(VG+ Ve/R).

To remove possible confounding effects of HD and HT on FHB traits, FHB traits were corrected by regression analysis using HD and HT as covariate factors.In greenhouse experiments, the corrected BLUP values of PSS were calculated using a mixed linear model: Y = R + E + G + G × E + HD + HT.In the field experiment,the corrected BLUP values of PSS, FDK and DON content were calculated using the mixed linear model: Y = R + G + HD + HT.Both models were implemented in the R packages lme4 [31] with all the variables as random effects.The corrected BLUP values were used to re-map QTL.The broad-sense heritability was corrected using the formula H2= VG/[VG+ VG×E/E + Ve/(R*E) + VHD+ VHT] for the greenhouse experiments but the formula H2= VG/(VG+ Ve/R) + VHD+ VHT] for the field experiment, where VHDand VHTwere the variance of HD and HT, respectively.

3.Results

3.1.Phenotypic variation for wheat FHB resistance and other traits

In the greenhouse experiments, genotypic (G), environmental(E) effects and the G × E interactions were significant (P < 0.01)for all the traits measured in the RIL population except KNS, FDK and DON(Table S3).In the field experiments,genotypic(G)effects were significant(P<0.01)for all traits(Table S3).The environmental (E) effects were significant (P < 0.01) for all agronomic traits.The G×E interactions were significant(P<0.05)for most of these traits except GL and KNS.Continuous distributions of the BLUP values were observed for three FHB traits (PSS, FDK and DON), and nine agronomic traits (HT, HD, SL, SC, KNS, SNS, GA, GW and GL)in the RIL population evaluated in both greenhouse and field experiments (Figs.1, 2).BLUP values for most of traits fit normal distribution with exceptions of HD and SNS that showed bimodal distributions in both greenhouse and field experiments.The heritability was high for FHB traits (65%–88%) and agronomic traits(70%–96%) based on BLUP values (Table S3), indicating that major portion of the variance for these traits was heritable.

HT significantly affected PSS in both the greenhouse and field environments as well as DON in the field,whereas HD showed significant effects on PSS in the greenhouses and FDK in the field(Table S4).Therefore, HD and HT were used as covariates for ANOVA of FHB traits.The results indicated that genotypic effects of PSS,FDK and DON remained significant(P<0.01)in both greenhouse and field experiments, but the heritability of PSS, FDK and DON were significantly reduced (Tables S3, S4), indicating that HD and HT might have confounding effects on FHB traits in greenhouse and field conditions.

3.2.Correlations among FHB resistance and agronomic traits

In the greenhouse experiments, PSS positively correlated with SC and kernel traits (TGW, GA, GW and GL) (0.31 < r < 0.42,P < 0.01), but negatively correlated with HD, HT, SNS and SL(-0.82 < r < - 0.55, P < 0.01).Since plant HD or HT showed confounding effect on FHB resistance,we adjusted the FHB phenotypic data using HD and HT data.After adjustment, the corrected correlation coefficients between PSS and those traits were dramatically reduced (-0.56 < r < 0.30) but remained significant (P < 0.01)except for SC, suggesting that FHB resistant lines in general had later HD, taller plants, more SNS and longer SL but lower TGW and grain size than FHB susceptible lines in the population under greenhouse conditions (Table 1).

Fig.1.Distribution of best linear unbiased prediction (BLUP) values of Fusarium head blight (FHB) and agronomic traits evaluated in the greenhouse experiments.PSS,percentage of symptomatic spikelets in a spike;HD,heading date;HT,plant height;SNS,spikelet number per spike;SL,spike length;SC,spike compactness;TGW,thousand grain weight; GA, grain area; GW, grain width; GL, grain length.

Fig.2.Distribution of best linear unbiased prediction(BLUP)values of Fusarium head blight(FHB)and agronomic traits evaluated in the field experiments.PSS,percentage of symptomatic spikelets in a spike;FDK,Fusarium damaged kernel;DON,deoxynivalenol;HD,heading date;HT,plant height;SNS,spikelet number per spike;SL,spike length;SC, spike compactness; TGW, thousand grain weight; GA, grain area; GW, grain width; GL, grain length; KNS, kernel number per spike.

In the field experiment, highly positive correlations(0.65 < r < 0.85, P < 0.01) were observed among PSS, FDK and DON(Table 1).The three FHB traits also showed significantly negative correlations (-0.64 < r < -0.41, P < 0.01) with HD, HT, SNS and SL, but positive correlations with GW (r = 0.26–0.31,P < 0.01).The correlations were not significant between PSS and other kernel traits.Additionally, DON content showed significant positive correlations with TGW, GA and GW (0.23 < r < 0.31,P < 0.01), and both FDK and DON content were negatively correlated with KNS (r = -0.29 and-0.32,respectively,P< 0.01).After correction using HD and HT data,the PSS and FDK remained correlated with HD (-0.28 and -0.24, P < 0.01), but not with HT(Table 1).The corrected DON content showed dramatically reduced correlations with HD(-0.32)and HT(-0.30),but remained significant (P < 0.01; Table 1).

3.3.QTL for FHB resistance

Only one QTL(QFhb.hwwg-2DS)with a major effect on PSS(FHB severity) between markers KASP2D58574820 and KASP-Ppd-D1 on chromosome arm 2DS was significant in all four greenhouse experiments and two BLUP datasets (Tables 2, S5), which explained 22.9% to 71.8% of the phenotypic variation for PSS in different experiments.KASP-Ppd-D1 is a diagnostic KASP marker for gene Ppd-D1, a major photoperiod response gene regulating wheat heading date [32].The QTL for FDK and DON overlapped withQFhb.hwwg-2DS and explained 38.8% and 45.1% of the phenotypic variation for the two traits, respectively, in the field experiment.G97252W contributed the resistance alleles at this QTL for all the three traits, suggesting QFhb.hwwg-2DS is a major FHB resistance QTL with a pleiotropic effect on all three FHB traits (Tables 2, S5).

Table 1 The correlation coefficients between Fusarium head blight (FHB) traits and agronomic traits based on best linear unbiased prediction (BLUP) values before and after correction with plant height (HT) and heading date (HD).

The QTL on 2DS from G97252W remained highly significant and explained 28.9% of the phenotypic variation for PSS (QFhb.hwwg-2DS)in the greenhouse environments and 17.1%of the phenotypic variation for DON(QDon.hwwg-2DS)after removing the confounding effects due to HT and HD(Table S6).In the field conditions,the FHB resistance QTL was overlapped with QDon.hwwg-2DS and explained 11.7% of the phenotypic variation, but the QTL for FHB severity was mapped in a slightly different position from QFhb.hwwg-2DS identified from greenhouse data (Table S6).

3.4.QTL for yield-related traits

Three QTL were detected for KNS on chromosomes 2AL,2DS and 4BS(Tables 2,S5).QKns.hwwg-2DS showed the largest effect in two field experiments and the field BLUP dataset, explained 19.1 to 25.1%of the phenotypic variation.Ppd-D1 is within the QTL region.QKns.hwwg-2AL explained 8.4% to 11.3% of the phenotypic variation and QKns.hwwg-4BS explained 9.6%to 10.3%of the phenotypic variation in one field experiment and the field BLUP dataset.G97252W contributes alleles for increased KNS at QKns.hwwg-2DS and QKns.hwwg-2AL but an allele for decreased KNS at QKns-4BS.The same QTL interval on 2DS also showed a major effect on SNS, explaining 27.9% to 75.1% of the phenotypic variation in all greenhouse and field experiments with the increased SNS allele from G97252W.

Three QTL were significant for TGW and GA(Tables 2,S5).QTgw.hwwg-2DS showed the largest effect on TGW and GA,explaining up to 35.8% and 34.7% of the phenotypic variation, respectively, in greenhouse and field experiments.QTgw.hwwg-2AL explained up to 21.3% and 28.7% of the phenotypic variation for TGW and GA,respectively.QTgw.hwwg-3AL explained up to 12.9% and 16.1% of the phenotypic variation for TGW and GA, respectively.G79252W contributes high TGW and large GA alleles at QTgw.hwwg-3AL, whereas G97380A contributes the positive alleles at the other two QTL.

Two QTL were detected for GW(Table S5).G97380A contributes the wide grain alleles at both loci.Five QTL were detected for GL(Table S5).G97252W contributes the long grain alleles at QGI.hwwg-3AL and QGl.hwwg-5AL and G97380A contributes the long grain alleles at other three QTL.

3.5.QTL for other traits

Six QTL were detected for plant HT: two each on chromosome arms 2DS and 3AL, and one each on 2DL and 6BL (Table S5).QHt.hwwg-2DS.2, close to Ppd-D1 and overlapping with QFhb.hwwg-2DS, showed the largest effect, explaining 22.0% to 48.9% of the phenotypic variation, and was significant in two greenhouse experiments, three field experiments and two BLUP datasets.QHt.hwwg-2DS.1 including Rht8 for reduced plant height in this region was significant in three greenhouse experiments,three field experiments and two BLUP datasets, explaining 6.4% to 25.3% of the phenotypic variation.QHt.hwwg-2DL explained 7.4% to 14.0%of the phenotypic variation in one greenhouse experiment, two field experiments and two BLUP datasets.Other QTL (QHt.hwwg-3AL.1, QHt.hwwg-3AL.2 and QHt.hwwg-6BL) explained 3.8% to 11%of phenotypic variation in some of the greenhouse and field experiments.G97252A contributes the short alleles at all loci except for

QHt.hwwg-2DL and QHt.hwwg-6BL.

Two QTL on chromosome arms 2DS and 7DS were significant for HD in at least two experiments (Table S5).Among them, QHd.hwwg-2DS showed the largest effect in all experiments,explaining 41.3% to 84.4% of the phenotypic variation and Ppd-D1 within the QHd.hwwg-2DS interval might be the causal gene.QHd.hwwg-7DS showed only a minor effect and explained only 4.1% to 4.9% of the phenotypic variation.G97252A carries early heading alleles at both loci.

Five SL QTL were mapped on chromosome arms 2DS, 3AL and 6BS, respectively (Table S5).G97252W contributes the long spike alleles at all loci except QSl.hwwg-6BS.Four QTL for SC were significant on chromosome arms 2DS, 3AL, 7BL and 7DS (Table S5).G97380A contributes the compactness alleles at QSc.hwwg-2DS,QSc.hwwg-3AL and QSc.hwwg-7BL.

3.6.QTL clusters for multiple traits

A total of four QTL clusters were discovered on chromosome arms 2AL, 2DS and 3AL in the mapping population (Table S7).Two QTL clusters for different traits were mapped on chromosome arm 2DS (Fig.3).The cluster 2DS-1 including Rht8 that is flanked by Xgwm261 (20.4 Mb) and KASP2D26715133 (26.7 Mb) in IWGSC RefSeq v2.1 [27] contains overlapping QTL for HT, SL and SC.The cluster 2DS-2 including Ppd-D1 that is flanked by KASP35014114(35.0 Mb) and KASP2D64237023 (64.2 Mb) contains QTL for PSS,FDK and DON and all nine agronomic traits (KNS, SNS, TGW, GA,GW,GL,HT,HD and SL).The cluster 2AL contains QTL for five kernel traits (KNS, TGW, GA, GW and GL) and was flanked by GBS2A_460238480 (460.2 Mb) and GBS2A_694523139 (694.5 Mb).The cluster 3AL flanked by GBS3A_522003888 (522.0 Mb) and GBS3A_667967807(668.0 Mb)contains QTL for six agronomic traits(TGW, GA, GL, HT, SL and SC).

Fig.3.Partial genetic map(left)and physical map(right)based on Chinese Spring RefSeq v2.1 for chromosome 2D to show the quantitative trait locus(QTL)regions (black bars in the linkage map) for multiple traits (QTL names and intervals on the left).

G97252W contributes positive alleles for FHB resistance and KNS and SNS, but negative alleles for TGW and grain size (GL and GW) at the cluster 2DS-2 (Table S7).Similarly, G97252W contributes alleles for more KNS,but lower TGW and smaller grain size at the cluster 2AL.However, G97252W contributes alleles for higher TGW and larger grain size at the cluster 3AL without adverse effects on FHB resistance,KNS and SNS.The QTL for different traits at the same locations may have pleiotropic effects on these traits or may be tightly linked.

4.Discussion

4.1.QFhb.hwwg-2DS is a stable major QTL for FHB type II resistance

Sumai3 and its derivatives are derived from Chinese landraces and have been extensively used as sources of FHB resistance in wheat breeding programs worldwide [33].However, exotic germplasm shows poor adaptability to U.S.wheat growing environments, which limits the deployment of exotic FHB resistance genes.Native resistance genes may provide a better alternative for developing locally adapted FHB resistant varieties [34,35].However, unfavorable association between FHB resistance and key agronomic traits makes more difficulty for their adoption in U.S.hard winter wheat improvement [36,37], because significant pre-breeding work needs to be done before they can be used in breeding.

In this study, marker analysis indicated that G97380A carries the semi-dwarfing Rht8 allele and the photoperiod-insensitive Ppd-D1a allele; while G97252W carries the alternative alleles for the two genes.QFhb.hwwg-2DS for three FHB resistance traits was mapped in the vicinity of the two genes regulating plant HT and HD, respectively, on chromosome arm 2DS of G97252W in the G97252W × G97380A population.QFhb.hwwg-2DS showed major effects on PSS, FDK and DON content, and explained up to 71.8%of the phenotypic variation for PSS in the greenhouses using the single floret inoculation [11,13] and 22.9% in the field where plants were inoculated by Fusarium infected corn spawn(Table S5).In the field condition, FHB resistance of QFhb.hwwg-2DS may be contributed by both type I (resistance to initial infection)and type II (resistance to FHB spread within a spike)because ascospores produced from infected corn spawn randomly landed on wheat spikes and each spike could have multiple initial infection sites.However, in greenhouse conditions, even a larger effect on FHB resistance was detected for the QTL than that in the field.In the greenhouse,the single floret inoculation evaluates type II resistance only, therefore QFhb.hwwg-2DS is most likely a QTL mainly for type II resistance.

QFhb.hwwg-2DS was mapped between markers KASP2D30932191 (30.9 Mb) and KASP2D58574820 (58.6 Mb)(Tables 2, S5).Previously, McCartney et al.[38] reported three FHB resistance QTL on 2DS (QFhb.crc-2D.1, QFhb.crc-2D.2 and QFhb.crc-2D.3) between Xgwm261 (20.4 Mb) and Xgwm484(50.6 Mb)based on IWGSC RefSeq v2.1.The Canadian spring wheat‘Kenyon’contributed the FHB resistance alleles at all three QTL.The QTL for multiple FHB traits were also mapped in the same 2DS region of several Canadian winter wheat cultivars and a U.S.soft winter wheat ‘Truman’ in other studies [39–41].Based on the physical positions of the flanking markers (Table S6), QFhb.hwwg-2DS identified in this study is most likely the same as those previously reported on 2DS.Consistent detection of QFhb.hwwg-2DS in diverse germplasm indicates that QFhb.hwwg-2DS is a stable QTL in North American wheat and QFhb.hwwg-2DS reduces not only FHB disease severity but also DON content in harvested grain in different genetic backgrounds and testing environments.The KASP markers, KASP2D35014114 and KASP-Ppd-D1, flanking QFhb.hwwg-2DS can be used to select QFhb.hwwg-2DS in wheat breeding.

4.2.Relationship between QFhb.hwwg-2DS and other agronomic traits

Fig.4.The overlapping quantitative trait locus(QTL)on 2DS based on best linear unbiased prediction(BLUP)values of Fusarium head blight(FHB)traits,heading date(HD)and plant height (HT) from greenhouse (GH) and field (FD) experiments without masking confounding effects.

In the QFhb.hwwg-2DS region, QTL were also detected for HT(QHt.hwwg-2DS.2) and HD (QHd.hwwg-2DS).QHd.hwwg-2DS was mapped at the Ppd-D1 locus (Figs.3, 4; Table S5), suggesting that Ppd-D1 is most likely the causal gene; whereas QHt.hwwg-2DS.1 for plant HT was detected in the Rht8 position, indicating Rht8 is the major contributor to the plant height variation in the QTL region.In field conditions, tall and late headed plants might have reduced FHB infection due to disease escaping mechanisms under natural infection conditions [11].However, QFhb.hwwg-2DS was still highly significant with a major effect on PSS and DON content in the field conditions after removing the effects of HD and HT(Tables S5, S6), demonstrating the QTL for FHB resistance is real.In the greenhouse trials,the plants were manually inoculated using point inoculation, therefore, disease infection and development should not be affected by HD and HT under relatively controlled environments.To validate this, QFhb.hwwg-2DS was re-mapped with the greenhouse FHB data after adjusted by HD and HT data.The results showed that the QTL was still highly significant, but the effect was significantly reduced (Tables S5, S6).Those results indicate that QFhb.hwwg-2DS is a real QTL for FHB resistance that may be either tightly linked to or pleiotropy of Ppd-D1b and Rht8.Previously, Xu et al.[19] reported the association between semi-dwarfing Rht8 allele and increased FHB susceptibility.Rht8 was physically located in the 25.6 Mb position on 2DS of Chinese cultivar ‘Y8679’ [42], which is about 10 Mb from QFhb.hwwg-2DS.Factors such as difference in mapping populations and FHB evaluation conditions may cause shift of QTL positions in different experiments.The results from this study cannot determine the causal gene for QFhb.hwwg-2DS and further fine mapping in the QTL region may provide insight on the relationship between QFhb.hwwg-2DS and Ppd-D1b or Rht8.

Besides HD and HT genes, three QTL (QSns.hwwg-2DS, QKns.hwwg-2DS and QTgw.hwwg-2DS) for SNS, KNS, TGW on 2DS also overlapped with QFhb.hwwg-2DS in this study (Table 2;Fig.3).Previously, QTL for SNS [43–45], KNS [46] and TGW [47,48]were detected in the QFhb.hwwg-2DS region in several wheat cultivars or landraces.Based on their physical positions,QSns.hwwg-2DS,QKns.hwwg-2DS and QTgw.hwwg-2DS are likely the same QTL as previously reported.QFhb.hwwg-2DS is likely the QTL with pleiotropic effects on SNS, KNS and TGW or linked to the QTL for these traits.

In the 2DS-2 QTL cluster where Ppd-D1 is located, 12 QTL were identified in the current study(Fig.3;Table S7).This QTL cluster in G97252W showed increased FHB resistance, improved KNS and SNS,but reduced TGW and grain size,suggesting that high spikelet fertility decreases host vulnerability to Fusarium infection in spikes and high KNS and SNS usually lower TGW.QFhb.hwwg-2DS in G97252W contributed to taller plants and later HD that are not desired in modern cultivars(Table S7).Fortunately,3AL QTL cluster in G97252W showed increased TGW and grain size without negative effects on FHB resistance, KNS and SNS (Table S7), thus this QTL can be pyramided with QFhb.hwwg-2DS to improve FHB resistance and reduce yield penalty (Table S8).To balance the adverse effects of later HD due to Ppd-D1,QFhb.hwwg-2DS can be deployed in climatic regions of high latitude where have strong winter,long daylength during wheat heading and late onset of hot-dry weather at grain filling stage to maximize yield potential.In addition,QFhb.hwwg-2DS can be pyramided with positive alleles for yield traits at different QTL from different sources to simultaneously improve FHB resistance and agronomic traits.Also, breaking the unfavorable linkage between QFhb.hwwg-2DS and Ppd-D1b using cytogenetic and genomic tools is possible if they are linked genes.

Table 2 Quantitative trait locus (QTL) mapped using best linear unbiased prediction (BLUP) values calculated for Fusarium head blight severity (FHB), Fusarium damaged kernel (FDK),deoxynivalenol (DON), spikelet number per spike (SNS), thousand grain weight (TGW) and kernel number per spike (KNS) evaluated in both greenhouse (GH) and field (FD)experiments.

4.3.Other QTL for grain yield component traits

In the current study, two overlapping QTL for KNS and TGW were mapped on 2AL (QKns.hwwg-2AL, QTgw.hwwg-2AL) between GBS2A_460238480 (460.2 Mb) and GBS2A_626218088 (626.2 Mb)(Table 2).The two QTL are either tightly linked or pleiotropic.G97252W contributed alleles for more kernels per spike but lower TGW.Shi et al.[49]identified a SNP IWB7310(616.6 Mb)that was associated with a QTL for KNS and Liu et al.[50]found a SNP B4170(522.6 Mb)for a TGW QTL in the region.These QTL were located in the same region, and are most likely the same QTL as the ones identified in this study.

QKns.hwwg-4BS was physically mapped between 18.5 Mb and 20.5 Mb (Table 2) in this study, which may be the same QTL reported by Li et al.[51]because IWB45065 is a tightly linked marker to both QTL (18.5 Mb).QTgw.hwwg-3AL was flanked by GBS3A_648390973 (648.4 Mb) and GBS3A_659255761 (659.3 Mb)on 3AL (Table 2), which is likely the same QTL reported by Yang et al.[52].

In this study,SNP and SSR markers were developed for the QTL for both FHB resistance and agronomic traits(Tables S2,S9).Some of the markers can be used in marker-assisted selection (MAS) in breeding or in fine mapping of those QTL to further dissect the genetic relationship between FHB resistance and agronomic traits.

Declaration of competing interest

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

CRediT authorship contribution statement

Yuzhou Xu:Conceptualization,Writing–original draft,Investigation, Data curation, Formal analysis.Yaoguang Li:Genotyping,Data curation.Ruolin Bian:Investigation, Data curation.Guorong Zhang:Field experiment.Allan K.Fritz:Field experiment.Yanhong Dong:DON assay.Lanfei Zhao:Investigation,Data curation.Yunfeng Xu:Investigation, Data curation.Nida Ghori:Investigation, Data curation.Amy Bernardo:Genotyping, Data curation.Paul St.Amand:Genotyping, Data curation.Jessica Rupp:Field experiment.Myron Bruce:Field experiment.Wei Wang:Data curation, Software.Eduard Akhunov:Data curation, Software.Brett Carver:Resources.Guihua Bai:Conceptualization,Resources, Supervision, Writing – reviewing & editing.All authors revised and approved final version of the manuscript.

Acknowledgments

This is contribution number 23-251-J from the Kansas Agricultural Experiment Station.This project was partially supported by the U.S.Wheat and Barley Scab Initiative and the National Research Initiative Competitive Grants (2022-68013-36439) from the National Institute of Food and Agriculture,U.S.Department of Agriculture(USDA).Mention of trade names or commercial products in this publication is solely for the purpose of providing specific information and does not imply recommendation or endorsement by the USDA.USDA is an equal opportunity provider and employer.

Appendix A.Supplementary data

Supplementary data for this article can be found online at https://doi.org/10.1016/j.cj.2023.09.004.