Identification and molecular mapping of resistance genes for adult-plant resistance to stripe rust in spring wheat germplasm line PI660076

Wheat is one of the major food crops worldwide. Stripe rust can cause a great loss of wheat yield, especially when the disease is prevalent. Chemical control not only causes the loss of resistance to stripe rust but also has a serious impact on the human body and environment. Therefore, the most economical measure to control wheat stripe rust is to cultivate resistant varieties. Rapid variation of stripe rust races often leads to rapid "loss" of resistance to stripe rust disease; therefore, breeders and researchers have to continuously explore new stripe rust resistance genes to provide new resistance sources for the rapid variation of stripe rust races. Previous studies have confirmed that PI660076, a spring wheat line, showed stripe rust resistance under natural conditions at the adult stage, which has great value in breeding programs. In this study, a recombinant inbred lines (RIL) population was constructed by crossing the wheat line PI660076 with the stripe rust-susceptible line AvS. Genotyping of the population was performed using a wheat 15 K SNP array. Three QTLs were identified using phenotypic data over four years across three environments. The resistance type of each QTL was determined by inoculating the RIL lines with single and homozygous QTL during the seedling and adult stages under controlled conditions. The all-stage resistance (ASR) QTL, QYr076.jaas-2A (flanked by SNP marker AX-11048464 and Kompetitive Allele-Specific PCR (KASP) marker KASP_4940 ) explained 7.13–16.58% and 6.95–7.25% of infection type (IT) and disease severity (DS), respectively. The adult-plant resistance (APR) resistance QTL, QYr076.jaas-4D.1 (flanked by KASP marker KASP_0795 and SNP marker AX-111567243 ,) explained 6.85–12.70% and 7.94–17.26% of IT and DS, respectively. The APR resistance QTL, QYr076.jaas - 4D.2 flanked by KASP markers KASP_9130 and KASP_6535 , explained 7.97–39.19% and 8.77–20.55% of the phenotypic variation in IT and DS, respectively. All the three QTLs are likely to be new. The obtained results lay a foundation for further utilization of the stripe rust-resistant line PI660076, as well as for fine mapping and molecular marker-assisted selection breeding.


Introduction
Stripe rust, caused by Puccinia striiformis f. sp.tritici (Pst), is a typical airborne disease that can spread with high air flow and occurs in almost all wheat growing areas in the world.Once wheat stripe rust breaks out, it will seriously interfere with the normal growth of wheat.In the epidemic year, it will cause serious yield loss and economic loss to wheat production (Chen 2014;Zhou et al. 2022a).Currently, there are two main control strategies for wheat stripe rust.One is chemical control, which has the advantages of quick effect and strong results.Especially in the year of the stripe rust epidemic, it can effectively reduce the loss of wheat yield.However, this method also has clear weakness; specifically, long-term use will not only make the pathogen produce fungicide resistance but will also have an impact on the environment and human health.Compared with chemical control, planting disease-resistant varieties is the most economic and environmentally friendly method; however, the loss of resistance to stripe rust is caused by virulence variation of stripe rust, which is a major problem in resistant breeding programs (Line 2002).
Many stripe rust resistance genes lost effectiveness within a few years due to racial variation in pathogen populations, such as Yr1-Yr4, Yr6-Yr10, Yr17, Yr20-Yr22, Yr24-Yr29 and Yr43 (Chen 2020).These resistance genes belong to the all-stage resistance (ASR) type (also referred to as seedling resistance), which has the disadvantage of being race specific.In contrast, the other type of resistance, adult-plant resistance (APR), has the characteristics of being nonrace-specific and therefore more durable, which can effectively delay the speed of 'loss' of resistance in wheat varieties, such as the wheat cultivar Libellula grown in Longnan, Gansu, China (Zhou et al. 2003) and Alpowa grown in the US Pacific Northwest (Lin & Chen 2007).To date, 84 officially named stripe rust resistance genes and more than 200 QTLs have been identified, most of which belong to ASR, and only a few are APR (Feng et al. 2018;Klymiuk et al. 2022;Ren et al. 2012a;Wang et al. 2022;Zhou et al. 2014Zhou et al. , 2022b)).Therefore, it is important to identify more APR resistance genes in wheat and its close wild relatives.
PI660076 is a spring wheat line obtained by crossing stripe rust resistance donor PI180957 and stripe rust susceptible wheat cultivar ' Avocet Susceptible (AvS)' in USDA-ARS, Wheat Genetics, Quality, Physiology and Disease Research Unit (Wang et al. 2012).Previous studies have found that PI660076 was susceptible at the seedling stage and resistant at the adult stage, with the exception of Tianshui, Gansu Province, China, in 2013, which showed that PI660076 was susceptible in the field at the adult stage (unpublished data).For this interesting situation, one conjecture was that there was probably a race in the field of Tianshui in 2013, which could interact with a host gene in PI660076 and cause susceptibility, but during the next year, this race did not become dominant, and there was no corresponding race to interact with the host gene, resulting in resistance.However, because the races that caused susceptibility to PI660076 in the Tianshui field in 2013 were not collected and identified, there was no way to verify this susceptible phenotype.On the other hand, PI660076 showed resistance in the adult stage but susceptibility at the seedling stage, and it is necessary to identify and map its resistance QTL.
In this study, 208 F 7 generation recombinant inbred lines (RILs) were constructed by crossing PI660076 with AvS and were genotyped with the 15 K SNP array on a whole genome scale.Combined with the genotypic variation and phenotypic identification of stripe rust in the field, molecular mapping of the stripe rust resistance gene in PI660076 was performed.

Plant materials and stripe rust evaluation in field
A RIL population with 208 lines was produced by crossing the stripe rust-susceptible wheat cultivar AvS as the female parent with PI660076 as the donor of stripe rust resistance.The RIL population and the two parents were tested for their reaction to naturally occurring stripe rust in the field at Mianyang (MY), Sichuan in 2019, 2020, 2021and 2022, and Yangling (YL), Shannxi in 2020.In these field trials, each trial was designed by a random complete block design, with three repetitions at each location.Each trial plot was composed of a single row with a length of 1m, and the distance between adjacent rows was 25cm.Each plot was seeded with approximately 40 seeds of a RIL, and a group of parents was seeded every 60 rows to detect the uniformity of stripe rust infection.A row of AvS was sown around each block and served as induced plants for stripe rust.The infection type (IT) and disease severity (DS) data were collected when the susceptible parent AvS and some susceptible lines of the RIL population were fully infected.The classification standard of infection type (IT) is based on the 0-9 scale.The disease severity (DS) refers to the percentage of disease area in leaves (classified by 0%, 1%, 5%, 10%, 20%, 30%, 40%, 50%, 60%, 70%, 80%, 90%, 100%).

Greenhouse test
Two parents and the RIL lines with single and homozygous QTL were selected for seedling test and adult test.Two-leaf stage seedlings and adult plant (flag leaf fully developed) of the lines were inoculated with mixed Pst races collected from infected wheat plants in field.About 10 seeds of each line and parents were planted in a 7 × 7 × 7cm pot filled with soil mixture and grown in a rust-free growth chamber.The inoculated plants were kept in a dew chamber for 24h at 10 C without light, and then the seedling plants were grown in a growth chamber using a low diurnal temperature cycle gradually changing between 4 C at 2:00 am and 20 C at 2:00 pm with 16h light/8h dark.For the adult plants after inoculation, they were grown under controlled conditons using a low diurnal cycle of temperatures gradually changing from 4 C to 20 C and a high diurnal cycle from 10 C to 35 C. Infection type (IT) based on the 0-9 scale was scored for each line 18 to 21days after inoculation when stripe rust was fully developed on AvS.

DNA extraction
Approximately 2 g of young leaf tissue from each RIL and the parents was harvested and dried in a freeze dryer (Thermo Savant, Holbrook, NY, USA) for 48 h and ground using a Mixer Mill (MM 300, Retsch, Germany) for DNA isolation.After the completion of the DNA extraction using a CTAB method as modified by Zhou et al. (2021), DNA was dissolved in Tris-EDTA (10mM Tris-HCl and 1mM EDTA, pH 8.0) buffer.All DNA samples were quantified by an ND-1000 Spectrophotometer (NanoDrop Technologies, Thermo Scientific, Wilmington, DE, USA) and adjusted to a final concentration of 50ng/μl as stock DNA solutions.Stock DNA solutions were further diluted with sterilized ddH 2 O depending on the requirements of the individual experiment.

Genetic map construction and QTL analysis
Genotyping analysis of the RIL population was performed using a wheat 15 K SNP chip assay, and a marker scoring matrix was created in Excel.For mapping analysis, the association between disease trait data and marker data was calculated by the inclusive composite interval mapping (ICIM) method with the software IciMapping v4.1 (Meng et al. 2015).Marker deletion was performed using the "Bin" function, and genetic mapping was performed with the Kosambi mapping function (Kosambi 2016).Phenotypic variation and correlation analysis were also calculated by SAS 9.0 (SAS Inc., Cary, NC).QTL analysis was performed on the basis of line means from individual experiments in each site-year environment.A threshold likelihood of odds (LOD) of 2.5 was set to claim the significance of the QTL (P < 0.01).
The physical location of the QTL was determined according to the physical locations of the SNP markers linked to the QTL.The IWGCS-RefSeq v1.1 was used as a reference to construct a physical map.MapChart v2.3 (Voorrips 2002) was used to draw the genetic and physical reference maps.

Exon capture sequencing analysis
DNA from the parental cultivars was used to perform exome capture sequencing using BGI T7 platform (150bp paired-end reads) according to the wheat exome capture sequencing protocol described by Dong et al. (2020) by Tiancheng Weilai Technology Co., Ltd., (Chengdu, China).After removing adapter sequences, low-quality bases, or undetected bases, clean reads were aligned to the Chinese Spring (CS) reference genome (IWGSC Ref-Seq v2.1) with the BWA software (default parameters) (Li & Durbin 2009) and single nucleotide polymorphism (SNP) and short insertions and deletions (INDELs) were called using the GATK software (default parameters).

Development of KASP markers
Based on the physical positions of the SNPs obtained from the exon capture sequencing analysis, the SNPs in the targeted QTL intervals were converted to KASP markers.KASP primers were developed following standard KASP guidelines.The allele-specific primers were designed carrying the standard FAM (5′-GAA GGT GAC CAA GTT CAT GCT-3′) and HEX (5′-GAA GGT CGG AGT CAA CGG ATT-3′) tails at the 3′ end of the targeted SNP.Common primers were designed so that the total amplicon length was less than 120 bp.KASP assays were performed in 96-well format in 10 μL reaction volumes containing 5 μL HiGeno 2 × Probe Mix, 0.14 μL KASP primer mix (allele-specific primer 1-FAM (12 μM), allelespecific primer 2-HEX (12μM) and common reverse primer (30 μM)), 2 μL genomic DNA at 30 ng μL −1 and 3 μL ddH 2 O. KASP reactions were carried out using Pherastar scanner (LGC Genomics, United Kingdom), with the following PCR cycling protocol: hot start at 95 °C for 10 min, followed by 10 touchdown cycles (95 C for 20 s; touchdown at 61 C initially and decreasing by 0.6 °C per cycle for 40s), followed by 30 additional cycles of annealing (95 C for 20 s; 55 C for 40 s).

Phenotypic characterization of APR Resistance
In all three field trials, the plants in each trial were fully infected.The average IT and DS of the susceptible parent AvS were 8-9 and 91-100%, respectively, while those of the resistant parent PI660076 were 2-3 and 0-10%, respectively (Figs. 1 and 2).The distribution of the average IT and DS of the F 7 RIL population at the adult stage showed continuous unimodal distribution characteristics (Fig. 2), indicating that the disease resistance at the adult stage belonged to quantitative trait inheritance.Pearson correlation coefficients of the mean IT and DS of the F 7 RILs between different environments were all positive and highly significant (P < 0.001) and ranged from 0.46 to 0.80 for IT and from 0.38 to 0.65 for DS (Table 1).The ANOVA showed that the variations in lines, environments and lines × environment for both the IT and DS (Table 2) were all significant (P < 0.001) (Table 2), suggesting that the expression of APR resistance was consistent across the different environments and over different years.The broad sense heritability of IT and DS was 0.9 for IT and 0.83 for DS (Table 2), which indicated that the disease resistance at the adult stage was mainly determined by genotype and had stable gene expression.

Genetic linkage map construction
The wheat 15 K chip assay scanning results showed that a total of 3494 SNP markers were used for linkage map construction, giving a total map length of 14,025.4cM, with individual chromosomes ranging from 473.5 cM for chromosome 6D to 828.9 cM for chromosome 3B (Table 3).The number of markers per chromosome ranged from 40 for chromosome 4B to 710 for chromosome 2A, with an average number of 166 SNP markers.The average distance between neighboring SNP markers ranged from 1.1 cM/marker for chromosome 2A to 16.8 cM/marker for chromosome 4B, with an average number of 6.9 cM/marker.The map was used to identify significant associations between SNP markers and stripe rust resistance.

Exon capture sequencing and development of KASP markers
In order to facilitate the use of QTL-linked markers, we first converted the flanking SNP markers of the QTLs into KASP markers.Although attempts were made to convert 6 SNP sites into KASP markers, only 3 SNP sites (all for the QTL on chromosomes 4D) could be successfully converted to KASP markers.The successfully converted SNP markers were AX-89349130 (converted to KASP_9130), AX-108766535 (converted to KASP_6535), and AX-110080795 (converted to KASP_0795).Therefore, in order to obtain more SNP markers, we did the exon capture sequencing of parental DNA.We selected the SNP sites within the mapping interval, and successfully developed three more KASP markers (KASP_4940, KASP_2719, KASP_0974).The newly developed KASP markers together with the other SNP markers were verified in the RIL population.

QTL for stripe rust resistance
Three major QTLs were consistently detected in the RIL population across 5 experiments on chromosomes 2A and 4D, and the one QTLs was temporarily named QYr076.jaas-2A(Fig. 3) and two QTLs on chromosome 4D were temporarily named QYr076.jaas-4D.1 and QYr076.jaas-4D.2(Fig. 4).QYr076.jaas-2A was located between the SNP markers AX-110484643 and KASP marker KASP_4940 at LOD scores of 3.32-8.24,and its genetic distance was 4.13 cM (Table 4).This QTL explained 7.13-16.58% of the phenotypic variance explained (PVE).The physical locations of the two closely linked markers were 21.13 Mb and 27.90 Mb, respectively (Table 5).The flanking markers for QTL QYr076.jaas-4D.1 were KASP marker KASP_0795 and SNP marker AX-111567243, with AX-111567243 as the closest marker to the QTL, and the PVEs were 6.85-12.70%(Table 4).Its physical location was 24.26 Mb and 46.99 Mb, respectively, and its genetic distance was 29.5 cM (Table 5).The flanking markers for QTL QYr076.jaas-4D.2were KASP markers KASP_9130 and KASP_6535, with KASP_9130 as the closest marker to the QTL, and the PVEs were 7.97-39.19%(Table 4).Its physical location was 379.78 Mb and 436.91 Mb, respectively, and its genetic distance was 7.41 cM (Table 5).The primer information of the KASP markers was provided in Table 6.

Discussion
In this study, we focused on the unidentified source of resistance in PI660076, a line originating from the USA.PI660076 maintains long-term disease resistance under natural conditions in an experimental field in Pullman, Washington, USA (unpublished).To test its resistance to Chinese predominant stripe rust races, PI660076  Through the identification of stripe rust resistance at the seedling stage and adult stage by Zhou et al. (2015), it was found that PI660076 showed no stripe rust symptoms on its leaves in many years from 2007 to the present in the infected field, whereas AvS consistently showed infection symptoms when grown at the same time and under the same conditions, except for the test in Tianshui in 2013, in which PI660076 showed a susceptible phenotype (IT = 8; DS = 80%).Interestingly, PI660076 still maintained good resistance in other years in Tianshui.Compared with other places at the same time and Tianshui in other years, one explanation is that there was a physiological race in the Tianshui area in 2013, and the stripe rust race could interact with genes in the host, which resulted in the original disease resistance genes becoming susceptible.This could be because that Tianshui is in the southwest China, which is suitable for suitable for overwintering and summering stripe rust.In other words, the Tianshui area belongs to the lair of Chinese stripe rust pathogens.Although this is an interesting phenomenon, there is currently no way to identify this susceptible gene.Because PI660076 maintained good stripe rust resistance over multiple years and locations in China, it is necessary to identify and map its resistance genes.This study identified three QTLs for stripe rust resistance in the germplasm PI660076.The QTL QYr076.jaas-2A was located on the short arm of chromosome 2A.Chromosome 2A is rich in resistance QTLs and resistance genes, but almost all of the resistance genes that have been reported are ASR genes.Wheat chromosome 2AS is also a hot spot where genes for stripe rust resistance are enriched.Currently, this region contains more than 20 genes/QTLs having the characteristics of APR, including one officially named APR gene, Yr17 (Bariana & McIntosh 1993), and 24 QTL with APR to stripe rust such as QYr.inra-2A_CampRemy (Mallard et al. 2005), QYr.inra_2AS.1_Recital(Dedryver et al. 2009), QYr.sun-2A_Kukri (Bariana et al. 2010), QYrst.orr-2AS_Stephens (Vazquez et al. 2015), QYr.According to the physical location of molecular markers linked to the genes/QTLs, the general chromosome physical location of the genes/QTLs can be indirectly determined.Among these genes, Yr17 was the only gene that showed both APR and ASR to stripe rust and was located in a region from 7.5 to 21.2 Mb (Beukert et al. 2020;Jia et al. 2011).QYr.inra-2AS_CampRemy was flanked by the SSR markers Xwgm382a and Xgwm359, which correspond to a physical location of approximately 28.20 Mb (Mallard et al. 2005).QYr.Two QTLs (QYr076.jaas-4D.1 and QYr076.jaas-4D.2) were mapped on chromosome 4D; however, there are few SNP markers on 4D; thus, the location range is still large.To date, a total of seven stripe rust QTLs or genes on chromosome 4D have been reported, including Yr28 (Singh et al. 2000) and YrAS2388 (Huang et al. 2011) on 4DS, QTL-4DL from Israeli wheat Oligoculm (Suenaga et al. 2003), Yr46 (Herrera-Foessel et al. 2011) (Sybil et al. 2010), QYr.sun-4DL (Chhetri et al. 2016) and QYr.caas-4DL (Lan et al. 2009;Ren et al. 2012b) on 4DL, and Yr22 located on 4D, but the gene was not designated to a particular chromosomal location (Chen et al. 1995).Among these QTLs, QYr.caas-4DL and Yr46 belong to adult plant resistance and are probably the same locus, but the relationship between the two loci has not yet been verified (Ren et al. 2012b), while the QYr.sun-4DL and Yr46 genes have been confirmed to be the same (Herrera-Foessel et al. 2011).The QTL-4DL from Israeli wheat Oligoculm also confers APR, and this QTL was approximately 26 cM from QYr.caas-4DL (Suenaga et al. 2003).Yr28 belongs to ASR, and Athiyannan et al. (2022) reported that Yr28, as well as the other two genes YrAS2388 and YrAet672, were identified as haplotypes of the locus, all encoding identical protein sequences but are polymorphic in nontranslated regions of the gene.The gene Yr22 also belongs to the ASR (Chen et al. 1995).Thus, based on the currently identified genes/QTLs, there are only 4 APR genes located on 4D, and the relationship between the QTLs on chromosome 4D (QYr076.jaas-4D.1 and QYr076.jaas-4D.2) and the three other QTLs still needs to be verified by subsequent relevant tests.

Fig. 1
Fig. 1 Frequency distributions of the mean infection type (IT) and disease severity (DS) for 208 RILs from the AvS × PI660076 cross grown at Mianyang (MY) and Yangling (YL) in 2019-2022.Arrows indicate the values of the parental lines

Fig. 3
Fig. 3 Stripe rust resistance QTLs QYr076.jaas-4D.1 (A) and QYr076.jaas-4D.2(B) on the genetic map of chromosomes 4D based on IT and DS data.The x-axis is in centimorgan distance (cM), and all genetic distance values shown are in the same scale; the y-axis indicates the LOD value.The red rectangle on the genetic map indicates the corresponding QTL region.AIT: The average of IT.ADS: average of DS

Table 1
Correlation coefficients (r) of the mean infection type (IT) and disease severity (DS) of the AvS × PI660076-derived recombinant inbred lines tested in different environments together with other wheat germplasm resources were introduced into China for stripe rust resistance testing.

Table 2
Analysis of variance and estimation of broad-sense heritability of the infection type (IT) and disease severity (DS) among 208 RILs from the AvS × PI660076 cross tested in Mianyang (MY) and Yangling (YL) in 2019-2022 a σ2g was estimated for genotypic (line) variances b H 2 indicates the estimated heritability in the broad sense on the basis of the mean across replications and environments (or heritability per mean) c Significance level at P < 0.05 (*), P < 0.01 (**), P < 0.001 (***)

Table 3
Summary of chromosome assignment, number of SNP markers, map length, and marker density of the SNP genetic map of the 208 RILs from the AvS × PI660076 cross

Table 4
Summary of stripe rust resistance QTLs identified using ICIM based on the mean disease severity (DS) and infection type (IT) of the 119 RILs from the cross of AvS × PI660076 tested in Mianyang (MY) and Yangling (YL) in 2019-2022 Add additive effect of the resistance allele a LOD logarithm of odds score b PVE percentage of the phenotypic variance explained by individual QTLs c

Table 5
Genetic (cM) and physical (Mb) positions of flanking markers of each QTL identified based on the mean infection type (IT) and disease severity (DS) of the 208 RILs from the AvS × PI660076 cross tested in Mianyang (MY) and Yangling (YL) in 2019-2022

Table 6
Information of the KASP markers developed in this study KASP_6535TTG AAT GGA AAT ACA GGC AGT GCA TGA ATG GAA ATA CAG GCA GTGCC GTT TCC TTT TTT AAT CGG TCA ACC GKASP_0795CGG CCT TCA TGT CTT CGC TA CGG CCT TCA TGT CTT CGC TC CAA AGA TAC ACA TGC ACA CGA ACA