Stripe rust, caused by Puccinia striiformis Westend. f. sp. tritici Erikss. (Pst) remains one of the most significant diseases of wheat worldwide. We investigated stripe rust resistance by genome-wide association analysis (GWAS) in 959 spring wheat accessions from the United States Department of Agriculture-Agricultural Research Service National Small Grains Collection, representing major global production environments. The panel was characterized for field resistance in multi-environment field trials and seedling resistance under greenhouse conditions. A genome-wide set of 5,619 informative SNP markers were used to examine the population structure, linkage disequilibrium and marker-trait associations in the germplasm panel. Based on model-based analysis of population structure and hierarchical Ward clustering algorithm, the accessions were clustered into two major subgroups. These subgroups were largely separated according to geographic origin and improvement status of the accessions. A significant correlation was observed between the population sub-clusters and response to stripe rust infection. We identified 11 and 7 genomic regions with significant associations with stripe rust resistance at adult plant and seedling stages, respectively, based on a false discovery rate multiple correction method. The regions harboring all, except three, of the QTL identified from the field and greenhouse studies overlap with positions of previously reported QTL. Further work should aim at validating the identified QTL using proper germplasm and populations to enhance their utility in marker assisted breeding.
Citation: Muleta KT, Bulli P, Rynearson S, Chen X, Pumphrey M (2017) Loci associated with resistance to stripe rust (Puccinia striiformis f. sp. tritici) in a core collection of spring wheat (Triticum aestivum). PLoS ONE 12(6): e0179087. https://doi.org/10.1371/journal.pone.0179087
Editor: Aimin Zhang, Institute of Genetics and Developmental Biology Chinese Academy of Sciences, CHINA
Received: January 24, 2017; Accepted: May 23, 2017; Published: June 7, 2017
This is an open access article, free of all copyright, and may be freely reproduced, distributed, transmitted, modified, built upon, or otherwise used by anyone for any lawful purpose. The work is made available under the Creative Commons CC0 public domain dedication.
Data Availability: All relevant data are within the paper, its Supporting Information files as well as the Triticae cap website: https://triticeaetoolbox.org/wheat/
Funding: This project was funded by the National Research Initiative Competitive Grant 2011-68002-30029 (Triticeae-CAP) from the USDA National Institute of Food and Agriculture, the Monsanto-Beachell Borlaug Fellowship program, the Washington Grain Commission, and through Hatch project funds to Washington State University. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors declare that they have no competing interest.
Stripe rust, caused by Puccinia striiformis Westend. f. sp. tritici Erikss. (Pst), is one of the most devastating diseases of wheat worldwide [1–3]. Historically, stripe rust was mainly a problem of wheat production in areas with cool and wet weather conditions. In recent years, however, the emergence of aggressive and high-temperature-adapted strains of the pathogen has played a significant role in changing the epidemiology of stripe rust to include areas with climates normally considered unfavorable for the disease development [4,5]. These new strains of Pst are currently widespread and reduce wheat production on the global scale, heightening a particular concern towards the economic impact of stripe rust on wheat yield and global food security.
Use of genetic resistance is the most effective and economic method for long-term management of stripe rust. Genetic resistance in wheat to stripe rust is categorized as either race-specific or non-race-specific. Race-specific resistance in most cases can be detected at the seedling stage and remains effective at all stages of plant growth, therefore, is often referred to as seedling or all-stage resistance based on “R” genes. Many cultivars with resistance controlled by R genes are usually short-lived due to the emergence of virulent races in the pathogen population. The transient nature of the R-gene mediated resistance has been responsible for the occurrence of many devastating epidemics throughout the world [2,6–9]. On the other hand, non-race-specific resistance genes are mainly expressed at the post-seedling stage of plant development. Therefore, it is often referred to as adult plant resistance (APR). In many cases, APR genes tend to be more durable (remain effective even when used over large acreage for several years) than seedling resistance genes in controlling stripe rust.
In response to “boom-and-bust” cycles, a phenomenon associated with the deployment of single R genes followed by the evolution of matching virulence in the pathogen, much effort has been invested in identifying new sources of durable resistance to stripe rust. Deployment of cultivars carrying combinations of race-specific and non-race-specific resistance genes is a long-term goal for many breeding programs. This approach provides a complex resistance gene landscape against the dynamics of pathogen virulence and should promote durability [7,8,10]. The prospect of deploying resistance gene combinations is being increasingly facilitated by the recent advances in genomics and statistical methods that provide highly effective marker tagging systems and efficient means of genome manipulation [8,11,12].
Employing efficient methods of genetic analysis to facilitate the identification of genomic regions underlying traits of economic and biological importance in a diverse germplasm accessions is a goal for effective utilization of diversity in crop improvement programs. Genome-wide association study (GWAS), a widely used approach for detecting quantitative trait loci (QTL) in plants [13–17], investigates genotype-phenotype correlation by taking advantage of linkage disequilibrium as well as historical recombination present within the gene pool of a species. An additional advantage of GWAS is the detection of QTL with greater resolution from populations of diverse origins, thereby eliminating the need for the time-consuming process of developing mapping populations.
Wheat germplasm collections of the primary gene pool maintained in germplasm banks include landraces, breeding lines and traditional cultivars, which offer access to a diverse range of phenotypes such as disease resistance [18,19]. In some cases, these accessions co-existed with the rust pathogens under a natural evolutionary arms race, which might have resulted in a diversifying selection and accumulation of complex resistance loci . These genetic materials may possess potentially untapped sources of useful genetic resistance owing to their limited use in modern plant breeding programs. The usefulness of wheat landraces and other germplasm maintained in gene banks as a good source of resistance to diseases in wheat have been demonstrated [13,14,16–18,21,22]. Here, we hypothesized that the global collection of the spring wheat accessions maintained by the United States Department of Agriculture-Agricultural Research Service National Small Grains Collection (USDA-ARS NSGC) are useful genetic resources that provide a wide range of diversity for Pst resistance. The present study addresses the following three objectives: (1) to assess the diversity of stripe rust resistance in a world-wide collection of spring wheat accessions, (2) to carry out a genome-wide search for single-nucleotide polymorphism (SNP) loci associated with resistance to current Pst populations in the Pacific Northwest region of the United States, and (3) to establish relationships between the Pst resistance loci identified in this study and previously identified Yr genes and QTL.
Materials and methods
Wheat germplasm resources
A total of 1,163 spring wheat (Triticum aestivum ssp. aestivum) accessions were provided by USDA-ARS National Small Grains and Potato Germplasm Research Unit (Aberdeen, Idaho). The accessions were collected from 91 countries in six contents, including Asia (38.8%), Europe (23.2%), Africa (18.9%), South America (14.0%), North America (2.8%), Australia and New Zealand (1.8%) (Fig 1). Accessions with unknown origin account for 0.5% of the population. We excluded accessions with greater than 10% missing genotypic data from the final analyses. Genetically redundant accessions were also culled based on Identity-By-Descent (IBD) kinship analysis, and represented by only one individual accession. Accordingly, a total of 959 non-duplicate accessions were used for the analyses.
Phenotypic trait evaluation
Field-based resistance screening.
Adult plants of the accessions were tested for response to Pst infection under field conditions at two locations in the Pacific Northwest (PNW) of the U.S. during 2012–2014 crop seasons. The locations were Mount Vernon, WA (48° 25’ 12”N; 122° 19’ 34”W), a high rainfall area located west of the Cascade Mountain range; and Pullman, WA (46° 43’ 59'' N; 117° 10’ 00”W), a semi-arid wheat belt area located east of the Cascade Mountain range. The nursery locations are subject to high disease pressure on an annual basis, but vary in Pst populations and weather patterns. The highly susceptible cv. 'Avocet S' was planted every 20 rows and on each side of the plot to ensure uniform disease pressure across the experimental plots. Infection type (IT) to stripe rust was recorded using a 0–9 scale . The SEV and IT data collected from Pullman in 2013 were excluded from subsequent analyses due to low and uneven disease development during that particular season. Phenotypic data are available at the Triticeae Toolbox https://triticeaetoolbox.org/wheat/.
Seedling resistance screening.
Races PSTv-14, PSTv-37 and PSTv-40, representing the prevalent races of Pst in the PNW and across the U.S. , were used to screen seedlings of the accessions under controlled conditions in a greenhouse. The virulence/avirulence formulae of the three races are given in S1 Table. The seedlings were evaluated following the standard procedure described in .
DNA extraction and genotyping
Genomic DNA was extracted from leaf tissues of 10–15 days old seedlings of the 1,163 spring wheat core collection at the USDA-ARS Small Grains and Potato Germplasm Research Unit. DNA extractions were performed using the CTAB protocol . The DNA was precipitated by adding isopropanol, followed by washing of the pellet with ice-cold 70% ethanol, and resuspension in 200 μL of Tris HCl ethylenediaminetetraacetic acid (pH 8.0).
Whole genome profiling of the accessions was performed using the Illumina iSelect 9K wheat assay at the USDA-ARS Wheat Genotyping Laboratory, Fargo, ND. Genotype calling of the original Illumina data was carried out using Illumina’s GenomeStudio v2011.1 software to optimize the SNP call rates for misclassification and ambiguous clusters. After removing SNPs with low-quality clustering and those with minor allele frequency (MAF) less than 5%, a total of 5,619 high quality SNP markers with genetic map information were used for GWAS analyses. The genetic positions of the SNP markers were based on the wheat 9K SNP consensus map . Genotypic data are available at the Triticeae Toolbox https://triticeaetoolbox.org/wheat/.
For controls and validation, the collection was screened with molecular markers for the known stripe rust resistance genes. These include, Lr34/Yr18/Pm38 (KASP wMAS000003) (http://maswheat.ucdavis.edu/protocols/Lr34/index.htm), Lr37/Yr17/Sr38 (marker Ventriup-LN2) (http://maswheat.ucdavis.edu/protocols/Lr37/index.htm), Lr67/Yr46 (Kasp856) , Yr30/Sr2 (KASP marker wMAS000005) (http://maswheat.ucdavis.edu/protocols/Sr2/index.htm) and Lr68 (CAPS marker cs7BLNLRR)  (http://maswheat.ucdavis.edu/protocols/Lr68/index.htm). The MAF for each control marker, except for Lr37/Yr17/Sr38 and Yr30/Sr2, were greater than 5%, and were thus included in the GWAS and linkage disequilibrium analyses.
Population structure and linkage disequilibrium
Population structure was investigated using the Bayesian model-based clustering algorithm implemented in the STRUCTURE software version 2.2.3 . A set of 425 SNP markers distributed across wheat genetic map with an inter-marker distance of >10 cM were used to infer population structure in the germplasm panel. A burn-in of 50,000 iterations and 100,000 Monte Carlo Markov Chain (MCMC) replicates were set to determine K values (number of subpopulations) in the range of 1 to 10. For each K, five independent runs were carried out. The Evanno method  was used to determine the likely number of subpopulations using STRUCTURE HARVESTER . Principal component analysis (PCA) was also performed using JMP GENOMICS software (JMP®, Version 6.1. SAS Institute Inc., Cary, NC) to further explore the pattern of population sub-structuring and compare with the STRUCTURE results.
Pairwise measures of linkage disequilibrium (LD) between pairs of SNP markers were estimated using JMP GENOMICS version 6.1. LD was estimated as squared allele frequency correlations (r2) between pairs of intra-chromosomal SNPs with known chromosomal position and minor allele frequency (MAF) > 0.05. To determine the average pattern of genome-wise LD decay over genetic distance, a scatterplot of r2 values against the corresponding genetic distance between markers was constructed. The second-degree locally weighted polynomial regression (LOESS)-based curve was fitted to estimate the extent of LD decay .
Estimation of BLUPs and variance components
Best linear unbiased predictors (BLUPs) were estimated for each accession at each location using the phenotypic trait values available for each test year as a predictor. BLUPs were estimated using a mixed effects model by fitting years as a fixed effect and accessions as a random effect using the lme4 package  in R (R Development Core Team). The mixed model was also used to estimate variance components for genotypes, environments, and genotype by environment interactions. Trait heritability (h2) estimates were calculated using the Restricted Maximum Likelihood (REML) method .
Marker-trait association was performed using the compressed mixed linear model (cMLM) method [35,36] implemented in GAPIT (Genomic Association and Prediction Integrated Tool) in R environment . In the mixed model analysis, the 959 × 959 kinship matrix and the first two principal components determined by PCA were used as covariates to control spurious associations due to cryptic relatedness and population structure, respectively. A marker-wise association probability value of P <0.01 in at least two of the five test environments was used to declare significant marker-trait associations (MTAs). To minimize the chance of false positive MTAs, associations significant at genome-wide adjusted P <0.1 based on the False Discovery Rate (FDR) multiple correction method  were also identified. Inter-marker distance ≥ 2.5 cM, which is genomic distance below which genome-wide LD was predicted to decay, was used to establish QTL confidence interval (CI) of ±2.5 cM. Among the adjacent SNPs, the QTL-tagging SNP was selected as the one that showed the strongest P value.
Stepwise regression analysis was also performed to identify the best combinations of significant alleles explaining the variation for stripe rust resistance in the germplasm panel. Stepwise regression also facilitates the selection of markers that have a major effect in a QTL region and simultaneously exclude other markers in LD with the major marker within the confidence interval of the QTL region. A significant P value threshold of 0.1 and 0.05 was used in the stepwise regression analysis for the inclusion and exclusion of the markers in the analysis, respectively. Finally, multiple linear regression analysis was performed to assess the amount of phenotypic variation explained by the significant markers. The markers were used as the independent variable and BLUP values of IT and SEV from the field trials were included as the response variable.
Relationship between number of favorable alleles and response to stripe rust
Pearson correlation analyses were performed to assess the relationship between the number of favorable alleles of the significant QTL-tagging SNPs accumulated in a single accession and stripe rust resistance. Favorable and alternate alleles for the QTL-tagging markers that were significant at P < 0.01 in at least two environments were coded as 1 and 0, respectively, and the cumulative numbers of favorable and the alternate alleles were counted for each accession for use in the correlation analysis.
No specific permits were required for the described field studies. No specific permissions were required for these locations/activities. The study locations are not privately-owned or protected in any way. In addition, the field studies did not involve endangered or protected species.
Estimates of variance components and trait heritability
Mean responses of the accessions to stripe rust as well as estimates of variance components and broad sense heritability are summarized in Table 1. The mixed linear model analysis using the SAS procedure PROC MIXED revealed highly significant differences among the genotypes and genotype × environment interactions for SEV and IT (P <0.0001). Using the BLUP values to group the accessions into categories based on IT, 18% of the accessions displayed high resistance reactions (IT = 0–3), while 24% of the accessions were considered highly susceptible (IT = 7–9) across all environments. The remaining 58% of the accessions showed either an intermediate reaction (IT = 4–5) across all environments, or variable responses at different test environments (Fig 2). Heritability (H2) values for IT were 0.84 and 0.86 in Pullman and Mount Vernon, respectively, while the heritability estimates for SEV were 0.80 in Pullman and 0.90 in Mount Vernon. Consistent with the high heritability estimates, we also observed highly significant correlations for IT and SEV among the five test environments (S2 Table).
A) Best linear unbiased prediction (BLUP) values of the infection type (IT) responses of adult plants under field condition, B) BLUP values of the severity (SEV) responses of adult plants under field condition, and C) seedling resistance screening against three Puccinia striiformis f. sp. tritici races (PSTv_14, PSTv_37 and PSTv_40) under greenhouse experiments. MTV_SEV = Mount Vernon severity, MTV_IT = Mount Vernon infection type, PLM_SEV = Pullman severity, PLM_IT = Pullman infection type.
Population structure and linkage disequilibrium
Both STRUCTURE and PC analyses grouped the accessions into two major clusters (Fig 3, S1 Fig). Furthermore, the distance-based Fast Ward hierarchical clustering also revealed a similar grouping pattern. Population structure was also evident when the accessions were grouped by geographic origin and improvement status (ACIMPT). The first sub-population (sub-population 1) composed of accessions mainly from Europe, North America and South America, while the second sub-population (sub-population 2) constituted landraces collected from Asia. Fig 3 (C) depicts a two dimensional scatterplot of the first two PCs of the spring wheat accessions superimposed with geographic origin and improvement status of the accessions. A significant correlation (P <0.001 and r ranging from 0.11 to 0.30) was observed between the population sub-clusters and response to stripe rust resistance, justifying the use of a GWAS model accounting for population structure in the panel.
A) Dendrogram based on Ward clustering of the spring wheat core collection, B) pairwise kinship matrix depicting clustering of the accessions based on identity-by-decent (IBD), C) population structure based on principal component analysis (PCA). Both genetic relatedness and principal component analyses grouped the accessions into two subpopulations, Subpopulation 1 (SP1) and Subpopulation 2 (SP2). Clustering pattern based on PCA also explained geographic origin and improvement status of the spring wheat core collection. D) Effect of population structure on stripe rust infection type (IT) and severity (SEV). Box plots show trait distribution and compare the levels of stripe rust between the two subpopulations. MTV_12 = Mount Vernon 2012, MTV_13 = Mount Vernon 2013, MTV_14 = Mount Vernon 2014, PLM_12 = Pullman 2012, PLM_13 = Pullman 2013, PLM_14 = Pullman 2014.
Genome-wide LD was predicted to decay below the critical r2 = 0.36 at an inter-marker distance of 2.5 cM (Fig 4). Average r2 value was 0.55 for the completely linked marker pairs (markers with zero inter-marker distance). Among the non-completely linked marker pairs whose genetic distance was less than 1 cM, the r2 value decreased to 0.3. This indicates a decrease in r2 value to half of its initial value within 1 cM. Among the marker pairs in LD due to linkage (r2 > 0.36), chromosome 2B contained the highest percentage (17.2%) of these markers, while chromosome 4D contained the lowest percentage (less than 1%). The proportion of marker pairs in the A, B and D genomes that are in LD at r2 >0.36 were 55.3, 41.7 and 7.0%, respectively (Fig 4).
Marker-trait association and annotation
Field-based resistance to stripe rust.
GWAS identified a total of 130 SNPs associated with resistance to stripe rust at adult plant stage when filtering loci significant at marker-wise P <0.01 in at least two of the five test environments. Among the 130 significant SNPs, multiple co-locating SNPs were assigned to a unique QTL region based on the following approaches. First, a confidence interval for assigning the putative resistance loci were determined based on a threshold of the inter-marker distance (cM) at which LD has decayed below the critical r2 value. Among the multiple SNPs within a confidence interval of a QTL, the SNP that showed the strongest association and significant in the stepwise regression analysis was selected as the QTL-tagging SNP. Of the 130 significant SNPs identified, 70 fit into the stepwise regression model and also corresponded with the SNPs that showed the strongest association (smallest P value) in the GWAS test; each represented a unique QTL region, which was determined based on the confidence interval for determining putative resistance loci (±2.5 cM around the QTL-tagging SNP) (S3 Table). The 70 QTL-tagging SNPs explained 51% and 52% of the total variation for stripe rust IT and SEV, respectively. QTL regions were detected on all chromosomes except on 3D and 7D. Detailed information from the GWAS and stepwise regression tests for the 70 QTL is presented in S3 Table and S4 Table, respectively.
Eleven of the 70 QTL-tagging SNPs were significant at FDR-adjusted P <0.10. The 11 genomic regions were represented by SNPs IWA1191 and IWA5861 (on chromosome 1B at 23.7 and 94.2 cM respectively), IWA3621 and IWA1040 (on chromosome 2B at 114.3 and 210.2 cM, respectively), IWA6843 (on chromosome 3B at 111.6 cM), IWA5707 (on chromosome 4D at 25.7 cM), IWA1755 (on chromosome 5B at 130.4 cM), IWA167 (on chromosome 6D at 16.8 cM), IWA7306 and IWA1845 (on chromosome 7A at 3.9 and 42.5 cM, respectively) and IWA3415 (on chromosome 7B at 164.9 cM). The resistance alleles of the 11 genomic regions showed average allelic effect of reducing stripe rust responses ranging from 0.3 to 1.8 for IT and 2.2 to 23.0 for SEV (Fig 5). Collectively, these QTL explained 24% of the observed variation for both IT and SEV. Detailed information on the 11 QTL is summarized in Table 2. The integrated map developed by Maccaferri et al.  was used to establish the relationship between the 11 QTL with significant genome-wide associations (FDR-adjusted P <0.1) with previously mapped Yr genes and QTL. Accordingly, any Yr genes and QTL previously mapped within the ±2.5 cM confidence interval of the 70 QTL detected in the present study were identified and are presented in Table 2. QTL with confidence intervals not overlapping with that of previously reported Yr gene or QTL were considered to represent newly-discovered Pst resistance loci. Two of the 11 loci identified from the field studies (IWA1755 on 5BL and IWA7306 on 7AS) were mapped far from previously identified Pst resistance genes and QTL and likely represent new resistance loci.
Blue colored boxes indicate effect due to the resistance associated alleles, while the red colored boxes indicated effect due to the unfavorable allele.
Seedling stage resistance.
Seven loci were significantly associated with IT response to the three races of Pst at seedling stage (Table 3). These SNPs were detected on chromosome 4B (IWA2194 mapped at 49.4 cM), 5A (IWA2145 and IWA1258 mapped at 19.7 cM and 235.1 cM, respectively), 5B (IWA7815 mapped at 167.8 cM), 6A (IWA2129 mapped at 212.2 cM) and 7B (IWA312 and IWA2770 mapped at 76.1 cM and 246.5 cM, respectively). SNPs IWA2194, IWA1258, IWA2129 and IWA2770 were also identified in GWAS of field resistance to Pst at the nominal probability of association. The putative resistance loci represented by IWA2194, IWA2145, IWA7815 and IWA312 were effective to the three races of Pst. Six of the seven genomic regions significant for conferring seedling resistance to stripe rust were mapped closely to known resistance genes and QTL, while one (IWA2145 on chromosome 5A) was mapped far from previously identified Pst resistance genes and QTL.
Relationship between number of favorable alleles and response to stripe rust
The number of favorable alleles of the SNPs representing the 70 genomic regions in each of the 959 accessions varied from 22–50. When accessions were sorted in descending order by the number of favorable alleles, stripe rust SEV and IT in the top 10% accessions (with average number of favorable alleles of 41) were 65.1 and 62.0% lower than the bottom 10% accessions (with average number of favorable alleles of 26). We investigated the relevance of the addition of every favorable allele in an accession in predicting the accessions’ stripe rust response (their additive effect to enhance Pst resistance) by correlating BLUP values of IT and SEV on the cumulative number of beneficial alleles of the accessions. There was a highly significant negative correlation (P <0.0001, r = 0.67) between the number of favorable alleles in individual accessions and the respective stripe rust IT and SEV (Fig 6).
Effect of previously mapped Yr genes
Allelic frequencies of markers linked to Lr67/Yr46, Lr34/Yr18, Yr17, Yr30 and the leaf rust resistance gene Lr68 are presented in S2 Fig. The Lr34/Yr18-diagnostic marker was present in 8.5% of the accessions. The resistant allele of this locus reduced IT and SEV by 1.5 (25%) and 18 (32%), respectively. Similarly, association analysis in which the Lr34 KASP marker data was included detected a strong association of the Lr34/Yr18 locus with reactions to Pst (FDR adjusted P value < 0.1) at multiple environments. No significant SNP markers were detected in the 7D region where Lr34/Yr18 resides. Likewise, none of the SNPs on chromosome 7D were in significant LD with wMAS000003, indicating the lack of significant MTAs on 7D in this study was due to inadequate marker coverage of the chromosome.
The minor allele frequency of the marker Kasp856 (tightly linked to the Yr gene Lr67/Yr46) in this germplasm panel was 7.8%. The Yr46 gene reduced IT and SEV by 2.3 (38%) and 24.6 (44%), and showed strong association with Pst resistance (FDR adjusted P value < 0.1) in multiple environments. Similar to the report by Maccaferri et al.  and Forrest et al. , we also observed that Kasp856 is in LD with IWA5707 (r2 = 0.40) and other associated SNPs within the confidence interval of the QTL tagged by IWA5707 (IWA6277, IWA5375 and IWA5766) (r2 = 0.31, 0.33 and 0.39, respectively).
Yr30/Sr2 was present in about 4% of the accessions (S2 Fig) that were collected largely from countries from South America. The resistance-associated allele of the marker for Yr30/Sr2 reduced stripe rust by an average of 12.1%. Since the minor allele frequency of the Yr30/Sr2 marker was less than 5%, it was excluded from the genome-wide association analyses.
In order to determine whether the slow rusting leaf rust resistance gene, Lr68, influences reaction to Pst, we performed GWAS analyses and calculated allelic effects associated with the marker closely linked to Lr68 (cs7BLNLRR). The marker had an effect of reducing stripe rust IT and SEV by 1.5 (25%) and 15 (27%), respectively. Similarly, GWAS showed significant association of the cs7BLNLRR with reactions to Pst (marker-wise P <0.01) in three of the five environments. Further genetic analysis is required to determine whether the observed allelic effect is conferred by Lr68 itself or another linked stripe rust gene/QTL.
Response of spring wheat accessions to stripe rust
The limited availability of well-characterized and effective sources of stripe rust resistance in the elite wheat germplasm pool has constrained the capacity for routine development of varieties with durable resistance. Identifying new sources of resistance and incorporating them into adapted cultivars is therefore a pressing priority to sustainably enhance stripe rust resistance breeding. The present study characterized a large and diverse sample of the NSGC spring wheat core collection for seedling and field-based resistance to the prevailing populations of Pst in the Pacific Northwest of the US. These accessions exhibit considerable variation for both seedling and field-based resistance to Pst. Among the 464 accessions that showed strong to moderate resistance, with IT ratings <5, 387 accessions were resistant only at post-seedling stage of growth. This may indicate that the resistance in these accessions is likely conferred by APR genes/QTL. Mining of the useful stripe rust resistance genes in such germplasm resources and incorporating them into breeding programs should enhance the durability of released cultivars and mitigate losses due to the disease.
Population structure of the spring wheat accessions
In the present study, population structure of the 959 accessions from the spring wheat core collection was used as covariate in the GWAS analysis to reduce the likelihood of false positive associations. Structure analysis was also used to describe the effect of geographic origin and improvement status of the accessions in generating patterns of adaptive structure and clustering related to stripe rust resistance. The model based clustering algorithm of structure analysis and the distance-based measure of cryptic relatedness revealed the clustering of the panel into two major subpopulations and discovered a substantial agreement between the patterns of population sub-clustering and information of geographic origin and accessions types of the NSGC spring wheat core collection. The two sub-populations reveal a major division between accessions from Europe, North America and South America (sub-population 1) and accessions from Asia (sub-population 2) and grouping between landrace lines and advanced breeding lines and cultivars. This pattern of population clustering and genetic relatedness among the spring wheat accessions may reflect the impact of unique selection pressures and adaptation of the accessions in each of the diverse environments during the process of domestication and expansion of cultivated T. aestivum [14,39].
GWAS of stripe rust resistance using the high-density molecular marker information and Pst resistance data from multi-environmental field trials and greenhouse experiments provided a basis for comprehensive analysis of the NSGC spring wheat core collection. A total of 70 genomic regions significant at marker-wise P <0.01 in at least two of the five environments were identified that were retained in the stepwise regression model. Considering the experiment-wise correction for multiple testing (FDR adjusted P <0.1), 11 of the 70 genomic regions were declared significant for field-based resistance to stripe rust. Seven additional regions were significant at FDR P <0.1 for seedling resistance. The 11 high-confidence genomic regions explained on average 24% of the total variation in stripe rust IT and SEV; the trait variation explained by the 70 marker-wise significant SNPs was 52%. Hence, although the 70 SNP loci may exhibit some proportion of false positives, the increase in the amount of variation explained by the 70 genomic regions from 24% to 52% suggests the presence of additional true associations. The highly significant negative correlation (P <0.0001, r = 0.67) between the number of favorable alleles in individual accessions and the respective stripe rust IT and SEV also support the above hypothesis.
Alignment of the putative QTL to previously mapped Yr genes/QTL
The integrated genetic map information constructed by Maccaferri et al.  was used to compare resistance loci detected in the current study with previously mapped Yr genes/ QTL. Three of the genomic regions that were significant at FDR <0.10, IWA1755 and IWA7306 from the field experiments and IWA2194 from the seedling screening, were mapped far from the previously identified stripe rust resistance genes and QTL. Hence, these three genomic regions most likely tag new stripe rust resistance loci. The genomic regions associated with the remaining 9 putative QTL from the field experiments and six loci from the seedling resistance screening were mapped close to known stripe rust resistance genes and QTL. The relationship of the 15 significant genomic regions with previously mapped Yr genes and QTL are discussed below.
SNP IWA1191 was mapped to the proximity of several previously mapped Yr genes on the short arm of chromosome 1B. The genetic map position of IWA1191 (23.7 cM) was at a relative genetic distance of 3.1–6.8, 4.6–6.3, 4.6–14, 6.3–7.7 and 5.9–23.4 cM from YrAlp, Yr15, Yr64, YrCH52 and Yr24/Yr26 [40–43]. Evidently, IWA1191 represents a genomic region associated with a resistance locus that was not effective against any of the three Pst races at seedling stage, indicating that the associated locus is likely an APR QTL. However, all of the Yr genes previously mapped in this region of 1BS confer major gene or all-stage resistance to stripe rust. Hence, the genomic region tagged by SNP IWA1191 is likely linked to a different QTL conferring effectiveness to stripe rust at post seedling stage. Another temporarily designated stripe rust APR QTL, QYr.caas-1BL.1RS_SHA3/CBRD, was also mapped close to the region of IWA1191 (4.6–6.3 cM) , which could be related to IWA1191.
SNP IWA5861 was mapped at 70.5 cM proximal to IWA1191. Based on its genetic map position, it is unlikely that IWA5861 is related to YrAlp, Yr15, Yr64, YrCH52 and Yr24/Yr26 stripe rust resistance genes. Yet, several other temporarily designated QTL have been identified within the confidence interval of the QTL represented by IWA5861. These include QYr.sun-1B_CPI133872, QYr.sun-1B_Kukri, QYr.sun-1B_Wollaroi, QYr.cim-1BL_Francolin, QYr.tam-1B_Quaiu and QYr-1B_Saar [45–49]. Allelism tests will be required to determine the genetic relationship of the two resistance loci linked to the genomic region represented by IWA5861 and IWA1191 with previously mapped QTL. The position of IWA5861 also overlaps with the race-specific gene YrExp1 . However, the current races of Pst in the Pacific Northwest region are virulent on YrExp1. Hence, IWA5861 is most likely different from YrExp1. The position of IWA5861 is also close to IWA3017 and IWA5915, which showed significant association in studies by Maccaferri et al.  and Bulli et al. , respectively. In the present wheat population, LD between IWA5861 and IWA3017 was determined to be 0.50, indicating that the two SNPs are tagging the same putative stripe rust resistance locus. In the study by Maccaferri et al. , IWA3017 showed only a marker-wise level of association with stripe rust resistance. The strong association of IWA5861 in the present study (FDR <0.1) confirms that IWA5861 and IWA3017 tag a validated QTL. However, LD between IWA5861 and IWA5915 is less than 0.002, indicating that the two SNPs are putatively linked to two different stripe rust resistance QTL.
IWA3621 was mapped on chromosome 2B at 114.3 cM, which is in the confidence interval of several previously identified stripe rust QTL. These QTL include YrKK, QYr.inra-2B.1_Camp Remy, QYr.ucw-2B_UC1110, QYr.tam-2BL2_TAM111, QYr-2B_Opata 85, Qyrlo.wpg-2B_Louise and QYrid.ui-2B.2_IDO444 [51–56]. It is therefore likely that IWA3621 represents a genomic region associated to either one or more of these QTL. On the long arm of 2B, IWA1040 mapped (210.1 cM) close to the region of Yr53 (5–11 cM) and Yr43 (9–18 cM) [57,58]. Yr53 confers major gene resistance to stripe rust that was derived from durum wheat. Yr43 is a single dominant gene conferring race-specific all-stage resistance to stripe rust. However, IWA1040 is likely linked to an APR QTL as it was ineffective at seedling stage and hence may represent a different QTL from the two genes. Another QTL, QYraq.cau-2BL_Aquileja, was mapped near IWA1040 by Guo et al. , which requires further genetic analysis to determine their relationship.
IWA6843 was mapped at 111.6 cM on chromosome 3B. Although no officially designated Yr genes have been reported near this locus, two temporarily designated stripe rust QTL were previously mapped within the confidence interval of the QTL tagged by IWA6843. These QTL are QYr.cim-3B_Pastor  and QYr.inra-3Bcentr_Renan . Further genetic analysis will be required to determine the relationship between IWA6843 and the two previously mapped QTL.
IWA2194 was significantly associated with resistance against races PSTv-14, PSTv-37 and PSTv-40, as well under field conditions at marker-wise level of probability < 0.01. It was mapped to the proximity of a previously mapped QTL, QYr.ufs-4B_Palmiet . However, QYr.ufs-4B_Palmiet is linked to a minor effect APR QTL unlike IWA2194, which tags a major effect seedling resistance to Pst. Hence, it is likely that IWA2194 and QYr.ufs-4B_Palmiet are different.
On the short arm of chromosome 4D, we detected a significant association of IWA5707 and other linked SNPs (IWA6277, IWA5375 and IWA5766). Forrest et al.  reported an identification of significant association of IWA5707, IWA5375 and IWA5766 with resistance to stripe rust. Similarly, Maccaferri et al. (2015) also reported a significant association of IWA5375 and another SNP (IWA5766) linked with stripe rust resistance. Based on moderate to strong LD between IWA5707, IWA5766 and the KASP marker csSNP856, Forrest et al.  determined that these markers are linked to the Lr67/Yr46 locus. The identification of such previously reported significant marker-trait association further validates our current GWAS analysis.
Two SNP loci were significantly associated with seedling resistance to Pst races; IWA2145, mapped at 19.7 cM, and IWA1258 mapped at 235.1 cM. IWA2145 is likely linked to a new Pst resistance locus as there are no previously mapped Yr genes or QTL close to this locus. IWA1258 is mapped close (4.3–25.0 cM) to two previously mapped QTL. These are QYr.cim-5AL_Pastor  and QYr-5A_Opata_85 . Both are minor QTL for APR, unlike IWA1258, which likely tags a major effect seedling resistance gene.
On 5BL, IWA7815 and several other associated SNPs represent a locus most strongly associated with seedling resistance to all three Pst races. This locus was not effective under field conditions; possibly, due to the presence of other virulent races than the three used in this study. Other QTL including QYr-5B_Oligoculm  and YrEXP2  were previously mapped close to the region of IWA7815. QYr-5B_Oligoculm represents a minor gene-based APR, while YrExp2 is a seedling resistance gene effective against some Pst races, but ineffective to the three races used in our GWAS analysis, indicating that the IWA7815 tags a locus different from the two previously mapped QTL.
On 6AL, we identified SNP IWA2129 that was significant for seedling resistance to PSTv-14 and PSTv-37, and effective under field conditions at a threshold of marker-wise P <0.01 in at least two of the five environments. Its effectiveness both at seedling and adult plant stages may indicate that it is a major seedling resistance gene effective against multiple races of the pathogen. Other Yr genes and QTL that have been previously reported close to IWA2129 include YrLM168 , QYr-6A_Avocet , QYr-6A_Saar , QYr.ufs-6A_Kariega  and QYr.cim-6AL_Francolin , which requires further genetic analysis to determine the relationship between the gene tagged by IWA2129 and previously identified QTL.
SNP IWA167, mapped on the short arm of 6D at 16.8 cM, is one of the 11 strongly associated genomic regions and has previously been reported by Maccaferri et al. . Detailed information on the IWA167 locus can be referenced from the previous study .
Two stripe rust resistance regions were identified on chromosome 7A; IWA7306 at 6.2 cM, and IWA1845 at 42.5 cM, Rosewarne et al.  and Crossa et al.  reported a temporarily designated QTL (QYr.cim-7AS_Avocet and QYr.cim_7A.1_GWAS, respectively), which were mapped in close proximity to the IWA7306 locus. Similarly, Zwart et al.  and Ren et al.  reported two QTL, QYr.sun-7A_CPI133872 and QYr.caas-7A_Jingshuan16, that were mapped close to IWA1845. The identity or similarity of the genomic region tagged by IWA7306 and IWA1845 with previously mapped QTL needs to be further investigated. Although Yr61  and Yrxy1  have already been mapped on the short arm of chromosome 7A, both IWA7306 and IWA1845 were mapped far from Yr61 and Yrxy1, and likely represent different resistance loci.
On the long arm of chromosome 7B, SNP IWA3415 was effective across four field trials and effective at the seedling stage. It was mapped to the vicinity of Yr59, YrC591, Yr52, Yr67 and YrZH84 [40,71–75]. Yr52 and Yr59 confer high-temperature adult-plant (HTAP) resistance, while YrC591, Yr67 and YrZH84 confer seedling resistance to stripe rust. IWA3415 is likely more related to the latter three Yr genes due to its effect at a seedling stage. On chromosome 7B IWA312 (76.1 cM) and IWA2770 (151.5 cM) were also significant for seedling resistance at experiment-wise threshold of P <0.1. IWA312 could be related to the previously mapped stripe rust QTL QYr-7B_Oligoculm , but further genetic analysis will be required to determine their relationship. IWA2770 was also significant under field conditions at marker-wise P <0.01 across multiple environments. LD r2 between IWA3415 and IWA2770 was 0.03, which indicates that they likely represent different loci. Allelism tests will be required to determine the relationship between the significant SNP loci on 7B and the previously mapped genes on this chromosome.
The results of the present studies emphasize the prospect of exploiting the high genetic diversity and extensive LD due to historical recombination in wheat germplasm collections to identify genomic regions underpinning resistance to stripe rust. The USDA NSGC spring wheat accessions exhibited a wide range of phenotypic diversity for field-based and seedling resistance to stripe rust. Accessions with a higher percentage of stripe rust resistance-associated alleles are valuable genetic resources that could serve as parental breeding lines to enable more efficient breeding for stripe rust resistance. The molecular markers linked to QTL identified in the current GWAS studies will be of considerable interest for marker-assisted selection in wheat breeding. These genomic regions provide the initial step towards a quantitative, methodical exploitation of untapped genetic diversity in germplasm collection for wheat improvement. Allelism tests will be required to validate these QTL by using bi-parental populations or near-isogenic lines (NILs) and to determine which of the identified QTL represent novel resistance genes and which ones are alleles of previously mapped genes.
S1 Fig. Estimation of the most likely number of clusters in the germplasm panel based on the Bayesian clustering model in structure software.
Mean of Estimated Log-likelihood and the magnitude of delta K (ΔK) were plotted against K values based on five independent runs and K ranging from 1 to 10.
S2 Fig. Allelic frequency of the markers linked to previously mapped stripe rust resistance genes in the spring wheat association mapping panel.
S1 Table. Virulence/avirulence formula of the stripe rust isolates used for seedling resistance screening.
S2 Table. Pearson’s correlation coefficients between the five test environments for stripe rust infection types (IT) and severity (SEV).
S3 Table. SNP markers tagging QTL for resistance to stripe rust under field condition in a global collection of spring wheat accessions identified based on marker-wise P value < 0.01 in at least two environments for both IT and SEV in the GWAS test.
This project was funded by the National Research Initiative Competitive Grant 2011-68002-30029 (Triticeae-CAP) from the USDA National Institute of Food and Agriculture, the Monsanto-Beachell Borlaug Fellowship program, the Washington Grain Commission, and through Hatch project funds to Washington State University. John Kuehner, Victor DeMacon and Dr. Kent Evans are appreciated for providing technical support.
- Conceptualization: MP.
- Formal analysis: KTM.
- Funding acquisition: MP.
- Investigation: KTM PB SR.
- Project administration: KTM MP.
- Resources: MP.
- Writing – original draft: KTM.
- Writing – review & editing: KTM PB SR XC MP.
- 1. Boyd LA. Can Robigus defeat an old enemy?–Yellow rust of wheat. J Agric Sci. 2005;143: 233.
- 2. Chen XM. Epidemiology and control of stripe rust [Puccinia striiformis f. sp. tritici] on wheat. Can J Bot. 2005;27: 314–337.
- 3. Wellings CR. Global status of stripe rust: A review of historical and current threats. Euphytica. 2011;179: 129–141.
- 4. Hovmøller MS, Sørensen CK, Walter S, Justesen AF. Diversity of Puccinia striiformis on cereals and grasses. Annu Rev Phytopathol. 2011;49: 197–217. pmid:21599494
- 5. Sørensen C, Hovmoller M, Leconte M, Dedryver F, de Vallavieille-Pope C. New races of Puccinia striiformis found in Europe reveal race-specificity of long-term effective adult plant resistance in wheat. Phytopathology. 2014;2011: 1–37.
- 6. Steele KA., Humphreys E, Wellings CR, Dickinson MJ. Support for a stepwise mutation model for pathogen evolution in australasian Puccinia striiformis f.sp. tritici by use of molecular markers. Plant Pathol. 2001;50: 174–180.
- 7. Hulbert S, Pumphrey M. A time for more booms and fewer busts? Unraveling cereal-rust interactions. Mol Plant-Microbe Interact. 2014;27: 207–214. pmid:24499028
- 8. Burdon JJ, Barrett LG, Rebetzke G, Thrall PH. Guiding deployment of resistance in cereals using evolutionary principles. Evol Appl. 2014;7: 609–624. pmid:25067946
- 9. Ellis JG, Lagudah ES, Spielmeyer W, Dodds PN. The past, present and future of breeding rust resistant wheat. Front Plant Sci. 2014;5: 641. pmid:25505474
- 10. Chen X. High-temperature adult-plant resistance, key for sustainable control of stripe rust. Am J Plant Sci Biotechnol. 2013;4: 608–627.
- 11. Tanksley SD. Mapping polygenes. Annu Rev Genet. 1993; 205–233. pmid:8122902
- 12. Bentley AR, Scutari M, Gosman N, Faure S, Bedford F, Howell P, et al. Applying association mapping and genomic selection to the dissection of key traits in elite European wheat. Theor Appl Genet. 2014;127: 2619–2633. pmid:25273129
- 13. Kertho A, Mamidi S, Bonman JM, McClean PE, Acevedo M. Genome-wide association mapping for resistance to leaf and stripe rust in winter-habit hexaploid wheat landraces. PLoS One. 2015;10: e0129580. pmid:26076040
- 14. Gurung S, Mamidi S, Bonman JM, Xiong M, Brown-Guedira G, Adhikari TB. Genome-wide association study reveals novel quantitative trait loci associated with resistance to multiple leaf spot diseases of spring wheat. PLoS One. 2014;9: e108179. pmid:25268502
- 15. Le Gouis J, Bordes J, Ravel C, Heumez E, Faure S, Praud S, et al. Genome-wide association analysis to identify chromosomal regions determining components of earliness in wheat. Theor Appl Genet. 2012;124: 597–611. pmid:22065067
- 16. Korte A, Farlow A. The advantages and limitations of trait analysis with GWAS: a review. Plant Methods; 2013;9: 29. pmid:23876160
- 17. Maccaferri M, Zhang J, Bulli P, Abate Z, Chao S, Cantu D, et al. A genome-wide association study of resistance to stripe rust (Puccinia striiformis f. sp. tritici) in a worldwide collection of hexaploid spring wheat (Triticum aestivum L.). G3; Genes|Genomes|Genetics. 2015;5: 449–465.
- 18. Vasudevan K, Vera Cruz CM, Gruissem W, Bhullar NK. Large scale germplasm screening for identification of novel rice blast resistance sources. Front Plant Sci. 2014;5: 1–9.
- 19. Endresen FTD, Street K, Mackay M, Bari A, De Pauw E’. Predictive association between biotic stress traits and eco-geographic data for wheat and barley landraces. Crop Sci. 2011;51:2036–2055.;
- 20. Newton AC, Fitt BDL, Atkins SD, Walters DR, Daniell TJ. Pathogenesis, parasitism and mutualism in the trophic space of microbe-plant interactions. Trends Microbiol. 2010;18: 365–373. pmid:20598545
- 21. Zurn JD, Newcomb M, Rouse MN, Jin Y, Chao S, Sthapit J, et al. High-density mapping of a resistance gene to Ug99 from the Iranian landrace PI 626573. Mol Breed. 2014; 1–11.
- 22. Bulli P, Zhang J, Chao S, Chen X, Pumphrey M, Health W. Genetic architecture of resistance to stripe rust in a global winter wheat germplasm collection. G3 Genes|Genomes|Genetics. 2016;6: 2237–2253. pmid:27226168
- 23. Line RF, Qayoum A. Virulence, aggressiveness, evolution, and distribution of races of Puccinia striiformis (the cause of stripe rust of wheat) in North America, 1968–1987. United States Dep Agric Tech Bull. 1992; 44.
- 24. Wan A, Chen X. Virulence characterization of Puccinia striiformis f. sp. tritici using a new set of Yr single-gene line differentials in the United States in 2010. Plant Dis. 2014;98: 1534–1542.
- 25. Stewart CJ, Via L. A rapid CTAB DNA isolation technique useful for RAPD fingerprinting and other PCR applications. Biotechniques. 1993;14: 748–750. pmid:8512694
- 26. Cavanagh CR, Chao S, Wang S, Huang BE, Stephen S, Kiani S, et al. Genome-wide comparative diversity uncovers multiple targets of selection for improvement in hexaploid wheat landraces and cultivars. Proc Natl Acad Sci USA. 2013;110: 8057–8062. pmid:23630259
- 27. Forrest K, Pujol V, Bulli P, Pumphrey M, Wellings C, Herrera-Foessel S, et al. Development of a SNP marker assay for the Lr67 gene of wheat using a genotyping by sequencing approach. Mol Breed. 2014;34: 2109–2118.
- 28. Herrera-Foessel SA, Singh RP, Huerta-Espino J, Rosewarne GM, Periyannan SK, Viccars L, et al. Lr68: A new gene conferring slow rusting resistance to leaf rust in wheat. Theor Appl Genet. 2012;124: 1475–1486. pmid:22297565
- 29. Pritchard JK, Stephens M, Donnelly P. Inference of population structure using multilocus genotype data. Genet Soc Am. 2000;155: 945–959.
- 30. Evanno G, Regnaut S, Goudet J. Detecting the number of clusters of individuals using the software structure: a simulation study. Mol Ecol. 2005;14: 2611–2620. pmid:15969739
- 31. Earl DA, VonHoldt BM. STRUCTURE HARVESTER: a website and program for visualizing STRUCTURE output and implementing the Evanno method. Conserv Genet Resour. 2012;4: 359–361.
- 32. Cleveland WS. Robust locally weighted regression and smoothing scatterplots. J Am Stat Assoc. 1979;74: 829–836.
- 33. Bates D, Machler M, Bolker BM, Walker SC. Fitting linear mixed-effects models using lme4. J. Stat Softw. 2014; 67:1–48
- 34. Corbeil RR, Searle SR. Restricted Maximum Likelihood (REML) Estimation of variance components in the mixed model. Technometrics. 1976;18: 31–38.
- 35. Zhang Z, Ersoz E, Lai C-Q, Todhunter RJ, Tiwari HK, Gore MA, et al. Mixed linear model approach adapted for genome-wide association studies. Nat Genet. 2010;42: 355–360. pmid:20208535
- 36. Yu J, Pressoir G, Briggs WH, Vroh Bi I, Yamasaki M, Doebley JF, et al. A unified mixed-model method for association mapping that accounts for multiple levels of relatedness. Nat Genet. 2006;38: 203–208. pmid:16380716
- 37. Lipka AE, Tian F, Wang Q, Peiffer J, Li M, Bradbury PJ, et al. GAPIT: Genome association and prediction integrated tool. Bioinformatics. 2012;28: 2397–2399. pmid:22796960
- 38. Benjamini Y, Hochberg Y. Controlling the false discovery rate: a practical and powerful approach to multiple testing. J Royal Stati Soc. 1995. pp. 289–300.
- 39. Wang S, Wong D, Forrest K, Allen A, Chao S, Huang BE, et al. Characterization of polyploid wheat genomic diversity using a high-density 90 000 single nucleotide polymorphism array. Plant Biotechnol. 2014;12: 787–796.
- 40. Li Y, Niu YC, Chen XM. Mapping a stripe rust resistance gene YrC591 in wheat variety C591 with SSR and AFLP markers. Theor Appl Genet. 2009;118: 339–346. pmid:18946654
- 41. Cheng P, Xu LS, Wang MN, See DR, Chen XM. Molecular mapping of genes Yr64 and Yr65 for stripe rust resistance in hexaploid derivatives of durum wheat accessions PI 331260 and PI 480016. Theor Appl Genet. 2014;127: 2267–77. pmid:25142874
- 42. Murphy LR, Santra D, Kidwell K, Yan G, Chen X, Campbell KG. Linkage maps of wheat stripe rust resistance genes Yr5 and Yr15 for use in marker-assisted selection. Crop Sci. 2009;49: 1786–1790.
- 43. Ma J, Zhou R, Dong Y, Wang L, Wang X, Jia J. Molecular mapping and detection of the yellow rust resistance gene Yr26 in wheat transferred from Triticum turgidum L. using microsatellite markers. Euphytica. 2001;120: 219–226.
- 44. Ren Y, He Z, Li J, Lillemo M, Wu L, Bai B, et al. QTL mapping of adult-plant resistance to stripe rust in a population derived from common wheat cultivars Naxos and Shanghai 3/Catbird. Theor Appl Genet. 2012;125: 1211–1221. pmid:22798057
- 45. Zwart RS, Thompson JP, Milgate AW, Bansal UK, Williamson PM, Raman H, et al. QTL mapping of multiple foliar disease and root-lesion nematode resistances in wheat. Mol Breed. 2010;26: 107–124.
- 46. Bariana HS, Bansal UK, Schmidt A, Lehmensiek A, Kaur J, Miah H, et al. Molecular mapping of adult plant stripe rust resistance in wheat and identification of pyramided QTL genotypes. Euphytica. 2010;176: 251–260.
- 47. Bansal UK, Kazi AG, Singh B, Hare RA, Bariana HS. Mapping of durable stripe rust resistance in a durum wheat cultivar Wollaroi. Mol Breed. 2014;33: 51–59.
- 48. Lan C, Rosewarne GM, Singh RP, Herrera-Foessel SA, Huerta-Espino J, Basnet BR, et al. QTL characterization of resistance to leaf rust and stripe rust in the spring wheat line Francolin#1. Mol Breed. 2014; 789–803.
- 49. Lillemo M, Asalf B, Singh RP, Huerta-Espino J, Chen XM, He ZH, et al. The adult plant rust resistance loci Lr34/Yr18 and Lr46/Yr29 are important determinants of partial resistance to powdery mildew in bread wheat line Saar. Theor Appl Genet. 2008;116: 1155–1166. pmid:18347772
- 50. Lin F, Chen XM. Genetics and molecular mapping of genes for race-specific all-stage resistance and non-race-specific high-temperature adult-plant resistance to stripe rust in spring wheat cultivar Alpowa. Theor Appl Genet. 2007;114: 1277–1287. pmid:17318493
- 51. Li Z, Singh S, Singh RP, Lopez-Vera EE, Huerta-Espino J. Genetics of resistance to yellow rust in PBW343 x Kenya Kudu recombinant inbred line population and mapping of a new resistance gene YrKK. Mol Breed. 2013;32: 821–829.
- 52. Mallard S, Gaudet D, Aldeia A, Abelard C, Besnard AL, Sourdille P, et al. Genetic analysis of durable resistance to yellow rust in bread wheat. Theor Appl Genet. 2005;110: 1401–1409. pmid:15841362
- 53. Lowe I, Cantu D, Dubcovsky J. Durable resistance to the wheat rusts: Integrating systems biology and traditional phenotype-based research methods to guide the deployment of resistance genes. Euphytica. 2011;179: 69–79. pmid:26900170
- 54. Basnet BR, Ibrahim AMH, Chen X, Singh RP, Mason ER, Bowden RL, et al. Molecular mapping of stripe rust resistance in hard red winter wheat TAM 111 adapted to the U.S. High Plains. Crop Sci. 2014;54: 1361–1373.
- 55. Boukhatem N, Baret P V., Mingeot D, Jacquemin JM. Quantitative trait loci for resistance against yellow rust in two wheat-derived recombinant inbred line populations. Theor Appl Genet. 2002;104: 111–118. pmid:12579435
- 56. Carter AH, Chen XM, Garland-Campbell K, Kidwell KK. Identifying QTL for high-temperature adult-plant resistance to stripe rust (Puccinia striiformis f. sp. tritici) in the spring wheat (Triticum aestivum L.) cultivar “Louise.” Theor Appl Genet. 2009;119: 1119–1128. pmid:19644666
- 57. Sui XX, Wang MN, Chen XM. Molecular mapping of a stripe rust resistance gene in spring wheat cultivar Zak. Phytopathology. 2009;99: 1209–1215. pmid:19740035
- 58. Xu LS, Wang MN, Cheng P, Kang ZS, Hulbert SH, Chen XM. Molecular mapping of Yr53, a new gene for stripe rust resistance in durum wheat accession PI 480148 and its transfer to common wheat. Theor Appl Genet. 2013;126: 523–33. pmid:23090143
- 59. Guo Q, Zhang ZJ, Xu YB, Li GH, Feng J, Zhou Y. Quantitative trait loci for high-temperature adult-plant and slow-rusting resistance to Puccinia striiformis f. sp. tritici in wheat cultivars. Phytopathology. 2008;98: 803–809. pmid:18943256
- 60. Rosewarne GM, Singh RP, Huerta-Espino J, Herrera-Foessel SA, Forrest KL, Hayden MJ, et al. Analysis of leaf and stripe rust severities reveals pathotype changes and multiple minor QTLs associated with resistance in an Avocet × Pastor wheat population. Theor Appl Genet. 2012;124: 1283–1294. pmid:22274764
- 61. Dedryver F, Paillard S, Mallard S, Robert O, Trottet M, Nègre S, et al. Characterization of genetic components involved in durable resistance to stripe rust in the bread wheat “Renan”. Phytopathology. 2009;99: 968–973. pmid:19594316
- 62. Agenbag GM, Pretorius ZA, Boyd LA, Bender CM, Prins R. Identification of adult plant resistance to stripe rust in the wheat cultivar Cappelle-Desprez. Theor Appl Genet. 2012;125: 109–120. pmid:22350093
- 63. Suenaga K, Singh RP, Huerta-Espino J, William HM. Microsatellite markers for genes Lr34/Yr18 and other quantitative trait loci for leaf rust and stripe rust resistance in bread wheat. Phytopathology. 2003;93: 881–890. pmid:18943170
- 64. Lin F, Chen XM. Molecular mapping of genes for race-specific overall resistance to stripe rust in wheat cultivar Express. Theor Appl Genet. 2008;116: 797–806. pmid:18214420
- 65. Feng J, Chen G, Wei Y, Liu Y, Jiang Q, Li W, et al. Identification and genetic mapping of a recessive gene for resistance to stripe rust in wheat line LM168-1. Mol Breed. 2014;33: 601–609.
- 66. William HM, Singh RP, Huerta-Espino J, Palacios G, Suenaga K. Characterization of genetic loci conferring adult plant resistance to leaf rust and stripe rust in spring wheat. Genome. 2006;49: 977–990. pmid:17036073
- 67. Prins R, Pretorius ZA, Bender CM, Lehmensiek A. QTL mapping of stripe, leaf and stem rust resistance genes in a Kariega × Avocet S doubled haploid wheat population. Mol Breed. 2011;27: 259–270.
- 68. Crossa J, Burgueno J, Dreisigacker S, Vargas M, Herrera-Foessel SA, Lillemo M, et al. Association analysis of historical bread wheat germplasm using additive genetic covariance of relatives and population structure. Genetics. 2007;177: 1889–1913. pmid:17947425
- 69. Zhou XL, Han DJ, Chen XM, Gou HL, Guo SJ, Rong L, et al. Characterization and molecular mapping of stripe rust resistance gene Yr61 in winter wheat cultivar Pindong 34. Theor Appl Genet. 2014; 127:2349–2358. pmid:25163935
- 70. Zhou XL, Wang WL, Wang LL, Hou DY, Jing JX, Wang Y, et al. Genetics and molecular mapping of genes for high-temperature resistance to stripe rust in wheat cultivar Xiaoyan 54. Theor Appl Genet. 2011;123: 431–8. pmid:21516354
- 71. Li ZF, Zheng TC, He ZH, Li GQ, Xu SC, Li XP, et al. Molecular tagging of stripe rust resistance gene YrZH84 in Chinese wheat line Zhou 8425B. Theor Appl Genet. 2006;112: 1098–103. pmid:16450183
- 72. Xu H, Zhang J, Zhang P, Qie Y, Niu Y, Li H, et al. Development and validation of molecular markers closely linked to the wheat stripe rust resistance gene YrC591 for marker-assisted selection. Euphytica. 2014;198: 317–323.
- 73. Zhou XL, Wang MN, Chen XM, Lu Y, Kang ZS, Jing JX. Identification of Yr59 conferring high-temperature adult-plant resistance to stripe rust in wheat germplasm PI 178759. Theor Appl Genet. 2014;127: 935–945. pmid:24487945
- 74. Ren RS, Wang MN, Chen XM, Zhang ZJ. Characterization and molecular mapping of Yr52 for high-temperature adult-plant resistance to stripe rust in spring wheat germplasm PI 183527. Theor Appl Genet. 2012;125: 847–857. pmid:22562146
- 75. McIntosh R, Yamazaki Y, Dubcovsky J, Rogers WJ, Morris C, Appels R, et al. Catalogue of gene symbols for wheat in 12th International Wheat Genetics Symposium. Ed by RA McIntosh, Yokohama, Japan. 2013.