Wheat genetic loci conferring resistance to stripe rust in the face of genetically diverse races of the fungus Puccinia striiformis f. sp. tritici

Abstract

Key message Analysis of a wheat multi-founder population identified 14 yellow rust resistance QTL. For three of the four most significant QTL, haplotype analysis indicated resistance alleles were rare in European wheat. Abstract Stripe rust, or yellow rust (YR), is a major fungal disease of wheat (Triticum aestivum) caused by Puccinia striiformis Westend f. sp. tritici (Pst). Since 2011, the historically clonal European Pst races have been superseded by the rapid incursion of genetically diverse lineages, reducing the resistance of varieties previously showing durable resistance. Identification of sources of genetic resistance to such races is a high priority for wheat breeding. Here we use a wheat eight-founder multi-parent population genotyped with a 90,000 feature single nucleotide polymorphism array to genetically map YR resistance to such new Pst races. Genetic analysis of five field trials at three UK sites identified 14 quantitative trait loci (QTL) conferring resistance. Of these, four highly significant loci were consistently identified across all test environments, located on chromosomes 1A (QYr.niab-1A.1), 2A (QYr.niab-2A.1), 2B (QYr.niab-2B.1) and 2D (QYr.niab-2D.1), together explaining ~ 50% of the phenotypic variation. Analysis of these four QTL in two-way and three-way combinations showed combinations conferred greater resistance than single QTL, and genetic markers were developed that distinguished resistant and susceptible alleles. Haplotype analysis in a collection of wheat varieties found that the haplotypes associated with YR resistance at three of these four major loci were rare (≤ 7%) in European wheat, highlighting their potential utility for future targeted improvement of disease resistance. Notably, the physical interval for QTL QYr.niab-2B.1 contained five nucleotide-binding leucine-rich repeat candidate genes with integrated BED domains, of which two corresponded to the cloned resistance genes Yr7 and Yr5/YrSp. Graphical abstract

Genetic resistance to stripe rust infection of the wheat ear is controlled by genes controlling foliar resistance and flowering time

“…The MAGIC RILs showed phenotypic variation over the full range of the assessment scale (Figure 1c): nearly 70 % of RILs showed little or no infection (GIF = 0) while ~10 % exhibited Pst infection in the majority of the ear (GIF = 3-4) (Supplementary Figure 2). Pearson’s correlation coefficient analysis of the MAGIC glume YR infection data alongside previously reported leaf YR infection data from the same trials (Bouvet et al 2021b) found a positive correlation at both environments, with OSG16 ( R = 0.62, p <2.2E-16) showing a stronger correlation than was observed at NIAB16 ( R = 0.49, p < 2.2E-16) (Figure 2). MAGIC RILs susceptible in the glumes generally also exhibited high leaf susceptibility (leaf infection >60%).…”

“…Statistical analysis of the glume infection phenotypic data using REML showed that the Wald statistic for genotype effect was highly significant (Chi pr <0.001) at both environments (Supplementary Table 3). High genotypic performance is further supported by high broad sense heritability values at NIAB16 ( h 2 = 0.80) and OSG16 ( h 2 = 0.82), mirroring values observed for leaf infection at the same trials ( h 2 = 0.79-0.94) (Bouvet et al 2021b). This evidence points towards stable genotypic performance for glume infection in the MAGIC population at each site.…”

“…Cross-comparison between the glume and YR resistance QTL showed three of our glume QTL to co-locate with leaf QTL: (1) glume QTL QYrg.niab-2D.2 and leaf QTL QYr.niab-2D.1 (~18% PVE; Bouvet et al 2021b), both identified by the same peak marker ( Ra_c21099_1781 ). (2) Glume QTL QYrg.niab-4D.1 and leaf QTL QYr.niab-4D.1 (4% PVE; Bouvet et al 2021b), both defined by the same peak marker ( D_GDRF1KQ02H66WD_341 ). (3) Glume QTL QYrg.niab-5A.1 and leaf QTL QYr.niab-5A.1 (3.6% PVE; Bouvet et al 2021b), for which the peak markers were 8 cM apart.…”

“…Therefore, the natural log was applied to improve the normality of the data. Phenotypic data for YR infection of the leaf in the NIAB16 and OSG16 trials has previously been reported, as part of a wider series of five trials investigating the genetics of foliar YR infection (Bouvet et al 2021b). To determine the consistency of the phenotype across test environments, and the relationship between YR glume infection with YR leaf infection previously reported in the same trials (Bouvet et al 2021b), Pearson correlation coefficients for glume infection were computed using the ‘cor.test()’ function in the statistical software R 3.6.1 (R Core Team, 2019).…”

“…Recently, several multi-founder experimental populations have been developed in wheat (reviewed by Scott et al 2020). Here we use an eight-parent wheat multifounder advanced generation inter-cross (MAGIC) population genotyped using a 90,000 feature single nucleotide polymorphism (SNP) array which has previously been used for genetic mapping of YR infection of the leaf (Bouvet et al 2021b) to: (1) characterise for the first time the genetic basis of glume YR resistance, and (2) compare glume resistance QTL to the genetic basis of yellow rust resistance in leaves.…”