Molecular genetic characterisation of triple rust resistance in Aegilops tauschii

Abstract

Bread wheat (Triticum aestivum) is one of the top three cultivated crops, and a major caloric source for humanity. Global wheat production is under threat due to the rapid evolution of highly virulent fungal pathogens such as Puccinia spp. that cause rust diseases. Losses due to rusts are routinely minimised through the deployment of host-mediated genetic resistance. However, the rust pathogens have the ability to evolve virulence and overcome the host resistance. Therefore, a continuous supply of new sources of resistance is essential for sustainable rust management. Wheat wild relatives are a valuable resource as they provide resistance against diverse rust forms.In this study, CPI110672, an accession of the D-genome progenitor Aegilops tauschii, was chosen for in-depth analysis as it resists three wheat rust diseases, namely leaf, stem and stripe rust. To determine whether the triple rust resistance is pleiotropic or involves multiple genes conferring specific resistance, we conducted genetic analysis using a mapping population derived from a cross between CPI110672 and CPI110717 (susceptible) accessions. Through rust infection screening, we determined the triple rust resistance was conferred by multiple genes. Two independent genes (Sr672a and Sr672b) segregated for stem rust resistance, while we identified monogenic segregation for stripe (Yr672) and leaf rust (Lr672) resistance. Genotyping by 90K Infinium single nucleotide polymorphism (90K SNP) chip analysis confirmed independent segregation where each of the resistances were linked with different SNP markers. Based on closely-associated SNPs and their physical position, the leaf rust resistance gene Lr672 and one of the stem rust resistance genes (Sr672a) were mapped to the short arm of chromosome 2D, whereas the stripe rust resistance gene mapped to chromosome arm 4DS. Converting the SNPs into Kompetitive Allele Specific (KASP) genotyping markers and mapping on the segregating population resulted in the identification of flanking markers for all three resistance genes.Anchoring the Sr672a flanking markers on to the Chinese Spring (CS) IWGSC RefSeq v1.0 and Ae. tauschii AL8/78 v4.0 reference genome sequences identified one candidate gene belonging to CC-NBS-LRR (CNL), the major class of resistance genes known in plants for rust resistance. The candidate gene was identified as a homologue of the recently cloned Sr46 gene. Screening using a gene specific marker for Sr46 confirmed that Sr672a was an allele of Sr46 with a single amino acid difference and hence designated as Sr46b. Validation of Sr46b by a transgenic complementation test confirmed that rust resistance is conferred by this allele, and indicated that the difference in one amino acid did not alter the rust resistance function. Further, we deployed Sr46b into a commercial cultivar through marker assisted selection. Also, we attempted to stack Sr46b with other stem rust resistance genes (Sr33 and Sr45) isolated from Ae. tauschii using speed breeding technology. Through speed breeding and marker assisted selection, we were able to select recombinant lines with multiple resistance genes combinations within 180 days.We used traditional map-based cloning in conjunction with a comparative genomics approach to fine mapping the leaf rust (Lr672) and stripe rust (Yr672) resistance genes. The whole genome sequence assemblies of parental accessions CPI110672, CPI110717 from the open wild wheat consortium (OWWC), John Innes Centre, UK and the recent version of reference sequences of CS-RefSeq v1.0 and AL8/78 v4.0 were used to fine map and identify candidate genes for Lr672 and Yr672.Based on previous studies, we identified the accession CPI110672 as synonymous with TA1675, the donor for the leaf rust resistance gene Lr39, thus confirming that Lr672 is Lr39. Anchoring the closely linked flanking markers in the CS-RefSeq v1.0 and AL8/78 v4.0 reference sequences delimited a 1.2 Mb genomic region in both reference genomes. Additionally, we also used the CPI110672 genome sequences to predict CNL genes mapped within the 1.2 Mb physical region of Lr39. Markers specific to the candidate genes co-segregated with the leaf rust phenotype in the CPI110672xCPI110717 F2:3 mapping population and hence was useful for marker assisted deployment of Lr39 resistance.Similarly, anchoring the Yr672 flanking markers narrowed down an approximately 500 kb region in both reference sequences. Evaluation of the CPI110672 CNL genes mapped collinear to the 500 kb reference sequences identified one candidate gene. Markers specific to the candidate CNL gene co-segregated with the stripe rust resistance phenotype in the F2:3 mapping population. Cloning and transgenic complementation tests confirmed the stripe rust resistance. Further, we deployed the Yr672 in the commercial cultivar using marker assisted backcross strategy.The major outcomes of this study include:• Understanding the genetic architecture of triple rust resistance in accession CPI110672• Cloning and validation of Sr46b and Yr672 rust resistance candidate genes• Fine mapping a leaf rust resistance (Lr39) gene• Development of breeder-friendly molecular markers for Sr46, Yr672 and Lr39• Marker assisted introgression of Sr46b and Yr672 into a commercial cultivar• Marker assisted pyramiding of Sr33/Sr45/Sr46b resistance genes using speed breeding technology