Dissecting the genetic basis of wheat yellow rust resistance in the NIAB Elite MAGIC population

Yellow rust, caused by the biotrophic fungus Puccinia striiformis f. sp. tritici (Pst), poses a major challenge for wheat breeders and growers globally. The past two decades have seen the rise of Pst populations that are more genetically diverse, more aggressive and that have adapted to warmer temperatures. These features, likely further aided by increased international travel, have led to important epidemic outbreaks and have jeopardised wheat yellow rust resistance levels in the main wheat-producing regions globally. Widespread epidemics have been further facilitated by the deployment of genetically uniform material underpinned by major resistance over large areas. With such a rapidly changing Pst population landscape, disease resistance breeding strategies must adapt accordingly, and this starts with the continued characterisation of adequate yellow rust resistance loci and accompanying molecular and genomic tools. It is crucial that these loci are of direct relevance to breeding programmes for rapid varietal deployment. To this end, I used the NIAB Elite Multi-parent Advanced Generation Inter-Cross (MAGIC) population, a multi-founder population that captures 80 % of the genetic variation present in key representative varieties used in UK wheat breeding (1970-2010s), to identify and characterise genetic loci controlling yellow rust resistance in replicated multi-environmental field trials, in both leaves and ears. This approach has further opened up the avenue for dissecting disease resistance beyond the limited scope of single varieties. I found that nine Quantitative Trait Loci (QTLs) conferred resistance to yellow rust, with four consistently detected across environments and explaining nearly 50 % of the phenotypic variation, and the other five explaining 15-20 % with inconsistent detection across environments. There was a strong indication of additivity effects between the four strong-effect QTL. Furthermore, all founders but the most susceptible one contributed towards resistance, indirectly demonstrating that UK breeding germplasm has high resistance potential against yellow rust. In the second part of my thesis, I focus on the physical interval of the eight most significant QTL previously identified, by examining gene annotations from the recently published IWGSC RefSeq v1.0 genome assembly. Five QTLs were characterised by NBS-LRR clusters. The presence of NBS-LRR-encoding genes with integrated domains revealed the potential for effector triggered immunity based on indirect recognition for a subset of those yellow rust resistance loci. The other three QTL were characterised by the absence of NBS-LRR-encoding genes in their physical interval, potentially indicating the role of non-race specific yellow rust resistance in the MAGIC population. Finally, I focus on glume infection, a phenotypic trait largely overlooked in QTL mapping studies, despite repeated reports of outbreaks. Despite high heritability (72 %), the five QTLs detected explained between 3 and 6 % of the phenotypic variation. Three QTLs co-located with QTLs for foliar resistance. The other two were associated with flowering time, suggesting that earlier ear emergence potentially leads to increased susceptibility to yellow rust in the glumes