k-mer-based GWAS reveals a candidate avirulence gene and structural variation in Puccinia triticina linked to gain of Lr20 virulence

Abstract Background Plant pathogens secrete effector proteins into their hosts to promote colonisation. Among these are avirulence (Avr) effectors, which can be recognised by specific host immune receptors, triggering an immune response that prevents pathogen progression. This recognition exerts strong evolutionary pressure on pathogens to alter and/or eliminate Avr genes to escape recognition. Consequently, understanding Avr gene evolution is critical for developing effective resistance deployment strategies. However, identifying and validating Avr effectors remains a significant challenge, especially for fungal plant pathogens, leading to a limited catalogue of Avr genes. This challenge is particularly pronounced for obligate biotrophic pathogens such as the wheat leaf (brown) rust fungus Puccinia triticina (Pt), where only two Avr genes have been confirmed to date. Results In this study, we conducted a k-mer-based genome-wide association study (GWAS) to detect a broad spectrum of structural genetic variations — including single nucleotide polymorphisms (SNPs), insertions and deletions (indels) and copy number variations (CNVs) — that may contribute to the gain of virulence in Pt. Analysis of k-mers linked to avirulence phenotypes of Pt isolates across eleven leaf rust resistance (Lr) loci, revealed a distinct association peak on chromosome 10B corresponding to avirulence against Lr20. Assembly of the associated k-mers produced a 50 bp sequence that was located near two candidate effector genes, one of which — termed Pt76_024702 — also displayed high levels of expression during both early and later stages of infection. Furthermore, the genomic region harbouring Pt76_024702 exhibited large-scale deletions in certain Pt lineages virulent to Lr20, particularly those infecting durum wheat (Triticum turgidum ssp. durum). Conclusion These findings highlight Pt76_024702 as a compelling candidate for AvrLr20 and demonstrate the significant potential of the presented k-mer-based GWAS approach to enhance the Avr gene catalogue. This strategy is particularly promising for complex fungal pathogens such as the notorious wheat rust pathogens where conventional approaches have previously proved challenging.