De Novo Assembly and Phasing of Dikaryotic Genomes from Two Isolates of Puccinia coronata f. sp. avenae , the Causal Agent of Oat Crown Rust

Abstract

Oat crown rust, caused by the fungus Pucinnia coronata f. sp. avenae, is a devastating disease that impacts worldwide oat production. For much of its life cycle, P. coronata f. sp. avenae is dikaryotic, with two separate haploid nuclei that may vary in virulence genotype, highlighting the importance of understanding haplotype diversity in this species. We generated highly contiguous de novo genome assemblies of two P. coronata f. sp. avenae isolates, 12SD80 and 12NC29, from long-read sequences. In total, we assembled 603 primary contigs for 12SD80, for a total assembly length of 99.16 Mbp, and 777 primary contigs for 12NC29, for a total length of 105.25 Mbp; approximately 52% of each genome was assembled into alternate haplotypes. This revealed structural variation between haplotypes in each isolate equivalent to more than 2% of the genome size, in addition to about 260,000 and 380,000 heterozygous single-nucleotide polymorphisms in 12SD80 and 12NC29, respectively. Transcript-based annotation identified 26,796 and 28,801 coding sequences for isolates 12SD80 and 12NC29, respectively, including about 7,000 allele pairs in haplotype-phased regions. Furthermore, expression profiling revealed clusters of coexpressed secreted effector candidates, and the majority of orthologous effectors between isolates showed conservation of expression patterns. However, a small subset of orthologs showed divergence in expression, which may contribute to differences in virulence between 12SD80 and 12NC29. This study provides the first haplotype-phased reference genome for a dikaryotic rust fungus as a foundation for future studies into virulence mechanisms in P. coronata f. sp. avenae.

Dissecting the first phased dikaryotic genomes of the wheat rust pathogenPuccinia triticinareveals the mechanisms of somatic exchange in nature

“…For each isolate, we predicted about 2,000 effectors on the primary contigs (Supplementary file 4) and haplotigs ( Table 2; Supplementary file 4), respectively, comprising about 8% of the total proteins in each genome, in line with previous reports for Pt Race1, Pst, and Pca (11,18,19). A majority (57-70%) of the effectors belonged to either synteny or chromosome allele pairs, and only 6-9% of the effectors were singletons ( Table 2).…”

“…While these resequencing studies have identified a small list of effector candidates for further validation, a more recent comparative study between an isolate of Pgt (Pgt279) and a derivative (Pgt632) isolate with virulence for Sr50 successfully identified AvrSr50 (17). Recently, third generation long-read sequencing has been used to assemble the first haplotype-phased reference genomes for the dikaryotic rust fungi Pst and Pca, which produced reference genomes with much higher contiguity than those based on short-read sequencing and uncovered novel genome features in a dikaryotic fashion (18,19).…”

A chromosome‐scale genome assembly reveals a highly dynamic effector repertoire of wheat powdery mildew

“…Although these genome assemblies allowed a detailed insight into the B. graminis gene content, only rudimentary analyses of the TE fraction and large-scale chromosomal structures were possible. Recent advances in long-read sequencing technology, along with new scaffolding methods, have enabled resolution of chromosomescale assemblies of an increasing number of plant pathogens' genomes (Van Kan et al, 2017;Miller et al, 2018). Only genomes of high contiguity allow the addressing of topics such as gene space organization and copy number variation (CNV).…”

Structural and functional insights into the modulation of the activity of a flax cytokinin oxidase by flax rust effector AvrL567‐A

“…For example, Ustilago maydis , a hemibiotrophic fungal pathogen, secretes the effector Cmu1, affecting salicylic acid (SA) biosynthesis in plants to inhibit SA‐mediated immunity (Djamei et al , ). In contrast with the relatively small set of effectors produced by individual species of bacterial pathogens, whole‐genome sequencing of biotrophic and hemibiotrophic fungi has identified hundreds of potential effectors (Dean et al , ; Duplessis et al , ; Kamper et al , ; Miller et al , ; Nemri et al , ; Pedersen et al , ; Plissonneau et al , ; Schwessinger et al , ; Spanu et al , ; Yoshida et al , ). Functional characterization of fungal effectors is challenging, mainly because fungal effector mutants often display no associated phenotype, either as a result of functional redundancy or of difficulty in accurately measuring small changes in phenotype (Selin et al , ).…”