Jana Sperschneider

Dikaryotic rust fungi maintain 2 distinct haploid nuclei for most of their life cycle, making their large, repeat-rich genomes difficult to assemble and phase. Here we present haplotype-phased, near chromosome-scale genome assemblies for the poplar rust pathogens Melampsora larici-populina 98AG31 and Melampsora allii-populina 12AY07, generated using PacBio HiFi sequencing and Hi-C-guided scaffolding. For each species, we resolved 18 chromosomes per haplotype, providing the first near chromosome-level representations of poplar rust fungal species. M. larici-populina diploid assembly spans ∼203 Mb, while M. allii-populina reaches ∼416 Mb, with high completeness and strong collinearity between haplotypes. Compared with previous fragmented or collapsed references, these assemblies greatly improve contiguity, recover centromeric and telomeric features, and support the transposable element-driven genome size expansion in M. allii-populina. The haplotype-aware annotations of genes and predicted effectors derived from these resources will enable detailed analyses of genome architecture, repeat dynamics, and key loci such as avirulence genes. Together, these assemblies provide a robust genomic resource for investigating host adaptation, virulence evolution, and population diversity in poplar rust fungi.

Stripe (yellow) rust is a devastating disease of wheat caused by the globally distributed plant pathogenic fungus Puccinia striiformis f. sp. tritici (Pst), which results in substantial annual yield losses 1, 2. The Yr28 gene from Aegilops tauschii encodes a nucleotide-binding leucine-rich repeat immune receptor (NLR) that provides broad spectrum resistance to Pst 3, 4. Here we identified the corresponding avirulence protein, AvrYr28-01, by screening a library of effector candidates from Pst. A paralogous gene adjacent to AvrYr28-01 at this locus encodes closely related effector variants that are not recognised by Yr28 due to variation in an N-terminal domain region prior to a predicted alpha-helical bundle. Suppression of Yr28-mediated seedling resistance in some wheat genetic backgrounds is alleviated by between 3 and 6 weeks, with older plants showing effective resistance. The AvrYr28-01 gene is homozygous and invariant in major global lineages of stripe rust, which explains the broad-spectrum resistance of Yr28 and suggests it could provide durable resistance in the field.

Interactions between plants and pathogens drive long-term co-evolution through cycles of effector diversification and immune recognition. Effectors play central roles in this molecular interplay, and zinc-binding folds have been identified in a subset of pathogen effectors, yet the contribution of zinc coordination to effector stability and immune recognition remains unclear. Here, we investigated the structure and recognition of the AvrSr33 effector from the wheat stem rust fungus (Puccinia graminis f. sp. tritici, Pgt). Structural and biochemical analyses show that AvrSr33 adopts a fold containing two zinc-binding sites. Transient expression assays in Nicotiana benthamiana show that AvrSr33 directly interacts with Sr33, and comparison of recognised and non-recognised AvrSr33 variants identifies a polymorphic loop associated with recognition. This loop is positioned adjacent to one of the zinc-binding sites with its orientation constrained by zinc coordination. Reciprocal mutations of key surface residues within this region alter recognition, whereas mutation of the zinc-binding site totally abolishes recognition. Our data suggest that zinc coordination in AvrSr33 provides a structurally constrained scaffold that supports the surface features associated with Sr33 recognition. These findings provide a mechanistic framework for understanding how zinc coordination contributes to effector recognition and may influence the evolutionary trajectories of pathogen effectors.

Rust fungi are significant threats to global food security, causing substantial damage to crops through their ability to adapt and evolve new strains that overcome resistance. These obligate biotrophs infect host plants by secreting effector proteins that manipulate host physiology to promote infection and colonisation. We used AlphaFold2 to investigate structural conservation among effector proteins for the secretomes of Melampsora lini and four Puccinia species. AlphaFold2 yielded high-confidence predictions for 45.7% of the 27,090 secreted proteins, while 19% were poor quality. Comparative analysis revealed extensive structural diversity across the rust secretomes, with all thirteen known rust Avr proteins belonging to different clusters apart from AvrSr13 and AvrSr33. Nevertheless, there were still numerous large clusters of structurally-related proteins, including 59 clusters with over 50 members each, three of which contained known Avr proteins. Of the major structural families defined in other fungi, the rust species studied here only contained FOLD and ToxA-like families. Structural analysis of cysteine-rich proteins revealed over a thousand effector candidates featuring zinc-binding sites, with approximately 75% predicted to be cytoplasmic effectors. In contrast, cysteine-rich apoplastic effector candidates were characterized by a high frequency of disulfide bonds. One family of predicted metal-binding proteins was greatly expanded in P. graminis f. sp. tritici and includes AvrSr13 and AvrSr33. We confirmed that purified AvrSr13 and AvrSr22 proteins bind to zinc in vitro using biochemical assays. Taken together, structural modeling provides new avenues to study sequence-unrelated effectors and highlights the high degree of diversity in the effector repertoires of rust species.

Barley leaf rust disease, caused by Puccinia hordei, leads to substantial yield losses and diminished malting quality of barley across temperate growing regions worldwide. To address the paucity of high-resolution genomic resources for this pathogen, we generated haplotype-phased, chromosome-scale assemblies for ten globally distributed isolates using PacBio HiFi and Hi-C sequencing. Phylogenomic analysis revealed seven distinct lineages of P. hordei, including evidence of nuclear exchange, with a shared nuclear haplotype detected between two US lineages. Nuclear genome sizes ranged from ∼140-147Mbp, with the exception of isolate 90ISR03 from Israel (∼163Mbp), which also harbored a 6.2Mbp extra scaffold in one nucleus exhibiting chromosomal characteristics. Consistent with its larger genome, P. hordei had a higher repeat content (∼70%) than related cereal rust fungi, driven primarily by the proliferation of LTR retroelements and DNA transposons. Across the global pan-genome of 13 unique nuclear haplotypes, approximately one-third of all protein orthogroups were conserved across all isolates. Only 18% of predicted effector orthogroups were conserved across all haplotypes, reflecting the highly dynamic and variable nature of the effector repertoire. The long-term propagation of clonal P. hordei lineages is apparent both within the US and globally, and nuclear exchange plays a role in generating novel diversity. Genome plasticity is evident in extensive structural variation, including large-scale translocations and inversions as well as an extra chromosome. These chromosome-level, haplotype-resolved genomes provide a foundational resource for exploring the evolution, diversity, and avirulence gene repertoire of P. hordei.

Accurate gene annotation is crucial for inference of biological knowledge from genomes. However, non-canonical genes such as orphan or single-exon genes as well as those residing in rapidly evolving regions are routinely dismissed in annotation pipelines. In filamentous pathogen genomes, this disproportionately affects the annotation of genes encoding disease-promoting effector proteins. We introduce EffectorGeneP, a machine learning tool that self-trains on transcript data, predicts the most likely coding sequence from transcripts and effectively separates bona fide genes from transcriptional noise. EffectorGeneP annotates over 95% of known effectors correctly, while other state-of-the-art methods annotate 15%-78%. We show that EffectorGeneP expands the predicted secretome of pathogens by over 50% and that high-throughput screening of an effector library in plant protoplasts uncovers the previously poorly annotated AvrSr26 gene family in the wheat stem rust fungus. EffectorGeneP decodes genomes at unprecedented resolution and will enable the study of biological processes in important pathogen species.

Dikaryotic rust fungi maintain two distinct haploid nuclei for most of their life cycle, making their large, repeat-rich genomes difficult to assemble and phase. Here we present haplotype-phased, near chromosome-scale genome assemblies for the poplar rust pathogens Melampsora larici-populina 98AG31 and Melampsora allii-populina 12AY07, generated using PacBio HiFi sequencing and Hi-C-guided scaffolding. For each species, we resolved 18 chromosomes per haplotype, providing the first near chromosome-level representations of poplar rust fungal species. M. larici-populina diploid assembly spans ~203 Mb, while M. allii-populina reaches ~416 Mb, with high completeness and strong collinearity between haplotypes. Compared with previous fragmented or collapsed references, these assemblies greatly improve contiguity, recover centromeric and telomeric features, and support the transposable element-driven genome size expansion in M. allii-populina . The haplotype-aware annotations of genes and predicted effectors derived from these resources will enable detailed analyses of genome architecture, repeat dynamics, and key loci such as avirulence genes. Together, these assemblies provide a robust genomic resource for investigating host adaptation, virulence evolution, and population diversity in poplar rust fungi.

Puccinia coronata f. sp. avenae ( Pca), a fungal pathogen causing crown rust of oat, demonstrates rapid virulence evolution and adaptation to newly released cultivars. To further capture the genetic variation of Pca, we generated nuclear haplotype-resolved genomes for 10 isolates from Europe, Africa, and the Middle East and compared these with existing references for U.S. and Australian isolates. Of the full collection of 52 haplotypes, 40 were unique. Importantly, the presence of a nearly identical haplotype in a U.K. isolate collected in 1984 and in U.S. isolates from 1990 and 2017 supports the existence of long-lived clonal haplotypes in the global population that have been exchanged between lineages. We identified infrequent recombination between haplotypes from geographically dispersed isolates, hinting toward a globally mobile population of Pca that is mostly composed of persistent clonal lineages with some influence from rare recombination events. One isolate contained an additional scaffold with telomeric and centromeric features, suggesting the presence of a supernumerary chromosome. Whereas the core pan-proteome is enriched for predicted secreted and effector proteins, sequence and expression variation are most prevalent in non-core orthogroups, which also displayed allele-specific expression. We anticipate that this expanded collection of haplotypes will facilitate the development of new surveillance technologies and identification of virulence loci. [ ] Copyright © 2026 The Author(s). This is an open access article distributed under the CC BY-NC-ND 4.0 International license .

Puccinia graminis f. sp. tritici (Pgt) causes wheat stem rust, a devastating disease of cereals. Recent approaches to examine populations at a genomic level have provided valuable information on the genotypic diversity of Pgt populations in Africa, Europe and North America and evolutionary mechanisms underlying the emergence of new races. However, an in-depth characterisation of Pgt populations in South America has been lacking. To bridge this knowledge gap, 91 Pgt isolates were collected from Argentina and Uruguay in 2020 and 2021 and used to generate transcriptome and whole genome sequence data and pathotype information. Phylogenetic analyses revealed that this South American Pgt population includes three clonal lineages, two of which have not been detected elsewhere. The third lineage is globally dispersed, including isolates from Africa and the Middle East. The predominant lineage, unique to South America, encompassed 90% of the samples and showed related pathotypes differing by virulence on single resistance genes consistent with evolution by stepwise mutation within the clonal lineage. There was no evidence of sexual recombination giving rise to new genetic diversity in this population. These observations urge to routinely incorporate and compare genotypic data from South American Pgt isolates to surveillance data from other geographic regions.

Wheat stem rust is a disease of global importance caused by the fungal pathogen Puccinia graminis f. sp. tritici ( Pgt ). Here we generate chromosome-level, nuclear-phased genome references for Pgt isolates ETH2013-1 and ITA2018-1, representing races TKTTF and TTRTF respectively, that have caused major epidemics in Africa and Europe. The nuclear haplotypes of ETH2013-1 and ITA2018-1 are unique and unrelated to those of Ug99 and Pgt21. Pgt nuclear haplotypes show extensive variation in sequence and copy number of six known Avr genes and AvrSr33 , which we identify through an effector gene library screen. Recognition properties of 22 novel Avr gene variants explain the race virulence phenotypes and the outbreak of TTRTF on durum cultivars containing Sr13b , since ITA2018-1 carries a homozygous deletion of AvrSr13 . This work establishes an Avr gene atlas for Pgt that can inform wheat breeding and enable development of sequence-based virulence diagnostic tools for pathogen surveillance.