Ash, G. J. | Search bioRxiv

bioRxiv2025

Multi-locus phylogenetic network analysis of Ampelomyces mycoparasites isolated from diverse powdery mildews in Australia and the generation of two de novo genome assemblies

The interactions between powdery mildews (Ascomycota, Erysiphaceae), obligate biotrophic pathogens of many plants, and pycnidial fungi belonging to the genus Ampelomyces, are classic examples of specific mycoparasitic relationships. These interactions are common and finely tuned tritrophic relationships amongst host plants, powdery mildews, and Ampelomyces mycoparasites wherever these organisms co-occur in the field. Selected Ampelomyces strains have already been developed as biocontrol agents of powdery mildew infections of some crops. In Australia, their study has received little attention so far. Only a single Ampelomyces strain, included in a whole-genome sequencing (WGS) project, was known from this continent. Here, we report the isolation of 20 more Ampelomyces strains from eight powdery mildew species in Australia. Multi-locus phylogenetic network analyses of all the 21 Australian Ampelomyces strains carried out in combination with 32 reference strains from overseas revealed that the Australian strains belonged to four molecular taxonomic units (MOTUs). All those MOTUs were delimited earlier based on Ampelomyces strains isolated in Europe, North America, and elsewhere. Analyses indicated possible recombination events amongst phylogenetically different Ampelomyces strains. Based on the phylogenetic analyses, two Australian strains belonging to different MOTUs were selected for WGS. Long-read (minION) and short-read (Illumina) technologies were used to provide genome assemblies with high completeness. Both assemblies reported here had bipartite structure, i.e., consisted of AT-rich, gene-sparse regions interspersed with GC-balanced, gene-rich regions. These new high-quality assemblies and evidence-based annotations are important resources for future analyses of mycoparasitic interactions to disentangle molecular mechanisms underlying mycoparasitism, possible new biocontrol applications, and natural tritrophic relationships.

microbiology↗

bioRxiv2019

Genome mining of the citrus pathogen Elsinoë fawcettii; prediction and prioritisation of candidate effectors, cell wall degrading enzymes and secondary metabolite gene clusters

AbstractElsinoe fawcettii, a necrotrophic fungal pathogen, causes citrus scab on numerous citrus varieties around the world. Known pathotypes of E. fawcettii are based on host range; additionally, cryptic pathotypes have been reported and more novel pathotypes are thought to exist. E. fawcettii produces elsinochrome, a non-host selective toxin which contributes to virulence. However, the mechanisms involved in potential pathogen-host interactions occurring prior to the production of elsinochrome are unknown, yet the host-specificity observed among pathotypes suggests a reliance upon such mechanisms. In this study we have generated a whole genome sequencing project for E. fawcettii, producing an annotated draft assembly 26.01 Mb in size, with 10,080 predicted gene models and low (0.37%) coverage of transposable elements. The assembly showed evidence of AT-rich regions, potentially indicating genomic regions with increased plasticity. Using a variety of computational tools, we mined the E. fawcettii genome for potential virulence genes as candidates for future investigation. A total of 1,280 secreted proteins and 203 candidate effectors were predicted and compared to those of other necrotrophic (Botrytis cinerea, Parastagonospora nodorum, Pyrenophora tritici-repentis, Sclerotinia sclerotiorum and Zymoseptoria tritici), hemibiotrophic (Leptosphaeria maculans, Magnaporthe oryzae, Rhynchosporium commune and Verticillium dahliae) and biotrophic (Ustilago maydis) plant pathogens. Genomic and proteomic features of known fungal effectors were analysed and used to guide the prioritisation of 77 candidate effectors of E. fawcettii. Additionally, 378 carbohydrate-active enzymes were predicted and analysed for likely secretion and sequence similarity with known virulence genes. Furthermore, secondary metabolite prediction indicated nine additional genes potentially involved in the elsinochrome biosynthesis gene cluster than previously described. A further 21 secondary metabolite clusters were predicted, some with similarity to known toxin producing gene clusters. The candidate virulence genes predicted in this study provide a comprehensive resource for future experimental investigation into the pathogenesis of E. fawcettii.

genomics↗