The dispersion of fungal inocula such as the airborne spores of rust fungi (Pucciniales) can be monitored through metabarcoding of the internal transcribed spacer 2 (ITS2) of the rRNA gene in environmental DNAs. This method is largely dependent on a high-quality reference database (refDB) and primers with proper taxonomic coverage and specificity. For this study, a curated ITS2 reference database (named CR-ITS2-refDB) comprising representatives of the major cereal rust fungi and phylogenetically related species was compiled. Interspecific and intraspecific variation analyses suggested that the ITS2 region had reasonable discriminating power for the majority of the Puccinia species or species complexes in the database. In silico evaluation of nine forward and seven reverse ITS2 primers, including three newly designed, revealed marked variation in DNA amplification efficiency for the rusts. We validated the theoretical assessment of rust-enhanced (Rust2inv/ITS4var_H) and universal fungal (ITS9F/ITS4) ITS2 primer pairs by profiling the airborne rust fungal communities from environmental samples via a metabarcoding approach. Species- or subspecies-level identification of the rusts was improved by use of CR-ITS2-refDB and the Automated Oligonucleotide Design Pipeline (AODP), which identified all mutations distinguishing highly conserved DNA markers between close relatives. A generic bioinformatics pipeline was developed, including all steps used in this study from in silico evaluation of primers to accurate identification of short metabarcodes at the level of interest for defining phytopathogens. The results highlight the importance of primer selection, refDBs that are resolved to reflect phylogenetic relationships, and the use of AODP for improving the reliability of metabarcoding in phytopathogen biosurveillance.
Keywords: Puccinia; aeromycobiota; bioinformatics; cereal crops; computational biology; fungal pathogens; herbarium specimens; microbiome; pathogen detection; rust fungi; spore sampler.