Fungi represent a common and diverse part of the microbial communities that associate with plants. They also commonly colonise various plant parts asymptomatically. The molecular mechanisms of these interactions are, however, poorly understood. In this study we use transcriptomic data from Eucalyptus grandis, to demonstrate that RNA-seq data are a neglected source of information to study fungal–host interactions, by exploring the fungal transcripts they inevitably contain. We identified fungal transcripts from E. grandis data based on their sequence dissimilarity to the E. grandis genome and predicted biological functions. Taxonomic classifications identified, amongst other fungi, many well-known pathogenic fungal taxa in the asymptomatic tissue of E. grandis. The comparison of a clone of E. grandis resistant to Chrysoporthe austroafricana with a susceptible clone revealed a significant difference in the number of fungal transcripts, while the number of fungal taxa was not substantially affected. Classifications of transcripts based on their respective biological functions showed that the fungal communities of the two E. grandis clones associate with fundamental biological processes, with some notable differences. To shield the greater host defence machinery in the resistant E. grandis clone, fungi produce more secondary metabolites, whereas the environment for fungi associated with the susceptible E. grandis clone is more conducive for building fungal cellular structures and biomass growth. Secreted proteins included carbohydrate active enzymes that potentially are involved in fungal–plant and fungal–microbe interactions. While plant transcriptome datasets cannot replace the need for designed experiments to probe plant–microbe interactions at a molecular level, they clearly hold potential to add to the understanding of the diversity of plant–microbe interactions.
Mannitol metabolism during pathogenic fungal–host interactions under stressed conditions
