Differential alteration of plant functions by homologous fungal candidate effectors

Abstract

Rust fungi are plant pathogens that cause epidemics that threaten the production of important plant species, such as wheat, soy, coffee and poplar. Melampsora larici-populina (Mlp) causes the poplar rust and encodes at least 1 184 candidate effectors (CEs), however their functions are poorly known. In this study, we used Arabidopsis plants constitutively expressing CEs of Mlp to discover processes targeted by these fungal proteins. For this purpose, we sequenced the transcriptome and used mass spectrometry to analyse the metabolome of Arabidopsis plants expressing individually one of the 14 selected CEs and of a control line. We found 2 299 deregulated genes across the experiment. Among the down-regulated genes, the KEGG pathways "MAPK signaling pathway" and "Plant-pathogen interaction" were respectively over-represented in six and five of the 14 transgenic lines. Moreover, genes related to hormone response and defense were down-regulated across all transgenic lines are. We further observed that there were 680 metabolites deregulated in at least one CE-expressing transgenic line, with highly unsaturated and phenolic compounds enriched in up-regulated metabolites and peptides enriched among down-regulated metabolites. Interestingly, we found that transgenic lines expressing unrelated CEs had correlated patterns of gene and metabolite deregulation, while expression of CEs belonging to the same family deregulated different genes and metabolites. Taken together, our results indicate that the sequence of effectors and their belonging to families may not be a good predictor of their impact on the plant. ImportanceRust fungi are plant pathogens that threaten the production of important crops, including wheat, soy, coffee and poplar. Effectors are used by pathogens to control the host, however in the case of Melampsora larici-populina, the causal agent of the poplar rust, and other rust fungi these proteins are poorly known. We used Arabidopsis plants expressing candidate effectors (CEs) of Mlp to better understand the interaction between this pathogen and its hosts. We found that expression of unrelated CEs led to similar patterns of gene and metabolite deregulation, while transgenic lines expressing CEs belonging to the same family showed different groups of different genes and metabolites deregulated. Thus, our results suggest that functional annotation of effectors based on sequence similarity may be misleading.