Résumé
Corynespora cassiicola, causing agent of the Corynespora Leaf Fall (CLF) of rubber tree, is an Ascomycete fungus able to infect a large range of plants, mostly in tropical areas or green houses. Over the past 20 years, the CLF disease has gained increasing impo1tance in all Asian and African rubber producing countries but has not been declared so far in South America. lt causes massive sporadic leaf falls, thus impacting the yield of rubber production. Susceptibility of the various rubber clones to the disease varies depending on the geographical location, suggesting an important diversity of the natural inoculum. The biology of the fungus is complex since it has been described as a necrotrophic pathogen as well as an endophyte or a saphophyte. In this presentation, I will give a broad description of the research program that we are developing for the characterization of fungal effectors involved in the H. brasiliensis/C. cassiicola interaction. Our objective is to describe the diversity of the C. cassiicola species, trying to connect the genetic races with specific effectors potentially involved in virulence or host specialization of the isolates. In this program, we have first used biochemical and molecular approaches, which lead to the characterization of the cassiicolin toxin, a small secreted protein (SSP) putatively involved in the pathogenicity of C. cassiicola (isolate CCP) (Barthe et al., 2007; Breton, Sanier, & d'Auzac, 2000; de Lamotte et al., 2007; Deon et al., 2012a). A phylogenetic analysis based on C. cassiicola isolates of various geographical origins and host plants identified at least 8 phylogenetic clades (Deon et al., 2014). In addition, a classification based on the cassiicolin gene diversity was proposed (toxin classes). We found a rather good coherence between the toxin classes and the phylogenetic clades. However, 53% of the isolates did not carry any casssicolin gene, although some of them are virulent, suggesting the existence of effectors other than cassiicolin. This analysis is currently being extended to new isolates recently collected in West Africa (preliminary results will be presented). We are also conducting a genomic/transcriptomic approach with the objective to identify all putative effectors involved in the interaction with rubber tree. We have obtained the first reference sequence of the C. cassiicola genome, in frame of the DOE-JGI "1000 Fungal Genome" program. The assembled genome is around 45 Mb and contains about 17000 genes. We are now conducting comparative genomic analysis. The analysis pipeline is presented below. Jnterspecific comparative genomics with 45 fungal genomes, mostly pathogens from the Dothideomycetes (Schoch et al. , 2009), aims to describe the potential pathogenesis effectors arsenal, including the secondary metabolites, SSPs, lipases, proteases and CAZymes. lntraspecific comparative genomics among 37 C. cassiicola isolates of our collection, covering the known diversity, will help defining new and complete effector patterns that will be analyzed comparatively togenome-wide phylogenetic profiles. Finally, a transcriptomic analysis (RNAseq) of fungal genes differentially expressed after inoculation of detached rubber tree leaves (sensitive clone PB260) with a spore suspension (isolate CCP) is being conducted, in order to identify functional and pertinent effector candidates. The best candidates will be validated by functional genetics (through the knock-down of effector encoding genes). This study will provide knowledge and molecular tools for the characterization of C. cassiicola populations in epidemiological studies. This knowledge will help adapting the breeding strategies, taking into account the diversity of the pathogen, for the selection of more tolerant rubber tree clones. (Résumé d'auteur)