Résumé
Mycotoxins are toxic secondary metabolites synthesized by molds that contaminate many foods of plant origin, including cereals which are the first factor of consumer exposure to these molecules. In particular, the species Fusarium graminearum and Fusarium verticillioides are responsible for the production of mycotoxins belonging to trichothecene and fumonisin classes respectively. To fight against these fungi, an alternative strategy to phytosanitary products is the use of antagonistic microorganisms which by various mechanisms will reduce the progression of these fungi and limit the accumulation of mycotoxins. Actinobacteria show some of the most promising features for the biocontrol applications. Carrying a large toolbox of enzymes and active metabolites that allow them to settle in the soil environment, they are also able to interact directly with the plant. Our studies aim to propose a global approach to evaluate and to characterize the potential of actinobacteria isolated from amended soils. To this aim, our studies will span the control of wheat fungal diseases, from in vitro testing to in planta efficacy, and will also evaluate their technological properties for a future industrial production. Sixty actinobacteria strains were confronted against the mycotoxigenic fungi. Several isolates have shown promising antimycotoxigenic potential and have allowed us to select 10 strains for further investigations. In planta assays will be implemented to challenge our in vitro results. Finally, technological properties such as sporulation yield and resistance to conservation processes will be studied for the most promising strains. This will open new perspectives for the potential formulation of a biocontrol product. Moreover, the implementation of such protection methods requires a good knowledge of the biological mechanisms that govern the interaction between mycotoxinogenic molds and biocontrol agents, in order to favor the antagonistic phenomena but also to facilitate the selection of new biocontrol candidates according to the most promising processes. Thus, at the same time, a toolbox to decrypt the mechanisms of action is under development on 3 commercial agents. The impacts of these BCAs on pathogens are analyzed at the microscopic level, but also on the growth kinetics of the pathogens and their ability to produce mycotoxins under different conditions. Further studies on in vitro pathosystem will use omics to identify the genes and molecules differentially expressed during these interactions. The preliminary results show that in all the interaction combinations (pathogen-BCA), the BCA significantly reduce the growth of the pathogens. However, a great variability of interaction profiles is observed with in particular a strong impact of the nutrient source. For example, with a limiting nutrient source, T. asperellum causes a reduction in specific mycotoxinogenesis, while the opposite phenomenon is observed in a non-limiting nutritive source. Confrontation experiments will then be carried out on wheat spikelets in order to validate the hypotheses observed on synthetic medium and put the plant back at the heart of the experimental setup. In the long term, this highly integrative and multidisciplinary approach will provide elements of answers on the biological mechanisms involved in the biocontrol of mycotoxigenic fungi, the biological pathways involved in this interaction.