CHARACTERIZATION OF BIOACTIVE PEPTIDES IDENTIFIED FROM GENETICALLY ENCODED LIBRARIES.

Modern agriculture relies on the application of huge amounts of pesticides to protect crops from pathogens and pests and prevent yield losses. Unfortunately, the continuous application of these active ingredients is causing the fast development of resistance, toxicity on non-target organisms and environmental accumulation. New solutions must be designed to increase sustainability of agricultural practices and dramatically reduce resistance insurgence. In this scenario, antimicrobial peptides (AMPs) can represent a solution as they show specificity, efficacy and rapid degradation, thus avoiding the long-term effects observed for chemical pesticides. During the last years, some AMPs have been successfully commercialized as plant protection products (PPPs) and some screening pipelines have been developed to provide a fast identification of new antimicrobial peptides. Genetically encoded peptide libraries (GEPLs) allow to screen millions of different amino acid combinations, to identify peptides that can bind a target protein or inhibit a protein-protein interaction (PPI) of choice. This technology has been mainly exploited to identify peptides to be used as human therapeutics, while its potential for PPPs development is poorly explored. In the frame of the “Novel PESTicides for a sustainable agriculture” (NoPEST) project, GEPLs have been constructed to identify peptides to protect crops from disruptive diseases caused by oomycetes. With this aim, a yeast two-hybrid suitable library was constructed to identify cyclic peptides (CPs), which provide enhanced levels of in vivo stability and affinity for the target protein respect to linear counterparts. The produced combinatorial library of cyclic peptides (CYCLIC) was firstly validated identifying G4CP2, a peptide capable to interfere with GAL4 activity and, consequently, with galactose metabolism in yeast cells, thus demonstrating that the library can be used to isolate bioactive molecules. After validation, 14 cell wall biosynthetic enzymes have been selected and screened with CYCLIC to identify CPs having the ability to interfere with the activity of these target proteins and, consequently, with the cell wall biosynthesis and/or integrity in target oomycetes. Different CPs with an antimicrobial activity have been identified, among those, CP32 has been identified based on its binding to the Phytophthora infestans cellulose synthase A2 (PiCesA2), having very good performances in inhibiting the related disease symptoms (late blight). Microscopy, chemical and biochemical analysis are now used to elucidate the precise mechanism of action of CP32 and to confirm its interaction with the target enzyme. By shading light on these aspects, together with a good toxicological profile, the developed pipeline can be finally validated. These data might also suggest CP32 suitability for future improvements and commercialisation.