Producción y utilización biotecnológica de nuevas proteínas antifúngicas de hongos filamentosos

Resumen

Antimicrobial peptides (AMPs) are promising antifungal alternatives to the fungicides used in agriculture. However, the high cost of chemical synthesis and the difficulties of large-scale production have limited their application. Antifungal proteins (AFPs) are a group of natural, small, cationic, secreted, cysteine-rich AMPs that offer a great potential to develop new biomolecules for the control of phytopathogenic fungi. AFPs are naturally present in filamentous fungi, are very stable, and can be produced in large amounts. However, the biological role of these AFPs in their producer fungus is still unclear. In this thesis, we first studied the diversity of AFPs in ascomycetous genomes and proposed a new classification in three different classes (A, B and C). Penicillium digitatum is the main citrus postharvest pathogen and encodes only one AFP from class B in its genome (AfpB), while Penicillium expansum is the main pome postharvest pathogen and encodes one AFP from each class (AfpA, AfpB and AfpC). In this work, we report the identification, efficient biotechnological production and characterization of these four AFPs. We characterized the biological role of the afpB gene in P. digitatum by the study of its gene expression pattern and the generation of null and constitutive expression mutants. Results indicated that afpB is dispensable for the biology and life cycle of this fungus, although expression of the afpB gene under the constitutive Aspergillus nidulans gpdA promoter is detrimental to growth and virulence to citrus. Surprisingly, neither the wild type nor the constitutive strains produced detectable amounts of AfpB in spite of the high afpB gene expression. Molecular modeling and rational design allowed us to predict the AfpB tertiary structure and design synthetic peptides to map antifungal motifs within the AfpB primary sequence. We confirmed that the cationic exposed loops L2 and L3 showed moderate antifungal activity and that they can act synergistically. With the objective of the biotechnological production of AfpB, we used an AFP expression cassette based on the promoter and terminator regions of the well-studied paf gene from Penicillium chrysogenum, which naturally produces high amounts of its own protein PAF. This paf cassette worked efficiently in P. digitatum and allowed the homologous production of AfpB. Data also showed that the signal peptide (SP) and pro-peptide sequences of the translated SP-Pro-AfpB do not determine protein production. We also demonstrated the thermal stability and resistance to proteolytic cleavage of the P. digitatum AfpB, and provided data that suggest that tertiary structure is not required for antifungal activity. Similar to P. digitatum, none of the three AFPs were detected in supernatants of cultures of P. expansum in rich medium. By contrast, AfpA was produced with very high yields in P. expansum cultures in minimal medium. To complete the repertoire of AFPs from P. expansum we produced the three AFPs from P. expansum (AfpA, AfpB and AfpC) in P. chrysogenum with the use of the paf cassette. With this combined approach, the three P. expansum proteins were successfully produced, purified and characterized. None of the four AFPs produced in this work were cytotoxic against mammal erythrocytes. The P. expansum AfpA followed by the P. digitatum AfpB were the most active AFPs against filamentous fungi, including plant and human pathogens, mycotoxin-producer fungi, and their own producers, a feature that had not been previously described for AFPs. Moreover, AfpA from P. expansum and AfpB from P. digitatum protected against fungal infection caused by Botrytis cinerea in tomato plants, and additionally the P. expansum AfpA protected against P. digitatum in orange fruits. These results confirm our previous hypothesis that AFPs are good candidates for the development of antifungals in plant protection and postharvest conservation, but also in clinic or food preservation.