Thursday, 4 July at 12:00pm
Biochemistry Seminar Room BIG13
Prof Octavio Luiz Franco, Universidade Católica de Brasília
Antimicrobial peptide digital optimization enables the generation of bioinspired products.
Antimicrobial peptides (AMPs) have attracted considerable attention because of their multiple and complex mechanisms of action toward resistant bacteria. However, reports have increasingly highlighted how bacteria can escape AMP administration. Here, we have described using multiple strategies, including genetic and Joker algorithms, and an artificial intelligence strategy to design bioinspired AMPs derived from bacteria, plants, and animals. This approach yielded different peptide classes, including those with an unusually high proportion of cationic residues and hydrophobic counterparts. At least dozens of peptides emerged as prototype AMP among natural analogs screened for their activity against an engineered luminescent Pseudomonas strain. Peptides were further characterized in structure, activity, mechanism of action, and biotechnological potential for developing new compounds beneficial for human and animal health. Most novel peptides were unstructured in water and underwent a coil-to-helix transition in hydrophobic environments. This conformation was corroborated by NMR analysis in dodecylphosphocholine micelles, which revealed an α-helical structure. The generated Peptides caused a bactericidal effect at low micromolar concentrations on several resistant bacteria, causing membrane disruption without triggering depolarization but rather hyperpolarization. Finally, the large-scale production strategies used to prepare such peptides for the market were also discussed. In summary, the present work presents a computational approach to explore natural products to design short and potent peptide antibiotics that could be used against resistant bacteria.
