Selection and redesign for high selectivity of membrane-active antimicrobial peptides from a dedicated sequence/function database

Abstract Antimicrobial peptides (AMPs) are plausible candidates for the development of novel classes of antibiotics with a low tendency to elicit resistance. They often form lesions in the bacterial membrane making it hard for bacteria to develop permanent resistance. However, a potent antibacterial activity is often accompanied by excessive cytotoxicity towards host cells. Modifying known natural sequences, based on desirable biophysical properties, is expensive and time-consuming and often with limited success. ‘Mutator’ is a freely available web-based computational tool for suggesting residue variations that potentially increase a peptide's selectivity, based on the use of quantitative structure activity relationship (QSAR) criteria. Although proven to be successful, it has never been used to analyze multiple sequences simultaneously. Modifying the Mutator algorithm allowed screening of many sequences in the dedicated Database of Anuran Defense Peptides (DADP) and by implementing limited amino acid substitutions on appropriate candidates, propose 8 potentially selective AMPs called Dadapins. Two were chosen for testing, confirming the prediction and validating this approach. They were shown to efficiently inactivate bacteria by disrupting their membranes but to be non-toxic for host cells, as determined by flow cytometry and confirmed by atomic force microscopy (AFM).