Novel combinatorial libraries consisting of simplified amino acid sequences were designed to screen for peptides active against theCandida albicans membrane. A novel decapeptide, KKVVFKVKFK, that had a unique primary amino acid sequence was identified in this work. This peptide irreversibly inhibited the growth of C. albicans and showed a broad range of antibacterial activity but no hemolytic activity. Circular dichroism spectra revealed that the predominant secondary structure of this peptide strongly depended on the membrane-mimetic environments; the peptide preferred to form an amphipathic α-helical structure in the presence of 50% trifluoroethanol, while it preferred to adopt a distorted α-helical structure in the presence of sodium dodecyl sulfate micelles. Experiments in which dye was released from vesicles indicated that this novel antimicrobial peptide killed microorganisms through the action on the membrane as its primary target. Replacement of amino acids in this active decapeptide on the basis of information from the libraries could provide unique information about factors affecting its antimicrobial activity such as its secondary structure, net positive charge, and hydrophobicity.

An active fragment was identified from tenecin 1, an antibacterial protein belonging to the insect defensin family, by synthesizing the peptides corresponding to the three regions of tenecin 1. Only the fragment corresponding to the C-terminal beta-sheet domain showed activity against fungi as well as Gram-positive and Gram-negative bacteria, whereas tenecin 1, the native protein, showed activity only against Gram-positive bacteria. CD spectra indicated that each fragment in a membrane-mimetic environment might adopt a secondary structure corresponding to its region in the protein. The leakage of dye from liposomes induced by this fragment suggested that this fragment acts on the membrane of pathogens as a primary mode of action. A comparison between the structure and the activity of each fragment indicated that a net positive charge was a prerequisite factor for activity. To the best of our knowledge this is the first report in which the fragment corresponding to the beta-sheet region in antibacterial proteins, which consists of alpha-helical and beta-sheet regions, has been identified as a primary active fragment.