Combinatorial libraries: a tool to design antimicrobial and antifungal peptide analogues having lytic specificities for structure-activity relationship studies.

In the race for supremacy, microbes are sprinting ahead. This warning by the World Health Organization clearly demonstrates that the spread of antibiotic-resistant bacteria leads to a global health problem and that antibiotics never seen before by bacteria are urgently needed. Antimicrobial peptides represent such a source for novel antibiotics due to their rapid lytic activity (within minutes) through disruption of cell membranes. However, due to the similarities between bacterial, fungal, and mammalian plasma cell membranes, a large number of antimicrobial peptides have low lytic specificities and exhibit a broad activity spectrum and/or significant toxic effect toward mammalian cells. Mutation strategies have allowed the development of analogues of existing antimicrobial peptides having greater lytic specificities, although such methods are lengthy and would be more efficient if the molecular mechanisms of action of antimicrobial peptides were clearly elucidated. Synthetic combinatorial library approaches have brought a new dimension to the design of novel biologically active compounds. Thus, a set of peptide analogues were generated based on the screening of a library built around an existing lytic peptide, and on a deconvolution strategy directed toward activity specificity. These peptide analogues also served as model systems to further study the effect of biomembrane mimetic systems on the peptides structural behavior relevant to their biological activities. Copyright 2000 John Wiley & Sons, Inc.