Antimicrobial Peptide Recognition of Bacterial Membranes

Abstract

In the face of increasing clinical resistance to existing antibiotic drugs, antimicrobial peptides (AMP) are seen as a source of new antibiotics. Since these AMP have high affinities for bacterial membranes, but with highly variable activities over different strains, a multidisciplinary approach combining solid-state NMR, microbiology and lipidomics has been taken to decipher their specific recognition. Most studies of AMP have been carried out using model membrane systems, such as phosphatidylcholine (PC) and phosphatidylglycerol (PG) bilayers, which bear questionable resemblance to bacterial membranes. We have expanded these studies to include other bacterial phospholipids, e.g. phosphatidylethanolamine, cardiolipin (CL) and natural lipid extracts. Model membranes prepared from bacterial lipid extracts and live pathogenic bacteria are being used to help determine the mechanism of action of a class of AMP obtained from Australian tree frogs. The aims are to determine at which growth stage of the bacteria the peptides are active and to correlate with the membrane composition at each stage using mass spectrometry of the extracted phospholipids (lipidomics). Ongoing solid-state NMR experiments are providing in vivo AMP activity spectra at atomic resolution. This includes monitoring of phospholipid-peptide association and identification of specific bacterial features, such as acyl chain lipid composition, as critical mechanisms of antibiotic resistance. We report AMPs that have shown a better affinity for CL-containing vesicles, than for PC or a PC/PG mixture, and promote lamellar lipid organization to break into isotropic, as yet unsolved, structures. These correlate with the lower minimum inhibitory concentration assays done on a range of bacteria and circular dichroism experiments of AMP with specific lipid composition as those used in preliminary in vivo NMR experiments.