Antimicrobial peptides (AMPs) are promising candidates for next generation antimicrobial therapeutics. In mammalian cell membranes cholesterol plays a key regulatory function in antibiotic drug resistance and the immune response. However, the mechanism of action and the selectivity of AMPs towards bacterial membranes are not yet well understood. We hypothesized that differences in biophysical properties of mammalian versus bacterial membranes give rise to differences in peptide-membrane interactions which underlie the AMP mechanism and selectivity. Solid-state 2H NMR spectroscopy was employed to study di-monounsaturated phosphatidylcholine and phosphatidylethanolaminelipids (DOPC versus DOPE) bilayers in the presence of the antimicrobial peptide LL37 (2−4 mol%) and cholesterol (20−30% mol) as model systems mimicking mammalian and bacterial membranes. The lipid systems were probed with POPC-d31 (10 mol%) where the residual quadrupolar couplings yielded segmental order parameters (SCD) for the individual acyl segments. The LL37-peptide decreased the quadrupolar splittings indicating it increased the area per lipid, while cholesterol showed the opposite. The mean-torque model was used to calculate the key bilayer properties of area per lipid and bilayer thickness [2]. The model-free functional dependence of spin-lattice relaxation rates on SCD (square-law plots) indicated greater bending rigidity with cholesterol for both DOPC and DOPE membranes. By contrast LL37 decreased the bending rigidity of the membranes. Additionally all-atom molecular dynamics simulations for bilayers comprising POPE/DMPG/cardiolipin (90:5:5) and DMPC/cholesterol (90:10) were conducted with LL37 peptides on the bilayer surface to capture the molecular-level interactions. We observed partial insertion of the LL37-peptides in the POPE system, while the peptides were repelled from the membrane in the DMPC/cholesterol system. These observations reveal the surprising rivalry between cholesterol and AMPs which underlies AMP mechanism and selectivity.
Read full abstract