Antimicrobial peptide activity in asymmetric bacterial membrane mimics

Abstract

We report on the response of asymmetric lipid membranes composed of palmitoyl oleoyl phosphatidylethanolamine and palmitoyl oleoyl phosphatidylglycerol, to interactions with the frog peptides L18W-PGLa and magainin 2 (MG2a), as well as the lactoferricin derivative LF11-215. In particular we determined the peptide-induced lipid flip-flop, as well as membrane partitioning of L18W-PGLa and LF11-215, and vesicle dye-leakage induced by L18W-PGLa. The ability of L18W-PGLa and MG2a to translocate through the membrane appears to correlate with the observed lipid flip-flop, which occurred at the fastest rate for L18W-PGLa. The higher structural flexibility of LF11-215 in turn allows this peptide to insert into the bilayers without detectable changes of membrane asymmetry. The increased vulnerability of asymmetric membranes to L18W-PGLa in terms of permeability, appears to be a consequence of tension differences between the compositionally distinct leaflets, but not due to increased peptide partitioning.

Fig. 1. Measurement principle and scattering contrast between (POPG-d31)ⁱⁿ/(POPE/POPG-d31)^out aLUVs and ScraLUVs in HBSD buffer, as observed by SANS at 37 °C. Scattering contrast was additionally enhanced by multiplying the scattered intensities with q³.

Fig. 2. Kinetics of L18W-PGLa-induced dye efflux from (POPG)ⁱⁿ/(POPE/POPG)^out aLUVs, ScraLUVs and OLM for [P]/[L] = 1 : 400 ([L] = 50 μM, T = 37 °C).

Fig. 3. Mole fraction of membrane-partitioned peptides (A), partitioning coefficient (B) and ratio of partitioned peptides (C) as a function of total LF11-215 (open symbols) and L18W-PGLa (filled symbols) concentrations and [L] = 50 μM. Data refer to aLUVs (circles), ScraLUVs (squares) and OLM (triangles). The gray line in (A) represents the limit [P]/[L] corresponding to f_B = 1; all other lines are guides for the eye. In the case of L18W-PGLa, the arrows indicate a realistic propagation for [P] < 0.5 μM.

Fig. 4. Decay of scattering contrast between aLUVs and scrambled LUVs due to L18W-PGLa-mediated lipid flip-flop at [L] = 9 mM and different [L]/[P]. As a control, aLUV data in the absence of peptide are also shown.