Membrane composition is a key factor that regulates the destructive activity of antimicrobial peptides and the non-leaky permeation of cell penetrating peptides in vivo. Hence, the choice of model membrane is a crucial aspect in NMR studies and should reflect the biological situation as closely as possible. Here, we explore the structure and dynamics of the short multifunctional peptide BP100 using a multinuclear solid-state NMR approach. The membrane alignment and mobility of this 11 amino acid peptide was studied in various synthetic lipid bilayers with different net charge, fluidity, and thickness, as well as in native biomembranes harvested from prokaryotic and eukaryotic cells. (19)F-NMR provided the high sensitivity and lack of natural abundance background that are necessary to observe a labelled peptide even in protoplast membranes from Micrococcus luteus and in erythrocyte ghosts. Six selectively (19)F-labeled BP100 analogues gave remarkably similar spectra in all of the macroscopically oriented membrane systems, which were studied under quasi-native conditions of ambient temperature and full hydration. This similarity suggests that BP100 has the same surface-bound helical structure and high mobility in the different biomembranes and model membranes alike, independent of charge, thickness or cholesterol content of the system. (31)P-NMR spectra of the phospholipid components did not indicate any bilayer perturbation, so the formation of toroidal wormholes or micellarization can be excluded as a mechanism of its antimicrobial or cell penetrating action. However, (2)H-NMR analysis of the acyl chain order parameter profiles showed that BP100 leads to considerable membrane thinning and thereby local destabilization.

Single D-amino acid substitutions can be used to suppress or slow down the aggregation of peptides into β-sheeted assemblies compared to the respective L-amino acids. Here, we investigate the influence of local stereochemistry in the model peptide [KIGAKI]3-NH2, which is known to form amyloid-like fibrils. To find out whether aggregation plays a role in various biologically relevant functions that involve peptide-lipid interactions, we studied the antimicrobial, hemolytic and fusogenic activities of this amphiphilic membrane-active molecule. The stiff and sterically constrained amino acid CF3-Bpg [3-(trifluoromethyl)-bicyclopent-[1,1,1]-1-ylglycine] was incorporated either as an L- or a D-enantiomer at different hydrophobic positions of the KIGAKI sequence. D-Epimers have a higher aggregation threshold than the L-epimers, yet the aggregation of both was confirmed using electron microscopy and circular dichroism. Solid-state (19)F-NMR analysis showed that the peptide aggregated in native membranes from human erythrocytes and bacterial protoplasts in the same way as in synthetic lipid bilayers. We then monitored the effect of the single L- or D-CF3-Bpg substitutions in KIGAKI on its distinct biological activities, which have to be measured at low peptide concentrations where the aggregation threshold cannot be directly assessed. These functional assays showed that the aggregation propensity of KIGAKI does not play a role in its antimicrobial action, but an increased tendency to aggregate promotes other undesirable effects such as hemolysis and membrane fusion. These results confirm the membranolytic and thereby toxic nature of amyloidogenic peptides, and emphasize the unpredictable role of peptide aggregation in the different assays used to study biological activities.

To understand the molecular mechanisms of amphiphilic membrane-active peptides, one needs to study their interactions with lipid bilayers under ambient conditions. However, it is difficult to control the pH of the sample in biophysical experiments that make use of mechanically aligned multilamellar membrane stacks on solid supports. HPLC-purified peptides tend to be acidic and can change the pH in the sample significantly. Here, we have systematically studied the influence of pH on the lipid interactions of the antimicrobial peptide PGLa embedded in oriented DMPC/DMPG bilayers. Using solid-state NMR (31P, 2H, 19F), both the lipid and peptide components were characterized independently, though in the same oriented samples under typical conditions of maximum hydration. The observed changes in lipid polymorphism were supported by DSC on multilamellar liposome suspensions. On this basis, we can present an optimized sample preparation protocol and discuss the challenges of performing solid-state NMR experiments under controlled pH. DMPC/DMPG bilayers show a significant up-field shift and broadening of the main lipid phase transition temperature when lowering the pH from 10.0 to 2.6. Both, strongly acidic and basic pH, cause a significant degree of lipid hydrolysis, which is exacerbated by the presence of PGLa. The characteristic re-alignment of PGLa from a surface-bound to a tilted state is not affected between pH of 7 to 4 in fluid bilayers. On the other hand, in gel-phase bilayers the peptide remains isotropically mobile under acidic conditions, displays various co-existing orientational states at pH7, and adopts an unknown structural state at basic pH.

Many antimicrobial peptides are synthesized non-ribosomally in bacteria, but little is known about their subcellular route of biosynthesis, their mode of intracellular accumulation, or their role in the physiology of the producer cells. Here, we present a comprehensive view on the biosynthesis of gramicidin S (GS) in Aneurinibacillus migulanus, having observed a peripheral membrane localization of its synthetases. The peptide gets accumulated in nano-globules, which mature by fusion into larger granules and end up within vacuolar structures. These granules serve as energy storage devices, as they contain GS molecules that are non-covalently attached to alkyl phosphates and protect them from dephosphorylation and premature release of energy. This finding of a fundamentally new type of high-energy phosphate storage mechanism can explain the curious role of GS biosynthesis in the physiology of the bacterial producer cells. The unknown role of the GrsT protein, which is part of the non-ribosomal GS synthetase operon, can thus be assumed to be responsible for the biosynthesis of alkyl phosphates. GS binding to alkyl phosphates may suggest its general affinity to phosphagens such as ATP and GTP, which can represent the important intracellular targets in pathogenic bacteria.

No abstract