Antimicrobial Peptide PGLa Research Articles

Conditions affecting the re-alignment of the antimicrobial peptide PGLa in membranes as monitored by solid state 2H-NMR

The cationic antimicrobial peptide PGLa is electrostatically attracted to bacterial membranes, binds as an amphiphilic α-helix, and is thus able to permeabilize the lipid bilayer. Using solid state 2H-NMR of non-perturbing Ala-d 3 labels on the peptide, we have characterized the helix alignment under a range of different conditions. Even at a very high peptide-to-lipid ratio (1:20) and in the presence of negatively charged lipids, there was no indication of a toroidal wormhole structure. Instead, PGLa re-aligns from a surface-bound S-state to an obliquely tilted T-state, which is presumably dimeric. An intermediate structure half-way between the S- and T-state was observed in fully hydrated multilamellar DMPC vesicles at 1:50, suggesting a fast exchange between the two states on the time scale of >50 kHz. We demonstrate that this equilibrium is shifted from the S- towards the T-state either upon (i) increasing the peptide concentration, (ii) adding negatively charged DMPG, or (iii) decreasing the level of hydration. The threshold concentration for re-alignment in DMPC is found to be between 1:200 and 1:100 in oriented samples at 96% humidity. In fully hydrated multilamellar DMPC vesicles, it shifts to an effective peptide-to-lipid ratio of 1:50 as some peptides are able to escape into the bulk water phase.

Solid-State NMR Analysis of the PGLa Peptide Orientation in DMPC Bilayers: Structural Fidelity of 2H-Labels versus High Sensitivity of 19F-NMR

The structure and alignment of the amphipathic α-helical antimicrobial peptide PGLa in a lipid membrane is determined with high accuracy by solid-state 2H-NMR. Orientational constraints are derived from a series of eight alanine-3,3,3-d 3-labeled peptides, in which either a native alanine is nonperturbingly labeled (4×), or a glycine (2×) or isoleucine (2×) is selectively replaced. The concentration dependent realignment of the α-helix from the surface-bound “S-state” to a tilted “T-state” by 30° is precisely calculated using the quadrupole splittings of the four nonperturbing labels as constraints. The remaining, potentially perturbing alanine-3,3,3-d 3 labels show only minor deviations from the unperturbed peptide structure and help to single out the unique solution. Comparison with previous 19F-NMR constraints from 4-CF 3-phenylglycine labels shows that the structure and orientation of the PGLa peptide is not much disturbed even by these bulky nonnatural side chains, which contain CF 3 groups that offer a 20-fold better NMR sensitivity than CD 3 groups.

Dynamics of the Antimicrobial Peptide PGLa Action on Escherichia coli Monitored by Atomic Force Microscopy

Atomic force microscopy (AFM) images of living cells in physiological solution were used to monitor the different stages involved in the interaction between Escherichia coli and the antimicrobial peptide PGLa. Damage on bacterial membranes was observed in the past using standard electron microscopy; stiffness measurements and images scanned in physiological solution demonstrate the advantage of AFM for such studies. From force versus separation curve measurements it is possible to determine the variation of the cellular stiffness. PGLa action on components of the cell structure like the outer membrane, the bacterial pili, the peptidoglycan wall and the inner membrane was determined by the comparison of AFM images of bacteria before and after PGLa addition. The interaction of Escherichia coli with PGLa in the culture medium has two stages. The first is characterized by the loss of surface stiffness and the formation of micelles probably originating from the disruption of the outer membrane and the loss of the bacteria’s ability to adhere to the substrates. In the second stage there is further damage, which resulted in total cell rupture. AFM images of bacteria in air and surface roughness measurements were also used to estimate peptide damage.

Concentration-Dependent Realignment of the Antimicrobial Peptide PGLa in Lipid Membranes Observed by Solid-State 19F-NMR

The membrane-disruptive antimicrobial peptide PGLa is found to change its orientation in a dimyristoyl-phosphatidylcholine bilayer when its concentration is increased to biologically active levels. The alignment of the alpha-helix was determined by highly sensitive solid-state NMR measurements of (19)F dipolar couplings on CF(3)-labeled side chains, and supported by a nonperturbing (15)N label. At a low peptide/lipid ratio of 1:200 the amphiphilic peptide resides on the membrane surface in the so-called S-state, as expected. However, at high peptide concentration (>/=1:50 molar ratio) the helix axis changes its tilt angle from approximately 90 degrees to approximately 120 degrees , with the C-terminus pointing toward the bilayer interior. This tilted "T-state" represents a novel feature of antimicrobial peptides, which is distinct from a membrane-inserted I-state. At intermediate concentration, PGLa is in exchange between the S- and T-state in the timescale of the NMR experiment. In both states the peptide molecules undergo fast rotation around the membrane normal in liquid crystalline bilayers; hence, large peptide aggregates do not form. Very likely the obliquely tilted T-state represents an antiparallel dimer of PGLa that is formed in the membrane at increasing concentration.

Orientation of the antimicrobial peptide PGLa in lipid membranes determined from 19F-NMR dipolar couplings of 4-CF 3-phenylglycine labels

A highly sensitive solid state 19F-NMR strategy is described to determine the orientation and dynamics of membrane-associated peptides from specific fluorine labels. Several analogues of the antimicrobial peptide PGLa were synthesized with the non-natural amino acid 4-trifluoromethyl-phenylglycine (CF 3-Phg) at different positions throughout the α-helical peptide chain. A simple 1-pulse 19F experiment allows the simultaneous measurement of both the anisotropic chemical shift and the homonuclear dipolar coupling within the rotating CF 3-group in a macroscopically oriented membrane sample. The value and sign of the dipolar splitting determines the tilt of the CF 3-rotational axis, which is rigidly attached to the peptide backbone, with respect to the external magnetic field direction. Using four CF 3-labeled peptide analogues (with l-CF 3-Phg at Ile9, Ala10, Ile13, and Ala14) we confirmed that PGLa is aligned at the surface of lipid membranes with its helix axis perpendicular to the bilayer normal at a peptide:lipid ratio of 1:200. We also determined the azimuthal rotation angle of the helix, which agrees well with the orientation expected from its amphiphilic character. Peptide analogues with a d-CF 3-Phg label resulting from racemization of the amino acid during synthesis were separately collected by HPLC. Their spectra provide additional information about the PGLa structure and orientation but allow only to discriminate qualitatively between multiple solutions. The structural and functional characterization of the individual CF 3-labeled peptides by circular dichroism and antimicrobial assays showed only small effects for our four substitutions on the hydrophobic face of the helix, but a significant disturbance was observed in a fifth analogue where Ala8 on the hydrophilic face had been replaced. Even though the hydrophobic CF 3-Phg side chain cannot be utilized in all positions, it allows highly sensitive NMR measurements over a wide range of experimental conditions and dynamic regimes of the peptide.

4-fluorophenylglycine as a label for 19F NMR structure analysis of membrane-associated peptides.

The non-natural amino acid 4-fluorophenylglycine (4F-Phg) was incorporated into several representative membrane-associated peptides for dual purpose. The (19)F-substituted ring is directly attached to the peptide backbone, so it not only provides a well-defined label for highly sensitive (19)F NMR studies but, in addition, the D and L enantiomers of the stiff side chain may serve as reporter groups on the transient peptide conformation during the biological function. Besides peptide synthesis, which is accompanied by racemisation of 4F-Phg, we also describe separation of the epimers by HPLC and removal of trifluoroacetic acid. As a first example, 18 different analogues of the fusogenic peptide "B18" were prepared and tested for induction of vesicle fusion; the results confirmed that hydrophobic sites tolerated 4F-Phg labelling. Similar fusion activities within each pair of epimers suggest that the peptide is less structured in the fusogenic transition state than in the helical ground state. In a second example, five doubly labelled analogues of the antimicrobial peptide gramicidin S were compared by using bacterial growth inhibition assays. This cyclic beta-sheet peptide could accommodate both L and D substituents on its hydrophobic face. As a third example, we tested six analogues of the antimicrobial peptide PGLa. The presence of d-4F-Phg reduced the biological activity of the peptide by interfering with its amphiphilic alpha-helical fold. Finally, to illustrate the numerous uses of l-4F-Phg in (19)F NMR spectroscopy, we characterised the interaction of labelled PGLa with uncharged and negatively charged membranes. Observing the signal of the free peptide in an aqueous suspension of unilamellar vesicles, we found a linear saturation behaviour that was dominated by electrostatic attraction of the cationic PGLa. Once the peptide is bound to the membrane, however, solid-state (19)F NMR spectroscopy of macroscopically oriented samples revealed that the charge density has virtually no further influence on the structure, alignment or mobility of the peptide.

Effects of the antimicrobial peptide PGLa on live Escherichia coli

The activity of PGLa, an antimicrobial peptide isolated from hemocytes of frog skin and its secretions on living Escherichia coli, was investigated by imaging the cells with atomic force microscopy (AFM) in physiological conditions and by measuring its cellular stiffness. The treatment of bacteria with the antimicrobial peptide PGLa in the culture medium had two stages. The first was characterized by the loss of surface stiffness and consequent loss of bacteria topographic features and the formation of micelles probably originating from the disruption of the outer membrane. The formation of outer membrane originated micelles is in agreement with the carpet-like mechanism of action proposed for antimicrobial peptides of the magainin family. The peptide action also resulted in the removal of bacterial pili. In a second stage there was further damage which resulted in total cell rupture. The addition of Mg 2+ ions prior to peptide treatment partially inhibited the effects of PGLa on bacteria. This result suggests that PGLa interacts with the outer membrane by displacing Mg 2+ from LPS, inserting itself into the bilayer and cross-bridging the negative charges of LPS lipids as proposed in the self-promoted pathway mechanism. The peptide effect on the bacteria was compared to the activity of the chelating agent EDTA that damages the bacterial outer membrane by removing Mg 2+ ions.

Susceptibility corrections in solid-state NMR experiments with oriented membrane samples. Part I: applications

Chemical shift referencing of solid-state NMR experiments on oriented membranes has to compensate for bulk magnetic susceptibility effects that are associated with the non-spherical sample shape, as described in the accompanying paper [J. Magn. Reson. 164 (2003) 115–127]. The resulting frequency deviations can be on the order of 10 ppm, which is serious for nuclei with a narrow chemical shift anisotropy such as 1H or 13C, and in some cases even 19F. Two referencing schemes are proposed here to compensate for these effects: A flat (0.4 mm) glass container with an isotropic reference molecule dissolved in a thin film of liquid is stacked on top of the oriented membrane sample. Alternatively, the intrinsic proton signal of the hydrated lipid can be used for chemical shift referencing. Further aspects related to magnetic susceptibility are discussed, such as air gaps in susceptibility-matched probeheads, the benefits of shimming, and limitations in the accuracy of orientational constraints. A biological application is illustrated by a series of experiments on the antimicrobial peptide PGLa, aimed at understanding its concentration-dependent membranolytic effect. To address a wide range of molar peptide/lipid ratios between 1:3000 and 1:8, multilayers of hydrated DMPC containing a 19F-labeled peptide were oriented between stacked glass plates. Maintaining an approximately constant amount of peptide gives rise to thick samples (18 plates) at low, and thin samples (3 plates) at high peptide/lipid ratio. Accurate referencing was critical to reveal a small but significant change over 5 ppm in the anisotropic chemical shift of the 19F label on the peptide, indicative of a change in the orientation and/or dynamics of PGLa in the membrane.

Lipid discrimination in phospholipid monolayers by the antimicrobial frog skin peptide PGLa. A synchrotron X-ray grazing incidence and reflectivity study.

We present a first study using synchrotron grazing incidence diffraction and X-ray reflectivity measurements on mixed phospholipid/peptide monolayers at the air/water interface. The thermodynamic properties of the pure and mixed monolayers were characterized using the classical film balance technique. Surface pressure/potential-area isotherms showed that the antimicrobial frog skin peptide PGLa formed a very stable monolayer with two PGLa molecules per kinetic unit and a collapse pressure of ~22 mN/m. X-ray grazing incidence diffraction indicated that the peptide-dimer formation did not lead to self-aggregation with subsequent crystallite formation. However, the scattering length density profiles derived from X-ray reflectivity measurements yield information on the PGLa monolayer that protrudes into the air phase by about 0.8 nm, suggesting that the peptide is aligned parallel to the air/water interface. The monolayers, composed of disaturated phosphatidylcholines or phosphatidylglycerols, were stable up to 60 mN/m and exhibited a first-order transition from a liquid-expanded to a liquid-condensed state around 10 mN/m. Structural details of the phospholipid monolayers in the presence and absence of PGLa were obtained from synchrotron experiments. Thereby, the X-ray data of distearoylphosphatidylcholine/PGLa can be analyzed by being composed of the individual components, while the peptide strongly perturbed the lipid acyl chain order of distearoylphosphatidylglycerol. These results are in agreement that PGLa mixes at a molecular level with negatively charged lipids, but forms separate islands in zwitterionic phosphatidylcholine monolayers and demonstrates that antimicrobial peptides can discriminate between the major phospholipid components of bacterial and mammalian cytoplasmic membranes.

Raman spectroscopy of synthetic antimicrobial frog peptides magainin 2a and PGLa

Magainin and PGLa are 23- and 21-residue peptides isolated from the skin of the African clawed frog Xenopus laevis. They protect the frog from infection and exhibit a broad-spectrum antimicrobial activity in vitro. The mechanism of this activity involves the interaction of magainin with microbial membranes. We have measured the secondary structure and membrane-perturbing ability of these peptides to obtain information about this mechanism. Our results show that mgn2a forms a helix with an average length of less than 20 A upon binding to liposomes. At high concentrations (50 mg/mL) mgn2a spontaneously solubilizes phosphatidylcholine liposomes at temperatures above the gel-liquid-crystalline phase transition. Mgn2a appears to bind to the surface of liposomes made of negatively charged lipids without spontaneously penetrating the bilayer. Finally, mgn2a and PGLa interact together with liposomes in a synergistic way that enhances the helix content of one or both of the peptides and allows the peptides to more easily penetrate the bilayer. PGLa mixed with a small nonperturbing amount of magainin 2 amide is 25-43 times as potent as PGLa alone at inducing the release of carboxyfluorescein from liposomes. The results suggest that the mechanism of antimicrobial activity does not involve a channel formed by transmembrane helical peptides.