Molecular mechanism of antimicrobial peptides: the origin of cooperativity

Based on very extensive studies on four peptides (alamethicin, melittin, magainin and protegrin), we propose a mechanism to explain the cooperativity exhibited by the activities of antimicrobial peptides, namely, a non-linear concentration dependence characterized by a threshold and a rapid rise to saturation as the concentration exceeds the threshold. We first review the structural basis of the mechanism. Experiments showed that peptide binding to lipid bilayers creates two distinct states depending on the bound-peptide to lipid ratio P/L. For P/L below a threshold P/L*, all of the peptide molecules are in the S state that has the following characteristics: (1) there are no pores in the membrane, (2) the axes of helical peptides are oriented parallel to the plane of membrane, and (3) the peptide causes membrane thinning in proportion to P/L. As P/L increases above P/L*, essentially all of the excessive peptide molecules occupy the I state that has the following characteristics: (1) transmembrane pores are detected in the membrane, (2) the axes of helical peptides are perpendicular to the plane of membrane, (3) the membrane thickness remains constant for P/L> or =P/L*. The free energy based on these two states agrees with the data quantitatively. The free energy also explains why lipids of positive curvature (lysoPC) facilitate and lipids of negative curvature (PE) inhibit pore formation.     

Molecular mechanism of antimicrobial peptides: The origin of cooperativity

Based on very extensive studies on four peptides (alamethicin, melittin, magainin and protegrin), we propose a mechanism to explain the cooperativity exhibited by the activities of antimicrobial peptides, namely, a non-linear concentration dependence characterized by a threshold and a rapid rise to saturation as the concentration exceeds the threshold. We first review the structural basis of the mechanism. Experiments showed that peptide binding to lipid bilayers creates two distinct states depending on the bound-peptide to lipid ratio P/L. For P/L below a threshold P/L*, all of the peptide molecules are in the S state that has the following characteristics: (1) there are no pores in the membrane, (2) the axes of helical peptides are oriented parallel to the plane of membrane, and (3) the peptide causes membrane thinning in proportion to P/L. As P/L increases above P/L*, essentially all of the excessive peptide molecules occupy the I state that has the following characteristics: (1) transmembrane pores are detected in the membrane, (2) the axes of helical peptides are perpendicular to the plane of membrane, (3) the membrane thickness remains constant for P/L ≥ P/L*. The free energy based on these two states agrees with the data quantitatively. The free energy also explains why lipids of positive curvature (lysoPC) facilitate and lipids of negative curvature (PE) inhibit pore formation.     

Cooperative antimicrobial action of melittin on lipid membranes: A coarse-grained molecular dynamics study

We conducted a series of coarse-grained molecular dynamics (CG-MD) simulations to investigate the complicated actions of melittin, which is an antimicrobial peptide (AMP) derived from honey bee venom, on a lipid membrane. To accurately simulate the AMP action, we developed and used a protein CG model as an extension of the pSPICA force field (FF), which was designed to reproduce several thermodynamic quantities and structural properties. At a low peptide-to-lipid (P/L) ratio (1/102), no defect was detected. At P/L = 1/51, toroidal pore formation was observed due to collective insertion of multiple melittin peptides from the N-termini. The pore formation was initiated by a local increase in membrane curvature in the vicinity of the peptide aggregate. At a higher P/L ratio (1/26), two more modes were detected, seemingly not controlled by the P/L ratio but by a local arrangement of melittin peptides: 1. Pore formation accompanied by lipid extraction by melittin peptides:a detergent-like mechanism. 2. A rapidly formed large pore in a significantly curved membrane: bursting. Thus, we observed three pore formation modes (toroidal pore formation, lipid extraction, and bursting) depending on the peptide concentration and local arrangement. These observations were consistent with experimental observations and hypothesized melittin modes. Through this study, we found that the local arrangements and population of melittin peptides and the area expansion rate by membrane deformation were key to the initiation of and competition among the multiple pore formation mechanisms.     

Evidence for membrane thinning effect as the mechanism for peptide-induced pore formation

Antimicrobial peptides have two binding states in a lipid bilayer, a surface state S and a pore-forming state I. The transition from the S state to the I state has a sigmoidal peptide-concentration dependence indicating cooperativity in the peptide-membrane interactions. In a previous paper, we reported the transition of alamethicin measured in three bilayer conditions. The data were explained by a free energy that took into account the membrane thinning effect induced by the peptides. In this paper, the full implications of the free energy were tested by including another type of peptide, melittin, that forms toroidal pores, instead of barrel-stave pores as in the case of alamethicin. The S-to-I transitions were measured by oriented circular dichroism. The membrane thinning effect was measured by x-ray diffraction. All data were in good agreement with the theory, indicating that the membrane thinning effect is a plausible mechanism for the peptide-induced pore formations.     

Free energy analysis of membrane pore formation process in the presence of multiple melittin peptides

Understanding the molecular mechanism underlying pore formation in lipid membranes by antimicrobial peptides is of great importance in biological sciences as well as in drug design applications. Melittin has been widely studied as a pore forming peptide, though the molecular mechanism for pore formation is still illusive. We examined the free energy barrier for the creation of a pore in lipid membranes with and without multiple melittin peptides. It was found that six melittin peptides significantly stabilized a pore, though a small barrier (a few k_BT) for the formation still existed. With five melittin peptides or fewer, the pore formation barrier was much higher, though the established pore was in a local energy minimum. Although seven melittins effectively reduced the free energy barrier, a single melittin peptide left the pore after a long time MD simulation probably because of the overcrowded environment around the bilayer pore. Thus, it is highly selective for the number of melittin peptides to stabilize the membrane pore, as was also suggested by the line tension evaluations. The free energy cost required to insert a single melittin into the membrane is too high to explain the one-by-one insertion mechanism for pore formation, which also supports the collective melittin mechanism for pore formation.     

On the mechanism of pore formation by melittin

The mechanism of pore formation of lytic peptides, such as melittin from bee venom, is thought to involve binding to the membrane surface, followed by insertion at threshold levels of bound peptide. We show that in membranes composed of zwitterionic lipids, i.e. phosphatidylcholine, melittin not only forms pores but also inhibits pore formation. We propose that these two modes of action are the result of two competing reactions: direct insertion into the membrane and binding parallel to the membrane surface. The direct insertion of melittin leads to pore formation, whereas the parallel conformation is inactive and prevents other melittin molecules from inserting, hence preventing pore formation.     

Toroidal pores formed by antimicrobial peptides show significant disorder

A large variety of antimicrobial peptides have been shown to act, at least in vitro, by poration of the lipid membrane. The nanometre size of these pores, however, complicates their structural characterization by experimental techniques. Here we use molecular dynamics simulations, to study the interaction of a specific class of antimicrobial peptides, melittin, with a dipalmitoylphosphatidylcholine bilayer in atomic detail. We show that transmembrane pores spontaneously form above a critical peptide to lipid ratio. The lipid molecules bend inwards to form a toroidally shaped pore but with only one or two peptides lining the pore. This is in strong contrast to the traditional models of toroidal pores in which the peptides are assumed to adopt a transmembrane orientation. We find that peptide aggregation, either prior or after binding to the membrane surface, is a prerequisite to pore formation. The presence of a stable helical secondary structure of the peptide, however is not. Furthermore, results obtained with modified peptides point to the importance of electrostatic interactions in the poration process. Removing the charges of the basic amino-acid residues of melittin prevents pore formation. It was also found that in the absence of counter ions pores not only form more rapidly but lead to membrane rupture. The rupture process occurs via a novel recursive poration pathway, which we coin the Droste mechanism.     

Melittin-induced bilayer leakage depends on lipid material properties: evidence for toroidal pores

The membrane-lytic peptide melittin has previously been shown to form pores in lipid bilayers that have been described in terms of two different structural models. In the "barrel stave" model the bilayer remains more or less flat, with the peptides penetrating across the bilayer hydrocarbon region and aggregating to form a pore, whereas in the "toroidal pore" melittin induces defects in the bilayer such that the bilayer bends sharply inward to form a pore lined by both peptides and lipid headgroups. Here we test these models by measuring both the free energy of melittin transfer (DeltaG degrees ) and melittin-induced leakage as a function of bilayer elastic (material) properties that determine the energetics of bilayer bending, including the area compressibility modulus (K(a)), bilayer bending modulus (k(c)), and monolayer spontaneous curvature (R(o)). The addition of cholesterol to phosphatidylcholine (PC) bilayers, which increases K(a) and k(c), decreases both DeltaG degrees and the melittin-induced vesicle leakage. In contrast, the addition to PC bilayers of molecules with either positive R(o), such as lysoPC, or negative R(o), such as dioleoylglycerol, has little effect on DeltaG degrees , but produces large changes in melittin-induced leakage, from 86% for 8:2 PC/lysoPC to 18% for 8:2 PC/dioleoylglycerol. We observe linear relationships between melittin-induced leakage and both K(a) and 1/R(o)(2). However, in contrast to what would be expected for a barrel stave model, there is no correlation between observed leakage and bilayer hydrocarbon thickness. All of these results demonstrate the importance of bilayer material properties on melittin-induced leakage and indicate that the melittin-induced pores are defects in the bilayer lined in part by lipid molecules.     

Influence of the arrangement and secondary structure of melittin peptides on the formation and stability of toroidal pores

Melittin interactions with lipid bilayers and melittin formed pores are extensively studied to understand the mechanism of the toroidal pore formation. Early experimental studies suggested that melittin peptide molecules are anchored by their positively charged residues located next to the C-terminus to only one leaflet of the lipid bilayer (asymmetric arrangement). However, the recent non-linear spectroscopic experiment suggests a symmetric arrangement of the peptides with the C-terminus of the peptides anchored to both bilayers. Therefore, we present here a computational study that compares the effect of symmetric and asymmetric arrangements of melittin peptides in the toroidal pore formation. We also investigate the role of the peptide secondary structure during the pore formation. Two sets of the symmetric and asymmetric pores are prepared, one with a helical peptide from the crystal structure and the other set with a less helical peptide. We observe a stable toroidal pore being formed only in the system with a symmetric arrangement of the less helical peptides. Based on the simulation results we propose that the symmetric arrangement of the peptides might be more favorable than the asymmetric arrangement, and that the helical secondary structure is not a prerequisite for the formation of the toroidal pore.     

Kinetics of melittin induced pore formation in the membrane of lipid vesicles.

We have investigated the permeabilization of POPC unilamellar vesicle bilayers upon the addition of melittin. This process was measured in an early time range of a few minutes by means of monitoring the release of an entrapped marker, the self-quenching fluorescent dye carboxyfluorescein. Pore formation is indicated by an apparent 'all-or-none' efflux out of individual vesicles and a higher than linear dependence on melittin concentration. Applying a recently developed evaluation procedure, the data are readily converted into the gross number of pores per vesicle formed within the elapsed measuring time t. The results can be generally described in terms of a fast initial rate of pore formation that slows down to a much lower value after a period of about 1 to 2 minutes, following a single exponential time course. The three rate parameters involved are shown to be power functions of the concentration of melittin that is actually associated with the vesicle membrane. These findings are in excellent quantitative agreement with a proposed scheme of reaction steps where the formation of lipid associated peptide dimers becomes rate determining once an initial fast deposit is exhausted.     

Membrane Bending Energy and Fusion Pore Kinetics in Ca-Triggered Exocytosis

A fusion pore composed of lipid is an obligatory kinetic intermediate of membrane fusion, and its formation requires energy to bend membranes into highly curved shapes. The energetics of such deformations in viral fusion is well established, but the role of membrane bending in Ca²⁺-triggered exocytosis remains largely untested. Amperometry recording showed that during exocytosis in chromaffin and PC12 cells, fusion pores formed by smaller vesicles dilated more rapidly than fusion pores formed by larger vesicles. The logarithm of 1/(fusion pore lifetime) varied linearly with vesicle curvature. The vesicle size dependence of fusion pore lifetime quantitatively accounted for the nonexponential fusion pore lifetime distribution. Experimentally manipulating vesicle size failed to alter the size dependence of fusion pore lifetime. Manipulations of membrane spontaneous curvature altered this dependence, and applying the curvature perturbants to the opposite side of the membrane reversed their effects. These effects of curvature perturbants were opposite to those seen in viral fusion. These results indicate that during Ca²⁺-triggered exocytosis membrane bending opposes fusion pore dilation rather than fusion pore formation. Ca²⁺-triggered exocytosis begins with a proteinaceous fusion pore with less stressed membrane, and becomes lipidic as it dilates, bending membrane into a highly curved shape.     

Investigation of toroidal pore and oligomerization by melittin using transmission electron microscopy

We studied the effects of melittin on various cell wall components and vesicles of various lipid compositions. To interact with the cytoplasmic membrane, melittin must traverse the cell wall, which is composed of oligosaccharides. Here, we found that melittin had a strong affinity for chitin, peptidoglycan, and lipopolysaccharide. We further examined the influence of lipid compositions on the lysis of the membranes by melittin. The result showed that melittin bound better to negatively charged than to zwitterionic lipid vesicles but was more potent at inducing leakage from zwitterionic lipid vesicles. Our studies further indicated that the oligomeric state of melittin varied between tetramers and octamers during the formation of toroidal pores. Dextran leakage experiments confirmed the formation and dimension of these toroidal pores. Finally, transmission electron microscopy revealed that melittin formed pores via peptide oligomerization by the toroidal pore-forming mechanism. The toroidal pores composed of 7-8 nm diameter rings that encircled 3.5-4.5 nm diameter cavities on zwitterionic lipid vesicles.     

Differential scanning calorimetry and X-ray diffraction studies of the specificity of the interaction of antimicrobial peptides with membrane-mimetic systems

Interest in biophysical studies on the interaction of antimicrobial peptides and lipids has strongly increased because of the rapid emergence of antibiotic-resistant bacterial strains. An understanding of the molecular mechanism(s) of membrane perturbation by these peptides will allow a design of novel peptide antibiotics as an alternative to conventional antibiotics. Differential scanning calorimetry and X-ray diffraction studies have yielded a wealth of quantitative information on the effects of antimicrobial peptides on membrane structure as well as on peptide location. These studies clearly demonstrated that antimicrobial peptides show preferential interaction with specific phospholipid classes. Furthermore, they revealed that in addition to charge-charge interactions, membrane curvature strain and hydrophobic mismatch between peptides and lipids are important parameters in determining the mechanism of membrane perturbation. Hence, depending on the molecular properties of both lipid and peptide, creation of bilayer defects such as phase separation or membrane thinning, pore formation, promotion of nonlamellar lipid structures or bilayer disruption by the carpet model or detergent-like action, may occur. Moreover, these studies suggest that these different processes may represent gradual steps of membrane perturbation. A better understanding of the mutual dependence of these parameters will help to elucidate the molecular mechanism of membrane damage by antimicrobial peptides and their target membrane specificity, keys for the rationale design of novel types of peptide antibiotics.     

A combined study of aggregation, membrane affinity and pore activity of natural and modified melittin

The pore activity of melittin and several chemically modified derivatives has been investigated using conductance measurements on planar lipid bilayers and marker release from small unilamellar vesicles. The modifications included N-terminal formylation, acetylation, succinylation and modification of the tryptophan residue. All of the compounds showed bilayer permeabilizing properties, though quantitative differences were evident. These comprised changes in the voltage dependence of the conductance, in the single-pore kinetics, in the concentration of aqueous peptide required to induce a given pore activity and in the apparent 'molecularity' reflected by the power law of its concentration dependence. A strong tendency for disrupting bilayers was not always correlated with strong pore activity. For a better understanding of these results, measurements of pore activity were complemented by studying the aggregation behavior in solution and the water-membrane partition equilibrium. Modifications of charged residues gave rise to significant changes in the aggregation properties, had virtually no influence on the partition coefficient. The latter decreased strongly, however, as a result of tryptophan modification. Analysis of the isotherms was consistent with the assumption that the arginine residues in melittin do not contribute very much to charge accumulation at the immediate membrane/water interface.     

Melittin-induced alterations in dynamic properties of human red blood cell membranes.

The interaction of bee venom melittin with erythrocyte membrane ghosts has been investigated by means of fluorescence quenching of membrane tryptophan residues, fluorescence polarization and ESR spectroscopy. It has been revealed that melittin induces the disorders in lipid-protein matrix both in the hydrophobic core of bilayer and at the polar/non-polar interface of melittin complexed with erythrocyte membranes. The peptide has been found to act most efficiently at the concentration of the order of 10(-10) mol/mg membrane protein. The apparent distance separating the membrane tryptophan and bound 1-anilino-8-naphthalenesulphonate (ANS) molecules is decreased upon melittin binding, which results in a significant increase of the maximum energy transfer efficiency. Significant changes in the fluorescence anisotropy of both 1,6-diphenyl-1,3,5-hexatriene and 1-anilino-8-naphthalenesulphonate bound to erythrocyte ghosts, which have been observed in the presence of melittin and crude venom, indicate membrane lipid bilayer rigidization. The effect of crude honey bee venom has been found to be of similar magnitude as the effect of pure melittin at the concentration of 10(-10) mol/mg membrane protein. Using two lipophilic spin labels, methyl 5-doxylpalmitate and 16-doxylstearic acid, we found that melittin at its increasing concentrations induces a well marked rigidization in the deeper regions of lipid bilayer, whereas the effect of rigidization near the membrane surface maximizes at the melittin concentration of 10(-10) mol/mg (10(-4) mol melittin per mole of membrane phospholipid). The decrease in the ratio hw/hs of maleimide and the rise in relative rotational correlation time (tau c) of iodacetamid spin label, indicate that melittin effectively immobilizes membrane proteins in the plane of the lipid bilayer. We conclude that melittin-induced rigidization of the lipid bilayer may induce a reorganization of lipid assemblies as well as the rearrangements in membrane protein pattern and consequently the alterations in lipid-protein interactions. Thus, the interaction of melittin with erythrocyte membranes is supposed to produce local conformational changes in membranes, which are discussed in the connection with their significance during the synergistic action of melittin and phospholipase of bee venom on red blood cells.     

Pore formation and translocation of melittin.

Melittin, a bee venom, is a basic amphiphilic peptide, which mainly acts on the lipid matrix of membranes, lysing various cells. To elucidate the molecular mechanism, we investigated its interactions with phospholipid vesicles. The peptide formed a pore with a short lifetime in the membrane, as revealed by the release of an anionic fluorescent dye, calcein, from the liposomes. Our new double-labeling method clarified that the pore size increased with the peptide-to-lipid ratio. Upon the disintegration of the pore, a fraction of the peptides translocated across the bilayer. The pore formation was coupled with the translocation, which was proved by three fluorescence experiments recently developed by our laboratory. A novel model for the melittin pore formation was discussed in comparison with other pore-forming peptides.     

Energetics and partition of two cecropin-melittin hybrid peptides to model membranes of different composition

The energetics and partition of two hybrid peptides of cecropin A and melittin (CA(1-8)M(1-18) and CA(1-7)M(2-9)) with liposomes of different composition were studied by time-resolved fluorescence spectroscopy, isothermal titration calorimetry, and surface plasmon resonance. The study was carried out with large unilamellar vesicles of three different lipid compositions: 1,2-dimyristoil-sn-glycero-3-phosphocholine (DMPC), 1,2-dimyristoyl-sn-glycero-3-phospho-rac-(1-glycerol) (DMPG), and a 3:1 binary mixture of DMPC/DMPG in a wide range of peptide/lipid ratios. The results are compatible with a model involving a strong electrostatic surface interaction between the peptides and the negatively charged liposomes, giving rise to aggregation and precipitation. A correlation is observed in the calorimetric experiments between the observed events and charge neutralization for negatively charged and mixed membranes. In the case of zwitterionic membranes, a very interesting case study was obtained with the smaller peptide, CA(1-7)M(2-9). The calorimetric results obtained for this peptide in a large range of peptide/lipid ratios can be interpreted on the basis of an initial and progressive surface coverage until a threshold concentration, where the orientation changes from parallel to perpendicular to the membrane, followed by pore formation and eventually membrane disruption. The importance of negatively charged lipids on the discrimination between bacterial and eukaryotic membranes is emphasized.     

Magainin 2 and PGLa in bacterial membrane mimics IV: Membrane curvature and partitioning

We previously reported that the synergistically enhanced antimicrobial activity of magainin 2 (MG2a) and PGLa is related to membrane adhesion and fusion. Here, we demonstrate that equimolar mixtures of MG2a and L18W-PGLa induce positive monolayer curvature stress and sense, at the same time, positive mean and Gaussian bilayer curvatures already at low amounts of bound peptide. The combination of both abilities—membrane curvature sensing and inducing—is most likely the base for the synergistically enhanced peptide activity. In addition, our coarse-grained simulations suggest that fusion stalks are promoted by decreasing the free-energy barrier for their formation rather than by stabilizing their shape. We also interrogated peptide partitioning as a function of lipid and peptide concentration using tryptophan fluorescence spectroscopy and peptide-induced leakage of dyes from lipid vesicles. In agreement with a previous report, we find increased membrane partitioning of L18W-PGLa in the presence of MG2a. However, this effect does not prevail to lipid concentrations higher than 1 mM, above which all peptides associate with the lipid bilayers. This implies that synergistic effects of MG2a and L18W-PGLa in previously reported experiments with lipid concentrations >1 mM are due to peptide-induced membrane remodeling and not their specific membrane partitioning.     

X-ray diffraction study of feline leukemia virus fusion peptide and lipid polymorphism

The structural effects of the fusion peptide of feline leukemia virus (FeLV) on the lipid polymorphism of N-methylated dioleoylphosphatidylethanolamine were studied using a temperature ramp with sequential X-ray diffraction. This peptide, the hydrophobic amino-terminus of p15E, has been proven to be fusogenic and to promote the formation of highly curved, intermediate structures on the lamellar liquid-crystal to inverse hexagonal phase transition pathway. The FeLV peptide produces marked effects on the thermotropic mesomorphic behaviour of MeDOPE, a phospholipid with an intermediate spontaneous radius of curvature. The peptide is shown to reduce the lamellar repeat distance of the membrane prior to the onset of an inverted cubic phase. This suggests that membrane thinning may play a role in peptide-induced membrane fusion and strengthens the link between the fusion pathway and inverted cubic phase formation. The results of this study are interpreted in relation to models of the membrane fusion mechanism.     

Sigmoidal concentration dependence of antimicrobial peptide activities: a case study on alamethicin.

The transition of the state of alamethicin from its inactive state to its active state of pore formation was measured as a function of the peptide concentration in three different membrane conditions. In each case the fraction of the alamethicin molecules occupying the active state, phi, showed a sigmoidal concentration dependence that is typical of the activities of antimicrobial peptides. Such a concentration dependence is often interpreted as due to peptide aggregation. However, we will show that a simple effect of aggregation cannot explain the data. We will introduce a model based on the elasticity of membrane, taking into consideration the membrane-thinning effect due to protein inclusion. The elastic energy of membrane provides an additional driving force for aggregation. The model produces a relation that not only predicts the correct concentration dependence but also explains qualitatively how the dependence changes with membrane conditions. The result shows that the membrane-mediated interactions between monomers and aggregates are essential for the strong cooperativity shown in pore formation.