Effect of cholesterol and charge on pore formation in bilayer vesicles by a pH-sensitive peptide.

The effect of cholesterol on the bilayer partitioning of the peptide GALA (WEAALAEALAEALAEHLAEALAEALEALAA) and its assembly into a pore in large unilamellar vesicles composed of neutral and negatively charged phospholipids has been determined. GALA undergoes a conformational change from a random coil to an amphipathic alpha-helix when the pH is reduced from 7.0 to 5.0, inducing at low pH leakage of contents from vesicles. Leakage from neutral or negatively charged vesicles at pH 5.0 was similar and could be adequately explained by the mathematical model (Parente, R. A., S. Nir, and F. C. Szoka, Jr., 1990. Mechanism of leakage of phospholipid vesicle contents induced by the peptide GALA. Biochemistry. 29:8720-8728) which assumed that GALA becomes incorporated into the vesicle bilayer and irreversibly aggregates to form a pore consisting of 10 +/- 2 peptides. Increasing cholesterol content in the membranes resulted in a reduced efficiency of the peptide to induce leakage. Part of the cholesterol effect was due to reduced binding of the peptide to cholesterol-containing membranes. An additional effect of cholesterol was to increase reversibility of surface aggregation of the peptide in the membrane. Results could be explained and predicted with a model that retains the same pore size, i.e., 10 +/- 2 peptides, but includes reversible aggregation of the monomers to form the pore. Resonance energy transfer experiments using fluorescently labeled peptides confirmed that the degree of reversibility of surface aggregation of GALA was significantly larger in cholesterol-containing liposomes, thus reducing the efficiency of pore formation.     

Interactions of peptides with liposomes: pore formation and fusion

Leakage from liposomes induced by several peptides is reviewed and a pore model is described. According to this model peptide molecules become incorporated into the vesicle bilayer and aggregate reversibly or irreversibly within the surface. When a peptide aggregate reaches a critical size, peptide translocation can occur and a pore is formed. With the peptide GALA the pores are stable and persist for at least 10 minutes. The model predicts that for a given lipid/peptide ratio, the extent of leakage should decrease as the vesicle diameter decreases, and for a given amount of peptide bound per vesicle less leakage would be observed at higher temperatures due to the increase in reversibility of surface aggregates of the peptide. Effect of membrane composition on pore formation is reviewed. When cholesterol was included in the liposomes the efficiency of inducation of leakage by the peptide GALA was reduced due to reduced binding and increased reversibility of surface aggregation of the peptide. Phospholipids which contain less ordered acyl-chains and have a slightly wedge-like shape, can better accommodate peptide surface aggregates, and consequently insertion and translocation of the peptide may be less favored. Demonstrations of antagonism between pore formation and fusion are presented. The choice of factors which promote vesicle aggregation, e.g., larger peptides, increased vesicle and peptide concentration results in enhanced vesicle fusion at the expense of formation of intravesicular pores. FTIR studies with HIV-1 fusion peptides indicate that in systems where extensive vesicle fusion occurred the beta conformation of the peptides was predominant, whereas the alpha conformation was exhibited in cases where leakage was the main outcome. Antagonism between leakage and fusion was exhibited by 1-palmitoyl-2-oleoylphosphatidylglycerol vesicles, where the order of addition of peptide (HIV(arg)) or Ca(2+)dictated whether pore formation or vesicle fusion would occur. The current study emphasizes that the addition of Ca(2+), which promotes vesicle aggregation can also reduce peptide translocation in isolated vesicles.     

Mechanisms of leakage

Abstract

Two mechanisms of leakage from liposomes are discussed, (i) Cations such as Ca²⁺ induce graded release whose rate depends mainly on vesicle collisions and is associated in the case of several acidic phospholipids with fusion events. A certain degree of leakage also occurs in between collisions. Consequently, the leakage per fusion is reduced at larger lipid and Ca concentrations, (n) Certain peptides induce leakage by pore formation, which shows selectivity to the size of the entrapped molecules and occurs by an all or none mechanism; vesicles either leak or retain all of their contents. A model for final extents and kinetics of leakage due to pore forming peptides is described. This model assumes that pore forming peptides become incorporated into the vesicle bilayer and aggregate to form a pore. Recent developments in the model enable considerations of a reversible or irreversible surface aggregation of peptides. Results of final extents and kinetics of leakage induced by pore forming peptides can be well explained and predicted by this formalism. Studies demonstrate that Ca can play a dual role in affecting leakage. A case is presented where Ca ⁺ inhibits and can even arrest pore formation by a peptide, while promoting vesicle fusion. Conversely, formation of pore structures by a peptide can inhibit vesicle fusion.     

Interaction of fluorescently labeled pardaxin and its analogues with lipid bilayers.

Fluorescence measurements were used to monitor the interaction of the neurotoxin pardaxin and its analogues with membranes. Eight peptides were selectively labeled with the fluorophore 7-nitrobenz-2-oxa-1,3-diazole-4-yl, either at their N-terminal or at their C-terminal. No detectable changes in membrane permeability or hemolytic activity were observed upon modification. Upon the titration of solutions containing the different peptides with small unilamellar vesicles, the fluorescent emission spectra of 7-nitrobenz-2-oxa-1,3-diazole-4-yl-labeled pardaxin and its analogues, but not those of control peptides, displayed blue shifts in addition to enhanced intensities upon relocation of the probe to a more apolar environment. The results revealed that the N terminus of pardaxin is buried within the lipid bilayer while the C terminus is located at the bilayer's surface. Binding isotherms were obtained from the observed increases in the fluorescence emission yields, from which surface partition constants, in the range of 10(4) M-1, were in turn derived. The existence of an aggregation process was suggested by the shape of the binding isotherms. Furthermore, the results show good correlation between the incidence of aggregation and the ability of the different analogues to induce the release of relatively large molecules from vesicles. As such, our results suggest that the mechanism of pore formation employed by pardaxin and its analogues could be described by the "barrel stave" model.     

Surface aggregation and membrane penetration by peptides: relation to pore formation and fusion

The peptide GALA undergoes a conformational change to an amphipathic alpha -helix when the pH is reduced, inducing leakage of contents from vesicles. Leakage from neutral or negativelycharged vesicles at pH 5.0 was similar and could be adequately explained by a mathematical model which assumed that GALA becomes incorporated into the vesicle bilayer and irreversibly aggregates to form a pore consisting of M =10+/-2 peptides. Increasing cholesterol content in the membranes resulted in reduced leakage, and increased reversibility of surface aggregation of the peptide. Employing fluorescently labelled peptides confirmed that the degree of reversibility of surface aggregation of GALA was significantly larger in cholesterol containing liposomes. Orientation of the peptide GALA in bilayers was determined by a bodipy-avidin/ biotin binding assay. The peptide was labelled by biotin at the N- or Cterminus and bodipy-avidin molecules were added externally or were preencapsulated in the vesicles. The peptides are arranged in the pore perpendicularly to the membrane, such that 3/4 of the N-termini are on the internal side of the membrane. The pores are stable and persist for at least 10 min. When the peptides form an aggregate of size smaller than M, the orientation of the peptide is mostly parallel to the surface and the biotinylated peptide does not translocate. When a critical size of the aggregate is attained, a rearrangement of the peptide occurs, which amounts to rapid penetration and formation of a pore structure. Induction of fusion by peptides may be antagonistic to pore formation, the outcome being dependent on vesicle aggregation.     

A synthetic peptide corresponding to the hydrophobic amino terminal region of pardaxin can perturb model membranes of phosphatidyl choline and serine

Peptides corresponding to the amino terminal region of pardaxin from Pardachirus pavoninus (Gly-Phe-Phe-Ala-Leu-Ile-Pro-Lys-Ile-Ile-Ser-Ser-Pro-Leu-Phe) have been synthesized and their interaction with model membranes of phosphatidyl choline and serine studied by 90 degrees C light scattering and fluorescence spectroscopy. The amino terminal 8-residue peptide and the protected 15-residue peptide cause only aggregation of lipid vesicles. The deprotected 15-residue peptide has the ability to cause aggregation and release of entrapped carboxyfluorescein with both phosphatidyl choline and serine lipid vesicles, like pardaxin. The membrane-perturbing ability of the amino terminal 15-residue peptide can be attributed to its ability to adopt an alpha-helical conformation which is amphiphilic in nature in a hydrophobic environment.     

Insertion and pore formation driven by adsorption of proteins onto lipid bilayer membrane-water interfaces

We describe the binding of proteins to lipid bilayers in the case for which binding can occur either by adsorption to the lipid bilayer membrane-water interface or by direct insertion into the bilayer itself. We examine in particular the case when the insertion and pore formation are driven by the adsorption process using scaled particle theory. The adsorbed proteins form a two-dimensional "surface gas" at the lipid bilayer membrane-water interface that exerts a lateral pressure on the lipid bilayer membrane. Under conditions of strong intrinsic binding and a high degree of interfacial converge, this pressure can become high enough to overcome the energy barrier for protein insertion. Under these conditions, a subtle equilibrium exists between the adsorbed and inserted proteins. We propose that this provides a control mechanism for reversible insertion and pore formation of proteins such as melittin and magainin. Next, we discuss experimental data for the binding isotherms of cytochrome c to charged lipid membranes in the light of our theory and predict that cytochrome c inserts into charged lipid bilayers at low ionic strength. This prediction is supported by titration calorimetry results that are reported here. We were furthermore able to describe the observed binding isotherms of the pore-forming peptides endotoxin (alpha 5-helix) and of pardaxin to zwitterionic vesicles from our theory by assuming adsorption/insertion equilibrium.     

Fluctuations and the Rate-Limiting Step of Peptide-Induced Membrane Leakage

Peptide-induced vesicle leakage is a common experimental test for the membrane-perturbing activity of antimicrobial peptides. The leakage kinetics is usually very slow, requiring minutes to hours for complete release of vesicle contents, and exhibits a biphasic behavior. We report here that, in the case of the peptaibol trichogin GA IV, all processes involved in peptide-membrane interaction, such as peptide-membrane association, peptide aggregation, and peptide translocation, take place on a timescale much shorter than the leakage kinetics. On the basis of these findings, we propose a stochastic model in which the leakage kinetics is determined by the discrete nature of a vesicle suspension: peptides are continuously exchanging among vesicles, producing significant fluctuations over time in the number of peptide molecules bound to each vesicle, and in the formation of pores. According to this model, the fast initial leakage is caused by vesicles that contain at least one pore after the peptides are randomly distributed among the liposomes, whereas the slower release is associated with the time needed to occasionally reach in an intact vesicle the critical number of bound peptides necessary for pore formation. Fluctuations due to peptide exchange among vesicles therefore represent the rate-limiting step of such a slow mechanism.     

Revisiting peptide amphiphilicity for membrane pore formation

It has previously been shown that an amphipathic de novo designed peptide made of 10 leucines and four phenylalanines substituted with crown ethers induces vesicle leakage without selectivity. To gain selectivity against negatively charged dimyristoylphosphatidylglycerol (DMPG) bilayers, one or two leucines of the peptide were substituted with positively charged residues at each position. All peptides induce significant calcein leakage of DMPG vesicles. However, some peptides do not induce significant leakage of zwitterionic dimyristoylphosphatidylcholine vesicles and are thus active against only bacterial model membranes. The intravesicular leakage is induced by pore formation instead of membrane micellization. Nonselective peptides are mostly helical, while selective peptides mainly adopt an intermolecular β-sheet structure. This study therefore demonstrates that the position of the lysine residues significantly influences the secondary structure and bilayer selectivity of an amphipathic 14-mer peptide, with β-sheet peptides being more selective than helical peptides.     

Channel formation properties of synthetic pardaxin and analogues.

Six analogues of teh 33-residue shark repellent neurotoxin pardaxin were synthesized by the solid phase method: [Ala13]pardaxin, [Gly14,Gly15]pardaxin, des[1----9]pardaxin, [N1-succinamido]pardaxin, C33-dihydroxyethylamido]pardaxin, and C33-[diaminoethylamido]pardaxin. The spectroscopic and functional characterizations of the analogues are described. The peptides were characterized spectroscopically by circular dichroism (CD) before and after binding to soybean vesicles. They were characterized functionally by measuring their potential to evoke the dissipation of diffusion potential and calcein release from sonicated unilamellar soybean liposomes, by determining their ability to create single channels in planar bilayers, and by measuring their cytolytic activity on human erythrocytes. The behavior of the analogues modified at the C terminus is similar to that of pardaxin. [N'-succinamido]Pardaxin, however, reveals an increase in alpha-helicity both alone and in the presence of liposomes. It has the same potency as pardaxin to dissipate diffusion potential, to evoke calcein release and to produce single channels in lipid bilayers, but at a slower rate than that of pardaxin. It has more than 70-fold less cytolytic activity than pardaxin. [Ala13] Pardaxin has twice the alpha-helical content than pardaxin, both alone and in the presence of vesicles, yet it has less effect on the diffusion potential and calcein release, and it does not have cytolytic activity on human erythrocytes. Both [Gly14,Gly15]pardaxin and des[1----9]pardaxin are much less potent than pardaxin in all effects. However des[1----9]pardaxin exhibits a slight change in alpha-helicity upon binding to vesicles, whereas [Gly14,Gly15]pardaxin does not. The results support a model in which pardaxin is composed of two putative alpha-helices separated by proline. The N-terminal alpha-helix is important for the insertion of the peptide to the lipid bilayer, and the C-terminal amphiphilic alpha-helix is the ion channel lining segment of pardaxin.     

Action mechanism of tachyplesin I and effects of PEGylation

PEGylation of protein and peptide drugs is frequently used to improve in vivo efficacy. We investigated the action mechanism of tachyplesin I, a membrane-acting cyclic antimicrobial peptide from Tachypleus tridentatus and the effects of PEGylation on the mechanism. The PEGylated peptide induced the leakage of calcein from egg yolk l-α-phosphatidylglycerol/egg yolk l-α-phosphatidylcholine large unilamellar vesicles similarly to the parent peptide. Both peptides induced lipid flip-flop coupled to leakage and was translocated into the inner leaflet of the bilayer, indicating that tachyplesin I forms a toroidal pore and that PEGylation did not alter the basic mechanism of membrane permeabilization of the parent peptide. Despite their similar activities against model membranes, the peptides showed very different biological activities. The cytotoxicity of tachyplesin I was greatly reduced by PEGylation, although the antimicrobial activity was significantly weakened. We investigated the enhancement of the permeability of inner membranes induced by the peptides. Our results suggested that outer membranes and peptidoglycan layers play an inhibitory role in the permeation of the PEG moiety. Furthermore, a reduction in DNA binding by PEGylation may also contribute to the weak activity of the PEGylated peptide.     

Pardaxin induces aggregation but not fusion of phosphatidylserine vesicles

The effects on membranes of pardaxin, an amphipathic polypeptide, purified from the gland secretion of the Red Sea Moses sole flatfish Pardachirus marmoratus are dose-dependent and range from formation of voltage-gated, cation-selective pores to lysis. We have now investigated the interactions of pardaxin with small unilamellar liposomes. Light scattering showed that pardaxin (10⁻⁷–10⁻⁹M) mediated the aggregation of liposomes composed of phosphatidylserine but not of phosphatidylcholine. Aggregation of phosphatidylserine vesicles was impaired by vesicle depolarization. Furthermore, pardaxin-mediated aggregation between fluorescent-labeled PS vesicles was accompanied by leakage of the vesicle contents, and not by fusogenic process within the aggregates. We suggest that pardaxin is a unique polypeptide, that induces vesicle aggregation and membrane destabilization, but not membrane fusion; the mechanism of the aggregation activity of pardaxin is related to its amphipathic properties.     

Amyloid Aggregation on Lipid Bilayers and Its Impact on Membrane Permeability

Fibrillar protein aggregates (amyloids) are involved in several common pathologies, e.g., Alzheimer's disease and type II diabetes. Accumulating evidence suggests that toxicity in amyloid-related diseases originates from the deposition of protein aggregates on the cell membrane, which results in bilayer disruption and cell leakage. The molecular mechanism of damage to the membrane, however, is still obscure. To shed light on it we have performed coarse-grained molecular dynamics simulations of fibril-forming amphipathic peptides in the presence of lipid vesicles. The simulation results show that highly amyloidogenic peptides fibrillate on the surface of the vesicle, damaging the bilayer and promoting leakage. In contrast, the ordered aggregation of peptides with low amyloidogenicity is hindered by the vesicles. Remarkably, leakage from the vesicle is caused by growing aggregates, but not mature fibrils. The simulation results provide a basis for understanding the range of aggregation behavior that is observed in experiments with fibril-forming (poly)peptides.     

Effect of magainin, class L, and class A amphipathic peptides on fatty acid spin labels in lipid bilayers

Magainins and other antimicrobial peptides increase ion flux across the membrane. They may do this by forming some type of pore or by perturbing lipid organization due to peptide lying on the bilayer surface. In order to determine if magainins perturb the lipid sufficiently to permeabilize the bilayer, their effect on the motion of fatty acid and lipid spin labels in phosphatidylcholine/phosphatidylglycerol (PC/PG) lipid vesicles was determined. Their effect was compared to two synthetic peptides, 18L and Ac-18A-NH(2), designed to mimic the naturally occurring classes of lytic (class L) and apolipoprotein (class A) amphipathic helices, respectively. We show that although magainins and 18L both had significant effects on lipid chain order, much greater than Ac-18A-NH(2), there was no correlation between these effects and the relative ability of these three peptide classes to permeabilize PC/PG vesicles in the order magainins=Ac-18A-NH(2) > 18L. This suggests that the perturbing effects of magainins on lipid chain order at permeabilizing concentrations are not directly responsible for the increased leakage of vesicle contents. The greater ability of the magainins to permeabilize PC/PG vesicles relative to 18L is thus more likely due to formation of some type of pore by magainins. The greater ability of Ac-18A-NH(2) relative to 18L to permeabilize PC/PG vesicles despite its lack of disordering effect must be due to its ability to cause membrane fragmentation. Effects of these peptides on other lipids indicated that the mechanism by which they permeabilize lipid bilayers depends both on the peptide and on the lipid composition of the vesicles.     

Beta-sheet pore-forming peptides selected from a rational combinatorial library: mechanism of pore formation in lipid vesicles and activity in biological membranes

In a previous report we described the selection of potent, beta-sheet pore-forming peptides from a combinatorial library designed to mimic membrane-spanning beta-hairpins (Rausch, J. M., Marks, J. R., and Wimley, W. C. (2005) Proc. Natl. Acad. Sci. U.S.A. 102, 10511-10515). Here, we characterize their mechanism of action and compare the structure-function relationships in lipid vesicles to their activity in biological membranes. The pore-forming peptides bind to membrane interfaces and self-assemble into beta-sheets that cause a transient burst of graded leakage across the bilayers. Despite the continued presence of the structured peptides in the bilayer, at most peptide concentrations leakage is incomplete and ceases quickly after peptide addition with a deactivation half-time of several minutes. Molecules up to 3,000 Da escape from the transient pores, but much larger molecules do not. Fluorescence spectroscopy and quenching showed that the peptides reside mainly on the bilayer surface and are partially exposed to water, rather than in a membrane-spanning state. The "carpet" or "sinking raft" model of peptide pore formation offers a viable explanation for our observations and suggests that the selected pore-formers function with a mechanism that is similar to the natural pore-forming antimicrobial peptides. We therefore also characterized the antimicrobial and cytotoxic activity of these peptides. All peptides studied, including non-pore-formers, had sterilizing antimicrobial activity against at least some microbes, and most have low activity against mammalian cell membranes. Thus, the structure-function relationships that were apparent in the vesicle systems are similar to, but do not correlate completely with, the activity of the same peptides in biological membranes. However, of the peptides tested, only the pore-formers selected in the high-throughput screen have potent, broad-spectrum sterilizing activity against Gram-positive and Gram-negative bacteria as well as against fungi, while having only small lytic effects on human cells.     

Action mechanism of tachyplesin I and effects of PEGylation

PEGylation of protein and peptide drugs is frequently used to improve in vivo efficacy. We investigated the action mechanism of tachyplesin I, a membrane-acting cyclic antimicrobial peptide from Tachypleus tridentatus and the effects of PEGylation on the mechanism. The PEGylated peptide induced the leakage of calcein from egg yolk L-alpha-phosphatidylglycerol/egg yolk L-alpha-phosphatidylcholine large unilamellar vesicles similarly to the parent peptide. Both peptides induced lipid flip-flop coupled to leakage and was translocated into the inner leaflet of the bilayer, indicating that tachyplesin I forms a toroidal pore and that PEGylation did not alter the basic mechanism of membrane permeabilization of the parent peptide. Despite their similar activities against model membranes, the peptides showed very different biological activities. The cytotoxicity of tachyplesin I was greatly reduced by PEGylation, although the antimicrobial activity was significantly weakened. We investigated the enhancement of the permeability of inner membranes induced by the peptides. Our results suggested that outer membranes and peptidoglycan layers play an inhibitory role in the permeation of the PEG moiety. Furthermore, a reduction in DNA binding by PEGylation may also contribute to the weak activity of the PEGylated peptide.     

Effect of phospholipid composition on an amphipathic peptide-mediated pore formation in bilayer vesicles

To better understand the influence of phospholipid acyl-chain composition on the formation of pores by cytotoxic amphipathic helices in biological membranes, the leakage of aqueous contents induced by the synthetic peptide GALA (WEAALAEALAE ALAEHLAEALAEALEALAA) from large unilamellar phospholipid vesicles of various compositions has been studied. Peptide-mediated leakage was examined at pH 5.0 from vesicles made of phosphatidylcholine (PC) and phosphatidylglycerol (PG) with the following acyl-chain compositions: 1-palmitoyl-2-oleoyl (PO), 1,2-dioleoyl (DO), 1, 2-dielaidoyl (DE), and 1,2-dipetroselinoyl (DPe). A mathematical model predicts and simulates the final extents of GALA-mediated leakage of 1-aminonaphthalene-3,6,8-trisulfonic acid (ANTS) and p-xylene-bis-pyridinium bromide (DPX) from 1-palmitoyl-2-oleoyl-phosphatidylcholine/1-palmitoyl-2-oleoyl-phospha tidylglycerol (POPC/POPG) and 1, 2-dielaidoyl-sn-glycero-3-phosphocholine/1, 2-dielaidoyl-phosphatidylglycerol (DEPC/DEPG) liposomes at pH 5.0 as a function of peptide concentration in the bilayer, by considering that GALA pores responsible for this leakage have a minimum size of 10 +/- 2 monomers and are formed by quasiirreversible aggregation of the peptide. With the phospholipid acyl-chain compositions tested, GALA-induced ANTS/DPX leakage follows the rank order POPC/POPG approximately DEPC/DEPG > DPePC/DPePG > DOPC/DOPG. Results from binding experiments reveal that this reduced leakage from DOPC/DOPG vesicles cannot be explained by a reduced binding affinity of the peptide to these membranes. As shown by monitoring the leakage of a fluorescent dextran, an increase in the minimum pore size also does not explain the reduction in ANTS/DPX leakage. The data suggest that surface-associated GALA monomers or aggregates are stabilized in bilayers composed of phospholipids containing a cis unsaturation per acyl chain (DO and DPe), while transbilayer peptide insertion is reduced. GALA-induced ANTS/DPX leakage is also decreased when the vesicles contain phosphatidylethanolamine (PE). This lends further support to the suggestion that factors stabilizing the surface state of the peptide reduce its insertion and subsequent pore formation in the bilayer.     

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.