Pore-forming toxins: experiments with S. aureus alpha-toxin, C. perfringens theta-toxin and E. coli haemolysin in lipid bilayers, liposomes and intact cells

Zinc Deprivation as a Promising Approach for Combating Methicillin-Resistant Staphylococcus aureus: A Pilot Study

Methicillin-resistant Staphylococcus aureus (MRSA) infections are a global health burden with an urgent need for antimicrobial agents. Studies have shown that host immune responses limit essential metals such as zinc during infection, leading to the limitation of bacterial virulence. Thus, the deprivation of zinc as an important co-factor for the activity of many S. aureus enzymes can be a potential antimicrobial approach. However, the effect of zinc deprivation on S. aureus and MRSA is not fully understood. Therefore, the current study aimed to dissect the effects of zinc deprivation on S. aureus hemolytic activity and biofilm formation through employing biochemical and genetic approaches to study the effect of zinc deprivation on S. aureus growth and virulence. Chemically defined media (CDM) with and without ZnCl2, was used to assess the effect of zinc deprivation on growth, biofilm formation, and hemolytic activity in methicillin-susceptible S. aureus (MSSA) RN6390 and MRSA N315 strains. Zinc deprivation decreased the growth of RN6390 and N315 S. aureus strains significantly by 1.5-2 folds, respectively compared to the zinc physiological range encountered by the bacteria in the human body (7-20 µM) (p < 0.05). Zinc deprivation significantly reduced biofilm formation by 1.5 folds compared to physiological levels (p < 0.05). Moreover, the hemolytic activity of RN6390 and N315 S. aureus strains was significantly decreased by 20 and 30 percent, respectively compared to physiological zinc levels (p < 0.05). Expression of biofilm-associated transcripts levels at late stage of biofilm formation (20 h) murein hydrolase activator A (cidA) and cidB were downregulated by 3 and 5 folds, respectively (p < 0.05) suggested an effect on extracellular DNA production. Expression of hemolysins-associated genes (hld, hlb, hla) was downregulated by 3, 5, and 10 folds, respectively, in absence of zinc (p < 0.001). Collectively the current study showed that zinc deprivation in vitro affected growth, biofilm formation, and hemolytic activity of S. aureus. Our in vitro findings suggested that zinc deprivation can be a potential supportive anti-biofilm formation and antihemolytic approach to contain MRSA topical infections.

Calcium-Mediated Liposome Fusion to Engineer Giant Lipid Vesicles with Cytosolic Proteins and Reconstituted Mammalian Proteins

Giant unilamellar lipid vesicles (GUVs) are widely used as model membrane systems and provide an excellent basis to construct artificial cells. To construct more sophisticated artificial cells, proteins-in particular membrane proteins-need to be incorporated in GUVs. However, current methods for protein reconstitution have limited throughput or are not generally applicable for all proteins because they depend on detergent solubilization. This limitation is addressed here by introducing calcium-mediated membrane fusion to transfer proteins between negatively charged GUVs and cell-derived plasma membrane vesicles (CDVs), derived from HEK293T cells overexpressing a membrane receptor protein. Fusion conditions are optimized using large unilamellar vesicles and GUVs containing phosphatidylserines and fusogenic lipids. The approach is then applied to induce lipid mixing and subsequent transfer of the overexpressed membrane receptor from CDVs into GUVs. The membrane receptor is detected by immunofluorescence on GUVs that underwent lipid mixing with CDVs. Those GUVs also exhibit esterase activity because cytosolic esterases entrapped in the CDVs are transferred during membrane fusion. Thus, content mixing is demonstrated. Using CDVs circumvents the need to purify or solubilize proteins. Moreover, calcium-mediated fusion allows transfer of lipids, water-soluble and membrane bound proteins in one step, resulting in a semi-synthetic cell.

Single-molecule conformational dynamics of viroporin ion channels regulated by lipid-protein interactions

Classic swine fever is a highly contagious and often fatal viral disease that is caused by the classical swine fever virus (CSFV). Protein p7 of CFSV is a prototype of viroporin, a family of small, highly hydrophobic proteins postulated to modulate virus-host interactions during the processes of virus entry, replication and assembly. It has been shown that CSFV p7 displays substantial ion channel activity when incorporated into membrane systems, but a deep rationalization of the size and dynamics of the induced pores is yet to emerge. Here, we use high-resolution conductance measurements and current fluctuation analysis to demonstrate that CSFV p7 channels are ruled by equilibrium conformational dynamics involving protein-lipid interactions. Atomic force microscopy (AFM) confirms the existence of a variety of pore sizes and their tight regulation by solution pH. We conclude that p7 viroporin forms subnanometric channels involved in virus propagation, but also much larger pores (1-10 nm in diameter) with potentially significant roles in virus pathogenicity. Our findings provide new insights into the sources of noise in protein electrochemistry and demonstrate the existence of slow complex dynamics characteristic of crowded systems like biomembrane surfaces.

Measuring kinetic drivers of pneumolysin pore structure

Most membrane attack complex-perforin/cholesterol-dependent cytolysin (MACPF/CDC) proteins are thought to form pores in target membranes by assembling into pre-pore oligomers before undergoing a pre-pore to pore transition. Assembly during pore formation is into both full rings of subunits and incomplete rings (arcs). The balance between arcs and full rings is determined by a mechanism dependent on protein concentration in which arc pores arise due to kinetic trapping of the pre-pore forms by the depletion of free protein subunits during oligomerization. Here we describe the use of a kinetic assay to study pore formation in red blood cells by the MACPF/CDC pneumolysin from Streptococcus pneumoniae. We show that cell lysis displays two kinds of dependence on protein concentration. At lower concentrations, it is dependent on the pre-pore to pore transition of arc oligomers, which we show to be a cooperative process. At higher concentrations, it is dependent on the amount of pneumolysin bound to the membrane and reflects the affinity of the protein for its receptor, cholesterol. A lag occurs before cell lysis begins; this is dependent on oligomerization of pneumolysin. Kinetic dissection of cell lysis by pneumolysin demonstrates the capacity of MACPF/CDCs to generate pore-forming oligomeric structures of variable size with, most likely, different functional roles in biology.

Protein-lipid interactions and non-lamellar lipidic structures in membrane pore formation and membrane fusion

Pore-forming proteins and peptides act on their targeted lipid bilayer membranes to increase permeability. This approach to the modulation of biological function is relevant to a great number of living processes, including; infection, parasitism, immunity, apoptosis, development and neurodegeneration. While some pore-forming proteins/peptides assemble into rings of subunits to generate discrete, well-defined pore-forming structures, an increasing number is recognised to form pores via mechanisms which co-opt membrane lipids themselves. Among these, membrane attack complex-perforin/cholesterol-dependent cytolysin (MACPF/CDC) family proteins, Bax/colicin family proteins and actinoporins are especially prominent and among the mechanisms believed to apply are the formation of non-lamellar (semi-toroidal or toroidal) lipidic structures. In this review I focus on the ways in which lipids contribute to pore formation and contrast this with the ways in which lipids are co-opted also in membrane fusion and fission events. A variety of mechanisms for pore formation that involve lipids exists, but they consistently result in stable hybrid proteolipidic structures. These structures are stabilised by mechanisms in which pore-forming proteins modify the innate capacity of lipid membranes to respond to their environment, changing shape and/or phase and binding individual lipid molecules directly. In contrast, and despite the diversity in fusion protein types, mechanisms for membrane fusion are rather similar to each other, mapping out a pathway from pairs of separated compartments to fully confluent fused membranes. Fusion proteins generate metastable structures along the way which, like long-lived proteolipidic pore-forming complexes, rely on the basic physical properties of lipid bilayers. Membrane fission involves similar intermediates, in the reverse order. I conclude by considering the possibility that at least some pore-forming and fusion proteins are evolutionarily related homologues. This article is part of a Special Issue entitled: Pore-Forming Toxins edited by Mauro Dalla Serra and Franco Gambale.

Giant MACPF/CDC pore forming toxins: A class of their own

Pore Forming Toxins (PFTs) represent a key mechanism for permitting the passage of proteins and small molecules across the lipid membrane. These proteins are typically produced as soluble monomers that self-assemble into ring-like oligomeric structures on the membrane surface. Following such assembly PFTs undergo a remarkable conformational change to insert into the lipid membrane. While many different protein families have independently evolved such ability, members of the Membrane Attack Complex PerForin/Cholesterol Dependent Cytolysin (MACPF/CDC) superfamily form distinctive giant β-barrel pores comprised of up to 50 monomers and up to 300Å in diameter. In this review we focus on recent advances in understanding the structure of these giant MACPF/CDC pores as well as the underlying molecular mechanisms leading to their formation. Commonalities and evolved variations of the pore forming mechanism across the superfamily are discussed. This article is part of a Special Issue entitled: Pore-Forming Toxins edited by Mauro Dalla Serra and Franco Gambale.

Suppression of cell membrane permeability by suramin: involvement of its inhibitory actions on connexin 43 hemichannels

BACKGROUND AND PURPOSE

Suramin is a clinically prescribed drug for treatment of human African trypanosomiasis, cancer and infection. It is also a well-known pharmacological antagonist of P2 purinoceptors. Despite its clinical use and use in research, the biological actions of this molecule are still incompletely understood. Here, we investigated the effects of suramin on membrane channels, as exemplified by its actions on non-junctional connexin43 (Cx43) hemichannels, pore-forming α-haemolysin and channels involved in ATP release under hypotonic conditions.

EXPERIMENTAL APPROACH

Hemichannels were activated by removing extracellular Ca(2+) . The influences of suramin on hemichannel activities were evaluated by its effects on influx of fluorescent dyes and efflux of ATP. The membrane permeability and integrity were assessed through cellular retention of preloaded calcein and LDH release.

KEY RESULTS

Suramin blocked Cx43 hemichannel permeability induced by removal of extracellular Ca(2+) without much effect on Cx43 expression and gap junctional intercellular communication. This action of suramin was mimicked by its analogue NF023 and NF449 but not by another P2 purinoceptor antagonist PPADS. Besides hemichannels, suramin also significantly blocked intracellular and extracellular exchanges of small molecules caused by α-haemolysin from Staphylococcus aureus and by exposure of cells to hypotonic solution. Furthermore, it prevented α-haemolysin- and hypotonic stress-elicited cell injury.

CONCLUSION AND IMPLICATIONS

Suramin blocked membrane channels and protected cells against toxin- and hypotonic stress-elicited injury. Our finding provides novel mechanistic insights into the pharmacological actions of suramin. Suramin might be therapeutically exploited to protect membrane integrity under certain pathological situations.

Incomplete pneumolysin oligomers form membrane pores

Pneumolysin is a member of the cholesterol-dependent cytolysin (CDC) family of pore-forming proteins that are produced as water-soluble monomers or dimers, bind to target membranes and oligomerize into large ring-shaped assemblies comprising approximately 40 subunits and approximately 30 nm across. This pre-pore assembly then refolds to punch a large hole in the lipid bilayer. However, in addition to forming large pores, pneumolysin and other CDCs form smaller lesions characterized by low electrical conductance. Owing to the observation of arc-like (rather than full-ring) oligomers by electron microscopy, it has been hypothesized that smaller oligomers explain smaller functional pores. To investigate whether this is the case, we performed cryo-electron tomography of pneumolysin oligomers on model lipid membranes. We then used sub-tomogram classification and averaging to determine representative membrane-bound low-resolution structures and identified pre-pores versus pores by the presence of membrane within the oligomeric curve. We found pre-pore and pore forms of both complete (ring) and incomplete (arc) oligomers and conclude that arc-shaped oligomeric assemblies of pneumolysin can form pores. As the CDCs are evolutionarily related to the membrane attack complex/perforin family of proteins, which also form variably sized pores, our findings are of relevance to that class of proteins as well.

Structural features of cholesterol dependent cytolysins and comparison to other MACPF-domain containing proteins

Five different cholesterol-dependent cytolysins (CDCs) have now had their atomic structures solved. Here their structures are compared and shown to vary less in the C-terminal region than they do in their N-terminal MACPF/CDC homology region. The most variable region of the C-terminal domain is the undecapeptide, which is observed in two clusters of conformations, and comparison of this domain with the C2 domain of perforin shows that the two structures have a common ancestor. Structural studies of CDC pre-pore and pore oligomers by cryo-electron microscopy and atomic force microscopy have revealed much about their mechanism of action. Understanding the activity of CDCs has required a combination of structural, biophysical and functional assays but current models of pore formation still require development to account for variable functional pore size.

Mechanisms of cytolysin-induced cell damage -- a role for auto- and paracrine signalling

Cytolysins inflict cell damage by forming pores in the plasma membrane. The Na(+) conductivity of these pores results in an ion influx that exceeds the capacity of the Na(+) /K(+) -pump to extrude Na(+) . This net load of intracellular osmolytes results in swelling and eventual lysis of the attacked cell. Many nucleated cells have the capacity to reduce the potential damage of pore-forming proteins, whereas erythrocytes have been regarded as essentially defenceless against cytolysin-induced cell damage. This review addresses how autocrine/paracrine signalling and the cells intrinsic volume regulation markedly influence the fate of the cell after membrane insertion of cytolysins. Moreover, it regards the various steps that may explain the relative large degree of diversity between cell types and species as well as highlights some of the current gaps in the mechanistic understanding of cytolysin-induced cell injury.

Effects of MACPF/CDC proteins on lipid membranes

Recent work on the MACPF/CDC superfamily of pore-forming proteins has focused on the structural analysis of monomers and pore-forming oligomeric complexes. We set the family of proteins in context and highlight aspects of their function which the direct and exclusive equation of oligomers with pores fails to explain. Starting with a description of the distribution of MACPF/CDC proteins across the domains of life, we proceed to show how their evolutionary relationships can be understood on the basis of their structural homology and re-evaluate models for pore formation by perforin, in particular. We furthermore highlight data showing the role of incomplete oligomeric rings (arcs) in pore formation and how this can explain small pores generated by oligomers of proteins belonging to the family. We set this in the context of cell biological and biophysical data on the proteins' function and discuss how this helps in the development of an understanding of how they act in processes such as apicomplexan parasites gliding through cells and exiting from cells.

5- and 4'-Hydroxylated flavonoids affect voltage gating of single alpha-hemolysin pore

Molecular mechanisms of the influence of flavonoids on the voltage gating of a single alpha-hemolysin channel in planar lipid membranes are studied. It is shown that the addition of flavonoids hydroxylated in position 5 of the A-ring and in position 4' of the B-ring into bilayer bathing solution shifts the voltage dependence of channel switching from high- to low-conductance states to voltages nearer zero. It is concluded that the effect is likely to be attributed to a specific interaction of at least three flavonoid molecules with the voltage sensor of an alpha-hemolysin pore. Possible flavonoid binding sites and identification of amino acid residues included into the voltage sensor domain of the alpha-hemolysin channel are discussed.

Biological relevance of natural alpha-toxin fragments from Staphylococcus aureus

Serine proteases represent an essential part of cellular homeostasis by generating biologically active peptides. In bacteria, proteolysis serves two different roles: a major housekeeping function and the destruction of foreign or target cell proteins, thereby promoting bacterial invasion. In the process, other virulence factors such as exotoxins become affected. In Staphylococcus aureus culture supernatant, the pore-forming alpha-toxin is cleaved by the coexpressed V8 protease and aureolysin. The oligomerizing and pore-forming abilities of five such spontaneously occurring N- and C-terminal alpha-toxin fragments were studied. (3)H-marked alpha-toxin fragments bound to rabbit erythrocyte membranes but only fragments with intact C termini, missing 8, 12 and 71 amino acids from their N-terminal, formed stable oligomers. All isolated fragments induced intoxication of mouse adrenocortical Y1 cells in vitro, though the nature of membrane damage for a fragment, degraded at its C terminus, remained obscure. Only one fragment, missing the first eight N-terminal amino acids, induced irreversible intoxication of Y1 cells in the same manner as the intact toxin. Four of the isolated fragments caused swelling, indicating altered channel formation. Fragments missing 12 and 71 amino acids from the N terminus occupied the same binding sites on Y1 cell membranes, though they inhibited membrane damage caused by intact toxin. In conclusion, N-terminal deletions up to 71 amino acids are tolerated, though the kinetics of channel formation and the channel's properties are altered. In contrast, digestion at the C terminus results in nonfunctional species.

Cholesterol-dependent cytolysins

The cholesterol-dependent cytolysins (CDCs) are part of a large family of pore-forming proteins that include the human proteins perforin and the complement membrane attack complex. The activity of all family members is focused on membranes, but the proteins are themselves involved in a diverse range of phenomena. An overview of some of these phenomena is provided here, along with an historical perspective of CDCs themselves and how our understanding of their mechanism of action has developed over the years. The way in which pore formation depends on specific characteristics of the membrane under attack as well as of the protein doing the attacking is emphasised. The cholesterol-dependent cytolysins (CDCs) have been the focus of a renewed keen research interest for over ten years now. Their importance has been even further enhanced by the homology now identified between them and the membrane attack complex/perforin (MACPF) family of proteins, which includes several components of the complement cascade as well as perforin itself. In this chapter I aim to provide an overview of our understanding of the interaction between CDCs and other members of what is now called the MACPF/CDC superfamily, with their target membranes. CDCs (also in the past known as thiol-activated toxins or cholesterol-binding toxins) were originally identified from four Gram-positive bacterial genera (Clostridium, Listeria, Bacillus and Streptococcus). Well-known examples include listeriolysin, perfringolysin, streptolysin and pneumoysin. Listeriolysin from L. monocytogenes is responsible for the escape of bacteria from the phagosome to colonise the cytoplasm and has been applied as a protein adjuvant in the development of vaccines against cancer and tuberculosis, for example. Perfringolysin from C. perfringens (Fig. 1A) has become perhaps the most studied CDC4 and has an important role in pathology associated with infection (gangrene). Streptolysin from S. pyogenes is another intensely studied CDC and has been applied widely in experimental permeabilisation of biological membranes. Pneumolysin is a major virulence determinant for S. pneumoniae, allowing bacterial invasion of tissues and mediating inflammation and the activation of the complement cascade. However, CDCs have now, for example, been identified in the bacteria Arcanobacterium pyogenes and Gardnerella vaginalis and there also appear to be homologues outside prokaryotes such as the sea anemone Metridium senile pore-forming toxin metridiolysin. The homology with the MACPF family was unknown until the first structures of the canonical fold of that family were solved, revealing the now characteristic MACPF/CDC fold of a twisted 3-sheet around which helices are clustered (Fig. 1A and D). Without any significant other sequence homology, the fold of this superfamily of pore-forming and membrane-binding proteins has been conserved by compensatory mutation around a handful of key conserved glycines. The glycines presumably act as critical hinges during the dramatic refolding that CDCs are known to undergo and which is presumably the selective advantage of this specific structure that has caused it to be conserved over such a vast evolutionary timescale. While not all MACPF domains are involved in pore formation-for example, C6 and C8beta--they are all apparently involved in action on membranes. The dramatic refolding undergone by CDCs is tightly coupled to their oligomerisation and results in the conversion of the helices hemming the core 3-sheet of the MACPF/CDC domain into a pair of beta-hairpins which in tandem and alongside those from other subunits within the oligomer insert into the membrane to create a pore (Fig. 1A-C). It is obviously the basic assumption that where nonCDC members of the superfamily-such as complement proteins and perforin-act on membranes they do so by a mechanism involving similar refolding.58 Even where a member of the MACPF/CDC superfamily is not known to form a pore, or has been shown not to-at least alone-the same conformational change could have other adaptive functions during activity on or at membranes. However, the bicomponent nature of some pore-forming toxins alerts us that showing an absence of activity for one pure protein does not mean that they do not contribute to pore formation quite directly, since that may require the presence of another MACPF/CDC family member or members from the same specific system. Complement acts by a combination of the C5b-8 complex of proteins preassembled on a target membrane recruiting C9 to form a lesion, which may be a complete ring of C9 associated with the C5b-8 or an arc-electron microscopy images show both possibilities.Perforin acts in concert with granzymes, to trigger apoptosis when delivered by cytotoxic cells at their targets (damaged, transformed and infected host cells). Incomplete rings are visible for perforin also and there are many unresolved issues concerning its mechanism and the dependence ofgranzymes on it for their delivery.

Inactivation and Activity of Cholesterol-Dependent Cytolysins: What Structural Studies Tell Us

The homologous bacterially expressed cholesterol-dependent cytolysins (CDCs) form pores via oligomerization; this must occur preferentially once the target membrane has been engaged. Conformational changes in CDCs then drive partition from an aqueous environment to a lipidic one. This review addresses how premature oligomerization is prevented, how conformational changes are triggered, and how cooperativity between subunits brings about new functionality absent from isolated protomers. Variations are found in the answers provided by the CDCs to these issues. Some toxins use pH as a trigger of activity, but recent results have shown that dimerization in solution is an alternative way of preventing premature oligomerization, in particular for the CDC from Clostridium perfringens, perfringolysin. More controversially, there is still no resolution to the debate as to whether incomplete (arciform) oligomers form pores: recent results again suggest that they do.

Purification and analysis of a phospholipase A2-like lytic factor of Trichomonas vaginalis

Trichomonas vaginalis produces soluble factors that have been reported to have the ability to damage target cells in vitro, and it has been hypothesized that these factors may play a role in the pathogenesis of human trichomoniasis. A lytic factor (LF) was purified from T. vaginalis, and the molecular characteristics of LF were determined. T. vaginalis extract was subjected to hydrophobic chromatography with a 10 to 60% N-propanol gradient in 0.1 M ammonium acetate, resulting in the elution of LF from the column at 30% N-propanol. Cytotoxicity assays revealed that LF was cytotoxic to WEHI 164 cells and bovine red blood cells, and inactivation of LF by treatment with trypsin suggested that the active component of LF was a protein. Size exclusion chromatography of LF produced two fractions at 144 and 168 kDa, and analysis by sodium dodecyl sulfate-polyacrylamide gel electrophoresis of LF under reducing conditions revealed two subunits of 57 and 60 kDa. Results of a fluorescence assay of LF on carboxyfluorescein-labeled liposomes composed of phosphatidylcholine-cholesterol showed that liposomes were hydrolyzed, suggesting that LF had phospholipase activity. Thin-layer chromatography analysis of BODIPY (4,4-difluoro-3a,4adiaza-s-indacene)-labeled phosphatidylcholine treated with LF demonstrated products that migrated identically to the products produced by treatment with phospholipase A(2) (PLA(2)). These results suggest that LF is a PLA(2) and may be an important virulence factor of T. vaginalis mediating the destruction of host cells and contributing to tissue damage and inflammation in trichomoniasis.

Fluctuation of surface charge in membrane pores

Surface charge in track-etched polyethylene terephthalate (PET) membranes with narrow pores has been probed with a fluorescent cationic dye (3,3'-diethyloxacarbocyanine iodide (diO-C2-(3))) using confocal microscopy. Staining of negatively charged PET membranes with diO-C2-(3) is a useful measure of surface charge for the following reasons: 1) the dye inhibits K(+) currents through the pores and reduces their selectivity for cations; 2) it inhibits [3H]-choline+ transport and promotes 36Cl- transport across the membrane in a pH- and ionic-strength-dependent fashion; and 3) staining of pores by diO-C2-(3) is reduced by low pH and by the presence of divalent cations such as Ca2+ and Zn2+. Measurement of the time dependence of cyanine staining of pores shows fluctuations of fluorescence intensity that occur on the same time scale as do fluctuations of ionic current in such pores. These data support our earlier proposal that fluctuations in ionic current across pores in synthetic and biological membranes reflect fluctuations in the surface charge of the pore walls in addition to molecular changes in pore proteins.

Venom from the ectoparasitic wasp Nasonia vitripennis increases Na+ influx and activates phospholipase C and phospholipase A2 dependent signal transduction pathways in cultured insect cells

The mode of action of venom from the ectoparasitic wasp Nasonia vitripennis in eliciting cell death was examined using an in vitro approach with BTI-TN-5B1-4 cells, and the cell responses were compared to those evoked by the extensively studied wasp toxin mastoparan. Wasp venom increased plasma membrane permeability to Na+, resulting in cellular swelling and death due to oncosis. When ouabain was used to disable Na+, K+-ATPases, the effects of venom were enhanced. Measurements of intracellular calcium using fluo-4 AM revealed a rearrangement and an increase in cytosolic [Ca+2]i within 30 min after exposure of BTI-TN-5B1-4 cells to venom. This venom-mediated increase in Ca+2 was apparently due to mobilization of intracellular stores since the changes occurred in the absence of extracellular Ca+2. Phospholipase C (PLC) inhibitors, neomycin and U-73122, blocked the venom-induced death temporarily (<3h), but by 24h, all venom-treated cells swelled and lysed. Pre-treatment of cells with caffeine or theophylline but not ryanodine attenuated the induction of oncosis by wasp venom. Anti-inflammatory peptide 1 (antiflammin 1) but not bromophenacyl bromide, agents that block phospholipase A2 (PLA2) activity, abolished the responsiveness of BTI-TN-5B1-4 cells to venom. These results suggest that venom initiates cell death by inducing Ca+2 release from intracellular stores probably via phospholipase C and IP3. A possible mode of action for venom from N. vitripennis requiring dual activation of PLC and PLA2 is discussed and compared to the pathways known to be activated by mastoparan.

Clostridium perfringens beta-toxin forms potential-dependent, cation-selective channels in lipid bilayers

Recombinant beta-toxin from Clostridium perfringens type C was found to increase the conductance of bilayer lipid membranes (BLMs) by inducing channel activity. The channels exhibited a distribution of conductances within the range of 10 to 380 pS, with the majority of the channels falling into two categories of conductance at 110 and 60 pS. The radii of beta-toxin pores found for the conductance states of 110 and 60 pS were 12.7 and 11.1 A, respectively. The single channels and the steady-state currents induced by beta-toxin across the BLMs exhibited ideal monovalent cation selectivity. Addition of divalent cations (Zn(2+), Cd(2+), or Mg(2+)) at a concentration of 2 mM increased the rate of beta-toxin insertion into BLMs and the single-channel conductance, while application of 5 mM Zn(2+) to a beta-toxin-induced steady-state current decreased the inward current by approximately 45%. The mutation of arginine 212 of beta-toxin to aspartate, previously shown to increase the 50% lethal dose of beta-toxin for mice nearly 13-fold, significantly reduced the ability of beta-toxin to form channels. These data support the hypothesis that the lethal action of beta-toxin is based on the formation of cation-selective pores in susceptible cells.

The mechanism of membrane insertion for a cholesterol-dependent cytolysin: a novel paradigm for pore-forming toxins

Perfringolysin O (PFO), a water-soluble monomeric cytolysin secreted by pathogenic Clostridium perfringens, oligomerizes and forms large pores upon encountering cholesterol-containing membranes. Whereas all pore-forming bacterial toxins examined previously have been shown to penetrate the membrane using a single amphipathic beta hairpin per polypeptide, cysteine-scanning mutagenesis and multiple independent fluorescence techniques here reveal that each PFO monomer contains a second domain involved in pore formation, and that each of the two amphipathic beta hairpins completely spans the membrane. In the soluble monomer, these transmembrane segments are folded into six alpha helices. The insertion of two transmembrane hairpins per toxin monomer and the major change in secondary structure are striking and define a novel paradigm for the mechanism of membrane insertion by a cytolytic toxin.

In vitro analysis of venom from the wasp Nasonia vitripennis: susceptibility of different cell lines and venom-induced changes in plasma membrane permeability

The lethal effects of crude venom prepared from the ectoparasitic wasp Nasonia vitripennis were examined with cultured cells from six insect and two vertebrate species. Venom caused cells from Sarcophaga peregrina (NIH SaPe4), Drosophila melanogaster (CRL 1963), Trichoplusia ni (TN-368 and BTI-TN-5B1-4), Spodoptera frugiperda (SF-21AE), and Lymantria dispar (IPL-Ldfbc1) to round up, swell, and eventually die. Despite similar sensitivities and overlapping LC50 values [0.0004-0.0015 venom reservoir equivalents (VRE)/microl], profound differences were noted at the onset of cytotoxicity among the six insect cell lines: over 80% of the NIH SaPe4 and SF21AE cells were nonviable within 1 h after addition of an LC99 dose of venom, whereas the other cells required a 5-10-fold longer incubation period to produce mortality approaching 100%. In contrast, cells from the grass frog, Rana pipiens (ICR-2A), and goldfish, Carassius auratus (CAR), showed little sensitivity to the venom: six venom reservoir equivalents were needed to induce 50% mortality in ICR-2A cells [50% lethal concentration (LC50) = 0.067 VRE/microl), and 9 VRE did not yield sufficient mortality in CAR cells for us to calculate an LC50. All susceptible cells showed similar responses when incubated with wasp venom: retraction of cytoplasmic extensions (when present), blebbing of the plasma membrane, swelling of the plasma and nuclear membranes, condensation of nuclear material, and eventual cell death attributed to lysis. The rate of swelling and lysis in NIH SaPe4 and BTI-TN-5B1-4 cells exposed to venom appeared to be dependent on the diffusion potential of extracellular solutes (Na+ = choline > sucrose > or = raffinose > K+), which is consistent with a colloid-osmotic lysis mechanism of cell death. When T. ni cells were cotreated with venom and the K+ channel blocker 4-aminopyridine, cell swelling and lysis increased with increasing drug concentration. In contrast, cells from S. peregrina were protected from the effects of the venom when treated in a similar manner. Addition of certain divalent cations (Zn+2 and Ca+2) to the extracellular media 1 h postvenom incubation rescued both BTI-TN-5B1-4 and NIH SaPe4 cells, suggesting that protection was gained from closure of open pores rather than prevention of pore formation. Venom from N. vitripennis displayed no hemolytic activity toward sheep erythrocytes, supporting the view that venom intoxication is not by a nondiscriminate mechanism. A possible mode of action of the venom is discussed.

Formation of ring-shaped structures on erythrocyte membranes after treatment with botulinolysin, a thiol-activated hemolysin from Clostridium botulinum

Damage to erythrocyte membranes by botulinolysin (BLY) was studied by electron microscopy, which revealed ring-shaped structures with inner diameters and widths of approximately 32 and 6.7 nm, respectively. BLY bound to membranes at 0 degrees C, but subsequent treatment with glutaraldehyde prevented ring formation during further incubation at 37 degrees C. Zn2+ ions inhibited ring formation but not binding of BLY to membranes.

A conserved tryptophan in pneumolysin is a determinant of the characteristics of channels formed by pneumolysin in cells and planar lipid bilayers

Pneumolysin is one of the family of thiol-activatable, cytolytic toxins. Within these toxins the amino acid sequence Trp-Glu-Trp-Trp is conserved. Mutations made in this region of pneumolysin, residues 433-436 inclusive, did not affect cell binding or the formation of toxin oligomers in the target cell membrane. However, the mutations did affect haemolysis, leakage of low-molecular-mass metabolites from Lettre cells and the induction of conductance channels across planar lipid bilayers. Of eight modified pneumolysins examined, Trp-433-->Phe showed the smallest amount of haemolysis or leakage (less than 5% of wild type). Pneumolysin-induced leakage from Lettre cells was sensitive to inhibition by bivalent cations but the extent of inhibition varied depending on the modification. Leakage by the mutant Trp-433-->Phe was least sensitive to cation inhibition. The ion-conducting channels formed across planar lipid bilayers exhibit small (less than 30 pS), medium (30 pS-1 nS) and large (more than 1 nS) conductance steps. Small- and medium-sized channels were preferentially closed by bivalent cations. In contrast with wild-type toxin, which formed predominantly small channels, the modified toxin Trp-433-->Phe formed large channels that were insensitive to cation-induced closure. Polysaccharides of molecular mass more than 15 kDa inhibited haemolysis by wild-type toxin, but polysaccharide of up to 40 kDa did not prevent haemolysis by Trp-433-->Phe. Electron microscopy revealed that Trp-433-->Phe formed oligomeric arc and ring structures with dimensions identical with those of wild-type toxin, and that the ratio of arcs to rings formed was the same for wild-type toxin and the Trp-433-->Phe variant. We conclude that the change Trp-433-->Phe affects channel formation at a point subsequent to binding to the cell membrane and the formation of oligomers, and that the size of arc and ring structures revealed by electron microscopy does not reflect the functional state of the channels.

Prelytic and lytic conformations of erythrocyte-associated Escherichia coli hemolysin

Flow cytometry was developed as a method to assess the conformation of erythrocyte-bound Escherichia coli hemolysin polypeptide (HlyA). Topology of membrane-associated hemolysin (HlyA(E)) was investigated by testing surface accessibility of HlyA regions in lytic and nonlytic bound states, using a panel of 12 anti-HlyA monoclonal antibodies (MAbs). Hemolysin associates nonlytically with erythrocytes at 0 to 2 degrees C. To test the hypothesis that the nonlytic HlyA(E) conformation at 0 to 2 degrees C differs from the lytic conformation at 23 degrees C, MAb epitope reactivity profiles at the two temperatures were compared by flow cytometry. Four MAbs have distinctly increased reactivity at 0 to 2 degrees C compared to 23 degrees C. HlyA requires HlyC-dependent acylation at lysine residues 563 and 689 for lytic function. Toxin with cysteine substitution mutations at each lysine (HlyA(K563C) and HlyA(K689C)) as well as the nonacylated form of hemolysin made in a HlyC-deficient strain were examined by flow cytometry at 0 to 2 and 23 degrees C. The three mutants bind erythrocytes at wild-type toxin levels, but there are conformational changes reflected by altered MAb epitope accessibility for six of the MAbs. To test further the surface accessibility of regions in the vicinity of MAb-reactive epitopes, HlyA(E) was proteolytically treated prior to testing for MAb reactivity. Differences in protease susceptibility at 0 to 2 degrees and 23 degrees C for the reactivities of three of the MAbs further support the model of two distinct conformations of cell-associated toxin.

Membrane pores--from biology to track-etched membranes

Flow of ions through narrow pores, either induced in biological membranes or created in synthetic membrane filters, exhibits, under appropriate conditions: 1) rapid switching of ion current between high and low conducting states; 2) selectivity between different ions; 3) inhibition by protons or divalent cations with an order of efficacy usually H(+)> Zn(2+)>Ca(2+)>Mg(2+). It seems reasonable to conclude that these common properties arise from a common cause-the nature of the flow of ions close to a charged surface.

Low conductance states of a single ion channel are not 'closed'

We have used a polymer-exclusion method to estimate the sizes of the high- and low-conductance states of Staphylococcus aureus alpha-toxin channels across planar lipid bilayers. Despite a > 10-fold difference in conductance between high- and low-conductance states, the size differs by < 2-fold. We conclude that factors other than the dimensions have a strong influence on the conductance of alpha-toxin channels. We also show that the high conductance state is destabilized by the presence of high molecular weight polymers outside the channel, compatible with the removal of channel water as the high conductance state "shrinks" to the low conductance state.

Binding of antibodies to functional epitopes on the pore formed by Escherichia coli hemolysin in cells and model membranes

Escherichia coli hemolysin (HlyA) inserts into target membranes producing a cation-selective pore. We approached the problem of determining which portions of this protein remain exposed on the side of attack by applying specific antibodies. Results obtained with resealed erythrocyte ghosts and planar phospholipid membranes were compared. The effects of one polyclonal and four monoclonal anti-hemolysin antibodies (mAbs) were studied. Using ghosts we found one mAb which strongly reduced the ion-permeability through the preinserted HlyA channels and one which clearly increased it. Experiments with planar bilayers corroborated these results by showing that the former mAb effectively promoted the closed state of the channel whereas the latter forced the HlyA channel into an open configuration. Anti-hemolysin polyclonal antibodies initially stimulated but then prevented channel opening, indicating they contained clones able to act on both these channel determinants. They were effective only when applied on the same side as the hemolysin indicating that the epitopes were exposed to that side. Finally, the antigenic epitopes of three of the mAbs were localised on the HlyA molecule by using different mutants (amber and frame shift mutants and hemolysin gene hybrids).

Membrane depolarization prevents cell invasion by Bordetella pertussis adenylate cyclase toxin

Adenylate cyclase toxin from Bordetella pertussis is a 177-kDa calmodulin-activated enzyme that has the ability to enter eukaryotic cells and convert endogenous ATP into cAMP. Little is known, however, about the mechanism of cell entry. We now demonstrate that intoxication of cardiac myocytes by adenylate cyclase toxin is driven and controlled by the electrical potential across the plasma membrane. The steepness of the voltage dependence of intoxication is comparable with that previously observed for the activation of K+ and Na+ channels of excitable membranes. The voltage-sensitive process is downstream from toxin binding to the cell surface and appears to correspond to the translocation of the catalytic domain across the membrane.

The nodulation-signaling protein NodO from Rhizobium leguminosarum biovar viciae forms ion channels in membranes

The secreted nodulation-signaling protein NodO was purified from the supernatant of cultures of Rhizobium leguminosarum biovar viciae. The native protein has a M(r) of approximately 67,000, suggesting that it exists as a dimer since the DNA sequence predicts a M(r) of 30,002. Pure NodO protein had no protease, pectinase, or cellulase activity, and no binding was observed to lipooligosaccharide nodulation factors. Although NodO is relatively hydrophilic, it appeared to insert into liposomes and was protected by liposomes from proteolytic cleavage. When added to planar lipid bilayers, NodO formed cation-selective channels that allowed the movement of monovalent cations (K+ and Na+) across the membrane. NodO is a Ca(2+)-binding protein; in the presence of high concentrations of Ca2+, channel activity was reduced. We hypothesize that NodO plays a role in nodulation signaling by stimulating uptake of nodulation factors or by forming cation-specific channels that function synergistically with the proposed lipooligosaccharide-induced depolarization of the plasma membrane of leguminous plants.

Triton channels are sensitive to divalent cations and protons

Addition of Triton X-100 to planar bilayers composed of dioleoyl phosphatidyl choline, diphytanoyl phosphatidyl choline or mono-oleoyl glycerol induces single channel-like events when electrical conductivity across the bilayer is measured. Addition of divalent cations or protons causes channels to disappear; single channel conductance of remaining channels is not significantly altered; addition of EDTA or alkali (respectively) reverses the effect. It is concluded that sensitivity to divalent cations and protons need not be dependent on specific channel proteins or pore-forming toxins, but may be a feature of any aqueous pore across a lipid milieu.

Improved purification and characterization of hemolysin BL, a hemolytic dermonecrotic vascular permeability factor from Bacillus cereus

Bacillus cereus causes diarrheal and emetic food poisoning syndromes as well as a variety of mild to severe infections. A dermonecrotic vascular permeability (VP) factor has been implicated as a virulence factor in these illnesses. Hemolysin BL was previously identified as a unique tripartite hemolysin possessing VP activity. In this study, a high-yield purification scheme, which allowed quantitative characterization of hemolysin BL activity and determination of the molecular weight, pI, and N-terminal sequence of each component, was developed. Milligram quantities of the B, L1, and L2 components were highly purified by a combination of anion-exchange and hydroxylapatite chromatographies. The combined components had VP activity at low doses and were necrotic at higher doses. The toxin exhibited an unusual dose-response zone phenomenon in turbidometric hemolysis assays. Activity increased at doses up to 200 ng/ml, then decreased at doses up to 350 ng/ml, and was constant at doses up to at least 2,500 ng/ml. This behavior may provide an explanation for the unusual discontinuous pattern typical of hemolysin BL in gel diffusion assays. At high concentrations of one or two components, the presence of low amounts of the complementary component(s) resulted in full hemolytic activity. Erythrocytes were protected from lysis by Zn2+ at micromolar concentrations but not by Ca2+ and Mg2+ at concentrations up to 25 mM. These data provide guidelines for future work on this toxin and indicate that hemolysin BL is the dermonecrotic VP factor implicated as a B. cereus virulence factor.

The cytolytic toxin aerolysin: from the soluble form to the transmembrane channel

Aerolysin is a cytolytic toxin which forms channels in the plasma membranes of eucaryotic cells. The protein is secreted by Aeromonas hydrophila as an inactive protoxin. Its stability and water solubility are conferred by its ability to dimerize. Maturation of the protein occurs through proteolytic removal of a C-terminal peptide outside the secreting cell. Although the aerolysin which is so produced is still a dimer, it then has the ability to oligomerize. The oligomer is the active form of the toxin, capable of forming the transmembrane channels that disrupt cells. We review here the present knowledge about the structure of aerolysin in relation to the various steps in channel formation.

Ionic factors regulating the interaction of Gardnerella vaginalis hemolysin with red blood cells

We have studied the hemolytic properties of an exotoxin released by Gardnerella vaginalis (Gvh). We found that hemolysis induced by Gvh is modulated by the composition of the isotonic buffer in which the red cells are suspended. In particular, low pH enhances its lytic activity, whereas low ionic strength and divalent cations diminish it. The inhibitory effects of reduced salt concentration and divalent cations occur despite normal binding of the toxin to the cells. This suggests that some post-binding step is impaired. The toxin is also able to damage cholesterol-containing lipid vesicles, and even on these model membranes it is more active at low pH. From this point of view, Gvh has some similarity to Clostridium perfringens theta-toxin, a membrane-damaging toxin belonging to the family of 'thiol-activated' cytolysins produced by Gram-positive bacteria.

Pore-forming and haemolytic properties of the Gardnerella vaginalis cytolysin

The pleomorphic bacterium Gardnerella vaginalis releases in the culture broth a haemolytic exotoxin (Gvh) which is probably a virulence determinant of this unique bacterium, implicated in gynaecological and urological disorders. This 59 kDa cytolysin was purified to homogeneity in just one chromatographic step directly from the culture supernatant, a final specific activity up to 1.9 x 10(6) HU mg-1 being obtained. The toxin-induced lesion on human erythrocytes results from the formation of a pore whose radius is approximately 2.4 nm. The damage is inhibited by osmotic protectants and shows a sigmoidal dose-response profile suggesting an aggregation process of haemolysin molecules on the target membrane to create the functional lesion. The extent and the kinetics of haemolysis are strongly dependent on temperature and an activation energy of 64.0 kJ mol-1 has been derived. Lipid membranes can be very efficient inhibitors of Gvh-haemolysis, being able to bind the toxin quite avidly. The inhibitory effect requires the presence of cholesterol and it is stronger when cholesterol is mixed with negatively charged phospholipids rather than with zwitterionic phospholipids, suggesting that a negative surface potential increases the affinity of the toxin for the lipid bilayer. The functional properties of Gvh have been compared with those of Clostridium perfringens thetatoxin (PFO) and Escherichia coli haemolysin (HlyA), which are representative of widespread haemolysins produced by Gram-positive and Gram-negative bacteria, respectively. The toxin shares several features with the family of the so-called 'sulphydryl-activated' cytolysins produced by Gram-positive bacteria, although Gvh does not truly belong to this family, being deactivated by beta-mercaptoethanol and being antigenically distinct from them. We report here for the first time the detection in the vaginal fluid of infected women of a specific IgA response against the toxin.

Interaction of sporidesmin, a mycotoxin from Pithomyces chartarum, with lipid bilayers

Sporidesmin, a mycotoxin from Pithomyces chartarum is a hydrophobic molecule. It can therefore be easily incorporated in the cell membrane, where it is likely to cause changes in the bilayer organization and the properties of membrane proteins. In order to understand the redox behaviour of sporidesmin in a hydrophobic environment, we have investigated the effects of oxidized and reduced sporidesmin on the phase transition properties of bilayers and on the susceptibility of bilayers to pancreatic phospholipase A2 (PLA2). The changes induced by sporidesmin in the thermotropic phase transition profiles of dimyristoyl-sn-3-phosphatidyl choline (DMPC) bilayers were similar to those caused by solutes known to localize in the glycerol-backbone region of the lipid bilayer, suggesting a similar localization for oxidized and reduced sporidesmin. Neither form of toxin disrupt the bilayer or membrane organization even at relatively high mole fractions. At concentrations < 10 mole% both forms partitioned equally well in the gel and liquid-crystalline phases, whereas at higher concentrations (approximately 30 mole%) reduced sporidesmin is preferentially localized in the liquid-crystalline phase. These effects of sporidesmin on the phase properties of DMPC vesicles were also reported by the fluorescence behavior of 10-pyrenedecanoic acid (PDA). The effects of oxidized and reduced sporidesmins on PLA2 kinetics are consistent with their ability to perturb bilayer organisation.

Membrane damage: common mechanisms of induction and prevention

Common features in the induction of pores by various agents are as follows: induction is stochastic and progressive; damage by different agents is often synergistic and limited. The prevention of membrane damage is affected by trivalent and divalent cations, by low pH, by low ionic strength and by high osmotic pressure. The inhibitory role of protons and divalent cations is considered in greater detail: pore-forming agents can be classified into two groups: channels across planar lipid bilayers induced by the first group display voltage-sensitive, reversible inhibition by divalent cations; channels of the second group show voltage-insensitive, irreversible inhibition by divalent cations. A search for the ligands to which divalent cations and protons bind has proved elusive. Comparison with the phenomenon of 'surface conductance' through narrow apertures, that is manifest in the absence of any pore-forming agent, may prove fruitful.

The Staphylococcus aureus alpha-toxin channel complex and the effect of Ca2+ ions on its interaction with lipid layers

Using the techniques of two-dimensional crystallization on supported lipid bilayers together with computer image processing, two distinct two-dimensional crystal types of staphylococcal alpha-toxin complex are formed depending on the presence or absence of Ca2+ ions. Without Ca2+, these are hexagonally packed (in A, a = b = 89.5 +/- 2.5 A; theta = 119.7 degrees) With Ca2+ present, rectangular crystal packing is seen (in A, a = 114.8 +/- 1.6 A, b = 140.2 +/- 0.7 A; theta = 89.1 degrees). A third, banded crystal type is also seen which is interpreted as a side-to-side packing of regular tubules. We use these tubular crystals for cross-correlation searches with top and side-on views of the complex from single particle reconstructions, and with the repeating units from the two-dimensional crystal types. The results lead us to propose a model in which the different two-dimensional crystal types are formed as a result of alpha-toxin hexamers packing in different orientations. In the hexagonal crystals the hexamers lie end-on with a 6-fold axis in projection. On the addition of Ca2+, the hexamers reorient to lie tilted with respect to the support, thus giving rise to a rectangular projection.

Differential sensitivity of pneumolysin-induced channels to gating by divalent cations

The induction of channels across planar lipid bilayers by purified, recombinant pneumolysin (a hemolytic protein from Streptococcus pneumoniae) has been studied by measuring increases in electrical conductivity. Pneumolysin-induced channels exhibit a wide range of single channel conductances (less than 50 pS to greater than 1 nS at 0.1 M KCl). Channels can be categorized on the basis of their K+:Cl- selectivity: the smallest channels are strongly cation selective, with t+ (the cation transference number) approaching 1.0; the largest channels are unselective (t+ approximately 0.5). Channels tend to remain open at all voltages (-150 to 150 mV); only the smallest channels exhibit any rectification. In the presence of divalent cations (1-5 mM Zn2+; 10-20 mM Ca2+), small (less than 50 pS) and medium-sized (50 pS to 1 nS) channels are closed in a voltage-dependent manner (more closure at higher voltages); at 0 voltage channels reopen. Overall selectivity is reduced by divalent cations, compatible with small, selective channels being closed preferentially to large, nonselective ones. It is concluded that a single molecular species (pneumolysin) induces multiple-sized channels that can be categorized by cation:anion selectivity and by their sensitivity to closure by divalent cations.

Effect of fatty acyl domain of phospholipids on the membrane-channel formation of Staphylococcus aureus alpha-toxin in liposome membrane

By use of carboxyfluorescein-loaded multilamellar liposomes prepared from synthetic phosphatidylcholine (PC) or sphingomyelin and cholesterol in a molar ratio of 1:1, we studied whether or not fatty acyl domain of the phospholipids affects the membrane-damaging action (or channel formation) of Staphylococcus aureus alpha-toxin on the phospholipid-cholesterol membranes. Our data indicated: (1) that toxin-induced carboxyfluorescein-leakage from the liposomes composed of saturated fatty acyl residue-carrying PC and cholesterol was decreased with increasing chain length of the acyl residues between 12 and 18 carbon atoms, although toxin-binding to the liposomes was not significantly affected by the length of fatty acyl residue; (2) that unsaturated fatty acyl residue in PC or sphingomyelin molecule conferred higher sensitivity to alpha-toxin on the phospholipid-cholesterol liposomes, compared with saturated fatty acyl residues; and (3) that hexamerization of alpha-toxin, estimated by SDS-polyacrylamide gel electrophoresis, occurred more efficiently on the liposomes composed of PC with shorter fatty acyl chain or unsaturated fatty acyl chain. Thus, hydrophobic domain of the phospholipids influences membrane-channel formation of alpha-toxin in the phospholipid-cholesterol membrane, perhaps by modulating packing of phospholipid, cholesterol and the toxin in membrane.

Divalent cation-sensitive pores formed by natural and synthetic melittin and by Triton X-100

Leakage of ions and low-molecular-weight metabolites from Lettre cells is induced by synthetic melittin, as effectively as by melittin isolated from bee venom; in each case leakage is inhibited by Ca2+, Zn2+ or H+. Inhibition of leakage by divalent cations is reversible in that Lettre cells incubated with melittin (or with Triton X-100) in the presence of inhibitory amounts of Zn2+, when freed of Zn2+ by EGTA or by centrifugation, begin to leak (in Zn2(+)-sensitive manner). Electrorotation of Lettre cells is altered by melittin, compatible with membrane permeabilization; melittin plus Zn2+ does not alter electrorotation until Zn2+ (and unbound melittin) are removed. Melittin or Triton X-100 added to calcein-loaded liposomes induces leakage of calcein; divalent cations inhibit. Energy transfer between liposome-associated melittin and 2-, 7- or 12-(9-anthroyloxy)stearate (AS) is maximal with 12-AS; addition of Zn2+ has little effect. Circular dichroism spectra of melittin plus liposomes are unaffected by Zn2+. These results show that the formation of divalent cation-sensitive pores is not dependent on the presence of endogenous membrane proteins and that the action of divalent cations is not by displacement of melittin (or Triton) from the lipid bilayer.