Membrane Interactions of Host-defense Peptides ... - IngentaConnect

0 downloads 0 Views 262KB Size Report
Department of Chemistry and Staedler Minerva Center for Mesoscopic, Macromolecular Engineering, Ben Gurion. University of the Negev, Beersheva 84105, ...
Current Protein and Peptide Science, 2005, 6, 103-114

103

Membrane Interactions of Host-defense Peptides Studied in Model Systems Raz Jelinek* and Sofiya Kolusheva Department of Chemistry and Staedler Minerva Center for Mesoscopic, Macromolecular Engineering, Ben Gurion University of the Negev, Beersheva 84105, Israel Abstract: Host-defense, antibiotic peptides are believed to generate their cytolytic effects by interacting with the membranes of bacterial cells. Direct analyses of peptide interactions with real cellular membranes are difficult, however, due to the high complexity of physiological membranes. This review summarizes experimental work aiming to understand peptide-membrane interactions and their relationships with the peptides' biological actions using specific model systems. Varied model assemblies have been constructed that generally aim to mimic the fundamental lipid bilayer organization of the membrane. The model systems we will describe include multilamellar and unilamellar vesicles, planar lipid bilayers, lipid monolayers and micelles, and colorimetric biomimetic membranes. The different artificial models have facilitated examination of specific biological or chemical parameters affecting peptide action, for example the effect of membrane lipid composition on peptide affinities and membrane penetration, the relationship between membrane fluidity and peptide interactions, the conformations of active peptides, and other factors. We evaluate the strengths and limitations of the various approaches, and point to future directions in the field.

1. INTRODUCTION The potential of host-defense peptides (also referred to as antimicrobial peptides, or AMPs) as promising therapeutic agents for combating bacterial infections has been a powerful driving force for conducting numerous studies aiming to decipher their mode of action. Early observations indicated a possible connection between membrane permeation and bacterial killing [1]. Indeed, most known AMPs are short amphipathic cationic sequences that interact strongly with bacterial membranes [2]. These interactions, often resulting in pore formation and/or cell lysis, are believed to constitute primary factors of the bactericidal capacity of AMPs. Thus, the study of peptide-membrane interactions and elucidating the effects of AMPs on the cell membrane have become increasingly important for understanding the biological activities of such peptides. Despite a significant amount of experimental data, many of the central tenets of AMP action remain unresolved and debated. Among the central issues pursued are the molecular mechanisms and characteristics of peptide-mediated cell lysis, degree of peptide insertion into the membranes, the extent and significance of pore formation, membrane micellization processes, and others. Such questions are almost universally addressed through examination of model membrane systems. Studying peptide-membrane complexes in model systems is primarily due to the molecular and structural complexity of the membrane of actual cells, combined with the difficulties encountered in working with complete cellular systems. The construction of artificial *Address correspondence to this author at the Department of Chemistry and Staedler Minerva Center for Mesoscopic, Macromolecular Engineering, Ben Gurion University of the Negev, Beersheva 84105, Israel; Tel: +972-86461747; Fax: +972-8-6472943; E-mail: [email protected]

1389-2037/05 $50.00+.00

models further allows examination of the contribution of specific parameters to membrane-peptide interactions and peptide activities such as lipid type and composition, peptide-lipid ratio, surface curvature, and others by diverse bio-analytical and spectroscopic methods. This Review summarizes the different model lipid systems for studying the action of host-defense peptides, including multilamellar and unilamellar vesicle assemblies, supported lipid bilayers, micelles, lipid monolayers at the air water interface, and recently developed chromatic vesicle platforms. We discuss the important properties of each approach and the type of information extracted by the experiments. Due to the large body of work in the field and limited space we only refer to representative publications. In particular, we have tried to provide here a critical summary of the subJect that would allow the reader to evaluate the biological insights provided on AMP-membrane interactions by using different model systems. Several excellent reviews on related topics have appeared in recent years [3-10]. The use of model systems for elucidating AMP biological actions has progressed hand in hand with a simultaneous development of bio-analytical techniques. The majority of spectroscopic methodologies applied for studying host defense peptides in model systems have focused on the structural modifications of the peptides following their interactions with the lipid assembly. Such studies have attempted to correlate the peptide conformational changes induced within the lipid environments with their bactericidal properties. Accordingly, we also summarize the analytical techniques applied in conjunction with biomimetic membranes, including fluorescence spectroscopy, nuclear magnetic resonance (NMR), circular dichroism (CD), infrared spectroscopy, and others.

© 2005 Bentham Science Publishers Ltd.

104

Current Protein and Peptide Science, 2005, Vol. 6, No. 1

2. LIPID VESICLES 2.1 Multilamellar Vesicles Multilamellar vesicles (MLVs) have been used for studying various aspects of membrane permeation by hostdefense peptides and their bactericidal mechanisms. MLVs retain the lipid bilayer organization of cellular membranes (Fig. 1) and thus could serve as a useful model for studying peptide association and induced effects within the bilayer. The primary advantages of MLVs for elucidating membrane properties and interactions are often related to the technical requirements of several bio-analytical techniques employed for investigating peptide-membrane interactions, such as solid state NMR, calorimetry, and X-ray diffraction.

Jelinek and Kolusheva 31

P solid-state NMR of DMPC/DMPG MLVs similarly suggested that the action of the wasp venom peptide mastoparan involves induction of non-lamellar phases in the bilayer [12]. Solid-state NMR techniques have been applied to investigate the modification of AMP molecular properties in membranes. High-resolution solid-state NMR experiments could provide information on molecular interactions as well as structural and dynamical parameters pertaining to both the lipid molecules as well as the inserted peptides. A solid-state NMR study revealed, for example, that the potent AMP protegrin-1 (PG-1) formed highly immobile large aggregates within lipid bilayers [17]. Another NMR study utilizing Mn 2+ ions incorporated within the lipid bilayer examined the depth of insertion of PG-1 through the induction of the distancedependent dipolar relaxation of the nuclear spins [18]. Solid-state NMR methods have been applied in MLV suspensions to address one of the most important question related to the bactericidal mechanisms of AMPs, specifically the significance of interactions of the peptides with the phospholipid headgroups vs. insertion into the hydrophobic interior of the bilayer. 31P and 2H investigations were carried out to determine the extent of headgroup disruption by short AMPs compared with their penetration into the hydrophobic environment of the acyl chains [19]. That research determined that short peptides preferably accumulated at the lipid surface, while longer sequences tended to penetrate deeper into the bilayer. Similar interpretation emerged from solidstate NMR studies of oriented lipid systems (see below) [14, 20]. Surface dimerization of the amphipathic AMP magainin-2 was suggested as the primary mechanism for the membrane permeabilization properties of the peptide, using a newly developed NMR technique denoted transferred nuclear Overhauser enhancement (TRNOE) spectroscopy [21].

Fig. (1). Schematic picture of a multilamellar vesicle assembly.

The use of MLVs has been essential in varied solid-state NMR studies of peptide-membrane assemblies. Lipid multilamellae provide sufficient sample quantities and acceptable signal-to-noise ratios, but also, critically, maintain a favorable time-regime for the solid-state NMR experiments [11-13]. Studies of structure, dynamics, and orientation of antimicrobial peptides in membranes, and their structurefunction relationships as probed by solid-state NMR spectroscopy were reviewed [14, 15]. Examples for specific NMR experiments for deciphering AMP-membrane interactions include the application of 31P NMR for studying pore formation of equinatoxin II, a eukaryotic toxin, and its specificity to phospholipid targets within bacterial membranes [16]. The appearance of toroidal pores within the bilayers was deduced from the observation of an isotropic component in the 31P NMR spectra of the MLV suspensions.

The effect of AMP permeation on membranes has been often described in terms of lipid disruption and decreased order. Several techniques, primarily differential scanning calorimetry (DSC) and x-ray diffraction, have been applied to probe the ordering and molecular packing of lipid bilayers within MLVs prior and following interactions with AMPs. Several studies utilizing these methods in model lipid systems have yielded quantitative information on the effects of host defense peptides on membrane structure, as well as on peptide localization within the lipid bilayer [22]. Specifically, DSC has been applied in multilamellar phospholipid systems to probe the thermodynamic profiles of the lipid assemblies and the consequences of peptide interactions in term of ordering and cooperative properties of the bilayers [22]. A detailed DSC study of membrane interactions of gramicidin S (GS) indicated that the peptide appeared to perturb packing in liquid crystalline lipid phases inducing the formation of inverted cubic phases in certain conditions [23]. That study suggested that this type of bilayer destabilization is a primary constituent of the antimicrobial mode of action of GS. Similar to NMR, DSC analyses require relatively large sample quantities for probing modifications of the lipid phase transitions induced by the associated peptides [23-25]. Grasso et al, for example, employed DSC to elucidate the

Membrane Interactions of Host-defense Peptides Studied in Model Systems

effect of a scrambled hydrophobic/hydrophilic sequence on the cooperative transitions of MLVs composed of DPPC [26]. A similar microcalorimetric study analyzed the binding properties of the peptide lactoferricin B to phosphatidylglycerol (PG) or phosphatidylcholine (PC) MLVs [27]. These studies pointed to the higher affinity of certain AMPs to negatively-charged phospholipids, generally more abundant within bacterial membranes. Particularly important, an x-ray diffraction and DSC investigation indicated that membrane curvature strain and hydrophobic mismatch between the interacting peptides and lipids are prominent factors in the mechanisms of membrane perturbation and cell lysis [24]. This interpretation was supported by a DSC study indicating that a short AMP containing 17 β-amino acids promoted a negative curvature within phospholipid assemblies [28]. Special MLV preparation methods have been introduced for facilitating investigation of peptide-membrane interactions using advanced structural techniques such as neutron diffraction. By carefully modifying sample hydration, neutron diffraction analyses could identify distinct lipid crystallization processes occurring within the multilamellae and relate this information to structural transformations induced by bound AMPs. Yang et al showed that crystallization of membrane pores occurred spontaneously following association of magainin and protegrin with MLVs [29]. Studies of peptide-membrane interactions have frequently utilized different model assemblies rather than a single system, lending better credence to the experimental conclusions. An example of this multiprong approach has been the application of Fourier transform IR spectroscopy to characterize interactions of gramicidin S (GS) with MLVs, lipid monolayers, and micelles (see subsections below) [30]. That systematic work examined the dependence of peptide interactions on temperature and lipid phase. Interestingly, the researchers found that GS was completely or partially excluded from the gel states of all of the lipid bilayers examined but strongly partitioned into micelles, lipid monolayers, or bilayers in the liquid crystalline phase, suggesting a possible mechanistic prerequisite for bactericidal action of the peptide. Gramicidin S represents an AMP that was extensively studied in diverse model systems to elucidate its bactericidal mechanisms. MLVs were mostly employed for deciphering GS-membrane interactions using electron spin resonance (ESR) [31], fluorescence spectroscopy using the diphenylhexatriene (DPH) probe for lipid dynamics [32], FTIR spectroscopy [33], and Raman spectroscopy [34]. X ray studies carried out in total membrane lipid extracts from microbial species indicated that GS induced the formation of cubic lipid phases, an important observation that could shed light on the underlying influence of the peptide on bacterial membranes [35]. Experiments utilizing various analytical techniques have indicated that cholesterol attenuates the interaction of GS with lipid bilayers [36]. MLV suspensions have made possible morphological characterization of membrane-peptide complexes using visualization techniques such as electron microscopy. Cryoelectron microscopy, complemented by 31P and 2H NMR,

Current Protein and Peptide Science, 2005, Vol. 6, No. 1

105

was applied to study the permeability and morphological perturbations induced by the pore-forming peptide nisin in phosphatidylcholine membranes [37]. The researchers observed dramatic modifications of the morphologies of the lipid assemblies following interactions with the peptide, and could further detect an inhibitory effect of cholesterol. The electron microscopy data suggested a distinct interplay between nisin and lipid constituents of the bilayer resulting in membrane permeability. 2.2 Unilamellar Vesicles Unilamellar vesicles have been used widely for studying membrane processes in general and peptide-membrane interactions in particular. This is due in most part to the fact that conceptually such vesicles mimic a cell assembly - in which the lipid bilayer forms an enclosed volume separated from the external solution. Traditionally, unilamellar vesicles have been divided into three main groups depending on their sizes: small unilamellar vesicles (SUVs), large unilamellar vesicles (LUVs), and giant unilamellar vesicles (GUVs) (Fig. 2). The preparation of the different vesicle classes depends upon the types of lipids used and experimental methods [3841]. Larger-size vesicles have been generally preferred as model systems for studying membrane processes since their curvature more closely resembles actual cells [42, 43]. Unilamellar vesicles have been used in combination with spectroscopic techniques to decipher membrane association and permeation by AMPs, and to correlate the biophysical data with the antibacterial activities of the peptides. CD spectroscopy has been an important tool for probing secondary structures of vesicle-associated peptides. This technique has been applied to probe the effect of negative phospholipids compared to zwitterionic phospholipids on AMP conformations [4,44,45]. Electron spin resonance (ESR) has been an important tool for studying AMP insertion into the bilayer and the resultant modification of bilayer fluidity and lipid motion [46]. ESR experiments have usually employed spin labels covalently attached to the lipid acyl chains at various positions. A large number of studies have employed fluorescence leakage techniques to probe AMP effect on the cell membrane. In such experiments the changes in the fluorescence of probes entrapped within unilamellar vesicles are monitored following interactions of the membrane-active peptides with the vesicles. Specifically, pore formation or complete bilayer destruction by the peptides result in leakage of the encapsulated fluorescence dyes, thereby significantly modifying [increasing or reducing] their fluorescence emission. Vesicle-encapsulated dye release by Hagfish intestinal antimicrobial peptides (HFIAPs) confirmed the high specificity of these peptides to bacterial membranes [47]. Leakage experiments were carried out in a study comparing the membrane activity of a synthetic amphipathic β-sheet cationic AMP to the action of known bactericidal sequences [48]. Oren and Shai investigated membrane + permeation by cyclic peptides using a K -sensitive fluorescent dye [49]. Among the most common experiments in that regard is the construction of vesicles encapsulating self-quenching fluorescent probes such as calcein and fluorescein. When

106

Current Protein and Peptide Science, 2005, Vol. 6, No. 1

Jelinek and Kolusheva

Fig. (2). Schematic pictures of unilamellar vesicles: small unilamellar vesicles (SUVs, left), large unilamellar vesicles (LUVs, middle), giant unilamellar vesicles (GUVs, right).

such labels are entrapped within the vesicle interior a considerable self-quenching occurs. However, bilayer destruction or formation of pores within the enclosing lipid surface gives rise to an increase in fluorescence due to leakage of the probe molecules into the external solution [50]. Mangoni et al recorded the leakage induced in vesicles containing different phospholipid compositions by temporins, short AMPs secreted by an amphibian species, and examined the link between the biophysical data and the bactericidal activities of the peptides [51]. A systematic study of the effect of different cationic AMP families, including α-helical, β-sheet, extended, or cyclic structures examined lipid disruption and calcein leakage from LUVs comprised mostly of negatively-charged phospholipids [52]. The experiments assessed peptide translocation across the lipid bilayers and the induction of lipid “flip-flop” processes at a range of peptide/lipid ratios, however no common structural peptide feature was identified as a central determinant for bactericidal activity. Beside of monitoring dye leakage, numerous experiments have recorded the fluorescence of trypthophane residues within AMP sequences, either Trp residues within the native sequences or inserted intentionally. The method is based on the high sensitivity of the spectral position and intensity of the fluorescence peak to the hydrophobicity of the Trp environment, thus providing a measure of the depth of bilayer insertion of the peptide [53, 54]. Other fluorescence techniques have utilized dyes embedded within the lipid bilayer, either physically incorporated or covalently bound to the lipid moieties. Such experiments were similarly designed to evaluate the penetration and bilayer localization of the peptide. 8-anilino-1-naphthalene sulfonic acid (ANS) has been a common fluorescence probe for studying the effect of AMPs [55]. Another example is the use of DPH- and pyrenelabeled phospholipids [56]. These fluorescence dyes were used for studying temporin-peptide interactions and revealed that dynamical modifications were induced by the peptides at

specific parts within the lipid bilayer [56]. The researchers detected that peptide penetration was enhanced by negatively-charged bacterial phospholipids, while an opposite effect was observed when the vesicles contained the eukaryotic lipid cholesterol. Indeed, one of the most important questions pertaining to the activities of host-defense peptides, studied extensively in unilamellar vesicle systems, concerns their specificity to bacterial rather than host-cell membranes. Several hypotheses emphasize AMP affinity to membranes containing an abundance of negatively-charged phospholipids [52, 57, 58], while other models focus on specificity towards particular components and domains on bacterial walls [59], or binding to molecular constituents on outer bacterial surfaces, such as lipopolysaccharides (LPS) [60]. A comparative study of magainin 2 penetration into LPScontaining SUVs vs. its insertion into LPS-free vesicles determined that LPS is most likely the primary molecular determinant for binding and bacterial lysis by this peptide [60]. Experiments utilizing LUVs formed exclusively from the anionic lipid palmitoyloleoylphosphatidylglycerol (POPG) suggested that multimeric pore formation is the predominant bactericidal mechanism of human defensins [61]. Unilamellar vesicles have been prepared not just from specific phospholipid units but also from total lipid extracts from varied microorganisms. Hristova et al discovered that rabbit neutrophil defensins permeate LUVs composed of E. coli lipid extracts to a significantly higher degree compared to vesicles constructed from mixtures of distinct components [62]. The researchers proposed that the overall composition of the bacterial membrane, in particular the presence and ratios between phosphatidylethanolamine, phosphatidylglycerol, and cardiolipin, play critical roles in affecting the bactericidal profiles of the peptides.

Membrane Interactions of Host-defense Peptides Studied in Model Systems

Valuable insights into the biological profiles and modes of action of AMPs could be gained from analysis of differences between the peptide activities as measured in model membranes (such as unilamellar vesicles) and their effects on intact bacterial cells. For example, significant differences in activities were observed between the interactions of nisin with unilamellar vesicles and its effects on varied microorganisms [63,64]. Intriguingly, the differences were traced to the presence of lipid II on the bacterial surface - a putative precursor for high affinity binding of nisin and its increased activity towards bacterial cells. Many investigations tailored the choice of vesicle type to the technique applied and the information desired. Epand et al, for example, characterized cytolytic peptides composed of leucines and lysines through application of CD spectroscopy on SUVs, while LUVs were prepared for determination of induced leakage [65]. While most spectroscopic studies of peptide insertion into unilamellar vesicles focused on SUVs or LUVs as the preferred model systems, giant unilamellar vesicles (GUVs) have been useful for actual visualization of the topological and morphological effects of AMPs. Optical microscopy was employed for studying the effects of magainin II and indolicidin on GUVs [66]. The experiments demonstrated that magainin 2 had essentially no influence on GUVs that contained either zwitterionic phospholipids or zwitterionic/cholesterol mixtures – echoing the minimal hemolytic capabilities of this peptide. Contraction and fractionation were observed, however, when acidic phospholipids were present in the giant vesicles. Similar shrinkage effects were detected when indolicidin was added to GUVs composed of zwitterionic phospholipids and cholesterol, pointing to the cytotoxic properties of this particular peptide [66]. 3. COLORIMETRIC BIO-MIMETIC MEMBRANE ASSEMBLIES A recent development in the field of biomimetic membrane systems has been the introduction of mixed vesicles composed of physiological lipids and

Current Protein and Peptide Science, 2005, Vol. 6, No. 1

107

polydiacetylene (PDA) that exhibit unique chromatic properties. This novel platform could be used for studying and rapid screening of membrane processes in general, and for analysis of AMP-membrane interactions in particular. The PDA polymer is formed through 1,4 addition of the diacetylenic monomers, initiated by UV irradiation, and is intensely blue to the naked eye due to absorption of light at around 650 nm in the visible region of the electromagnetic spectrum [66]. Conjugated PDA vesicles have attracted attention due to their versatile electronic and optical properties [67]. Specifically, rapid blue-red transformations of the PDA can be induced by varied external perturbations, such as temperature, surface pressure, ion concentration, and others [68-70]. Relevant to biological membrane applications, it was recently demonstrated that mixed vesicles can be prepared that contain both PDA as well as varied lipid species. Biophysical analyses confirmed that such assemblies comprise of phospholipid bilayers interspersed within the framework of the PDA without affecting the unique chromatic properties of the polymer [71]. Fig. 3 depicts a schematic description of the lipid/PDA vesicle assemblies, showing the mixed domains of phospholipids and the polymer. Several studies have shown that interactions of membrane-active peptides with the mixed vesicles induce dramatic colorimetric transitions [72-77]. The visible transformations of the vesicles could both report upon the occurrence of peptide-membrane binding as well as provide information on the mechanisms of peptide interaction with the lipid bilayer. The new lipid/PDA platform exhibits several important advantages for use as an assay for studying peptidemembrane interactions and processes, the most significant being the presence of organized lipid bilayers – the biomimetic membrane target - incorporated within the PDA matrix which acts as a colorimetric reporter. Importantly, it was shown that the lipid domains within the mixed vesicles can be composed of a variety of synthetic and natural phospholipids, glycolipids, lipopolysaccharides, cholesterol, and other molecules, and that these mixtures functionally

Fig. (3). Schematic picture of a lipid/polydiacetylene chromatic vesicle. The polymer domains are shown in blue, while lipid domains in green.

108

Current Protein and Peptide Science, 2005, Vol. 6, No. 1

mimic biological membrane platforms [71]. Moreover, diverse biological processes that occur exclusively at the lipid bilayer moieties induce the visible colorimetric transformations within the adJacent PDA. The phospholipid/PDA assay has been applied for studying varied aspects of AMP-membrane interactions and AMP biological action. Underlying the color change mechanism is the disruption of the lipid interface by membrane-interacting compounds and the depth of penetration into the lipid layer [72-74]. In particular, correlation was observed between the relative perturbation of the lipid bilayer surface by membrane-active compounds and the degree of color changes [72-74, 76, 77]. Peptides that preferably disrupt the lipid head-group region were shown to induce more pronounced color transitions, while deeper penetration into the hydrophobic bilayer core generally gave rise to more moderate blue-red transformations [73]. These structural relationships were confirmed through applications of analytical and spectroscopic techniques [72-77]. Specifically, the colorimetric transitions observed in the PDA assemblies have been ascribed to the perturbation of the pendant side-chains and modification of molecular packing of the PDA framework induced by the interfacial effects [71,74]. Thus, greater lipid surface interactions would result in more pronounced perturbations of the adJacent polymer domains, giving rise to higher colorimetric transitions. The new colorimetric assay has been applied to characterize membrane interactions and relate those interactions to the biological activities of several AMP families. Several studies elucidated parameters such as bilayer localization and the relationship with the antimicrobial and cytolytic properties of melittin and its derivatives [72,73], polymyxin B analogues [74], and human defensins [76,77]. A colorimetric study, complemented with advanced fluorescence techniques, recently determined the extent of interface localization of model AMPs within the biomimetic vesicles [76,77]. Phospholipid/PDA vesicles have also exhibited sensitivity to synergy effects occurring in certain antimicrobial peptide families [74]. The colorimetric membrane assay can also provide important insight into the cellular pathways affecting the bactericidal activity of host defense peptides, experiments utilizing the phospholipid/ PDA platform demonstrated that proteolytic cleavage of the prosegment of human defensin cryprdin-4 is essential for conferring membrane interaction to the mature biologicallyactive peptide [76]. 4. PLANAR LIPID SYSTEMS Planar lipid systems have been employed in varied applications for studying peptide association with membranes. Among the assemblies constructed were Langmuir monolayers [78], Langmuir-Blodgett films [79, 80], supported planar bilayers [81,82], self-assembled monolayers [83, 84], polymer-tethered membranes [85, 86], and black lipid membranes [87, 88]. Planar systems generally model the phospholipid ordering within cellular membranes, and can be also designed to mimic the lateral organization of cell surfaces [89]. The use of planar lipid models for studying peptide-membrane interactions often

Jelinek and Kolusheva

stems from the availability of varied experimental techniques for surface characterization, including microscopy methods [90], scattering [85], thermodynamic analysis [91, 92], and others. 4.1 Langmuir Monolayers Phospholipid monolayers deposited at the air/water interface of aqueous solutions (Langmuir monolayers, Fig. 4) have been used as model membrane surfaces for studying peptide interactions [92]. In general, investigations using such systems have yielded information both on the cooperative properties of the lipid-interacting peptides, as well as the organization and disruption of the lipid layers following AMP interactions. In many instances, physical and morphological analyses of Langmuir film / AMPs assemblies have been conducted by thermodynamic methods (pressurearea isotherms) and microscopy techniques (fluorescence microscopy, Brewster angle microscopy). A systematic study of a series of bombolitins, bumblebee-venom peptides, and their synthetic analogs employed Langmuir monolayers coupled with fluorescence microscopy to investigate the structural disruption and phase separation induced by the incorporated peptides within the phospholipid monolayers [93]. Zhang et al. investigated the relationships between the antibacterial activities and membrane interactions of several variants of the horseshoe crab antimicrobial peptide polyphemusin I, in which some of the analogues displayed entirely different amphipathic properties [94]. Langmuir monolayer analysis showed that the degree of negative charge on the membrane surface was closely related to bacterial membrane permeabilization. Importantly, binding of the peptides to monolayers composed solely of LPS pointed to preferential peptide binding to the saccharide headgroups, and re-emphasized the significance of electrostatic interactions for promoting peptide insertion into the external leaflet of the bacterial membrane. Other experimental approaches for studying AMP association with Langmuir monolayers have been described. A synchrotron grazing incidence diffraction and X-ray reflectivity study of frog skin antimicrobial peptides added to Langmuir monolayers was reported [95]. The study concluded that the peptides were fully interspersed within negatively-charged phospholipids but formed distinct domains in zwitterionic monolayers, suggesting that AMPs can discriminate among the major phospholipid components of bacterial and mammalian membranes. Recent developments have enabled the application of surfacesensitive X-ray and neutron scattering techniques for characterization of molecularly thin peptide crystal sheets constructed via Langmuir monolayers [85]. 4.2 Solid-supported Lipid Systems Solid-supported lipid monolayers and bilayers (Fig. 5) have been popular biomimetic membrane systems since they allow examining peptide interactions at the lipid/water interface while at the same time limiting lipid motion through surface immobilization. Constricting lipid mobility in supported lipid layers might make such systems a more realistic model compared to Langmuir films since this better

Membrane Interactions of Host-defense Peptides Studied in Model Systems

Fig. (4). Schematic picture of a Langmuir monolayer.

Fig. (5). Schematic picture of a hybrid bilayer system.

Current Protein and Peptide Science, 2005, Vol. 6, No. 1

109

110

Current Protein and Peptide Science, 2005, Vol. 6, No. 1

resembles lipid dynamics within real membranes. In addition, supported bilayers on solid substrates have become particularly useful for measuring secondary structures and orientations of peptides in lipid environments by using spectroscopic techniques such as attenuated total reflection Fourier transform IR (ATR-FTIR) [78, 96]. Solid-supported lipid bilayers have served as a platform for studying the effect of AMPs on the cooperative properties of the lipids. As an example, planar lipid monolayers with different compositions were constructed for investigation of the bactericidal mechanism of protegrin-1 (PG-1) [58]. The researchers applied epifluorescence microscopy and surface x-ray scattering in order to examine whether the strong surface interactions of PG-1 indeed depend on the lipid composition of the membrane, and whether such interactions affect the lipid order. The analysis showed a pronounced monolayer destabilization by the peptide when the lipid assembly consisted mainly of lipid A - the major constituent of the outer membrane of Gram negative bacteria [97, 98]. Many studies have applied electrophysiological techniques for characterizing AMP-membrane interactions using reconstituted lipid bilayers on solid surfaces. A recent study revealed that several α-helical CAP18-derived peptides induced lesion formation within different types of artificial membranes at clamp voltages below the transmembrane potential of the natural membrane [99]. That study suggested that the interplay between the physicochemical properties of both the peptides as well as the target membranes is an important factor for determination of the antibacterial activity. A similar electrophysiological study of the action of Hagfish intestinal antimicrobial peptides reported disordered current fluctuations following addition of the peptides to supported lipid bilayers, which was interpreted as an evidence for pore formation induced by the peptides [47]. A lipid multibilayer assembly deposited on a polyion/alkylthiol solid-stabilized layer was fabricated for studying the topological and electrochemical consequences of the insertion of protegrin-1 (PG-1) into membranes [83]. This intriguing approach explored the relationship between peptide binding, pore formation, and modification of surface roughness by the adsorbed peptide using cyclic voltametry, surface plasmon resonance (SPR, see below) and atomic force microscopy (AFM). Oriented lipid layers have been useful models for providing structural information on peptide conformations in membrane and their insertion processes. Such assemblies have facilitates the use of spectroscopic techniques primarily solid-state NMR for extraction of structural parameters. Specifically, application of solid-state NMR on oriented samples yields varied structural parameters that cannot be extracted in conventional solid samples [14,100]. The orientation of the peptide molecules within the organized lipid assemblies reduces anisotropic broadening, thereby simplifying analysis of the NMR spectra while additionally allows extraction of various distance and angle parameters pertaining to the peptide. Solid-state NMR applied to oriented bilayers provided structural analyses of membraneincorporated AMPs that included transmembrane helical bundles, wormholes, carpets, as well as detergent-like effects

Jelinek and Kolusheva

or the in-plane diffusion of peptide-induced bilayer instabilities [14]. Oriented planar bilayers have been also used in other experimental approaches. CD experiments on alamethicin and magainin-1 in oriented lipid layers yielded interesting insights into transitions of the peptides from surface-bound into transmembrane structures [101]. Oriented lipid multibilayers were also employed in neutron reflectometry experiments. Neutron reflectometry indicated that magainin II significantly modified the thermal fluctuations of the lipid layers [102]. 5.3 Other Planar Lipid Models Various other planar models have been developed for studying AMP-membrane interactions. Black lipid membranes (BLMs) constitute of thin lipid films placed across small apertures separating two chambers containing ionic solutions (Fig. 6) [103]. The BLM technique relies upon conductance measurements to study the formation of pores and channels within the model membrane and their relative permeability. The use of BLMs also provides information on the ionic selectivity of the pores/channels formed by peptides. BLMs were instrumental in exposing the formation of permanent and transient channels by salmon and human calcitonins, host defense peptides that exhibit remarkable therapeutic capabilities [104]. The experiments also indicated that the peptide-induced channels were equally permeable to cations and anions. Tethered lipid bilayer membranes conceptually resemble BLMs as a planar biomimetic model system [105]. The artificial membrane in this arrangement is created by covalently binding a lipid monolayer onto gold electrodes, followed by deposition of another lipid film resulting in the formation of a tethered bilayer [106]. Incorporation of channel-forming AMPs into tethered bilayers has been demonstrated, and electrical impedance spectroscopy was applied for studying channel formation induced by the peptides and screening of blocking inhibitors [105]. A planar asymmetric bilayer designed for electrical application has been developed [107]. The assembly consists of one leaflet composed solely of LPS and the other layer containing phospholipids of bacterial plasma membranes. The goal of this design was to examine the role of LPS in targeting of antibacterial peptides to invading bacterial cells. The researchers mostly applied electrical measurements to study the association of the bactericidal/permeabilityincreasing protein (BPI) with the reconstituted bilayer, as a representative model for bacterial membrane interactions of common AMPs such as cathelicidins, defensins, and polymyxin B (PMB). One of the main conclusions drawn from these studies was that the negative charges of LPS are most likely the primary determinants for peptide-membrane interactions. However, the detailed mechanisms of bacterial killing may involve further steps, which could differ for specific AMPs. Hybrid bilayer systems have been the main experimental arrangement for studying AMP-membrane interactions using surface plasmon resonance (SPR), a technique that is increasingly applied for characterizing peptide binding and

Membrane Interactions of Host-defense Peptides Studied in Model Systems

Current Protein and Peptide Science, 2005, Vol. 6, No. 1

111

Fig. (6). Schematic picture of a black lipid membrane (BLM) system.

association with lipid bilayers [108]. Sample organization for SPR experiments usually consist of covalently-attached self-assembled hydrophobic films (in most instances alkanethiol monolayers), onto which lipid SUVs are adsorbed, putatively forming homogeneous monolayers [4]. The preformed lipid monolayers could also include receptor molecules and proteins [109]. SPR analysis applied to hybrid bilayers containing mucopeptides detected a correlation between the binding strengths of several glycopeptides and their antimicrobial activities [109]. A detailed SPR investigation explored the differences between pore formation and detergent-like membrane micellization (the “carpet model”) as the two membrane permeation mechanisms of cytotoxic peptides [110]. That study characterized and compared the binding profiles of native melittin on the one hand and diastereomeric melittin (Dmelittin) on the other hand. These two peptides represented the two models for membrane interactions with zwitterionic vs negatively-charged phospholipid assemblies, and indeed the SPR experiments demonstrated distinct behavior of the two peptides in the membrane models constructed. The dependence of binding affinity upon membrane composition of the hemolytic peptide lysenin was similarly uncovered by a kinetic SPR analysis [111]. SPR experiments have been carried out not only on phospholipid monolayer arrangements but also on supported lipid bilayers [110]. Such comparisons are important since they allow exploration of the affinity of AMPs to the outer membrane surface vs. the inner bilayer leaflet, since the lipid compositions of these monolayers is generally different in many cells. In particular, the quantitative ratio between the SPR-derived binding constants of specific peptides to the lipid bilayer and the monolayer, respectively, could give indications on the peptide's binding process and the depth of penetration into the membrane core [110].

5. Micelles Micellar assemblies are typically formed in aqueous solutions by organic co-solvents, lysophospholipids, and detergents below their critical micelle concentrations (CMC's). Among the most popular micelle-forming compounds have been trifluoroethanol (TFE), dodecylphosphocholine (DPC), and sodium dodecyl sulfate (SDS). Despite the fact that micelles constitute of single lipid or lipid-like layers rather than bilayers, and their surface curvatures are significantly higher than those encountered in real cellular membranes, micelles have been extensively used as biomimetic systems for membrane environments and for deciphering structural features of host-defense peptides. One of the advantages of micelles for studying peptidemembrane association has been the observation that such assemblies often stabilize a single peptide conformation, rather than an ensemble of structures [112]. Furthermore, rapid molecular tumbling of micelles in aqueous solutions has facilitated structural characterization of the incorporated peptides using important bio-analytical techniques such as high-resolution NMR, Fourier transform infrared spectroscopy (FTIR), and circular dichroism (CD). High-resolution solution NMR spectroscopy, in particular, has been applied to determine the conformations of AMPs in micelle environments [15,113]. NMR studies have generally suggested that peptide conformations induced through micelle binding resemble the physiological membrane-bound structures. Even though this claim is probably reliable only for simple secondary structural elements, in many cases the NMR structural information was important in providing information on the bactericidal and hemolytic activities of the peptides, aiding efforts for rational design of improved analogs for clinical use. Two-dimensional 1H NMR was applied to determine the structure of micelle-associated alamethicin [114], arginine-

112

Current Protein and Peptide Science, 2005, Vol. 6, No. 1

tryptophane sequence [25], and others. Circular dichroism (CD) spectroscopy was also applied for studying peptide conformations in micelles, in particular exposing the differences between micelle-bound vs. aqueous solutions [4]. In that regard, CD analyses have mostly confirmed the prominence of helical conformations of AMPs in micelle environments. CD experiments carried out in lysophospholipid micelles probed the effect of helix destabilization in a series of doubly D-residue labeled sequences [115]. 6. Miscellaneous Model Systems A variety of additional membrane-mimicking systems designed to elucidate structure-function relationships of hostdefense peptides was described in the literature. Similar to the techniques described above, the goals for introducing new model systems have been both to reconstitute lipid/protein assemblies that would mimic a cell membrane, as well as facilitating application of specific bio-analytical techniques for probing peptide-membrane interactions. A recent development, inspired partly by NMR studies of micelles, was the introduction of oriented phospholipid aggragates, or “bicelles”. These magnetically-oriented phospholipid/detergent micelles facilitate high-resolution structure determination of membrane-associated peptides and proteins using advanced NMR techniques [116]. Other model systems consisting of membrane lipids and proteins have been tailored for specific bio-analytical techniques. Poly(ethylene glycol)-lipid conjugates were found to promote bilayer formation in various lipid systems [117]. Reversed-phase high performance liquid chromatography (RP-HPLC) was applied to evaluate peptide conformations at hydrophobic-hydrophilic interfaces [4, 118, 119]. In such experiments, the hydrocarbon moieties within the retention column mimicked the lipid assemblies, inducing conformational changes of the peptides in the mobile phase [4,119]. RP-HPLC analyses could further probe in detail the relative affinities of membrane-active peptides to different phospholipid compositions [4]. The obvious goal of most studies utilizing model systems has been to best mimic the “real cellular world”. Accordingly, there has been an ongoing effort to evaluate AMP activities in whole cell systems. It could be argued that such analyses constitute less of “model systems” more fitting the description of microbiological assays, however several elegant studies thread nicely the line between the model and the actual biological assembly (see further the Review by A. Tossi in this volume). The laboratory of H.-G. Sahl, for example, has introduced several whole-cell approaches for determination of bacterial permeation kinetics and mechanisms of varied AMPs [120,121]. CONCLUSIONS Biomimetic membrane models have laid the groundwork for deciphering many biophysical and mechanistic aspects of the antibacterial properties of host defense peptides. Various proposed mechanisms for AMP action have originated from model system studies, including the “barrel stave” and “carpet” mechanisms, the significance of electrostatic attraction between the negative membrane-surface charge and cationic AMPs, the induction of transient holes and

Jelinek and Kolusheva

permanent pores in lipid bilayers by AMPs, and other models. Diverse bio-analytical, spectroscopic, and microscopic techniques were applied for studying host defense peptides in model membranes, aiming to elucidate the characteristics of peptide-mediated cell lysis. The experimental approaches have provided a wealth of information upon the extent and significance of pore formation, the contributions of headgroup interactions vs degree of peptide insertion into the membranes, the structure and orientation of AMPs in the membrane environment, and other factors believed to be central to the biological actions of AMPs. It is important to emphasize, however, that while in many cases important correlations between peptide-membrane interactions and peptides’ biological activities were detected in model systems, other works have unearthed the absence of such relationships – pointing to putative molecular targets of AMPs either on the membrane surface of the intact bacteria, or inside the bacterial cells. Furthermore, it has been often demonstrated that high peptide activity observed in model membranes is not necessarily predictive of its actual biological activity and vice versa. Overall, the availability of diverse biomimetic membrane models, each illuminating different aspects of peptidemembrane interactions, and the considerable amount of experimental data accumulated using artificial membrane models have provided us with a comprehensive insight into the properties and bactericidal activities of AMPs. The molecular and functional information acquired could further help design new therapeutic substances with improved antimicrobial properties and reduced toxicity. REFERENCES [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18]

Lehrer, R.I., Barton, A., Daher, K.A., Harwig, S.S., Ganz, T., Selsted, M.E. (1989) Journal of Clinical Investigation, 84, 553. Papo, N., Shai, Y. (2003) Peptides, 24, 1693. Sansom, M.S.P. (1998) Current Opinion in Colloid & Interface Science, 3, 518. Blondelle, S.E., Lohner, K., Aguilar, M.-I. (1999) Biochimica et Biophysica Acta, 1462, 89. Shai, Y. (1999) Biochimica et Biophysica Acta, 1462, 55. Dathe, M., Wieprecht, T. (1999) Biochimica et Biophysica Acta, 1462, 71. Tossi, A., Sandri, L. , iangaspero, A. (2000) Biopolymers, 55, 4. Shai, Y. (2002) Biopolymers, 66, 236. Hancock, R.E.W., Rozek, A. (2002) FEMS Microbiology Letters, 206, 143. Epand, R.M., Epand, R. F. (2003) Methods in Enzymology, 372, 124. Lee, D.K., Santos, J. S., Ramamoorthy, A. (1999) Journal of Physical Chemistry B, 103, 8383. Hori, Y., Demura, M., Niidome, T., Aoyagi, H., Asakura, T. ( 1999) FEBS Letters, 455. Picard, F., Paquet, M.-J., Dufourc, E. J., Auger, M. (1998) Biophysical Journal, 74, 857. Bechinger, B. (1999) Biochimica et Biophysica Acta, 1462, 157. Hwang, P.M., Vogel, H. J. (1998) Biochemistry and Cell Biology, 76, 235. Anderluh, G., Serra, M. D., Viero, G., Guella, G., Macek, P., Menestrina, G. (2003) Journal of Biological Chemistry, 278, 45216. Buffy, J.J., Waring, A. J., Lehrer, R. I., Hong, M. (2003) Biochemistry, 42, 13725. Buffy, J.J., Hong, T., Yamaguchi, S., Waring, A. J., Lehrer, R. I., Hong, M. (2003) Biophysical Journal, 85, 2363.

Membrane Interactions of Host-defense Peptides Studied in Model Systems [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36] [37] [38] [39] [40] [41] [42] [43] [44] [45] [46] [47] [48] [49] [50] [51] [52] [53] [54] [55]

Marcotte, I., Wegener, K. L., Lam, Y.-H., Chia, B. C. S., De Planque, M. R. R., Bowie, J. H., Auger, M., Separovic, F. (2003) Chemistry and Physics of Lipids, 122, 107. Bechinger, B. (1997) J. Membrane Biol., 156, 197. Wakamatsu, K., Takeda, A., Tachi, T., Matsuzaki, K. (2002) Biopolymers, 64, 314. Lohner, K., Staudegger, E., Prenner, E. J., Lewis, R. N. A. H., Kriechbaum, M., Degovics, G., McElhaney, R. N. (1999) Biochemistry, 38, 16514. Prenner, E.J., Lewis, R. N. A. H., KondeJewski, L. H., Hodges, R. S., McElhaney, R. N. (1999) Biochimica et Biophysica Acta , 1417, 211. Lohner, K., Prenner, E. J. (1999) Biochimica et Biophysica Acta, 1462, 141. Jing, W., Hunter, H. N., Hagel, J., Vogel, H. J. (2003) Journal of Peptide Research, 61, 219. Grasso, D., Milardi, D., La Rosa, C., Impellizzeri, G., Pappalardo, G. (2002) Thermochimica Acta, 390, 73. Jing, W., Strom, M. B., Svendsen, J. S., Vogel, H. J. (2002) Biochemistry and Cell Biology, 80, 161. Epand, R.F., Umezawa, N., Porter, E. A., Gellman, S.H., Epand, R.M. (2003) European Journal of Biochemistry, 270, 1240. Yang, L., Weiss, T. M., Lehrer, R. I., Huang, H. W. (2000) Biophysical Journal, 79, 2002. Lewis, R.N.A.H., Prenner, E.J., KondeJewski, L. H., Flach, C. R., Mendelsohn, R., Hodges, R. S., McElhaney, R. N. (1999) Biochemistry, 38, 15193. Mihailescu, D., Horvath, L.I. (1998) Eur. Biophys., 28, 216. Katsu, T., Kobayashi, H., FuJita, Y. (1986) Biochimica et Biophysica Acta, 860, 608. Prenner, E.J., Lewis, R. N.A.H., McElhaney, R. N. (1999) Biochimica et Biophysica Acta, 1462, 201. Susi, H., Sampugna, J., Hampson, J.W., Ard, J.S. (1979) Biochemistry, 18, 297. Staudegger, E., Prenner, E. J., Kriechbaum, M., Degovics, G., Lewis, R. N. A. H., McElhaney, R. N., Lohner, K. (2000) Biochimica et Biophysica Acta, 1468, 213. Prenner, E.J., Lewis, R. N. A. H., Jelokhani-Niaraki, M., Hodges, R. S., McElhaney, R. N. (2001) Biochimica et Biophysica Acta, 1510, 83. El Jastimi, R., Edwards, K., Lafleur, M. (1999) Biophysical Journal, 77, 842. Vanlerberghe, G. (1994) Oleagineux, Corps Gras, Lipides, 1, 174. De Kroon, A.I.P.M., de Gier, J., de KruiJff, B. (1993) New Comprehensive Biochemistry, 25, 107. Hope, M.J., Nayar, R., Mayer, L. D., Cullis, P. R. (1993) Liposome Technologies, 1, 123. Lichtenberg, D., Barenholz, Y. (1988) Methods of Biochemical Analysis, 33, 337. Kahya, N., Pecheur, E. I., de BoeiJ, W. P., Wiersma, D. A., Hoekstra, D. (2000) Biophysical Journal, 81, 1464. Matsuda, R., Kaneko, N., Horikawa, Y. (1997) Biochemical and biophysical research communications, 237, 499. Jelokhani-Niaraki, M., KondeJewski, L. H., Chen, Y., Hancock, R. E. W., Kay, C. M., Hodges, R. S. (1999) Peptide Science, 35, 81. Mangoni, M.L., Papo, N., Mignogna, G., Andreu, D., Shai, Y., Barra, D., Simmaco, M. (2003) Biochemistry, 42, 14023. Marsh, D. (2002) Biochem. Soc. Trans., 29, 582. Basanez, G., Shinnar, A. E., Zimmerberg, J. (2002) FEBS Letters, 532, 115. Blazyk, J., Wiegand, R., Klein, J., Hammer, J., Epand, R. M., Epand, R.l F., Maloy, W. L., Kari, U.P. (2001) Journal of Biological Chemistry, 276, 27899. Oren, Z., Shai, Y. (2000) Biochemistry, 39, 6103. Ladokhin, A.S., Wimley, W. C., Hristova, K., White, S. H. (1997) Methods in Enzymology, 278, 474. Mangoni, M.L., Rinaldi, A. C., Di Giulio, A., Mignogna, G., Bozzi, A., Barra, D., Simmaco, M. (2000) European Journal of Biochemistry, 267, 1447. Zhang, L., Rozek, A., Hancock, R. E. (2001) Journal of Biological Chemistry, 276, 349. Zhao, H., Kinnunen, P.K.J. (2002) J. Biol. Chem., 277, 25170. Matsuzaki, K., Fukui, M., FuJii, N., MiyaJima, K. (1991) Biochim. Biophy. Acta, 1070, 259. Yagi, Y., Kimura, S., Imanishi, Y. (1990) International Journal of Peptide and Protein Research, 36, 18.

Current Protein and Peptide Science, 2005, Vol. 6, No. 1 [56] [57] [58] [59] [60] [61] [62] [63] [64] [65] [66] [67] [68] [69] [70] [71] [72] [73] [74] [75] [76] [77] [78] [79] [80] [81] [82] [83] [84] [85] [86] [87] [88] [89] [90] [91] [92] [93] [94] [95]

113

Zhao, H., Rinaldi, A. C., Di Giulio, A., Simmaco, M., Kinnunen, P. K. J. (2002) Biochemistry, 41, 4425. Peschel, A. (2002) Trends in Microbiology, 10, 179. Gidalevitz, D., Ishitsuka, Y., Muresan, A. S., Konovalov, O., Waring, A. J. , Lehrer, R. I., Lee, K. Y. C. (2003) PNAS, 100, 6302. Matsuzaki, K., Sugishita, K.-i., Harada, M., FuJii, N., MiyaJima, K. (1997) Biochimica et Biophysica Acta, 1327, 119. Matsuzaki, K., Sugishita, K.-i., MiyaJima, K. (1999) FEBS Letters, 449, 221. Wimley, W.C., Selsted, M.E., White, S. H.. (1994) Protein Science, 3, 1362. Hristova, K., Selsted, M. E., White, S. H. (1997) Journal of Biological Chemistry, 272, 24224. Breukink, E., de KruiJff, B. (1999) Biochimica et Biophysica Acta , 1462, 223. van Heusden, H.E., de KruiJff, B., Breukink, E. (2002) Biochemistry, 41, 12171. Epand, R.F., Lehrer, R. I., Waring, A., Wang, W., Maget-Dana, R., Lelievre, D., Epand, R. M. (2003) Biopolymers, 71, 2. Zhao, H., Mattila, J.-P., Holopainen, J. M., Kinnunen, P. K. J. (2001) Biophysical Journal, 81, 2979. Ringsdorf, H., Schlarb, B., Venzmer, J. (1988) Angew. Chem., Int, 27, 113. Tanaka, H., Gomez, M.A., Tonelli, A.E., Thakur, M. (1989) Macromolecules, 22, 1208. Wenzel, M., Atkinson, G.H. (1989) J. Am. Chem. Soc., 111, 4123. Chance, R.R. (1980) Macromolecules, 13, 396. Kolusheva, S., Wachtel, E., Jelinek, R. (2003) Journal of Lipid Research, 44, 65. Kolusheva, S., Boyer, L., Jelinek, R. (2000) Nature Biotechnology, 18, 225. Kolusheva, S., Shahal, T., Jelinek, R. (2000) Biochemistry, 39, 15851. Katz, M., Tsubery, H., Kolusheva, S., Shames, A., Fridkin, M., Jelinek, R. (2003) Biochemical Journal, 375, 405. Rozner, S., Kolusheva, S., Cohen, Z., Dowhan, W., Eichler, J., Jelinek, R. (2003) Analytical Biochemistry, 319, 96. Satchell, D.P., Sheynis, T., ShirafuJi, Y., Kolusheva, S., Ouellette, A. J., Jelinek, R. (2003) Journal of Biological Chemistry, 278, 13838. Sheynis, T., Sykora, J, Benda, A., Kolusheva, S., Hof, M., Jelinek, R. (2003) European Journal of Biochemistry, 270, 4478. Tamm, L.K. (2002) Current Topics in Membranes, 52, 191. Fragneto-Cusani, G. (2001) Journal of Physics: Condensed Matter, 13, 4973. Gabrielli, G. (1991) Advances in Colloid and Interface Science, 34, 31. Baumgart, T., Offenhaeusser, A. (2003) Langmuir, 19, 1730. Frey, S., Tamm, L. K. (1991) Biophysical Journal, 60, 922. Zhang, L., Vidu, R., Waring, A. J., Lehrer, R. I., Longo, M. L., Stroeve, P. (2002) Langmuir, 18, 1318. Noinville, S., Bruston, F., El Amri, C., Baron, D., Nicolas, P. (2003) Biophysical Journal, 85, 1196. Losche, M. (2002) Current Topics in Membranes, 52, 117. Sinner, E.-K., Knoll, W. (2001) Current Opinion in Chemical Biology, 5, 705. Gabrielli, G., Caminati, G., Puggelli, M., Gilardoni, A., Margheri, E. (1993) Journal de Physique IV: Proceedings, 3, 279. Tien, H.T., Diana, A. Louise. (1968) Chemistry and Physics of Lipids, 2, 55. Greenhall, M.H., Yarwood, J., Brown, R., Swart, R. M. (1998) Langmuir, 14, 2619. Chen, H.M., Leung, K. W., Thakur, N. N., Tan, A., Jack, R. W. (2003) European Journal of Biochemistry, 270, 911. Thomas, C.J., Surolia, N., Surolia, A. (2001) Journal of Biological Chemistry, 276, 35701. Maget-Dana, R. (1999) Biochimica et Biophysica Acta, 1462, 109. Peggion, E., Mammi, S., Schievano, E. (1998) Biopolymers, 43, 419. Zhang, L., Scott, M. G., Yan, H., Mayer, L. D., Hancock, R. E. W. (2000) Biochemistry, 39, 14504. Konovalov, O., Myagkov, I., Struth, B., Lohner, K. (2002) Biophysics J., 31, 428.

114 [96] [97] [98] [99] [100] [101] [102] [103] [104] [105] [106] [107] [108]

Current Protein and Peptide Science, 2005, Vol. 6, No. 1 Axelsen, P.H., Braddock, W. D., Brockman, H. L., Jones, C. M., Dluhy, R. A., Kaufman, B. K., Puga, F. J. II. (1995) Applied Spectroscopy, 49, 526. Alexander, C., Rietschel, E. Th. (2001) Journal of Endotoxin Research, 7, 167. Yethon, J.A., Whitfield, C. (2001) Current Drug Targets: Infectious Disorders, 1, 91. Gutsmann, T., Hagge, S. O., Larrick, J. W., Seydel, U., Wiese, A. (2001) Biophysical Journal, 80, 2935. Marassi, F.M., Opella, S. J., Juvvadi, P., Merrifield, R. B. (1999) Biophysical Journal, 77, 3152. Ludtke, S.J., He, K., Heller, W. T., Harroun, T. A., Yang, L., Huang, H. W. (1996) Biochemistry, 35, 13723. Salditt, T., Muenster, C., Mennicke, U., Ollinger, C., Fragneto, G. (2003) Langmuir, 19, 7703. Winterhalter, M. (2000) Curr. Opin. Coll. & Interf. Sci., 5, 250. Stipani, V., Gallucci, E., Micelli, S., Picciarelli, V., Benz, R. (2001) Biophysical Journal, 81, 3332. Yin, P., Burns, C. J., Osman, P. D.J., Cornell, B. A. (2003) Biosensors and Bioelectronics, 18, 389. Stora, T., Lakey, J.H., Vogel, H. (1999) Angew. Chem. Intern., 38, 389. Wiese, A., Gutsmann, T., Seydel, U. (2003) Journal of endotoxin research, 9, 67. Mozsolits, H., Aguilar, M.-I. (2002) Biopolymers, 66, 3.

Jelinek and Kolusheva [109] [110] [111] [112] [113] [114] [115] [116] [117] [118] [119] [120] [121]

Cooper, M.A., Williams, D. H., Cho, Y. R. (1997) Chemical Communications, 17, 1625. Papo, N., Shai,Y. (2003) Biochemistry, 42, 458. YamaJi, A., Sekizawa, Y., Emoto, K., Sakuraba, H., Inoue, K., Kobayashi, H., Umeda, M. (1998) Journal of Biological Chemistry, 273, 5300. Ikura, T., Go, N., Inagaki, F. (1991) Proteins: Struct. Funct. Genet., 81. Henry, G.D., Sykes, B.D. (1992) Methods Enzymology, 239, 515. Franklin, J.C., Ellema, J.F., Jayasinghe, S., Kelsh, L.P., Cafiso, D.S. (1994) Biochemistry, 33, 4036. Maeda, M., Melnyk, R. A., Partridge, A. W., Liu, L.-P., Deber, C. M. (2003) Biopolymers, 71, 77. Sanders, C.R., Hare, B.J., Howard, K.P., Prestegard, J.H. (1994) Progress NMR Spectros., 26, 421. Holland, J.W., Cullis, P.R., Madden, T.D. (1996) Biochemistry, 35, 2610. Regnier, F.E. (1983) Analytical Chemistry, 55, 1298A. Blondelle, S.E., Forood, B., Perez-Paya, E., Houghten, R. A. (1996) International Journal of Bio-Chromotography, 2, 133. Ruhr, E. & Sahl, H.-G. (1985). Antimic. Agents and Chemotherapy, 27, 841. Pag, U., Oedenkoven, M., Papo, N., Oren, Z., Shai, Y. and Sahl, H.-G. (2004). Journal of Antimicrobial Chemotherapy, 53, 203.