Membrane binding and translocation of cell-penetrating peptides

Cell-penetrating peptides (CPPs) have been extensively studied during the past decade, because of their ability to promote the cellular uptake of various cargo molecules, e.g., oligonucleotides and proteins. In a recent study of the uptake of several analogues of penetratin, Tat(48-60) and oligoarginine in live (unfixed) cells [Thorén et al. (2003) Biochem. Biophys. Res. Commun. 307, 100-107], it was found that both endocytotic and nonendocytotic uptake pathways are involved in the internalization of these CPPs. In the present study, the membrane interactions of some of these novel peptides, all containing a tryptophan residue to facilitate spectroscopic studies, are investigated. The peptides exhibit a strong affinity for large unilamellar vesicles (LUVs) containing zwitterionic and anionic lipids, with binding constants decreasing in the order penetratin > R(7)W > TatP59W > TatLysP59W. Quenching studies using the aqueous quencher acrylamide and brominated lipids indicate that the tryptophan residues of the peptides are buried to a similar extent into the membrane, with an average insertion depth of approximately 10-11 A from the bilayer center. The membrane topology of the peptides was investigated using an assay based on resonance energy transfer between tryptophan and a fluorescently labeled lysophospholipid, lysoMC, distributed asymmetrically in the membranes of LUVs. By determination of the energy transfer efficiency when peptide was added to vesicles with lysoMC present exclusively in the inner leaflet, it was shown that none of the peptides investigated is able to translocate across the lipid membranes of LUVs. By contrast, confocal laser scanning microscopy studies on carboxyfluorescein-labeled peptides showed that all of the peptides rapidly traverse the membranes of giant unilamellar vesicles (GUVs). The choice of model system is thus crucial for the conclusions about the ability of CPPs to translocate across lipid membranes. Under the conditions used in the present study, peptide-lipid interactions alone cannot explain the different cellular uptake characteristics exhibited by these peptides.     

Comparison of the membrane association of two antimicrobial peptides, magainin 2 and indolicidin

Interactions of two antimicrobial peptides, magainin 2 and indolicidin, with three different model biomembranes, namely, monolayers, large unilamellar vesicles (LUVs), and giant liposomes, were studied. Insertion of both peptides into lipid monolayers was progressively enhanced when the content of an acidic phospholipid, 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphoglycerol (POPG) in a film of 1-stearoyl-2-oleoyl-sn-glycero-3-phosphocholine (SOPC) was increased. Indolicidin and magainin 2 penetrated also into lipid monolayers containing cholesterol (mole fraction, X = 0.1). Membrane association of magainin 2 attenuated lipid lateral diffusion in POPG-containing LUVs as revealed by the decrease in the excimer/monomer fluorescence ratio I(e)/I(m) for the pyrene fatty-acid-containing phospholipid derivative 1-palmitoyl-2-[10-(pyren-1-yl) decanoyl]-sn-glycero-3-phospho-rac-glycerol (PPDPG). Likewise, an increase in steady-state fluorescence anisotropy of the membrane-incorporated diphenylhexatriene (DPH) was observed, revealing magainin 2 to increase acyl chain order and induce segregation of acidic phospholipids. Similar effects were observed for indolicidin. The topological effects of magainin 2 and indolicidin on phospholipid membranes were investigated using optical microscopy of giant vesicles. Magainin 2 had essentially no influence on either SOPC or SOPC:cholesterol (X = 0.1) giant liposomes. However, effective vesiculation was observed when acidic phospholipid (X(PG) = 0.1) was included in the giant vesicles. Indolicidin caused only a minor shrinkage of giant SOPC vesicles whereas the formation of endocytotic vesicles was observed when the giant liposome contained POPG (X(PG) = 0.1). Interestingly, for indolicidin, vesiculation was also observed for giant vesicles composed of SOPC/cholesterol (X(chol) = 0.1). Possible mechanisms of membrane transformation induced by these two peptides are discussed.     

Bilayer interaction and localization of cell penetrating peptides with model membranes: a comparative study of a human calcitonin (hCT)-derived peptide with pVEC and pAntp(43-58)

Cell-penetrating peptides (CPPs) are able to translocate problematic therapeutic cargoes across cellular membranes. The exact mechanisms of translocation are still under investigation. However, evidence for endocytic uptake is increasing. We investigated the interactions of CPPs with phospholipid bilayers as first step of translocation. To this purpose, we employed four independent techniques, comprising (i) liposome buffer equilibrium dialysis, (ii) Trp fluorescence quenching, (iii) fluorescence polarization, and (iv) determination of zeta-potentials. Using unilamellar vesicles (LUVs) of different phospholipid composition, we compared weakly cationic human calcitonin (hCT)-derived peptides with the oligocationic CPPs pVEC and penetratin (pAntp). Apparent partition coefficients of hCT-derived peptides in neutral POPC LUVs were dependent on amino acid composition and secondary structure; partitioning in negatively charged POPC/POPG (80:20) LUVs was increased and mainly governed by electrostatic interactions. For hCT(9-32) and its derivatives, D values raised from about 100-200 in POPC to about 1000 to 1500 when negatively charged lipids were present. Localization profiles of CPPs obtained by Trp fluorescence quenching were dependent on the charge density of LUVs. In POPC/POPG, hCT-derived CPPs were located on the bilayer surface, whereas pVEC and pAntp resided deeper in the membrane. In POPG LUVs, an increase of fluorescence polarization was observed for pVEC and pAntp but not for hCT-derived peptides. Generally, we found strong peptide-phospholipid interactions, especially when negatively charged lipids were present.     

Bilayer interaction and localization of cell penetrating peptides with model membranes: A comparative study of a human calcitonin (hCT)-derived peptide with pVEC and pAntp(43-58)

Cell-penetrating peptides (CPPs) are able to translocate problematic therapeutic cargoes across cellular membranes. The exact mechanisms of translocation are still under investigation. However, evidence for endocytic uptake is increasing. We investigated the interactions of CPPs with phospholipid bilayers as first step of translocation. To this purpose, we employed four independent techniques, comprising (i) liposome buffer equilibrium dialysis, (ii) Trp fluorescence quenching, (iii) fluorescence polarization, and (iv) determination of ζ-potentials. Using unilamellar vesicles (LUVs) of different phospholipid composition, we compared weakly cationic human calcitonin (hCT)-derived peptides with the oligocationic CPPs pVEC and penetratin (pAntp). Apparent partition coefficients of hCT-derived peptides in neutral POPC LUVs were dependent on amino acid composition and secondary structure; partitioning in negatively charged POPC/POPG (80:20) LUVs was increased and mainly governed by electrostatic interactions. For hCT(9-32) and its derivatives, D values raised from about 100-200 in POPC to about 1000 to 1500 when negatively charged lipids were present. Localization profiles of CPPs obtained by Trp fluorescence quenching were dependent on the charge density of LUVs. In POPC/POPG, hCT-derived CPPs were located on the bilayer surface, whereas pVEC and pAntp resided deeper in the membrane. In POPG LUVs, an increase of fluorescence polarization was observed for pVEC and pAntp but not for hCT-derived peptides. Generally, we found strong peptide-phospholipid interactions, especially when negatively charged lipids were present.     

A membrane filtering method for the purification of giant unilamellar vesicles

The use of giant unilamellar vesicles (GUVs) for investigating the properties of biomembranes is advantageous compared to the use of small-sized vesicles such as large unilamellar vesicles (LUVs). Experimental methods using GUVs, such as the single GUV method, would benefit if there was a methodology for obtaining a large population of similar-sized GUVs composed of oil-free membranes. We here describe a new membrane filtering method for purifying GUVs prepared by the natural swelling method and demonstrate that, following purification of GUVs composed of dioleoylphosphatidylglycerol (DOPG)/dioleoylphosphatidylcholine (DOPC) membranes suspended in a buffer, similar-sized GUVs with diameters of 10–30 μm are obtained. Moreover, this method enabled GUVs to be separated from water-soluble fluorescent probes and LUVs. These results suggest that the membrane filtering method can be applied to GUVs prepared by other methods to purify larger-sized GUVs from smaller GUVs, LUVs, and various water-soluble substances such as proteins and fluorescent probes. This method can also be used for concentration of dilute GUV suspensions.     

CD Spectroscopy of Peptides and Proteins Bound to Large Unilamellar Vesicles

Circular dichroism (CD) spectroscopy is an essential tool for determining the conformation of proteins and peptides in membranes. It can be particularly useful for measuring the free energy of partitioning of peptides into lipid vesicles. The belief is broadly held that such CD measurements can only be made using sonicated small unilamellar vesicles (SUVs) because light scattering associated with extruded large unilamellar vesicles (LUVs) is unacceptably high. We have examined this issue using several experimental approaches in which a chiral object (i.e., peptide or protein) is placed both on the membrane and outside the membrane. We show that accurate CD spectra can be collected in the presence of LUVs. This is important because SUVs, unlike LUVs, are metastable and consequently unsuitable for equilibrium thermodynamic measurements. Our data reveal that undistorted CD spectra of peptides can be measured at wavelengths above 200 nm in the presence of up to 3 mM LUVs and above 215 nm in the presence of up to 7 mM LUVs. We introduce a simple way of characterizing the effect on CD spectra of light scattering and absorption arising from suspensions of vesicles of any diameter. Using melittin as an example, we show that CD spectroscopy can be used to determine the fractional helical content of peptides in LUVs and to measure their free energy of partitioning of into LUVs.     

Membrane perturbation effects of peptides derived from the N-termini of unprocessed prion proteins

Peptides derived from the unprocessed N-termini of mouse and bovine prion proteins (mPrPp and bPrPp, respectively), comprising hydrophobic signal sequences followed by charged domains (KKRPKP), function as cell-penetrating peptides (CPPs) with live cells, concomitantly causing toxicity. Using steady-state fluorescence techniques, including calcein leakage and polarization of a membrane probe (diphenylhexatriene, DPH), as well as circular dichroism, we studied the membrane interactions of the peptides with large unilamellar phospholipid vesicles (LUVs), generally with a 30% negative surface charged density, comparing the effects with those of the CPP penetratin (pAntp) and the pore-forming peptide melittin. The prion peptides caused significant calcein leakage from LUVs concomitant with increased membrane ordering. Fluorescence correlation spectroscopy (FCS) studies of either rhodamine-entrapping (REVs) or rhodamine-labeled (RLVs) vesicles, showed that addition of the prion peptides resulted in significant release of rhodamine from the REVs without affecting the overall integrity of the RLVs. The membrane leakage effects due to the peptides had the following order of potency: melittin > mPrPp > bPrPp > pAntp. The membrane perturbation effects of the N-terminal prion peptides suggest that they form transient pores (similar to melittin) causing toxicity in parallel with their cellular trafficking.     

Real-time transmembrane translocation of penetratin driven by light-generated proton pumping

Cell penetrating peptides (CPPs) are small peptides that are able to penetrate the plasma membrane of mammalian cells. Because these peptides can also carry large hydrophilic cargos such as proteins, they could potentially be used to transport biologically active drugs across cell membranes to modulate in vivo biology. One characteristic feature of the CPPs is that they typically have a net positive charge. Therefore, a key issue associated with the transport mechanism is the role of the transmembrane electrochemical potential in driving the peptides across the membrane. In this study, we have reconstituted bacteriorhodopsin (bR) in large unilamellar vesicles (LUVs) with fluorescein-labeled CPP penetratin enclosed within the LUVs under conditions when the fluorescence is quenched. Illumination of the bacteriorhodopsin-containing LUVs resulted in creation of a transmembrane proton electrochemical gradient (positive on the inside). Upon generation of this gradient, an increase in fluorescence was observed, which shows that the proton gradient drives the translocation of penetratin. The mechanism most likely can be generalized to other CPPs.     

Measuring peptide translocation into large unilamellar vesicles

There is an active interest in peptides that readily cross cell membranes without the assistance of cell membrane receptors(1). Many of these are referred to as cell-penetrating peptides, which are frequently noted for their potential as drug delivery vectors(1-3). Moreover, there is increasing interest in antimicrobial peptides that operate via non-membrane lytic mechanisms(4,5), particularly those that cross bacterial membranes without causing cell lysis and kill cells by interfering with intracellular processes(6,7). In fact, authors have increasingly pointed out the relationship between cell-penetrating and antimicrobial peptides(1,8). A firm understanding of the process of membrane translocation and the relationship between peptide structure and its ability to translocate requires effective, reproducible assays for translocation. Several groups have proposed methods to measure translocation into large unilamellar lipid vesicles (LUVs)(9-13). LUVs serve as useful models for bacterial and eukaryotic cell membranes and are frequently used in peptide fluorescent studies(14,15). Here, we describe our application of the method first developed by Matsuzaki and co-workers to consider antimicrobial peptides, such as magainin and buforin II(16,17). In addition to providing our protocol for this method, we also present a straightforward approach to data analysis that quantifies translocation ability using this assay. The advantages of this translocation assay compared to others are that it has the potential to provide information about the rate of membrane translocation and does not require the addition of a fluorescent label, which can alter peptide properties(18), to tryptophan-containing peptides. Briefly, translocation ability into lipid vesicles is measured as a function of the Foster Resonance Energy Transfer (FRET) between native tryptophan residues and dansyl phosphatidylethanolamine when proteins are associated with the external LUV membrane (Figure 1). Cell-penetrating peptides are cleaved as they encounter uninhibited trypsin encapsulated with the LUVs, leading to disassociation from the LUV membrane and a drop in FRET signal. The drop in FRET signal observed for a translocating peptide is significantly greater than that observed for the same peptide when the LUVs contain both trypsin and trypsin inhibitor, or when a peptide that does not spontaneously cross lipid membranes is exposed to trypsin-containing LUVs. This change in fluorescence provides a direct quantification of peptide translocation over time.     

Laurdan and di-4-ANEPPDHQ do not respond to membrane-inserted peptides and are good probes for lipid packing

Laurdan and di-4-ANEPPDHQ are used as probes for membrane order, with a blue shift in emission for membranes in liquid-ordered (lo) phase relative to membranes in liquid-disordered (ld) phase. Their use as membrane order probes requires that their spectral shifts are unaffected by membrane proteins, which we have examined by using membrane inserting peptides and large unilamellar vesicles (LUVs). The transmembrane polypeptides, mastoparan and bovine prion protein-derived peptide (bPrPp), were added to LUVs of either lo or ld phase, up to 1:10 peptide/total lipid ratio. The excitation and emission spectra of laurdan and di-4-ANEPPDHQ in both lipid phases were unaltered by peptide addition. The integrity and size distribution of the LUVs upon addition of the polypeptides were determined by dynamic light scattering. The insertion efficiency of the polypeptides into LUVs was determined by measuring their secondary structure by circular dichroism. Mastoparan had an α-helical and bPrPp a β-strand conformation compatible with insertion into the lipid bilayer. Our results suggest that the presence of proteins in biological membranes does not influence the spectra of laurdan and di-4-ANEPPDHQ, supporting that the dyes are appropriate probes for assessing lipid order in cells.     

Different membrane behaviour and cellular uptake of three basic arginine-rich peptides

Cell penetrating peptides (CPPs) are peptides displaying the ability to cross cell membranes and transport cargo molecules inside cells. Several uptake mechanisms (endocytic or direct translocation through the membrane) are being considered, but the interaction between the CPP and the cell membrane is certainly a preliminary key point to the entry of the peptide into the cell. In this study, we used three basic peptides: RL9 (RRLLRRLRR-NH(2)), RW9 (RRWWRRWRR-NH(2)) and R9 (RRRRRRRRR-NH(2)). While RW9 and R9 were internalised into wild type Chinese Hamster Ovary cells (CHO) and glycosaminoglycan-deficient CHO cells, at 4°C and 37°C, RL9 was not internalised into CHO cells. To better understand the differences between RW9, R9 and RL9 in terms of uptake, we studied the interaction of these peptides with model lipid membranes. The effect of the three peptides on the thermotropic phase behaviour of a zwitterionic lipid (DMPC) and an anionic lipid (DMPG) was investigated with differential scanning calorimetry (DSC). The presence of negative charges on the lipid headgroups appeared to be essential to trigger the peptide/lipid interaction. RW9 and R9 disturbed the main phase transition of DMPG, whereas RL9 did not induce significant effects. Isothermal titration calorimetry (ITC) allowed us to study the binding of these peptides to large unilamellar vesicles (LUVs). RW9 and R9 proved to have about ten fold more affinity for DSPG LUVs than RL9. With circular dichroism (CD) and NMR spectroscopy, the secondary structure of RL9, RW9 and R9 in aqueous buffer or lipid/detergent conditions was investigated. Additionally, we tested the antimicrobial activity of these peptides against Escherichia coli and Staphylococcus aureus, as CPPs and antimicrobial peptides are known to share several common characteristics. Only RW9 was found to be mildly bacteriostatic against E. coli. These studies helped us to get a better understanding as to why R9 and RW9 are able to cross the cell membrane while RL9 remains bound to the surface without entering the cell.     

Modeling the endosomal escape of cell-penetrating peptides: transmembrane pH gradient driven translocation across phospholipid bilayers

Cell-penetrating peptides (CPPs) are able to mediate the efficient cellular uptake of a wide range of cargoes. Internalization of a number of CPPs requires uptake by endocytosis, initiated by binding to anionic cell surface heparan sulfate (HS), followed by escape from endosomes. To elucidate the endosomal escape mechanism, we have modeled the process for two CPPs: penetratin (pAntp) and the N-terminal signal peptide of the unprocessed bovine prion protein (bPrPp). Large unilamellar phospholipid vesicles (LUVs) were produced encapsulating either peptide, and an ionophore, nigericin, was used to create a transmembrane pH gradient (DeltapH(mem), inside acidic) similar to the one arising in endosomes in vivo. In the absence of DeltapH(mem), no pAntp escape from the LUVs is observed, while a fraction of bPrPp escapes. In the presence of DeltapH(mem), a significant amount of pAntp escapes and an even higher degree of bPrPp escape takes place. These results, together with the differences in kinetics of escape, indicate different escape mechanisms for the two peptides. A minimum threshold peptide concentration exists for the escape of both peptides. Coupling of the peptides to a cargo reduces the fraction escaping, while complexation with HS significantly hinders the escape. Fluorescence correlation spectroscopy results show that during the escape process the LUVs are intact. Taken together, these results suggest a model for endosomal escape of CPPs: DeltapH(mem)-mediated mechanism, following dissociation from HS of the peptides, above a minimum threshold peptide concentration, in a process that does not involve lysis of the vesicles.     

The Antimicrobial Domains of Wheat Puroindolines Are Cell-Penetrating Peptides with Possible Intracellular Mechanisms of Action

The puroindoline proteins (PINA and PINB) of wheat display lipid-binding properties which affect the grain texture, a critical parameter for wheat quality. Interestingly, the same proteins also display antibacterial and antifungal properties, attributed mainly to their Tryptophan-rich domain (TRD). Synthetic peptides based on this domain also display selectivity towards bacterial and fungal cells and do not cause haemolysis of mammalian cells. However, the mechanisms of these activities are unclear, thus limiting our understanding of the in vivo roles of PINs and development of novel applications. This study investigated the mechanisms of antimicrobial activities of synthetic peptides based on the TRD of the PINA and PINB proteins. Calcein dye leakage tests and transmission electron microscopy showed that the peptides PuroA, Pina-M and Pina-W→F selectively permeabilised the large unilamellar vesicles (LUVs) made with negatively charged phospholipids mimicking bacterial membranes, but were ineffective against LUVs made with zwitterionic phospholipids mimicking eukaryotic membranes. Propidium iodide fluorescence tests of yeast (Saccharomyces cerevisiae) cells showed the peptides were able to cause loss of membrane integrity, PuroA and Pina-M being more efficient. Scanning electron micrographs of PINA-based peptide treated yeast cells showed the formation of pits or pores in cell membranes and release of cellular contents. Gel retardation assays indicated the peptides were able to bind to DNA in vitro, and the induction of filamental growth of E. coli cells indicated in vivo inhibition of DNA synthesis. Together, the results strongly suggest that the PIN-based peptides exert their antimicrobial effects by pore formation in the cell membrane, likely by a carpet-like mechanism, followed by intracellular mechanisms of activity.     

New insight into the interaction of TRAF2 C-terminal domain with lipid raft microdomains

In this study we provide the first evidence of the interaction of a truncated-TRAF2 with lipid raft microdomains. We have analyzed this interaction by measuring the diffusion coefficient of the protein in large and giant unilamellar vesicles (LUVs and GUVs, respectively) obtained both from synthetic lipid mixtures and from natural extracts. Steady-state fluorescence measurements performed with synthetic vesicles indicate that this truncated form of TRAF2 displays a tighter binding to raft-like LUVs with respect to the control (POPC-containing LUVs), and that this process depends on the protein oligomeric state. Generalized Polarization measurements and spectral phasor analysis revealed that truncated-TRAF2 affects the membrane fluidity, especially when vesicles are heated up at physiological temperature. The addition of nanomolar concentration of TRAF2 in GUVs also seems to exert a mechanical action, as demonstrated by the formation of intraluminal vesicles, a process in which ganglioside GM1 plays a crucial role.     

Influence of molecular weight and pH on adsorption of chitosan at the surface of large and giant vesicles

This paper describes the mechanisms of adsorption of chitosan, a positively charged polyelectrolyte, on the DOPC lipid membrane of large and giant unilamellar vesicles (respectively, LUVs and GUVs). We observe that the variation of the zeta potential of LUVs as a function of chitosan concentration is independent on the chitosan molecular weight (Mw). This result is interpreted in terms of electrostatic interactions, which induce a flat adsorption of the chitosan on the surface of the membrane. The role of electrostatic interactions is further studied by observing the variation of the zeta potential as a function of the chitosan concentration for two different charge densities tuned by the pH. Results show a stronger chitosan-membrane affinity at pH 6 (lipids are negatively charged, and 40% chitosan amino groups are protonated) than at pH 3.4 (100% of protonated amino groups but zwitterionic lipids are positively charged) which confirms that adsorption is of electrostatic origin. Then, we investigate the stability of decorated LUVs and GUVs in a large range of pH (6.0 < pH < 12.0) in order to complete a previous study made in acidic conditions [Quemeneur et al. Biomacromolecules 2007, 8, 2512-2519]. A comparative study of the variation of the zeta potential as a function of the pH (2.0 < pH < 12.0) reveals a difference in behavior between naked and chitosan-decorated LUVs. This result is further confirmed by a comparative observation by optical microscopy of naked and chitosan-decorated GUVs in basic conditions (6.0 < pH < 12.0): at pH > 10.0, in the absence of chitosan, the vesicles present complex shapes, contrary to the chitosan-decorated vesicles which remain spherical, confirming thus that chitosan remains adsorbed on vesicles in basic conditions up to pH = 12.0. These results, in addition with our previous data, show that the chitosan-decorated vesicles are stable over a very broad range of pH (2.0 < pH < 12.0), which holds promise for their in vivo applications. Finally, the quantification of the chitosan adsorption on a LUV membrane is performed by zeta potential and fluorescence measurements. The fraction of membrane surface covered by chitosan is estimated to be lower than 40 %, which corresponds to the formation of a flat layer of chitosan on the membrane surface on an electrostatic basis.     

Cationic poly(ethyleneglycol) lipids incorporated into pre-formed vesicles enhance binding and uptake to BHK cells

This paper describes a new method for enhancing the interaction of liposomes with cells. A novel class of cationic poly(ethyleneglycol) (PEG)-lipid (CPL) conjugates have been characterized for their ability to insert into pre-formed vesicles and enhance in vitro cellular binding and uptake of neutral and sterically-stabilized liposomes. The CPLs, which consist of a distearoylphosphatidylethanolamine (DSPE) anchor, a fluorescent dansyl moiety, a heterobifunctional PEG polymer (M(r) 3400), and a cationic headgroup composed of lysine derivatives, have been described previously [Bioconjug. Chem. 11 (2000) 433]. Five separate CPL, possessing 1-4 positive charges in the headgroup (referred to as CPL(1)-CPL(4), respectively), were incubated (as micellar solutions) in the presence of neutral or sterically-stabilized cationic large unilamellar vesicles (LUVs), and were found to insert into the external leaflet of the LUVs in a manner dependent on temperature, time, CPL/lipid ratio, and LUV composition. For CPL/lipid molar ratios < or =0.1, optimal insertion levels of approximately 70% of initial CPL were obtained following 3 h at 60 degrees C. The insertion of CPL resulted in aggregation of the LUVs, as assessed by fluorescence microscopy, which could be prevented by the presence of 40 mM Ca(2+). The effect of CPL-insertion on the binding of LUVs to cells was examined by fluorescence microscopy and quantified by measuring the ratio of rhodamine fluorescence to protein concentration. Neither control LUVs or LUVs containing CPL(2) displayed significant uptake by BHK cells. However, a 3-fold increase in binding was observed for LUVs possessing CPL(3), while for CPL(4)-LUVs values as high as 10-fold were achieved. Interestingly, the increase in lipid uptake did not correlate with total surface charge, but rather with increased positive charge density localized at the CPL distal headgroups. These results suggest that incorporation of CPLs into existing liposomal drug delivery systems may lead to significant improvements in intracellular delivery of therapeutic agents.     

Thermodynamics of the interactions of tryptophan-rich cathelicidin antimicrobial peptides with model and natural membranes

Tritrpticin and indolicidin are short 13-residue tryptophan-rich antimicrobial peptides that hold potential as future alternatives for antibiotics. Isothermal titration calorimetry (ITC) has been applied as the main tool in this study to investigate the thermodynamics of the interaction of these two cathelicidin peptides as well as five tritrpticin analogs with large unilamellar vesicles (LUVs), representing model and natural anionic membranes. The anionic LUVs were composed of (a) 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphoethanolamine/1-palmitoyl-2-oleoyl-sn-glycero-3-phosphoglycerol (POPE/POPG) (7:3) and (b) natural E. coli polar lipid extract. 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine (POPC) was used to make model zwitterionic membranes. Binding isotherms were obtained to characterize the antimicrobial peptide binding to the LUVs, which then allowed for calculation of the thermodynamic parameters of the interaction. All peptides exhibited substantially stronger binding to anionic POPE/POPG and E. coli membrane systems than to the zwitterionic POPC system due to strong electrostatic attractions between the highly positively charged peptides and the negatively charged membrane surface, and results with tritrpticin derivatives further revealed the effects of various amino acid substitutions on membrane binding. No significant improvement was observed upon increasing the Tritrp peptide charge from + 4 to + 5. Replacement of Arg residues with Lys did not substantially change peptide binding to anionic vesicles but moderately decreased the binding to zwitterionic LUVs. Pro to Ala substitutions in tritrpticin, allowing the peptide to adopt an α-helical structure, resulted in a significant increase of the binding to both anionic and zwitterionic vesicles and therefore reduced the selectivity for bacterial and mammalian membranes. In contrast, substitution of Trp with other aromatic amino acids significantly decreased the peptide's ability to bind to anionic LUVs and essentially eliminated binding to zwitterionic LUVs. The ITC results were consistent with the outcome of fluorescence spectroscopy membrane binding and perturbation studies. Overall, our work showed that a natural E. coli polar lipid extract as a bacterial membrane model was advantageous compared to the simpler and more widely used POPE/POPG lipid system.     

Vesicle membrane interactions of penetratin analogues

Reports on serious artifacts associated with the use of cell fixation in studies of the cellular uptake of cell-penetrating peptides, also denoted protein transduction domains, have demonstrated the need for a reevaluation of the current understanding of peptide-mediated cellular delivery of large, hydrophilic molecules. In a recent study on the internalization in unfixed cells of penetratin and its analogues in which tryptophans are substituted for phenylalanines (Pen2W2F), lysines for arginines (PenArg), and arginines for lysines (PenLys), we revealed large dissimilarities in cell interactions among the peptides [Thorén et al. (2003) Biochem. Biophys. Res. Commun. 307, 100-107]. We here investigated possible correlations with their respective affinities for the lipid membranes of large unilamellar vesicles. The variations found in membrane affinity correlated qualitatively with differences in hydrophobicity among the peptides but were by far too small to account for the striking differences in cell membrane binding. Interestingly, we found that the inclusion of a small fraction of lipids conjugated to poly(ethylene glycol) (PEG) in the vesicles both stabilized the vesicle dispersion against peptide-induced aggregation and, furthermore, enhanced the binding of the peptides to the membrane. By use of PEG-conjugated lipids, it could be shown that vesicle aggregation drives an alpha-helix to beta-sheet conformational transition for these peptides. A similar transition was discovered at submicellar concentrations of sodium dodecyl sulfate in aqueous solution for all peptides except PenLys. Finally, significant changes of the contributions to CD spectra from aromatic residues due to their insertion into the membrane were observed.     

Biophysical properties of membrane lipids of anammox bacteria: I. Ladderane phospholipids form highly organized fluid membranes

Anammox bacteria that are capable of anaerobically oxidizing ammonium (anammox) with nitrite to nitrogen gas produce unique membrane phospholipids that comprise hydrocarbon chains with three or five linearly condensed cyclobutane rings. To gain insight into the biophysical properties of these ‘ladderane’ lipids, we have isolated a ladderane phosphatidylcholine and a mixed ladderane phosphatidylethanolamine/phosphatidylglycerol lipid fraction and reconstituted these lipids in different membrane environments. Langmuir monolayer experiments demonstrated that the purified ladderane phospholipids form fluid films with a relatively high lipid packing density. Fluid-like behavior was also observed for ladderane lipids in bilayer systems as monitored by cryo-electron microscopy on large unilamellar vesicles (LUVs) and epi-fluorescence microscopy on giant unilamellar vesicles (GUVs). Analysis of the LUVs by fluorescence depolarization revealed a relatively high acyl chain ordering in the hydrophobic region of the ladderane phospholipids. Micropipette aspiration experiments were applied to study the mechanical properties of ladderane containing lipid bilayers and showed a relatively high apparent area compressibility modulus for ladderane containing GUVs, thereby confirming the fluid and acyl chain ordered characteristics of these lipids. The biophysical findings in this study support the previous postulation that dense membranes in anammox cells protect these microbes against the highly toxic and volatile anammox metabolites.     

Thermodynamics of the interactions of tryptophan-rich cathelicidin antimicrobial peptides with model and natural membranes

Tritrpticin and indolicidin are short 13-residue tryptophan-rich antimicrobial peptides that hold potential as future alternatives for antibiotics. Isothermal titration calorimetry (ITC) has been applied as the main tool in this study to investigate the thermodynamics of the interaction of these two cathelicidin peptides as well as five tritrpticin analogs with large unilamellar vesicles (LUVs), representing model and natural anionic membranes. The anionic LUVs were composed of (a) 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphoethanolamine/1-palmitoyl-2-oleoyl-sn-glycero-3-phosphoglycerol (POPE/POPG) (7:3) and (b) natural E. coli polar lipid extract. 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine (POPC) was used to make model zwitterionic membranes. Binding isotherms were obtained to characterize the antimicrobial peptide binding to the LUVs, which then allowed for calculation of the thermodynamic parameters of the interaction. All peptides exhibited substantially stronger binding to anionic POPE/POPG and E. coli membrane systems than to the zwitterionic POPC system due to strong electrostatic attractions between the highly positively charged peptides and the negatively charged membrane surface, and results with tritrpticin derivatives further revealed the effects of various amino acid substitutions on membrane binding. No significant improvement was observed upon increasing the Tritrp peptide charge from +4 to +5. Replacement of Arg residues with Lys did not substantially change peptide binding to anionic vesicles but moderately decreased the binding to zwitterionic LUVs. Pro to Ala substitutions in tritrpticin, allowing the peptide to adopt an alpha-helical structure, resulted in a significant increase of the binding to both anionic and zwitterionic vesicles and therefore reduced the selectivity for bacterial and mammalian membranes. In contrast, substitution of Trp with other aromatic amino acids significantly decreased the peptide's ability to bind to anionic LUVs and essentially eliminated binding to zwitterionic LUVs. The ITC results were consistent with the outcome of fluorescence spectroscopy membrane binding and perturbation studies. Overall, our work showed that a natural E. coli polar lipid extract as a bacterial membrane model was advantageous compared to the simpler and more widely used POPE/POPG lipid system.