Self-association process of a peptide in solution: from beta-sheet filaments to large embedded nanotubes.

Lanreotide is a synthetic octapeptide used in the therapy against acromegaly. When mixed with pure water at 10% (w/w), Lanreotide (acetate salt) forms liquid crystalline and monodisperse nanotubes with a radius of 120 A. The molecular and supramolecular organization of these structures has been determined in a previous work as relying on the lateral association of 26 beta-sheet filaments made of peptide noncovalent dimers, the basic building blocks. The work presented here has been devoted to the corresponding self-association mechanisms, through the characterization of the Lanreotide structures formed in water, as a function of peptide (acetate salt) concentration (from 2% to 70% (w/w)) and temperature (from 15 degrees C to 70 degrees C). The corresponding states of water were also identified and quantified from the thermal behavior of water in the Lanreotide mixtures. At room temperature and below 3% (w/w) Lanreotide acetate in water, soluble aggregates were detected. From 3% to 20% (w/w) long individual and monodisperse nanotubes crystallized in a hexagonal lattice were evidenced. Their molecular and supramolecular organizations are identical to the ones characterized for the 10% (w/w) sample. Heating induces the dissolution of the nanotubes into soluble aggregates of the same structural characteristics as the room temperature ones. The solubilization temperature increases from 20 degrees C to 70 degrees C with the peptide concentration and reaches a plateau between 15% and 25% (w/w) in peptide. These aggregates are proposed to be the beta-sheet filaments that self-associate to build the walls of the nanotubes. Above 20% (w/w) of Lanreotide acetate in water, polydisperse embedded nanotubes are formed and the hexagonal lattice is lost. These embedded nanotubes exhibit the same molecular and supramolecular organizations as the individual monodisperse nanotubes formed at lower peptide concentration. The embedded nanotubes do not melt in the range of temperature studied indicating a higher thermodynamic stability than individual nanotubes. In parallel, the thermal behaviors of water in mixtures containing 2-80% (w/w) in peptide have been studied by differential scanning calorimetry, and three different types of water were characterized: 1), bulk water melting at 0 degrees C, 2), nonfreezing water, and 3), interfacial water melting below 0 degrees C. The domains of existence and coexistence of these different water states are related to the different Lanreotide supramolecular structures. All these results were compiled into a binary Lanreotide-water phase diagram and allowed to propose a self-association mechanism of Lanreotide filaments into monodisperse individual nanotubes and embedded nanotubes.

Origin of laurdan sensitivity to the vesicle-to-micelle transition of phospholipid-octylglucoside system: a time-resolved fluorescence study.

The fluorescent probe laurdan has been shown to be sensitive to the vesicle-to-micelle transition of phosphatidylcholine/octylglucoside (M. Paternostre, O. Meyer, C. Grabielle-Madelmont, S. Lesieur, and, Biophys. J. 69:2476-2488). On the other hand, a study on the photophysics of laurdan in organic solvents has shown that the complex de-excitation pathway of the probe can be described by two successive processes, i.e., an intramolecular charge transfer followed by dielectric relaxation of the solvent if polar. These two excited-state reactions lead to three emitting states, i.e., a locally excited state, a charge transfer state, and a solvent relaxed state (M. Viard, J. Gallay, M. Vincent, B. Robert and, Biophys. J. 73:2221-2234). Experiments have been performed using time-resolved fluorescence on the probe inserted in amphiphile aggregates (mixed liposomes, mixed micelles) different in detergent-to-lipid ratios. The results have been compared with those obtained for laurdan inserted in dipalmitoyl phosphatidylcholine liposomes in the gel and in the fluid lamellar phase. Except for laurdan in dipalmitoyl phosphatidylcholine liposomes in the gel lamellar phase, the red part of the emission spectra originates from the de-excitation of the relaxed excited state of laurdan, indicating that indeed the dielectric relaxation process is an important phenomena in the ground-state return pathway of this probe. On the other hand, the maximization entropy method (MEM) analysis of the fluorescence decay recorded in the blue part of the emission spectra indicates that the dielectric relaxation is not the only reaction occurring to the excited state of laurdan. Moreover, the analysis of the fluorescence decays of laurdan inserted in gel lamellar dipalmitoylphosphatidylcholine (DPPC) liposomes indicates excited-state reactions, although dielectric relaxation is impossible. These results are in agreement with the de-excitation pathway determined from laurdan behavior in organic solvent even if, in most of the aggregates studied in this work, the major phenomenon is the dielectric relaxation of the solvent. All along the vesicle-to-micelle transition, we have observed that the lifetime of the relaxed excited state of laurdan continuously decreases probably due to a dynamic quenching process by water molecules. On the other hand, the time constant of the dielectric relaxation process remains almost unchanged in the lamellar part of the transition but abruptly decreases as soon as the first mixed micelle is formed. This decrease is continuous all over the rest of the transition even if it is more pronounced in the mixed liposomes' and mixed micelles' coexistence. The increase of the octylglucoside-to-lipid ratio of the mixed micelles via the change of the size and the shape of the aggregates may facilitate the penetration and the mobility of water molecules. Therefore, during the vesicle-to-micelle transition, laurdan probes the evolution of both the amphiphile packing in the aggregates and the increase of the interface polarity. This study finally shows that the detergent-to-lipid ratio of the mixed micelles is an important parameter to control to limit the penetration and the mobility of water within the amphiphile aggregates and that laurdan is a nice tool to monitor this phenomenon.

Vesicle reconstitution from lipid-detergent mixed micelles.

The process of formation of lipid vesicles using the technique of detergent removal from mixed-micelles is examined. Recent studies on the solubilization and reconstitution of liposomes participated to our knowledge of the structure and properties of mixed lipid-detergent systems. The mechanisms involved in both the lipid self assembly and the micelle-vesicle transition are first reviewed. The simplistic three step minimum scheme is described and criticized in relation with isothermal as well as a function of the [det]/[lip] ratio, phase diagram explorations. The techniques of detergent elimination are reviewed and criticized for advantages and disadvantages. New methods inducing micelle-vesicle transition using enzymatic reaction and T-jump are also described and compared to more classical ones. Future developments of these techniques and improvements resulting of their combinations are also considered. Proper reconstitution of membrane constituents such as proteins and drugs into liposomes are examined in the light of our actual understanding of the micelle-vesicle transition.

Liposomal formulations for oral immunotherapy: in-vitro stability in synthetic intestinal media and in-vivo efficacy in the mouse.

The aim of this work was to develop a liposomal formulation which could act as a carrier for allergens during oral desensitization therapy. A model protein, ovalbumin, was associated with negatively charged, multilamellar vesicles of various compositions and their stability in the presence of synthetic intestinal media (bile salt, pancreatic enzymes and their combination) was investigated. Liposomes containing soya phosphatidylcholine as the main lipid, regardless of their cholesterol content (20-40%), were unable to protect ovalbumin against the combined action of pancreatic enzymes and bile salt. In contrast, liposomes prepared from distearoylphosphatidylcholine and cholesterol (6:3.5 molar ratio) were more stable: about 50% of the lipid remained as liposomes after a 4-h incubation at 37 degrees C and intact ovalbumin could be demonstrated therein by immunoblotting. The immunomodulating properties of liposomes were tested by following changes in serum IgE levels (by passive cutaneous anaphylaxis) in Balb/C mice sensitized to ovalbumin, after feeding various preparations. In this model, free ovalbumin was able to provoke a premature fall in IgE levels, and liposomes, whatever their composition, contributed no further effect.

A comparative study of the action of melittin on sphingomyelin and phosphatidylcholine bilayers.

To investigate whether lipid solubilization is of relevance in describing the interaction between melittin and biological membranes, we studied melittin-induced polymorphism using model membranes composed of the biological lipid sphingomyelin (bovine brain). The behavior of the system was monitored by solid state 31P-NMR and turbidity measurements and compared to the peptides well-characterized action on the synthetic lipid dipalmitoylphosphatidylcholine. It was found that melittin-induced macroscopic changes of sphingomyelin membranes are qualitatively the same as in the case of dipalmitoyl-phosphatidylcholine bilayers. The sphingomyelin/melittin system is thus proposed to show a reversible vesicle-to-disc transition (fluid-to-gel phase) through an intermediate fusion or aggregation event centered at the main transition temperature, Tm, as reported in the case of saturated phosphatidylcholine. In the case of spontaneous disc formation at 37 degrees C, the lipid-to-peptide molar ratio in the discoidal objects was determined to be approximately 20 for dipalmitoylphosphatidylcholine and about 12 in the case of natural sphingomyelin. Melittin partition coefficients between membranes and the aqueous medium at 37 degrees C were found to be 6.1 +/- 0.8 mM-1 and 3.7 +/- 0.4 mM-1 for sphingomyelin and dipalmitoylphosphatidylcholine, respectively. For very high peptide quantities (lipid-to-peptide molar ratio, Ri < or = 5) mixed micelles are formed over the entire temperature range (20 degrees to 60 degrees C) for both kinds of lipids.

Laurdan solvatochromism: solvent dielectric relaxation and intramolecular excited-state reaction.

Absorption, steady-state, and time-resolved fluorescence measurements have been performed on laurdan dissolved either in white viscous apolar solvents or in ethanol as a function of temperature. The heterogeneity of the absorption spectra in white oils or in ethanol is consistent with semiempirical calculations performed previously on Prodan. From steady-state and time-resolved fluorescence measurements in apolar media, an excited state reaction is evidenced. The bimodal lifetime distribution determined from the maximum entropy method (MEM) analysis is attributed to the radiative deexcitation of a "locally excited" (LE) state and of a "charge transfer" (CT) state, whereas a very short component (20 ps), the sign and the amplitude of which depend on the emission wavelength, is attributed to the kinetics of the interconvertion reaction. The observation of an isoemissive point in the temperature range from -50 degrees C to -110 degrees C in ethanol suggests an interconvertion between two average excited-state populations: unrelaxed and solvent-relaxed CT states. A further decrease in temperature (-190 degrees C), leading to frozen ethanol, induces an additional and important blue shift. This low temperature spectrum is partly attributed to the radiative deexcitation of the LE state. Time-resolved emission spectra (TRES) measurements at -80 degrees C in the ethanol liquid phase show a large spectral shift of approximately 2500 cm(-1) (stabilization energy of the excited state: 7.1 kcal x M(-1)). The time-dependent fluorescence shift (TDFS) is described for its major part by a nanosecond time constant. The initial part of the spectral shift reveals, however, a subnanosecond process that can be due to fast internal solvent reorientation and/or to intramolecular excited-state reactions. These two relaxation times are also detected in the analysis of the fluorescence decays in the middle range of emission energy. The activation energy of the longest process is approximately 3 kcal x M(-1). At -190 degrees C, one subnanosecond and one nanosecond excited-state reactions are also evidenced. They are likely due to intramolecular rearrangements after the excitation, leading to the CT state and not to solvent relaxation, which is severely hindered in these temperature conditions. Therefore, both intramolecular and solvent relaxations are responsible for the large Stokes shift displayed by this probe as a function of solvent polarity. A possible scheme is proposed for the deexcitation pathway, taking into account the kinetics observed in these different solvents.

Solubilization and reconstitution of vesicular stomatitis virus envelope using octylglucoside.

Reconstituted vesicular stomatitis virus envelopes or virosomes are formed by detergent removal from solubilized intact virus. We have monitored the solubilization process of the intact vesicular stomatitis virus by the nonionic surfactant octylglucoside at various initial virus concentrations by employing turbidity measurements. This allowed us to determine the phase boundaries between the membrane and the mixed micelles domains. We have also characterized the lipid and protein content of the solubilized material and of the reconstituted envelope. Both G and M proteins and all of the lipids of the envelope were extracted by octylglucoside and recovered in the reconstituted envelope. Fusion activity of the virosomes tested either on Vero cells or on liposomes showed kinetics and pH dependence similar to those of the intact virus.

Partition coefficient of a surfactant between aggregates and solution: application to the micelle-vesicle transition of egg phosphatidylcholine and octyl beta-D-glucopyranoside.

The mechanism of the solubilization of egg phosphatidylcholine containing 10% (M/M) of egg phosphatidic acid unilamellar vesicles by the nonionic detergent, octyl beta-D-glucopyranoside, has been investigated at both molecular and supramolecular levels by using fluorescence and turbidity measurements. In the lamellar region of the transition, the solubilization process has been shown to be first a function of the initial size before reaching an equilibrium aggregation state at the end of this region (the onset of the micellization process). The analysis during the solubilization process of the evolution of both the fluorescence energy transfer between N-(7-nitro-2,1,3-benzoxadiazol-4-yl)-phosphatidylethanolamine (NBD-PE) and N-(lissamine rhodamine B sulfonyl)-phosphatidylethanolamine (Rho-PE) and the fluorescence of 6-dodecanoyl-2-dimethylaminoaphtalene (Laurdan) has allowed us to determine the evolution of the detergent partitioning between the aqueous and the lipidic phases, i.e., the evolution of the molar fraction of OG in the aggregates (XOG/Lip) with its monomeric detergent concentration in equilibrium ([OG]H2O), throughout the vesicle-to-micelle transition without isolating the aqueous medium from the aggregates. The curve described by XOG/Lip versus [OG]H2O shows that the partition coefficient of OG is changing throughout the solubilization process. From this curve, which tends to a value of 1/(critical micellar concentration), five different domains have been delimited: two in the lamellar part of the transition (for 0 < [OG]H2O < 15.6 mM), one in the micellization part, and finally two in the pure micellar region (for 16.5 < [OG]H2O < 21 mM). The first domain in the lamellar part of the transition is characterized by a continuous variation of the partition coefficient. In the second domain, a linear relation relates XOG/Lip and [OG]H2O, indicating the existence of a biphasic domain for which the detergent presents a constant partition coefficient of 18.2 M-1. From the onset to the end of the solubilization process (domain 3), the evolution of (XOG/Lip) with [OG]H2O can be fitted by a model corresponding to the coexistence of detergent-saturated lamellar phase with lipid-saturated mixed micelles, both in equilibrium with an aqueous phase, i.e., a three-phase domain. The micellar region is characterized first by a small two-phase domain (domain 4) with a constant partition coefficient of 21 M-1, followed by a one-phase mixed-micellar domain for which XOG/Lip no longer linearly depends on [OG]H2O. The results are discussed in terms of a phase diagram.

Solubilization steps of dark-adapted purple membrane by Triton X-100. A spectroscopic study.

Ultraviolet-visible spectroscopy has been used to follow the solubilization of the dark-adapted purple membrane of Halobacterium halobium by Triton X-100. Turbidity of purple membrane fragments and absorbance of bacteriorhodopsin variations during continuous addition of detergent give solubilization profiles exhibiting several break points corresponding to different equilibrium stages of the solubilization process. The present method allows the determination of the detergent to protein+lipid ratio in mixed aggregates at the corresponding break points. It was concluded that, when performed systematically, this technique is a very convenient and powerful tool for the quantitative study of biomembrane-to-micelle transition.

Size analysis and stability study of lipid vesicles by high-performance gel exclusion chromatography, turbidity, and dynamic light scattering.

Vesicles of egg phosphatidylcholine (EPC) and phosphatidic acid (EPA) were prepared by reverse-phase evaporation (REV) followed either by sequential extrusion through polycarbonate membranes with pore diameters of 0.8, 0.4, 0.2, 0.1, and 0.05 micron or by filtration through 0.8-micron cellulosic or 0.22-micron polyvinylidene fluoride (PVF) membranes. The resulting vesicles ranging from 130 to 640 nm in mean diameter (REVs) were characterized by high-performance liquid chromatography (HPLC) using a TSK G6000 PW gel exclusion column. The efficiency of this technique to determine vesicle size parameters was studied by the analysis of the chromatograms in combination with dynamic light scattering (DLS) determination of the mean diameters (MD) of the fractionated vesicles in the region of the elution profile maxima. The HPLC TSK G6000 PW gel exclusion provides a reproducible and fast method of size characterization for lipid vesicles having MD up to 1 micron, the best selectivity being obtained in the 20- to 500-nm MD range. HPLC analysis of REV's demonstrates that: (i) both the average size and polydispersity of the vesicles decrease with decreasing pore size of the membranes, cellulosic or PVF "tortuous" ones being less efficient than "straight bores" polycarbonate ones; (ii) mixed EPC/EPA REVs sequentially extruded down through 0.2-micron polycarbonate membranes are highly deformable without rupture of the bilayer; and (iii) the mean size of extruded REV's is stable for at least 1 week. The role of EPA on the size stability of mixed EPC/EPA vesicles was studied by coupling HPLC gel exclusion and turbidity analysis of pure EPC and EPC/EPA (mole ratio: 91/9) sonicated small unilamellar vesicles as a function of time. The apparent size variation of EPC vesicles observed over a week, is mainly due to their aggregation which is significantly reduced by the introduction of a small amount of EPA in the vesicle membrane.

pH-dependent fusion of reconstituted vesicular stomatitis virus envelopes with Vero cells. Measurement by dequenching of fluorescence.

Reconstituted vesicular stomatitis virus (VSV) envelopes were formed by solubilization of the viral envelope with Triton X-100 followed by removal of detergent by direct addition of SM2 biobeads. We provide direct demonstration of fusion of reconstituted VSV with cells using fluorescent lipid and aqueous probes incorporated into the VSV virosomes during reconstitution. We show a direct comparison of the kinetics and pH profile of fusion with cells between reconstituted VSV and fluorescently labeled intact virus. With this preparation it is now possible to gain additional information about the role of cooperativity in viral protein-mediated fusion, and to permit construction of efficient vehicles for delivery of drugs and other materials into cells.

Mechanisms of membrane protein insertion into liposomes during reconstitution procedures involving the use of detergents. 1. Solubilization of large unilamellar liposomes (prepared by reverse-phase evaporation) by triton X-100, octyl glucoside, and sodium cholate.

The mechanisms governing the solubilization by Triton X-100, octyl glucoside, and sodium cholate of large unilamellar liposomes prepared by reverse-phase evaporation were investigated. The solubilization process is described by the three-stage model previously proposed for these detergents [Lichtenberg, D., Robson, R.J., & Dennis, E.A.(1983) Biochim. Biophys. Acta 737, 285-304]. In stage I, detergent monomers are incorporated into the phospholipid bilayers until they saturate the liposomes. At that point, i.e., stage II, mixed phospholipid-detergent micelles begin to form. By stage III, the lamellar to micellar transition is complete and all the phospholipids are present as mixed micelles. The turbidity of liposome preparations was systematically measured as a function of the amount of detergent added for a wide range of phospholipid concentrations (from 0.25 to 20 mM phospholipid). The results allowed a quantitative determination of RSat, the effective detergent to lipid molar ratios in the saturated liposomes, which were 0.64, 1.3, and 0.30 for Triton X-100, octyl glucoside, and sodium cholate, respectively. The corresponding ratios in the mixed micelles, RSol, were 2.5, 3.8, and 0.9 mol of detergent/mol of phospholipid. The monomer concentrations of the three detergents in the aqueous phase were also determined at the lamellar to micellar transitions (0.18, 17, and 2.8 mM, respectively). These transitions were also investigated by 31P NMR spectroscopy, and complete agreement was found with turbidity measurements. Freeze-fracture electron microscopy and permeability studies in the sublytic range of detergent concentrations indicated that during stage I of solubilization detergent partitioning between the aqueous phase and the lipid bilayer greatly affects the basic permeability of the liposomes without significantly changing the morphology of the preparations. A rough approximation of the partition coefficients was derived from the turbidity and permeability data (K = 3.5, 0.09, and 0.11 mM-1 for Triton X-100, octyl glucoside, and sodium cholate, respectively). It is concluded that when performed systematically, turbidity measurements constitute a very convenient and powerful technique for the quantitative study of the liposome solubilization process by detergents.

Mechanisms of membrane protein insertion into liposomes during reconstitution procedures involving the use of detergents. 2. Incorporation of the light-driven proton pump bacteriorhodopsin.

A method has been developed for identifying the step in a detergent-mediated reconstitution procedure at which an integral membrane protein can be associated with phospholipids to give functional proteoliposomes. Large liposomes prepared by reverse-phase evaporation were treated with various amounts of the detergents Triton X-100, octyl glucoside, or sodium cholate as described in the preceding paper [Paternostre, M.-T., Roux, M., & Rigaud, J. L. (1988) Biochemistry (preceding paper in this issue)]. At each step of the solubilization process, we added bacteriorhodopsin, the light-driven proton pump from Halobacterium halobium. The protein-phospholipid detergent mixtures were then subjected to SM2 Bio-Beads treatments to remove the detergent, and the resulting vesicles were analyzed with respect to protein insertion and orientation in the membrane by freeze-fracture electron microscopy, sucrose density gradients, and proton pumping measurements. The nature of the detergent used for reconstitution proved to be important for determining the mechanism of protein insertion. With sodium cholate, proteoliposomes were formed only from ternary phospholipid-protein-detergent micelles. With octyl glucoside, besides proteoliposome formation from ternary mixed micelles, direct incorporation of bacteriorhodopsin into preformed liposomes destabilized by saturating levels of this detergent was observed and gave proteoliposomes with optimal proton pumping activity. With Triton X-100, protein insertion into destabilized liposomes was also observed but involved a transfer of the protein initially present in phospholipid-Triton X-100-protein micelles into Triton X-100 saturated liposomes. Our results further demonstrated that protein orientation in the resulting proteoliposomes was critically dependent upon the mechanism by which the protein was incorporated.

Effects of N-alcohols on potassium conductance in squid giant axons.

The effect of bath application of several short chain N-alcohols on voltage-dependent potassium conductance has been studied in intact giant axons of Loligo forbesi under voltage-clamp conditions. All tested alcohols (methanol, ethanol, propanol, butanol, heptanol and octanol) were found to depress potassium conductance only at concentrations much larger than those necessary to reduce sodium conductance. The efficacy of the different molecules was correlated with the carbon-chain length. In all cases the effects were found to be at least partly reversible. Low concentrations of propanol (100 mM) or heptanol (1 mM) were found to increase potassium conductance whereas higher concentrations had the usual depressing effect. The two alcohols were found to induce a slow inactivation of the potassium conductance. A detailed analysis of the time course of the turning-on of the potassium current for various pulse potentials in the presence of TTX revealed that, for membrane potential values more positive than -20 mV, the time constant of activation was reduced in the presence of propanol or heptanol. The delay which separates the change in potential and the turning-on of the potassium current, which was systematically analysed for different pulse and prepulse potential values, was increased by the two alcohols, the curve relating this delay to prepulse potential being shifted towards larger (positive) delays. This high degree of complexity in the effects on potassium conductance suggests that the alcohol molecules modify several more or less independent mechanisms associated with the turning-on of the potassium current.