Unilamellar Phospholipid Vesicles Research Articles

Increasing Packing Density of Hydrated Dipalmitoylphosphatidylcholine Unilamellar Vesicles Induced by Trehalose

On the basis of the observation of 31P{1H} nuclear Overhauser enhancement, trehalose is shown to increase the packing density of the polar head groups of phospholipid unilamellar vesicles. This finding is not consistent with previous results obtained for Langmuir-type membranes or multilamellar vesicles. The present results explain the anti-fusogenic ability of trehalose.

The effects of lipid composition on the binding of lasalocid A to small unilamellar vesicles

The binding of the carboxylic ionophore lasalocid A (X537A) to small unilamellar phospholipid vesicles of varying composition was examined in an effort to determine what structural features of the phospholipid membrane influence the ionophore-membrane interaction. Apparent dissociation constants ( K app) were calculated for both the acidic and anionic forms of the ionophore using the change in fluorescence intensity observed for lasalocid A upon addition of phospholipid vesicles. The K app for binding to fluid phase dimyristoylphosphatidylcholine (DMPC) vesicles is 46 μM for the anion and 14 μM for the acid. While the phase transition of DMPC had no effect on the K app of the anion, an increase was observed in the K app of the acid below the phase transition temperature. The K app of the anion was not affected by the incorporation of 10% dimyristoylphosphatidylethanolamine (DMPE), but increased slightly upon incorporation of cholesterol. The p K a values of the ionophore were the same in DMPC and DMPC/DMPE membranes. Incorporation of the negative lipids phosphatidylglycerol, phosphatidic acid, or phosphatidylethanolamine (at pH 9.4 where PE carries a negative charge) decreases binding of the anion in accord with the increase in surface potential estimated from Gouy-Chapman theory. The CD spectrum of membrane-bound lasalocid A anion indicated the ionophore to be in an extended acyclic conformation on the membrane surface with the C-1 carboxylate rotated out of the plane of the salicylate ring. The out-of-plane rotation of the carboxylate may be the result of facial binding by the amphiphilic ionophore on the membrane surface or of weak ion pairing to the polar lipid head groups. These results suggest that the primary determinants of binding of the anionic ionophore on the membrane surface are packing density of the polar head groups and membrane surface potential. There is no evidence of strong hydrogen bond formation between the lipid polar head groups and the ionophore as has previously been suggested.

A novel method for the efficient entrapment of calcium in large unilamellar phospholipid vesicles

A technique for the efficient entrapment of high concentrations of Ca 2+ in large unilamellar phospholipid vesicles (LUVs), using the carboxylic acid antibiotic ionophore A23187 (calcimycin) is demonstrated. It is shown that rapid A23187-mediated entrapment of Ca 2+, corresponding to essentially 100% sequestration of the extravesicular cation may be achieved for egg yolk phosphatidylcholine LUVs (100 nm) in the presence of a transmembrane proton gradient (acidic interior). Interior-exterior concentration cation gradients of over 400-fold may be readily achieved, with interior Ca 2+ concentrations in excess of 250 mM. It is shown that the extent and efficiency of the A23187-mediated uptake process is affected by the intravesicular buffering capacity and the extravesicular Ca 2+ concentration in a manner that is consistent with a Ca 2+-H + exchange process. In the absence of a pH gradient, or the presence of a reversed gradient (basic interior), only background levels of cation uptake are detected. The driving force for A23187-mediated uptake of Ca 2+ is shown to depend on the intravesicular proton pool rather than on a chelation process. This protocol provides a novel method for the efficient entrapment of high concentrations of Ca 2+ and other cations in phospholipid vesicles.

Accelerated thrombolysis and reperfusion in a canine model of myocardial infarction by liposomal encapsulation of streptokinase.

The aim of thrombolytic therapy for acute myocardial infarction with plasminogen activators such as streptokinase is to lyse the coronary thrombus and reestablish blood flow as quickly as possible so that heart tissue loss is minimized and mortality rates are improved. Streptokinase has been encapsulated in large unilamellar phospholipid vesicles and tested in an animal model of acute myocardial infarction. The time required to restore vessel patency has been reduced more than 50% when compared with findings for free streptokinase. The total dosage of streptokinase required was lower, and smaller remnant thrombi were observed with the encapsulated agent. Results from this initial unoptimized study may have significant implications for further reduction in mortality from heart attacks by therapy with plasminogen activators.

Frequency dependent2H N.M.R. relaxation rates of small unilamellar phospholipid vesicles

Deuterium (2H) N.M.R. spectroscopy has been used to investigate experimentally the molecular dynamics of phospholipid vesicles. 1,2-dimyristoyl-sn-glycero-3-phosphocholine (DMPC) was deuterated specifically in both acyl chains, and aqueous suspensions of vesicles were studied above the main orderdisorder transition temperature. 2H spin-lattice relaxation rates (R 1) were measured at nine different magnetic field strengths, corresponding to Larmor frequencies (v o = ωo/2π) spanning 2·5 MHz to 61·4 MHz. Of three models considered, a model in which R 1 ∝ ω-1/2 0 best described the 2H N.M.R. results, whereas fits to R 1 ∝ ω-1 0 or to a lorentzian plus a constant dispersion were less satisfactory. Thus a broad distribution of motions exists, which must be explained by any model for the N.M.R. relaxation of liquid-crystalline phospholipids.

Binding of human factor VIII to phospholipid vesicles.

Factor VIII, a protein cofactor involved in blood coagulation, functions in vitro on a phospholipid membrane surface to greatly increase the rate of factor X activation by factor IXa. Using gel filtration, rapid sedimentation, and resonance energy transfer we have studied the interaction of recombinant-derived human factor VIII with small and large unilamellar phospholipid vesicles composed of phosphatidylserine and phosphatidylcholine. Resonance energy transfer, from intrinsic fluorophores in factor VIII to dansyl-phosphatidylethanolamine incorporated into vesicles, has been adapted for quantitative equilibrium measurements. Factor VIII binds rapidly and reversibly to small and large vesicles. At 8 degrees C the interaction of factor VIII with small vesicles fits a simple bimolecular model with a KD of 2 nM and a phospholipid binding site defined by 180 phospholipid monomers. At 25 degrees C the binding of factor VIII to small vesicles containing 20% phosphatidylserine can be described by an apparent KD of 4 nM; the phospholipid/protein ratio at saturation was 170. Binding to large vesicles was demonstrated with a KD of 2 nM and a phospholipid/protein ratio at saturation of 385. Binding was dependent upon the phosphatidylserine mole fraction and was nonlinear from 0 to 30% phosphatidylserine content. A direct comparison of factor VIII and factor V binding indicated that the affinity of factor V to phospholipid vesicles was equivalent to that of factor VIII and that the phosphatidylserine requirement was lower. A model is proposed to explain the nonlinear phosphatidylserine dependence of binding for factor VIII.

Thrombolysis using liposomal-encapsulated streptokinase: an in vitro study.

The clot-lysing ability of streptokinase (SK) was examined using membrane-bound thrombi. Encapsulation of SK in large unilamellar phospholipid vesicles (liposomes) resulted in entrapping approximately 30% of its original activity. Measurements of streptokinase activity for liposomal-encapsulated streptokinase (LESK) indicated little loss of activity or leakage in Tris-buffered saline over a 24-hr period at temperatures of 4 and 23 degrees C. However, incubation of free SK and LESK in platelet-poor plasma (PPP) at 37 degrees C resulted in a decrease of SK activity. The retention of SK activity in LESK was considerably higher than that of unentrapped SK. Clot-dissolving time (CDT) was measured by monitoring the pressure drop during slow filtration in plasma through membrane-bound thrombi. The results indicated that both LESK and free SK were able to activate the fibrinolytic system. Without prior incubation in PPP at 37 degrees C, the CDT of a SK and PPP mixture (SK/PPP) was 10.7 +/- 1.9 min (n = 12), while that of a LESK and PPP mixture (LESK/PPP) was 12.4 +/- 1.7 min (n = 12). The CDT-detected clot-lysing abilities of both SK and LESK were diminished by incubation in PPP, but to different extents. After 15- and 30-min incubations, the CDT of SK/PPP increased significantly to 15.5 +/- 1.5 and 24.1 +/- 2.4 min (n = 5, P less than 0.05), respectively. In contrast, the CDT of LESK/PPP increased to 13.3 +/- 0.8 min (n = 5) after 15 min of incubation and to 16.0 +/- 1.1 min (n = 5, P less than 0.05) after a 30-min incubation. These results suggest that entrapment of SK in liposomes preserves the thrombolytic potential of the plasminogen activator by limiting its exposure to the components of the plasma.

Cholesterol metabolism by purified cytochrome P-450scc is highly stimulated by octyl glucoside and stearic acid exclusively in large unilamellar phospholipid vesicles.

Cholesterol side-chain cleavage (CSCC) catalyzed by purified bovine adrenal mitochondrial cytochrome P-450scc is highly dependent on the vesicles that supply cholesterol. Six-fold higher rates are achieved with large unilamellar dioleoylphosphatidylcholine vesicles (diameter 150 nm) prepared by octyl glucoside (OG) dialysis (DOPC-LUV) than with small sonicated vesicles (diameter 30 nm) (DOPC-SUV) (Vmax = 25 and 4 min-1, respectively. Extensive dialysis that may remove OG decreased Vmax rates for DOPC-LUV almost to rates seen with DOPC-SUV. These dialyzed DOPC-LUV were, however, very sensitive to addition of OG (EC50 = 2.5 microM, 4.3-fold stimulation) while DOPC-SUV were only weakly affected (EC50 = 100 microM, 1.6-fold stimulation). This enhancement of CSCC in LUV by OG only occurred when the cholesterol:DOPC exceeded 0.1 and was associated with a 15-fold increase in the Km for cholesterol. Structural changes in both SUV and LUV at high cholesterol:DOPC ratios (0.1-1) were indicated by decreases in internal volume that were insensitive to OG and did not affect the external diameters. Stearic acid produced a similar stimulation of CSCC in LUV (EC50 = 50 microM) and had no effect on SUV. The Vmax for CSCC, produced by OG activation of DOPC-LUV, is comparable to the highest attained for cytochrome P-450scc (Tween 20/cholesterol). In LUV, a minor proportion of OG (1-5% of cholesterol) is thus sufficient to generate a domain of reactive cholesterol that maintains a near-optimum turnover. This increased CSCC was paralleled by increased binding of cholesterol to P-450scc, suggesting that this cholesterol is more readily donated by the membrane to the cytochrome.

Cholesterol movement between bovine rod outer segment disk membranes and phospholipid vesicles

The ability of cholesterol to move between bovine rod outer segment disk membranes and phospholipid membranes was examined. Disk membranes were incubated with small unilamellar phospholipid vesicles containing varying amounts of cholesterol. Aliquots were removed at specific times, and then the disks and the vesicles were separated by centrifugation and assayed for phospholipid and cholesterol content. When incubated with vesicles containing no cholesterol, the cholesterol to phospholipid ratio in the disk membrane was reduced due to migration of cholesterol from the disks into the vesicles. The cholesterol content of these cholesterol depleted disks could be readily returned to the normal disk cholesterol content by incubation of the cholesterol-depleted disks with small unilamellar vesicles containing high cholesterol. An apparent partition coefficient K was calculated as the quotient of the cholesterol/phospholipid mole ratio in the donor membranes and the cholesterol/phospholipid mole ratio in the acceptor membranes. The value of K was approximately 1 at cholesterol levels below normal disk cholesterol content, for disk membranes and phosphatidylcholine small unilamellar vesicles. Inclusion of phosphatidylethanolamine in the small unilamellar vesicle acceptor raised K, indicating that phosphatidylethanolamine creates an unfavourable environment for cholesterol. The cholesterol to phospholipid ratio of native disks could be increased by incubation with phosphatidylcholine small unilamellar vesicles (donor) which contained higher amounts of cholesterol than the disk membrane acceptor. In these experiments the distribution of cholesterol between disks and small unilamellar vesicles always favored the vesicles. The apparent partition coefficient was 1.7 at several cholesterol levels above the native disk cholesterol content. Liposomes made from lipid extracted from the disk membrane behaved in the same manner as intact disks with respect to cholesterol distribution at equilibrium. The phospholipid content of the disk membrane may be an important factor in determining the cholesterol content of the disk membrane.

Binding of crotoxin, a presynaptic phospholipase A2 neurotoxin, to negatively charged phospholipid vesicles.

Crotoxin, isolated from the venom of Crotalus durissus terrificus, is a potent neurotoxin consisting of a basic and weakly toxic phospholipase A2 subunit (component B) and an acidic nonenzymatic subunit (component A). The nontoxic component A enhances the toxicity of the phospholipase subunit by preventing its nonspecific adsorption. The binding of crotoxin and of its subunits to small unilamellar phospholipid vesicles was examined under experimental conditions that prevented any phospholipid hydrolysis. Isolated component B rapidly bound with a low affinity (Kapp in the millimolar range) to zwitterionic phospholipid vesicles and with a high affinity (Kapp of less than 1 microM) to negatively charged phospholipid vesicles. On the other hand, the crotoxin complex did not interact with zwitterionic phospholipid vesicles but dissociated in the presence of negatively charged phospholipid vesicles; the noncatalytic component A was released into solution, whereas component B remained tightly bound to lipid vesicles, with apparent affinity constants from 100 to less than 1 microM, according to the chemical composition of the phospholipids. On binding, crotoxin or its component B caused the leakage of a dye entrapped in vesicles of negatively charged but not of zwitterionic phospholipids. The selective binding of crotoxin suggests that negatively charged phospholipids may constitute a component of the acceptor site of crotoxin on the presynaptic plasma membrane.

The use of immunosome technology for vaccines against rabies and other viral diseases

Subunit viral vaccines present several advantages. They are free of nucleic acids (of viral and/or cellular origin) and proteins of cellular and/or serum origin; they contain only the relevant antigen. For rabies virus, the antigen which induces the virus-neutralizing antibody (VNAb) is the glycoprotein (GP), which is anchored, in form of spikes, to the viral membrane. The GP may be extracted from the virion but, after solubilization and purification, it appears to be poorly immunogenic. In order to restore its immunogenicity, GP molecules are anchored to preformed liposomes (unilamellar phospholipid vesicles) to mimic their native structure and environment. The subunit vaccine obtained by this technique is called an "immunosome" (IMS). Rabies immunosomes exhibit structural and immunological properties very similar to those of the viral particle. The rabies glycoprotein molecules, anchored to the lipid bilayer of the liposome, correctly expose the immunodominant epitope involved in VNAb induction and induce a strong specific humoral immune response. They also induce a specific cellular immune response. As a result IMS have a highly protective activity when tested with either pre- or post-exposure potency tests. Immunosome technology may be applied to other purified membrane proteins or amphiphilic peptides to restore their immunogenicity.

Lipid bilayer permeation by neutral aluminum citrate and by three α-hydroxy carboxylic acids

Several groups have proposed that aluminum (Al) may permeate biologival membranes as a neutral complex with citrate. We tested this hypothesis by measuring aluminum citrate flux across unilamellar phospholipid vesicles (liposomes). Results from two independent procedures show that lipid bilayer permeation by the neutral aluminum-citrate complex is slow ( P∼=1·10 −11 cm·s −1). We then compared aluminum-citrate permeation with permeation by a series of α-hydroxy carboxylic acids and by trimethylcitrate. This comparison showed that the aluminum-citrate flux is limited by diffusion across the water/lipid interface. This is due to hydrogen bonding between water and the citrate carboxyl groups, and by hydration of the bound metal in the aqueous phase. By analogy with citric acid, steric hindrance of diffusion within the bilayer does not affect the permeation rate of aluminum citrate. Elevated tissue levels of Al in subjects fed a diet supplemented with citric acid and Al(OH) 3 cannot be explained by lipid bilayer permeation of the neutral complex.

Spontaneous transfer between phospholipid bilayers of dehydroergosterol, a fluorescent cholesterol analog

The spontaneous interbilayer transfer of dehydroergosterol, a fluorescent cholesterol analog, was examined using small unilamellar phospholipid vesicles. The kinetic data were best fit by an equation of the form Aexp (− kt) + B. Qualitatively, the general trend of the half-time for transfer and the base values ( B) obtained for dehydroergosterol resemble the corresponding values obtained in the earlier studies of cholesterol transfer. However, quantitative differences, which reflect the molecular structure of the sterol, were observed. Acrylamide quenching performed on the donor vesicles at different stages of the transfer indicated that a time-dependent organization of DHE within the vesicles occurs.

A fluorescence study examining how 14-valerate side chain substitution modulates anthracycline binding to small unilamellar phospholipid vesicles

The intrinsic fluorescence properties of the anthracycline antitumor antibiotics were studied in an effort to understand how 14-valerate side chain substitution modulates drug associations with small unilamellar phospholipid vesicles (SUVs) under near physiological conditions. Drug location and dynamics in fluid-phase dimyristoylphosphatidylcholine (DMPC) bilayers were evaluated for several analogs; accessibilities of bound fluorophores to membrane-impermeable iodide were evaluated in quenching experiments, while the diffusive motions of these agents were studied using lifetime-resolved anisotropy plots. The bulky and hydrophobic valerate substituent was found to further hinder the rotations experienced by a bound drug molecule, with apparent limiting anisotropy ( α ∞) values showing increases of 13–82% upon valerate group substitution. In addition, the bimolecular quenching rate constants (unit, 10 9 M −1 · s −1) for membrane-bound adriamycin (1.4), N-trifluoroacetyladriamycin (0.4), and their corresponding valerate-substituted analogs ( k q values of 1.1 and 0.5, respectively) reveal that the side chain is a weak modulator of fluorophore penetration into the bilayer, with stronger modulation being achieved through amino group substitution. Similar results were obtained for drugs bound to negatively-charged dimyristoylphosphatidylglycerol (DMPG) bilayers. Finally, comparison of the equilibrium binding affinities of the various congeners for electroneutral DMPC versus negatively-charged DMPG bilayers demonstrate that positively-charged parent anthracyclines display high levels of selective binding to negatively-charged phospholipids, unlike valerate-substituted analogs which display no such selectivity. The modulation of anthracycline-membrane interactions achieved through valerate substitution offers potential explanations, at least in part, for some of the novel biological properties of valerate-containing anthracyclines.

Reconstitution and characterization of the human neutrophil N-formyl peptide receptor and GTP binding proteins in phospholipid vesicles

We have developed a unilamellar phospholipid vesicle system which contains the N-formyl peptide receptor and GTP binding proteins. Several detergents were investigated but only two, octyl glucoside (35 mM) and deoxycholate (7.5 mM), were capable of extracting N-formyl peptide receptor from neutrophil membranes in a form which remained functionally active upon reconstitution into phospholipid vesicles. Extracted proteins were reconstituted into phosphatidylcholine vesicles by passage over a Sephadex G-50-80 column. The reconstituted formylpeptide receptor could bind [ 3H]FMLP ( 3H-labeled fMet-Leu-Phe) and [ 125I]FMLPL-SASD ( 125I-labeled ( 125I-labeled N-formylmethionylleucylphenylalanyl- N 8-(2-( p-azidosalicylamido(ethyl-1,3′-dithiopropionyl)lysine) while the endogenous G protein could bind [ 35S]GTPγS. Furthermore, the functional interaction of the two proteins was preserved. Addition of the nonhydrolyzable guanine nucleotide, GTPγS, shifted the N-formyl peptide receptor from a high- to a low-affinity binding state for ligand. The development of this in vivo reconstitution system should provide a basis to study the mechanism of interaction of the N-formyl peptide receptor and the G protein.

ESR study of proton transport across phospholipid vesicle membranes

A new method for measuring the rates of proton transfer through bilayer phospholipid membranes using pH-sensitive nitroxyl radicals is suggested. The pH-sensitive alkylating radical was covalently bound to glutathione. This modified glutathione is pH sensitive at pH 1.5–4.5 and does not penetrate across phospholipid membranes. Using ESR this probe was applied to register the kinetics of pH variations inside large unilamellar phospholipid vesicles after creation ofa transmembrane proton gradient. In the acidic region (pH ≈ 3) the main mechanism of transmembrane proton transfer is that via transport of a proton in the form of an undissociated acid. The membrane permeability coefficients have been determined for a series of acids (HCl, HClO 4, HNO 3, upper estimate for H 2SO 4). Taking into account that imidazoline and imidazolidine nitroxyl radicals can be used as pH probes in a wide range of pH, the present method can be developed for measuring the rates of transmembrane proton transfer in neutral and alkaline media.

Entrapment of high-molecular-mass DNA molecules in liposomes for the genetic transformation of animal cells.

We modified the Ca/EDTA procedure for the production of liposomes [Papahadjopoulos, Vail, Jacobson & Poste (1975) Biochim. Biophys. Acta 394, 483-491] to entrap intact DNA molecules of very high molecular mass into large unilamellar phospholipid vesicles. The use of DNA-protein complexes and phage particles instead of naked linear DNA increases the efficiency of entrapment and protects the integrity of DNA molecules. We investigated the interaction of mammalian cells with liposome-encapsulated recombinant lambda bacteriophages carrying marker genes. The liposomes bind surprisingly fast to the cellular surface and are taken up by the cells. A significant proportion of the encapsulated DNA is transported to and soon located in or around the nuclei. Experiments prove that these liposomes can be used for the genetic transformation of mammalian cells.

Interaction of Concanavalin A and a Divalent Derivative with Lymphocytes and Reconstituted Lymphocyte Membrane Glycoproteins

Both concanavalin A (con A) and its divalent derivative, succinyl-concanavalin A (S-con A) are mitogenic for porcine lymph node lymphocytes. We have compared the binding of these two lectins to intact porcine lymphocytes and phospholipid vesicles containing reconstituted lymphocyte membrane glycoproteins. Both con A and S-con A showed high- and low-affinity binding to intact cells, as indicated by LIGAND analysis of Scatchard plots of binding data. Despite the apparently identical saccharide specificities of the two lectins, high-affinity binding sites for S-con A were only one-third as numerous as high-affinity sites for the parent lectin. Large numbers of low-affinity binding sites existed for con A, while many fewer were present for S-con A. It is suggested that these sites result from hydrophobic association. Con A bound to lymphocytes in a positively cooperative fashion, while S-con A showed noncooperative behavior. Lectin binding to large unilamellar phospholipid vesicles containing reconstituted lymphocyte membrane glycoproteins was measured using a rapid filtration assay, and was linear with the glycoprotein content of the vesicles. Almost all of the outward-facing glycoprotein was functional in terms of lectin binding. Reconstituted glycoproteins showed only a single class of high-affinity binding sites for both con A and S-con A, with association constants similar to those measured for intact cells. Con A, but not S-con A, showed positively cooperative binding to reconstituted vesicles. Cooperativity was observed in both gel phase and liquid crystalline phase lipid, and was thus not dependent on long-range lateral rearrangement of glycoprotein receptors. Results suggested that con A induces a microredistribution of receptors on the lymphocyte membrane surface, leading to the exposure of glycoproteins that were previously inaccessible to the lectin. S-Con A does not cause glycoprotein redistribution, and a large fraction of the receptors remain cryptic.

Equilibrium Uptake of D-Glucose by Osmotically Stressed Unilamellar Phospholipid Vesicles

The assumptions inherent in the use of osmotic manipulation to determine the extent of solute binding to brush border membrane vesicles (the ideal osmotic responsiveness of the vesicles and the independence of solute binding from the incubation medium osmotic pressure) were examined in a model system (large unilamellar lipid vesicles). The equilibrium uptake of D-glucose by unilamellar vesicles composed of egg lecithin (PC), phosphatidic acid (PA), and cholesterol (Chol) was measured as a function of the osmotic concentration of the incubation medium. The variation of the encapsulated aqueous volume of PC:PA and PC:PA:Chol vesicles with the osmotic stress was directly determined by a fluorescence self-quenching technique. Encapsulated volume changes of both PC:PA and PC:PA:Chol vesicles were found to be resistant to the osmotic stress, exhibiting positive deviations from ideal behavior. Equilibrium uptake experiments with these vesicles showed that glucose was taken up in excess of that amount predicted on the basis of the encapsulated volume when the vesicles were subjected to osmotic stress <0.25 osmol/kg. At osmotic stresses >0.75 osmol/kg, equilibrium uptake could be predicted solely on the basis of the encapsulated volume. These results, based on a model vesicle system, strongly suggest that osmotic manipulation may be an inappropriate method to assess the extent of solute binding to natural membrane vesicle preparations, such as brush border membrane vesicles, without more direct evidence.

Facilitation of the permeability of phospholipid vesicles to Tl + by prothrombin

The rate of release of Tl + from unilamellar phospholipid vesicles of various compositions was measured by pulse polarography as a function of prothrombin concentration, in the presence and in the absence of Ca 2+. The experiments were carried out at different temperatures, in the broad transition temperature region of the lipid mixtures. No significant release was observed from mixtures of 70% DPPC + 30% PC interacting with prothrombin. When it interacted with mixtures of 50% DPPC + 30% PC + 20% PS, the release was facilitated significantly with a first-order release rate constant. It was higher in the presence of Ca 2+ than in its absence, at 24° C, indicating a more pronounced penetration of the protein molecules in the lipid bilayers with Ca 2+.