Membrane-damaging action of staphylococcal alpha-toxin on phospholipid-cholesterol liposomes

The mechanism of membrane damage by staphylococcal alpha-toxin was studied using carboxyfluorescein (internal marker)-loaded multilamellar liposomes prepared from various phospholipids and cholesterol. Liposomes composed of phosphatidylcholine or sphingomyelin and cholesterol bound alpha-toxin and released carboxyfluorescein in a dose dependent manner, when they were exposed to alpha-toxin of concentrations higher than 1 or 8 micrograms/ml, respectively. In contrast, the other liposomes composed of phosphatidylethanolamine, phosphatidylserine, phosphatidylglycerol or phosphatidylinositol plus cholesterol were not susceptible to the toxin even at high concentrations up to 870 micrograms/ml. The insensitive liposomes containing either phosphatidylserine or phosphatidylglycerol were made sensitive to alpha-toxin by inserting phosphatidylcholine into the liposomal membranes. In addition, phosphorylcholine inhibited the toxin-induced marker release from liposomes. These results indicated that the choline-containing phospholipids are required for the interaction between alpha-toxin and liposomal membranes. Susceptibility of liposomes containing phosphatidylcholine or sphingomyelin increased with the increase in cholesterol contents of the liposomes. Based on these results, we propose that the choline-containing phospholipids are possible membrane components or structures responsible for the toxin-membrane interaction, which leads to damage of membranes. Furthermore, cholesterol may facilitate the interaction between alpha-toxin and membrane as a structural component of the membrane.     

Haptenic activity of galactosyl ceramide and its topographical distribution on liposomal membranes. I. Effect of cholesterol incorporation

The relation between the immune-reaction of phosphatidylcholine liposomes containing spin-labeled galactosyl ceramide with or without cholesterol and the topographical distribution of the glycolipid in membranes was studied. In egg yolk phosphatidylcholine liposomes, both immune agglutination and antibody binding occurred, irrespectively of the presence of cholesterol, though the motion of the fatty acyl chain of spin-labeled galactosyl ceramide was restricted by cholesterol. In dipalmitoyl phosphatidylcholine liposomes, unlike in egg yolk phosphatidylcholine liposomes, the immune-reaction depended on the cholesterol content. The electron spin resonance (ESR) spectra of spin-labeled galactosyl ceramide in dipalmitoyl phosphatidylcholine liposomes indicated that cholesterol affected the topographical distribution of spin-labeled galactosyl ceramide in the liposomes. Without cholesterol, most of the spin-labeled galactosyl ceramide was clustered on the dipalmitoyl phosphatidylcholine membrane, but with increase of cholesterol, random distribution of hapten on the membrane increased. The cholesterol-dependent change in the topographical distribution of hapten on the membranes was parallel with that of immune reactivity. 'Aggregates' composed solely of galactosyl ceramide did not show any binding activity with antibody. The findings suggest that the recognition of galactosyl ceramide by antibody depended on the topographical distribution of hapten molecules. Phosphatidylcholine and/or cholesterol may play roles as 'spacers' for the proper distribution of 'active' haptens on the membranes. The optimum density of haptens properly distributed on liposomal membranes is discussed.     

The dispersion of cholesterol with phospholipids and glycolipids

Of the polar lipids studied (phospholipids and glycolipids), only phosphatidylcholine and sphingomyelin can disperse in water with up to 2 mol cholesterol/mol polar lipid. However, mixtures of phosphatidylethanolamine with small amounts of phosphatidylcholine and mixed lipids from mitochondria and myelin will also form sterol-rich dispersions. Steroids in which the 3β-OH group is replaced by an oxo function do not form such steroid-rich dispersions. Electron microscopy and optical rotatory dispersion (ORD) show that sterols disperse with cerebrosides and gangliosides to form cylindrical structures with the regions around C atoms 3 and 7 of the sterol in less polar environments than those they occupy in phospholipid liposomes.

It is proposed that choline-containing phospholipids facilitate entry of sterol molecules into the outer leaflet of cell surface membranes but that the phospholipid composition itself will not give rise to an asymmetric distribution of sterol in membranes with a high cholesterol content.     

Protein-induced formation of cholesterol-rich domains

A major protein of neuronal rafts, NAP-22, binds specifically to cholesterol. We demonstrate by circular dichroism that NAP-22 contains a significant amount of beta-structure that is not sensitive to binding of the protein to membranes, suggesting that a major portion of the protein does not insert deeply into the membrane. The free energy of binding of NAP-22 to liposomes of dioleoylphosphatidylcholine with 40% cholesterol is -7.3 +/- 0.5 kcal/mol. NAP-22 mixed with dipalmitoylphosphatidylcholine and 40% cholesterol partitions into the detergent insoluble fraction in the presence of 1% Triton X-100. NAP-22 also causes this insoluble fraction to become enriched in cholesterol relative to phospholipid, again demonstrating the ability of this protein to segregate cholesterol and phospholipids into domains. Differential scanning calorimetry results demonstrate that NAP-22 promotes domain formation in liposomes composed of cholesterol and phosphatidylcholine. This is shown by NAP-22-promoted changes in the shape and enthalpy of the phase transition of phosphatidylcholine as well as by the appearance of cholesterol crystallite transitions in membranes composed of phosphatidylcholine with either saturated or unsaturated acyl chains. In situ atomic force microscopy revealed a marked change in the surface morphology of a supported bilayer of dioleoylphosphatidylcholine with 0.4 mole fraction of cholesterol upon addition of NAP-22. Prior to the addition of the protein, the bilayer appears to be a molecularly smooth structure with uniform thickness. Addition of NAP-22 resulted in the rapid formation of localized raised bilayer domains. Remarkably, there was no gross disruption or erosion of the bilayer but rather simply an apparent rearrangement of the lipid bilayer structure due to the interaction of NAP-22 with the lipid. Our results demonstrate that NAP-22 can induce the formation of cholesterol-rich domains in membranes. This is likely to be relevant in neuronal membrane domains that are rich in NAP-22.     

Binding of antibiotic amphotericin B to lipid membranes: a 1H NMR study

The (1)H NMR technique was applied to study binding of AmB, an antifungal drug, to lipid membranes formed with egg yolk phosphatidylcholine. The analysis of (1)H NMR spectra of liposomes, containing also cholesterol and ergosterol (at 40 mol%), shows that AmB binds preferentially to the polar headgroups. Such a binding restricts molecular motion of the choline fragment in the hydrophilic region at the surface of liposomes but increases the segmental motional freedom in the hydrophobic core. The same effects are also observed in the sterol-containing membranes, except that the effect on the hydrophobic core was exclusively observed in the membranes containing ergosterol.     

Effect of streptolysin S on liposomes. Influence of membrane lipid composition on toxin action

The effect of the bacterial cytolytic toxin, streptolysin S, on liposomes composed of various phospholipids was investigated. Large unilamellar vesicles containing [14C]sucrose were prepared by reverse-phase evaporation, and membrane damage produced by the toxin was measured by following the release of labeled marker. The net charge of the liposomes had little or no effect on their susceptibility to steptolysin S and the toxin was about equally effective on liposomes composed of phosphatidylcholine, phosphatidylethanolamine, phosphatidylserine and phosphatidylglycerol. Experiments with liposomes composed of synthetic phospholipids showed that the ability of the toxin to produce membrane damage depended on the degree of unsaturation of the fatty acyl chains. The order of sensitivity was C18 : 2 phosphatidylcholine greater than C18: I phosphatidylcholine greater than C18 : 0 phosphatidylcholine = C16 : 0 phosphatidylcholine. Liposomes containing the latter two phospholipids were virtually unaffected by streptolysin S, and experiments with C18 : 0 phosphatidylcholine suggested that toxin activity does not bind to liposomes composed of phospholipids with saturated fatty acyl chains. The inclusion of 40 mol% cholesterol in C16 : 0 phosphatidylcholine and C18 : 0 phosphatidylcholine liposomes made these vesicles sensitive to streptolysin S. Egg phosphatidylcholine liposomes, which were unaffected at 0 degrees C and 4 degrees C became susceptible to the toxin at these temperatures when cholesterol was included. Liposomes composed of C14 : 0 phosphatidylcholine were unaffected by streptolysin S at temperatures below the chain-melting transition temperature (23 degrees C) of this phospholipid, but became increasingly susceptible above this temperature. The results suggest that the fluidity of the phospholipid hydrocarbon chains in the membrane is important in streptolysin S action.     

Modulation of the in vitro activity of lysosomal phospholipase A1 by membrane lipids

Lysosomal phospholipases play a critical role for degradation of cellular membranes after their lysosomal segregation. We investigated the regulation of lysosomal phospholipase A1 by cholesterol, phosphatidylethanolamine, and negatively-charged lipids in correlation with changes of biophysical properties of the membranes induced by these lipids. Lysosomal phospholipase A1 activity was determined towards phosphatidylcholine included in liposomes of variable composition using a whole-soluble lysosomal fraction of rat liver as enzymatic source. Phospholipase A1 activity was then related to membrane fluidity, lipid phase organization and membrane potential as determined by fluorescence depolarization of DPH, 31P NMR and capillary electrophoresis. Phospholipase A1 activity was markedly enhanced when the amount of negatively-charged lipids included in the vesicles was increased from 10 to around 30% of total phospholipids and the intensity of this effect depended on the nature of the acidic lipids used (ganglioside GM1相似文献   

Effect of streptolysin S on liposomes Influence of membrane lipid composition on toxin action

The effect of the bacterial cytolytic toxin, streptolysin S, on liposomes composed of various phospholipids was investigated. Large unilamellar vesicles containing [¹⁴C]sucrose were prepared by reverse-phase evaporation, and membrane damage produced by the toxin was measured by following the release of labeled marker. The net charge of the liposomes had little or no effect on their susceptibility to steptolysin S and the toxin was about equally effective on liposomes composed of phosphatidylcholine, phosphatidylethanolamine, phosphatidylserine and phosphatidylglycerol. Experiments with liposomes composed of synthetic phospholipids showed that the ability of the toxin to produce membrane damage depended on the degree of unsaturation of the fatty acyl chains. The order of sensitivity was C18 : 2 phosphatidylcholine > C18 : 1 phosphatidylcholine > C18 : 0 phosphatidylcholine = C16 : 0 phosphatidylcholine. Liposomes containing the latter two phospholipids were virtually unaffected by streptolysin S, and experiments with C18 : 0 phosphatidylcholine suggested that toxin activity does not bind to liposomes composed of phospholipids with saturated fatty acyl chains. The inclusion of 40 mol% cholesterol in C16 : 0 phosphatidylcholine and C18 : 0 phosphatidylcholine liposomes made these vesicles sensitive to streptolysin S. Egg phosphatidylcholine liposomes, which were unaffected at 0°C and 4°C became susceptible to the toxin at these temperatures when cholesterol was included. Liposomes composed of C14 : 0 phosphatidylcholine were unaffected by streptolysin S at temperatures below the chain-melting transition temperature (23°C) of this phospholipid, but became increasingly susceptible above this temperature. The results suggest that the fluidity of the phospholipid hydrocarbon chains in the membrane is important in streptolysin S action.     

Melittin-lipid bilayer interactions and the role of cholesterol

The membrane-destabilizing effect of the peptide melittin on phosphatidylcholine membranes is modulated by the presence of cholesterol. This investigation shows that inclusion of 40 mol % cholesterol in 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine or 1,2-dioleoyl-sn-glycero-3-phosphocholine liposomes reduces melittin's affinity for the membrane. It is significant that the presence of cholesterol does not increase the amount of membrane-associated melittin needed to cause maximum leakage from, or major structural rearrangements of, the liposomes. Furthermore, comparison of microscopy and leakage data suggests that melittin-induced leakage occurs via different mechanisms in the cholesterol-free and cholesterol-supplemented systems. In the absence of cholesterol, leakage of carboxyfluorescein takes place from intact liposomes in a manner compatible with the presence of small melittin-induced pores. In the presence of cholesterol, on the other hand, adsorption of the peptide causes complete membrane disruption and the formation of long-lived open-bilayer structures. Moreover, in the case of cholesterol-supplemented systems, melittin induces pronounced liposome aggregation. Cryotransmission electron microscopy was used, together with ellipsometry, circular dichroism, turbidity, and leakage measurements, to investigate the effects of melittin on phosphatidylcholine membranes in the absence and presence of cholesterol. The melittin partitioning behavior in the membrane systems was estimated by means of steady-state fluorescence spectroscopy measurements.     

Cholesterol-dependent insertion of glycosylphosphatidylinositol-anchored enzyme

Evidence is now accumulating that the plasma membrane is organized in different lipid and protein subdomains. Thus, glycosylphosphatidylinositol (GPI)-anchored proteins are proposed to be clustered in membrane microdomains enriched in cholesterol and sphingolipids, called rafts.By a detergent-mediated method, alkaline phosphatase, a GPI-anchored enzyme, was efficiently inserted into the membrane of sphingolipids- and cholesterol-rich liposomes as demonstrated by flotation in sucrose gradients. We have determined the enzyme extraluminal orientation. Using defined lipid components to assess the possible requirements for GPI-anchored protein insertion, we have demonstrated that insertion into membranes was cholesterol-dependent as the cholesterol addition increased the enzyme incorporation in simple phosphatidylcholine liposomes.     

Cholesterol-dependent insertion of glycosylphosphatidylinositol-anchored enzyme

Evidence is now accumulating that the plasma membrane is organized in different lipid and protein subdomains. Thus, glycosylphosphatidylinositol (GPI)-anchored proteins are proposed to be clustered in membrane microdomains enriched in cholesterol and sphingolipids, called rafts.By a detergent-mediated method, alkaline phosphatase, a GPI-anchored enzyme, was efficiently inserted into the membrane of sphingolipids- and cholesterol-rich liposomes as demonstrated by flotation in sucrose gradients. We have determined the enzyme extraluminal orientation. Using defined lipid components to assess the possible requirements for GPI-anchored protein insertion, we have demonstrated that insertion into membranes was cholesterol-dependent as the cholesterol addition increased the enzyme incorporation in simple phosphatidylcholine liposomes.     

Effect of liposomal phospholipid composition on cholesterol transfer between microsomal and liposomal vesicles.

Preincubation of rat liver microsomal vesicles at 37 degrees C in the presence of [3H]cholesterol/phospholipid liposomes results in a net transfer of cholesterol from liposomes to microsomal vesicles. This transfer follows first-order kinetics. For similar concentrations of the donor vesicles, rates of transfer are about 6-8 times lower with cholesterol/sphingomyelin liposomes compared with cholesterol/phosphatidylcholine liposomes. Also, transfer of cholesterol from cholesterol/sphingomyelin liposomes to microsomal vesicles reveals a larger activation energy than for the process from cholesterol/phosphatidylcholine liposomes. There is a significant correlation between the amount of liposomal cholesterol transferred to microsomal vesicles during preincubation and the increase found with acyl-CoA:cholesterol acyltransferase activity in these microsomes over their corresponding controls. If, however, liposomes made solely of phospholipids are substituted for the cholesterol/phospholipid liposomes in the preincubation system containing microsomal vesicles, then the acyl-CoA:cholesterol acyltransferase activity is decreased compared with the corresponding control system. Both sphingomyelin and phosphatidylcholine liposomes are equally effective in decreasing the enzyme activity. These results offer direct kinetic evidence for the positive correlation between cholesterol and sphingomyelin found in vivo in biological membranes.     

Transfer of cholesterol between liposomal membranes.

The transfer of cholesterol between liposomal membranes was examined. On incubation of liposomes compsoed of egg yolk phosphatidylcholine, phosphatidic acid and cholesterol (molar percentage, 65.8 : 1.3 : 32.9 or 65.5 : 6.3 : 31.2), almost complete equilibration of the cholesterol pools was achieved within 6 to 8 h at 37 degrees C. The rate of transfer of cholesterol from the liposomes, in which cholesterol was introduced by 'the exchange reaction', was not significantly different from that from liposomes prepared in the presence of cholesterol, in which the cholesterol was distributed homogenously. These findings indicate that half life for 'flip-flop' of cholesterol molecules in egg yolk phosphatidylcholine liposomes is less than 6 h at 37 degrees C. The transfer of cholesterol between liposomes was strongly dependent on temperature and was affected by the fatty acid composition of the phospholipid, suggesting that the 'fluidity' of the membranes strongly influences the transfer rate. A preferential distribution of cholesterol molecules was observed in heterogeneous liposomes with different classes of phospholipids. The 'affinity order' of cholesterol for phospholipid deduced from the present experiments is as follows: beef brain sphingomyelin greater than dipalmitoylglycerophosphocholine = dimyristoylglycerophosphocholine greater than egg yolk phosphatidylcholine.     

Fusion of Sendai virions or reconstituted Sendai virus envelopes with liposomes or erythrocyte membranes lacking virus receptors

Incubation of intact Sendai virions or reconstituted Sendai virus envelopes with phosphatidylcholine/cholesterol liposomes at 37 degrees C results in virus-liposome fusion. Neither the liposome nor the virus content was released from the fusion product, indicating a nonleaky fusion process. Only liposomes possessing virus receptors, namely sialoglycolipids or sialoglycoproteins, became leaky upon interaction with Sendai virions. Fusion between the virus envelopes and phosphatidylcholine/cholesterol liposomes was absolutely dependent upon the presence of intact and active hemagglutinin/neuraminidase and fusion viral envelope glycoproteins. Fusion between Sendai virus envelopes and phosphatidylcholine/cholesterol liposomes lacking virus receptors was evident from the following results. Anti-Sendai virus antibody precipitated radiolabeled liposomes only after they had been incubated with fusogenic Sendai virions. Incubation of N-4-nitrobenzo-2-oxa-1,3-diazole-labeled fusogenic reconstituted Sendai virus particles with phosphatidylcholine/cholesterol liposomes resulted in fluorescence dequenching. Incubation of Tb3+-containing virus envelopes with phosphatidylcholine/cholesterol liposomes loaded with sodium dipicolinate resulted in the formation of the chelation complex Tb3+-dipicolinic acid, as was evident from fluorescence studies. Virus envelopes fuse efficiently also with neuraminidase/Pronase-treated erythrocyte membranes, i.e. virus receptor-depleted erythrocyte membranes, although fusion occurred only under hypotonic conditions.     

Differential scanning calorimetry and 31P NMR studies on sonicated and unsonicated phosphatidylcholine liposomes.

1. Phase transitions in sonicated (vesicles) and unsonicated liposomes composed of various synthetic phosphatidylcholines are monitored using differential scanning calorimetry and 31P NMR. 2. The temperature (Tc), heat content and width of the phase transition are comparable in both vesicles and liposomes prepared from 1,2-dipalmitoyl phosphatidylcholine and 1,2-dimyristoyl phosphatidylcholine. In vesicles composed of a (1 : 1) mixture of 1,2-dipalmitoyl phosphatidylcholine and 1,2-dioleoyl phosphatidylcholine phase separation occurs as in the bilayers of the unsonicated liposomes. 3. The linewidth of the 31P resonances in vesicles is not greatly dependent upon the fatty acid composition when the lipids are in the disordered liquid crystalline state (above Tc). When the lipids are in the gel state (below Tc), however, there is a marked increase in linewidth, demonstrating a reduction in motion of the phosphate group. 4. The ratio of the amounts of phosphatidylcholine present in the outside and inside monolayter of the vesicle membrane was determined with 31P NMR using Nd3+ as a non-permeating shift reagent. 5. The outside/inside ratio is dependent upon the hydrocarbon chain length. Increasing chain length gives a lower outside/inside ratio and a larger vesicle. Introduction of cis or trans double bonds in the chain influences the outside/inside ratio slightly. 6. The incorporation of cholesterol decreases the outside/inside ratio and increases the size of 1,2-dimyristoyl phosphatidylcholine vesicles. The cholesterol concentration in the outside and inside monolayer is approximately the same. The size of the 1,2-dioleoyl phosphatidylcholine vesicles is also increased by cholesterol incorporation but the outside/inside distribution is also increased, especially between 30 and 50 mol% cholesterol. In these vesicles cholesterol is asymmetrically distributed and strongly prefers the inside monolayer of the vesicle.     

Asymmetry and transposition rates of phosphatidylcholine in rat erythrocyte ghosts.

Purified phospholipid exchange protein from beef heart cytosol is used to accelerate the exchange of phospholipids between labeled sealed ghosts and phosphatidylcholine/cholesterol liposomes. The purified protein accelerates the transfer of phosphatidylcholine and, to a lesser degree, that of sphingomyelin, phosphatidylinositol, and lysophosphatidylcholine. The presence of exchange protein does not accelerate the exchange of phospholipids between intact red blood cells and liposomes, but 75% of the phosphatidylcholine of sealed ghosts is readily available for exchange. The remaining 25% is also exchangeable but at a slower rate. When the exchange is assayed between inside-out vesicles and liposomes, 37% of the phosphatidylcholine is readily available, and 63% is exchanged at a slower rate. These results are consistent with an asymmetric distribution of phosphatidylcholine in isolated erythrocyte membrane fractions. The sum of the forward and backward transposition of phosphatidylcholine between the inside and outside layers of sealed ghost membranes amounts to 11% per hour, and the half-time for equilibration is 2.3 h. Significatnly lower values are obtained for the inside-out vesicles (half-time for equilibration: 5.3 h). These results suggest that, during the formation of the vesicles, the asymmetry of phosphatidylcholine is partially preserved, but structural changes occur in the membrane that affect the rate of membrane transposition of phosphatidylcholine.     

Letter to the editor: Fusion of Sendai viruses with model membranes

Sendai virus membranes fuse with liposomes containing phosphatidylcholine, cholesterol, sphingomyelin, phosphatidylethanolamine and gangliosides. After fusion the viral glycoprotein spikes are found in patches in the surface of the liposomes.     

Cholesterol-phospholipid interactions: The role of the 3β-position in membrane ordering and intermembrane exchange

The role of the 3β-hydroxy substituent of cholesterol in sterol-lipid interactions has been examined by incorporating the 3β-thiol analogue, thiocholesterol into egg phosphatidylcholine membranes. Thiocholesterol concentration reaches a maximum at 19% on a molar basis. The degree of phospholipid ordering, as judged by a cholestane spin probe, is significantly weaker than cholesterol but is concentration-dependent up to 20 mol%, a concentration that correlates well with that for the maximum thiocholesterol incorporation into liposomes. The apparent rate constants for exchange between liposomes and erythrocytes of cholesterol and thiocholesterol are indistinguishable. The results suggest a role of hydrogen bonding between the 3β-hydroxy group of cholesterol and phospholipids in determining the concentration and membrane ordering properties of cholesterol.     

Reconstitution of a porcine submaxillary gland beta-D-galactoside alpha 2----3 sialyltransferase into liposomes

Form A of the beta-D-galactoside alpha 2----3 sialyltransferase from porcine submaxillary glands was incorporated into liposomes. Incorporation was achieved by gel filtration of the enzyme in the presence of octylglucoside-phospholipid micelles. As detergent was removed during gel filtration, liposomes (average diameter, 370 A) with bound enzyme were formed and emerged unretarded from the column. The recovery of enzyme activity in the liposomes was about 40% of the initial activity starting with as little as 9 micrograms of transferase. Chromatography on Sepharose CL6B and sucrose density gradient centrifugation confirmed the association of enzyme with liposomes. In contrast to Form A, Form B of the sialyltransferase, which lacks the proposed lipid-binding domain of Form A, cannot be incorporated into liposomes. Form A of the transferase was also incorporated into liposomes composed of phosphatidylcholine, cholesterol, and a mixture of phospholipids from the membranes of the Golgi apparatus from porcine submaxillary glands. Although the transferase was distributed about equally on the internal and external surface of the phosphatidylcholine liposomes, most of the transferase was on the external surface in liposomes containing cholesterol (72%) or in liposomes containing Golgi apparatus phospholipids (88%). The enzyme bound to phosphatidylcholine liposomes was shown by kinetic analysis to have the same activity as that found in the presence of activity-stimulating detergents such as Triton X-100. Enzyme incorporated into cholesterol-containing liposomes had the same activity. In contrast, enzyme bound to liposomes formed from the Golgi apparatus mixed phospholipids had a lower activity, but one similar to that of the transferase in Golgi apparatus membranes. These studies suggest that the composition of a biological membrane may well influence the orientation of the transferase in the membrane as well as modulate its enzymic activity.     

Effects of membrane-stabilizing agents,cholesterol and cepharanthin,on radiation-induced lipid peroxidation and permeability in liposomes

Effects of two membrane-stabilizing agents, cholesterol and cepharanthin, on radiation-induced lipid peroxidation and membrane permeability were examined. Radiation-induced lipid peroxidation caused an increase in membrane permeability in phosphatidylcholine liposomes. The presence of cholesterol in liposomal membranes caused a decrease in the degree of membrane permeability in spite of an increased lipid peroxidation. On the other hand, cepharanthin suppressed both lipid peroxidation and the changes in permeability induced by radiation. The membrane-stabilizing effect of cholesterol against radiation-induced changes in permeability seemed to depend on the rigidification of membranes, which was estimated by ESR studies. Cepharanthin suppressed the degree of membrane permeability mainly by inhibiting the radiation-induced lipid peroxidation. However, cepharanthin did not exhibit a radical-trapping ability.