Interactions between a paramagnetic analogue of cholesterol and filipin

A paramagnetic analogue of cholesterol (called 25-doxyl-27-norcholesterol (CNO)), labeled near the w-end of the hydrophobic tail, was used to study interactions of cholesterol with filipin. We observed by electron microscopy that CNO- and cholesterol-filipin complexes are structurally equivalent. Two kinds of complexes were seen by ESR spectroscopy and electron microscopy, depending on the stoichiometric R ratio between the antibiotic and sterol. When R was high, an immobilized ESR spectrum appeared, showing strong imbrication between CNO and filipin. When R was nearer to unity, an exchange-broadened spectrum emerged, corresponding to a new phase that was very rich in CNO (a fast exchange between spins could occur by nearest contacts). CNO was easily displaced from its complex (i) by gradual addition of genuine cholesterol; and (ii) by an excess of phospholipids, owing to the very poor affinity of CNO (and cholesterol, by extension) for filipin in the lipidic phase. Almost no difference appeared between the ESR spectra of oriented samples, i.e. the probe showed no long-range order in any complex of CNO with filipin.     

Interaction of a spin-labelled cholesterol derivative with the cytochrome P-450scc active site

The cholesterol analogue 25-doxyl-27-nor-cholesterol (CNO), was found to be a substrate for cytochrome P-450scc. Upon incubation with the cytochrome P-450scc electron transfer system, CNO is transformed to pregnenolone (Km = 33 microM, Vmax = 0.32 min-1). The pregnenolone formation from endogenous cholesterol is strongly inhibited by CNO (50% at 5 microM). It binds tightly to cytochrome P-450scc as evidenced by a reversed type I spectral absorbance change (Kd = 5.9 microM) which is paralleled by a greater hyperfine splitting of the room-temperature CNO ESR spectrum due to an enhanced probe immobilization (Kd = 1.9 microM). This finding is in accord with a rotational correlation time of about 10(-7) s, which is close to the tumbling rate of the protein. At 110 K the CNO-bound cytochrome P-450scc displays the ESR g-values gx = 2.404/2.456, gy = 2.245 and gz = 1.916; these are different from those of cholesterol-liganded cytochrome P-450scc and may thus serve as a marker for cytochrome P-450scc. Our data indicate that the stereospecificity of the cytochrome P-450scc side-chain-cleaving activity is not dependent on the nature of the cholesterol side-chain termination (C25 to C27). The substrate binding site is however rather sensitive to a modification of the side chain. The doxyl ring confers a stronger affinity of the substrate to the enzyme. Upon binding it becomes embedded in the protein matrix, and we estimate that its final position is 0.6-1.0 nm from the heme moiety.     

Phosphatidylcholine liposomes containing cholesterol analogues with side chains of various lengths

The effect of the length of the side chain of sterols on their interaction with phosphatidylcholine was studied by measuring the permeability properties of liposomes constituted with sterol analogues with side chains of various lengths. The sensitivities of liposomes constituted with these sterol analogues toward digitonin and polyene antibiotics were also examined.The effects of sterols on phase transition of phosphatidylcholine were examined by measuring their effects on permeability increase due to perturbation of phase equilibrium and by differential scanning calorimetry. An analogue with a short side chain, isopropyl (C-22), had a very similar effect to cholesterol in suppressing the permeability increase, suggesting that the full length of the side chain is not necessary for this effect.The permeability of egg yolk phosphatidylcholine at 42°C was suppressed as much by the analogue C-22 as by cholesterol. Androstene-3-β-ol, an analogue without a side chain, however, had little suppressive effect. Thus it is concluded that the condensing effect of sterol requires a side chain, but not the full length of side chain.Liposomes constituted with analogues having a side chain with more than 5 carbon atoms showed maximum reactivity with a polyene antibiotic, amphotericin B, whereas those constituted with analogues having a side chain with less than 4 carbon atoms showed weaker reactivity. These findings indicate that a side chain with more than 5 carbon atoms is essential for the maximum interaction of liposomes with amphotericin B. Unlike amphotericin B, filipin reacted almost equally well with liposomes containing C-22 and with those containing cholesterol. Thus the chain length of the side chain of sterol is less important for interaction of liposomes with filipin than for their interaction with amphotericin B.Liposomes containing analogues having a side chain with more than 6 carbon atoms showed maximum reactivity with digitonin. Thus for the maximum interaction of liposomes with digitonin, the side chain of sterol should be longer than 6 carbon atoms.     

Interaction of the polyene antibiotics with lipid bilayer vesicles containing cholesterol

The interaction of the polyene antibiotics, amphotericin B, nystatin and filipin with cholesterol-containing single bilayer lipid vesicles has been characterized using gel permeation chromatography and proton magnetic resonance. All three antibiotics bind to vesicles at low concentrations without causing a large amount of vesicle destruction. The strength of binding as determined by gel permeation studies is greater for filipin and amphotericin than for nystatin. Nystatin and amphotericin B at these low concentrations induce a rapid loss of internal vesicle contents consistent with pore formation. Filipin induces no leakage beyond that expected from partial vesicle destruction or general detergent action.At antibiotic levels above 1 : 1 antibiotic : cholesterol ratios the NMR results show all three antibiotics to cause extensive vesicle destruction. The onset of this behavior, which appears to be independent of the total antibiotic concentration, indicates a well defined antibiotic : cholesterol interaction stoichiometry. Despite the fact that cholesterol is required for antibiotic activity, the NMR spectra prior to vesicle destruction show no changes indicative of an antibiotic-induced reversal of cholesterol restriction of phosphatidylcholine mobility. The contrast with polyene antibiotic behavior in more extended bilayers is discussed.     

Interaction of the polyene antibiotics with lipid bilayer vesicles containing cholesterol.

The interaction of the polyene antibiotics, amphotericin B, nystatin and filipin with cholesterol-containing single bilayer lipid vesicles has been characterized using gel permeation chromatography and proton magnetic resonance. All three antibiotics bind to vesicles at low concentrations without causing a large amount of vesicle destruction. The strength of binding as determined by gel permeation studies is greater for filipin and amphotericin than for nystatin. Nystatin and amphotericin B at these low concentrations induce a rapid loss of internal vesicle contents consistents consistent with pore formation. Filipin induces no leakage beyond that expected from partial vesicle destruction or general detergent action. At antibiotic levels above 1:1 antibiotic: cholesterol ratios the NMR results show all three antibiotics to cause extensive vesicle destruction. The onset of this behavior, which appears to be independent of the total antibiotic concentraion, indicates a well defined antibiotic : cholesterol interaction stoichiometry. Despite the fact that cholesterol is required for antibiotic activity, the NMR spectra prior to vesicle destruction show no changes indicative of an antibiotic-induced reversal of cholesterol restriction of phosphatidylcholine mobility. The contrast with polyene antibiotic behavior in more extended bilayers is discussed.     

The potential of fluorescent and spin-labeled steroid analogs to mimic natural cholesterol

Cholesterol analogs are often used to investigate lipid trafficking and membrane organization of native cholesterol. Here, the potential of various spin (doxyl moiety) and fluorescent (7-nitrobenz-2-oxa-1,3-diazol-4-yl (NBD) group) labeled cholesterol analogs as well as of fluorescent cholestatrienol and the naturally occurring dehydroergosterol to mimic the unique properties of native cholesterol in lipid membranes was studied in 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine (POPC) membranes by electron paramagnetic resonance, nuclear magnetic resonance, and fluorescence spectroscopy. As cholesterol, all analogs undergo fluctuating motions of large amplitude parallel to the bilayer normal. Native cholesterol keeps a strict orientation in the membrane with the long axis parallel to the bilayer normal. Depending on the chemical modification or the position of the label, cholesterol analogs may adopt an "up-side-down" orientation in the membrane or may even fluctuate between "upright" and up-side-down orientation by rotational motions about the short axis not typical for native cholesterol. Those analogs are not able to induce a comparable condensation of phospholipid membranes as known for native cholesterol revealed by 2H nuclear magnetic resonance. However, cholesterol-induced lipid condensation is one of the key properties of native cholesterol, and, therefore, a well suited parameter to assess the potential of steroid analogs to mimic cholesterol. The study points to extreme caution when studying cholesterol behavior by the respective analogs. Among seven analogs investigated, only a spin-labeled cholesterol with the doxyl group at the end of the acyl chain and the fluorophore cholestatrienol mimic cholesterol satisfactorily. Dehydroergosterol has a similar upright orientation as cholesterol and could be used at low concentration (about 1 mol %), at which its lower potential to enhance lipid packing density does not perturb membrane organization.     

Influence of PAF-acether on lecithin-oriented multibilayers monitored by ESR: interaction of PAF-acether with cholesterol

The influence of PAF-acether on natural ovolecithin oriented multibilayers is detected by ESR of intercalated 5 doxyl stearic acid. Simulation of lineshapes demonstrates an enlarged orientational distribution of the local director of the phospholipid phase and a small increase of the order parameter. The amount of PAF-acether required to destabilize the ovolecithin lamellar phase depends on the degree of hydration. By contrast, cholesterol displays an organizing effect on the PAF-acether phase. Simulation establishes a sharp orientational distribution of the local director when cholesterol reaches stoichiometric ratio relative to PAF-acether. At the same time, cholesterol increases the order parameter.     

Dynamic fluorescence quenching studies on lipid mobilities in phosphatidylcholine-cholesterol membranes

Bimolecular collision rate of 8-anilinonaphthalene-1-sulfonic acid (ANS) and the nitroxide doxyl group attached to various carbons on stearic acid spin labels (n-SASL) in phosphatidylcholine-cholesterol membranes in the fluid phase was studied by observing dynamic quenching of ANS fluorescence by n-SASL's. The excited-state lifetime of ANS and its reduction by the n-SASL doxyl group were directly measured by the time-correlated single photon counting technique to observe only dynamic quenching separately from static quenching and were analyzed by using Stern-Volmer relations. The collision rate of ANS with the n-SASL doxyl group ranges between 1 X 10(7) and 6 X 10(7), and the extent of dynamic quenching by n-SASL is in the order of 5-much much greater than 6- greater than 7- less than 9- less than 10- less than 12- less than 16-SASL (less than 5-SASL) in dimyristoylphosphatidylcholine (DMPC) membranes. Collision rate of 16-SASL is only 10% less than that of 5-SASL. Since the naphthalene ring of ANS is located in the near-surface region of the membrane, these results indicate that the methyl terminal of SASL appears in the near surface area frequently, probably due to extensive gauche-trans isomerism of the methylene chain. The presence of 30 mol% cholesterol decreases the collision rate of ANS with 12- and 16-SASL doxyl groups but not with the 5-SASL doxyl group in DMPC membranes. On the other hand, in egg-yolk phosphatidylcholine membranes, inclusion of 30 mol% cholesterol does not affect the collision of ANS with either 5-SASL or 16-SASL doxyl groups, in agreement with our previous observation that alkyl chain unsaturation moderates cholesterol effects on lipid motion in the membrane (Kusumi et al., Biochim. Biophys. Acta 854, 307-317). It is suggested that dynamic quenching of ANS fluorescence by lipid-type spin labels is a useful new monitor of membrane fluidity that reports on various lipid mobilities in the membrane; a class of motion can be preferentially observed over others by selecting a proper spin label, i.e., rotational diffusion of lipid about its long axis and translational diffusion by using 5-SASL, wobbling motion of the lipid long axis by using 7-SASL or androstane spin label, and gauche-trans isomerism by using 16-SASL.     

Rates of amphotericin B and filipin association with sterols. A study of changes in sterol structure and phospholipid composition of vesicles

The influence of structural modifications in sterols and phospholipids on the rate of polyene antibiotic-sterol interaction was studied. For filipin and amphotericin B association with sterols in vesicles, a preferential interaction was found with sterols whose side chain length is close to that of cholesterol. Introduction of trans double bonds into the sterol side chain did not alter the rate of interaction in vesicles. The delta 7-bond of the sterol appears to be of critical importance in amphotericin B-sterol interaction, whereas the delta 5-bond is not essential. These observations are relevant to the well-known effects of amphotericin B on cell membranes containing ergosterol compared with those containing cholesterol. The dependence of the rates of sterol-polyene antibiotic interaction on the phospholipid composition of the vesicles indicates that phospholipid vesicles may be an inadequate model for reaching a comprehensive understanding of the effects exerted on biological membranes by these agents.     

Interaction of polyene antibiotics with sterols in phosphatidylcholine bilayer membranes as studied by spin probes

Interaction of filipin and amphotericin B with sterols in phosphatidylcholine membranes has been studied using various spin probes; epiandrosterone, cholestanone, phosphatidylcholine with 12-nitroxide or 5-nitroxide stearate attached to 2 position and also with tempocholine at the head group. Filipin caused increase in the fluidity of cholesterol-containing phosphatidylcholine membranes near the center, while it rather decreased the fluidity near the polar surface. On the other hand, amphotericin B did not apparently affect the fluidity. In the electron spin resonance spectrum of steroid spin probes in the antibiotic-containing membranes, both bound and free signals were observed and the association constant was calculated from the siganal intensity. In the binding of steroids with filipin, both 3 and 17 positions were involved, while the 17 position was less involved in the binding with amphotericin B. Phase change in the host membrane markedly affected the interaction of filipin with epiandrosterone probe. The bound fraction jumped from 0.4 to 0.8 on going to the crystalline state and increased further with decrease in temperature. The overall splitting of the bound signal also increased on lowering the temperature below phase transition. This change was attributed to aggregate formation of filipin-steroid complexes in the crystalline state. On the other hand, effect of phase transition was much smaller on the interaction of amphotericin B with the steroid probe.     

Filipin as a flow microfluorometry probe for cellular cholesterol

The polyene antibiotic filipin, which forms specific complexes with 3 beta-hydroxysterols, displays spectral properties compatible with its use in flow microfluorometry (FMF). The purpose of this study was to test the suitability of filipin as an FMF probe for unesterified cellular cholesterol. The following experimental conditions appeared optimal for cells with an average unesterified cholesterol content of less than 3 nmol per 10(6) cells: 2 X 10(6) fixed cells (1-4% p-formaldehyde, 30 min, 21 degrees C) stained for 2-4 h with 100 micrograms/ml filipin and excited at 350.7/356.7 nm. Fluorescence emission (Em) was measured above 510 nm. Less suitable conditions involved excitation at 488 nm or using cells which had not been fixed. Fixation preserved the live-dead cell discrimination provided by forward light scatter measurements, so that dead cells could be excluded from the FMF analysis of cellular cholesterol. Under the above conditions FMF analysis of a variety of murine cell types showed that in all cases the fluorescence intensity of filipin-stained cells was clearly increased above autofluorescence levels of the unstained control cells. The increase in fluorescence signal in different filipin stained cell types correlated (P less than or equal to .001) with the cellular content of unesterified cholesterol determined by an independent enzymatic assay. The sensitivity of the FMF assay was in the femtomole (10(-15) ) range. Mixing experiments with cells of different cholesterol levels showed that the technique distinguishes cell populations with distinctive levels of unesterified cholesterol. We therefore concluded that filipin is a useful FMF probe for determining relative levels of unesterified cholesterol in cells.     

The fidelity of response by nitroxide spin probes to changes in membrane organization: The condensing effect of cholesterol

The response of doxyl fatty acid spin probes in egg lecithin bilayers to added cholesterol is compared with results from ²H-NMR. Large differences are found between the profiles of order parameter vs. label position and cholesterol concentration.At constant cholesterol content, the ESR spin probe order parameter decreases continuously as the label position is moved toward the terminal methyl region of the bilayer whereas an order parameter ‘plateau’ is observed for the upper region of the bilayer by ²H-NMR. In addition, the spin probe order parameters are smaller than those observed by ²H-NMR.Differences are also observed in the profiles of order parameter vs. cholesterol content for each label position. The spin probes detect a maximal response to added cholesterol for the central portion of the chains with much weaker responses near both ends of the chains. In contrast, the ²H-NMR results indicate a large, approximately constant response for the first ten positions in the chains with a decreasing response toward the terminal methyl group. For all the positions examined, the spin probes show a weaker response than that observed by ²H-NMR.A direct measure of the perturbing effect of a spin label is made by comparing the deuterium quadrupole splittings in egg lecithin-cholesterol bilayers for stearic acid with and without an attached doxyl moiety. The spin-labelled fatty acid has a much reduced quadrupole splitting and an opposite response to cholesterol addition.     

The effects of cholesterol on lateral diffusion and vertical fluctuations in lipid bilayers. An electron-electron double resonance (ELDOR) study.

Electron-electron double resonance (ELDOR) and saturation-recovery spectroscopy employing 14N:15N stearic acid spin-label pairs have been used to study the effects of cholesterol on lateral diffusion and vertical fluctuations in lipid bilayers. The 14N:15N continuous wave electron-electron double resonance (CW ELDOR) theory has been developed using rate equations based on the relaxation model. The collision frequency between 14N-16 doxyl stearate and 15N-16 doxyl stearate, WHex (16:16), is indicative of lateral diffusion of the spin probes, while the collision frequency between 14N-16 doxyl stearate and 15N-5 doxyl stearate, WHex (16:5), provides information on vertical fluctuations of the 14N-16 doxyl stearate spin probe toward the membrane surface. Our results show that: (a) cholesterol decreases the electron spin-lattice relaxation time Tle of 14N-16 doxyl stearate spin label in dimyristoylphosphatidylcholine (DMPC) and egg yolk phosphatidylcholine (egg PC). (b) Cholesterol increases the biomolecular collision frequency WHex (16:16) and decreases WHex (16:5), suggesting that incorporation of cholesterol significantly orders the part of the bilayer that it occupies and disorders the interior region of the bilayer. (c) Alkyl chain unsaturation of the host lipid moderates the effect of cholesterol on both vertical fluctuations and lateral diffusion of 14N-16 doxyl stearate. And (d), there are marked differences in the effects of cholesterol on lateral diffusion and vertical fluctuations between 0-30 mol% and 30-50 mol% of cholesterol that suggest an inhomogeneous distribution of cholesterol in the membrane.     

Effect of polyene antibiotics on the lectin-induced agglutination of transformed and untransformed cell lines.

Treatment of transformed Py3T3, SV101-3T3, and L1210 cells, as well as mitotic and Pronase-treated untransformed 3T3 cells, with the polyene antibiotics filipin, nystatin, and amphotericin B inhibited agglutination by wheat germ agglutinin. The effect of polyene antibiotic treatment was lectin and cell specific. Concanavalin A induced agglutination was not inhibited, wheat germ agglutination induced agglutination of untransformed 3T3 interphase cells was not influenced, and other aggregation phenomena, including those of erythrocytes with blood group specific antibodies or divalent cations, were unaffected by polyene treatments. This suggests that the formation of polyene-cholesterol complexes in transformed and erythrocyte cell membranes may specifically affect wheat germ agglutinin receptors and/or secondary events necessary for wheat germ agglutinin induced agglutination. Fluorescence studies of membrane filipin-cholesterol complexes showed that pretreating the cells with wheat germ agglutinin, but not concanavalin A, perturbed the fluorescence properties of filipin. Electron spin resonance studies with spin-labeled fatty acids revealed at best only a slight decrease in fatty acyl chain flexibility following filipin treatment. These studies indicate that there are not only quantitative differences between the agglutinability of transformed and untransformed cells with wheat germ agglutinin but that qualitative differences exist as well.     

Effects of Filipin on the Structure and Biological Activity of Enveloped Viruses

The interaction of the polyene antibiotic filipin with membrane-bound cholesterol in vesicular stomatitis (VS), influenza, and Rauscher leukemia virions was studied. Exposure of virions to filipin resulted in a series of depressions and ridges in the envelope of VS virions, with a periodicity of 15 to 20 nm perpendicular to the long axis of the particle; similar morphological alterations were observed in negatively stained preparations, in thin-sectioned virions, and in protease-treated virions that lack surface glycoproteins. This morphological effect was specific for filipin, since the envelopes of VS virions that had been treated with another polyene antibiotic, amphotericin B, exhibited markedly different morphology. Morphological alterations induced by filipin in influenza and Rauscher leukemia virions differed from those seen in VS virions. The infectivity of filipin-treated VS virions was reduced up to 500-fold, whereas influenza virions were resistant to filipin treatment. Incorporation of filipin into the virions was demonstrated, and no release of either lipids or proteins from virions was detected after filipin treatment. A stoichiometry of approximately 1 mol of bound filipin per mol of cholesterol was found in both intact and protease-treated VS virions. The equilibrium dissociation constant for filipin-cholesterol interaction was approximately 74-fold larger in intact than in protease-treated VS virions. The initial rate of association of filipin with cholesterol in intact virions was slower than that in protease-treated particles. The fluidity of lipids in VS viral membranes, as probed by a stearic acid derivative spin label, was markedly reduced when either intact or protease-treated virions were treated with filipin.     

Absorption and fluorescence spectra of polyene antibiotics in the presence of cholesterol.

The alterations in the absorption and fluorescence spectra observed for the polyene antibiotics filipin and nystatin in the presence of cholesterol are due to an exciton interaction (polyene aggregates) and cannot be attributed to a specific sterol-antibiotic complex. Filipin and nystatin molecules partition into the sterol aggregates, these structures being very efficient to induce exciton interaction; the observed splitting profile indicates that the chromophores are in a stacked arrangement (parallel transition dipoles). For filipin incorporated in lipid bilayers, the sterol is able to induce the same type of aggregate, at variance with nystatin.     

Acute effects of filipin on the plasmic, hepatic, and biliary cholesterol of the rat

Twenty-one male Wistar rats, 13 weeks old, were fed ad libitum hyperlipidic diets (28% fats) loaded with cholesterol (1.2%) for 5 weeks. One group of 11 rats was fed saturated fats (diet group "S") and another group of 10 rats was fed polyunsaturated fats (diet group "PU"). On the day they were sacrificed 10 of the rats were injected intravenously with 1 mg of filipin. Contrary to the rats in diet group "PU," the rats in diet group "S" treated with filipin presented certain characteristics that were not found in the nontreated group: They provided evidence of biliary cholestasis accompanied by a decline in the level of secretion of bile salts and phospholipids into bile. The concentrations of both free and esterified cholesterol in plasma fell and the amount of (esterified) hepatic cholesterol rose, although there was no change due to the filipin in the amounts of hepatic phospholipids. Explanatory hypotheses for these phenomena were considered, first, at the site of plasma membranes where filipin binds selectively to the cholesterol in the membrane, causing a disruption which probably disturbs the absorbance of circulating lipoproteins, especially that of hepatocyte cells, particularly in diet group "PU." Second, the effects of filipin on subcellular membranes seem to disturb the secretion of lipids and lipoproteins into bile and plasma, especially in diet group "S." Last, at the intracellular level, filipin appears to have a blocking effect on the organelles involved in biliary lipid secretion. The activity of certain enzymes such as cholesterol esterase may also be blocked, particularly in diet group "S," which would explain the accumulation of esterified cholesterol in liver.     

Filipin as a cholesterol probe. I. Morphology of filipin-cholesterol interaction in lipid model systems

We report some novel morphological observations on the interaction of the polyene antibiotic filipin (crude complex) with cholesterol, studied in non-cellular systems with replication, freeze-fracture, and negative stain techniques. Cholesterol crystals, lecithin liposomes containing 0 to 20 mole% of cholesterol, and liposomes containing 10 mole% of cholesterol and 5 to 40 mole% of sphingomyelin were incubated for varying lengths of time with filipin at different cholesterol: filipin molar ratios. The resulting filipin-induced lesions (FIL) were pleomorphic in all systems studied. In replicas of crystals, FIL appeared as ridges which were either straight, or curved into C- and S-shaped figures or closed circles. Negatively stained preparations showed FIL as white lines of the same configurations and in addition revealed a delicate veil attached to individual FIL. FIL, fused by their veils into clusters or large sheets ("holey sheets"), were shed from crystals. Incubation of liposomes for 1 h at cholesterol:filipin molar ratios of 4:1, 2:1, 1:1, and 1:5, demonstrated that cholesterol detection (i.e. formation of FIL) depend upon the ratio of cholesterol to filipin. At a 1:1 molar ratio FIL formed on liposomes containing 10 mole% cholesterol or more, but detectability increased to 5 mole% at the 1:5 ratio. Increasing the molar ratio of cholesterol:filipin to 2:1 and 4:1 decreased cholesterol detectability to between 10 and 20 mole%. Increasing concentrations of sphingomyelin decreased cholesterol detectability at the 1:1 cholesterol:filipin ratio; further, FIL in sphingomyelin-containing liposomes tended towards larger diameters. Filipin induced aggregation of liposomes and linked them together by holey sheets, providing evidence for filipin-induced extraction of cholesterol from liposomes. Taken together our morphological observations on filipin-cholesterol interaction in non-cellular systems raise pertinent questions as to the feasibility of filipin as a cholesterol probe in cellular systems.     

The use of amphotericin B to detect inhibitors of cellular cholesterol biosynthesis

Pores formed in the membranes of animal cells by complexes of sterols and the polyene antibiotic amphotericin B can efficiently kill the cells. Thus, in the absence of exogenous sources of cholesterol, inhibitors of enzymes in the cholesterol biosynthetic pathway render cells resistant to amphotericin B. Preincubation of Chinese hamster ovary cells with compactin or 25-hydroxycholesterol, inhibitors of the synthesis of the key intermediate mevalonate, protected cells from amphotericin B killing and this protection was reversed by the addition of exogenous mevalonate. The ability of compactin to confer amphotericin B resistance on normal cells was abolished when cells were provided exogenous cholesterol by the receptor-mediated endocytosis of low density lipoprotein. Low density lipoprotein receptor-defective Chinese hamster ovary cells were not subject to this low density lipoprotein-dependent amphotericin B killing. Exogenous mevalonate did not prevent 4,4,10 beta-trimethyl-trans-decal-3 beta-ol, an inhibitor of mevalonate conversion to sterols, from protecting cells from amphotericin B. A simple two-step protocol in which cells are preincubated (15-24 h) with potential inhibitors and then treated (3-6 h) with amphotericin B was devised to provide a sensitive method for detecting direct (e.g., competitive) and regulatory inhibitors of cholesterol biosynthesis. This protocol may prove useful in detecting potential antihypercholesterolemia drugs and is currently being used to isolate mutants in receptor-mediated endocytosis.