Search

Candida albicans and human erythrocytes were treated with liposomal amphotericin B (AmB) obtained by incubation of free AmB with small unilamellar vesicles (SUV) composed of unsaturated fatty acyl chains phosphatidylcholine (egg-yolk PC). Cellular effects were determined by changes in the K+ internal content of cells and in the number of colonies formed by fungal cells or as hemolysis, measured as a decrease in haemoglobin retention by erythrocytes. Dose-response curves were obtained by two procedures: either the ratio of AmB to phospholipids was kept constant over the AmB concentration range used (R = 10(-2] or the phospholipid concentration was kept constant (C = 0.2 mM) and the concentration of AmB varied. The liposomal preparations of AmB were as active against fungi as AmB in dimethylsulfoxide but less active (internal K+ changes) or inactive (hemolysis) against erythrocytes. On the other hand the binding of AmB to the SUV, as a function of the AmB concentration, was monitored by circular dichroism, fluorescence and UV absorption, in the two conditions used for the cellular studies. The amount of AmB bound when the total concentration of antibiotic was 2.10(-7) M was very low but increased with concentration and reached 90% at 10(-5) M. In all the assays we used, the anticellular effects could be attributed to the levels of AmB remaining free (unbound to the lipids). The variations of these levels with total AmB concentration could therefore explain the increased selectivity of liposomal AmB in toxicity against fungi and erythrocytes as compared to that of AmB added as a solution in dimethylsulfoxide. Indeed fungal cells are sensitive to low concentrations of AmB in dimethylsulfoxide; at these concentrations, in liposomal preparations, AmB is not bound to phospholipids and therefore as active as if added in dimethylsulfoxide. By contrast erythrocytes are only sensitive to much higher concentrations of AmB in dimethylsulfoxide; at these concentrations AmB is almost totally bound to phospholipids and therefore much less active.

The interaction of the polyene antibiotic, filipin, with individual or mixed plant sterols (stigmasterol, sitosterol, campesterol and 24-methylpollinastanol) incorporated into large unilamellar vesicles (LUV) of soybean phosphatidylcholine (PC) as well as the filipin interaction with purified membrane fractions from maize roots containing these sterols was investigated by ultraviolet (UV) absorption and and circular dichroism (CD) spectroscopy. With both types of membrane preparation, dramatic changes in the UV absorption and CD spectra of the antibiotic were evidenced. When LUV containing stigmasterol, sitosterol and/or campesterol were incubated with low filipin concentrations (i.e., for filipin/sterol molar ratios (rst) lower than 1), CD signal characteristic of the formation of filipin-sterol complexes were observed. At higher rst values, the filipin-sterol interaction was shown to be in competition with a filipin-phospholipid interaction. With 24-methylpollinastanol-containing LUV, the filipin-phospholipid interaction was detected even at rst values lower than 1, which suggests a lower affinity of filipin for this sterol and emphasizes the structural differences between delta 5-sterols and 9 beta,19-cyclopropylsterols. With sterol-free soybean PC LUV, a filipin-phospholipid interaction could also be evidenced. With maize root cell membranes containing either delta 5-sterols or 9 beta,19-cyclopropylsterols, CD spectra similar to those obtained in the presence of LUV having these sterols as components were observed. Thus, the protein component of the membranes does not appear to be an important feature.

The interaction of phorbol 12,13-dibutyrate (PDBu), 12-O-retinoylphorbol 13-acetate (RPA) and 12-O-tetradecanoylphorbol 13-acetate (TPA) with L-alpha-phosphatidylserine-containing small unilamellar vesicles or erythrocyte ghosts was monitored by circular dichroism (CD). No change in the CD spectra of PDBu was observed upon binding, while RPA and TPA spectra were slowly affected by the interaction. The changes in RPA and TPA spectra were assigned to the embedding of these molecules in the membrane bilayers. In the presence of 10(8) cells/ml, after one minute incubation, about 2 to 5% of the amount of phorbol ester added is embedded in the membrane. It is suggested that either phorbol esters entering the membrane is not a prerequisite for protein kinase C activation or the amount of phorbol esters necessary to activate protein kinase C is very small.

The interactions of unilamellar lipid vesicles with vacidin A, an aromatic heptaene antibiotic, and with amphotericin B, a nonaromatic heptaene antibiotic were compared. Uptake of both antibiotics, monitored by circular dichroism, was found to be faster with small vesicles than with large ones. By combining permeability measurements (Gary-Bobo and Cybulska, J. Antibiotics 35, 1068 (1982)) and circular dichroism spectra, we found that for vacidin A, the same permeability inducing species is formed regardless of vesicles size. However, at a given concentration of antibiotic, less of the permeability inducing species is formed in the presence of small vesicles than in the presence of large vesicles. This may account for the differences between small and large vesicles in antibiotics-induced permeability. For amphotericin B, the permeability inducing species formed in the presence of small vesicles differs from that formed in the presence of large vesicles.

The effect of the polyene antibiotic etruscomycin on the permeability of large unilamellar lipid vesicles was investigated. Proton leakage was induced in egg-yolk phosphatidylcholine (EPC) vesicles only when sterol was present in the membrane; the extent of leakage was limited. High etruscomycin/lipid ratios (R) were necessary (R greater than 0.1). Higher percentages of sterol increased the permeability, slightly more strongly for ergosterol than for cholesterol. Dipalmitoylphosphatidylcholine (DPPC) vesicles were more sensitive to permeability inducement, even in the absence of sterol in the bilayer (inducement for R greater than 0.06). The interactions of etruscomycin with the vesicles were examined by circular dichroism, fluorescence and 31P-NMR. In the range of antibiotic concentration where permeability was induced, R greater than 0.1 for EPC vesicles, R greater than 0.06 for DPPC vesicles, etruscomycin exhibited characteristic circular dichroism spectra independent of the presence of sterol. Under the same conditions, 31P-NMR and fluorescence studies indicated a destruction or a fusion of the vesicle bilayer. At lower etruscomycin concentrations (R less than 0.03), the etruscomycin circular dichroism spectra were different, indicating that the interaction with membranes containing ergosterol differed from that with membranes containing cholesterol. From correlating the increase in fluorescence intensity with this interaction, as well as from exchange experiments, it was inferred that etruscomycin at a low antibiotic/lipid ratio is more strongly bound to ergosterol-containing vesicles than to cholesterol-containing vesicles. These results and their comparison with the results obtained with other polyene antibiotics indicate that at low R etruscomycin resembles amphotericin rather than filipin in its preferential binding to ergosterol-containing vesicles. At higher R, that is in conditions where permeability is induced, the selectivity is different. The corresponding mechanism seems not to involve the formation of an etruscomycin-sterol channel, since the hydrophobic chain of the complex would be too short to form a channel.

The selective toxicity of the polyene antibiotic amphotericin B between pathogenic eukaryotic organisms and animal cells has often been said to originate in the presence of ergosterol in fungal membranes instead of cholesterol, found in membranes of animal cells. We have tested this hypothesis by measuring the proton efflux induced by amphotericin B in egg yolk phosphatidylcholine small unilamellar vesicles. By measuring circular dichroism under the same conditions, we monitored the interaction of the antibiotic and its conformational changes. Sterol-free vesicles are sensitive to amphotericin B, but the sensitivity of sterol-containing vesicles is always greater and increasingly so with increasing sterol concentration. Ergosterol-containing vesicles are more sensitive than cholesterol-containing vesicles. On the other hand, numerous amphotericin B conformers can be detected in sterol-containing vesicles, depending upon both the concentration of sterol and the amphotericin B sterol ratio. It appears that one conformer, or maybe two at high amphotericin B concentration, is responsible for the induced permeability. From their circular dichroism spectra, these two conformers are the same in the presence of ergosterol or cholesterol. The concentration of amphotericin B necessary to obtain the two conformers is higher with cholesterol than with ergosterol, which agrees with the permeability results.

The amphotericin B (AmB) - ergosterol complex, formed by interaction of the antibiotic with ergosterol-containing phospholipid vesicles, is associated with the lipid bilayer. It has been shown by circular dichroism studies that the AmB-ergosterol complex formed in water-propanol binary mixtures has a similar structure to that observed in phospholipid vesicles. A positive cooperativity is found for the interaction of AmB with ergosterol. The similar AmB-cholesterol complex is much less stable and rearranges rapidly to a different conformation.

Circular dichroism (CD) of the antifungal antibiotic amphotericin B (AmB) can be used to characterize the liposomal preparations of the drug with regard to the levels of drug bound to the lipids. The very intense dichroic doublet centered around 340 nm of free amphotericin B in water or the dichroism observed above 435 nm can be used to determine the percentages of bound AmB and free AmB in preparations containing high antibiotic/lipid ratios (ranging from 10(-2) to 10(-1] used in these carrier systems. Examples are given for AmB in the presence of small unilamellar vesicles prepared from four saturated fatty acyl chain phosphatidylcholines of different chain lengths, with or without cholesterol. The transfer of AmB from vesicles to two blood components, serum albumin, and lipoproteins can also be monitored by CD under particular conditions.

Amphotericin B (AmB) was shown to induce a Ca2+ influx across ergosterol- and cholesterol-containing large unilamellar liposomes, by following spectrophotometrically the formation of the Arsenazo III-Ca2+ complex. At equivalent antibiotic concentrations the Ca2+ influx was much more extensive through ergosterol-containing membranes (almost 100% with 1 microM AmB, 160 microM lipid) than through cholesterol-containing membranes (below 0.5 microM the influx of Ca2+ was negligible). In the presence of ergosterol-containing membranes the initial rate of Ca2+ influx had the same linear dependence on the ratio antibiotic/lipid whatever the lipid concentration, which was not the case in cholesterol-containing membranes. These results suggest that the channels responsible for the AmB-induced Ca2+ permeability across cholesterol- and ergosterol-containing liposomes have different structures.

The interaction of the antitumor compound adriamycin with negatively charged unilamellar phospholipid vesicles was studied. The negative charges were provided by cardiolipin or phosphatidic acid. By analyzing the changes in the circular dichroism spectrum of adriamycin, we demonstrated the presence of two different spectral patterns corresponding to two different binding sites (I and II) on the vesicles. In site I, the amino sugar of adriamycin is bound to the ionized phosphate of either cardiolipin or phosphatidic acid, and the dihydroxyanthraquinone lies outside the bilayer. In site II, the amino sugar is still bound to the phosphate, but the dihydroxyanthraquinone moiety is embedded in the bilayer. This has been shown by measuring spectroscopically the binding of the aglycon part to an external probe and by measuring the susceptibility of bound adriamycin to reduction by NADH dehydrogenase.