Fluorescence study of the macrolide pentaene antibiotic filipin in aqueous solution and in a model system of membranes

Interaction of the antifungal ketoconazole and its diphenylphosphine derivatives with lipid bilayers: Insights into their antifungal action

Ketoconazole (Ke) is an important antifungal drug, and two of its diphenylphosphinemethyl derivatives (KeP: Ph2PCH2-Ke and KeOP: Ph2P(O)CH2-Ke) have shown improved antifungal activity, namely against a yeast strain lacking ergosterol, suggesting alternative modes of action for azole compounds. In this context, the interactions of these compounds with a model of the cell membrane were investigated, using POPC (1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine) large unilamellar vesicles and taking advantage of the intrinsic fluorescence of Ke, KeP and KeOP. Steady-state fluorescence spectra and anisotropy, including partition and aggregation studies, as well as fluorescence lifetime measurements, were carried out. In addition, the ability of the compounds to increase membrane permeability was assessed through carboxyfluorescein leakage. The membrane/water mole fraction partition coefficients (Kp,x): (3.31 ± 0.36) x105, (8.31 ± 1.60) x105 and (4.66 ± 0.72) x106, for Ke, KeP and KeOP, respectively, show that all three compounds have moderate to high affinity for the lipid bilayer. Moreover, KeP, and particularly KeOP interact more efficiently with POPC bilayers than Ke, which correlates well with their in vitro antifungal activity. Furthermore, although the three compounds disturb the lipid bilayer, KeOP is the quickest and most efficient one. Hence, the higher affinity and ability to permeabilize the membrane of KeOP when compared to that of KeP, despite the higher lipophilicity of the latter, points to an important role of Ph2P(O)CH2- oxygen. Overall, this work suggests that membrane interactions are important for the antifungal activity of these azoles and should be considered in the design of new therapeutic agents.

Polyene Antibiotics Physical Chemistry and Their Effect on Lipid Membranes; Impacting Biological Processes and Medical Applications

This review examined a collection of studies regarding the molecular properties of some polyene antibiotic molecules as well as their properties in solution and in particular environmental conditions. We also looked into the proposed mechanism of action of polyenes, where membrane properties play a crucial role. Given the interest in polyene antibiotics as therapeutic agents, we looked into alternative ways of reducing their collateral toxicity, including semi-synthesis of derivatives and new formulations. We follow with studies on the role of membrane structure and, finally, recent developments regarding the most important clinical applications of these compounds.

Sterol dynamics during endocytic trafficking in Arabidopsis

Sterols are lipids found in membranes of eukaryotic cells. Functions of sterols have been demonstrated for various cellular processes including endocytic trafficking in animal, fungal, and plant cells. The ability to visualize sterols at the subcellular level is crucial to understand sterol distribution and function during endocytic trafficking. In plant cells, the polyene antibiotic filipin is the most extensively used tool for the specific detection of fluorescently labeled 3-β-hydroxysterols in situ. Filipin can to some extent be used to track sterol internalization in live cells, but this application is limited, due to the inhibitory effects filipin exerts on sterol-dependent endocytosis. Nevertheless, filipin-sterol labeling can be performed on aldehyde-fixed cells which allows for sterol detection in endocytic compartments. This approach can combine studies correlating sterol distribution with experimental manipulations of endocytic trafficking pathways. Here, we describe step-by-step protocols and troubleshooting for procedures on live and fixed cells to visualize sterols during endocytic trafficking. We also provide a detailed discussion of advantages and limitations of both methods. Moreover, we illustrate the use of the endocytic recycling inhibitor brefeldin A and a genetically modified version of one of its target molecules for studying endocytic sterol trafficking.

Effects of central metal on the photophysical and photochemical properties of non-transition metal sulfophthalocyanine

The photophysical and photochemical properties and quenching (by 1,4-benzoquinone) of metallophthalocyanine sulfonates of aluminium ( AlPcSmix), zinc ( ZnPcSmix), silicon ( SiPcSmix), germanium ( GePcSmix) and tin ( SnPcSmix) are presented. The quantum yield values of fluorescence (ΦF), triplet state (ΦT), singlet oxygen (ΦΔ) and photodegradation (Φd) were determined and the observed trends in their variation among the complexes discussed in terms of aggregation and the heavy atom effect. 1,4-benzoquinone effectively quenched the fluorescence of the complexes. Quenching analyses gave positive deviations from Stern-Volmer behavior, suggesting the existence of static quenching in addition to dynamic quenching. The static and dynamic components of the quenching were separated using a modified Stern-Volmer equation and the “sphere of action quenching model”. The quenching constant was found to be a function of the radius of the central metal ion.

Real-time monitoring of membrane cholesterol reveals new insights into epidermal differentiation

Cholesterol is organized in distinctive liquid-ordered micro-domains within biological membranes called lipid rafts. These micro-domains direct multiple physiological functions in mammalian cells by modulating signaling processes. Recent findings suggest a role for lipid rafts in cellular processes in human keratinocytes such as early differentiation and apoptosis. However, research of lipid rafts is hindered by technological limitations in visualizing dynamic cholesterol organization in plasma membranes. This study addresses a real-time, non-invasive method for the long-term observation of cholesterol reorganization in plasma membranes. In addition, this study also addresses the dynamic process of cholesterol depletion and repletion in primary human keratinocytes. Cholesterol reorganization was measured by observed changes in cellular impedance. Disruption of lipid rafts with low concentrations of methyl-beta-cyclodextrin (MbetaCD) resulted in an increase in the proliferative capacity of keratinocytes, which was assessed using real-time proliferation curves and adenosine triphosphate (ATP)-based proliferation assays. Quantitative PCR showed a concomitant decrease in messenger RNA (mRNA) expression of the early differentiation markers keratins 1 and 10. Conversely, specific cholesterol reintegration led to a 4.5-fold increase in keratin 2 mRNA expression, a marker for late keratinocyte differentiation, whereas depletion resulted in a significant downregulation. These findings imply a strictly controlled mechanism for the regulation of membrane cholesterol composition in both early and terminal keratinocyte differentiation. The impedance-based method that this study addresses further enhances our understanding of how physiological processes in keratinocytes are controlled by membrane cholesterol.

Lipid membrane-induced optimization for ligand-receptor docking: recent tools and insights for the "membrane catalysis" model

Cells in living organisms are regulated by chemical and physical stimuli from their environment. Often, ligands interact with membrane receptors to trigger responses and Sargent and Schwyzer conceived a model to describe this process, "membrane catalysis". There is a notion that the physical organization of membranes can control the response of cells by speeding up reactions. We revisit the "membrane catalysis" model in the light of recent technical, methodological and theoretical advances and how they can be exploited to highlight the details of membrane mediated ligand-receptor interactions. We examine the possible effects that ligand concentration causes in the membrane catalysis and focus our attention in techniques used to determine the partition constant. The hypothetical diffusional advantage associated with membrane catalysis is discussed and the applicability of existing models is assessed. The role of in-depth location and orientation of ligands is explored emphasizing the contribution of new analysis methods and spectroscopic techniques. Results suggest that membranes can optimize the interaction between ligands and receptors through several different effects but the relative contribution of each must be carefully investigated. We certainly hope that the conjugation of the methodological and technical advances here reported will revive the interest in the membrane catalysis model.

Filipin orientation revealed by linear dichroism. Implication for a model of action

The organization of the polyene antibiotic filipin in membranes containing cholesterol is a controversial matter of debate. Two contradictory models exist, one suggesting a parallel and the other perpendicular organization of filipin with respect to the plane of the membrane. UV-vis linear dichroism, ATR-FTIR, and fluorescence anisotropy decay techniques were combined to study the orientation of filipin in model systems of membranes composed of 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine (POPC) or 1,2-palmitoyl-sn-glycero-3-phosphocholine (DPPC) with and without cholesterol. Filipin's orientation is determined by the presence/absence of cholesterol when it is inserted in gel crystalline phase model membranes. When cholesterol (33%) is present in DPPC bilayers, filipin stands perpendicular to the membrane surface as expected in "pore-forming" models. At variance, absence of cholesterol leaves filipin in an essentially random organization in the lipidic matrix. In liquid crystalline phase bilayers (POPC) filipin's orientation is perpendicular to the membrane surface even in absence of cholesterol. Thus filipin's activity/organization depends not only on cholesterol presence but also in the lipid phase domain it is inserted in. These findings were combined with spectroscopy and microscopy data in the literature, solving controversial matters of debate.

Cholesterol Oxidase Senses Subtle Changes in Lipid Bilayer Structure

We investigated the dependence of cholesterol oxidase catalytic activity and membrane affinity on lipid structure in model membrane bilayers. The binding affinities of cholesterol oxidase to 100-nm unilamellar vesicles composed of mixtures of DOPC or DPPC and cholesterol are not sensitive to cholesterol mole fraction if the phase of the membrane is in a fluid state. When the membrane is in a solid-ordered state, the binding affinity of cholesterol oxidase increases approximately 10-fold. The second-order rate constants (kcat*/Km*) for different lipid mixtures show a 2-fold substrate specificity for cholesterol in the l(d) phase of high cholesterol chemical activity over cholesterol in the l(o) phase. Moreover, the enzyme is 2-fold more specific for cholesterol in the l(o) phase than in the s(o) phase. Likewise, there is 2-fold substrate specificity for the high cholesterol chemical activity l(d) phase over the low chemical activity l(d) phase. The specificities for the l(d) phase of low cholesterol chemical activity and the l(o) phase are the same. These data indicate that the more ordered the lipid cholesterol structure in the bilayer, the lower the catalytic rate. However, under all of the conditions investigated, the enzyme is never saturated with substrate. The enzymatic activity directly reflects the facility with which cholesterol can move out of the membrane, whether changes in cholesterol transfer facility are due to phase changes or more localized changes in packing. We conclude that the activity of cholesterol oxidase is directly and sensitively dependent on the physical properties of the membrane in which its substrate is bound.

Intrinsic tyrosine fluorescence as a tool to study the interaction of the shaker B "ball" peptide with anionic membranes

Steady-state and time-resolved fluorescence from the single tyrosine in the inactivating peptide of the Shaker B potassium channel (ShB peptide) and in a noninactivating peptide mutant, ShB-L7E, has been used to characterize their interaction with anionic phospholipid membranes, a model target mimicking features of the inactivation site on the channel protein. Partition coefficients derived from steady-state anisotropy indicate that both peptides show a high affinity for anionic vesicles, being higher in ShB than in ShB-L7E. Moreover, differential quenching by lipophilic spin-labeled probes and fluorescence energy transfer using trans-parinaric acid as the acceptor confirm that the ShB peptide inserts deep into the membrane, while the ShB-L7E peptide remains near the membrane surface. The rotational mobility of tyrosine in membrane-embedded ShB, examined from the decay of fluorescence anisotropy, can be described by two different rotational correlation times and a residual constant value. The short correlation time corresponds to fast rotation reporting on local tyrosine mobility. The long rotational correlation time and the high residual anisotropy suggest that the ShB peptide diffuses in a viscous and anisotropic medium compatible with the aliphatic region of a lipid bilayer and support the hypothesis that the peptide inserts into it as a monomer, to configure an intramolecular beta-hairpin structure. Assuming that this hairpin structure behaves like a rigid body, we have estimated its dimensions and rotational dynamics, and a model for the peptide inserted into the bilayer has been proposed.

Quantifying molecular partition into model systems of biomembranes: an emphasis on optical spectroscopic methods

Optical spectroscopies have been intensively used to determine partition coefficients by a plethora of methodologies. The present review is intended to give detailed and useful information for the determination of partition coefficients and addresses several relevant aspects, namely: (i) definition and calculation of the partition coefficient between aqueous and lipidic phases; (ii) partition coefficients vs. "binding" formalisms; (iii) advantages of spectroscopic methodologies over separation techniques; (iv) formalisms for various experimental approaches based on UV-Vis absorption or fluorescence parameters (fluorescence intensity, lifetime, anisotropy and quenching); (v) experimental hints, artifacts and model limitations; and (vi) a brief survey of nonoptical techniques.

Cooperative partition model of nystatin interaction with phospholipid vesicles

Nystatin is a membrane-active polyene antibiotic that is thought to kill fungal cells by forming ion-permeable channels. In this report we have investigated nystatin interaction with phosphatidylcholine liposomes of different sizes (large and small unilamellar vesicles) by time-resolved fluorescence measurements. Our data show that the fluorescence emission decay kinetics of the antibiotic interacting with gel-phase 1,2-dipalmitoyl-sn-glycero-3-phosphocholine vesicles is controlled by the mean number of membrane-bound antibiotic molecules per liposome, <A>. The transition from a monomeric to an oligomeric state of the antibiotic, which is associated with a sharp increase in nystatin mean fluorescence lifetime from approximately 7-10 to 35 ns, begins to occur at a critical concentration of 10 nystatin molecules per lipid vesicle. To gain further information about the transverse location (degree of penetration) of the membrane-bound antibiotic molecules, the spin-labeled fatty acids (5- and 16-doxyl stearic acids) were used in depth-dependent fluorescence quenching experiments. The results obtained show that monomeric nystatin is anchored at the phospholipid/water interface and suggest that nystatin oligomerization is accompanied by its insertion into the membrane. Globally, the experimental data was quantitatively described by a cooperative partition model which assumes that monomeric nystatin molecules partition into the lipid bilayer surface and reversibly assemble into aggregates of 6 +/- 2 antibiotic molecules.

Real-time analysis of the effects of cholesterol on lipid raft behavior using atomic force microscopy

Cholesterol plays a crucial role in cell membranes, and has been implicated in the assembly and maintenance of sphingolipid-rich rafts. We have examined the cholesterol-dependence of model rafts (sphingomyelin-rich domains) in supported lipid monolayers and bilayers using atomic force microscopy. Sphingomyelin-rich domains were observed in lipid monolayers in the absence and presence of cholesterol, except at high cholesterol concentrations, when separate domains were suppressed. The effect of manipulating cholesterol levels on the behavior of these sphingomyelin-rich domains in bilayers was observed in real time. Depletion of cholesterol resulted in dissolution of the model lipid rafts, whereas cholesterol addition resulted in an increased size of the sphingomyelin-rich domains and eventually the formation of a single raftlike lipid phase. Cholesterol colocalization with sphingomyelin-rich domains was confirmed using the sterol binding agent filipin.

A photophysical study of the polyene antibiotic filipin. Self-aggregation and filipin--ergosterol interaction

Filipin, a macrolide polyene antibiotic, is known to interact selectively with ergosterol, a constituent of fungi membranes. In this work, the fluorescence resonance energy transfer (FRET) between a fluorescent analog of ergosterol, dehydroergosterol (DHE), and filipin was measured in small unilamellar vesicles of dipalmitoylphosphatidylcholine at 25 degrees C. The time-resolved FRET results were rationalized in the framework of the mean concentration model, and were complemented with steady-state fluorescence intensity, anisotropy and absorption measurements. The results point to the formation of both DHE--filipin aggregates (evidence from static quenching of DHE fluorescence by filipin) and filipin--filipin aggregates (evidence from: (i) the FRET acceptor concentration distributions; (ii) spectral changes of filipin absorption in the vesicles, the excitonic interaction suggesting a stack arrangement; (iii) filipin fluorescence self-quenching), even in presence of DHE and low antibiotic mole fractions (<1 mol%). These results point out that apparently contradictory biochemical models for the action of filipin (some based on the presence of sterols, others not) can be equally valid. Moreover, since results (ii) and (iii) are also observed when a sterol is present, both models of action can actually coexist in membranes with a low sterol content.

The pentaene macrolide antibiotic filipin prefers more rigid DPPC bilayers: a fluorescence pressure dependence study

Filipin is a pentaene macrolide antibiotic which was previously shown to incorporate more extensively into DPPC bilayers below the main phase transition temperature than above this temperature. This result was extremely unusual because drugs tend to be expelled from ordered gel phases. However, such results could not be safely attributed to the phase change of the bilayer itself because the temperature was changing concomitantly. In this work we changed the bilayer phase isothermally (53 degrees C) by hydrostatic pressure variation and discovered that filipin has a slightly more extensive incorporation in the pure DPPC gel phase (P>ca. 54.4 MPa): Kp,lc approximately 3x10(3) vs. Kp,gel approximately 6x10(3). The presence of sterols (45% molar ergosterol or cholesterol) caused an increase in the partition coefficients, regardless of pressure, ergosterol having a more pronounced effect (Kp approximately 2x10(4)-6x10(4)). Kp was pressure dependent in both cases, but mainly with cholesterol (Kp approximately 2x10(3)-2x10(4)). At variance with cholesterol, when ergosterol was used, no phase transition was detected. This difference cannot be due to a more extended uptake of filipin by cholesterol-containing membranes, and so must be due to specific interactions with cholesterol. In agreement with this finding, we discovered that filipin is more tightly packed (lower partial molar volume) in the cholesterol-rich phase than in the ergosterol-rich phase. Our results also point to a 2:1 DPPC:cholesterol stoichiometry in the cholesterol-rich phase (17% molar cholesterol). All partition coefficients were calculated from steady-state fluorescence anisotropy measurements.

Analysis of vertical fluorescence resonance energy transfer from the surface of a small-diameter sphere

Fluorescence resonance energy transfer (FRET) measurements have been used to analyze fluorophore separations in a number of varying geometries, including small particles and extended surfaces. This study focuses on the geometry created by a donor extended above the surface of a small sphere (radius < R0), where the acceptors are integrated into the sphere surface. The model of this geometry was based on an amphipathic molecule with its lipophilic region integrated into a detergent micelle and its hydrophilic region extending outward from the micelle surface, where the donor fluorophore is attached to the hydrophilic region of the molecule. Based on random acceptor incorporation into the micelle, a Poisson distribution was used to calculate the distribution of acceptor probes across the micelle population. The model converges to RET on a flat surface when the radius of the micelle exceeds 0.8 R0. The model was also used to simulate FRET data showing that the positions of donors above the micelle surface could be uniquely resolved. Experimental verification of the model was achieved in a sulfobetaine palmitate micelle with fluorescein isothiocyanate donors attached to detergent-solubilized lipopolysaccharide (LPS) and lipophilic Fast-DiI acceptors. The use of steady-state analysis allowed resolution of cases in which donors were located at different distances from the surface. Combining steady-state with excited-state lifetime analysis allowed resolution of cases where there was a combination of distances. Given the large number of biomolecules that interact with lipids, this approach may prove generally useful for defining molecular conformation.

Filipin-induced lesions in planar phospholipid bilayers imaged by atomic force microscopy

Filipin is a macrolide polyene with antifungal activity belonging to the same family of antibiotics as amphotericin B and nystatin. Despite the spectroscopy and electron microscopy studies of its interaction with natural membranes and membrane model systems, several aspects of its biochemical action, such as the role of membrane sterols, remain to be completely understood. We have used atomic force microscopy (AFM) to study the effect of filipin on dipalmitoylphosphatidylethanolamine bilayers in the presence and absence of cholesterol. The bilayers were prepared by Langmuir-Blodgett deposition over mica and imaged under water. It was shown that filipin-induced lesions could only be found in membranes with cholesterol. In close agreement with electron microscopy results, we have reported the presence of densely packed circular protrusions in the membrane with a mean diameter of 19 nm (corrected for convolution with AFM tip) and 0.4 nm height. Larger circular protrusions (90 nm diameter and 2.5 nm height) and doughnut-shaped lesions were also detected. These results demonstrate that filipin-induced lesions in membranes previously observed by electron microscopy are not biased by artifacts resulting from sample preparation. Filipin aggregates in aqueous solution could also be imaged for the first time. These polydisperse spherical structures were observed in samples with and without cholesterol.

Fluorescence quenching data interpretation in biological systems. The use of microscopic models for data analysis and interpretation of complex systems

In micro-heterogeneous media (e.g. membranes, micelles and colloidal systems), the fluorescence decay in the absence of quencher is usually intrinsically complex, e.g. due to the existence of several sub-populations with different micro-environments. In this case it is impossible to analyze data in detail (accounting for transient effects) and simpler formalisms are needed. The objective of the present work is to present and discuss such simpler formalisms. The goal is to achieve simple data analysis and meaningful, clear data interpretation in complex systems using microscopic models that consider several sub-populations of chromophores. Two points are dealt with in detail. (i) It is shown that the approximation of the transient effects by the quenching sphere-of-action model is not always possible. The quenching sphere-of-action concept can be regarded as a valuable tool, although crude, only in a limited range of experimental conditions, namely time resolution. (ii) The Stern-Volmer equation usually used for data analysis is only valid for a limited range of small and moderate equilibrium association constants, Ka, although this is frequently overlooked in the literature. Self-consistency criteria are presented for the proposed methods. The well-known downward curvature due to a fraction of fluorophores which is not accessible to the quencher is only a limiting case from a set of possible situations which result in deviations to linearity. A systematic classification of the different types of quenching is presented.

Filipin III: Configuration Assignment and Confirmation by Synthetic Correlation

The stereochemical configuration of filipin III (1) was determined using the (13)C acetonide analysis. The relative configurations for the nine stereogenic centers in the top half of filipin were initially identified using just three acetonide derivatives (2, 3, and 4) arising from a two-step protection sequence. The structure was confirmed by synthesis and direct correlation of degradation products 8 (C26-C28) and 10 (C1-C16). Filipin tetraacetonide 2 and triacetonide 4 each contain an anti acetonide in a highly unusual chair conformation. Molecular modeling successfully reproduced the preference for a chair conformation over the normally more stable twist-boat conformation.

The transverse location of the fluorescent probe trans-parinaric acid in lipid bilayers

The transverse location of trans-parinaric acid in spherical vesicles made up from dipalmitoylphosphatidylcholine has been investigated by the differential quenching of the probe fluorescence by 5- and 16-doxylstearic acid derivatives. The quenching data are interpreted in terms of a local fluorophore concentration factor. In this way it was found that the polyene of t-PnA is located within the inner part of the bilayer (presumably aligned with the bilayer lipids), both in the gel and in the liquid crystalline phases.

Association of polyene antibiotics with sterol-free lipid membranes: I. Hydrophobic binding of filipin to dimyristoylphosphatidylcholine bilayers

The interaction of filipin III with multilamellar vesicles (MLV) of dimyristoylphosphatidylcholine (DMPC ) was studied by four complementary methods leading to the following results: (1) The modifications of the filipin dichroic spectrum, by adding preformed fluid DMPC MLV, provide evidence of a saturable association with the stoichiometry DMPC/filipin = 4.2 +/- 0.5, constant between 24 and 35 degrees Celsius. (2) Thermograms obtained by differential scanning calorimetry (DSC) on mixtures where filipin is incorporated during the formation of MLV exhibit a high-temperature tail the more marked the higher the filipin content and some structures at temperatures which depend on this content. The corresponding evolution with the temperature of the CD spectra reveals that the characteristic bound filipin spectrum appears at the temperature at which a structure emerges. (3) Titration calorimetry measurements reveal that the association process is exothermic in the temperature range of the DSC endotherms in agreement with the filipin-induced ordering of the lipid chains, previously established by 2H-NMR in the same temperature range (Milhaud et al.(1989) Eur. Biophys. J. 17, 151-158). A discussion of the relevancy of this exothermicity to the hydrophobic effect is developed by referring to the paper by Wimley and White ((1993) Biochemistry 32, 6307-6312).

Filipin fluorescence quenching by spin-labeled probes: studies in aqueous solution and in a membrane model system

A detailed photophysical study of the fluorescence quenching (transient and steady state) of the macrolide antibiotic filipin by nitroxide-substituted fatty acids and a cholesterol derivative was carried out, aimed at determining its transverse position in a model system of membranes (multilamellar vesicles of dipalmitoylphosphatidylcholine). Filipin partitions efficiently into membranes (Kp = (5.0 +/- 1.0).10(3), 20 degrees C) and it was concluded that the antibiotic is buried in the membrane, away from the lipid-water interface. In addition, information on the organization of the quenchers was also obtained. The 5-nitroxide derivative of the fatty acid is essentially randomly distributed, while the 16-nitroxide is aggregated at concentrations higher than approximately 5% molar. For the cholesterol compound the results point to a phase separation at concentrations higher than 3% molar (below this limit concentration filipin associates with the derivatized sterol with KA = 20 M-1, assuming a 1:1 interaction). We propose that this phase separation and the aggregation state of filipin in the aqueous solution may be key processes in the antibiotic mode of action. A systematic and general approach to fluorescence quenching data analysis in complex (e.g., biochemical) systems is also presented.

Filipin and its interaction with cholesterol in aqueous media studied using static and dynamic light scattering

Aggregation of filipin in aqueous medium and filipin-induced changes in cholesterol micelles have been studied using intensity and dynamic light scattering. The dependencies of filipin aggregate dimensions on concentration, solvent, and temperature were studied, and revealed that the aggregates do not have a well-defined geometry, i.e., a critical micelle concentration cannot be detected and stable structures are not formed. The aggregates are of size Rg approximately 110 nm and Rh approximately 63 nm, referring to the radius of gyration and hydrodynamic radius, respectively. In the concentration range studied (1 microM < C < 30 microM), a low molecular weight species (monomer/dimer) is always present together with the aggregates. In ethanol/water mixtures, large (Rg approximately 500 nm), narrow distribution aggregates are formed in the water volume fraction range 0.45 < phi H2O < 0.65. Aggregation also occurs on changing the temperature; In the range 7-37 degrees C, smaller aggregates (10-30 nm form and the process is only partially reversible. No pronounced effect of filipin on the structure of the cholesterol micelles was observed (a small increase in Rg and Rh is noted). These results rule out any "specificity" for the filipin interactions with cholesterol, which has been considered a key event in the filipin biochemical mode of action. A reevaluation of this question is suggested and some alternatives are advanced.