Amphotericin B channel conductance inactivation

Effects induced in bilayer lipid membranes by amphotericin B and its alkyl derivatives was analysed. Inactivation of the antibiotic-dependent multichannel membrane conductance was discovered. Kinetics of membrane conductivity was shown to depend on the antibiotic concentration in the membrane. At concentrations between 10(-8) and 10(-7) M, the resulting conductance appeared to the transient. We suggest that the phenomenon of biphasic kinetics of membrane conductance is the result of a consecutive transformation of polyene channels in the membrane: half-pores are assembled on either side of membrane-nonconducting 1; two half-pores combine to build up a conducting channels-conducting 2, and the conducting channels are disassemled to monomers and nonconducting self-associated forms inside the membrane-disassembled state (nonconducting 3). To explain the transient characteristics of the induced conductance, it is proposed that the antibiotic, present in the solution under self-associated form, binds the membrane and forms pores, then dissociates in the bilayer in a non-active monomeric form. The existence of definite monomers and nonconducting self-associated forms of amphotericin B molecules inside the membrane was estimated from the dependence of kinetic conductance of lipid membranes of amphotericin B and its alkyl derivatives, when the antibiotics are washed out from aqueous medium. Equilibrium between different antibiotic assemblies inside the membrane was demonstrated by the kinetics of conductance decrease following washing the antibiotic. Using circular dichroism measurements, we observed that amphotericin B alkyl derivatives were in self-associated form being susceptible to form pores across cholesterol-containing membranes. The phenomenon of biophasic kinetics was observed only in the cholesterol-containing membrane. The substitution of membrane cholesterol for ergosterol provides monotonic kinetics of membrane conductance at any antibiotic concentration.     

Amphotericin B channels in the bacterial membrane: role of sterol and temperature

Amphotericin B is an antibiotic that forms ion channels in the membrane of a host cell. The change in permeability produced by these channels is greatly improved by sterols; nevertheless, the single channel conductivity remains invariant. Hence, it is proposed that sterols do not act directly, but rather through the modulation of the membrane phase. We look at the formation of these channels in the bacterial membrane to determine the mechanism of its known antibiotic resistance. We found that channels can indeed be formed in this membrane, but a substantial amount of amphotericin B is required. We also study the effects of the antibiotic concentration needed for channel expression as well as the dynamics of channels affected by both sterol and temperature in phosphatidylcholine membranes. The results support the idea that membrane structure is a determining factor in the action of the antibiotic.     

The Properties of Ion Channels in Lipid Membranes Modified by the Aromatic Antibiotic Levorin А2

It has been shown that the main components of levorin A, that is, A₀, A₁, A₂, or A₃, that contain an aromatic group increase the permeability of membranes in the series A₃ > A₂ > A₁ > A₀ when they are on the same side of the membrane. All levorin components have cationic selectivity. The most studied levorin, А₂, promotes the almost ideal permeability of membranes to potassium ions. The membrane potential for a ten-fold change in the KCl concentration gradient is 56 ± 2 mV. It has been shown that the injection of the same concentration of levorin А₂ into one side of the membrane and then, after achieving the typical membrane permeability, into the other side of the membrane generates a two-fold increase in the total membrane permeability. This means that independent levorin-induced conductive semi-pores are formed on each side of the membrane. It has been found that the injection of levorin А₂ only into one side of the membrane enhances the membrane permeability to monosaccharides and other neutral molecules. The presence of levorin А₂ in cholesterol-, ergosterol-, and stigmasterol-containing phospholipid membranes has been shown to lead to the single-channel conductivity of typical ion channels of 0.2–0.5 pS. The properties of these channels have been studied. The levorin channels exist in two states, open and closed. Most of the time, the channel remains in the open state in the KBr solution. In solutions of different salts of the same concentration, the conductivity value of the levorin channels is approximately the same (0.4–0.5 pS). An increase in the dimethyl sulfoxide concentration in aqueous solutions facilitates the transition of polyene antibiotic molecules from dispersed to monomolecular form. The molecules of polyene antibiotics in the associated form exhibit high membrane activity.

     

Amphotericin B channels in phospholipid membrane-coated nanoporous silicon surfaces: implications for photovoltaic driving of ions across membranes

The antimycotic agent amphotericin B (AmB) functions by forming complexes with sterols to form ion channels that cause membrane leakage. When AmB and cholesterol mixed at 2:1 ratio were incorporated into phospholipid bilayer membranes formed on the tip of patch pipettes, ion channel current fluctuations with characteristic open and closed states were observed. These channels were also functional in phospholipid membranes formed on nanoporous silicon surfaces. Electrophysiological studies of AmB-cholesterol mixtures that were incorporated into phospholipid membranes formed on the surface of nanoporous (6.5 nm pore diameter) silicon plates revealed large conductance ion channels ( approximately 300 pS) with distinct open and closed states. Currents through the AmB-cholesterol channels on nanoporous silicon surfaces can be driven by voltage applied via conventional electrical circuits or by photovoltaic electrical potential entirely generated when the nanoporous silicon surface is illuminated with a narrow laser beam. Electrical recordings made during laser illumination of AmB-cholesterol containing membrane-coated nanoporous silicon surfaces revealed very large conductance ion channels with distinct open and closed states. Our findings indicate that nanoporous silicon surfaces can serve as mediums for ion-channel-based biosensors. The photovoltaic properties of nanoporous silicon surfaces show great promise for making such biosensors addressable via optical technologies.     

Effect of amphotericin B on cholesterol-containing liposomes of egg phosphatidylcholine and didocosenoyl phosphatidylcholine. A refinement of the model for the formation of pores by amphotericin B in membranes

(1) Binding and K+-permeability measurements were performed on egg and 22 : 1c/22 : 1c-phosphatidylcholine liposomes with or without cholesterol. (2) Amphotericin B binds specifically to cholesterol in both types of liposome despite the difference in bilayer thickness. (3) Addition of amphotericin B to one side of the cholesterol-containing egg phosphatidylcholine bilayers induces a fast K+ efflux from the outermost compartment of the liposomes. In contrast, the total K+ content of sonicated unilamellar cholesterol-containing egg phosphatidylcholine vesicles is released by amphotericin B. (4) Amphotericin B addition to one side of the cholesterol-containing 22 : 1c/22 : 1c-phosphatidylcholine liposomes does not cause a change in K+ permeability. The presence of amphotericin B on both sides of the bilayer, however, induces an increase in K+ permeability. (5) A model is proposed which accounts for the effect of bilayer thickness on the amphotericin B-induced permeability changes in membranes.     

Structural rearrangements induced by glycerol increase the permeability of bilayer lipid membranes for amphotericin

It has been shown that structural rearrangements induced by glycerol in bilayer lipid membranes (BLM) containing cholesterol facilitate the transmembrane transport of amphotericin B molecules in the direction of glycerol gradient. The addition of amphotericin B to the same side with glycerol results in a change in bilayer selectivity from the cation to the anion one. Besides, the final conductivity is blocked by tetraethylammonium from the solution with no amphotericin B added. It testifies to the transport of amphotericin molecules to the opposite side of the membrane. The transport effect depends on the cholesterol content in bilayer, ionic strength of the medium and slightly depends on temperature. It is concluded that transport of amphotericin B in such conditions differs from the diffusive one and is due to the formation of intermediate lipid phases in the course of structural rearrangements of bilayers.     

Complexes of Amphotericin B and Cholesteryl Sulfate

Abstract

To improve the clinical utility of amphotericin B, we have developed a novel formulation of amphotericin B, Amphocil® (also known as Amphotericin B Colloidal Dispersion, or ABCD). Amphocil is a uniform disc-shaped complex of amphotericin B and sodium cholesteryl sulfate in a molar ratio of 1:1. The complex has a mean hydrodynamic diameter of approximately 115 nm and is thermodynamically stable. In an extensive series of pharmacodynamic, pharmacokinetic and toxicology studies, Amphocil was found to be less toxic than conventional amphotericin B (Fungizone®), providing a four- to five-fold improvement in safety, while remaining effective in treating a variety of fungal infections. Plasma pharmacokinetics and tissue disposition of amphotericin B differ in several respects after administration of Amphocil and conventional amphotericin B, due to a rapid uptake of Amphocil by liver. Animals receiving Amphocil demonstrated reduced peak levels in plasma, prolonged residence time and lowered levels of amphotericin B in most tissues including the kidney, the major target organ for toxicity, compared with animals receiving conventional amphotericin B.

In healthy male subjects receiving a single dose of Amphocil, ranging from 0.25-1.5 mg/kg, mild to moderate dose-dependent acute side effects typically seen with conventional amphotericin B were observed but there was no sign of renal or hepatic toxicities. In two dose-escalating studies, multiple daily doses of Amphocil up to 4.5 mg/kg were well tolerated in patients who had previously failed to tolerate or respond to conventional amphotericin B. In addition, complete clearance of fungal infection was observed with the Amphocil therapy. Thus, Amphocil is a safe and effective agent for treating systemic mycoses. Toleration and efficacy of higher doses of Amphocil in patients with life-threatening mycoses is currently being evaluated.     

Amphotericin and model membranes. The effect of amphotericin B on cholesterol-containing systems as viewed by 2H-NMR

The interactions between amphotericin B and sterol-containing model membranes were monitored by 2H-NMR of deuterium-labelled dimyristoylphosphatidylcholine (DMPC), cholesterol and epicholesterol. The addition of amphotericin B to a cholesterol/DMPC (3:7) system was perceived differently by the lipid, depending upon the depth in the bilayer: no structural change was manifest in the acyl chain region associated with the plateau in molecular ordering (C4'), whereas the lipid clearly senses two environments near the center of the bilayer (C13', C14'). The amount as well as the ordering properties of the more ordered antibiotic-induced component, sensed at C14', increased with decreasing temperature. The structural parameters of deuterium-labelled cholesterol in cholesterol/DMPC mixtures were unchanged upon addition of amphotericin B, regardless of the bilayer depth. Upon addition of amphotericin B, the lipid T1 values are unchanged, whereas the T2 values are reduced by a factor of 2. The minimum in T1 observed for cholesterol in DMPC at 32-35 degrees C was shifted towards 38-40 degrees C in the presence of amphotericin B. Epicholesterol-containing dispersions of DMPC had properties similar to those of their cholesterol-containing analogs; a noticeable difference between the systems was an approx. 10% increase in the segmental order parameters on the addition of amphotericin B to the system containing the alpha-isomer of cholesterol. The concept of a dynamic complexation between amphotericin B and sterol is discussed.     

Effects of amphotericin B on membrane permeability--kinetics of spin probe reduction

The effect of the polyene antibiotic amphotericin B on the permeability of both unilamellar and multilamellar model membranes is investigated. The method measures the loss of the electron paramagnetic resonance signal of a spin probe, trapped in the aqueous compartment of a lipid dispersion, upon addition of ascorbate ions to the bulk aqueous phase. Amphotericin B causes large increases in the permeability of cholesterol-containing egg phosphatidylcholine membranes, whereas the effects are small in the absence of sterol and do not depend on surface charge. The effect of amphotericin depends upon the antibiotic:sterol mole ratio. The antibiotic appears to be unable to cross the membrane, acting only on the outermost bilayer of a multibilayer dispersion. When a phospholipid in the gel phase is used, amphotericin B causes large increases in permeability, independently of the presence or absence of sterol. It is suggested that the mechanism of action of amphotericin B is different for lipids in the liquid crystalline or gel states.     

A kinetics study of pig erythrocyte hemolysis induced by polyene antibiotics

The kinetics of the hemolysis induced by filipin is of the damage type, indicating the formation of large nonselective perforations of erythrocyte membranes. The process is relatively independent of the ionic composition of the incubation medium, and the differences between the hemolysis induced by filipin in pig and human erythrocytes are not significant. In a sucrose medium, filipin-induced hemolysis is inhibited in humans, whereas it is stimulated in pig erythrocytes. It is suggested that low ionic strength is the reason for the different modifications of complexation of filipin in pig and human erythrocyte membranes in a sucrose medium. The kinetics of the hemolysis induced in pig erythrocytes by amphotericin B and nystatin is of the permeability type, indicating the formation of selective channels in erythrocyte membranes and colloid osmotic hemolysis. The rate of the hemolysis, which is high in a KCl medium, is decreased in all the other media tested (CaCl2, MgCl2, potassium phosphate buffer, K2SO4, sucrose), although there are no changes in the kinetics of hemolysis. The results are interpreted as the formation of highly selective channels at a low concentration of the antibiotics. At increasing concentrations, channels of decreasing selectivity occur. The resistances of pig erythrocytes to amphotericin B and nystatin are lower than those of human erythrocytes.     

Effect of amphotericin B added to one side of a membrane

Amphotericin B which was added to the one side of a membrane in aqueous solutions with low pH 3.0 sharply increases the membrane conductance by the mechanism of channel forming. Long living ionic channels (20 sec) with large conductance dispersion (2-20 Ps) were found at the concentration 2 X 10(-8) M. Amphotericin B concentration which is necessary for obtaining single channels does not depend on cholesterol concentration in the membrane-forming solution in the range from 2 to 10 mg/ml. The effect of amphotericin B on the membrane depends on the phospholipid concentration in the membrane. The one-side selective permeability is the same as one under the symmetric action of the antibiotic. It is assumed that amphotericin B induced the membrane conductance by the mechanism of half-pores formation.     

Ion channels, induced by amphotericin B introduced unilaterally into lipid bilayers

Single ionic channels of approximately 10 pS in magnitude and approximately 100 ms duration (in 2 M KCl solution) were recorded when amphotericin B (AB) was added to one side of a lipid bilayer. Using blocking ions it has been shown that these channels are asymmetrical half-pores (similar to those postulated by Marty and Finkelstein) which are capable of forming long-living symmetrical pores if AB is added to both sides of the membrane.     

Polyene--sterol interaction and selective toxicity

From permeability experiments carried out with series of amphotericin B derivatives in both biological and model membranes, it was concluded that derivatives, whose carboxyl group at the C18 position is blocked by substitution, are much more efficient at inducing permeability in ergosterol-containing than in cholesterol-containing membranes, whereas derivatives whose carboxyl group is free and ionizable are equally efficient in both membranes types. Binding measurements on erythrocyte membranes showed that all amphotericin B derivatives simply partition between membrane lipids and aqueous medium, according to their lipid solubility. There is no relationship between binding and efficiency in inducing permeability. Permeability studies carried out on lipidic vesicles containing various sterols showed that: 1) derivatives having their carboxyl free induced permeability of the 'channel' type, regardless of the sterol present, and no detectable permeability in sterol-free membranes; 2) derivatives whose carboxyl group is blocked induce channels only in membranes containing ergosterol or sterols having an alkyl side chain identical to that of ergosterol. In the presence of other sterols or in sterol-free membranes, their ionophoric activity is poor and always of the 'mobile-carrier' type. A model of polyene-sterol interaction is proposed, accounting for the data obtained with both biological and model membranes.     

Formation of Two Different Types of Ion Channels by Amphotericin B in Human Erythrocyte Membranes

The polyene antibiotic amphotericin B (AmB) is known to form aqueous pores in lipid membranes and biological membranes. Here, membrane potential and ion permeability measurements were used to demonstrate that AmB can form two types of selective ion channels in human erythrocytes, differing in their interaction with cholesterol. We show that AmB induced a cation efflux (negative membrane polarization) across cholesterol-containing liposomes and erythrocytes at low concentrations (≤1.0 × 10⁻⁶ M), but a sharp reversal of such polarization was observed at concentrations greater than 1.0 × 10⁻⁶ M AmB, an indication that aqueous pores are formed. Cation-selective AmB channels are also formed across sterol-free liposomes, but aqueous pores are only formed at AmB concentrations 10 times greater. The effect of temperature on the AmB-mediated K⁺ efflux across erythrocytes revealed that the energies of activation for channel formation are negative and positive at AmB concentrations that lead predominantly to the formation of cation-selective channels and aqueous pores, respectively. These findings support the conclusion that the two types of AmB channels formed in human erythrocytes differ in their interactions with cholesterol and other membrane components. In effect, a membrane lipid reorganization, as induced by incubation of erythrocytes with tetrathionate, a cross-linking agent of the lipid raft–associated protein spectrin, led to differential changes in the activation parameters for the formation of both types of channels, reflecting the different lipid environments in which such structures are formed.     

Ion channels in mammalian proximal renal tubules.

In the plasma membranes of mammalian proximal renal tubules single ion channels were investigated mainly in isolated tubules perfused on one side, in isolated nonperfused (collapsed) tubules and in primary cell cultures. With these techniques, the following results were obtained: in the luminal membrane of isolated one-sided perfused tubules of rabbit and mouse S3 segments, K(+)-selective channels with single-channel conductance (g) of 33 pS and 63 pS, respectively, were recorded. In primary cultures of rabbit S1 segments, a small-conductance (42 pS) as well as a large-conductance (200 pS) K+ channel were observed. The latter was Ca2(+)- and voltage-sensitive. In cultured cells a Ca2(+)-activated, nonselective cation channel with g = 25 pS was also recorded. On the other hand, an amiloride-sensitive channel with g = 12 pS, which was highly selective for Na+ over K+, was observed in the isolated perfused S3 segment. In the basolateral membrane of isolated perfused S3 segments, two types of K+ channels with g = 46 pS and 36 pS, respectively, were observed. The latter channel was not dependent on cytosolic Ca2+ in cell-excised patches. A K+ channel with g = 54 pS was recorded in isolated nonperfused S1 segments. This channel showed inward rectification and was more active at depolarizing potentials. In isolated perfused S3 segments, in addition to the K+ channels also a nonselective cation channel with g = 28 pS was observed. This channel was highly dependent on cytosolic Ca2+ in cell-free patches. It can be concluded that the K+ channels both in the luminal and contraluminal cell membrane are involved in the generation of the cell potential. Na+ channels in the luminal membrane may participate in Na+ reabsorption, whereas the function of a basolateral cation channel remains unclear. Recently, single anion-selective channels were recorded in membranes of endocytotic vesicles, isolated from rat proximal tubules. Vesicles were enlarged by the dehydration/rehydration method and investigated with the patch clamp technique. The Cl- channel had a conductance of 73 pS, the current-voltage curve was linear and the channel inactivated at high negative clamp potentials. It is suggested that this channel is responsible for charge neutrality during active H+ uptake into the endosomes.     

Amphotericin B-phospholipid interactions responsible for reduced mammalian cell toxicity.

When interacting with phospholipid in an aqueous environment, amphotericin B forms unusual structures of markedly reduced toxicity (Janoff et al. (1988) Proc. Natl. Acad. Sci. USA 85, 6122-6126). These structures, which appear ribbon-like by freeze-fracture electron microscopy (EM), are found exclusively at amphotericin B to lipid mole ratios of 1:3 to 1:1. At lower mole ratios they occur in combination with liposomes. Circular dichroism (CD) spectra revealed two distinct modes of lipid-amphotericin B interaction, one for liposomes and one for the ribbon-like structures. In isolated liposomes, amphotericin B which comprised 3-4 mole percent of the bulk lipid was monomeric and exhibited a hemolytic activity comparable to amphotericin B suspended in deoxycholate. Above 3-4 mole percent amphotericin B, ribbon-like structures emerged and CD spectra indicated drug-lipid complexation. Minimal inhibitory concentrations for Candida albicans of liposomal and complexed amphotericin B were comparable and could be attributed to amphotericin a release as a result of lipid breakdown within the ribbon-like material by a heat labile extracellular yeast product (lipase). Negative stain EM of the ribbon-like structures indicated that the ribbon-like appearance seen by freeze-fracture EM arises as a consequence of the cross-fracturing of what are aggregated, collapsed single lamellar, presumably interdigitated, membranes. Studies examining complexation of amphotericin B with either DMPC or DMPG demonstrated that headgroup interactions played little role in the formation of the ribbon-like structures. With these results we propose that ribbon-like structures result from phase separation of amphotericin B-phospholipid complexes within the phospholipid matrix such that amphotericin B release, and thus acute toxicity, is curtailed. Formation of amphotericin B-lipid structures such as those described here indicates a possible new role for lipid as a stabilizing matrix for drug delivery of lipophilic substances, specifically where a highly ordered packing arrangement between lipid and compound can be achieved.     

Probing amphotericin B single channel activity by membrane dipole modifiers

The effects of dipole modifiers and their structural analogs on the single channel activity of amphotericin B in sterol-containing planar phosphocholine membranes are studied. It is shown that the addition of phloretin in solutions bathing membranes containing cholesterol or ergosterol decreases the conductance of single amphotericin B channels. Quercetin decreases the channel conductance in cholesterol-containing bilayers while it does not affect the channel conductance in ergosterol-containing membranes. It is demonstrated that the insertion of styryl dyes, such as RH 421, RH 237 or RH 160, in bilayers with either cholesterol or ergosterol leads to the increase of the current amplitude of amphotericin B pores. Introduction of 5α-androstan-3β-ol into a membrane-forming solution increases the amphotericin B channel conductance in a concentration-dependent manner. All the effects are likely to be attributed to the influence of the membrane dipole potential on the conductance of single amphotericin B channels. However, specific interactions of some dipole modifiers with polyene-sterol complexes might also contribute to the activity of single amphotericin B pores. It has been shown that the channel dwell time increases with increasing sterol concentration, and it is higher for cholesterol-containing membranes than for bilayers including ergosterol, 6-ketocholestanol, 7-ketocholestanol or 5α-androstan-3β-ol. These findings suggest that the processes of association/dissociation of channel forming molecules depend on the membrane fluidity.     

Lepidopteran-specific crystal toxins from Bacillus thuringiensis form cation- and anion-selective channels in planar lipid bilayers

Summary Previous studies in our laboratory have shown that CryIC, a lepidopteran-specific toxin from Bacillus thuringiensis, triggers calcium and chloride channel activity in SF-9 cells (Spodoptera frugiperda, fall armyworm). Chloride currents were also observed in SF-9 membrane patches upon addition of CryIC toxin to the cytoplasmic side of the membrane. In the present study the ability of activated CryIC toxin to form channels was investigated in a receptor-free, artificial phospholipid membrane system. We demonstrate that this toxin can partition in planar lipid bilayers and form ion-selective channels with a large range of conductances. These channels display complex activity patterns, often possess subconducting states and are selective to either anions or cations. These properties appeared to be pH dependent. At pH 9.5, cation-selective channels of 100 to 200 pS were most frequently observed. Among the channels recorded at pH 6.0, a 25–35 pS anion-selective channel was often seen at pH 6.0, with permeation and kinetic properties similar to those of the channels previously observed in cultured lepidopteran cells under comparable pH environment and for the same CryIC toxin doses. We conclude that insertion of CryIC toxin in SF-9 cell native membranes and in artificial planar phospholipid bilayers may result from an identical lipid-protein interaction mechanism.The assistance of A. Mazza and G.A.R. Mealing is gratefully acknowledged. The trypsin-activated, HPLC-purified CryIC toxin isolated from B. thuringiensis var. entomocidus crystal was a kind gift from M. Pusztai, Institute for Biological Sciences, NRC, Ottawa.     

Water permeability of gramicidin A-treated lipid bilayer membranes

In membranes containing aqueous pores (channels), the osmotic water permeability coefficient, P f, is greater than the diffusive water permeability coefficient, P d. In fact, the magnitude of P f/P d is commonly used to determine pore radius. Although, for membranes studied to date, P f/P d monotonically declines with decreasing pore radius, there is controversy over the value it theoretically assumes when that radius is so small that water molecules cannot overtake one another within the channel (single-file transport). In one view it should equal 1, and in another view it should equal N, the number of water molecules in the pore. Gramicidin A forms, in lipid bilayer membranes, narrow aqueous channels through which single-file transport may occur. For these channels we find that P f/P d approximately 5. In contrast, for the wider nystatin and amphotericin B pores, P f/P d approximately 3. These findings offer experimental support for the view that P f/P d = N for single-file transport, and we therefore conclude that there are approximately five water molecules in a gramicidin A channel. A similar conclusion was reached independently from streaming potential data. Using single-channel conductance data, we calculate the water permeability of an individual gramicidin A channel. In the Appendix we report that there is a wide range of channel sizes and lifetimes in cholesterol-containing membranes.     

The Water and Nonelectrolyte Permeability Induced in Thin Lipid Membranes by the Polyene Antibiotics Nystatin and Amphotericin B

Nystatin and amphotericin B increase the permeability of thin (<100 A) lipid membranes to ions, water, and nonelectrolytes. Water and nonelectrolyte permeability increase linearly with membrane conductance (i.e., ion permeability). In the unmodified membrane, the osmotic permeability coefficient, P_f, is equal to the tagged water permeability coefficient, (P_d)_w; in the nystatin- or amphotericin B-treated membrane, P_f/(P_d)_w ≈ 3. The unmodified membrane is virtually impermeable to small hydrophilic solutes, such as urea, ethylene glycol, and glycerol; the nystatin- or amphotericin B-treated membrane displays a graded permeability to these solutes on the basis of size. This graded permeability is manifest both in the tracer permeabilities, P_d, and in the reflection coefficients, σ (Table I). The "cutoff" in permeability occurs with molecules about the size of glucose (Stokes-Einstein radius 4 A). We conclude that nystatin and amphotericin B create aqueous pores in thin lipid membranes; the effective radius of these pores is approximately 4 A. There is a marked similarity between the permeability of a nystatin- or amphotericin B-treated membrane to water and small hydrophilic solutes and the permeability of the human red cell membrane to these same molecules.