Functional reconstitution of influenza virus envelopes.

We have examined several procedures for the reconstitution of influenza virus envelopes, based on detergent removal from solubilized viral membranes. With octylglucoside, no functionally active virosomes are formed, irrespective of the rate of detergent removal: in the final preparation the viral spike proteins appear predominantly as rosettes. Protein incorporation in reconstituted vesicles is improved when a method based on reverse-phase evaporation of octylglucoside-solubilized viral membranes in an ether/water system is employed. However, the resulting vesicles do not fuse with biological membranes, but exhibit only a non-physiological fusion reaction with negatively charged liposomes. Functional reconstitution of viral envelopes is achieved after solubilization with octaethyleneglycol mono(n-dodecyl)ether (C12E8), and subsequent detergent removal with Bio-Beads SM-2. The spike protein molecules are quantitatively incorporated in a single population of virosomes of uniform buoyant density and appear on both sides of the membrane. The virosomes display hemagglutination activity and a strictly pH-dependent hemolytic activity. The virosomes fuse with erythrocyte ghosts, as revealed by a fluorescence resonance energy transfer assay. The rate and the pH dependence of fusion are essentially the same as those of the intact virus. The virosomes also fuse with cultured cells, either at the level of the endosomal membrane or directly with the cellular plasma membrane upon a brief exposure to low pH.     

Reconstitution and fusogenic properties of Sendai virus envelopes

Sendai virus membranes were reconstituted by detergent dialysis, using the non-ionic detergents Triton X-100 and octyl glucoside. Membrane reassembly was determined by measuring the surface-density-dependent efficiency of resonance energy transfer between two fluorescent phospholipid analogues, which were co-reconstituted with the viral envelopes. The functional incorporation of the viral proteins was established by monitoring the ability of the reconstitution products to fuse with erythrocyte membranes, utilizing assays based on either resonance energy transfer or on relief of fluorescence selfquenching. The persistent adherence of residual Triton X-100 with the reconstituted membrane was revealed by an artificial detergent-effect on the resonance energy transfer efficiency and the occurrence of hemolysis of human erythrocytes under conditions where fusion does not occur. Properly reconstituted Sendai virus envelopes were obtained with octyl glucoside. The fusion activity of the viral envelopes was dependent on the initial concentration of octyl glucoside used to disrupt the virus and the rate of detergent removal. Rapid removal of detergent by dialysis against large volumes of dialysis buffer (ratio 1:850) or by gel filtration produced reconstituted membranes capable of inducing hemagglutination but significant fusion activity was not detected. By decreasing the volume ratio of dialysate versus dialysis buffer to 1:250 or 1:25, fusogenic viral envelopes were obtained. The initial fusion kinetics of the reconstituted viral membrane and the parent virus were different in that both the onset and the initial rate of fusion of the reconstituted membranes were faster, whereas the extents to which both particles eventually fused with the target membrane were similar. The differences in the initial fusion kinetics lead us to suggest that the details of the fusion mechanism between Sendai virus and the target membrane involve factors other than the mere presence of glycoproteins F and HN in the viral bilayer. Finally, the results also indicate that determination of the viral fusion activity in a direct manner, rather than by an indirect assay, such as hemolysis, is imperative for a proper evaluation of the functional properties retained upon viral reconstitution.     

Fusion of Sendai virus with vesicles of oligomerizable lipids: a microcalorimetric analysis of membrane fusion

Sendai virus fuses efficiently with small and large unilamellar vesicles of the lipid 1,2-di-n-hexadecyloxypropyl-4- (beta-nitrostyryl) phosphate (DHPBNS) at pH 7.4 and 37 degrees C, as shown by lipid mixing assays and electron microscopy. However, fusion is strongly inhibited by oligomerization of the head groups of DHPBNS in the bilayer vesicles. The enthalpy associated with fusion of Sendai virus with DHPBNS vesicles was measured by isothermal titration microcalorimetry, comparing titrations of Sendai virus into (i) solutions of DHPBNS vesicles (which fuse with the virus) and (ii) oligomerized DHPBNS vesicles (which do not fuse with the virus), respectively. The observed heat effect of fusion of Sendai virus with DHPBNS vesicles is strongly dependent on the buffer medium, reflecting a partial charge neutralization of the Sendai F and HN proteins upon insertion into the negatively-charged vesicle membrane. No buffer effect was observed for the titration of Sendai virus into oligomerized DHPBNS vesicles, indicating that inhibition of fusion is a result of inhibition of insertion of the fusion protein into the target membrane. Fusion of Sendai virus with DHPBNS vesicles is endothermic and entropy-driven. The positive enthalpy term is dominated by heat effects resulting from merging of the protein-rich viral envelope with the lipid vesicle bilayers rather than by the fusion of the viral with the vesicle bilayers per se.     

Reconstitution of the fusogenic activity of vesicular stomatitis virus.

Enveloped virus glycoproteins exhibit membrane fusion activity. We have analysed whether the G protein of vesicular stomatitis virus, reconstituted into liposomes, is able to fuse nucleated cells in a pH-dependent fashion. Proteoliposomes produced by octylglucoside dialysis did not exhibit cell fusion activity of the G protein. However, by making use of n-dodecyl octaethylene monoether (C12E8) as the solubilizing agent and by removal of the detergent in two steps, we were able to produce fusogenic G protein liposomes. These G protein liposomes fuse to the BHK-21 cell surface at pH 5.7-6.0 with an efficiency of fusion comparable with that of the parent virus. Physical and chemical analysis revealed that the fusogenic liposomes exhibited a protein to lipid weight ratio of 0.67 and showed an average diameter of 130 nm.     

Membrane fusion activity of reconstituted vesicles of influenza virus hemagglutinin glycoproteins

Reconstituted vesicles of hemagglutinin glycoproteins into egg yolk phosphatidylcholine/spin-labeled phosphatidylcholine/cholesterol (molar ratio 1.6:0.4:1) were prepared by dialysis. Preparations at appropriate protein-to-lipid ratios (1:44 and 1:105 mol/mol) contained vesicles with a diameter of 100-300 nm and a high density of spikes on the surface. These vesicles showed low pH-induced membrane fusion activity. At pH 5.2 and 37 degrees C, fusion with erythrocyte membranes took place very rapidly within 1-2 min and reached a plateau at 63-66% fusion. The fusion was negligibly small at neutral pH and was induced to occur at pH values lower than 6.0. The reconstituted vesicles caused hemolysis and fusion of human erythrocyte cells in the same pH range as that of the fusion with erythrocyte membranes. The low pH-induced fusion activity of the reconstituted vesicles is essentially the same as that of the parent virus. These vesicles can be used to deliver some reagents or drugs into target cell cytoplasm via fusion at lysosomes.     

Electrostatic control by lipids upon the membrane-bound (Na+ + K+)-ATPase. II. The influence of surface potential upon the activating ion equilibria

Sarcoplasmic reticulum Ca2+-ATPase has been reconstituted with excess lipid (25-150 g egg phosphatidylcholine per g sarcoplasmic reticulum protein) by a procedure combining the use of a non-ionic detergent with cholate dialysis. The reconstituted vesicles were analyzed by sucrose density fractionation and freeze-fracture electron microscopy. At the lowest lipid to protein ratios some vesicles containing aggregated protein were observed. At a lipid to protein ratio of 150:1 (w/w) only 30-40% of the reconstituted protein sedimented through 7% (w/v) sucrose. The remainder of the latter preparation was characterized by a high Ca2+-uptake capacity and a coupling ratio of 1.6 mol Ca2+ transported per mol ATP hydrolyzed. Intramembranous particles in this preparation occurred isolated in the membrane. In most cases only one particle could be seen on a fracture face. Cross-linking with cupric phenanthroline indicated that protein-protein contacts were drastically reduced by reconstitution. It is concluded that aggregation of intramembranous particles is not required for optimal Ca2+-transport function. The dispersed preparation obtained by a combined reconstitution and sucrose density fractionation procedure is useful for further characterization of the Ca2+ pump.     

Kinetics of pH-dependent fusion between influenza virus and liposomes

The pH-dependent fusion between influenza virus and liposomes (large unilamellar vesicles) has been investigated as a model for the fusion step in the infectious entry of the virus into cells. Fusion was monitored continuously, with a fluorescence assay based on resonance energy transfer (RET) [Struck, D. K., Hoekstra, D., & Pagano, R. E. (1981) Biochemistry 20, 4093-4099], which allows an accurate quantitation of the fusion process. Evidence is presented indicating that the dilution of the RET probes from the liposomal bilayer into the viral membrane is not due to transfer of individual lipid molecules. The initial rate and final extent of the fusion reaction increase dramatically with decreasing pH, fusion being virtually complete within 1 min at pH 4.5-5.0. From experiments in which the ratio of virus to liposomes was varied, it is concluded that virus-liposome fusion products continue to fuse with liposomes, but not with virus. Fusion is most efficient with liposomes consisting of negatively charged phospholipids, while phosphatidylcholine and sphingomyelin are inhibitory. The reaction is completely blocked by an antiserum against the virus and inhibited by pretreatment of the virus with trypsin. The effect of proteolytic pretreatment at pH 7.4 is enhanced after preincubation of the virus at pH 5.0, consistent with the occurrence of a low pH induced, irreversible, conformational change in the viral fusion protein, the hemagglutinin (HA), exposing trypsin cleavage sites. When, after initiation of the fusion reaction at pH 5.0, the pH is readjusted to neutral, the process is arrested instantaneously, indicating that the low pH induced conformational change in the HA protein, in itself, is not sufficient to trigger fusion activity.     

Apocytochrome c induces pH-dependent vesicle fusion

The ability of apocytochrome c and the heme containing respiratory chain component, cytochrome c, to induce fusion of phosphatidylcholine (PC) small unilamellar vesicles containing 0-50 mol % negatively charged lipids was examined. Both molecules mediated fusion of phosphatidylserine (PS):PC 1:1 vesicles as measured by energy transfer changes between fluorescent lipid probes in a concentration- and pH-dependent manner, although cytochrome c was less potent and interacted over a more limited pH range than the apocytochrome c. Maximal fusion occurred at pH 3, far below the pKa of the 19 lysine groups contained in the protein (pI = 10.5). A similar pH dependence was observed for vesicles containing 50 mol % cardiolipin (CL), phosphatidylglycerol (PG), and phosphatidylinositol (PI) in PC but the apparent pKa values varied somewhat. In the absence of vesicles, the secondary structure of apocytochrome c was unchanged over this pH range, but in the presence of negatively charged vesicles, the polypeptide underwent a marked conformational change from random coil to alpha-helix. By comparing the pH dependencies of fusion induced by poly-L-lysine and apocytochrome c, we concluded that the pH dependence derived from changes in the net charge on both the vesicles and apocytochrome c. Aggregation could occur under conditions where fusion was imperceptible. Fusion increased with increasing mole ratio of PS. Apocytochrome c did induce some fusion of vesicles composed only of PC with a maximum effect at pH 4. Biosynthesis of cytochrome c involves translocation of apocytochrome c from the cytosol across the outer mitochondrial membrane to the outer mitochondrial space where the heme group is attached. The ability of apocytochrome c to induce fusion of both PS-containing and PC-only vesicles may reflect characteristics of protein/membrane interaction that pertain to its biological translocation.     

Parameters affecting fusion between Sendai virus and liposomes. Role of viral proteins, liposome composition, and pH

A kinetic and quantitative characterization of the fusion process between Sendai virus and phospholipid vesicles is presented. Membrane fusion was monitored in a direct and continuous manner by employing an assay which relies on the relief of fluorescence self-quenching of the probe octadecylrhodamine B chloride which was located in the viral membrane. Viral fusion activity was strongly dependent on the vesicle lipid composition and was most efficient with vesicles solely consisting of acidic phospholipids, particularly cardiolipin (CL). This result implies that the fusion of viruses with liposomes does not display an absolute requirement for specific membrane receptors. Incorporation of phosphatidylcholine (PC), rather than phosphatidylethanolamine (PE), into CL bilayers strongly inhibited fusion, suggesting that repulsive hydration forces interfere with the close approach of viral and target membrane. Virus-liposome fusion products were capable of fusing with liposomes, but not with virus. In contrast to fusion with erythrocyte membranes, fusion between virus and acidic phospholipid vesicles was triggered immediately, did not strictly depend on viral protein conformation, and did not display a pH optimum around pH 7.5. On the other hand, with vesicles consisting of PC, PE, cholesterol, and the ganglioside GD1a, the virus resembled more closely the fusogenic properties that were seen with erythrocyte target membranes. Upon decreasing the pH below 5.0, the viral fusion activity increased dramatically. With acidic phospholipid vesicles, maximal activity was observed around pH 4.0, while with GD1a-containing zwitterionic vesicles the fusion activity continued to increase with decreasing pH down to values as low as 3.0.(ABSTRACT TRUNCATED AT 250 WORDS)     

Functional reconstitution of the integral membrane proteins of influenza virus into phospholipid liposomes

The integral membrane proteins of influenza virus, a hemagglutinin and a neuraminidase, have been incorporated into liposomes composed of either phosphatidylcholine or a mixture of phosphatidylcholine and phosphatidylethanolamine (2:1 w/w) using detergent dialysis. The virus spike glycoproteins for reconstitution were selectively solubilized by using cetyltrimethylammonium bromide to leave a "core particle", which lacked a lipid bilayer but possessed quaternary structure as observed by electron microscopy. The viral spike proteins were combined with exogenous phospholipid in excess sodium cholate followed by exhaustive dialysis for 150 h. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis showed that only the viral glycoproteins were associated with all the complexes formed. The level of sodium cholate remaining after dialysis was shown to be reduced to less than 1 molecule per 80 protein molecules. Viral proteins reconstituted into dimyristoylphosphatidylcholine liposomes were shown to have retained hemagglutination, low-pH-dependent hemolysis, and neuraminidase activities and were associated with a lipid bilayer in two types of complexes with average lipid to protein mole ratios after sucrose density gradient purification of either 590:1 or 970:1. The bilayer vesicles formed were of similar sizes and were shown by negative-stain electron microscopy to be 150-300 nm in diameter with well-defined spikes on their surface. Reconstituted liposomes of dimyristoylphosphatidylcholine were found to be unstable with respect to their trapped volume and therefore were unsuitable for fusion studies, unlike complexes formed with phosphatidylcholine or a mixture of phosphatidylcholine/phosphatidylethanolamine derived from hen eggs.(ABSTRACT TRUNCATED AT 250 WORDS)     

pH-dependent fusion of reconstituted vesicular stomatitis virus envelopes with Vero cells. Measurement by dequenching of fluorescence

Reconstituted vesicular stomatitis virus (VSV) envelopes were formed by solubilization of the viral envelope with Triton X-100 followed by removal of detergent by direct addition of SM2 biobeads. We provide direct demonstration of fusion of reconstituted VSV with cells using fluorescent lipid and aqueous probes incorporated into the VSV virosomes during reconstitution. We show a direct comparison of the kinetics and pH profile of fusion with cells between reconstituted VSV and fluorescently labeled intact virus. With this preparation it is now possible to gain additional information about the role of cooperativity in viral protein-mediated fusion, and to permit construction of efficient vehicles for delivery of drugs and other materials into cells.     

Incorporation of bacterial membrane proteins into liposomes: factors influencing protein reconstitution

Meningococcal and gonococcal outer membrane proteins were reconstituted into liposomes using detergent-mediated dialysis. The detergents octyl glucopyranoside (OGP), sodium cholate and Empigen BB were compared with respect to efficiency of detergent removal and protein incorporation. The rate of OGP removal was greater than for cholate during dialysis. Isopycnic density gradient centrifugation studies showed that liposomes were not formed and hence no protein incorporation occurred during dialysis from an Empigen BB containing reconstitution mixture. Cholate-mediated reconstitution yielded proteoliposomes with only 75% of the protein associated with the vesicles whereas all of the protein was reconstituted into the lipid bilayer during OGP-mediated reconstitution. Essentially complete protein incorporation was achieved with an initial protein-to-lipid ratio of 0.01:1 (w/w) in the reconstitution mixture; however, at higher initial protein-to-lipid ratios (0.02:1) only 75% protein incorporation was achieved. Reconstituted proteoliposomes were observed as large (>300 nm), multilamellar structures using cryo-electron microscopy. Size reduction of these proteoliposomes by extrusion did not result in significant loss of protein or lipid. Extruded proteoliposomes were unilamellar vesicles with mean diameter of about 100 nm.     

Characterization of the unidirectional transport of carnitine catalyzed by the reconstituted carnitine carrier from rat liver mitochondria

The carnitine carrier from rat liver mitochondria was purified by chromatography on hydroxyapatite and celite and reconstituted in egg yolk phospholipid vesicles by adsorbing the detergent on polystyrene beads. In the reconstituted system, in addition to the carnitine/carnitine exchange, the purified protein catalyzed a uni-directional transport (uniport) of carnitine measured as uptake into unloaded proteoliposomes as well as efflux from prelabelled proteoliposomes. In both cases the reaction followed a first-order kinetics with a rate constant of 0.023-0.026 min-1. Besides carnitine, also acylcarnitines were transported in the uniport mode. N-Ethylmaleimide inhibited the uni-directional transport of carnitine completely. The uniport of carnitine is not influenced by the delta pH and the electric gradient across the membrane. The activation energy for uniport was 115 kJ/mol and the half-saturation constant on the external side of the proteoliposomes was 0.53 mM. The maximal rate of the uniport at 25 degrees C was 0.2 mumol/min per mg protein, i.e. about 10 times lower than that of the reconstituted carnitine transport in exchange mode.     

Low pH-induced fusion of liposomes with membrane vesicles derived from Bacillus subtilis

We have investigated the pH-dependent interaction between large unilamellar phospholipid vesicles (liposomes) and membrane vesicles derived from Bacillus subtilis, utilizing a fluorescent assay based on resonance energy transfer (RET) (Struck, D. K., Hoekstra, D., and Pagano, R. E. (1981) Biochemistry 20, 4093-4099). Efficient interaction occurs only with negatively charged liposomes, containing cardiolipin or phosphatidylserine, as revealed by the dilution of the RET probes from the liposomal bilayer into the bacterial membrane. The initial rate of fluorophore dilution increases steeply with decreasing pH. The interaction involves a process of membrane fusion, as indicated by the proportional transfer of cholesteryl-[1-14C]oleate, 14C-labeled egg PC, and the RET probes from the liposomes to the bacterial vesicles, the formation of interaction products with an intermediate buoyant density, and the appearance of colloidal gold, initially encapsulated in the liposomes, in the internal volume of fused structures as revealed by thin-section electron microscopy. Treatment of B. subtilis vesicles with trypsin strongly inhibits the fusion reaction, indicating the protein dependence of the process. Vesicles derived from Streptococcus cremoris or from the inner membrane of Escherichia coli also show low pH-dependent fusion with liposomes. The fusion process described in this paper may well be of considerable importance to studies on the mechanisms of membrane fusion and to studies on the structure and function of bacterial membranes. In addition, the fusion reaction could be utilized to deliver foreign substances into bacterial protoplasts.     

Kinetic characterization of the reconstituted tricarboxylate carrier from rat liver mitochondria

The tricarboxylate carrier from rat liver mitochondria was purified by chromatography on hydroxyapatite/celite and reconstituted in phospholipid vesicles by removing the detergent using hydrophobic chromatography on Amberlite. Optimal transport activity was obtained by using a Triton X-114/phospholipid ratio of 0.8, 6% cardiolipin and 24 passages through a single Amberlite column. In the reconstituted system the incorporated tricarboxylate carrier catalyzed a first-order reaction of citrate/citrate or citrate/malate exchange. The activation energy of the exchange reaction was 70.1 kJ/mol. The rate of the exchange had a pH optimum between 7 and 8. The half-saturation constant was 0.13 mM for citrate and 0.76 mM for malate. All these properties were similar to those described for the tricarboxylate transport system in intact mitochondria. In proteoliposomes the maximum exchange rate at 25 degrees C reached 2000 mumols/min per g protein. This value was independent of the type of substrate present at the external or internal space of the liposomes (citrate or malate).     

Preparation and characterization of F-protein vesicles isolated from Sendai virus by means of octyl glucoside

We have demonstrated that Triton X-100 is always present in F-protein vesicles at concentrations that can provoke cell lysis. In order to avoid any misinterpretation of the fusogenic capacity of this protein, we solubilized the Sendai virus using octyl glucoside, which can be totally removed from the F protein preparation in less than 16 h by dialysis in the presence of absorbent beads. F-glycoprotein preparations preserved their ability to lyse erythrocytes in the presence of lectins and to induce cell-vesicle fusion as demonstrated by ESR studies. These vesicles were characterized by electron microscopy and SDS-polyacrylamide gel electrophoresis. Lipid analysis of these preparations by thin-layer chromatography indicated that they had the same proportion of lipids as virus envelopes, with slight variations in the sphingomyelin content and the cholesterol/phospholipid molar ratio. F-protein vesicles of different sizes can be obtained by adding exogenous lipids before detergent removal. The hemolytic activity of the vesicles was retained over a large range of lipid concentrations. We conclude that F-protein vesicles prepared with octyl glucoside are convenient tools for studying the fusogenic mechanism of this protein and improving the fusion process between liposomes and cells.     

Novel preparation of functional Sindbis virosomes

Lipid and protein factors important for the preparation and stability of reconstituted membranes prepared by the insertion of detergent-solubilized Sindbis virus glycoproteins into preformed lipid vesicles have been defined. It was found that both the state of aggregation of the membrane protein and the phase of the lipid are critical for the insertion of proteins into preformed lipid vesicles. The membranes prepared with the insertion technique were characterized in terms of residual detergent, protein orientation, and whether or not they were sealed. Binding and fusion experiments have been carried out with the insertion membranes and virus. It was found that BHK-21 cells at 4 degrees C bound one-fifth to one-tenth the number of insertion membranes as intact virus, and binding was saturable in both cases. Variation of the lipid/protein ratio did not result in significant differences in binding. The insertion membranes were found to fuse to a model lipid bilayer at least as well as the virus. These results are discussed in terms of structural factors important for the biological functionalities of the viral spike glycoproteins.     

The reconstituted isolated uncoupling protein is a membrane potential driven H+ translocator.

The isolated uncoupling protein (UCP) from brown fat adipose tissue mitochondria has been reconstituted into artificial phospholipid vesicles. Because of the high lability of H+ transport, several new steps have been introduced in the reconstitution; the detergent octyl-POE, the addition of phospholipids to mitochondria prior to solubilization and purification, the vesicle formation by rapid removal of detergent with polystyrene beads and of external salts by a mixed ion exchange. In the K+-loaded proteoliposomes, H+ influx can be induced by a diffusion potential on addition of valinomycin. H+ influx is inhibited to more than 90% by GTP addition, in the assay for UCP activity. By reversing delta psi with external K+, H+ efflux is measured, however, at a four times lower rate. In vesicles loaded with internal GTP, H+ influx is fully inhibited but can be activated by Dowex-OH treatment to an even higher rate than that found in the GTP-free vesicles. Binding studies with GTP show that most of the active UCP are oriented with the binding site outside as in mitochondria, and that in GTP-loaded vesicles GTP is also bound at the outside. The rate of H+ transport is linearly dependent on the membrane potential. Despite the ordered orientation, there is no 'valve' mechanism, since there is H+ efflux with a reversed potential. pH dependency is only small between pH 6.5 and 7.5, indicating that the H+-translocating site differs from the highly pH-dependent nucleotide-binding site. The turnover number of reconstituted UCP is commensurate with mitochondrial function and indicates a carrier instead of a channel-type H+ transport.(ABSTRACT TRUNCATED AT 250 WORDS)     

Kinetics of the reconstituted dicarboxylate carrier from rat liver mitochondria

The dicarboxylate carrier from rat liver mitochondria was purified by the Amberlite/hydroxyapatite procedure and reconstituted in egg yolk phospholipid vesicles by removing the detergent with Amberlite. The efficiency of reconstitution was optimized with respect to the ratio of detergent/phospholipid, the concentration of phospholipid and the number of Amberlite column passages. In the reconstituted system the incorporated dicarboxylate carrier catalyzed a first-order reaction of malate/phosphate exchange. V of the reconstituted malate/phosphate exchange was determined to be 6000 mumol/min per g protein at 25 degrees C. This value was independent of the type of substrate present at the external or internal space of the liposomes (malate, phosphate or malonate). The half-saturation constant was 0.49 mM for malate, 0.54 mM for malonate and 1.41 mM for phosphate. The activation energy of the exchange reaction was determined to be 95.8 kJ/mol. The transport was independent of the external pH in the range between pH 6 and 8.     

CLIC1 inserts from the aqueous phase into phospholipid membranes, where it functions as an anion channel

CLIC1 is a member of the CLIC familyof proteins, which has been shown to demonstrate chloride channelactivity when reconstituted in phospholipid vesicles. CLIC1 exists incells as an integral membrane protein and as a soluble cytoplasmicprotein, implying that CLIC1 might cycle between membrane-inserted andsoluble forms. CLIC1 was purified and detergent was removed, yieldingan aqueous solution of essentially pure protein. Pure CLIC1 was mixedwith vesicles, and chloride permeability was assessed with a chloride efflux assay and with planar lipid bilayer techniques. Soluble CLIC1confers anion channel activity to preformed membranes that isindistinguishable from the previously reported activity resulting fromreconstitution of CLIC1 into membranes by detergent dialysis. Theactivity is dependent on the amount of CLIC1 added, appears rapidly onmixing of protein and lipid, is inhibited by indanyloxyacetic acid-94,N-ethylmaleimide, and glutathione, is inactivated by heat,and shows sensitivity to pH and to membrane lipid composition. Weconclude that CLIC1 in the absence of detergent spontaneously insertsinto preformed membranes, where it can function as an anion-selective channel.