Inhibition by gangliosides of Bacillus cereus phospholipase C activity against monolayers, micelles and bilayer vesicles

The effect of complex glycosphingolipids (gangliosides) on the activity of phospholipase C from Bacillus cereus was studied using lipid monolayers, mixed micelles and small unilamellar vesicles containing phosphatidylcholine as substrate. In all artificial membrane systems assayed, gangliosides exhibit qualitatively similar inhibitory properties. Gangliosides decrease the enzyme activity irrespective of the aggregation structure in which the substrate is offered to B. cereus phospholipase C, and they do not affect the adsorption process of the enzyme. The modulatory effect of gangliosides occurs at the level of the interface, affecting both the maximum rate of catalysis of the enzyme already adsorbed and the availability of the substrate in a suitable organization for enzyme catalysis to take place.

Effect of long-chain acyl-CoAs and acylcarnitines on gel-fluid and lamellar-hexagonal phospholipid phase transitions

The effect of a series of long-chain acyl-CoAs and acylcarnitines has been tested by differential scanning calorimetry on the gel-fluid transition of saturated phosphatidylcholines and of dietai-doylphosphatidylethanolamine (DEPE), and on the lamellar-Hexagonal transition of DEPE. Both series of acylderivatives have similar effects (the acylcarnitines being more potent): a decrease in the gel-fluid transition enthalpy and an increase in the gel-fluid transition width. Mixtures of dipalmitoylphosphatidylcholine with palmitoyl-CoA or palmitoylcarnitine (i.e. when all three hydrocarbon chains are 16C in length) display a peculiar behaviour, in that the main endotherm remains unchanged until a high proportion of palmitoylderivative is present, then it collapses suddenly. The disappearance of the gel-fluid main endotherm in the presence of palmitoylcarnitine is due to the fragmentation of the bilayer below the cooperative unit size of the phospholipid, while the same effect is caused by palmitoyl-CoA through the interaction of the coenzyme A polar moiety with the lipid-water interface, the overall bilayer structure being maintained. The effect of both series of compounds on the lamellar to inverted hexagonal phase transition of DEPE is also similar: they both stabilize the lamellar phase, increase the transition temperature and smear out the transition endotherm. Their behaviour may be rationalized considering that they are compounds with a bulky polar head, relative to their single hydrophobic chain, that would favour a positive curvature in the monolayer, while the inverted hexagonal phase requires a negative curvature.

Dual inhibitory effect of gangliosides on phospholipase C-promoted fusion of lipidic vesicles

The effect of a variety of gangliosides has been tested on the phospholipase C-induced fusion of large unilamellar vesicles. Bilayer composition was phosphatidylcholine:phosphatidylethanolamine: cholesterol (2:1:1 mole ratio) plus the appropriate amounts of glycosphingolipids. Enzyme phosphohydrolase activity, vesicle aggregation, mixing of bilayer lipids and mixing of liposomal aqueous contents were separately assayed. Small amounts ( < 1 mol %) of gangliosides in the lipid bilayer produce a significant inhibition of the above processes. The inhibitory effect of gangliosides increases with the size of the oligosaccharide chain in the polar head group. Inhibition depends in a nonlinear manner on the ganglioside proportion, and is complete at approximately 5 mol %. Inhibition is not due to ganglioside-dependent changes in vesicle curvature or size. Ganglioside inhibition of vesicle fusion is due to two different effects: inhibition of phospholipase C activity and stabilization of the lipid lamellar phase. Enzyme inhibition leads to a parallel decrease of vesicle aggregation and lipid mixing rates. Mixing of aqueous contents, though, is depressed beyond the enzyme inhibition levels. This is explained in terms of the fusion pore requiring a local destabilization of the lipid bilayer, the lamellar structure being stabilized by gangliosides. 31P-NMR and DSC experiments confirm the inhibitory effect of gangliosides in various lamellar-to-nonlamellar transitions.

Purification of Escherichia coli pro-haemolysin, and a comparison with the properties of mature alpha-haemolysin

Pro-haemolysin (approximately 110 kDa), the inactive precursor of the membrane-lytic toxin alpha-haemolysin, has been purified from an overproducing strain of Escherichia coli. Pro-haemolysin forms aggregates in aqueous media, like the mature protein, suggesting an amphipathic structure. Direct measurements of protein binding to liposomal membranes, following a novel procedure, show that pro-haemolysin can bind the lipid bilayers to a similar extent as alpha-haemolysin. This is confirmed by the observed changes in the intrinsic fluorescence emission of the protein upon binding the bilayers. However, pro-haemolysin is totally unable to induce liposomal membrane lysis. Binding of Ca2+, that is essential for the lytic activity of alpha-haemolysin, is greatly diminished in the precursor protein, as shown both by direct measurements of 45Ca(2+)-binding and by fluorescence measurements. The results suggest that binding of a fatty acyl residue in the activation step brings about an important conformational change in the protein that involves the Ca(2+)-binding domain.

Infrared evidence of a beta-hairpin peptide structure in solution

The IR spectrum of an 16-amino acid peptide corresponding, according to NMR studies, to a beta-hairpin has been analysed. Two characteristic features distinguish its spectrum from that of an antiparallel beta-sheet: the low-frequency band that in a beta-sheet structure is located at approximately 1632 cm-1 appears here at approximately 1620 cm-1, and the high-frequency component does not undergo the isotopic shift typical of beta-sheet from 1690 to 1675 cm-1 when transferred to D2O. The infrared characteristics associated with beta-hairpins have been described so far in two proteins, in one of which, whose three-dimensional structure is known from X-ray diffraction, a beta-hairpin has actually been detected.

The Turbidity of Cell Nuclei in Suspension: A Complex Case of Light Scattering

The turbidity of a suspension of cell nuclei isolated from animal tissue homogenates is a complex case of non-Rayleigh scattering. As a first approximation to this system, we have characterized a number of factors that may contribute to the observed turbidity: cation-dependent chromatin condensation, thermal denaturation of chromatin, nuclear shrinking, and changes in the optical properties of the membrane bilayer. Small differences in cation concentration, particularly in the case of divalent cations, lead to large changes in chromatin supramolecular organization, thus to large turbidity effects; thermally-induced changes in turbidity have a similar origin, although they are less pronounced. Under certain circumstances, either salts or heat may induce condensation of chromatin, the latter being connected to the inner side of the nuclear envelope, nuclear shrinking ensues, and this in turn modifies the suspension turbidity. Finally, changes in the physical properties of the lipid bilayers or of the phospholipids in the nuclear envelopes may also have significant effects, though smaller than chromatin changes, in the overall turbidity of the nuclear suspension.

Subunit III of cytochrome c oxidase influences the conformation of subunits I and II: an infrared study

The secondary structure of wild-type Paracoccus denitrificans cytochrome c oxidase obtained by decomposition of the infrared amide I band contains 44% alpha-helix, 18% beta-sheet, 14% beta-turns, 18% loops, and 6% nonordered segments. The mutant lacking subunit III presents a small but significant increase (from 18% to 24%) in the percentage of loops and slight differences in the other components. Using band/area ratios and tyrosine side chain absorption as an inner standard, it is shown that in the absence of subunit III the structure of subunits I and II is altered although no changes in their alpha-helix or beta-sheet content are observed. In the bacterial oxidase, thermal infrared studies show a complex denaturation pattern characterized by the presence of a partially denatured intermediate state. Of the seven predicted subunit III alpha-helices, only four are resistant toward the thermal challenge and behave as expected for typical transmembrane helices. The observation that the absence of subunit III influences the conformation of loop regions in the two other subunits suggests that part of the interaction surface between subunit III and the catalytic subunits might be located outside the lipid bilayer.

Interaction of the bacterial protein toxin alpha-haemolysin with model membranes: protein binding does not always lead to lytic activity

alpha-Haemolysin interaction with model membranes has been investigated by a 2-fold procedure. First, protein binding has been measured, by a direct method as well as through changes in the intrinsic fluorescence of the protein when incubated with liposomes and divalent cations. Then, the above results have been correlated with the protein lytic activity. The extent of protein binding is not significantly modified by the presence or absence of Ca2+, or by changes in lipid composition, although these factors influence greatly the membrane lytic activity of the protein. Moreover, Ca2+ binding to the toxin must occur prior to protein binding to the bilayer, for a lytic effect to take place.

The membrane-perturbing properties of palmitoyl-coenzyme A and palmitoylcarnitine. A comparative study

Fatty acyl-coenzyme A's are temporarily converted into fatty acylcarnitines while transferred across the inner mitochondrial membrane, in their catabolic pathway. In search of an explanation for the need of this coenzyme exchange, the present work describes comparatively the abilities of both kinds of fatty acyl derivatives (represented by palmitoyl-coenzyme A and palmitoylcarnitine) in binding to and perturbing the structure of phosphatidylcholine bilayers in the form of large unilamellar vesicles. Both palmitoyl-coenzyme A and palmitoylcarnitine partition preferentially into the bilayer lipids, so that their free concentration in water is in practice negligible. However, palmitoylcarnitine is able to disrupt the membrane barrier to solutes, leading to vesicle leakage, and, at higher concentrations, it produces complete membrane solubilization, while palmitoyl-coenzyme A produces neither leakage nor solubilization. Palmitoylcarnitine has the properties of many commonly used biochemical detergents. The different behavior of both fatty acyl derivatives helps to explain the need for the transitory coenzyme A/carnitine exchange, and provides a pathogenic mechanism for some genetic defects of mitochondrial fatty acid transport. Other pathophysiological processes in which palmitoylcarnitine has been putatively involved are examined in light of the above results.

Topological properties of two cubic phases of a phospholipid:cholesterol:diacylglycerol aqueous system and their possible implications in the phospholipase C-induced liposome fusion

Water dispersions of phospholipid:cholesterol:diacylglycerol may, under certain conditions, originate either the lipid- and water-permeable Q224 cubic phase, or the lipid-permeable but water-impermeable Q227 cubic phase. These results are discussed within the framework of the phospholipase C-induced fusion of liposomes [Nieva et al. (1993) Biochemistry 32, 1054]. It is suggested that the cubic phases Q224 and Q227 represent two classes of lipid organization, one promoting, the other hindering the mixing of aqueous contents that is characteristic of membrane fusion. In this context, inverted micelles appear to be the end point of the fusion process, rather than fusion intermediates.

A pathway for the thermal destabilization of bacteriorhodopsin

A variety of structural techniques, including IR spectroscopy, reveals that thermal denaturation of bacteriorhodopsin follows a given pathway (successively rearrangement of helical structures, extensive deuterium exchange, and finally protein aggregation) irrespective of heating rate, pH or ionic strength conditions. In all cases, thermal denaturation leads to a 'compact denatured state' which retains a large proportion of ordered structure.

An infrared investigation of palmitoyl-coenzyme A and palmitoylcarnitine interaction with perdeuterated-chain phospholipid bilayers

Mixtures of di-(perdeuteropalmitoyl)-sn-glycero-3 choline ([2H62]Pam2GroPCho) with palmitoylcarnitine or palmitoyl-CoA in aqueous suspension have been examined by Fourier-transform infrared spectroscopy. The C-H (or C-D) region of the spectrum shows that an order-disorder transition exists in pure aqueous palmitoylcarnitine at 45 degrees C; palmitoylcarnitine mixes with [2H62]Pam2GroPCho without perturbing the gel-fluid transition of the phospholipid even at [2H62]Pam2GroPCho/palmitoylcarnitine 1:2 molar ratios; and palmitoyl-CoA, however, at similar proportions, smears out the [2H62]Pam2GroPCho transition as detected from C-D stretching vibrations. Relevant data from the carbonyl region include; the high-frequency (non-hydrogen bound) carbonyl subpopulation, but not the low frequency one, detects the gel-to-fluid transition of the phospholipid; the carbonyl region detects the thermotropic transition over a wider temperature range than the methylene stretching region, i.e. detects changes starting well below and ending several degrees above the methylene transition temperature, and a significant interaction may occur between some coenzyme A group and the carbonyl groups of the phospholipid. The latter interaction may contribute to explain the coenzyme A/carnitine exchange during mitochondrial fatty acid import.

Electrokinetic charge of the anesthetic-induced bR480 and bR380 spectral forms of bacteriorhodopsin

The translational and rotational electrokinetics of the anesthetic-induced spectral transitions bR568-->bR480-->bR380 of bacteriorhodopsin have been investigated. Formation of the bR480 form is associated with an increase of the purple membrane negative electrokinetic charge, while the transformation of bR480 into bR380 is accompanied by a decrease of the membrane negative charge as compared to that of the 480 nm-absorbing form. Removal of anesthetics leads to the back transitions bR480-->bR568 and (in part) bR380-->bR568; however, the electrokinetic charge of the native membranes is not restored. A strong decrease in the electric polarizability and the appearance of a slow polarizability component are also observed in anesthetic-treated membranes. Comparison with the electrokinetic behaviour of partially delipidated membranes and with that of liposomes composed of purple membrane total lipids suggests that: (i) anesthetic molecules partition mainly at the protein/lipid interface inducing irreversible rearrangement of the boundary lipid layer, and (ii) different mode(s) or site(s) of interaction are responsible for the spectral and surface charge effects. The data are compatible with the hypothesis of anesthetics acting through partial dehydration of the membrane surface.

Surface-core relationships in human low density lipoprotein as studied by infrared spectroscopy

The secondary structure of human apolipoprotein B at 37 degrees C is estimated to be 24% alpha-helix, 23% beta-sheet, 6% beta-turns, 24% unordered structure, and 24% "beta-strands," characterized by a band around 1618 cm-1, and consistent with extended string-like chains in contact with the lipid moiety not forming beta-sheets. When cooled to a temperature below the cholesteryl ester transition at 30 degrees C, the ordering of the low density lipoprotein core results in reversible changes in the protein conformation, decreasing the apparent amount of alpha-helix, beta-strand, and unordered structure below 30 degrees C and increasing beta-sheet and beta-turns. Lowering the ionic strength affects the core-associated transitions, shifting their temperature from 30 to 20 degrees C, and modifying protein conformation below the transition. An additional thermal event is observed at 75 degrees C, leading to irreversible protein denaturation. In the broad temperature range between the 30 and 75 degrees C transitions, apolipoprotein B is stable toward both temperature and ionic strength changes. After thermal denaturation, the protein retains a certain degree of ordered structure.

Liposome destabilization induced by the HIV-1 fusion peptide effect of a single amino acid substitution

The 23-residue synthetic peptide representing the N-terminus of HIV-1 gp41 is known to induce either leakage or fusion of lipid vesicles depending on the experimental conditions. In this paper we report that a polar amino acid substitution V-->E at position 2, known to block gp41 activity in vivo, makes the peptide unable to destabilize and/or fuse membranes. Moreover this variant, unlike the parent peptide, is never found in the membrane-associated beta conformation.

The binding of divalent cations to Escherichia coli alpha-haemolysin

alpha-haemolysin, an extracellular protein toxin of Escherichia coli, is known to disrupt eukaryotic cell membranes. In spite of genetic evidence of Ca(2+)-binding motifs in its sequence, conflicting results are found in the literature on the requirement of divalent cations for the membranolytic activity of the toxin. Moreover, Ca(2+)-binding sites have not been characterized to date in the native protein. The results in this paper show that when Ca2+ levels are kept sufficiently low during bacterial growth and toxin purification, membrane lysis does not occur in the absence of added divalent cations. Ca2+ and, at higher concentrations, Sr2+ and Ba2+, support the lytic activity, but Mg2+, Mn2+, Zn2+ and Cd2+ appear to be inactive in this respect. Binding of metal ions can be followed by changes in the intrinsic fluorescence of alpha-haemolysin; ions supporting lytic activity produce changes in the intrinsic fluorescence that are not caused by the inactive ones. Scatchard analysis of 45Ca2+ binding reveals three equivalent, independent sites, with Kd approximately 0.11 mM. No 45Ca2+ binding is observed when the protein is incubated with Zn2+; conversely, incubation with Ca2+ prevents subsequent binding of 65Zn2+. In the light of three-dimensional data available for a structurally related protein, alkaline protease of Pseudomonas aeruginosa [Baumann, U., Wu, S., Flaherty, K. M. & McKay, D. B. (1993) EMBO J. 12, 3357-3364] it is suggested that alpha-haemolysin may bind a larger number of Ca2+ than the three that are more easily exchangeable and are thus detected in the 45Ca(2+)-binding experiments. In addition, structural similarities and conservation of ion-binding motifs support the hypothesis that His 859 is involved in the mutually exclusive binding of Zn2+ and Ca2+.

Differential penetration of fatty acyl-coenzyme A and fatty acylcarnitines into phospholipid monolayers

The ability of fatty acyl-CoA's and fatty acylcarnitines to penetrate phospholipid monolayers was comparatively studied, in view of the important role of both kinds of derivatives in fatty acid transport across mitochondrial membranes. The interaction occurs predominantly through hydrophobic forces. Acylcarnitines penetrate phospholipid monolayers more strongly than acyl-CoAs; in addition the former show a positive cooperativity when they bind to the interface. These properties would facilitate membrane transfer of fatty acylcarnitines over that of their CoA homologues.

Early and delayed stages in the solubilization of purple membrane by a polyoxyethylenic surfactant

The purpose of this paper is to explore the reasons by which purple membrane solubilization by detergents takes hours, or even days, to reach equilibrium, while most biomembranes are solubilized in a matter of seconds, or minutes. With that aim, changes in the purple membrane absorption spectrum produced by hydrogenated Triton X-100 under equilibrium conditions (24 h) have been compared to those caused by the same surfactant in the minute, second and sub-second time scale. It is found that the various processes that accompany, or lead to, solubilization are already detected, and even reach an apparent equilibrium, in the 10 s that follow detergent addition. No new phenomena are detected in the following minutes, or hours, that are relevant to the process under study. This leads to the conclusion that the long solubilization process consists of the repeated operation of simple phenomena that are relatively fast in themselves. A hypothesis is proposed according to which the tight crystalline organization of the purple membrane prevents the insertion of detergent monomers in the lipid bilayer; instead, the surfactant would bind the periphery of the patches, i.e., the hydrocarbon-water contact region, and solubilization would take place gradually, from the periphery towards the core of the membrane patches, at a progressively lower rate as the amounts of free detergent and detergent-binding sites are decreased by the previous solubilization steps.

Structure and thermal denaturation of crystalline and noncrystalline cytochrome oxidase as studied by infrared spectroscopy

Fourier-transform infrared spectroscopy has been applied to the study of lipid vesicle-supported two-dimensional crystals and noncrystalline preparations of beef heart cytochrome oxidase. At room temperature, no conformational differences are seen between the noncrystalline and crystalline proteins, whose conformation is shown to consist of ca. 40% alpha-helix, 20% extended structures (including beta-sheet), 17% beta-turns, and 22% open loops plus nonstructured conformations. A novel infrared approach that combines quantitative spectral band decomposition with the study of the thermal behavior of each component has been applied. The procedure allows the independent examination of temperature-induced changes in individual structural elements (alpha-helix, beta-sheet, beta-turns, and unordered). All these reflect, upon heating the protein from 20 to 80 degrees C, a major irreversible thermal event centred at 55-60 degrees C, leading to a molecular state devoid of enzyme activity but with a defined secondary structure; in addition, when the band position, percent area (integrated intensity), and bandwidth of the various amide I components are separately plotted versus temperature, each component is seen to behave in a characteristic way. Thermal denaturation in D2O buffer shows a decrease in nonstructured conformations and an increase in beta-turns without major changes in the proportion of alpha-helix. Temperature-induced changes are not the same in amorphous and crystalline structures, the latter being in general more stable toward the thermal challenge. The above data extend and confirm previous structural studies on cytochrome oxidase using cryo-electron microscopy.

Thermodynamic and structural stability of cytochrome c oxidase from Paracoccus denitrificans

The structural stability of the integral membrane protein cytochrome c oxidase from Paracoccus denitrificans has been measured by high-sensitivity differential scanning calorimetry and Fourier transform infrared spectroscopy. Contrary to the mammalian enzyme or the yeast enzyme, which are composed of 13 subunits, the bacterial enzyme has only three or four subunits, thus providing a unique opportunity to examine the magnitude of the forces that stabilize this enzyme and to establish accurate structural assignments of events observed calorimetrically. In this paper, experiments have been performed with the wild-type enzyme and with a mutant enzyme lacking subunit III. Our results show that subunits I and II form a highly cooperative complex which denatures as a single cooperative unit at 67 degrees C, while subunit III is less stable and denatures 20 degrees C earlier. Reduction of the enzyme causes a large increase in the stability of subunits I and II but has absolutely no effect on subunit III. Despite the lack of a strong interaction between subunit III and the catalytic subunits, the absence of subunit III leads to a turnover-induced loss of electron-transfer activity. The magnitude of the energetic parameters and the infrared spectroscopic experiments indicate that the enzyme does not completely unfold upon thermal denaturation and that significant degrees of structure are preserved. The amount of native alpha-helix structure, which is 45% in the native state, decreases only to 30% after thermal denaturation. Presumably, the residual helical structure existing after thermal denaturation belongs to the intramembranous portions of the protein. The calorimetric behavior of subunit III does not fully conform to that expected for a highly alpha-helical membrane protein. The picture that emerges from these experiments is that, in the temperature-denatured form of the enzyme, most of the extramembranous structural elements are denatured while most of the intramembranous secondary structure is maintained even though native tertiary interactions appear to be disrupted.

An assessment of the biochemical applications of the non-ionic surfactant Hecameg

A number of properties and effects of the novel non-ionic detergent Hecameg (6-O-(N-heptylcarbamoyl)-methyl-alpha-D-glucopyranoside) have been examined in view of its possible biochemical applications. In particular, its critical micellar concentration has been measured, and its effects on pure lipid membranes, soluble and membrane-bound enzymes have been recorded. Hecameg has some advantageous and some less advantageous properties; its relatively high critical micellar concentration (16.5 mM), almost insensitive to pH or ionic strength changes, makes it suitable for reconstitution procedures in which detergent must be removed by dialysis. It is also an effective lipid-solubilizing agent, producing leakage of vesicle contents at detergent concentrations well below the solubilizing range. Among the drawbacks, the presence of an amide group in the molecule may interfere with the protein amide group in spectroscopic measurements. It also appears to be less gentle than other nonionic surfactants towards certain enzyme activities.

Interaction of the HIV-1 fusion peptide with phospholipid vesicles: different structural requirements for fusion and leakage

This paper presents a study on the membrane fusion activity of a 23-residue synthetic peptide, representing the N-terminus of gp41 of the human immunodeficiency virus type I (HIV-1; LAV1a strain), in a model system involving large unilamellar vesicles (LUV) composed of the negatively charged 1-palmitoyl-2-oleoylphosphatidylglycerol (POPG). The peptide (HIVarg) induced fusion of POPG LUV as evidenced by (i) mixing of membrane lipids, (ii) mixing of aqueous vesicle contents, and (iii) an irreversible increase in vesicle size. Fusion could be induced only in the presence of millimolar concentrations of Ca2+ or Mg2+, needed for induction of vesicle aggregation; the divalent cations by themselves did not induce any fusion. The rate constant of the fusion reaction, as determined by simulation of the process according to a kinetic model, increased dramatically with the peptide-to-lipid molar ratio, indicating that the peptide was the mediator of the process. In the absence of divalent cations, the HIVarg peptide induced leakage of small molecules due to formation of pores in the membrane of single vesicles. Final extents and kinetics of this leakage process could be simulated adequately by model calculations for peptide-to-lipid ratios ranging from 1:25 to 1:750. Experiments, in which the order of peptide and Ca2+ addition to the vesicles was varied, indicated that the peptide is likely to adopt two different structures, one in the absence of Ca2+, primarily supporting leakage by formation of pores in separate vesicles, and one in the presence of Ca2+, primarily supporting fusion. Once a final structure had been established, it persisted even upon addition or removal of Ca2+.(ABSTRACT TRUNCATED AT 250 WORDS)

Real-time measurements of chemically-induced membrane fusion in cell monolayers, using a resonance energy transfer method

Fusion of mouse melanoma cells grown in monolayers has been directly monitored by fluorescence resonance energy transfer between fluorescein and rhodamine probes attached to octadecanoic acid. Various poly(ethylene glycol)s (PEG), either alone or in combination with amphipathic molecules, have been used as fusogens. Fusion starts at a maximum rate as soon as PEG is removed from the medium and reaches a plateau after 20-30 min. Both the initial rate and extent of fusion have been recorded for each experiment. The extent of fusion shows in general a positive correlation with the initial rate, although PEGs with different molar masses appear to induce fusion at different rates, but to a similar extent. A good correlation has been found between the extent of fusion, as measured by fluorescence, and the 'fusion index' computed from cell and nucleus counting; a calibration curve is provided for the interconversion of both parameters. Optimum fusion values are obtained with 50% (w/v) PEG 1500. The effect of pre-treatments with surfactants (Triton X-100, sodium dodecylsulphate) on PEG-induced fusion has also been tested. Sodium dodecylsulphate, but not Triton, enhances considerably both the rate and extent of cell fusion. The in situ generation of the amphipathic molecule diacylglycerol, through the catalytic activity of a phospholipase C, also enhances significantly the fusion parameters. These results are in good agreement with previous studies based on syncytia counting.