Lipid-protein interactions. A comparative study of the binding of cardiotoxins and neurotoxins to phospholipid vesicles

Galectin-9 trafficking regulates apical-basal polarity in Madin-Darby canine kidney epithelial cells

Galectins are unconventionally secreted lectins that participate in the formation of glycoprotein lattices that perform a variety of cell surface functions. Galectins also bind glycosphingolipid headgroups with as yet unclear implications for cellular physiology. We report a specific interaction between galectin-9 and the Forssman glycosphingolipid (FGL) that is important for polarizing Madin-Darby canine kidney epithelial cells. Galectin-9 knockdown leads to a severe loss of epithelial polarity that can be rescued by addition of the recombinant protein. The FGL glycan is identified as the surface receptor that cycles galectin-9 to the Golgi apparatus from which the protein is recycled back to the apical surface. Together our results suggest a model wherein such glycosphingolipid-galectin couples form a circuit between the Golgi apparatus and the cell surface that in an epithelial context facilitates the apical sorting of proteins and lipids.

Structure of a cardiotoxic phospholipase A(2) from Ophiophagus hannah with the "pancreatic loop"

The crystal structure of an acidic phospholipase A(2) from Ophiophagus hannah (king cobra) has been determined by molecular replacement at 2.6-A resolution to a crystallographic R factor of 20.5% (R(free)=23.3%) with reasonable stereochemistry. The venom enzyme contains an unusual "pancreatic loop." The conformation of the loop is well defined and different from those in pancreas PLA(2), showing its structural variability. This analysis provides the first structure of a PLA(2)-type cardiotoxin. The sites related to the cardiotoxic and myotoxic activities are explored and the oligomer observed in the crystalline state is described.

Model of interaction between a cardiotoxin and dimyristoylphosphatidic acid bilayers determined by solid-state 31P NMR spectroscopy

The interaction of cardiotoxin IIa, a small basic protein extracted from Naja mossambica mossambica venom, with dimyristoylphosphatidic acid (DMPA) membranes has been investigated by solid-state 31P nuclear magnetic resonance spectroscopy. Both the spectral lineshapes and transverse relaxation time values have been measured as a function of temperature for different lipid-to-protein molar ratios. The results indicate that the interaction of cardiotoxin with DMPA gives rise to the complete disappearance of the bilayer structure at a lipid-to-protein molar ratio of 5:1. However, a coexistence of the lamellar and isotropic phases is observed at higher lipid contents. In addition, the number of phospholipids interacting with cardiotoxin increases from about 5 at room temperature to approximately 15 at temperatures above the phase transition of the pure lipid. The isotropic structure appears to be a hydrophobic complex similar to an inverted micellar phase that can be extracted by a hydrophobic solvent. At a lipid-to-protein molar ratio of 40:1, the isotropic structure disappears at high temperature to give rise to a second anisotropic phase, which is most likely associated with the incorporation of the hydrophobic complex inside the bilayer.

Proton NMR studies of the structural and dynamical effect of chemical modification of a single aromatic side-chain in a snake cardiotoxin. Relation to the structure of the putative binding site and the cytolytic activity of the toxin

This paper presents the comparative comprehensive analysis of NMR structural parameters (NOEs, scalar coupling, chemical shifts) of toxin gamma, a cardiotoxin isolated from the venom of Naja nigricollis, and three chemical derivatives, i.e. the 2-nitrophenylsulphonyl (NPS)-Trp11, 3-nitro-Tyr22 and 3-nitro-Tyr51 derivatives. In previous work, the chemical modifications of single side chains have suggested that these aromatic residues, in association with several lysine residues, contributed to the cytotoxicity of toxin gamma. Analysis of these results based on the refined solution structure of the toxin has resulted in the proposal of a conserved phospholipid binding site through which cardiotoxins are likely to interact with the membrane of target cells. The present work shows that modifications of either the tryptophan residue or the tyrosine residues, which are within or near the proposed binding site, have no influence on the three-dimensional structure of the protein. On the other hand, the proton exchange study of the backbone amides indicates that the structural core of the protein is destabilized in the three derivatives. This corresponds to a decrease of the overall stability of the protein as indicated by the comparative solvent denaturation study of the unmodified toxin gamma and the Trp11 derivative. More specifically, the dynamics of the three-stranded beta sheet, a part of the structural core, are highly perturbed by the chemical modifications. This sheet was previously proposed as a part of the phospholipid binding site of cardiotoxins. The dynamical perturbation of this site appears to be correlated with the decrease in toxicity of the chemical derivatives.

Binding of crotoxin, a presynaptic phospholipase A2 neurotoxin, to negatively charged phospholipid vesicles

Crotoxin, isolated from the venom of Crotalus durissus terrificus, is a potent neurotoxin consisting of a basic and weakly toxic phospholipase A2 subunit (component B) and an acidic nonenzymatic subunit (component A). The nontoxic component A enhances the toxicity of the phospholipase subunit by preventing its nonspecific adsorption. The binding of crotoxin and of its subunits to small unilamellar phospholipid vesicles was examined under experimental conditions that prevented any phospholipid hydrolysis. Isolated component B rapidly bound with a low affinity (Kapp in the millimolar range) to zwitterionic phospholipid vesicles and with a high affinity (Kapp of less than 1 microM) to negatively charged phospholipid vesicles. On the other hand, the crotoxin complex did not interact with zwitterionic phospholipid vesicles but dissociated in the presence of negatively charged phospholipid vesicles; the noncatalytic component A was released into solution, whereas component B remained tightly bound to lipid vesicles, with apparent affinity constants from 100 to less than 1 microM, according to the chemical composition of the phospholipids. On binding, crotoxin or its component B caused the leakage of a dye entrapped in vesicles of negatively charged but not of zwitterionic phospholipids. The selective binding of crotoxin suggests that negatively charged phospholipids may constitute a component of the acceptor site of crotoxin on the presynaptic plasma membrane.

Localization of the toxic site of Naja mossambica cardiotoxins: small synthetic peptides express an in vivo lethality

Cardiotoxins are small basic proteins which cause heart failure when they are injected in vivo. In order to better understand their molecular mode of action, short peptides designed on the model of the first loop of the molecule of cardiotoxin IV from Naja mossambica mossambica venom have been synthetized by the solid-phase procedure of Merrifield. These peptides express lethality in mouse when they are injected intravenously. Taking into account the respective molecular weights, they are 3.5 to 5% as toxic as the cardiotoxin. Furthermore, the symptomatology they induce is undistinguishable from that induced by cardiotoxins. These results strongly support our previous hypothesis that the first loop of the molecule is the toxic site of cardiotoxins.

Mixed phospholipid-cardiotoxin monomolecular films studied by intrinsic polarized surface fluorescence

Surface fluorescence spectroscopy was used to study mixed phospholipid-cardiotoxin monomolecular films. Using polarized incident light we were able to detect the intrinsic fluorescence emission at a wavelength of 339 nm of the single Trp11 in the cardiotoxin molecule. Its fluorescence intensity increased continuously upon film compression up to a maximal value reached at 42 mN x m-1 in surface pressure. A relative polarization index can be used as an indication of the orientation of the indole ring. A transition at around 25 +/- 5 mN x m-1 in the orientation of the Trp11 relative to the plane of the interface was clearly evidenced during film compression, in agreement with our independent data showing a flipping of the whole toxin molecule (Bougis et al., (1981), Biochemistry 20, 4915-4920).

Characterization of elapidae snake venom components using optimized reverse-phase high-performance liquid chromatographic conditions and screening assays for alpha-neurotoxin and phospholipase A2 activities

The vast majority of Elapidae snake venoms, genus Naja, includes three classes of toxic polypeptides: alpha-neurotoxins, phospholipases A2, and cardiotoxins. A new experimental approach using reverse-phase high-performance liquid chromatography in particular has been developed, allowing their respective resolution, identification, and quantitation from milligram quantities of venom. First, definition of optimal chromatographic conditions for Naja mossambica mossambica toxins has been ascertained. Different column packing and solvent systems were compared for their efficiency, with particular attention to the ionic strength of the aqueous solvent. A medium-chain alkyl support (octyl) in conjunction with a volatile ammonium formate (0.15 M, pH 2.70)/acetonitrile solvent system was found to be particularly effective. All the components known until now from this venom could be resolved in a single step, and the elution order was alpha-neurotoxins, phospholipases A2, and cardiotoxins with a total recovery of absorbance and toxicity. Then, with these suitable conditions, we describe a new major cardiotoxin molecule in this venom by hydrophobic and not ionic-charge discrimination. Second, specific assays were designed to detect alpha-neurotoxin and phospholipase A2 activities in chromatographic fractions: alpha-neurotoxin activity was determined by competition for the binding of a radiolabeled alpha-neurotoxin to the acetylcholine receptor of the ray electric organ, and phospholipase A2 activity was defined by the enzymatic activity of these toxins with a fluorescent phospholipid as substrate. Finally, the applicability of these new methods to study other Naja snake venoms was demonstrated.

The binding of snake venom cardiotoxins to heart cell membranes

Cobra venom cardiotoxins have the effect, inter alia, of causing systolic arrest of the heart. We have observed significant binding in vitro of 35S-labelled cardiotoxins to mouse heart cell membranes. Part of the binding was saturable and could be displaced with homologous unlabelled cardiotoxins but not by neurotoxins or cardiotoxins inactivated by chemical modification. The specifically bound component represented more than 70% of total binding at saturation. Inclusion of Triton X-100 and NaCl in the phosphate-buffered incubation medium prevented nonspecific adsorption to centrifuge tube walls, and gave lower but more reproducible specific binding results, respectively. An apparent dissociation constant of 5 . 10(-7) M and a binding density of 500 pmol toxin/mg membrane protein were derived from the saturation isotherms.

A possible orientation change of cardiotoxin molecule during its interaction with phospholipid monolayer

The monomolecular film technique was used in order to study the specific interactions of 4 cardiotoxins from Elapidae snake venom Naja mossambica mosambica with different phospholipids. The interaction, at pH 7.5, of cardiotoxin (10(-7) M) with both neutral and negatively charged phospholipids occurs up to a very high critical surface pressure (pi = 45 dynes/cm with the latest). The apparent molecular area of cardiotoxin molecule, during its insertion into a negatively charged phospholipid film, presents only two characteristic values: 1400 A2 for pi less than 20 dynes/cm and 420 A2 for pi greater than 30 dynes/cm, the transition occurring in a very narrow range of surface pressure (25 +/- 5 dynes/cm). Thus, according to the surface pressure, the cardiotoxin may present two different orientations: "flat" or "edgewise" and the transition between both could account for lytic activity of cardiotoxin.

Structure-function relationships for cardiotoxins interacting with phospholipids

Four cardiotoxins (CTX I-IV) from Naja mossambica mossambica were compared for their ability to interact with phospholipid vesicles and their capacity to bind erythrocytes. It is concluded that the affinity of the toxins always increases in the order: I approximately equal to II less than III less than IV. The binding is specific for charged lipids even in lipid mixtures. Proteolytic attack of the free and lipid-bound cardiotoxin indicates that at least the first loop Leu1-Thr13 is at the lipid contact. Tryptic and synthetic peptides constitutive of this loop are shown to interact with lipids. Arg5 residue increases the affinity toward the bilayer. The Raman spectra of lipid-bound cardiotoxin indicate a secondary and tertiary structure mainly similar to that of the free toxin. On charged lipids cardiotoxins induce a decrease of the enthalpy and an increase of disorder without change in the transition temperature; at saturating amounts of toxin the transition is abolished. In binary mixtures of phosphatidylcholine and charged lipids the observed effects can be accounted by a phase separation induced by the toxin.

Isolation of rhodopsin by the combined action of cardiotoxin and phospholipase A2 on rod outer segment membranes

Freeze-fracture electron microscopy was used to follow morphological changes induced by Naja mossambica mossambica venom VII4 cardiotoxin in rod outer segment membrane preparations. The extent of the morphological changes depend on the purity of the cardiotoxin. Pure cardiotoxin had no detectable effect upon the preparation, but, when contaminated with venom phospholipase A2, led to a rapid disintegration of the membrane vesicles. With trace amounts (up to about 0.5% of the cardiotoxin) of phospholipase A2, the membrane vesicles disintegrated into smooth lamellae and particles in solution. These two components were separated by centrifugation. The pellet, which showed the presence of smooth lamellae and aggregated particles, was composed of unbleached rhodopsin, initial membrane lipids, lysolipids and cardiotoxin. The supernatant, which showed only the presence of dispersed particles, was composed of unbleached rhodopsin, lysolipids and cardiotoxin. With cardiotoxin contained larger amounts of phospholipase A2 (more than 0.5% of the cardiotoxin), membrane vesicles were disintegrated into large aggregates of amorphous material, composed of bleached rhodopsin, initial membrane lipids, lysolipids and cardiotoxin. These results confirm our previous observation on the release of integral membrane proteins from membrane vesicles by the action of cardiotoxin containing traces of phospholipase A2 (Gulik-Krzywicki, T., Balerna, M., Vincent, J.P. and Lazdunski, M. (1981) Biochim. Biophys. Acta 643, 101-114) and suggest it possible use for isolation and purification of integral membrane proteins.

Freeze-fracture study of cardiotoxin action on axonal membrane and axonal membrane lipid vesicles

Freeze-fracture electron microscopy was used to follow morphological changes induced by Naja mossambica mossambica venom cardiotoxins on crab axonal membranes and thier lipids. It was shown that the extent of morphological changes depended drastically on the purity of cardiotoxin preparations and on their nature. Highly purified cardiotoxin induced mainly fusion of membrane or lipid vesicles. The extent of fusion and other morphological changes depended on the nature of cardiotoxin used: VII4 cardiotoxin induced only fusion while VII1 led to further modifications of membranes and liposomes. The most spectacular morphological changes were observed with axonal membranes treated with cardiotoxin containing traces of venom phospholipase A2. At low cardiotoxin concentration (10(-7)-10(-5) M) important intramembrane particle aggregation was observed and at higher concentrations (more than 10(-4) M) intramembrane particles disappeared from the membrane and were found in solution. The membrane vesicles, devoid of intramembrane particles, were observed to fuse rapidly into liposome-like aggregates. These morphological changes are interpreted as being due to the removal of intrinsic membrane proteins from the membrane by the combined action of cardiotoxin and phospholipase A2.