Membrane interactions of lipopeptides and saponins, two groups of natural amphiphiles

Lorent, Joseph H. [UCL]

(eng) We study in this work the interaction of some amphiphilic molecules of natural origin with membranes. Two groups of amphiphiles (saponins and lipopeptides) have been selected because they showed interesting activities towards cancer cells and are known to interact with lipid membranes. Saponins possess a hydrophilic sugar moiety and a hydrophobic triterpenoid or steroid ring structure. They are mainly extracted from plants but can also be found in holotorians and star fish and would constitute defensive molecules. We investigated the membrane interactions of α-hederin, a monodesmosidic pentacyclic triterpenoid saponin, as a representative member of saponins because it has shown some selectivity towards cancer cells (Swamy and Huat, 2003). We also investigated the effects of its aglycone hederagenin, a triterpenic acid, to point out the importance of the sugar chain. Lipopeptides are composed of a polar polypeptide ring and an apolar acyl chain. They are produced mainly by prokaryotes of the Bacillus species. We investigated the effect of surfactin since it has a strong amphiphilic character, is known to induce membrane lysis and to be active on cancer cells. Membrane interactions of amphiphiles were investigated by several membrane models and complementary techniques. Large unilamellar (LUV) and multilamellar vesicles (MLV) were used to obtain quantifiable results of the effects induced with regard to the amphiphile/lipid ratio and incubation time. They were used to study binding (ANS affinity), permeabilization (calcein release), membrane order (Laurdan fluorescence), phase separation (Laurdan fluorescence and Förster resonance energy transfer) and formation of new lipid mesophases (31P-NMR, dynamic light scattering). To visualize the effects on a microscopic scale, giant unilamellar vesicles (GUV) were investigated by fluorescence/confocal microscopy. Supported planar bilayers (SPB) were analyzed by atomic force microscopy (AFM) to give information about the effects at a nanoscopic scale. In parallel to the use of membrane models, we also investigated apoptosis and non-apoptotic cell death on cells derived from monocytic leukemiae. α-hederin was able to interact with sterols in aqueous solution and to form mixed aggregates. In LUV, when no cholesterol was present in the membrane, the molecule inhibited phase separation most probably by a linactant effect. In the presence of cholesterol, α-hederin induced membrane permeabilization and an increase of phospholipid mobility. This was associated with the formation of a non-bilayer mesophase as confirmed by AFM on SPB, and local separation of phospholipids and sterols. In GUV, the effect depended on the critical micellar concentration of the saponin. In its monomeric form, the molecule induced mainly budding, the formation of worm shaped domains and a gradual permeabilization. Above the CMC (critical micellar concentration), budding and domain formation were less important but we observed the appearance of macroscopic pores and instantaneous permeabilization. The presence of a sugar chain at C3 on the triterpenoid ring structure was important for domain formation and rapid permeabilization. We confirmed on cancer cells the cholesterol dependence of cytotoxicity. α-hederin and hederagenin induced apoptosis and non-apoptotic cell death most probably by formation of membrane defects or induction of the extrinsic apoptotic pathway and activity on lipid rafts. Surfactin activity was also CMC dependent. In SPB and LUV, at concentrations below its CMC, the molecule inserted at the boundary between gel and fluid lipid domains, inhibited phase separation and stiffened the bilayer. It also induced transient defects and permeabilization of LUV without global morphological changes. At concentrations close to the CMC, surfactin solubilized the fluid phospholipid phase and increased order in the remaining lipid bilayer. At higher concentrations, all bilayer structures were dissolved and transformed most probably into micelles. On the basis of these results, we develop a new interaction mechanism for α-hederin and hederagenin with lipid membranes which may be applied to other saponins and triterpenic acids as well. We also compare the effects of surfactin, α-hederin and hederagenin on membrane models and discuss possible consequences regarding their effect on cancer cells. Finally, we give some short term perspectives about ongoing and future experiments and explore long term perspectives about the possibility of further pharmacological use of the molecules we used in this work.