The repeats-in-toxin (RTX) family of toxins consists of large, membrane-interacting proteins produced by several Gram-negative bacteria, including
Bordetella pertussis,
Escherichia coli,
Vibrio cholerae, and the oral pathogen,
Aggregatibacter actinomycetemcomitans (
Aa). The RTX toxin produced by
Aa is a leukotoxin that specifically kills host immune cells via interactions with the integrin receptor lymphocyte function-associated antigen-1 (LFA-1). In addition to this specific interaction with a membrane receptor, we have observed several essential interactions of LtxA with the host cell membrane. Using surface plasmon resonance (SPR), we demonstrated that LtxA has a significantly stronger affinity for cholesterol-containing membranes than for those lacking cholesterol or containing a different sterol. Removal of cholesterol from the host cell membrane with methyl-b-cyclodextrin or blocking binding to cholesterol with peptides inhibits the ability of LtxA to kill host cells. We also discovered that LtxA mediates membrane disruption in a unique manner. LtxA is able to mediate leakage from liposomes only when the liposomes contain some fraction of lipid with a negative spontaneous curvature. Addition of bilayer-stabilizing lipids inhibits this activity. Because these types of lipids are involved in the formation of nonlamellar lipid phases, we used P
31 nuclear magnetic resonance (P
31 NMR) to investigate if membrane disruption by LtxA proceeds through the disruption of the stable bilayer phase. In these experiments, we observed that LtxA strongly promotes the formation of a nonlamellar, inverted hexagonal (H
II) phase, thus supporting our hypothesis that membrane disruption by LtxA occurs not through the formation of a pore but rather through a unique process of bilayer disruption.
More recently, we have used these two specific membrane interactions of LtxA to design a “Trojan Horse†liposomal antibiotic delivery vehicle. We envisioned this antibiotic-encapsulating liposome to be composed of specific lipids to endow it with two distinct properties. The liposome will be composed of (1) cholesterol to allow it to adsorb LtxA, acting as a “sponge†and (2) a nonlamellar lipid that will enable release of antibiotic only upon LtxA-mediated membrane disruption. We have demonstrated that LtxA, at physiologically relevant concentrations, efficiently permeabilizes liposomes composed of 1,2-dioleoyl-sn-glycero-3-phosphoethanolamine-N-methyl (N-methyl-DOPE), and inclusion of cholesterol enhances this effect. Control liposomes, composed of 1,2-dioleoyl-sn-glycero-3-phosphocholine (DOPC), which are unable to undergo a bilayer-to-HII transition, were not affected by LtxA. Finally, we found that moxifloxacin-containing liposomes inhibit the growth of an LtxA-producing strain of Aa but not that of a non-LtxA-producing strain. Together, our results demonstrate that the mechanism of a membrane interacting toxin can be used to inspire the design new therapeutic options.
