The pH-dependent induction of lipid membrane ionic permeability by N-terminally lysine-substituted analogs of gramicidin A

Function Investigations and Applications of Membrane Proteins on Artificial Lipid Membranes

Membrane proteins play an important role in key cellular functions, such as signal transduction, apoptosis, and metabolism. Therefore, structural and functional studies of these proteins are essential in fields such as fundamental biology, medical science, pharmacology, biotechnology, and bioengineering. However, observing the precise elemental reactions and structures of membrane proteins is difficult, despite their functioning through interactions with various biomolecules in living cells. To investigate these properties, methodologies have been developed to study the functions of membrane proteins that have been purified from biological cells. In this paper, we introduce various methods for creating liposomes or lipid vesicles, from conventional to recent approaches, as well as techniques for reconstituting membrane proteins into artificial membranes. We also cover the different types of artificial membranes that can be used to observe the functions of reconstituted membrane proteins, including their structure, number of transmembrane domains, and functional type. Finally, we discuss the reconstitution of membrane proteins using a cell-free synthesis system and the reconstitution and function of multiple membrane proteins.

Effect of Site-Specific Intermolecular Lysine-Tryptophan Interactions on the Aggregation of Gramicidin-Based Peptides Leading to Pore Formation in Lipid Membranes

In contrast to the parent pentadecapeptide gramicidin A (gA), some of its cationic analogs have been shown previously to form large-diameter pores in lipid membranes. These pores are permeable to fluorescent dyes, which allows one to monitor pore formation by using the fluorescence de-quenching assay. According to the previously proposed model, the gA analog with lysine substituted for alanine at position 3, [Lys3]gA, forms pores by a homopentameric assembly of gramicidin double-stranded β-helical dimers. Here, we studied the newly synthesized analogs of [Lys3]gA with single, double and triple substitutions of isoleucines for tryptophans at positions 9, 11, 13, and 15. Replacement of any of the tryptophans of [Lys3]gA with isoleucine resulted in suppression of the pore-forming activity of the peptide, the effect being significantly dependent on the position of tryptophans. In particular, the peptide with a single substitution of tryptophan 13 showed much lower activity than the analogs with single substitutions at positions 9, 11, or 15. Of the peptides with double substitutions, the strongest suppression of the leakage was observed with tryptophans 13 and 15. In the case of triple substitutions, only the peptide retaining tryptophan 11 exhibited noticeable activity. It is concluded that tryptophans 11 and 13 contribute most to pore stabilization in the membrane, whereas tryptophan 9 is not so important for pore formation. Cation-π interactions between the lysine and tryptophan residues of the peptide are suggested to be crucial for the formation of the [Lys3]gA pore.

Gramicidin A disassembles large conductive clusters of its lysine-substituted derivatives in lipid membranes

N-terminally substituted lysine derivatives of gramicidin A (gA), [Lys1]gA and [Lys3]gA, but not glutamate- or aspartate-substituted peptides have been previously shown to cause the leakage of carboxyfluorescein from liposomes. Here, the leakage induction was also observed for [Arg1]gA and [Arg3]gA, while [His1]gA and [His3]gA were inactive at neutral pH. The Lys3-containing analogue with all tryptophans replaced by isoleucines did not induce liposome leakage, similar to gA. This suggests that the presence of both tryptophans and N-terminal cationic residues is critical for pore formation. Remarkably, the addition of gA blocked the leakage induced by [Lys3]gA. By examining with fluorescence correlation spectroscopy the peptide-induced leakage of fluorescent markers from liposomes, we estimated the diameter of pores responsible for the leakage to be about 1.6 nm. Transmission electron cryo-microscopy imaging of liposomes with [Lys3]gA showed that the liposomal membranes contained high electron density particles with a size of about 40 Å, suggesting the formation of peptide clusters. No such clusterization was observed in liposomes incorporating gA or a mixture of gA with [Lys3]gA. Three-dimensional reconstruction of the clusters was compatible with their pentameric arrangement. Based on experimental data and computational modeling, we suggest that the large pore formed by [Lys3]gA represents a barrel-stave oligomeric cluster formed by antiparallel double-stranded helical dimers (DH). In a tentative model, the pentamer of dimers may be stabilized by aromatic Trp-Trp and cation-π Trp-Lys interactions between the neighboring DHs. The inhibiting effect of gA on the [Lys3]gA-induced leakage can be attributed to breaking of cation-π interactions, which prevents peptide clusterization and pore formation.

Channel-forming bacterial toxins in biosensing and macromolecule delivery

To intoxicate cells, pore-forming bacterial toxins are evolved to allow for the transmembrane traffic of different substrates, ranging from small inorganic ions to cell-specific polypeptides. Recent developments in single-channel electrical recordings, X-ray crystallography, protein engineering, and computational methods have generated a large body of knowledge about the basic principles of channel-mediated molecular transport. These discoveries provide a robust framework for expansion of the described principles and methods toward use of biological nanopores in the growing field of nanobiotechnology. This article, written for a special volume on "Intracellular Traffic and Transport of Bacterial Protein Toxins", reviews the current state of applications of pore-forming bacterial toxins in small- and macromolecule-sensing, targeted cancer therapy, and drug delivery. We discuss the electrophysiological studies that explore molecular details of channel-facilitated protein and polymer transport across cellular membranes using both natural and foreign substrates. The review focuses on the structurally and functionally different bacterial toxins: gramicidin A of Bacillus brevis, α-hemolysin of Staphylococcus aureus, and binary toxin of Bacillus anthracis, which have found their "second life" in a variety of developing medical and technological applications.

N-terminally glutamate-substituted analogue of gramicidin A as protonophore and selective mitochondrial uncoupler

Limited uncoupling of oxidative phosphorylation could be beneficial for cells by preventing excessive generation of reactive oxygen species. Typical uncouplers are weak organic acids capable of permeating across membranes with a narrow gap between efficacy and toxicity. Aimed at designing a nontoxic uncoupler, the protonatable amino acid residue Glu was substituted for Val at the N-terminus of the pentadecapeptide gramicidin A (gA). The modified peptide [Glu1]gA exhibited high uncoupling activity in isolated mitochondria, in particular, abolishing membrane potential at the inner mitochondrial membrane with the same or even larger efficacy as gA. With mitochondria in cell culture, the depolarizing activity of [Glu1]gA was observed at concentrations by an order of magnitude lower than those of gA. On the contrary, [Glu1]gA was much less potent in forming proton channels in planar lipid bilayers than gA. Remarkably, at uncoupling concentrations, [Glu1]gA did not alter cell morphology and was nontoxic in MTT test, in contrast to gA showing high toxicity. The difference in the behavior of [Glu1]gA and gA in natural and artificial membranes could be ascribed to increased capability of [Glu1]gA to permeate through membranes and/or redistribute between different membranes. Based on the protective role of mild uncoupling, [Glu1]gA and some other proton-conducting gA analogues may be considered as prototypes of prospective therapeutic agents.