pH-triggered pore-forming peptides with strong composition-dependent membrane selectivity

Abstract

Peptides that self-assemble into nanometer-sized pores in lipid bilayers could have utility in a variety of biotechnological and clinical applications if we can understand their physical chemical properties and learn to control their membrane selectivity. To empower such control, we have used synthetic molecular evolution to identify the pH-dependent delivery peptides, a family of peptides that assemble into macromolecule-sized pores in membranes at low peptide concentration but only at pH < $6. Further advancements will also require better selectivity for specific membranes. Here, we determine the effect of anionic headgroups and bilayer thickness on the mechanism of action of the pH-dependent delivery peptides by measuring binding, secondary structure, and macromolecular poration. The peptide pHD15 partitions and folds equally well into zwitterionic and anionic membranes but is less potent at pore formation in phosphatidylserine-containing membranes. The peptide also binds and folds similarly in membranes of various thicknesses, but its ability to release macromolecules changes dramatically. It causes potent macromolecular poration in vesicles made from phosphatidylcholine with 14 carbon acyl chains, but macromolecular poration decreases sharply with increasing bilayer thickness and does not occur at any peptide concentration in fluid bilayers made from phosphatidylcholine lipids with 20-carbon acyl chains. The effects of headgroup and bilayer thickness on macromolecular poration cannot be accounted for by the amount of peptide bound but instead reflect an inherent selectivity of the peptide for inserting into the membrane-spanning pore state. Molecular dynamics simulations suggest that the effect of thickness is due to hydrophobic match/mismatch between the membrane-spanning peptide and the bilayer hydrocarbon. This remarkable degree of selectivity based on headgroup and especially bilayer thickness is unusual and suggests ways that pore-forming peptides with exquisite selectivity for specific membranes can be designed or evolved.

The effect of membrane thickness on the membrane permeabilizing activity of the cyclic lipopeptide tolaasin II

“…This was explained by the increasing energy requirement of membrane deformations to hydrophobically match peptide pores shorter than the relaxed membrane thickness (Bobone et al, 2013). A similar effect of membrane thickness was also observed for pHD15 (Kim et al, 2021) and maculatin 1.1 (Lee et al, 2018), other AMPs that are proposed to span the membrane to induce leakage.…”

“…brown blotch disease (Soler-Rivas et al, 1999). To probe membrane thickness effects, an established strategy was followed and the activity of tolaasin II against membranes of a homologous series of phosphatidylcholines with symmetric, monounsaturated chains with lengths from 14:1 to 20:1 was tested (Ashrafuzzaman et al, 2008;Bobone et al, 2012;Bobone et al, 2013;Lee et al, 2018;Henderson et al, 2019;Kim et al, 2021). These lipids form stable, fluid bilayers at ambient temperature, are well characterized (Kucerka et al, 2008;Kucerka et al, 2009;Frallicciardi et al, 2022) and commercially available at high purity.…”

Integrated Design of a Membrane‐Lytic Peptide‐Based Intravenous Nanotherapeutic Suppresses Triple‐Negative Breast Cancer

“…[ 11 ] To overcome the translational challenges of peptide‐based drugs, including low apparent activity as compared to small‐molecule drugs and poor proteolytic stability and pharmacokinetics, a wide array of cross‐disciplinary approaches have emerged in recent literature, including the development of peptidomimetics, [ 12 ] peptide‐drug conjugates, [ 13 , 14 ] stapled peptides, [ 15 , 16 ] and peptide‐based nanoparticles. [ 17 ] Despite the tremendous progress made, clinical translation of ACPs still faces steep barriers, including selective targeting of the subtle distinctions between cancerous and healthy cell membranes [ 18 , 19 , 20 , 21 ] and enabling clinically relevant routes of administration. [ 22 , 23 ] In particular, the clinically preferred intravenous delivery route remains elusive for most of the proposed peptide and peptide‐analog chemistries as the journey between the site of injection and site of action present many biological barriers that can undermine peptide‐based therapeutics.…”

In Search of a Molecular View of Peptide–Lipid Interactions in Membranes

“…There are clearly many cases where inserted hydrophobic peptides tilt away from the membrane normal, as for example the Vpu peptide of HIV-1. , However, peptides are not prone to matching P L to H c by tilting or changing conformation. , Peptide tilting occurs mainly because the configurational entropy increases if the helix ends can span a cone centered on the membrane normal . Even if P L < H c , peptide helices still tilt away from the membrane normal, despite exacerbating the hydrophobic mismatch, because the configurational (orientational) entropy increases. In addition, tilting occurs to allow aromatic residues that cap the helix to reside in the membrane interface or, but to a lesser extent, to allow cationic residues to reside just above the lipid phosphate groups.…”

“…Some naturally occurring peptides with these properties have been studied, such as glycophorin A from human erythrocytes and the phage M13-coat protein. More convenient for laboratory study are peptides especially designed to insert in membranes, such as TMX-1 and -3, ,, the KALP and WALP peptides, ,,, the pH-sensitive peptides pHLIP , and pHD15, peptides based on poly-Leu, ,,− ,, or the macrolittins. , …”