Investigating Penetration and Antimicrobial Activity of Vector-Bicycle Conjugates

Abstract

Growing antibiotic resistance is rapidly threatening the efficacy of treatments for Gram-negative infections. Bicycle molecules, constrained bicyclic peptides from diverse libraries generated by bacteriophage display that bind with high affinity to a chosen target are a potential new class of antibiotics. The generally impermeable bacterial outer membrane currently limits the access of peptides to bacteria. The conjugation of membrane active peptides offers an avenue for outer membrane penetration. Here, we investigate which physicochemical properties of a specific membrane active peptide (MAP), derived from ixosin-B, could be tweaked to enhance the penetration of conjugates by generating multiple MAP-Bicycle conjugate variants. We demonstrate that charge and hydrophobicity are important factors, which enhance penetration and, therefore, antimicrobial potency. Interestingly, we show that induction of secondary structure, but not a change in amphipathicity, is vital for effective penetration of the Gram-negative outer membrane. These results offer insights into the ways vectors could be designed to deliver Bicycle molecules (and other cargos) through biological membranes.

Keywords: antibiotics; antimicrobial peptides; antimicrobial resistance; bicycle; membrane active peptides; outer membrane.

Figure 1

Schematic representation of the molecules used in this study. (A) Bicycle molecules are bicyclic peptides (blue–purple) constrained by a central scaffold (gray). (B) Membrane active peptides (MAPs) are peptides that can penetrate biological membranes. MAPs in this study were made from d-amino acids. (C) MAP-Bicycle conjugates (blue) are molecules in which the MAP is covalently attached to the Bicycle molecule. (D) MAP-d-Bicycle conjugates (green) have the same Bicycle molecule sequence but with d-amino acids.

Figure 2

CD spectra and helical projection wheels of MAPs investigated. (A–D) CD spectra for the MAP vectors AV1 (A), SV1 (B), SV3 (C), and SV4 (D) measured at a concentration of 0.15 mg/mL in a 10 mM sodium phosphate buffer, pH 7.0. The spectra were measured in either 0 mM (blue line) or 10 mM SDS (red line). (E) Helical projection wheels for MAP vectors AV1, SV3, and SV4. The d-amino acid residue and its number are shown close to its corresponding position. Positively charged residues are shown in blue pentagons, charged residues are shown in orange circles, and hydrophobic residues are shown in deep or light green squares or yellow circles, with higher hydrophobicity residues being closer to green and lower hydrophobicity residues being closer to yellow. μ_Η corresponds to the hydrophobic moment of each peptide, a measure of the amphipathicity of the helix (the greater the value, the more amphipathic the helix).

Figure 3

NPN uptake of different MAP vectors and MAP-Bicycle conjugates. (A) Representation of fluorescence timecourse experiment data generated using the NPN assay. An increase in fluorescence represents NPN incorporation in a lipid environment and, therefore, OM penetration. Percentage (%) NPN uptake (B–G) was calculated using the fluorescence at 25 min while comparison for the maximum fluorescence (F_max, H) was done using fluorescence after 1 h. %NPN uptake for MAP vectors or their equivalent MAP-Bicycle conjugate; (B) A1/AV1, (C) C4/CV4, (D) C10/CV10, (E) H3/HV3, (F) H6/HV6, (G) S3/SV3. The standalone MAP vector is in red, while the MAP-Bicycle conjugate is in blue. Error bars represent the standard deviation (n = 3). (H) Maximum fluorescence (F_max) for the MAP-Bicycle conjugate (pink) and its corresponding standalone MAP vector (red). In gray, the F_max of PMBN, a known OM penetrator, is shown for comparison. Error bars represent standard deviation (n = 3).

Figure 4

Aspects which affect penetration of the OM and inhibition of a periplasmic target for enhanced antimicrobial activity. (A) Bicycle molecules without a MAP vector are unable to penetrate the OM. MAP vectors can penetrate the OM but do not target the periplasmic target. MAP-Bicycle conjugates penetrate (albeit less efficiently) and inhibit the periplasmic target. Increased hydrophobicity and charge enhance penetration of the OM. (B) MAP amphipathicity did not affect penetration. (C) Secondary structure induction by MAP is important for penetration.