What Goes around Comes around-A Comparative Study of the Influence of Chemical Modifications on the Antimicrobial Properties of Small Cyclic Peptides

Abstract

Tryptophan and arginine-rich cyclic hexapeptides of the type cyclo-RRRWFW combine high antibacterial activity with rapid cell killing kinetics, but show low toxicity in human cell lines. The peptides fulfil the structural requirements for membrane interaction such as high amphipathicity and cationic charge, but membrane permeabilisation, which is the most common mode of action of antimicrobial peptides (AMPs), could not be observed. Our current studies focus on elucidating a putative membrane translocation mechanism whereupon the peptides might interfere with intracellular processes. These investigations require particular analytical tools: fluorescent analogues and peptides bearing appropriate reactive groups were synthesized and characterized in order to be used in confocal laser scanning microscopy and HPLC analysis. We found that minimal changes in both the cationic and hydrophobic domain of the peptides in most cases led to significant reduction of antimicrobial activity and/or changes in the mode of action. However, we were able to identify two modified peptides which exhibited properties similar to those of the cyclic parent hexapeptide and are suitable for subsequent studies on membrane translocation and uptake into bacterial cells.

Figure 1

Chemical structures of cWFW and the four fluorescent labels used in this study. Fluorophores are depicted coupled to the respective amino acid (latter given in brackets) used to introduce the labels into the peptide ring.

Figure 2

Retention times t_R of the cyclic hexapeptides. The influence of fluorescent labels (dark grey), lysine substitution (light grey) and double modifications (shaded) on interaction with the hydrophobic column is shown compared to the parent peptide cWFW (black).

Figure 3

CD spectra of the cyclic hexapeptides in different solvent systems. (A) Lysine-substituted and (B) fluorescent-labeled peptides compared to cWFW in phosphate buffer. The effect of membrane-mimicking additives on the structure of selected lysine peptides is shown in (C): phosphate buffer (solid lines), 25 mM SDS (dashed lines) and 10 mM POPG-SUVs (dotted lines). Peptide concentration was 100 µM.

Figure 4

Bacterial membrane permeabilisation determined with flow cytometry using propidium iodide (PI). PI influx after incubation with cWFW, fluorescent-labeled and K-substituted cyclic hexapeptides was tested into B. subtilis DSM 347 (A,C) and E. coli DH5α (B,D) at respective MICs. The antibiotic polymyxin B (5 µM) and the helical model peptide KLA-1 (KLALKLALKALKAALKLA-NH₂, 5 µM) obtain a high membrane permeabilising potential and served as positive control in Gram-negative and Gram-positive cells, respectively (solid line) [20,34]. Preparations without peptide were used as negative control (dotted line).

Figure 5

HPLC-based investigation of peptide uptake into HeLa S cells. (A) Amount of non-modified, free peptide in the supernatant after 1 h incubation at 37 °C with cR2[Cu] (18 nmol) and Fluos-penetratin (9 nmol), (B) non-modified peptide after extensive washing of cells and exposure to diazotized 2-nitroaniline. For each peptide, two independent experiments were performed in triplicates.