Specificity and mechanism of action of alpha-helical membrane-active peptides interacting with model and biological membranes by single-molecule force spectroscopy

Abstract

In this study, to systematically investigate the targeting specificity of membrane-active peptides on different types of cell membranes, we evaluated the effects of peptides on different large unilamellar vesicles mimicking prokaryotic, normal eukaryotic, and cancer cell membranes by single-molecule force spectroscopy and spectrum technology. We revealed that cationic membrane-active peptides can exclusively target negatively charged prokaryotic and cancer cell model membranes rather than normal eukaryotic cell model membranes. Using Acholeplasma laidlawii, 3T3-L1, and HeLa cells to represent prokaryotic cells, normal eukaryotic cells, and cancer cells in atomic force microscopy experiments, respectively, we further studied that the single-molecule targeting interaction between peptides and biological membranes. Antimicrobial and anticancer activities of peptides exhibited strong correlations with the interaction probability determined by single-molecule force spectroscopy, which illustrates strong correlations of peptide biological activities and peptide hydrophobicity and charge. Peptide specificity significantly depends on the lipid compositions of different cell membranes, which validates the de novo design of peptide therapeutics against bacteria and cancers.

Figure 1. Tryptophan fluorescence emission spectra of MAPs with LUVs model membranes at 25 °C.

PC/PG (7:3 w/w) LUVs were used to mimic normal eukaryotic cell membranes in Panel (a,d); PC/Chol (8:1 w/w) LUVs were used to mimic prokaryotic cell membranes in Panel (b,e); PC/SM/PE/PS/Chol (4.35:4.35:1:0.3:1 w/w) LUVs were used to mimic cancer cell membranes in Panel (c,f). Symbols used are as follows: ▲ for V13L; ▼ for V13A; ◆ for V13S; ● for V13K; ★ for V13E; △ for A12L/A20L; ▽ for A20L; ◊ for L6A; and ☆ for L6A/L17A.

Figure 2. The schematic illustration of the peptide-functionalized AFM tip on different cell membranes.

The peptide was covalently coupled to the AFM tip via a heterobifunctional PEG interact with LUVs mimicking membranes (a), A. laidlawii (b), and HeLa cells (c).

Figure 3. The force measurement of the interaction between single peptide V13L and different LUVs model membranes and biological cell membranes.

Panel (a) denotes typical force curve presenting the interaction between V13L and chicken erythrocyte; Panel (b) denotes the interaction force curve of V13L interacting with chicken erythrocyte after blocking by Tris-HCl buffer; Panel (c) denotes the histogram of unbinding force between V13L and chicken erythrocyte; Panel (d) denotes the interaction probability of V13L with chicken erythrocyte; Panel (e) denotes the interaction force curve of V13L interaction with the PC/PG (7:3 w/w) after blocking by Tris-HCl buffer; Panel (f–h) denote the histogram of unbinding force between V13L and PC/PG (7:3 w/w), PC/Chol (8:1 w/w) and PC/SM/PE/PS/Chol (4.35:4.35:1:0.3:1 w/w), respectively; Panel (i) denotes the interaction force curve of V13L interaction with the A. laidlawii cells after blocking by Tris-HCl buffer; Panel (j–l) denotes the histogram of unbinding force between V13L and A. laidlawii cells, 3T3-L1 cells, HeLa cells, respectively.

Figure 4. The interaction probability of MAPs with different LUVs model membranes and biological cell membranes, and relationships of interaction probability and biological activity.

Panel (a) denotes the interaction probability of MAPs with PC/PG (7:3 w/w); Panel (b) denotes the interaction probability of MAPs with PC/Chol (8:1 w/w); Panel (c) denotes the interaction probability of MAPs with PC/SM/PE/PS/Chol (4.35:4.35:1:0.3:1 w/w); Panel (d) denotes the interaction probability of MAPs with A. laidlawii cells; Panel (e) denotes the interaction probability of MAPs with 3T3-L1 cells; and Panel (f) denotes the interaction probability of MAPs with HeLa cells. Panel (g) denotes the antibacterial activity (MIC) against gram-negative bacteria and the interaction probability on PC/PG (7:3 w/w). Panel (h) denotes IC₅₀ values against 3T3-LI cells and the interaction probability on PC/Chol (8:1 w/w). Panel (i) denotes IC₅₀ values against HeLa cells and the interaction probability on PC/SM/PE/PS/Chol (4.35:4.35:1:0.3:1 w/w). Panel (j) denotes the antibacterial activity (MIC) against gram-negative bacteria and the interaction probability on A. laidlawii. Panel (k) denotes IC₅₀ values against 3T3-LI cells and the interaction probability on 3T3-L1. Panel (l) denotes IC₅₀ values against HeLa cells and the interaction probability on HeLa cells. The peptides in this figure from left to right are V13L (1), A12L/A20L (2), V13K (3), V13E (4), and L6A/L17A (5). The experimental data are from Tables 1 and 2 and Fig. 4. All the correlation parameters of interaction probability and biological activity are >0.9.

Figure 5. The model for the interaction mechanism between single peptide and different types of cell membranes.

Panel (a) denotes peptide of V13K with positive charge interacting with the negatively charged lipid composition of membranes such as PC/PG (7:3 w/w) and PC/SM/PE/PS/Chol (4.35:4.35:1:0.3:1 w/w). Panel (b) denotes peptide of V13E with negative charged acid residue interaction with the negative charged lipid composition of membranes such as PC/PG (7:3 w/w) and PC/SM/PE/PS/Chol (4.35:4.35:1:0.3:1 w/w). Panel (c) denotes peptide of V13K with positive charge interaction with zwitterionic lipid composition of membranes such as PC/Chol (8:1 w/w).