Mechanistic Insight into the Early Stages of Toroidal Pore Formation by the Antimicrobial Peptide Smp24

Abstract

The antimicrobial peptide Smp24, originally derived from the venom of Scorpio maurus palmatus, is a promising candidate for further drug development. However, before doing so, greater insight into the mechanism of action is needed to construct a reliable structure-activity relationship. The aim of this study was to specifically investigate the critical early stages of peptide-induced membrane disruption. Single-channel current traces were obtained via planar patch-clamp electrophysiology, with multiple types of pore-forming events observed, unlike those expected from the traditional, more rigid mechanistic models. To better understand the molecular-level structures of the peptide-pore assemblies underlying these observed conductance events, molecular dynamics simulations were used to investigate the peptide structure and orientation both before and during pore formation. The transition of the peptides to transmembrane-like states within disordered toroidal pores occurred due to a peptide-induced bilayer-leaflet asymmetry, explaining why pore stabilization does not always follow pore nucleation in the experimental observations. To fully grasp the structure-activity relationship of antimicrobial peptides, a more nuanced view of the complex and dynamic mechanistic behaviour must be adopted.

Keywords: antimicrobial peptides; disordered toroidal pore; early-stage pore formation; mechanism of action; membrane pore; molecular dynamics simulations; patch-clamp electrophysiology.

Figure 1

Smp24 pore formation kinetics via patch clamp. (A) Time between addition of peptide to the bilayer and the occurrence of an irreversible disruption of the bilayer resistance. (B) Time between the addition of peptide to the bilayer and the observation of the first conductance event. Data shown are mean and standard deviation (n = 5), * indicates significant difference based on unpaired t-test (p < 0.05).

Figure 2

Representative examples of the current trace and amplitude histograms for the three different event types observed in the patch-clamp experiments. (A) Multilevel events, (B) spike events, (C) erratic events.

Figure 3

Characteristics of the insertion of Smp24 into the negative bilayer. (A) Configuration of the peptide in the N-terminal anchored stage of the insertion. Interactions occur via two of the four lysine residues (blue), the N-terminal isoleucine (green) and the position 4 phenylalanine (purple). (B) Changes over time in the Z-axis centre of mass of peptide (yellow) and N-terminal (blue) relative to the phosphor atoms of the top leaflet (red) and centre of the bilayer (black). (C) Changes over time in the tilt angle relative to the bilayer norm of the helical region from residues 1–12. (D) Cumulative changes in the local helical rotation of residue 2–10 during the rotational stage of the insertion process. Figures shown are based on one simulation; corresponding figures for all repeats can be found in Supplementary Materials S2 (Figure S2.2–S2.4).

Figure 4

(A) Three-dimensional structure of Smp24 inserted into a PCPG bilayer. Blue = primary helix (r1–13), magenta = secondary helix (r14–17), green = glycine linker (r18–20), red = tail (r21–24), orange = lipid phosphor atoms. (B) Partial density profiles of Smp24 inserted into a PCPG bilayer, with positions relative to the centre of the bilayer. Black = lipid acyl chains, yellow = lipid glycerol esters, brown = lipid headgroups with phosphates; peptide regions/colours are the same as in (A). (C) Per residue RMSF of Smp24 after insertion into the negative bilayer. Figures shown are based on one simulation, corresponding figures for all repeats can be found in Supplementary Materials S3 (Figure S2.1–S2.2).

Figure 5

Three-dimensional models of the two pore-associated configurations seen for Smp24 in the long_pbcg_1–3 simulations. (A) The helical regions of Smp24 are aligned with the curvature of the pore interface with the secondary helix positioned within the top of the pore. (B) The helical regions are aligned in reverse such that the primary helix is positioned within the top half of the pore.

Figure 6

Examples of transmembrane peptide configurations in the multi-peptide pore simulations. (A,B) Example of peptide with the primary helical region positioned the deepest in the pore lumen (frontal and sideways view), (C,D) example of peptide with the secondary helix positioned the deepest in the pore lumen (frontal and sideways views), (E) two peptides in transmembrane configurations in the same pore, (F) example of a top-down view of a pore with multiple peptides associated with the pore interface.

Figure 7

Translocation of lipids through the membrane pore during the simulations. Results are shown as the average ± SD lipid density in the bottom leaflet over time for the pore simulations. (A) Single peptide pore models (n = 3), (B) multi-peptide pore models (n = 5). Black = combined density for both DOPC and DOPG lipids divided by 2, red = density of DOPC lipids, green = density of DOPG lipids.

Figure 8

Proposed molecular-level structures corresponding to the conductance events observed in the patch-clamp experiments. (A) The short-lived spike events are caused by unsupported toroidal pores without peptides in the pore lumen. (B) The longer-lived multilevel events are caused by supported toroidal pores with peptides taking part in the pore structure. (C) General disruptions to the bilayer structure such as membrane thinning could be the reason for the erratic events. (D) Micellar-like aggregates within the bilayer could also be the reason for erratic events.