Discovery, development and optimisation of a novel frog antimicrobial peptide with combined mode of action against drug-resistant bacteria

Abstract

Antimicrobial peptides (AMP) have emerged as promising candidates for addressing the clinical challenges posed by the rapid evolution of antibiotic-resistant microorganisms. Brevinins, a representative frog-derived AMP family, exhibited broad-spectrum antimicrobial activities, attacking great attentions in previous studies. However, their strong haemolytic activity and cytotoxicity, greatly limit their further development. In this work, we identified and characterised a novel brevinin-1 peptide, brevinin-1pl, from the skin secretions of the northern leopard frog, Rana pipiens. Like many brevinins, brevinin-1pl also displayed strong haemolytic activity, resulting in a lower therapeutic index. We employed several bioinformatics tools to analyse the structure and potential membrane interactions of brevinin-1pl, leading to a series of modifications. Among these analogues, des-Ala¹⁶-[Lys⁴]brevinin-1pl exhibited great enhanced therapeutic efficacy in both in vitro and in vivo tests, particularly against some antibiotics-resistant Escherichia coli strains. Mechanistic studies suggest that des-Ala¹⁶-[Lys⁴]brevinin-1pl may exert bactericidal effects through multiple mechanisms, including membrane disruption and DNA binding. Consequently, des-Ala¹⁶-[Lys⁴]brevinin-1pl holds promise as a candidate for the treatment of drug-resistant Escherichia coli infections.

Keywords: Antimicrobial peptide; Brevinin-1; DNA binding; Drug-resistant; LPS binding.

Graphical abstract

Fig. 1

(a) Brevinin-1pl precursor cDNA and corresponding amino acid sequence. The nucleotide sequence of signal peptide is labelled with double-underline, and the nucleotide sequence of the deduced mature peptide sequence is labelled with single-underline. The asterisk represents the stop code. (b) Multiple sequence alignment results of brevinin-1pl with similar peptides found in NCBI-BLAST database. Zappo colour scheme is used to label and reveal patterns of variations.

Fig. 2

(a) Predicted secondary structure of breinin-1pl. (b) Distribution of predicted secondary structure in residues. Red, blue and green represents possible distribution of helical, coil and extended structure, respectively. (c) Helical wheel diagram of possible helix part of brevinin-1pl. The hydrophobic face calculated by the HeliQuest server is labelled in red. (d) Possible binding mode of brevinin-1pl with the plasma membrane (mammalian) model (Red represents the extracellular side; blue represents the intracellular side) predicted by PPM server and virtualised in PyMOL program. Residues embedded in the membrane are labelled with three-letter code.

Fig. 3

CD spectra of brevinin-1pl and its analogues. The red line represents the results of peptides in 10 mM NH₄AC solution. The blue line represents the results of peptides in 50 % TFE buffer.

Fig. 4

Time-killing kinetic curves of brevinin-1pl (a) and des-Ala¹⁶-[Lys⁴]brevinin-1pl (b) against E. coli 8739, E. coli 13846, and E. coli 2340. The tested time is 180 min, and the treated concentrations of peptide range from 1 × MIC to 4 × MIC. Untreated cells are set as the growth control. DMSO (1 %) treated cells are set as the vehicle control group.

Fig. 5

(a) Binding affinity of brevinin-1pl, des-Ala¹⁶-[Lys⁴]brevinin-1pl and melittin with LPS. (b) CD spectrum of brevinin-1pl (left) and des-Ala¹⁶-[Lys⁴]brevinin-1pl (right) with or without LPS.

Fig. 6

Effects of brevinin-1pl and des-Ala¹⁶-[Lys⁴]brevinin-1pl on the membrane of three E. coli strains. (a) Changes in OM permeability in the presence or absence of peptides. The positive control is E. coli cells treated with melittin. (b) Changes in cytoplasmic membrane permeability in the presence or absence of peptides. Melittin is served as the positive control. (c) ONPG assay results of breinin-1pl and des-Ala¹⁶-[Lys⁴]brevinin-1pl against three E. coli strains. The negative group is untreated E. coli cells. The positive control is E. coli cells treated with Melittin. The addition group include only the solvent.

Fig. 7

Interaction of brevinin-1pl and des-Ala¹⁶-[Lys⁴]brevinin-1pl with plasmid DNA of E. coli. (a) Effects of des-Ala¹⁶-[Lys⁴]brevinin-1pl and buforin II on the migration of plasmid DNA (pBR322). (b) CD spectra result of des-Ala¹⁶-[Lys⁴]brevinin-1pl (left) and buforin II (right) binding with plasmid DNA. (c) The normalised changes of des-Ala¹⁶-[Lys⁴]brevinin-1pl and buforin II to the plasmid DNA in CD signal at 275 nm. (d) Molecular docking model of des-Ala¹⁶-[Lys⁴]brevinin-1pl with the DNA (PDB ID: 1BNA). The grey dash bond represents the hydrophobic interaction between the peptide and DNA. The blue bond represents formed hydrogen bonds. The green dash bond represents the π-stacking.

Fig. 8

The Kaplan-Meier-curves of des-Ala¹⁶-[Lys⁴]brevinin-1plin wax moth larvae infected with E. coli 13846 (a) and E. coli 2340 (b). The group of infected larvae treated with Rifampin (20 mg/kg) was set as the positive control.

Fig. 9

Effects of brevinin-1pl (a) and des-Ala¹⁶-[Lys⁴]brevinin-1pl (b) on the growth of tested cancerous cells H838 and MCF-7 and human healthy cells HaCaT. The concentrations of tested peptides range from 10^-9 to 10^-4 M. Cells without treatment employed as the growth control (shown as growth in the graph). DMSO (0.5 %) was set as the vehicle control, and Triton x-100 was employed as the positive control. Growth medium was employed as the blank control. The data represented show the S.E.M. calculated from nine replicates conducted in three independent experiments. No significance (ns), * (p < 0.05), and * ** * (p < 0.0001) compared to the vehicle control group, determined by One-Way ANOVA.