Mining and characterization of novel antimicrobial peptides from the large-scale microbiome of Shanxi aged vinegar based on metagenomics, molecular dynamics simulations and mechanism validation

Abstract

The study aimed to mine and characterize novel antimicrobial peptides (AMPs) from the Shanxi aged vinegar microbiome. Utilizing machine learning techniques, AlphaFold2 structure prediction and molecular dynamics simulations, six novel AMPs were innovatively mined from 98,539 peptides based on metagenomic data, of which one peptide secreted by Lactobacillus (named La-AMP) was experimentally validated to have remarkable bactericidal effects against Staphylococcus aureus (S. aureus) and Escherichia coli (E. coli) with high stability and no hemolytic activity. Scanning electron microscopy revealed that La-AMP caused irreversible damage to cell membranes of S. aureus and E. coli, a finding further confirmed by calcein-AM/propidium iodide staining. Additionally, La-AMP induced nucleic acid leakage and reactive oxygen species accumulation in bacterial cells. It was found to bind to DNA gyrase through salt bridges, hydrogen bonds, and hydrophobic interactions, ultimately inducing apoptosis. Thus, La-AMP exhibited encouraging promise as a valuable bioactive component for the development of natural preservatives.