Global gene-mining strategy for searching nonribosomal peptides as antimicrobial agents from microbial sources

Abstract

Foodborne pathogenic microorganism contamination and the emergence of drug-resistant bacteria are serious threats to human public health, and cationic nonribosomal peptides (CNRPs) represent a potential solution to this threat. In this report, a global gene mining strategy was used to evaluate 6563 bacterial genomes, including Bacillus sp, Paenibacillus sp and Brevibacillus sp. Using antiSMASH with a Python script filter, we discovered that the Bacillus subtilis group, Paenibacillus polymyxa, Brevibacillus laterosporus and Brevibacillus brevis most likely produce a larger number and greater variety of CNRPs than other Bacilli bacteria. Amino acids lysine, ornithine, 2,4-diaminobutyric acid, leucine, valine, threonine and serine were incorporated in higher abundance (amino acid preference) into these antimicrobial NRPs. Furthermore, flat colony counting, acid-base titrations and Ubbelohde viscometry measurements indicated that brevilaterins from Brevibacillus laterosporus inhibited the growth of various microbes during the first four days and showed exceptional antimicrobial activity against various microbial contaminations in raw milk. Raw milk containing brevilaterins maintained its initial acidity and viscosity over two days. This study should facilitate the identification of new antimicrobial peptides derived from bacteria and their artificial design and modification, allowing those suitable microorganisms identified to be fully utilized in antimicrobial agent production and food industry applications.