The global spread of antibiotic resistance is forcing the scientific community to find new molecular strategies to counteract it. Deep functional profiling of microbiomes provides an alternative source for the discovery of novel antibiotic producers and probiotics. Recently, we implemented this ultrahigh-throughput screening approach for the isolation of Bacillus pumilus strains efficiently producing the ribosome-targeting antibiotic amicoumacin A (Ami). Proteomics and metabolomics revealed essential insight into the activation of Ami biosynthesis. Here, we applied omics to boost Ami biosynthesis, providing the optimized cultivation conditions for high-scale production of Ami. Ami displayed a pronounced activity against Lactobacillales and Staphylococcaceae, including methicillin-resistant Staphylococcus aureus (MRSA) strains, which was determined using both classical and massive single-cell microfluidic assays. However, the practical application of Ami is limited by its high cytotoxicity and particularly low stability. The former is associated with its self-lactonization, serving as an improvised intermediate state of Ami hydrolysis. This intramolecular reaction decreases Ami half-life at physiological conditions to less than 2 h, which is unprecedented for a terminal amide. While we speculate that the instability of Ami is essential for Bacillus ecology, we believe that its stable analogs represent attractive lead compounds both for antibiotic discovery and for anticancer drug development.

Gram-negative pathogens represent an urgent threat due to their intrinsic and acquired antibiotic resistance. Many recent drug candidates display prominent antimicrobial activity against Gram-positive bacteria being inefficient against Gram-negative pathogens. Ultrahigh-throughput, microfluidics-based screening techniques represent a new paradigm for deep profiling of antibacterial activity and antibiotic discovery. A key stage of this technology is based on single-cell cocultivation of microbiome biodiversity together with reporter fluorescent pathogen in emulsion, followed by the selection of reporter-free droplets using fluorescence-activated cell sorting. Here, a panel of reporter strains of Gram-negative bacteria Escherichia coli was developed to provide live biosensors for precise monitoring of antimicrobial activity. We optimized cell morphology, fluorescent protein, and selected the most efficient promoters for stable, homogeneous, high-level production of green fluorescent protein (GFP) in E. coli. Two alternative strategies based on highly efficient constitutive promoter pJ23119 or T7 promoter leakage enabled sensitive fluorescent detection of bacterial growth and killing. The developed live biosensors were applied for isolating potent E. coli-killing Paenibacillus polymyxa P4 strain by the ultrahigh-throughput screening of soil microbiome. The multi-omics approach revealed antibiotic colistin (polymyxin E) and its biosynthetic gene cluster, mediating antibiotic activity. Live biosensors may be efficiently implemented for antibiotic/probiotic discovery, environmental monitoring, and synthetic biology.

The biodiversity of microorganisms is maintained by intricate nets of interactions between competing species. Impaired functionality of human microbiomes correlates with their reduced biodiversity originating from aseptic environmental conditions and antibiotic use. Microbiomes of wild animals are free of these selective pressures. Microbiota provides a protecting shield from invasion by pathogens in the wild, outcompeting their growth in specific ecological niches. We applied ultrahigh-throughput microfluidic technologies for functional profiling of microbiomes of wild animals, including the skin beetle, Siberian lynx, common raccoon dog, and East Siberian brown bear. Single-cell screening of the most efficient killers of the common human pathogen Staphylococcus aureus resulted in repeated isolation of Bacillus pumilus strains. While isolated strains had different phenotypes, all of them displayed a similar set of biosynthetic gene clusters (BGCs) encoding antibiotic amicoumacin, siderophore bacillibactin, and putative analogs of antimicrobials including bacilysin, surfactin, desferrioxamine, and class IId cyclical bacteriocin. Amicoumacin A (Ami) was identified as a major antibacterial metabolite of these strains mediating their antagonistic activity. Genome mining indicates that Ami BGCs with this architecture subdivide into three distinct families, characteristic of the B. pumilus, B. subtilis, and Paenibacillus species. While Ami itself displays mediocre activity against the majority of Gram-negative bacteria, isolated B. pumilus strains efficiently inhibit the growth of both Gram-positive S. aureus and Gram-negative E. coli in coculture. We believe that the expanded antagonistic activity spectrum of Ami-producing B. pumilus can be attributed to the metabolomic profile predetermined by their biosynthetic fingerprint. Ultrahigh-throughput isolation of natural probiotic strains from wild animal microbiomes, as well as their metabolic reprogramming, opens up a new avenue for pathogen control and microbiome remodeling in the food industry, agriculture, and healthcare.

Oxydifficidin is a natural polyketide antibiotic that has long been recognized as a ribosome-targeting agent that inhibits protein synthesis. In this paper, we describe Bacillus velezensis strain EV17 and compare its complete genome sequence with that of the previously characterized B. velezensis strain K-3618 and the difficidin biosynthetic gene cluster (BGC) combined with mass spectrometry to elucidate the production of oxydifficidin by these strains. Toeprinting and small fluorescent peptide assays showed that isolated oxydifficidin induces a generalized inhibition of translation at every step of elongation in protein biosynthesis. In previous studies, it has been demonstrated that oxydifficidin targets bL12 protein. Although spontaneous mutations conferring resistance to oxydifficidin in ribosomal protein bL12 located relatively close to the thiostrepton binding site on uL11, our data show that oxydifficidin binding does not interfere with thiostrepton, thereby refining previous findings about its putative ribosomal target. We are the first to show that this compound does not affect eukaryotic translation and has two orders of magnitude lower effect on eukaryotic cells compared to bacteria. These facts are important to further investigate its potential as a bioprotectant against phytopathogens or even as a therapeutic agent.

Natural compounds continue to serve as the most fruitful source of new antimicrobials. Analysis of bacterial genomes have revealed that the biosynthetic potential of antibiotic producers by far exceeds the number of already discovered structures. However, due to the repeated discovery of known substances, it has become necessary to change both approaches to the search for antibiotics and the sources of producer strains. The pressure of natural selection and the diversity of interactions in symbiotic communities make animal microbiomes promising sources of novel substances. Here, microorganisms associated with various animals were examined in terms of their antimicrobial agents. The application of alternative cultivation techniques, ultrahigh-throughput screening, and genomic analysis facilitated the investigation of compounds produced by unique representatives of the animal microbiota. We believe that new strategies of antipathogen defense will be discovered by precisely studying cell–cell and host–microbe interactions in microbiomes in the wild.

The transition to organic farming is one of the most desirable achievements of our time. Rational use of modern organic farming approaches not only enables the reduction of costs and increased yields but also limits the risks associated with the use of pesticides and chemicals. Despite the widest practical application of numerous biocontrol agents based on Bacillus strains, their metabolome, including the main active substances, often remains unknown. In order to understand the basic principles of the functioning of the Bacillus velezensis K-3618 strain, widely used in organic farming, we studied its spectrum of antimicrobial metabolites in detail. It was shown that the main antimicrobial agents of the B. velezensis K-3618 are representatives of the macrolactin family. Macrolactins are a group of 24-membered lactones, including macrolactin A (MLN A) and its acylated analogs 7-O-malonyl macrolactin A (mal-MLN A) and 7-O-succinyl macrolactin A (suc-MLN A), mostly active against Gram-positive bacterial pathogens, including multidrug-resistant strains. Suc-MLN A was the most potent antimicrobial, highly active (MIC = 0.1 μg/mL) against the common human pathogen methicillin-resistant Staphylococcus aureus (MRSA). It was shown that the primary mechanism of action of MLN A-based macrolactins is protein translation inhibition. Acylated macrolactins outperform MLN A in the prokaryotic cell-free system, displaying high efficiency in low micromolar concentrations. We observed that acylated MLN A analogs undergo pathogen-mediated biotransformation to MLN F analogs, having the antimicrobial activity reduced by two orders of magnitude. Hence, both acylation of MLNs and stabilization of the MLN A core are essential for the creation of new synthetic MLNs with improved antimicrobial activity and stability. While the counteraction against the observed biotransformations is essential to create new macrolactin-based antimicrobials for clinical practice, we speculate that these degradability modes are essential for bacterial ecology, and they are highly conserved in many Bacillus species from various ecological niches.