Rombocin, a Short Stable Natural Nisin Variant, Displays Selective Antimicrobial Activity against Listeria monocytogenes and Employs a Dual Mode of Action to Kill Target Bacterial Strains

Abstract

Nisin, with its unique mode of action and potent antimicrobial activity, serves as a remarkable inspiration for the design of novel antibiotics. However, peptides possess inherent weaknesses, particularly their susceptibility to proteolytic degradation, such as by trypsin, which limits their broader applications. This led us to speculate that natural variants of nisin produced by underexplored bacterial species can potentially overcome these limitations. We carried out genome mining of two Romboutsia sedimentorum strains, RC001 and RC002, leading to the discovery of rombocin A, which is a 25 amino acid residue short nisin variant that is predicted to have only four macrocycles compared to the known 31-35 amino acids long nisin variants with five macrocycles. Using the nisin-controlled expression system, we heterologously expressed fully modified and functional rombocin A in Lactococcus lactis and demonstrated its selective antimicrobial activity against Listeria monocytogenes. Rombocin A uses a dual mode of action involving lipid II binding activity and dissipation of the membrane potential to kill target bacteria. Stability tests confirmed its high stability at different pH values, temperatures, and in particular, against enzymatic degradation. With its gene-encoded characteristic, rombocin A is amenable to bioengineering to generate novel derivatives. Further mutation studies led to the identification of rombocin K, a mutant with enhanced bioactivity against L. monocytogenes. Our findings suggest that rombocin A and its bioengineered variant, rombocin K, are promising candidates for development as food preservatives or antibiotics against L. monocytogenes.

Keywords: four lanthionine rings; mode of action; mutagenesis; short nisin variant; specificity; stability.

Figure 1

Scheme of two major approaches involved new lantibiotic discovery and characterization. In the first approach (shown in green arrow), bacteriocin producing strains are identified through culture-based approaches. The second approach involves first the whole genome sequencing of bacterial isolates. This is followed by detection of bacteriocin biosynthetic gene clusters through genome mining tools such as BAGEL4. Identified bacteriocins are cloned into suitable expression systems such as nisin-controlled gene expression (NICE) system for expression in suitable host such as L. lactis. The culture-based and genome-based approaches then incorporate similar downstream processes involving peptide purification, mass and structural analysis, and antimicrobial activity analysis. A major advantage of the second approach is that it allows identified bacteriocins to be bioengineered for improved antimicrobial and physicochemical properties as shown by the yellow arrow. To avoid rediscovery of already known bacteriocins and take advantage of bioengineering strategies of the second approach, culture-based approaches can be linked to the later approach as shown by the blue arrow.

Figure 2

Rombocin A is the first bacteriocin discovered in the Romboutsia genus. (A) Two strains, RC001 and RC002, isolated from chilled vacuum packed meat were identified as members of species Romboutsia sedimentorum as shown by a phylogenetic tree created from the alignment of 16S rRNA sequences (1216–1217 bp) of the two strains and types strains of known species within the Romboutsia genus. (B) Genome mining identified a class I lanthipeptide gene cluster in the genome of R. sedimentorum RC002. The encoded bacteriocin has been named rombocin A. The gene cluster of rombocin A encodes the precursor peptide, romA, biosynthetic genes, romBTC, regulatory genes, romRK, immunity genes, romFEG and maturation protease, romP. Genes with similar functions are present in the biosynthetic gene cluster of nisin A as shown through color coding. (C) Amino acid sequences alignment of rombocin A and nisin variants. Gray, similar residues; black, identical residues. The rings and hinge region (NMK) are indicated. (D) Phylogenetic analysis of rombocin A and nisin variants. The order in which they branch shows the relatedness between them, and the branch length represents the phylogenetic distance (0.05 represents a scale for the phylogenetic distance). (E) The alignment of the amino acid sequences of rombocin A and nisin A. The functional domains, including lipid II binding site, pore formation domain, and hinge region are indicated. The first three rings of rombocin A differ from nisin A, as indicated by green-highlighted amino acid differences.

Figure 3

Heterologous expression of rombocin in L. lactis using the nisin modification machinery NisBTC. (A) The heterologous expression of rombocin core peptide using nisin leader peptide. (B) The new developed nisin-controlled gene expression (NICE) system. The traditional system involves two plasmids, pil3eBTC and pNZ-rombocin, where the modification machinery nisBTC and rombocin A are induced simultaneously by nisin. In contrast, our novel system utilizes a two-step induction process, where nisBTC is first induced by Zn²⁺ and nisin is added 3 h later. The rombocin, encoding the peptide rombocin with nisin leader attached; sczA, encoding the repressor of P_czcD; P_czcD, a Zn²⁺ inducible promoter; P_nisA, a nisin inducible promoter; nisB, nisT, and nisC, encoding nisin modification machinery NisBTC; rep, encoding plasmid replication protein; Cm^R, chloramphenicol resistance gene; and Em^R, erythromycin resistance gene. (C) Tricine-SDS-PAGE analysis of peptide expression using the two expression systems. Each lane contained peptides from 0.2 mL of supernatant. (D) MALDI-TOF MS analysis to evaluate the dehydration efficiency of the peptides expressed using both systems. (E) Screening the antibacterial activity of peptides after cleavage of nisin leader part using NisP. The white circle indicates the antibacterial halo caused by the rombocin peptide, leading to the inhibition of bacterial strain growth.

Figure 4

Antimicrobial activity of rombocin and its analogues against L. monocytogenes LMG10470. (A) The structures of wild-type (WT) rombocin and four bioengineered analogues, namely, rombocin A/P, rombocin I/K (referred to as rombocin K), rombocin M/NMK, and rombocin K/A, which were generated through amino acid substitutions at positions A9, I12, M20, and K25, respectively, as indicated by the yellow residues. (B) Tricine-SDS-PAGE gel analysis of the purified rombocin and its analogues. (C) Relative antimicrobial activity of the four rombocin analogues and wild-type rombocin against L. monocytogenes. (D) The structure of the bioengineered rombocin analog fused with the nisin A pore-forming domain (denoted as rom/nisin-tail). (E) Antimicrobial activity of the wild-type rombocin and rom/nisin-tail against L. monocytogenes. (F) Relative antimicrobial activity of the wild-type rombocin and rom/nisin-tail against L. monocytogenes.

Figure 5

Rombocin A binds to cell wall synthesis precursor lipid II and lipoteichoic acid (LTA). (A) A spot-on-lawn assay to assess the ability of rombocin to bind to the cell wall synthesis precursor lipid II. Nisin was used as the positive control, and daptomycin and water used as the negative control. (B) A spot-on-lawn assay to investigate the binding of rombocin to the LTA. Nisin was used as the positive control. (C) Growth curve-based binding assay to determine if rombocin binds to the cell wall synthesis precursor lipid II and LTA. The small peak at 2.5 h was caused by a slight interruption when turning off the VarioskanTM LUX microplate reader.

Figure 6

Time-dependent killing assay to determine the bacteriostatic or bactericidal activity of rombocin. 5-fold MIC of the lantibiotics was used to against L. lactis, along with nisin as a bactericidial control, and nisin(1–22) as a bacteriostatic control. The experiment was repeated three times, and standard deviation (SD) was calculated.

Figure 7

Effect of rombocin on the cellular membrane. (A) Fluorescence microscope pictures of L. lactis treated by 5× MIC antimicrobials for 15 min and stained with membrane-permeable SYTO-9 and membrane-impermeable propidium iodide stains. (B) Potassium leakage, as detected by the increase in fluorescence of the PBFI probe, after the addition of 5× MIC antimicrobials. At 2 min, antibiotics were added. (C) Changes in membrane potential of L. lactis as indicated by the increase in the fluorescence of DiSC3(5) probe after treatment of cells with 5× MIC antimicrobial. At 5 min, antimicrobial peptides were added.

Figure 8

Stability of rombocin and nisin. (A) pH stability of nisin Z. (B) pH stability of rombocin A. (C) Thermal stability of nisin Z. (B) Thermal stability of rombocin A. (E) Relative antimicrobial activity of the nisin Z after exposure to different proteolytic enzymes. (F) Relative antimicrobial activity of the rombocin A after exposure to different proteolytic enzymes.