Cesin, a short natural variant of nisin, displays potent antimicrobial activity against major pathogens despite lacking two C-terminal macrocycles

Abstract

Nisin is a widely used lantibiotic owing to its potent antimicrobial activity and its food-grade status. Its mode of action includes cell wall synthesis inhibition and pore formation, which are attributed to the lipid II binding and pore-forming domains, respectively. We discovered cesin, a short natural variant of nisin, produced by the psychrophilic anaerobe Clostridium estertheticum. Unlike other natural nisin variants, cesin lacks the two terminal macrocycles constituting the pore-forming domain. The current study aimed at heterologous expression and characterization of the antimicrobial activity and physicochemical properties of cesin. Following the successful heterologous expression of cesin in Lactococcus lactis, the lantibiotic demonstrated a broad and potent antimicrobial profile comparable to that of nisin. Determination of its mode of action using lipid II and lipoteichoic acid binding assays linked the potent antimicrobial activity to lipid II binding and electrostatic interactions with teichoic acids. Fluorescence microscopy showed that cesin lacks pore-forming ability in its natural form. Stability tests have shown the lantibiotic is highly stable at different pH values and temperature conditions, but that it can be degraded by trypsin. However, a bioengineered analog, cesin R15G, overcame the trypsin degradation, while keeping full antimicrobial activity. This study shows that cesin is a novel (small) nisin variant that efficiently kills target bacteria by inhibiting cell wall synthesis without pore formation. IMPORTANCE The current increase in antibiotic-resistant pathogens necessitates the discovery and application of novel antimicrobials. In this regard, we recently discovered cesin, which is a short natural variant of nisin produced by the psychrophilic Clostridium estertheticum. However, its suitability as an antimicrobial compound was in doubt due to its structural resemblance to nisin(1-22), a bioengineered short variant of nisin with low antimicrobial activity. Here, we show by heterologous expression, purification, and characterization that the potency of cesin is not only much higher than that of nisin(1-22), but that it is even comparable to the full-length nisin, despite lacking two C-terminal rings that are essential for nisin's activity. We show that cesin is a suitable scaffold for bioengineering to improve its applicability, such as resistance to trypsin. This study demonstrates the suitability of cesin for future application in food and/or for health as a potent and stable antimicrobial compound.

Keywords: Clostridium estertheticum; bacteriocin; cesin; lipid II; nisin.

Fig 1

Primary structures of cesin and nisin. Regions and amino acid residues involved in the lantibiotics’ antimicrobial activity are highlighted.

Fig 2

Heterologous expression of cesin in Lactococcus lactis using the nisin-controlled gene expression system. (A) The hybrid design for heterologous expression of cesin core peptide using nisin leader peptide (hybrid peptide). (B) SDS-polyacrylamide gel electrophoresis of the expressed hybrid peptide. (C) MALDI-TOF MS analysis of the hybrid peptide. The observed mass was 4,462.9 Da compared to a predicted mass of 4,459.2 Da of a modified hybrid peptide with five dehydrations. (D) NEM alkylation assay to determine the level of cyclization. The mass of the major peak after addition of NEM (below) was consistent with the mass of the hybrid peptide before addition of NEM (top), confirming majority of the peptide lacked free cysteines. (E) Antimicrobial activity assay of modified cesin core peptide after cleavage of nisin leader peptide using NisP.

Fig 3

Antimicrobial mode of action of cesin in comparison to nisin against Lactococcus lactis MG1363. (A) Spot-on-lawn-based lipid II binding assay. Cesin (8 µg) and nisin (2 µg) were spotted adjacent to lipid II (300 µm, 3 µL) and the antimicrobial activity determined. Daptomycin (2 µg) and water were used as controls. (B) Growth curve-based lipid II binding assay. Cesin, nisin, and nisin(1–22) (5 × MIC) and 2 µL lipid II (0.6 mol/L) were added and microbial growth was monitored through spectrophotometry. DMSO was used as a control. (C) Membrane pore-forming ability of cesin, nisin, and nisin(1–22) (2 × MIC) in microbial cells were determined using a combination of microscopy and fluorescent dyes, SYTO-9 (membrane permeable), and propidium iodide (membrane impermeable). (D) Time-killing curves of cesin, nisin, and nisin(1–22) (10 × MIC) against the bacteria. (E) Spot-on-lawn-based LTA binding assay. Cesin and nisin (8 µg) were spotted adjacent to LTA (1 mg/mL, 3 µL) and the antimicrobial activity. Water was used as a control.

Fig 4

Effects of the dltA (lmo0974) gene deletion in L. monocytogenes. (A) The ΔdltA mutant has increased sensitivity to both cesin and nisin than the WT strain. (B) The ΔdltA mutant has increased binding affinity to the positively charged cytochrome c than the WT strain. The data support interaction between the positively charged cesin and negatively charged cell wall components.

Fig 5

Bioengineered analogs of cesin with the C-terminal pore-forming domains of nisin. (A) Top: cesin NK is a hybrid of cesin and nisin(23–34). Bottom: cesin NMK is a hybrid of cesin(1–20) and nisin(21–34). (B) Antimicrobial activity of the bioengineered analogs of cesin against L. lactis MG1363 and E. faecium.

Fig 6

Antimicrobial activity of cesin analogs against L. lactis MG1363. Cesin W/G, cesin R/G, and cesin K/G analogs were bioengineered after W4G, R15G, and K21G amino acid substitution, respectively.

Fig 7

Thermal stability of cesin (A) compared to nisin (B). pH stability of cesin (C) compared to nisin (D).

Fig 8

Variations in stability of cesin and nisin to proteolytic cleavage. Relative antimicrobial activity of the lantibiotics after exposure to (A) nisin resistance protein and (B) different proteolytic enzymes. (C) Residual antimicrobial activity of cesin and its analogs W/G, cesin R/G, and cesin K/G with glycine substitution of W4, R15, and K21, respectively, relative to cesin after exposure to trypsin.

Fig 9

Hemolytic activity of cesin in comparison to nisin. (A) The release of hemoglobin from the sheep blood erythrocytes incubated with cesin and nisin at concentrations ranging from 1 to 64 µg/mL. Cells treated with phosphate-buffered saline (PBS) only were used as no-lysis controls. (B) The level of hemolysis of cesin and nisin relative to 10% Triton X-100, which was used as the positive control.