A novel bacteriocin against multiple foodborne pathogens from Lacticaseibacillus rhamnosus isolated from juice ferments: ATF perfusion-based preparation of viable cells, characterization, antibacterial and antibiofilm activity

Abstract

Foodborne pathogens and their biofilms pose a risk to human health through food chain. However, the bacteriocin resources combating this threat are still limited. Here, Lacticaseibacillus rhamnosus, one of the most used probiotics in food industry, was prepared on a large scale using alternating tangential flow (ATF) perfusion-based technology. Compared to the conventional fed-batch approach, ATF perfusion remarkably increased the viable cells of L. rhamnosus CLK 101 to 11.93 ± 0.14 log CFU/mL. Based on obtained viable cells, we purified and characterized a novel bacteriocin CLK_01 with a broad spectrum of activity against both Gram-positive and Gram-negative foodborne pathogens. LC-MS/MS analysis revealed that CLK_01 has a molecular mass of 701.49 Da and a hydrophobic amino acid composition of I–K–K–V-T-I. As a novel bacteriocin, CLK_01 showed high thermal stability and acid-base tolerance over 25–121 °C and pH 2–10. It significantly reduced cell viability of bacterial pathogens (p < 0.001), and strongly inhibited their biofilm formation. Scanning electron microscopy demonstrated deformation of pathogenic cells caused by CLK_01, leading to cytoplasmic content leakage and bacterial death. Summarily, we employed ATF perfusion to obtain viable L. rhamnosus, and presented that bacteriocin CLK_01 could serve as a promising biopreservative for controlling foodborne pathogenic bacteria and their biofilms.

Graphical abstract

Highlights

• •Bacteriocin-producing L. rhamnosus was prepared by alternating tangential flow.
• •Bacteriocin CLK_01 with thermal stability and pH tolerance was characterized.
• •CLK_01 exhibited a broad antimicrobial spectrum against foodborne pathogens.
• •The bacterial biofilm formation was inhibited upon CLK_01 treatment.
• •CLK_01 caused permeability and deformation of bacterial pathogenic cells.

Abbreviations

L. rhamnosus

Lacticaseibacillus rhamnosus

ATF

alternating tangential flow

LAB

lactic acid bacteria

BSA

bovine serum albumin

MIC

minimal inhibitory concentration

SEM

scanning electron microscopy

pI

isoelectric point

SD

standard deviation

MRS

medium: de Man: Rogosa and Sharp Medium

1. Introduction

In recent years, increasing cases of food contamination and food poisoning caused by foodborne pathogens have led to a growing awareness on the seriousness of microbial contamination in foods (Martin et al., 2022; Melo and Quintas, 2023). For example, as a serious public health issue, nontyphoidal Salmonella (NTS) is responsible for food-borne illnesses with an estimated 94 million cases globally (Ramatla et al., 2022). Pires et al. reported that foodborne infections caused by seven pathogens in Denmark resulted in a total expenditure of 434 million Euro in 1 year in a country with 5.8 million citizens (Pires et al., 2022). Bacterial biofilm refers to a special form of bacterial community in which bacteria adhere to the surface of an object during growth and are encapsulated in their own extracellular polysaccharide matrix, lipoprotein and other components (Highmore et al., 2022). Due to the spatial structure and barrier effect of the biofilm, it can increase the drug resistance by 10–1000 times, highly stable upon treatment of desiccation, food disinfectants, and antibiotics. Biofilms are easily formed on the surfaces of food products, various processing equipment, conveying pipes and ventilation equipment, and then contaminated by hand or air to cause food poisoning (Olanbiwoninu and Popoola, 2023). With the increasing demands on food quality, there is a high expectation for safe and healthy food additives and biopreservatives to control foodborne pathogens and their biofilms in the food industry (Barroug et al., 2021; Mantovani et al., 2022). Therefore, developing new antimicrobial strategies is becoming important and urgent in the area of food safety.

Lactic acid bacteria (LAB) are an inherent component of many foods and can enhance food quality by producing a variety of metabolites, such as organic acids, ethanol oxides, bacteriocins, etc. (Peter et al., 2021; Martin et al., 2022). Lacticaseibacillus rhamnosus (L. rhamnosus), a Gram-positive probiotic, was first isolated from the intestine of healthy humans, functioning in regulation of intestinal flora, elimination of toxins in foods, and improvement of body immunity, etc. (Gorbach, 2000; Khailova et al., 2017). L. rhamnosus strains are usually resistant to gastric juice and bile, being able to easily colonize the intestinal tract (Oberhelman et al., 1999; Westerik et al., 2018). The commercial products related to L. rhamnosus include fermented dairy products, baby food, and juice drinks. As a highly accepted probiotic, L. rhamnosus has great value in food industry (Gorbach, 2000; Westerik et al., 2018).

Although bacteriocins are suggested to be potential food preservatives because of their beneficial properties, the production of bacteriocins is limited by high cost and low-performance (Johnson et al., 2018). Generally, concentrated bacteriocins are prepared by costly cultivation of microorganisms and subsequent peptide purification, with low titer and limitations of expensive media (Teusink and Molenaar, 2017). In common probiotic fermentation and bacteriocin production, the density and biomass of the bacterial cells can be improved by optimizing such parameters as medium composition, pH, dissolved oxygen, and temperature (Yang et al., 2018). However, the accumulation of metabolites often leads to growth inhibition or death of the strains and reduction of viable bacterial cells (Selder et al., 2020; Teke and Pott, 2021). Alternating tangential flow (ATF) perfusion-based technology, applying a hollow fiber column intercepting system, allows for continuous drainage of the metabolites, while a peristaltic pump continuously replenishes the fresh medium for replacement, so as to mitigate the negative effects of metabolites on bacterial strain growth. This technology is characterized by short production time, high biomass and excellent cell viability (Granicher et al., 2020). In recent years, it has been widely applied to the high-density culture of mammalian cells, achieving great yields in small-size biosystems (Hein et al., 2021; Su et al., 2021). However, to our knowledge, this technology has not yet been introduced into the fermenting production of probiotic bacteria, including L. rhamnosus.

At present, the inappropriate use of chemical preservatives makes them easy to remain in human body, posing a risk to human health. In contrast, LAB bacteriocins are gradually used as antimicrobial substances in food industry (Guo et al., 2020), which are essentially natural macromolecular proteins or short peptides with strong antibacterial activity produced by food-grade LAB and synthesized by their ribosomes (Garcia-Gutierrez et al., 2020). LAB bacteriocins can protect foods from specific pathogenic bacteria as biological preservatives, conferring foods antimicrobial capacity like congenital immunity, and also offering the possibility of artificially regulating food microorganisms (Guo et al., 2020; Barroug et al., 2021). However, very few bacteriocins with antibacterial activity derived from L. rhamnosus have been reported. Moreover, the use of bacteriocin in the prevention and control of biofilm contamination in food processing is also rather unexplored.

In this study, we performed the large-scale fermentation of L. rhamnosus based on the ATF perfusion technology, and subsequently purified a novel bacteriocin CLK_01 from the obtained viable cells. The molecular mass and amino acid sequence composition of CLK_01 were determined by LC-MS/MS; its thermal and pH resistance within broad range, enzyme sensitivity, antimicrobial and antibiofilm activity against different Gram-positive and Gram-negative pathogens were also examined and analyzed.

2. Materials and methods

2.1. Pathogenic bacteria and screening of potential probiotic Lacticaseibacillus

All pathogenic bacterial strains used in this study, and their culture conditions, are shown in Table 1. L. rhamnosus CLK 101 was isolated from fermented juice samples. Briefly, mixed juice samples of 20 mL were diluted with sterile water and prepared into different concentration gradients, and 100 μL of five dilutions (10⁻¹, 10⁻⁴, 10⁻⁵, 10⁻⁶ and 10⁻⁷) were spread evenly on CaCO₃-added MRS medium (Solarbio, Beijing, China) at 37 °C for 48 h. To obtain pure strains, a single colony with calcium lysis circles was selected on an ultra-clean bench (Optec, Chongqing, China) with a sterile inoculation loop and incubated in MRS medium at 37 °C for 48 h. The morphological identification was performed by Gram staining and optical microscope (Zeiss Primo star, Jena, Germany) examination (Ullah et al., 2017). According to “Begey's Manual of Systematic Bacteriology” and “Manual for Systematic Identification of Common Bacteria” (Guerrero, 2001), LAB strains have several morphological characteristics, such as rod-like appearance, no flagella, no envelope, and Gram-positive staining. Therefore, bacteria with these characteristics were initially identified as candidate strains and stored in 30% (v/v) glycerol in a −20 °C refrigerator. All bacterial strains used in these experiments were stored in Faculty of Life Science and Technology, Kunming University of Science and Technology, Kunming, China.

Table 1.

List of pathogenic strains used in this study.

Bacteria	aStrain number	Conditions (°C)	Reason for selection
Gram-negative species
Vibrio Parahemolyticus	ATCC 17802	LB, 37	Foodborne pathogen
Escherichia coli	CMCC(B) 44102	LB, 37	Foodborne pathogen
Shigella dysenteriae	KUST_DS8	LB, 37	Foodborne pathogen
Salmonella ser. Choleraesuis	ATCC 13312	LB, 37	Foodborne pathogen
Salmonella ser. Enteritidis	CMCC(B) 50335	LB, 37	Foodborne pathogen
Salmonella ser. Paratyphi B	CMCC(B) 50094	LB, 37	Foodborne pathogen
Pseudomonas aeruginosa	ATCC 27853	LB, 37	Drinking water pathogen
Klebsiella pneumonia	ATCC 10031	BHI, 37	Pathogenic bacterium
Aeromonas hydrophila	ATCC 49140	RS, 30	Foodborne pathogen
Gram-positive species
Staphylococcus aureus	ATCC 6538	LB, 37	Food-borne pathogen
Staphylococcus argenteus	ATCC 17802	LB, 37	Foodborne pathogen
Micrococcus luteus	ATCC 4698	NA, 30	Foodborne pathogen
Streptococcus agalactiae	CMCC(B) 32116	BHI, 37	Livestock and aquatic product pathogen

^aATCC, American Type Culture Collection; CMCC, China Center of Medicine Culture Collection; KUST, Kunming University of Science and Technology.

2.2. Detection of antibacterial activity

The antibacterial activity of the candidate Lacticaseibacillus strain CLK 101 was determined by applying the Oxford cup double-plate method (Lü et al., 2014) with minor modifications. Briefly, the bacterial cells, after being cultured in MRS liquid medium at 37 °C, were centrifuged at 8000×g for 10 min at 4 °C to discard the sediment. The cell-free supernatant was filtered through a 0.22-μm sterile membrane for preliminary antimicrobial testing. The cell-free filtrate was adjusted to a pH at around 6.5 to exclude the interference of organic acids. Gram-positive S. aureus and Gram-negative P. aeruginosa were used as indicator strains. The Oxford cup used in the experiment has a diameter of 8 mm and a volume of 200 μL. First, 100-μL resuspension solution of the indicator strains at a concentration of 10⁷ CFU/mL were spread evenly on LB soft agar; then, the Oxford cup was placed on the surface of the soft agar LB, and 200-μL cell-free supernatant from the strain CLK 101 were added. Control Oxford cups were also set for 24-h incubation at 37 °C. Finally, the size of the inhibition zone around the Oxford cup was measured using a vernier caliper. The experiments were repeated three times.

2.3. Molecular identification and growth curve analysis of the Lacticaseibacillus rhamnosus strain CLK 101

The morphology of the target strain CLK 101 was further confirmed as described in section 2.1, and its taxonomic position was verified based on 16S rRNA gene marker (universal primers: 27F and 1492R). Briefly, the bacterial DNA was extracted according to the instructions of the DNA Extraction Kit (Tiangen, Beijing, China), and PCR amplification was performed for the 16S rDNA sequence of target Lacticaseibacillus strain CLK 101. The PCR amplification procedure was as follows: pre-denaturation at 94 °C for 5 min, followed by “denaturation at 94 °C for 30s, annealing at 52 °C for 30 s and extension at 72 °C for 50 s”. These three steps were performed for a total of 35 cycles, following extension at 72 °C for 10 min. The PCR products were detected by 1.5% agarose gel electrophoresis at 80 v, and then sent to Sangong Biotech Co., Ltd. (Shanghai, China) for Sanger sequencing. The obtained sequences were submitted to NCBI GenBank database (accession number: OM370898) and compared based on the NCBI non-redundant nucleotide collection (nt) database and 16S ribosomal RNA database through the BLAST tool. The phylogenetic tree was constructed by applying the latest Molecular Evolutionary Genetics Analysis (MEGA) version 11.0.10 (Tamura et al., 2021).

For growth curve analysis, L. rhamnosus strain CLK 101 was inoculated with 1% inoculum into sterilized 500 mL MRS liquid medium. After incubation at 37 °C for 1 h, 2 h, 4 h, 6 h, 8 h, 10 h, 12 h, 14 h, 16 h, 18 h and 20 h, respectively, samples were taken and the OD₅₉₅ values were measured as end product readings at each sample point on a microplate reader (Wisdom Applied Science, mode 6500, Newark, DE, USA). Subsequently, the growth curve was drawn.

2.4. Large-scale fermentation of L. rhamnosus strain CLK 101 in 20-liter fermenters based on fed batch or ATF perfusion bioprocesses

Firstly, the pH electrode and dissolved oxygen electrode of BLB10-20SJ automatic control fermenter (Shanghai BaiLun biological technology Co., Ltd, China) were cleaned and calibrated. Subsequently, the fermentation experiments were carried out based on fed batch or ATF perfusion technology separately. For fed batch process, the seed bacterial cells were inoculated into the fermenter containing 8 L MRS liquid medium at an inoculation percentage of 5%. By continuously adding 1 mol/L NaOH, the fermentation pH was maintained at 4.5–5.5. The aerobic fermentation process was carried out at 15% of dissolved oxygen, 37 °C, and a stirring rate of 200 rpm. When the logarithmic growth period (at 8 h) was reached, 2 × MRS liquid medium was supplemented intermittently to maintain the glucose concentration at 6–8 g/L, and the process was automatically controlled. The number of viable cells and OD₅₉₅ values were determined at every 2 h. For ATF perfusion production, the seed cells were inoculated into the fermenter containing 15 L MRS liquid medium at an inoculation percentage of 5%. The cultivation pH was maintained at 4.5–5.5 by continuously pumping 1 mol/L NaOH, and the aerobic fermentation was performed at 15% of dissolved oxygen, 37 °C, and a stirring rate of 200 rpm. When the logarithmic growth period (at 8 h) was reached, the hollow fiber column closure system (model: WA94510INV41L, SARTORIUS, Germany) was initiated. The 2 × MRS medium was supplemented automatically, and the perfusion rate was set as 20 mL/min. The recycling flow rate was 10 times the perfusion rate. In parallel to the fed batch approach, the number of viable cells and OD₅₉₅ values were also measured at every 2 h.

2.5. Purification and determination of bacteriocin CLK_01

After fermentation, L. rhamnosus strain CLK 101 was centrifuged under 8000×g for 10 min at 4 °C to remove bacterial cells, and the cell-free supernatant was filtered using a 0.22-μm sterile membrane. An ÄKTA purifier platform with automatic chromatograph in tandem coupled with a Superdex™ 30 Increase 10/300 GL Exclusion Chromatography column (GE Healthcare, Marlborough, MA, USA) was used to purify the bacteriocins in the cell-free supernatant obtained, as previously published by our laboratory (Jiang et al., 2021; Xiang et al., 2021b). Briefly, the following operational parameters were used: chromatographic column equilibrium volume of 2 column volumes (CV), elution at pH 6.2, elution volume of 1.5 CV, detection of the corresponding products in elution peaks by UV 280 nm, flow rate of 0.3 mL/min, and collection of the products from each elution peak in separate centrifuge tubes of 0.5 mL. Subsequently, the antibacterial activity of the collected products was determined by the Oxford cup double-plate method as described in section 2.2. After the initial round, the collection showing the highest antibacterial activity was loaded for a secondary purification by ÄKTA purifier and Superdex™ 30 Chromatography column to obtain the purified bioactive substance (named CLK_01) whose antibacterial activity was verified once again. Finally, the bacteriocin CLK_01 obtained was freeze-dried by an FD-2A freeze dryer (Biocoll, Beijing, China) and stored in a refrigerator at 4 °C.

2.6. LC-MS/MS analysis of bacteriocin CLK_01

Molecular mass and amino acid sequence analysis of the purified CLK_01 were performed on a Nano liquid chromatograph coupled with tandem mass spectrometry (LC-MS/MS) system, as described previously (Jiang et al., 2021; Xiang et al., 2021b). Briefly, the freeze-dried powder of CLK_01 was first dissolved in ddH2O and treated with 10 mM DL-dithiothreitol (DTT, Sigma-Aldrich, MO, USA) at 56 °C for 1 h. Then, the powder was alkylated with 50 mM iodoacetamide (IAA, Sigma-Aldrich, MO, USA) for 40 min at room temperature in the dark. Prior to LC-MS/MS analysis, CLK_01 was lyophilized and resuspended in 20 μL of 0.1% formic acid (Sigma-Aldrich, MO, USA). Peptide analysis was performed in an Ultimate 3000 system (Thermo Scientific, MA, USA) coupled to a Q Exactive™ Hybrid Quadrupole-Orbitrap™ mass spectrometer (Thermo Scientific, MA, USA). A 150 μm × 15 mm, made-in-house reverse-phase nanocolumn was used, which was packaged with ReproSil-Pur C18-AQ 1.9 μm resin (100 Å; Dr. Maisch GmbH, Germany). The LC linear gradient elution was performed at a flow rate of 0.6 μL/min. The primary mass spectrometric analysis was carried out with a single full scan (MS). The secondary mass spectrum data were obtained by employing 27% step-normalized collision energy. Mass spectrometry was achieved with Xcalibur 2.1.2 software in Orbitrap (using a spray voltage of 2.2 kV and a capillary temperature of 270 °C). After mass spectra signals were presented by MS scanning, Peaks Studio X (Bioinformatics Solutions Inc., Waterloo, ONT, Canada) was used to identify the sequences of peptides. Using the BLASTP tool, the bacteriocin CLK_01 sequence was searched against the NCBI non-redundant protein database, the UniProt database, and the APD3 antimicrobial peptide database, respectively. Finally, the antibacterial activity of bacteriocin CLK_01 was further verified using the Oxford cup double-plate method.

2.7. Measurement of bacteriocin concentration and MIC values

The concentration of bacteriocin CLK_01 was determined with the Pierce® BCA assay kit (Thermo Scientific, Waltham, MA, USA). In brief, the bovine serum albumin (BSA) standard solution was diluted according to the assay kit. The BSA working reagent was produced by mixing reagent A and reagent B in the assay kit at a ratio of 50:1. Subsequently, 25 μL of each standard and sample to be tested were then added to the microplate and mixed thoroughly for 30 s with 200 μL of working reagent, and then incubated at 37 °C for 30 min. The absorbance was measured at 562 nm using a microplate reader (MR-96A, Mindray, Shenzhen, China) at room temperature. The experiment was repeated three times.

MIC denotes minimal inhibitory concentration and its determination was performed as previously described (Jiang et al., 2021). The indicator strain, S. aureus ATCC 6538, was pre-cultured in LB liquid medium for 12 h, and the bacteriocin CLK_01 (1 mg/mL) was serially diluted in sterile PBS (pH 7.2). 10 μL of bacteriocin dilution were mixed with 90 μL of culture solution containing S. aureus (final concentration of bacterial cells: 10⁷ CFU/mL), added to a 96-well plate, and co-cultured at 37 °C for 24 h. The absorbance was measured at 595 nm to show the growth of the indicator strain. Three replicate wells were set for each sample and the experiments were repeated three times independently.

2.8. Examination of enzyme sensitivity, thermal stability, pH tolerance and antibacterial spectrum of bacteriocin CLK_01

The activity of CLK_01 was assayed after treatment under different enzymes, temperatures, and pH values, as previously described (Jiang et al., 2021). Briefly, different proteases (papain, trypsin, proteinase K, pepsin) or non-proteases (lipase, hydrogen peroxidase, α-amylase) were added to the bacteriocin CLK_01 solution at a concentration of 1 mg/mL. The mixtures were incubated in a water bath at 37 °C for 2 h and then treated at 80 °C for 10 min to inactivate the enzymes. The Oxford cup double-plate method was applied to test the effects of different proteases on inhibition of the indicator strain (S. aureus strain ATCC 6538) by bacteriocin CLK_01, and CLK_01 without being treated with enzymes was used as a control. For the thermal stability analysis, bacteriocin CLK_01 (1 mg/mL) was treated at room temperature (25 °C), 40 °C, 60 °C, 80 °C, 100 °C and 121 °C for 30 min, respectively, and subsequently the antibacterial effect of the samples was measured. As to the stability of CLK_01 under varying acid and alkali conditions, the pH of CLK_01 solution was adjusted to 2.0, 4.0, 6.0, 8.0, and 10.0 with 1 mol/L HCL solution and NaOH solution, respectively, and incubated at 37 °C for 2 h. After incubation, the antibacterial effect of the samples was evaluated. For the antibacterial spectrum determination of CLK_01, its activity against various pathogenic strains (Table 1) was measured using the Oxford cup double-plate method. The diameter of the inhibition circle was measured using vernier caliper, and three replicates of each sample were set; the experiment was repeated three times independently.

2.9. Bacterial cell proliferation and viability assays

The effect of bacteriocin CLK_01 on Gram-positive S. aureus and Gram-negative P. aeruginosa cells was determined by the methods reported by Xu et al. (2016). According to the instructions of XTT assay kit (Abcam, Cambridge, UK), 190 μL of bacterial cells (10⁷ CFU/mL) were added to a 96-well plate, associated with 10 μL of the bacteriocin CLK_01 solution (final concentration: 10 μg/mL). After incubation at 37 °C for 1 h, 10 μL of XTT mixture were added to each well and incubated at 37 °C for 2 h. The absorbance values were then measured at 450 nm. The CLK_01 treated cells were also examined with the Cell-Check™ viability/cytotoxicity kit (ABP Biosciences, Rockville, USA). Firstly, the bacterial cells were collected by centrifugation at 1000×g for 5 min and washed three times. After washing, 100 μL of 0.85% NaCl solution were added to the washed bacteria cells and 1 μL of NucView Green or Propidium Iodide was added according to the kit instruction, and the cells were stained for 15 min at room temperature in the dark and placed under an inverted fluorescence microscope (Boschida, Shenzhen, China) for observation (Ex/Em: NucView Green: 500/530; Propidium Iodide: 535/617 nm). Samples following the same procedure without CLK_01 treatment were used as controls.

2.10. Determination of antibiofilm activity

To test the antibiofilm activity of CLK_01 against Gram-positive S. aureus and Gram-negative P. aeruginosa, 100 μL of bacterial cell suspension were added to a 6-well micro-titration plate with 2000 μL of TSB medium (for biofilm formation), and then CLK_01 was added at different concentrations of 10 μg/mL, 30 μg/mL and 50 μg/mL respectively. After incubation at 37 °C for 24 h, the supernatant was discarded and the micro-titration plate was washed with PBS to remove the planktonic bacteria cells. The biofilm-coated bacterial cells were fixed with 200 μL of methanol for 15 min, and then the fixative was discarded. After PBS washing, a MycoLight™ Live Bacteria Fluorescence Imaging Kit (AAT Bioquest, CA, USA) was used to stain the biofilm (20 min at room temperature in the dark) and washed three times with PBS to remove excess dye. Finally, the microwell plate was placed in a fluorescent microscope (Boschida, Shenzhen, China) for observation (Ex/Em: 488/530). CLK_01-free samples were set as control groups and the experiments were repeated three times.

2.11. Scanning electron microscopy (SEM) analysis

For SEM examination, 2 mL of bacterial cells of Gram-positive S. aureus or Gram-negative P. aeruginosa (10⁷ CFU/mL) were used. After washing three times with PBS buffer, the bacterial cells were collected by centrifugation. In the experimental group, 1 mL of bacteriocin CLK_01 at 0.5 mg/mL was added, and in the control group, an equal amount of PBS was added. After incubation at 37 °C for 1 h, the supernatant was removed, and the bacterial cells were collected. After fixing with 2.5% glutaraldehyde at 4 °C for 6 h, the collected bacterial cells were washed three times in PBS. Then the bacteria cells were dehydrated with a gradient of 30%, 50%, 60%, 70%, 80% and 90% ethanol for 20 min each, followed by two dehydrations with 100% ethanol. The cells were coated onto 10 mm × 10 mm polished silicon wafers, dried naturally at room temperature for 24 h, treated with gold powder for 250 s and then subjected for SEM (FlexSEM1000, Hitachi, Tokyo, Japan) observation.

2.12. Statistical analysis

Using R software, statistical significances among quantitative data were determined by Student's t-test (between two groups) or One-way analysis of variance (ANOVA, among three or more groups). All results were presented as mean ± standard deviation (SD). Two-tailed p-value <0.05 indicates statistical significance.

3. Results

3.1. Isolation and identification of potential bacteriocin-producing probiotic Lacticaseibacillus rhamnosus from food-grade juice ferments

Previously, a potential probiotic strain, CLK 101, was identified from food-grade juice ferments, as belonging to the genus Lacticaseibacillus. The strain showed antibacterial activity against both pathogenic Gram-positive S. aureus and Gram-negative P. aeruginosa, thus it was selected for further investigation. CLK 101 was incubated on MRS agar plate at 37 °C for 48 h, and the colonies formed were round, creamy white with a smooth surface (Fig. S1A). The microscopic morphology of the strain was rod-shaped (at a length of 1–4 μm), with a positive Gram staining (Fig. S1B). After PCR amplification, a 16S rRNA gene sequence was obtained from the strain CLK 101, which was submitted to the NCBI GenBank database (accession number OM370898). The sequence similarity between CLK 101 and Lacticaseibacillus rhamnosus strain IGM3-10 (GenBank: MT197234.1) was found to be as high as 99.55% (Query Coverage: 99%). It was shown by further comparison with NCBI 16S ribosomal RNA database that the sequence similarity between CLK 101 and the type strain NBRC 3425 of Lacticaseibacillus rhamnosus (NCBI Reference Sequence: NR_113332.1) was 99.28% (Query Coverage: 99%). The phylogenetic tree further demonstrated a close cluster of strain CLK 101 with L. rhamnosus (Fig. 1). Based on these morphological and molecular studies, CLK 101 was identified as L. rhamnosus.

Fig. 1.

Phylogenic tree analysis of L. rhamnosus CLK 101 based on 16 rDNA sequences. The analysis was performed with MEGA11 program, and the numbers at the nodes denote bootstrap probabilities that were calculated using 1000 replicates according to the statistical method of neighbor-joining.

3.2. Comparison between fed-batch and ATF perfusion approaches for large-scale preparation of L. rhamnosus CLK 101

Further, we analyzed the growth curve of L. rhamnosus strain CLK 101 in a 500 mL shake flask (Fig. S2). We found that after about 6–8 h, the strain CLK 101 could enter the logarithmic growth phase, with a considerable increase in bacterial cells. Based on this, we carried out a fed-batch-based large-scale production of L. rhamnosus CLK 101 in a 20 L fermenter and compared it with the ATF perfusion-based technology. Automatic intermittent fed-batch cultivation was initiated at 8 h of the fermentation, when the pH was essentially constant at around 4.5 for both fed-batch and ATF perfusion-based cultures (Fig. S3). We measured changes in OD₅₉₅ and viable cell counts during the production and found that both indices were greatly improved in the ATF perfusion-based process than in the fed-batch (Fig. 2A and B). At the end of fermentation (20th h), the ATF perfusion-based technology resulted in a viable cells of 11.93 ± 0.14 log CFU/mL for L. rhamnosus CLK 101, which was significantly better than the fed-batch-based approach (p = 0.00056). Moreover, at the later stage of fermentation (16–20 h), due to rapid consumption of carbon and nitrogen sources, and the growth space had reached its limit, the bacteria gradually died under conventional fed-batch fermentation conditions, with the viable bacteria decreasing significantly from 11.01 ± 0.10 log CFU/mL at 16 h to 10.55 ± 0.19 log CFU/mL at 20 h (p = 0.022). In contrast, the ATF perfusion-based bioprocess maintained a remarkably stable number of viable cells (Fig. 2B) at the later stage of fermentation (16–20 h). These results demonstrate the great advantages of ATF perfusion-based technology and its potential application in the preparation of viable probiotic cells and microbial products in food industry.

Fig. 2.

Large-scale production of L. rhamnosus stain CLK 101 in 20 L fermenter based on fed-batch and ATF perfusion approaches. (A) Measurement of OD₅₉₅ readings. (B) Viable cell numbers were examined upon fed-batch or ATF perfusion-based system.

3.3. Bacteriocin purification, molecular mass, and amino acid sequence of CLK_01

On the basis of large-scale production of viable L. rhamnosus CLK 101 cells, we further purified and characterized a novel bacteriocin with a broad-spectrum antimicrobial activity derived from the strain. Among the five peaks (A1-A5) identified in the chromatogram, the initially purified peak A3 showed the greatest antibacterial activity (Fig. 3A). The product peak B1 in Fig. 3B was further obtained by a secondary purification of the peak A3. The corresponding antibacterial substance in peak B1, CLK_01, was collected and verified to have an inhibition zone size of 18.00 ± 0.40 mm against the important drinking water pathogen P. aeruginosa ATCC 27853. The molecular mass of CLK_01 was 701.49 Da (M + H mode, m/z = 351.25, z = 2, Fig. S4) and its amino acid composition was I–K–K–V-T-I (Fig. 4), as identified by LC-MS/MS. The NCBI BLAST search presented no similar matches with CLK_01, thus confirming the purified bacteriocin as novel.

Fig. 3.

Purification of antibacterial substances derived from L. rhamnosus stain CLK 101. Base peak chromatograms of (A) initial purification and (B) further secondary purification of the bacteriocin extracts, accompanied by inhibition zone analysis of the corresponding peak samples, by using the Oxford cup double-plate method.

Fig. 4.

Tandem secondary mass spectrometry analysis of CLK_01 for deciphering its amino acid composition.

3.4. Minimal inhibitory concentration and antibacterial spectrum of CLK_01

The minimal inhibitory concentration (MIC) for CLK_01 against S. aureus strain ATCC 6538 was 10 μg/mL. The antibacterial activity of CLK_01 against different Gram-positive and Gram-negative foodborne pathogens is further shown in Fig. 5. The inhibition zone diameter of CLK_01 against Gram-positive foodborne bacteria such as S. aureus, S. argenteus, M. luteus, and S. agalactiae reached more than 15 mm, about twice the size of control (all statistical p-values were significant less than 0.001); among the Gram-negative bacteria, V. Parahemolyticus, E. coli, S. dysenteriae, Salmonella ser. Paratyphi B, P. aeruginosa, K. pneumonia, and A. hydrophila all had an inhibition zone diameter of approximately 17 mm or larger. Among these pathogenic bacteria, CLK_01 was most effective against M. luteus, with an inhibition zone diameter of 26.73 ± 0.35 mm. These results indicate that CLK_01 has a broad-spectrum antibacterial activity, fighting against both Gram-positive and Gram-negative pathogens, especially foodborne pathogenic bacteria.

Fig. 5.

Antibacterial spectrum of CLK_01. V. Parahemolyticus, E. coli, S. dysenteriae, Salmonella ser. Choleraesuis, Salmonella ser. Enteritidis, Salmonella ser. Paratyphi B, P. aeruginosa, K. pneumonia, and A. hydrophila are Gram-negative bacteria, while S. aureus, S. argenteus, M. luteus, and S. agalactiae are Gram-positive bacteria.

3.5. Enzyme sensitivity, thermal and pH stability of CLK_01

As illustrated in Fig. 6, CLK_01 treated by hydrogen peroxidase retained 94.81 ± 2.96% of its activity (no significance compared to the untreated control, p > 0.05), which excluded the catalase-based antibacterial activity of CLK_01. However, the antimicrobial activity of CLK_01 was significantly reduced after treatment with different proteases (p < 0.001), which decreased by more than 60% after treatment with Trypsin, Proteinase K and Pepsin. In particular, the activity was almost completely lost after treatment with Proteinase K and Pepsin (Fig. 6). Compared to the untreated control, the p-values for papain, trypsin, proteinase K, and pepsin groups were 1.93E-04, 1.79E-05, 1.07E-07, and 1.07E-07, respectively, indicating the significant sensitivity of CLK_01 to proteases. In contrast, CLK_01 activity was not affected by the treatment of non-proteases such as lipase and α-amylase.

Fig. 6.

Analysis of enzyme sensitivity of CLK_01. ***, p < 0.001; NS, not significant.

As shown in Fig. 7A, CLK_01 has a high thermal stability. Even when heated to 121 °C (sterilization temperature), CLK_01 still retained 94.69 ± 0.76% of its antibacterial activity. In addition, we found that the activity of CLK_01 tended to decrease with increasing pH values, reducing to 66.78 ± 2.23% at pH 10 (p < 0.001 compared to the initial pH 4). Overall, CLK_01 maintained its antimicrobial activity of more than 65% in the wide pH range of 2–10 (Fig. 7B).

Fig. 7.

Thermal stability (A) and pH stability (B) of CLK_01. All values were normalized to the maximal activity among the dataset (= 100%) and illustrated as mean ± SD of n = 3 independent experiments. RT, room temperature (25 °C).

3.6. Proliferation and cell viability of Gram-positive S. aureus and Gram-negative P. aeruginosa cells upon CLK_01 treatment

Changes in cell proliferation and metabolic viability of Gram-positive S. aureus and Gram-negative P. aeruginosa upon CLK_01 exposure were detected by XTT reduction method, and it was found that the cell viabilities of S. aureus and P. aeruginosa were significantly reduced to 35.61 ± 2.96% (p = 6.25E-06) and 52.09 ± 1.85% (p = 6.94E-06) (Fig. 8A and Fig. S5A) respectively after 1 x MIC CLK_01 treatment, compared with the control group. By cell fluorescence staining, we found that S. aureus and P. aeruginosa viable cells without the bacteriocin treatment showed a strong green fluorescence (Fig. 8B and Fig. S5B); while S. aureus and P. aeruginosa cells treated with 1 x MIC CLK_01 showed a considerable decrease in the green fluorescence, and dead cells stained with red fluorescence appeared (Fig. 8C and Fig. S5C), indicating that cell proliferation and cell viability were significantly affected after CLK_01 treatment. These results confirmed the effective inhibitory effects of CLK_01 on Gram-positive S. aureus and Gram-negative P. aeruginosa cell viability.

Fig. 8.

Analysis of cell viability of Gram-positive S. aureus upon exposure to CLK_01. (A) Viability of S. aureus strain ATCC 6538 after treatment by CLK_01 in comparison to the untreated control; ***, p < 0.001. Fluorescence microscopic images of control (B) and cells treated by CLK_01 (C).

3.7. Antibiofilm activity of CLK_01 and SEM observation

Treatment of Gram-positive S. aureus and Gram-negative P. aeruginosa with CLK_01 at concentrations of 10 μg/mL, 30 μg/mL, and 50 μg/mL considerably reduced the density of biofilm formed by them. The illustrations of the control group showed that the biofilm density of S. aureus and P. aeruginosa was very high (Fig. 9A and Fig. S6A). The biofilm formation of Gram-positive S. aureus and Gram-negative P. aeruginosa was clearly inhibited after treatment with 10 μg/mL CLK_01 (Fig. 9B and Fig. S6B). The bacteriocin was found to be more effective against biofilms at a concentration of 30 μg/mL (Fig. 9C and Fig. S6C); When CLK_01 concentration was 50 μg/mL (Fig. 9D and Fig. S6D), the biofilm formation of both Gram-positive S. aureus and Gram-negative P. aeruginosa could be essentially eliminated.

Fig. 9.

Effect of different concentrations of CLK_01 (A: 0 μg/mL; B: 10 μg/mL; C: 30 μg/mL; D: 50 μg/mL) on the biofilm formation of Gram-positive S. aureus ATCC 6538, as detected by the fluorescence microscopy.

Next, the scanning electron microscopy (SEM) analysis further suggested that the CLK_01-free S. aureus and P. aeruginosa had regular cell morphology, complete cell structure with clear outline and neat surface (Fig. 10A and C, Figs. S7A and C). However, compared to the control cells, Gram-positive S. aureus and Gram-negative P. aeruginosa cells treated by CLK_01 exhibited wrinkled surface, damaged cells, blurred outline of the cell body, perforation of the cell surface and leakage of cellular contents (Fig. 10B and D, Figs. S7B and D). These results suggest the disruptive effect of CLK_01 on the cellular integrity of bacterial pathogens.

Fig. 10.

Antibacterial effect of CLK_01 against S. aureus ATCC 6538 cells, as analyzed by scanning electron microscopy (SEM). (A) and (C) demonstrate the control, while (B) and (D) present the CLK_01 treatment.

4. Discussion

Numerous foodborne pathogenic bacteria, such as S. aureus, V. parahemolyticus, E. coli, and different species of Salmonella, cause food contamination, reduce food quality, and pose a significant risk to human health (Tsuzuki et al., 2020; Xiang et al., 2021a). In recent years, bacteriocins derived from food-grade probiotics, especially those isolated from fruits, juices, fermented foods, etc., inhibiting pathogenic microorganisms while improving the flavor of food, have drawn much attention, and are particularly valuable for their edibility and safety (Kumari et al., 2018; Xiang et al., 2021a). However, large-scale preparation of viable probiotics and bioactive bacteriocins is still a challenging task.

To the best of our knowledge, for the first time, this study is concerned with establishing an ATF perfusion-based bioprocess, for large-scale recovery of the food probiotic L. rhamnosus. Comparing to the conventional fed-batch approach, we found that the preparation of viable L. rhamnosus was greatly improved by using ATF perfusion technology. After 20 h of fermentation, L. rhamnosus was obtained with a viable cell counts of 11.93 ± 0.14 log CFU/mL, which is technically superior to most studies in current literature (Subramaniam et al., 2019; Wang et al., 2020, de Assis et al., 2022), thus showing a promising potential for applications of probiotic and bacteriocin production in food industry.

Based on large-scale recovery of viable L. rhamnosus CLK 101 cells, we further purified a novel bacteriocin, CLK_01. Compared to the molecular weights of other LAB bacteriocins reported in previous studies, e.g., 5592.22 Da from Lactobacillus pentosus 31-1 (Zhang et al., 2009), 1221.074 Da from L. salivarius SPW1 (Wayah and Philip, 2018b), 1049.56 Da from L. plantarum (Jiang et al., 2021), 10–15 kDa from L. rhamnosus ATCC 53103 (Zhou and Zhang, 2018), CLK_01 is a relatively smaller bacteriocin. In general, bacteriocins with small molecular mass are easier to synthesize at low cost and their spatial structure is more stable under environmental stresses (e.g. temperature, pH and humidity) (Wayah and Philip, 2018a), which means that CLK_01 may be suitable for a wide range of food processing conditions. Moreover, because they can play an antibacterial role not only by disrupting the integrity of bacterial cell membranes but also by interacting with the enzyme system or the cellular DNA, bacteriocins with a small size such as CLK_01 tend to have better bactericidal efficiency (Zhao et al., 2016; Pei et al., 2020). In addition, the sequence composition of CLK_01, which shows no similarity to other bacteriocins previously reported, provides useful information for further synthesis of natural food preservatives through genetic engineering. These advantages demonstrate the biopreservative potential of CLK_01 in food industry.

Besides the characteristics of CLK_01 discussed above, the thermal and pH resistance, protease sensitivity and antibacterial properties of a bacteriocin are also important parameters for its applications. In this study, we found that CLK_01 showed antibacterial activity against a variety of pathogens, including both Gram-positive and Gram-negative strains. Due to the presence of a protective outer membrane in Gram-negative bacteria, it is difficult for many antimicrobial agents to penetrate and enter the interior of the cells. Thus, current number of antimicrobial agents targeting Gram-negative bacteria remains much lower than those killing Gram-positive bacteria (Cleveland et al., 2001; Briers and Lavigne, 2015; Du et al., 2018). Similarly, many bacteriocins produced by lactic acid bacteria and used in food industry also have poor antimicrobial activity against Gram-negative bacteria (Cleveland et al., 2001; G. Liu et al., 2015; Du et al., 2018), hindering the development and utilization of bacteriocin products. By comparison, CLK_01 shows a broad antimicrobial spectrum against both Gram-positive and Gram-negative foodborne pathogens, suggesting its potential as a broad-spectrum biopreservative. Additionally, the sensitivity of CLK_01 to protease treatment suggests that it is a bacteriocin with protein-like properties, rather than carbohydrate and lipid properties; therefore, it could be more safely applied (Sung and Jo, 2020).

Recently, Ren et al. also presented that purification and characterization of bacteriocin from Genus Lacticaseibacillus rhamnosus (Ren et al., 2022). Very similarly, the bacteriocin produced by the strain A5 showed antibacterial activity against a wide range of food pathogens and high thermo-stability. However, its activity was completely lost above pH 5.0 and the authors did not determine the amino acid composition of the bacteriocin. In comparison, the antimicrobial activity of CLK_01 remained almost constant between pH 2 and pH 6, and only decreased by 21.4% and 33.2% at pH 8 and 10, respectively. According to previous studies, LAB bacteriocins (e.g., bacteriocins from L. salivarius SPW1, L. fermentum BZ532, and L. paracasei HD1-7, etc.) often showed a reduction in activity by more than 90% at pH above 8 (Wayah and Philip, 2018b; Rasheed et al., 2020). Pmka et al. also confirmed the inactivation of LAB bacteriocin from L. lactis under alkaline conditions (Pmka et al., 2019). Here, more than 65% activity of CLK_01 could be maintained in the broad pH range of 2–10, suggesting its high acid-base tolerance. For heat treatment, the antibacterial activity of bacteriocins from L. rhamnosus ZRX01 and L. plantarum B21 was reduced by at least 50% at 100 °C for 30 min (Golneshin et al., 2020; Zhao et al., 2020). Comparatively, the bacteriocins of L. acidophilus NX2-6 only had at most 20% residual activity after treatment at 121 °C for 1 h (Meng et al., 2020). In this study, we found that CLK_01 retained 95.9% and 94.7% of their antibacterial activity after treatment at 100 °C and 121 °C for 30 min, respectively, suggesting its remarkable superiority. Overall, these results indicate that CLK_01 has great acid-base and temperature tolerance over a wide range, and demonstrate its potential applications in industrial food processing.

Especially, in the XTT experiment, cell viability of Gram-positive S. aureus and Gram-negative P. aeruginosa was significantly reduced upon CLK_01 treatment (p < 0.001). Similarly, cell viability of S. aureus could be reduced by 47% upon treatment by the bacteriocin purified from L. salivarius (Li et al., 2021). Further, in this study, different MICs of CLK_01 were used to inhibit the formation of biofilms of Gram-positive S. aureus ATCC 6538 and Gram-negative P. aeruginosa strain ATCC 27853 cells. Our observations presented that the density of S. aureus and P. aeruginosa bacterial biofilms gradually decreased as the concentration of CLK_01 gradually increased, and at 50 μg/mL, biofilm formation of both Gram-positive S. aureus ATCC 6538 and Gram-negative P. aeruginosa ATCC 27853 were eliminated. These results suggest that CLK_01 could be developed as an effective inhibitor for controlling pathogenic bacteria and their biofilms from foods or food processing equipment surfaces.

To further explore the antimicrobial mechanism, SEM observations were carried out and we found that cell membrane permeability of pathogenic cells was affected by CLK_01. Cell surface perforations and cytoplasmic content leakage were also detected. These results are consistent with those of previous studies (Li et al., 2021; Zhu et al., 2021), revealing that the pathogenic bacteria showed wrinkled, damaged cells after exposure to the bacteriocins, leading to loss of viability and death of bacterial pathogens. Together, we hypothesize that these antimicrobial effects exhibited by CLK_01 are closely related to its intrinsic properties and composition of its amino acid sequence. As we know, the hydropathy index of an amino acid describes the degree of hydrophilicity or hydrophobicity of its branched chains, which was first introduced by Jack Kyte and Russell Doolittle in 1982 and is widely applied (Kyte and Doolittle, 1982). The larger the index, the more hydrophobic for the corresponding amino acid. It is noteworthy that the bacteriocin CLK_01 IKKVTI sequence contains two top hydrophobic Ile (one Ile at the N-terminal and one at the C-terminal, ranking 1st out of 20 common amino acids by a hydropathy index of 4.5), and a Val (ranking 2nd out of 20 common amino acids by a hydropathy index of 4.2), which ensures CLK_01 structural stability and may partly explain its high activity at wide range of temperature (25–121 °C) and pH (2–10). Besides, the hydrophobicity of the overall protein is usually predicted from the GRAVY value (Kyte and Doolittle, 1982). A positive GRAVY value usually indicates a hydrophobic protein. The SMS2 tool (Stothard, 2000) predicted that the total GRAVY value for the CLK_01 sequence was 0.78, in particular, the VTI peptide at its C-terminal has a very high GRAVY value of 2.67. These results revealed strong hydrophobic property of the bacteriocin CLK_01.

Further analysis by SMS2 tool showed that the protein isoelectric point (pI) of bacteriocin CLK_01 was as high as 10.81, indicating CLK_01 as a special cationic LAB bacteriocin. As we know, bacterial surfaces are usually negatively charged due to the high lipid content and the presence of secondary cell wall polymers (SCWPs) in the lipid bilayer and peptidoglycan layer (Low et al., 2011). A strong correlation between the positive net charge of an antimicrobial agent and its bactericidal spectrum has been highlighted by many studies (Donovan et al., 2006; Low et al., 2011). Moreover, this positive charge plays a key role in penetrating the bacterial outer membrane and the peptidoglycan layer. Given the hydrophobic structural feature and high pI of CLK_01, we hypothesized that there is a charge-driven interaction between CLK_01 and the bacterial surface, which greatly facilitates the penetration and localization of CLK_01 on the bacterial surface. This localization may be achieved by electrostatic steering and hydrophobic membrane channels (Kozack et al., 1995), thereby enabling CLK_01 to break the barrier of the cell wall, thus explaining the antimicrobial findings of CLK_01 observed in this study.

However, the current study still has some limitations. Firstly, the advanced structure including chemical bond composition of CLK_01 was not revealed. Secondly, the detailed mechanism underlying the interaction between CLK_01 and bacterial cell surface certainly merits further investigation.

5. Conclusion

In conclusion, we performed the large-scale preparation of the bacteriocin-producing L. rhamnosus CLK 101 based on the ATF perfusion technology for the first time, and further purified a novel bacteriocin, CLK_01. The molecular mass, amino acid composition, and antibacterial and antibiofilm activities of CLK_01 were analyzed. CLK_01 exhibited thermal stability, acid and alkaline resistance, and broad antibacterial spectrum against foodborne bacteria. Furthermore, it could reduce the cell viability of both Gram-positive and Gram-negative pathogens, and inhibit their biofilm formation. The effect of CLK_01 on killing foodborne pathogens was possibly attributed to its intrinsic cationic and hydrophobic properties, causing perforation of the cell membrane, cell structure destruction, and leakage of the intracellular components. These results suggest that CLK_01 has great potential as a natural biopreservative for the control of multiple pathogens and their biofilms in food industry.

CRediT authorship contribution statement

Shi-Yu Chen: Investigation, Methodology, Data curation, Software, Writing – original draft. Rui-Si Yang: Investigation, Methodology, Data curation. Bai-Quan Ci: Methodology, Data curation. Wei-Gang Xin: Methodology, Investigation. Qi-Lin Zhang: Methodology. Lian-Bing Lin: Conceptualization, Methodology, Resources, Investigation. Feng Wang: Conceptualization, Methodology, Investigation, Resources, Data curation, Writing – original draft, Writing – review & editing.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Handling Editor: Dr. Yeonhwa Park

Appendix A. Supplementary data

The following are the Supplementary data to this article:

Fig. S1

Isolation and Identification of the Lacticaseibacillus rhamnosus CLK 101. (A) Colony morphology and microscopic characteristics of Gram staining (B) of the L. rhamnosus strain CLK 101.

Fig. S2

The growth curve of Lacticaseibacillus rhamnosus stain CLK 101 in 500 mL culture flask was analyzed.

Fig. S3

Large-scale fermentation of L. rhamnosus stain CLK 101 was carried out in 20 L fermenter based on fed-batch or ATF perfusion technology, and pH changes during the processes were recorded.

Fig. S4

LC-MS analysis of CLK_01.

Fig. S5

Analysis of cell viability of Gram-negative P. aeruginosa upon exposure to CLK_01. (A) Viability of P. aeruginosa cells after treatment with CLK_01 in comparison to the untreated control; ***, p < 0.001. Fluorescence microscopic images of control (B) and cells treated by CLK_01 (C).

Fig. S6

Effect of different concentrations of CLK_01 (A: 0 μg/mL; B: 10 μg/mL; C: 30 μg/mL; D: 50 μg/mL) on the biofilm formation of Gram-negative P. aeruginosa ATCC 27853, as detected by the fluorescence microscopy.

Fig. S7

Antibacterial effect of CLK_01 against P. aeruginosa ATCC 27853 cells, as analyzed by scanning electron microscopy (SEM). (A) and (C) demonstrate the control, while (B) and (D) present the CLK_01 treatment.

Data availability

Data will be made available on request.