Inhibition of Listeria monocytogenes by bacteriocin-producing Bacillus velezensis isolated from silage

The aim of this study was to investigate the production of anti-Listeria bacteriocins by naturally occurring bacteria on silage and to characterize the most promising bacteriocin. The production of bacteriocins was evaluated by the spot-on-lawn method. The presence of interfering factors and the sensitivity to proteinase K were analyzed. The spectrum of activity was determined and the most promising bacteriocin-producing isolate was identified, being selected for the subsequent experiments. The antimicrobial peptide was extracted, analyzed regarding temperature and pH sensitivities, and purified. Of the 37 isolates evaluated, seven showed anti-Listeria activity. The presence of bacteriophages and production of acidic compounds were not detected. The selected antimicrobial compound was sensitive to proteinase K and the producing-bacterium was identified as Bacillus velezensis. The crude extract maintained the antimicrobial activity in different temperatures and pH conditions. In conclusion, the bacteriocin produced by Bacillus velezensis showed strong activity against Listeria monocytogenes, and great stability to elevated temperature and adverse pH, desirable features for future biotechnological applications.


Introduction
Listeria monocytogenes is widely distributed in nature and can be found in soil, vegetation, water and intestinal tract of some animals (Reis et al., 2011). This pathogen is able to grow and adapt to different stress conditions, including a wide range of temperature (1 °C to 45 °C), acidic pH and high salt concentration (Reis et al., 2011).
The infectious disease caused by L. monocytogenes is called listeriosis, and eating contaminated food containing this bacterium is the main route of infection (Junior et al., 2015). Listeriosis is one of the most serious and severe foodborne diseases, affecting livestock animals, especially ruminants, but also swine, poultry and fish, as well as humans. Although the number of cases of listeriosis is not so expressive, the high rate of death associated with this infection, especially among elderly and immunocompromised patients, makes it a significant public health concern (Reis et al., 2011;Buchanan et al., 2017;Heir et al., 2018).
Regarding the animal health, the consequences of listeriosis range from a decrease in animal production and milk quality, to encephalitis, abortion and infertility (Junior et al. 2015). One of the main sources of L. monocytogenes for ruminants is the consumption of silage contaminated with this bacterium. Although lactic acid bacteria usually dominate silage microbiota, some pathogens, including Listeria, can contaminate and grow during the ensilage, a process that includes fermentation and storage of silage. In this way, the consumption of low quality silage has been associated to listeriose cases in ruminants (Durmaz et al., 2015).
As a diverse environment, with broad microbial diversity, the silage is a suitable environment for the isolation of bacteria that are able to produce antagonistic compounds (Castro et al., 2006), such as bacteriocins. Bacteriocins are ribosomally synthesized peptides, which show bactericidal or bacteriostatic activity against the target cells (Cotter et al., 2005;Drider et al., 2006;Todorov et al., 2010). Bacteriocin-producing strains have been proposed as an alternative for inhibiting spoilage microorganisms in silages, and could be used as effective inoculants to produce good quality and safe silages (Flythe and Russell, 2004;Amado et al., 2012).
The aim of this study was to investigate the production of anti-Listeria bacteriocins by naturally occurring bacteria on silage. The most promising bacteriocin was characterized and the producer cell was genetically identified.

Microorganisms and growth conditions
Thirty seven bacteria, isolated from silage, and belonging to the culture collection of the Rumen Microbiology Laboratory, Embrapa Gado de Leite, Brazil, were randomly selected and screened for specific antagonism, in order to assess their usefulness in suppressing growth of L. monocytogenes. The isolates were maintained at -80 °C in glycerol containing Brain Heart Infusion (BHI) medium (10% final concentration), until use. The selected bacteria were grown in BHI medium, under aerobic conditions, at 37 °C.

Screening of antagonistic activity against indicator bacteria
Antagonistic activity was evaluated by the spot-on-lawn method (Booth et al. 1977). The silage bacteria were spotted onto BHI agar and incubated in aerobiosis, at 37 °C, for 18 h. Molten semi-solid BHI (0.8% agar), previously inoculated with L. monocytogenes (10 5 CFU mL -1 ), was poured over the agar plates containing the bacterial colonies. Plates were incubated at a suitable temperature and atmosphere for each indicator organism. The antimicrobial activity was determined by the presence of inhibition zones of the indicator organism growth around the colonies of selected bacteria (>6 mm diameter).

Determination of interfering factors on the antagonistic activity
In order to verify if the antagonistic activity observed on the spot-on-lawn method was indeed due to the production of antimicrobial peptides and not to the activity of other factors, we evaluated the presence of interfering factors, such as bacteriophage presence and production of acidic compounds by the selected silage isolates.
To evaluate the presence of bacteriophage, an agar piece (3 mm) of the inhibition zone observed in the spot-on-lawn method was aseptically removed, macerated in saline (NaCl 0.9%, pH 7.0), and centrifuged (12000 g, 20 min). An aliquot of the supernatant (200 μL) was transferred to a culture medium (5 mL), previously inoculated with the indicator microorganism.
After incubation period (10 min, at room temperature), an aliquot of the culture medium containing the supernatant (350 μL) was added to a semisolid BHI (3.5 mL), which was poured over a prior layer of solid BHI. After incubation in aerobiosis, at 37 °C, the plates were evaluated regarding the presence of lytic zones, which would indicate the bacteriophage presence in the selected bacteria.
To exclude the possibility of the production of acidic compounds be responsible for the antagonism, after spot-onlawn experiment, the pH inside and outside of the growth inhibition zones were determined using pH measuring tapes.

Spectrum of action of the antagonistic substances produced by silage bacteria
The inhibitory activity of the antagonistic compounds was evaluated by the spot-on-lawn method (described above) using, as indicator microorganisms, different bacterial strains, including the culture collection strains Escherichia coli ATCC Research, Society and Development, v. 10, n. 9, e2610917783, 2021 (CC BY 4. The plates were incubated in appropriate temperature and atmosphere for each indicator microorganism, for up to 48 h. The antimicrobial activity was verified by the presence of inhibition zones larger than 6 mm in diameter.
Based on the antimicrobial activity against L. monocytogenes and on the activity spectrum, the most promising silage bacterium was selected for the next experiments.

Bacterial crude extracts with antimicrobial activity
Extracts of the selected silage bacteria were obtained as described by Mantovani et al. (2002), with modifications.
Briefly, the pH of the stationary phase bacterial cultures was adjusted to 6.5, the cultures were heat treated (70 °C, 30 min), and the cells were harvested by centrifugation (4500 rpm, 15 min, 5 °C). The cell pellets were washed with sodium phosphate solution (5 mM, pH 6.5), centrifuged (4500 rpm, 15 min), and resuspended in acidic calcium chloride solution (100 mM, pH 2.0). After this stage, the suspension was incubated at room temperature, under stirring, for 2 h. At the end of the extraction step, the cell suspensions were centrifuged (12500 rpm, 20 min, 5 °C), and the cell free supernatant was lyophilized. The lyophilized material was resuspended in sodium phosphate solution (5 mM, pH 6.5), and named "crude extract".
The antimicrobial activity of the crude extracts was assessed by agar well-diffusion assay (Tagg et al. 1976). Molten semi-solid BHI (0.8% agar), previously inoculated with L. monocytogenes (10 7 CFU mL -1 ), was poured onto sterile Petri dishes, and after solidification, 5 mm wells were drilled in the culture medium. The crude extracts were serially diluted in sodium phosphate solution (5 mM, pH 6.5), and applied (20 µL) inside the wells. The plates were kept overnight at 4 °C (for diffusion of the antimicrobial compounds from the wells), and incubated in aerobiosis, at 37 °C, for 24 h. The antimicrobial activity was expressed as arbitrary units (AU), defined as the reciprocal of the highest dilution that showed an inhibition zone of at least 9 mm in diameter (Lewus and Montville, 1991).

Proteinase K sensitivity of the crude extracts with antimicrobial activity
In order to evaluate the peptide nature of the antimicrobial compounds produced by the selected silage bacteria, a proteinase K sensitivity assay was performed. BHI-containing plates, previously inoculated with L. monocytogenes, were used in an agar well-diffusion assay. Proteinase K (20 mg mL -1 ) was added (20 µL) next to each well in the culture medium; after 30 min, the crude extracts (20 µL) were applied inside the wells. The plates were kept overnight under refrigeration, and then aerobically incubated, at 37 °C, for 24 h. Non-treated crude extracts were used as a positive control for antimicrobial activity.
The proteinase K sensitivity was observed by alterations of the original inhibition halos, i.e., evidence of growth of the target organism where the proteinase K was added.

Influence of temperature and pH on the biological activity of the crude extracts with antimicrobial activity
The crude extracts with antimicrobial activity obtained from the silage bacterium were submitted to different treatments to evaluate their stability under variations of temperature and pH.
In order to evaluate the thermal stability, aliquots of the crude extracts diluted in sodium phosphate solution (5 mM, pH 6.5) were heated to 45 °C, 60 °C, 80 °C and 100 °C, for 30 min. The antimicrobial activity of the crude extracts after heat treatment was evaluated by agar well-diffusion assay, using L. monocytogenes as indicator bacteria. Afterwards, the residual antimicrobial activity after heat treatment was compared with control samples of the crude extracts, which were kept refrigerated. The residual antimicrobial activity was calculated according to the following equation (Cladera- Oliveira et al. 2004):

RA (%) = (HT/CT) x 100
Where RA is the residual antimicrobial activity, HT is the halo diameter (mm) of L. monocytogenes growth inhibition obtained after heat treatment of the crude extracts (measured on the agar well-diffusion assay), and CT is the halo diameter (mm) of the control samples (non-treated crude extracts).
The Britton-Robinson buffer, composed by ortho phosphoric acid (0.04 M), acetic acid (0.04 M) and boric acid (0.04 M) was used to evaluate the stability of the crude extracts at different pH values (2.0, 4.0, 8.0, 10.0 and 12.0). The pH adjustment was performed with NaOH solution (2M) and HCl (2M), and the buffers were sterilized by filtration. The crude extracts with antimicrobial activity were diluted (1:2) in the buffer with the desired pH, and the mixture was kept at room temperature, for 2 h. The antimicrobial activity of the crude extracts was then evaluated on agar well-diffusion assay. Residual antimicrobial activity was determined as described above.
The amplicons were purified using Invisorb® Fragment CleanUp kit (Stratec Molecular), following the manufacturer instructions. The samples were quantified with NanoDrop and the sequencing was performed by the company Myleus Facility, using the chain termination method (Sanger Method), and the ABI3730 automated sequencer (Thermo Fisher Scientific). All consensus sequences obtained were compared to those available in GenBank database (National Center for Biotechnology Information -NCBI), and the alignment of the sequences was performed using the Basic Local Alignment Search Tool algorithm for nucleotide (BLASTn). The sequences obtained were deposited in GenBank database.

Results
From the 37 bacteria previously isolated from silage and selected to the evaluation of antagonist activity, seven isolates, defined as 111, 117, 120, 129, 135, 139, and 150, were capable to inhibit the growth of L. monocytogenes. Moreover, the presence of bacteriophages was not detected in the bacterial isolates, since no lysis plates were observed after culturing.
The pH inside and outside of the growth inhibition halo was 7.0, confirming that the antagonist activity was not due the production of acidic compounds by the isolates.
Since the interference of bacteriophage presence and production of acidic metabolites on the inhibitory activity against L. monocytogenes were excluded, we evaluated the spectrum of activity of the antagonist compounds produced by the silage isolates. The majority of the indicator bacteria evaluated, including the Gram-negative strains, were inhibited in the spoton-lawn test, demonstrating the broad spectrum of action of the antagonist compounds produced by the silage bacteria (Table   1).  On the other hand, E. coli ATCC 25922 was inhibited only by the antagonistic compound produced by the isolate 150. Research, Society andDevelopment, v. 10, n. 9, e2610917783, 2021 (CC BY 4.0) | ISSN 2525-3409 | DOI: http://dx.doi.org/10.33448/rsd-v10i9.17783 7 Based on the anti-Listeria activity and on the broader spectrum of action, the isolate 150 was selected for the subsequent experiments. The crude extract obtained from the isolate 150 showed strong anti-Listeria activity, and a zone of Listeria growth inhibition measuring 21 mm of diameter was observed in the agar well-diffusion assay (Figure 1). The crude extract obtained from the isolate 150 was sensitive to proteinase K, losing its antagonist activity against L. monocytogenes after the treatment with the proteolytic enzyme. This effect was observed after changing in the form of the inhibition zone in the agar well-diffusion assay (Figure 2).  Once the peptide character has been confirmed by the protease K sensitivity, the antibacterial activity of the crude extract was determined in arbitrary units (AU mL -1 ), being obtained an activity of 1600 AU mL -1 against L. monocytogenes (presence of Listeria-growth inhibition zones until the dilution 1:32) (Figure 3). Listeria monocytogenes as indicator bacterium. From the crude extract, a serial dilution was prepared with twice increment (1:2 until 1:1024). Central well: crude extract; well 1: crude extract diluted 1:2; well 10: dilution 1:1024.
The crude extract obtained from the isolate 150 showed good stability in different temperatures, maintaining its anti-Listeria activity after heating and freezing (Fig. 4A). Considering the crude extract stored at 4 °C as a control (100% of anti-Listeria activity), when the extract was submitted to 45 °C, the residual activity (RA) observed was 98%; a RA of 93% was detected at 60 °C, and at 80 °C the RA was 90%. A bigger reduction in the anti-Listeria activity was observed when the crude extract was submitted to 100 °C (RA = 64%).The crude extract was also stable at lower temperatures, showing a RA of 92% and 97% when stored at -20 °C and -80 °C, respectively ( Figure 4A).

Source: Authors.
Analyses of similarity among the 16S rRNA gene sequence showed that the silage isolate evaluated showed high identity to Bacillus velezensis strain FZB42 (NR075005.2). The 16S rRNA gene sequence of the silage isolate (isolate 150) was deposited in the GenBank sequence database under accession number MK658866.
The crude extract obtained from B. velezensis (isolate 150) was purified in FPLC, using C18 column ( Figure 5). The fractions were collected and the predominant peak, which appeared in the chromatogram after 7 minutes of purification, exhibited strong anti-Listeria activity, showing a halo of growth inhibition of 18 mm diameter in the agar well-diffusion assay (the crude extract before purification showed a zone of Listeria growth inhibition measuring 21 mm of diameter). The minor peaks observed in the chromatogram did not show antimicrobial activity. Research, Society andDevelopment, v. 10, n. 9, e2610917783, 2021 (CC BY 4.0) | ISSN 2525-3409 | DOI: http://dx.doi.org/10.33448/rsd-v10i9.17783

Discussion
Listeriosis can cause severe sequels to animal health. The clinical form of the disease can be characterized in three ways: septicemia, with abscesses in viscera, such as liver and spleen, abortion and meningoencephalitis (Campero et al., 2002).
The acquisition of L. monocytogenes by animals can occur by the ingestion of water contaminated with feces or secretions of infected animals (Brugere-Picoux 2008), and, to ruminants, the main sources of this microorganism is soil and contaminated food (especially poorly fermented silage) (Fentahum and Freseberat, 2012).
Lactic acid bacteria (LAB) have been extensively used in order to inhibit the growth of undesirable microorganisms (Dolenčić Špehar et al., 2020). In addition to microbial control, LAB can also be used to improve the fermentation process in silage, contributing for silage quality (Cheng et al., 2021). Many of LAB are able to produce antimicrobial compounds, such as bacteriocins, that can control the bacterial community during the ensilage process (Contreras-Govea et al., 2011). The addition of Lactobacillus buchneri, a producer of the bacteriocin buchnericin LB, on sugarcane silage, for example, has been reported to inhibit the growth of undesirable bacteria, such as Listeria sp., Bacillus sp., Micrococcus sp., Enterococcus sp., Streptococcus sp., Lactobacillus sp., Leuconostoc sp. and Pediococcus sp., improving the quality of the silage and reducing the loss of dry matter during the ensilage process (Yildirim 2001;Zopollatto et al., 2009). However, these bacterial inoculants might not prevent secondary fermentation in moist silages, sometimes impairing the aerobic stability of grass and small grain silages (Weinberg and Muck, 1996). Therefore, it is important to evaluate other microorganisms than LAB as alternative inoculants to silage.
In the present study, 18.9% of the bacterial strains isolated from silage showed antibacterial activity, inhibiting the growth of L. monocytogenes, and other Gram-positive bacteria, including the pathogenic strains S. aureus and B. cereus, as well as Gram-negative bacteria, such as E. coli, Pseudomonas and Klebsiella. The activity spectrum of bacteriocins is generally associated to phylogenetically-related bacteria (Bordignon-Junior et al. 2013;Yang et al. 2014), although, as also demonstrated in the present study, some bacteriocins have shown a broad spectrum of action (Ivanova et al., 1998;Todorov et al., 2010;De Oliveira et al., 2012). It is important to highlight the activity of the bacteriocins produced by the silage isolates evaluated against S. aureus and E. coli, important pathogens that cause bovine mastitis, a disease still prevalent in the Brazilian herd that leads to considerable economic losses (Fagundes and Oliveira, 2004).
The most promising bacteriocin-producing bacterium evaluated in the present study was identified as B. velezensis.
In this study, the protein character of the antibacterial compound produced by B. velezensis was confirmed by the proteinase K sensibility. The bacteriocin was stable to different wide of temperature and pH, as expected for a bacterium capable of withstanding extreme environmental conditions (Abrioeul et al., 2011;Baruzzi et al. 2011;Ansari et al. 2012). The exposure of B. velezensis bacteriocin to low or high temperatures resulted in a slight reduction in its anti-Listeria activity. At temperatures below 80 °C, the residual anti-Listeria activity remained above 90%, while at 100 °C this activity decreased to 64%. Lim et al., (2016) demonstrated similar results, being the residual activity of 100% when the bacteriocin produced by Bacillus amyloliquefaciens was exposed to 4 °C, 37 °C, 50 °C and 80 °C for 30 min, while in 100 °C this activity decreased to 80%.
Even showing higher antimicrobial activity at neutral pH (pH = 6.5), the bacteriocin produced by B. velezensis showed considerable stability in different pH values, maintaining over than 80% of its anti-Listeria activity in the pH range of 2 to 12. Similarly, stability to pH levels ranging from 3.0 to 9.0 was reported for mejucin, produced by Bacillus subtilis SN7 (Lee and Chang, 2018). On the other hand, a pH-dependent antimicrobial activity has been demonstrated for other bacteriocins, such as nisin, lactoestrepcin (Bromberg et al. 2006), and bovicin HC5 (Paiva et al., 2011). The stability of the bacteriocin prospected in our study to different pH and temperature value is important to its potential application.

Final Considerations
From the literature survey, it has been found that B. velezensis, a growing fast bacterium, is harmless to human and animal, and does not pollute the environment, playing an important role as a probiotic in animal feed. These properties, associated with the production of a broad inhibitory spectrum, pH-tolerant and thermo-stable bacteriocin, characterized in our study, supports and suggests the use of B. velezensis and its bacteriocin as inoculants for silage, alone or in combination with other methods, specially focusing on L. monocytogenes proliferation control.